iwb i/m I -Ui JOURNAL OF SHELLFISH RESEARCH VOLUME 16, NUMBER 1 JUNE 1997 The Journal of Shellfish Research (formerly Proceedings of the National Shellfisheries Association ) is the official publication of the National Shellfisheries Association Editor Dr. Sandra E. Shumway Natural Science Division Southampton College, LIU Southampton, NY 11968 Dr. Standish K. Allen. Jr. (1998) Rutgers University Haskin Laboratory for Shellfish Research P.O. Box 687 Port Norris, New Jersey 08349 Dr. Peter Beninger (1997) Department of Biology University of Moncton Moncton. New Brunswick Canada El A 3E9 Dr. Andrew Boghen (1997) Department of Biology University of Moncton Moncton. New Brunswick Canada El A 3E9 Dr. Neil Bourne (1997) Fisheries and Oceans Pacific Biological Station Nanaimo. British Columbia Canada V9R 5K6 Dr. Andrew Brand (1997) University of Liverpool Marine Biological Station Port Erin. Isle of Man Dr. Eugene Burreson (1997) Virginia Institute of Marine Science Gloucester Point. Virginia 23062 Dr. Peter Cook (1998) Department of Zoology University of Cape Town Rondebosch 7700 Cape Town, South Africa EDITORIAL BOARD Dr. Simon Cragg (1998) Institute of Marine Sciences University of Portsmouth Ferry Road Portsmouth P04 9LY United Kingdom Dr. Leroy Creswell (1997) Harbor Branch Oceanographic Institute US Highway 1 North Fort Pierce, Florida 34946 Dr. Lou D'Abramo (1998) Mississippi State University Dept of Wildlife and Fisheries Box 9690 Mississippi State. Mississippi 39762 Dr. Ralph Elston ( 1997) Battelle Northwest Marine Sciences Laboratory 439 West Sequim Bay Road Sequim, Washington 98382 Dr. Susan Ford (1998) Rutgers University Haskin Laboratory for Shellfish Research P.O. Box 687 Port Norris, New Jersey 08349 Dr. Raymond Grizzle (1997) Randall Environmental Studies Center Taylor University Upland, Indiana 46989 Dr. Robert E. Hillman (1998) Battelle Ocean Sciences New England Marine Research Laboratory Duxbury, Massachusetts 02332 Dr. Mark Luckenbach (1997) Virginia Institute of Marine Science Wachapreague, Virginia 23480 Dr. Bruce MacDonald (1997) Department of Biology University of New Brunswick P.O. Box 5050 Saint John, New Brunswick Canada E2L 4L5 Dr. Roger Mann (1998) Virginia Institute of Marine Science Gloucester Point, Virginia 23062 Dr. Islay D. Marsden (1996) Department of Zoology Canterbury University Christchurch, New Zealand Dr. Kennedy Paynter (1998) 1200 Zoology Psychology Building College Park, Maryland 20742-4415 Dr. Michael A. Rice (1996) Dept. of Fisheries. Animal & Veterinary Science The University of Rhode Island Kingston, Rhode Island 0288 1 Dr. Tom Soniat (1998) Biology Department Nicholls State University Thibodaux, Louisiana 70310 Susan Waddy (1997) Biological Station St. Andrews, New Brunswick Canada, EOG 2XO Dr. Gary Wikfors (1998) NOAA/NMFS Rogers Avenue Milford, Connecticut 06460 Journal of Shellfish Research Volume 16, Number 1 ISSN: 00775711 June 1997 Journal of Shellfish Research, Vol. 16. No. I, 1-6, 1997 RECRUITMENT OF STROMBUS VELIGERS TO THE FLORIDA KEYS REEF TRACT: RELATION TO HYDROGRAPHIC EVENTS ALLAN W. STONER, 1 * NIKHIL MEHTA, 1 AND THOMAS N. LEE 2 'Caribbean Marine Research Center 805 E. 46th Place Vera Beach, Florida 32963 Rosenstiel School of Marine and Atmospheric Science University of Miami 4600 Rickenbacker Causeway Miami, Florida 3314V ABSTRACT Recruitment of the veliger larvae of two strombid gastropods was investigated during the reproductive season (May to September) at two stations in the Looe Key National Marine Sanctuary. Florida Keys, in 1992 and 1994. The adult population of Strombus gigas Linne (queen conch ) has been severely depleted by fishing, and despite protection of the species in Florida since 1 985. there has been no recovery. Strombus costatus Gmelin (milk conch) is a closely related but an unexploited gastropod in Florida waters. Although the two species have very similar reproductive and larval life histories, frequent sampling showed that larval recruitment patterns were different. Early-stage veligers of S. costatus were very abundant, and the number of veligers decreased rapidly with size. in a typical survivorship pattern. For S. gigas, early- and late-stage veligers were collected in approximately equal numbers, and mid-size larvae were rare. Competent, late-stage veligers of S. gigas arrived in the vicinity of nursery grounds in association with thermal stratification and eastward current, indicating the nearshore presence of the Florida Current. In contrast, late-stage veligers of 5. costatus were most often present in association with westerly flow, during periods when the Florida Current was well offshore, and the recruitment source appears to be local. These observations suggest that populations of S. gigas may now depend primarily on larval sources upstream from the Florida Keys in the western Caribbean Sea. Infrequent and irregular supply of larvae to the nurseries may explain the lack of population recovery for S. gigas in the Florida Keys. KEY WORDS: Florida, larval supply, oceanography, recruitment, Strombus gigas, Strombus costatus INTRODUCTION The supply of larvae to potential juvenile habitats is well known to be an important variable in the population dynamics of benthie marine invertebrates and fishes (e.g.. Yoshioka 1982, Wethey 1984, Gaines et al. 1985. Lipcius et al. 1990. Milicich et al. 1992. Peterson and Summerson 1992. Doherty and Fowler 1994. Stoner et al. 1996). Given the significance of larval supply in determining the distribution, abundance, and year-class strength of many economically significant marine species, it is imperative for effective resource management that larval sources be identi- fied. An understanding of the physical and biological processes that mediate delivery of larvae to the subject populations is also critical. The queen conch (Strombus gigas Linne) is a large, economi- cally significant gastropod that inhabits Bermuda and southern Florida through the greater Caribbean region to Venezuela (Ran- dall 1964). In Florida, the species was reduced to such an extent that all fishing was banned in 1985. Since that time, the population has shown no sign of recovery (Glazer and Berg 1994, Glazer and Anderson unpubl. data) and the fishing moratorium remains in effect. Larvae of other invertebrates, including certain species of lobsters and shrimp spawned in the Florida Keys, may be retained in mesoscale gyres and returned to local nurseries (Yeung and McGowan 1991, Lee et al. 1992. Lee et al. 1994. Criales and McGowan 1994). However, because there were so few queen conch in the local reproductive stock (5.800 adults along the 180 km of reef tract in 1992), Stoner et al. (1996) postulated that the populations in the Keys are probably replenished with larvae spawned in Cuba and the western Caribbean Sea (Mexico and Belize) that are delivered via the Florida Current. In contrast to the very small spawning population of S. gigas in the Florida Keys, adult Strombus costatus Gmelin are abundant in shallow coastal waters both north and south of the island chain. The co-occurrence of two closely related Strombus species in the Florida Keys and western Caribbean presents the opportunity to compare recruitment processes in the heavily exploited queen conch (5. gigas) and the milk conch [S. costatus). which is unex- ploited in Florida. Larvae of the two species are similar in size and general appearance, live primarily in the upper water column, and have variable but relatively similar developmental periods of 16- 35 days (Davis et al. 1993). Differences in larval size-frequency between the two species and temporal differences in their delivery to Looe Key National Marine Sanctuary, combined with physical oceanographic data for the site, provide important new insights into larval sources and recruitment processes in the Keys popula- tion. METHODS Study Site *Present address: Northeast Fisheries Science Center. National Marine Fisheries Service. 74 Magruder Road. Highlands. NJ. The Florida Keys are a chain of islands running south and west from the southern tip of the Florida peninsula to Key West. The reef tract, a series of shallow coral reefs, lies 5-10 km offshore from the islands and runs parallel to the Keys (Jaap 1984). Sam- pling was conducted in the lower Florida Keys near the coral reef in Looe Key National Marine Sanctuary. This particular reef is 0.8 km long, nearly emergent at low tide (range = -1 m), oriented east-west, and located -10 km south of Big Pine Key (Fig. 1). South of the reef, depth increases rapidly to 20 m. Areas of coral Stoner et al. .26' -vh Flo ?1" idaW Cuba ■20" 82" 76* FLORIDA BAY BIG PINE ' ^ '< A'Vv *• vaca FLORIDA STRAIT Figure 1. Map showing the lower Florida Keys and the locations of the two veliger-sampling sites near Looe Key reef. rubble extend north from both ends of the reef, separated by a shallow ( 1- to 2-m-deep) bed of seagrass (Thalassia testudinum). Juvenile queen conch occupy these shallow rubble and seagrass areas. Adults are present in the deeper (3- to 20-m) sand and rubble habitats, where spawning occurs (Stoner et al. 1996). Plankton samples were collected at LK1 and LK2. two sites previously studied by Stoner et al. (1996) (Fig. 1). LK1 was lo- cated in shallow (-1.5-m) water north of the reef in sand, coral rubble, and seagrass habitat. LK2 was located -0.5 km offshore from the reef in deeper water (20-30 m), where the substrate was sand, macroalgae, and sponges. In the laboratory, plankton samples were sorted in their entirety for veligers of S. gigas and S. costatus with a dissecting micro- scope (20x). Veligers were identified to species using the descrip- tions of Davis et al. (1993), counted, measured for maximum shell length (SL). and divided into three size classes: early-stage (<500 u-m SL). mid-size (500-900 u,m SL), and late-stage (>900 (xm SL) veligers. Abundance was calculated as veligers per unit volume of water sampled (veligers/10 m 3 ) for the three size classes and for total number. Size-specific density data are useful tools for inter- preting larval production and understanding transport processes. For example, early-stage S. gigas veligers are only a few days old (Davis et al. 1993) and reflect local larval production, whereas veligers >900 u.m represent conch ready to recruit to the benthos, are 3^4 wk old, and may have originated from a distant reproduc- tive population (Stoner et al. 1996). Physical Measurements On each sampling date, data were recorded on wave height and direction, wind speed and direction, water clarity, and surface- water temperature. Water temperature and current direction and speed were measured with three General Oceanics winged current meters moored at the 30-m isobath off of Looe Key reef (24°C32.5'N, 81°C24.1'W) at depths of 7, 17, and 27 m. Data were filtered with a 3- and 40-h low-pass Lancoz filter and sub- sampled at 1- and 6-h intervals, respectively. Current components were rotated into a local isobath coordinate system with +v toward 73° (i.e., alongshore in a generally eastward direction) and +u toward 163° (i.e., offshore). RESULTS Biological Collections Plankton samples were collected on 33 dates at LK1 and 35 dates at LK2 (two replicates at each site) between late May and late September in 1992 and 1994. At each station, conical nets (0.5 m in diameter, 2.5 m long, 202-p.m mesh size) were towed ( 15 min at -1.0 m s~') near the water surface during daylight hours. Samples were collected at the surface because 5. gigas veligers are known to be photopositive (Barile et al. 1994) and are most abun- dant near the surface when conditions are relatively smooth (Stoner and Davis 1997). Tow volume was calculated from a cali- brated General Oceanics flowmeter attached in the mouth of the net. Plankton samples were preserved in a buffered 5% formalin- seawater mixture. Veliger Abundance and Length-Frequency In 1992. mean densities of 5. gigas were 0.60 veligers/10 m 3 at the shallow site LK1 and 0.09 veligers/10 m 3 at the offshore site LK2 (Table 1). Few S. costatus were collected in 1992. Mean density was 0.04 veligers/10 nr at LK2. and none were collected at LKI. In 1994. densities of S. gigas were 0.13 veligers/10 m 3 at both LKI and LK2. and S. costatus had mean densities of >15 veligers/10 m 3 at both stations (Table 1). However, presence was sporadic for both species. Densities of 5. gigas were high (>1 veliger/10 m 3 ) on only 9% (3 of 33 dates) at LKI. and only 3% (1 of 35 dates) at LK2. High densities of 5. costatus veligers were found in 30% of the collections at LKI and 34% at LK2, but the two species never occurred in high density at the same time. TABLE 1. Counts and densities of veligers of Stromhus spp. {all stages) collected at two stations in the Florida Keys, May through September, 1992 and 1994. 1992 1994 S. gigas S. costatus S. gigas S. costatus Station No. of Veligers Density (no. ■ 10 m * No. of Veligers Density (no. • 10 m -3 ) No. of Veligers Density (no. ■ 10 m " No. of Veligers Density (no.- 10 nT 3 ) LooeKeyl(LKl) LooeKey2(LK2) Total 144 24 168 0.60 ± 1.14 0.09 ±0.17 13 13 0±0 0.04 ±0.11 118 135 253 0.13 ±0.43 0.13 ±0.30 15.102 15,533 30,635 16.79 + 48.15 15.88 ±51.12 Density values are mean ± standard deviation. In 1994. 50 tows were made at station LK 1 and 54 tows were made at station LK 2. Sixteen tows were made at each station in 1992. Recruitment of Queen Conch Larvae Size-frequency distributions for 5. gigas and S. costatus were very different (Fig. 2). At station LK1, 89% of all 5. gigas were early stage (<500 u.m) and 10% were late stage (>900 u.m). In contrast, at LK2, only 8% of the veligers of 5. gigas were early stage and 91% were late stage. Only three mid-size (500- to 900- u.m) S. gigas were collected. At both LK1 and LK2, over 60% of the S. costatus were early stage, 33-37% were mid-size, and only 0.2 (LK1) to 1.2% (LK2) were late stage (Fig. 2). Early-stage S. gigas were collected sporadically and on just a few dates, primarily at station LKI (Fig. 3). Highest densities occurred on June 13, 1992(3.1 veligers/10 m 3 ), July 18, 1992(1.5 veligers/10 nr), and August 24, 1994 (2.0 veligers/10 m 3 ). Early stages were collected only twice at LK2, on June 1, 1992 (0.2 veligers/10 m 3 ). and August 29, 1994 (0.14 veligers/10 m 3 ) (not shown). The spatial pattern for the late-stage veligers of S. gigas was opposite that for the early stages. Late stages were very rare at station LKI, except on September 6, 1994 (0.7 veligers/10 m 3 ), concurrent with a similar density at LK2 in 1994. Late-stage 5. gigas were collected sporadically at LK2 (Fig. 3), with the highest densities on August 19. 1992 (0.47 veligers/10 m 3 ), June 9, 1994 (0.85 veligers/10 m 3 I.September 1. 1994 ( 1.2 veligers/10 m 3 ). and September 6. 1994 (0.75 veligers/10 m 3 ). In 1992, no S. costatus were collected at LKI. and only 13 individuals (0.04 veligers/10 nr ) were collected at LK2, all on July 1 8. High densities of early-stage veligers were collected in 1994 at station LKI. Maxima occurred on June 20 ( 140 veligers/10 nr) and June 23 (58 veligers/10 nr ), with other sporadic occur- rences (Fig. 4). Late stages of S. costatus were rare at LKI but present in 12 of the 27 collections made at LK2 in 1994. The densities were generally <1.0 veliger/10 nr\ except on July 26. 1994 (8.1 veligers/10 m 3 ) (Fig. 4). Relationships Between Veliger Abundance and Hydrography The most prominent feature of currents off of Looe Key reef was the regular reversal in alongshore flow, with velocities occa- sionally exceeding 50 cm/sec at 7-m depth (Fig. 5). Cross-shelf currents were typically <5 cm/sec and were not particularly useful in this analysis. Strong easterly flow (+v) and increased vertical stratification in water temperature characterized periods when the Florida Current was close to Looe Key. Periods with westerly flow (-v) and low thermal stratification indicated that the Florida Cur- rent front was offshore, beyond the 30-m isobath. There were strong indications of the Florida Current at Looe Key in early July, mid- to late-August, and in mid-September 1992. In 1994. condi- tions indicative of the Florida Current were observed primarily in late-May through mid-June and later in mid-August to mid- September. Because late-stage larvae provide the best indication of conch ready to recruit to shallow-water habitats, we examined their pres- ence in terms of the position of the Florida Current front at LK2 where late stages were most abundant. On 9 of the 13 dates when late-stage S. gigas were collected, flow and temperature conditions indicated that the Florida Current front was close to the reef tract. On 8 of 12 dates, late-stage S. costatus were collected when the Florida Current was farther offshore. Fisher's exact test indicated that there was a significant (p = 0.036) interaction between spe- cies and the presence of the Florida Current. All high-density (>0.4 Strombus gigas Strombus costatus to 4U 35 30 25 20 15 - 10 I 5 n LK1 n = 262 ooooooooooooo ouiomomomomomo o o o o u> o m o CO rJ> O) o o o o o m o m o o v- v- e\J o o o o inomo w n n oo>-'-NCMna>^ o c ^r*«cOCOCTlOOO*— t- CM CM PI C) Tf juiomoinoi/iom< j cm n n »} -t in in id id i Shell Length (um) 40 35 30 25 20 • 15 10 - LK2 n= 15,533 OOOOOOOOOOOOOOOOOOOOOOOOO omomoinouiomoLnoinoinomomoLfioioo t>jc\icr>co^Tfi/)intD«>i*--r~~cotooia>oO'--»-c\i^\jc*)ro^' Shell Length (um) Figure 2. Size-frequency distribution for veligers of .S\ gigas and .S'. costatus at two stations in the Florida Keys. The frequency distributions for S. gigas represent all individuals collected in both 1992 and 1994. For S. costatus. only the data for 1994 are shown because only 13 early-stage veligers were collected on one date in 1992. Asterisks represent size classes comprising <0.01%. CO c (1) , — , Q I— E Q) o o> CD > o r * — * co A A A N CNI C\J 1992 Early-stage LK1 K. < < < CO CO CO r- CSJ Late-stage LK2 ^L 4 ^ r- eg < < CO CO CO CO 1994 Figure 3. Density of early-stage veligers (<500 urn SLl of S. gigas at station LK1 and late-stage veligers <<900 um SL) at station LK2 in 1992 and 1994. Values shown are mean ± standard error (n = 2). Note that the scales are different for early- and late-stage larvae. veligers/10 m 3 ) occurrences of late-stage 5. gigas (August 19, 1992. and June 9. September 1. and September 6. 1994) were associated with distinct easterly flow and vertical stratification of the water column (Figs. 3 and 5). The highest densities of late- stage S. costatus (July 18, 1992, and June 27. July 26. and August 1. 1994) always occurred when currents were westerly and there was little stratification (Figs. 4 and 5). Late-stage 5. costatus were present on one occasion at LK2. during a period of high easterly flow (June 9. 1994), but the density was only 0.36 veligers/10 m 3 ". DISCUSSION Mesoscale gyres can affect the retention and recruitment of larval fish and invertebrates in the nearshore environment of the lower Florida Keys (Lee et al. 1992, Lee et al. 1994). Evidence exists for the retention of penaeid shrimps on the Tortugas and Pourtales Gyres (Criales and McGowan 1994. Criales and Lee 1995), and similar mechanisms may be responsible for retaining the larvae of certain scyllarid lobsters in the vicinity of the Florida Keys shelf (Yeung and McGowan 1991). With a 2- to 4-wk-long larval period for Strombus spp. (Davis et al. 1993). similar to that of these shrimps and lobsters, it is plausible that the retention and recruitment of Strombus spp. would be affected by mesoscale gyres in the Straits of Florida. However, two strong lines of evi- dence suggest that this was not the case for larvae of Strombus spp. in the two years surveyed. First, it is very unlikely that most of the veligers of S. gigas were produced by spawners in the Florida Keys. Estimates for the total number of adult queen conch in the entire Keys region was just 5.800 individuals in 1992 and 9,200 in 1994 (Glazer and Anderson unpubl. data). There were relatively few newly hatched veligers of S. gigas at Looe Key, despite the fact that a large proportion of the total Keys reproductive stock ( 10% in 1992 and 20% in 1994) occurred in the well-patrolled environment of Looe Key National Marine Sanctuary. It is very unlikely that the den- sities of late-stage S. gigas, often exceeding the densities of early stages collected at numerous locations surveyed in the Florida Keys between 1992 and 1994 (Stoner et al. 1996, this study), could be derived from these low concentrations of early stages, particu- larly when mid-sized larvae are extremely rare. Second, there are large spawning stocks of S. gigas in Mexico and Belize, and recruitment of late-stage veligers to Looe Key during periods of high eastward flow is consistent with the hy- pothesis that they have a source in the western Caribbean Sea. The plausibility of this source of larvae and the associated transport mechanism is supported by examination of the larval development period in combination with what is known about near-surface cir- culation between the Yucatan Strait and Florida Keys. Trajectories of satellite-tracked drifters show that surface water passing through the Yucatan Strait and into the Loop Current of the eastern Gulf of Mexico can reach the Florida Keys area in 30-35 days (Kinder 1983). When the Loop Current is less well developed and surface flow is more direct from the Caribbean coast of Mexico to Florida (approx. 700 km), the transit time could be as short as 10 days, with a current velocity of 0.8 m/sec, observed by Kinder Recruitment of Queen Conch Larvae Strombus costatus 140±38 (20-Jurw) 58 ±7 (23-Jun«) Early-stage LK1 Late-stage LK2 1994 Figure 4. Density of early-stage veligers (<500 um SL) of S. costatus at station LK1 and late-stage veligers (3=900 Jim SL( at station LK2 in 1994. Values shown are mean ± standard error (n = 21. Numbers collected in 1992 Here too low to plot. Note that the scales are different for early- and late-stage larvae. (1983). Consequently, the larval period of 2—+ wk in S. gigas (Davis et al. 1993) is sufficiently long for transport from large reproductive populations in Mexico. Belize, and Honduras. The scarcity of early-stage veligers of 5. gigas in the Florida Current south of Looe Key precludes the north shore of Cuba as a source because drifter model trajectory estimates of arrival times are on the order of 4-6 days, if at all (Chen 1996). Concentrations of late-stage 5. gigas are known to be high in the Florida Current 35 km south of the middle Keys (Stoner et al. 1996), and the arrival of 5. gigas in association with easterly flow at Looe Key indicates that larvae of Caribbean origin are being delivered by the Florida Current. Although genetic studies have shown a high degree of similarity among populations of S. gigas in the Caribbean and Florida (Mitton et al. 1989, Campton et al. 1992), this study pro- vides the first oceanographic data indicating that the Florida Keys population of queen conch is now at least partially dependent on upstream sources. 5. costatus has a larval life history (Davis et al. 1993) and vertical migration behavior (Stoner and Davis 1997) similar to that of S. gigas. and it could be expected that some veligers of 5. costatus recruit from reproductive populations in the Yucatan. Nevertheless, recruitment of late-stage 5. costatus larvae was as- sociated with Florida Current conditions on only one occasion. Recruitment of S. costatus in the Keys may be influenced by mechanisms similar to those affecting alpheid shrimps. Criales and McGowan (1994) found that alpheid larvae were abundant only in nearshore waters in the Florida Keys, and they proposed that the entire larval development occurred in coastal waters, unaffected by the Florida Current or mesoscale gyres. The size-frequency distri- bution for veligers of S. costatus suggests a typical survivorship pattern, with a high abundance of early stages and relatively rare late stages. Also, 5. costatus occurred at Looe Key primarily when temperature and current conditions indicated that the Florida Cur- rent front was well offshore. This may explain the low numbers of larvae collected at Looe Key in 1992. Only one collection was made during a period of strong westerly flow in 1992, on July IS. when all of the larvae were collected. Periods of low thermal stratification were rare in 1992. However, in 1994. there were long periods of westerly flow and low stratification, and the vast ma- jority of S. costatus was collected during these periods (Fig. 5). The inshore distribution and spawning of S. costatus may ex- plain the association of larvae with westerly current and the shelf water mass. Although populations of S. costatus have not been quantified, the species is very abundant near the islands and in Florida Bay and is relatively rare near the reef tract (pers. observ.). Consequently, the larvae are generally outside the influence of the Florida Current, and the most important source of recruitment for 5. costatus is probably local. Our findings have important management implications. It is Figure 5. Alongshore current and water temperatures (Temp) off- shore from Looe Key reef. May through September, 1992 and 1994. Solid-line plots represent readings at 7 m, dotted lines are readings at 17 m, and dashed lines are readings at 27 m in 1992. In 1994, the meter at 27-m depth failed to record. Solid vertical lines indicate the sam- pling dates when high densities (>0.4 veligers/10 m 3 ) of late-stage S. gigas were collected. Dashed vertical lines show the dates when high densities of late-stage S. costatus were collected. Stoner et al. very likely that the large populations of 5. gigas once known in the Florida Keys were sustained by local spawners, as are populations of S. costatus. Today, however, with severe overfishing, the only reproductive stocks of S. gigas lie several kilometers offshore, along the reef tract (Stoner et al. 1996). The larvae produced at the outer shelf are easily lost to the Straits of Florida, the densities of late-stage larvae in the coastal water mass are practically zero, and the Florida Current is now the primary recruitment source for S. gigas. Recruitment to juvenile populations in S. gigas is known to be dependent on larval supply (Stoner et al. 1996). but deliveries of settlement-stage larvae by meanders of the Florida Current are probably too infrequent and irregular to sustain or rebuild a large spawning population in the Florida Keys. For example, Florida Current conditions did not occur at Looe Key between mid-June and mid- August 1994. and almost no late-stage 5. gigas were collected during that period. Lack of recruitment during the warm- est part of the summer season is likely to have a significant nega- tive effect on year-class strength for 1994, particularly at this northern extreme of the species' geographic range. Rehabilitation of queen conch stocks in the Florida Keys may now depend on transplants of spawners or the release of hatchery- reared juveniles. Unfortunately, stock enhancement through juve- nile release is difficult and expensive and has a history of low success (Stoner 1994, Stoner and Glazer in press). Wise manage- ment and transgenerational enhancement of marine fishery re- sources will depend on extensive knowledge of recruitment pro- cesses and metapopulation dynamics. ACKNOWLEDGMENTS This research was supported by grants from the National Un- dersea Research Program of NOAA (Department of Commerce) and NOAA/CIMAS (SEFCAR) through Contract NA85-WC-H- 06134. and USGS (SFOSRC Agreement No. RD-93-02). The Looe Key National Marine Sanctuary and the Marine Research Institute of the Florida Department of Environmental Protection provided boat time for larval collections. Research within the Florida Keys National Marine Sanctuary was conducted under National Marine Sanctuary Research Permits KLNM5 and LKNMS-1 1-89, LKNMS-05-92. and LKNMS-01-94. We thank P. Barile and the Looe Key National Marine Sanctuary's team of volunteers for assistance in the field and in the laboratory. M. Ray and anonymous reviewers provided careful readings and criticisms of the manuscript. LITERATURE CITED Barile. P. J., A. W. Stoner & C. M. Young. 1994. Phototaxis and vertical migration of the queen conch [Strombus gigas Linne) veliger larvae. J. Exp. Mar. Biol. Ecol. 183:147-162. Campton. D. E.. C.J. Berg. Jr.. L. M. Robison & R. A. Glazer. 1992. Genetic patchiness among populations of queen conch. Strombus gigas. in the Florida Keys and Biraini. Fish. Bull. U.S. 90:250-259. Chen. P. 1996. Current-modified wind-forced ageostrophic transport in the Straits of Florida: a numerical simulation study. M.S. thesis. Rosenstiel School of Marine and Atmospheric Science, University of Miami. Mi- ami. FL. 69 pp. Criales. M. M. & T. N. Lee. 1995. Larval distribution and transport of peneoid shrimps during the presence of the Tortugas Gyre in May-June 1991. Fish. Bull. U.S. 93:471-482. Criales. M. M. & M. F. McGowan. 1994. Horizontal and vertical distribu- tion of penaeidean and caridean larvae and micronektonic shrimps in the Florida Keys. Bull. Mar. Sci. 54:843-856. Davis, M., C. A. Bolton & A. W. Stoner. 1993. A comparison of larval development, growth, and shell morphology in three Caribbean Strom- bus species. Veliger 36:236-244. Doherty, P. & T. Fowler. 1994. An empirical test of recruitment limitation in a coral reef fish. Science 263:935-939. Gaines, S., S. Brown & J. Roughgarden. 1985. Spatial variation in larval concentrations as a cause of spatial variation in settlement for the barnacle. Balanus glandula. Oecologia 67:267-272. Glazer, R. A. & C. J. Berg, Jr. 1994. Queen conch research in Florida: an overview, pp. 79-95. In: R. S. Appeldoorn and B. Rodriguez (eds.). Queen Conch Biology. Fisheries and Mariculture. Fundacion Cientffica Los Roques. Caracas, Venezuela. Jaap, W. 1984. The ecology of the south Florida coral reefs: a community profile. U.S. Department of the Interior. Minerals Management Service. Washington. D.C. MMS 84-0038. 139 pp. Kinder. T. H. 1983. Shallow currents in the Caribbean Sea and Gulf of Mexico as observed with satellite-tracked drifters. Bull. Mar. Sci. 33: 239-246. Lee. T. N.. M. E. Clark. E. Williams. A. F. Szmant & T. Berger. 1994. Evolution of the Tortugas Gyre and its influence on recruitment in the Florida Keys. Bull. Mar. Sci. 54:621-646. Lee. T. N., C. Rooth. E. Williams, M. McGowan, A. F. Szmant & M. E. Clark. 1992. Influence of Florida Current, gyres and wind-driven cir- culation on transport of larvae and recruitment in the Florida Keys coral reefs./ Contin. Shelf Res. 12:971-1002. Lipcius. R. N„ E. J. Olmi & J. Van Montffans. 1990. Planktomc availabil- ity, molt stage and settlement of blue crab postlarvae. Bull. Mar. Sci. 46:230-243. Milicich, M. J., M. G. Meekan & P. J. Doherty. 1992. Larval supply: a good predictor of recruitment of three species of reef fish (Pomacen- tridae). Mar. Ecol. Prog. Ser. 86:153-166. Mitton. J. B.. C. J. Berg. Jr. & K. S. Orr. 1989. Population structure, larval dispersal, and gene flow in the queen conch. Srrombus gigas, of the Caribbean. Biol. Bull. 177:56-362. Peterson. C. H. & H. C. Summerson. 1992. Basin-scale coherence of popu- lation dynamics of an exploited marine invertebrate, the bay scallop: implications of recruitment limitation. Mar. Ecol. Prog. Ser. 90:257- 272. Randall. J. E. 1964. Contributions to the biology of the queen conch. Strombus gigas. Bull. Mar. Sci. Gulf. Carib. 14:246-295. Stoner. A. W. 1994. Significance of habitat and stock pre-testing for en- hancement of natural fisheries: experimental analyses with queen conch. /. World Aquacult. Soc. 25:155-165. Stoner. A. W. & M. Davis. 1997. Abundance and distribution of queen conch veligers {Strombus gigas Linne) in the central Bahamas: 2. Ver- tical patterns in nearshore and deep-water habitats. J. Shellfish Res. 16:19-29. Stoner, A. W. & R. A. Glazer. (In press). Variation in natural mortality: implications for queen conch stock enhancement. Bull. Mar. Sci. Stoner. A. W„ R. A. Glazer & P. J. Barile. 1996. Larval supply to queen conch nurseries: Relationships with recruitment process and population size in Florida and the Bahamas. J. Shellfish Res. 15:407-420. Wethey, D. S. 1984. Spatial pattern in barnacle settlement: day to day changes during the settlement season. J. Mar. Biol. Assoc. U.K. 64: 687-698. Yeung. C. & M. F. McGowan. 1991. Differences in inshore-offshore and vertical distribution of phyllosoma larvae of Panulirus. Scyllarus and Scyllarides in the Florida Keys in May-June. 1989. Bull. Mar. Sci. 49:699-714. Yoshioka, P. M. 1982. Role of planktomc and benthic factors in the popu- lation dynamics of the bryozoan Membranipora membranacea. Ecol- ogy 63:457^+68. Journal of Shellfish Research. Vol. 16, No. 1. 7-18, 1997. ABUNDANCE AND DISTRIBUTION OF QUEEN CONCH VELIGERS (STROMBUS GIGAS LINNE) IN THE CENTRAL BAHAMAS. I. HORIZONTAL PATTERNS IN RELATION TO REPRODUCTIVE AND NURSERY GROUNDS ALLAN W. STONER* AND MEGAN DAVIS T Caribbean Marine Research Center 805 E. 46th Place Vera Beach, Florida 32963 ABSTRACT Veliger larvae of the large gastropod Strumous gigas (queen conch) were collected over a 7-y period at a reproductive site in the Exuma Sound, in adjacent tidal inlets, and on the Great Bahama Bank near Lee Stocking Island, Exuma Cays, Bahamas. Although the spawning season for queen conch in the region occurs from April through October, larval abundance was highest during mid summer, between late June and August, when temperatures were >28°C. Metamorphically competent larvae were most abundant in July and August and made up a small percentage of the total count, reflecting high natural mortality. Although there was considerable interannual variation, veliger densities near Lee Stocking Island (typically 1-2 individuals/10 m 3 ) were much higher than estimates made in the eastern Caribbean and Florida Keys. Because most of the collected larvae were newly hatched, regional differences in larval abundance appear to be associated with size of local spawning stock. Highest densities of larvae were found on the Great Bahama Bank directly associated with axes of tidal currents and decreased with distance onto the bank. Intensive sampling in one tidal flow field showed that total larval densities as well as densities of late-stage veligers were highest at a well-established nursery site. Larval transport and retention may explain the general occurrence of nurseries at locations where water from the Exuma Sound flows onto the Great Bahama Bank on flood tides. Large, stable aggregations of juvenile queen conch were consistently supplied with high densities of larvae and were directly associated with tidal pathways. In contrast, more ephemeral aggregations were characterized by low or inconsistent veliger densities (particularly late-stage larvae) and were generally outside primary tidal current pathways. Queen conch distribution appears to be directly related to the horizontal supply of larvae. KEY WORDS: Bahamas, larval transport, oceanography, recruitment, Strombus gigas INTRODUCTION The large gastropod Strombus gigas Linne (queen conch) is an important fisheries resource in the wider Caribbean region (Berg and Olsen 1989. Appeldoorn 1994). Despite its culture in hatch- eries for nearly 20 y (Creswell 1994. Davis 1994a), knowledge of the natural history and field distribution of the veliger larva is limited to the most basic facts. The spawning season lasts 24-36 wk and varies slightly throughout the Caribbean, depending on temperature and photoperiod (Stoner et al. 1992). After 3-5 days, conch veligers hatch from their benthic egg masses at 300 u.m shell length (SL) (D'Asaro 1965. Davis 1994a); they then live in the water column for 18-26 days, depending on physical and trophic conditions (Laughlin and Weil 1983. Siddall 1983, Davis et al. 1993). Laboratory experiments showed that queen conch larvae are positively phototactic and negatively geotactic, suggest- ing that most will be found near the sea surface (Barile et al. 1994). At 0.9-1.0 mm SL, the veligers are competent to undergo meta- morphosis on contact with a variety of substrata found in natural nursery areas (Davis 1994b, Davis and Stoner 1994). The morphological development of queen conch veligers was first described by D'Asaro (1965), but a comparative description of larval morphology adequate to identify different Caribbean Strombus species has only recently become available (Davis et al. 1993). The only published data on larval abundance in the field are from Lee Stocking Island (LSI), in the central Bahamas (Chaplin 1989, Chaplin and Sandt 1992. Stoner et al. 1992. Stoner et al. •Present address: Northeast Fisheries Science Center. National Marine Fisheries Service. 74 Magruder Road Highlands. NJ 07732. Present address: Harbor Branch Oceanographic Institution. 5600 Old Dixie Highway, Fort Pierce, FL 34946. 19941. and from a one-time survey of 14 stations in the eastern Caribbean Sea (Posada and Appeldoorn 1994). With the rapid decline in queen conch populations throughout the species' geographic range (Appeldoorn et al. 1987, Appel- doorn 1994). it is increasingly important to understand larval trans- port and recruitment processes on the local and regional scale. Berg and Olsen (19891 pointed out the possibility that the conch fishery in many nations might depend on upstream sources of larvae and that this should be taken into account for effective management of the species. New data, for example, suggest that recovery of the severely depleted Florida conch population will depend on the transport of larvae from Caribbean nations (Stoner et al. 1996a, Stoner et al. 1997). Here, we summarize the findings from 7 y (1988-1994) of field sampling for queen conch veligers near LSI. Our primary purpose is to provide seasonal data on the abundance of conch larvae and examine distribution over a range of habitats, from spawning sites on the island shelf to shallow regions on the Great Bahama Bank near historically important nursery grounds. We also compare an- nual abundance and size structure of veligers collected at some nursery areas that contain large and stable juvenile populations and at some that have been ephemeral. These data are useful in inter- preting the relationships between larval supply and juvenile popu- lation size and help to explain patterns of nursery distribution. METHODS Study Sites During seven spawning seasons ( 1988-1994). plankton collec- tions were made in the vicinity of LSI, Bahamas (23°46'N. 76°06'W) (Fig. 1 ), where there is only light fishing pressure. Large populations of queen conch juveniles occur on the Great Bahama Stoner and Davis 23*50' ^ EXUMA SOUND NBC 1 \ NEIGHBOR CAY *VOv 76°05' Florida W - 23"^ ^' , } Bahamas Cuba -20° \l^_ > 82" ^%W76° ^^^^^7? GRE4r BAHAMA BANK ADDER LEY ^21 CAY ■rs ,LEE STOCKING ISLAND NORMAN'S POND CAY -»^Sfl4 S7 .. t N SR3 Figure 1. Map showing 13 plankton collection stations in the vicinity of LSI. Exuma Cays, Bahamas. Stations were located in and around known nursery grounds. Station labels with asterisks indicate sites that were occupied by juvenile conch. Solid lines and arrows indicate the primary pathways for flood tidal currents. Dashed lines show secondary pathways. Elliptical areas represent the general locations of long-term nurseries near Shark Rock (SR2*I and Children's Bay Cay (CBC2*). Stippled areas represent shallow sand bars. Bank (Stoner et al. 1994). and adults are abundant in the Exuma Sound (Stoner and Schwarte 1994, Stoner and Ray 1996). The islands of the Exuma chain are bordered on the west by shallow banks (mean depth, -4 m) and on the east by the deep Exuma Sound. On flood tides, oceanic waters from the Exuma Sound flow onto the bank through numerous inlets on the flood tide and mix with bank water. For the purpose of this study, we assumed that queen conch larvae were carried from the offshore spawning sites to the bank on the tidal currents. Velocities through the 5- to 8-m-deep inlets typically reach 50-100 cm/sec at maximum flood. The tide is semidiurnal with a range of approximately 1 m. Winds are predominately from the ESE during the summer spawning season, with wind speeds typically 3-6 m/sec (Caribbean Marine Research Center, unpubl. data). To observe seasonal variation in veliger density in a reproduc- tive area, plankton collections were made at station RS. located approximately 1 km east of LSI on the island shelf in the Exuma Sound (Fig. I ). Adult conch are abundant at RS on an 1 8-m-deep platform covered with sand and algae (Stoner and Sandt 1992). Sampling schedules are explained under Plankton Collections. Drifter studies have shown that tidal waters from the Exuma Sound flood through inlets north and south of LSI and into two corresponding flow fields. The north tidal system passes over conch nurseries near Shark Rock and Tugboat Rock, and the south system passes over a nursery west of Children's Bay Cay (Stoner et al. 1994, Stoner et al. 1996b) (Fig. I ). Plankton collections were made to determine seasonal and geographic distribution of veligers with respect to the tidal current patterns. Plankton collections were made at four stations along the primary axis of tidal flow between Adderley Cay and Cook's Cay (stations SRI, SR2*. SR3, and SR4) and at two stations along a secondary axis of tidal flow between LSI and Tugboat Rock (SR5 and SR6*). Plankton col- lections were also made at three stations in the Children's Bay Cay flow field between the inlet and Windsock Cay (CBC1. CBC2*. and CBC3). One additional station was sampled in the middle of the bank (MB), between the two primary flow fields but well outside the primary tidal currents. No juvenile conch have ever been observed at the nonnursery stations, although adults are oc- casionally observed over most of the Great Bahama Bank and island shelf adjacent to LSI. Stations located in known nursery grounds are indicated with asterisks in the station code (e.g., SR2*). The Shark Rock (SR2*) and Children's Bay Cay (CBC2*) nurseries have been occupied by large, stable juvenile queen conch populations that often contain between 10 4 and 10 5 individuals (Stoner et al. 1996b). Juvenile populations at SR6* usually contain <10 4 individuals and have been more ephemeral than those at SR2* and CBC2* (Stoner et al. 1996b). All of these nursery sites are located in shallow water (2-3 m) and are covered primarily with sparse to medium-density sea- grass (Thalassia testudimtm), a similar habitat on most of the bank. Plankton collections were made at three other nursery sites known to be occupied by low numbers of queen conch juveniles. Station NPC* was west of Norman's Pond Cay and is located in a tidal flow field that begins at the north end of this cay (Fig. 1). Station NBC* was located just off the north beach of Neighbor Cay, which is approximately 8 km north of LSI. Both of these nurseries are occupied by a few thousand juvenile conch that ap- pear in some years and are absent in others (Sandt and Stoner 1993. Stoner, unpubl. data). The third station was off Charlie's Beach (CHB*) on the windward side of LSI (Fig. 1). This island Hi iKI/u\l \l DlSIKIBl'IION ()l Ql 1 I \ CliMII I \KV-\I shelf site is occupied by only a few hundred juvenile conch, which appeur irregularly. All three sites are protected from the prevailing wind (ESE), and they are in shallow (1- to 2-m) water, where the bottom is mixed seagrass and sand. Plankton Collections Surface plankton collections were made by towing simple coni- cal nets (0.5 m in diameter, 2.5 m in length) from a small boat. The nets were towed in the upper 1 m of the water column for 15-20 min at approximately 1 m/sec. The volume of water sampled, typically 200-250 nr\ was calculated with a calibrated General Oceanics flowmeter suspended off-center in the mouth of the net. Unless otherwise specified, replicate tows were made at each sta- tion. A mesh size of 202 |xm was used to collect all larval stages including newly hatched larvae. During the first year of sampling ( 1988). collections were made at RS. SRI. and CBC1 from March to October at 2-wk intervals to examine larval densities over the entire spawning season. Because very few larvae were collected in March. April, and October (see Results), sampling was concentrated between May and September in subsequent years at all stations. Nursery sites were sampled every 2 wk in 1989 and 1990 and on an approximately weekly schedule in 1992-1994. Only the reproductive site was sampled in 1991. Collections were made at nursery station SR2* in 1989 to test for possible day-night variation in estimates of veliger density. Plankton tows were made at the high tide during midday and at midnight on August 2 and 17. Because daytime sampling yielded higher densities of conch larvae (see Results), all subsequent col- lections were made during midday. The effects of tidal period on veliger density were examined at the inlet station, SRI. north of LSI, where tidal current velocities are in direct phase with tide height (i.e., zero velocity occurs with high and low water). Plankton samples were collected on 20 dates in 2-day pairs between July 14 and August 28. 1990. at approxi- mately 1-wk intervals. Collections were made every hour, when possible, between low and high tide, with greatest effort concen- trated on the midtide period. Sixteen to 18 collections were made between 2 and 4 h after low water. Seven sets of tows were made 1 h after low water, and five sets were collected 5 h after low- water. The results of this analysis provided the rationale for sam- pling time at the bank sites. All collections used in seasonal and annual analysis of veliger density in the inlets (including 1988 and 1989) were made 2 h after the beginning of the flood tide. Because of the time required for Exuma Sound water to pass onto the Great Bahama Bank, collections at stations on the bank were made dur- ing the last 2 h of the flood tide. Collections at RS. in the open sound, were made independent of tide. iMrral Identification and Staging From 1988 to 1991. plankton samples were sorted live within 4 h of collection. With the aid of a dissecting microscope (20x), live queen conch veligers could be positively identified by distinct orange pigment cells on the propodium and purplish-brown pig- ment on the edges of the velar lobes (Davis et al. 1993). Labora- tory-reared veligers were used to verify distinguishing character- istics of the most similar and abundant veligers, S. gigas and Strombus costatus. From 1992 to 1994, the samples were pre- served in 59c buffered formalin immediatelv after collection, and sorting was accomplished within 6 mo. using shell features de- scribed by Davis et al. (1993). Each sample was sorted by pouring multiple subsamples into gridded Petri dishes. If the volume of settled plankton was high, the sample was split once with a Folsom plankton splitter before being sorted. Conch veligers from each plankton tow were counted and measured for SL (apex to siphonal canal) with an ocular mi- crometer. For this study, veligers were divided into three size classes: newly hatched (300-500 p.m SL). midsize (500-900 p.m), and late stage (>900 p.m). Late-stage larvae were either competent for metamorphosis or very nearly so. Data Analysis Densities are reported as number of veligers/ 10 m 3 . The mean of the replicate tows was calculated, and in some analyses, means of means were used to describe density for a particular month or season. One-way analysis of variance ( ANOVA) was used to com- pare larval densities estimated for different tidal stages at SRI and for day-night comparisons at SR2*. RESULTS Temporal \ ariation Monthly plankton collections made in 1988 at the reproductive site and in the tidal inlets (Fig. 2) were consistent with the known summer spawning season of S. gigas in the Exuma Cays. Veligers were first collected on June 2 at RS. on June 6 at SRI. and on June 20 at CBC1. Larval densities were highest (0.26-4.46 veligers/10 m ) in the surface water during midsummer (June through Au- gust), and no veligers were found after the end of September. These results provided the rationale for sampling between May and September in subsequent years. Larval densities were 3-10 times higher during the day than at night during the peak reproductive season in 1989 at nursery site SR2* (Fig. 3). The difference was significant on both August 2 | ANOVA. F (l ,, = 16.0. p = 0.057] and on August 17 [F tl2) = 106.1. p = 0.009]. All of the veligers collected were newly hatched with no difference in SL between day and night (day: mean = 436 |j,m; SD = 28; n = 49. night: mean = 426 p.m: SD = 13; n = 5). Veliger density at the reproductive site was generally low (<1.5 veligers/10 m 3 ) and variable. High abundance (2.6-7.4 veligers/10 m 3 ) occurred only during June 1990 and early July 1991 (Fig. 4). Most veligers were newly hatched, although midsize veligers were abundant in 1991. and late-stage veligers were present later in the spawning season during both years. The highest observed density for newly hatched veligers was 7.44 veligers/10 m\ which was 13 times the maximum for midsize larvae (0.56 veligers/10 m 1 ) and almost 30 times higher than the maximum for late-stage larvae (0.26 veligers/10 nr 1 ). No veligers were collected at the reproduc- tive site on several sampling dates during the spawning season in either 1990 or 1991. In the inlet north of LSI (station SRI ). where tidal variation was examined, the highest density of queen conch veligers occurred 2 h after the onset of flood tide (Fig. 5). Because of large variation in densities, tidal phase did not have a significant effect on esti- mates of veliger density [ANOVA, F (45g) = 0.652, p = 0.628]. Nevertheless, all subsequent sampling at SRI was conducted at the time of maximum density (i.e.. 2 h after slack low water) in the davtime. 10 Stoner and Davis E o o c 2 3 a> 1 2 «^ O 'I 1 0) Q dRS • SR1 ▲ CBC1 -a4- a • a Mar Apr May Jun Jul Aug Sep Oct 1988 Figure 2. Density of queen conch veligers during 8 mo in 1988 at the reproductive site (RS) (12 sampling dates) and at two inlet sites (17 dates at SRI: 15 dates at CBC1). Values are means of two replicate tows. Spatial Variation Veliger density during three spawning seasons was consistently higher at inlet station SRI (Fig. 6) than at RS (Fig. 4). Veligers were present in the inlet from May through September, with peak abundance in July and August. Highest veliger densities occurred in 1989 (5.5 veligers/10 nr ) and 1992 (7.2 veligers/10 m 3 ). In 1989 and 1990, veligers ranged from newly hatched to 600 u.m SL. In 1992. 95% of the veligers collected in the inlet were newly to 12 E D Day o IT" ■ Night d c 8 ^^ CO k_ a> O) a> > 4 <^ o T '55 c a> a 2-Aug 17-Aug 1989 Figure 3. Day versus night comparisons for density of queen conch veligers in surface waters at station SR2* in August 1989 (mean ± SE, n = two tows per sampling period). hatched; midsize veligers were occasionally found in the inlet, with a high value of 0.24 veligers/10 m 3 (Fig. 7). Late-stage ve- ligers were collected at SRI in June, in July, and at the end of the season in September, with maximum density at 0.26 veligers/10 m 3 , similar to maxima at the reproductive site (Figs. 4 and 7). In 1988, densities of queen conch veligers in the Children's Bay Cay tidal system decreased with distance from the Exuma Sound onto the Great Bahama Bank. The mean density of veligers in July and August was four times higher in the conch nursery (station CBC2*) than at the nearby bank station CBC3. but was approximately half the density observed at inlet station CBC1 (Table 1 ). Veligers in this flow field ranged in size from newly hatched to 600 u.m SL, except for three late-stage veligers (1,350 u.m SL) collected at CBC2* in mid-July. The mean density of veligers at CBC1 (1.90 veligers/10 m 3 ) was nearly identical to the density at the adjacent inlet site SRI (1.99 veligers/10 m 3 ) (Table 1 ). Relatively low veliger densities were found at station SR5. where juvenile conch have never been observed. In 1989. veliger distribution was examined along the Shark Rock flow field, where the tidal current is confined to a more distinct channel than that near Children's Bay Cay. In July and August, mean larval densities were high (>2.3 veligers/10 m 3 ) all along the flow field from the inlet (SRI) to Cook's Cay (SR4) (Table 1 ). with highest density (4.20 veligers/10 m 3 ) at the conch nursery (SR2*). Only two larvae were collected during the entire season at the station in the middle of the bank (MB), yielding a very low mean density (0.01 veligers/10 m 3 ) Veligers collected during 1989 were 300-600 u.m in SL, with no late-stage veligers collected at any site during this season. Sampling in the Shark Rock flow field during 1992 yielded a trend similar to that observed in 1989 (Table 1 ). The highest mean density during peak reproductive season (July to August) in 1992 E o c CO l_ O) o5 > o 3? to c CD Q Horizontal Distribution of Queen Conch Larvae 8 r ii 6 - 4 - r~t newly-hatched mid-size late-stage pa T- CVJ -9- CM 3 3 < CO CD 00 Q. "~ o CO 1990 3 r 2 1 JW ??? T? fT ? C\J r-»-r-'toO''-'c\jr~.c\jcO'-coco CM— i-CVJCVJg, r- v- CM CM a -2 3 CO < CO c 3 1991 Figure 4. Mean density of queen conch veligers collected at the reproductive site (RS) during the 1990 and 1991 reproductive seasons (n = two tows per sampling date). Densities are reported by size categories. Newly hatched veligers were 300-500 um SL, midsize veligers were 500-900 um SL, and late-stage veligers were >900 um SL. occurred at SR2* (3.61 veligers/10 m 3 ). The mean density at SR6* (1.98 veligers/10 m 3 ) was similar to that at the inlet site SRI (1.94 veligers/10 m 3 ) (Table 1). High numbers of larvae collected in the Shark Rock flow field in 1992 permitted analysis of size frequency (Fig. 7). Conch ve- ligers were present at the station farthest from the sound (SR4). but all were newly hatched larvae. Midsize and late-stage veligers were found only in the vicinity of existing nurseries (SR2*. SR6*) and at the inlet (SRI) (Tables 2 and 3). There was an influx of midsize and late-stage veligers through the inlet (SRI) to Shark Rock (SR2*) and Tugboat Rock (SR6*) nurseries on one date. July 2. 1992 (Fig. 7). A high mean density of veligers was also found at the Children's Bay Cay nursery (CBC2*) during the peak larval season, twice the value recorded in 1989 (Table 1). Veliger Densities in Stable Versus Ephemeral Nurseries Annual variation in the density of queen conch veligers was observed both at the stable, long-term nursery sites (SR2* and CBC2*) and near the most ephemeral populations (SR6* and NBC*) (Table I; Figs. 7-9). All densities increased from 1992 to 1993: however, during these years when all four nursery sites were Stoner and Davis E o o CO i- O) 1 "3 > o >» c 0) Q ■ 18 l 16 l V 7 17 1 1 1 i j „ i 1 1 1 1 1 1 Hours after Low Water Figure 5. Density of queen conch veligers collected at inlet station SRI during flood tides in July and August 1990 (mean ± SE). The values above the error bars are the numbers of sampling periods. Two replicate tows were made for each period. sampled, annual variation was less at the two stable nurseries (1.1 and 2.2 times, at CBC2* and SR2*. respectively) than at the ephemeral sites (4.0 times) (Table 1 ). Spatial patterns of veliger density over the Great Bahama Bank were not always consistent. For example, in 1994, veliger density was very high at SR2*. but low in the adjacent flow field at CBC2* (Table I ). Nevertheless, SR2* and CBC2* had the highest mean densities of veligers E o 8 6 - O CO 1 4 s > o >» 2 "55 c o Q a 1989 • 1990 A 1992 among all sites sampled in every year and the highest densities of late-stage veligers (Tables 2 and 4). Midsize and late-stage veligers were relatively rare at sites with ephemeral populations of queen conch. Veligers >0.5 mm SL were collected in the vicinity of NBC*; however, this occurred only during peak reproductive months (June through August) (Table 5. Fig. 9). and densities were typically an order of magnitude less D * A "flip* A« May Jun Jul Aug Sep Months Figure 6. Density of queen conch veligers collected at inlet station SRI during three spawning seasons (1989. 199(1, and 1992). Each point represents the mean for two plankton collections. nj Horizontal Distribution of Queen Conch Larvae SR1 13 T'l' T O — 0> 9 Oa newly-hatched mid-size late-stage E o o c g> a5 > o £> (A C 0) o 12 10 • 8 6 4 2 SR2* IL T I I I I I 1 I I I I I I Ot-Bi8(\(80'-N80NIO S — R n ^ n v SR4 2 • I T 'l " l " l " l " l SR2* nnn^llnllnllll ala TT CDIO'-NID^P'-OOllDOliO CM »- CM CM «- CM C*> «- «- OJ ■- SR2* nnnQn, 1 r-owooioo>ioi/>«-c\j{\joir 5 1 c _ en i 1994 6 r SR6* T T ' l ' T 'l nllnlln cm m 1992 6 r 4 - 2 - SR6* ol t i rftrrfWTff 8"fO"-NlD^n--OC'IID01IO 1993 Figure 7. Mean density of queen conch veligers collected at seven stations in the Shark Rock flow field during three spawning seasons ( 1992. 1993, and 1994). Densities are reported by size category. Each column represents the mean for two plankton collections made on 13-14 sampling dates between May and September. Newly hatched veligers were 30(1-5(10 uni SL, midsize veligers were 500-900 um SI., and late-stage veligers were >900 um SL. than those at the sites with larger juvenile populations. In the 2 y of sampling at SR6*. late-stage larvae were common (>0. 1 ve- liger/10 m 3 ) only on single dates in both 1992 and 1993 (Table 3). The ephemeral nursery at CHB* was sampled during only one season (1992) and yielded the lowest mean density of veligers among stations sampled that year (Table 1 ); midsize veligers were never collected, and only two late-stage veligers were collected (Fig. 9). DISCUSSION The reproductive season for queen conch at LSI extends from mid-April to early October (Stoner et al. 1992): however, veligers were collected only between the end of May and late September, with the vast majority occurring in a relatively narrow period between June and August. Although no correlation has been found between water temperature and reproductivity of queen conch at the study site (Stoner et al. 1992). seasonality of larval production associated with high summer temperature may be an adaptive strategy to shorten the time to metamorphosis and improve survi- vorship through the planktonic stage (Scheltema 1986). In labora- tory culture, growth rates of queen conch veligers were highest in temperatures between 28 and 32°C, slowed at 24°C. and rapidly declined to near zero at 20°C (Stoner and Davis, unpubl. data). It is possible that production or types of phytoplankton food avail- 14 Stoner and Davis TABLE 1. Density of S. gigas veligers (no. of veligers/10 m 3 ) during the peak reproductive months (July and August) of 5 y at 13 stations near LSI, Bahamas. Years Sites 1988 1989 1992 1993 1994 Shark Rock flow field SRI SR2* SR3 SR4 SR5 SR6* Children's Bay Cay flow field CBC1 CBC2* CBC3 Middle Bank (MB) ephemeral nurseries NPC* NBC* CHB* 1.99 ±0.33 (4) 0.43 ±0.02 (2) 1.90 ±0.91 (4) 0.98 + 0.15(3) 0.23 + 0.06 (2) 0.47 ±0.18 (2) 2.41 ±0.82(4) 4.20 + 1.77(5) 2.31 ±0.80(4) 2.52 ± 1.06(4) 0.01 ±0.01 (3) 2.56 ±1.03 (3) 1.94 ±0.70 (7) 3.61 ± 1.00(7) 1.12 + 0.37(7) 1.98 ± 0.41 (7) 2.00 ±0.46 (7) 0.68 ±0.16 (7) 0.16 ±0.05 (7) 1.61 ±0.39(7) 0.50 ±0.34 (7) 1.81 ±0.87(7) 0.17 ±0.08 (7) 3.89 ± 1.22(7) 1.00 + 0.42(7) Stations located in known nursery grounds are indicated with asterisks. Values are mean ± SE (n = number of sampling dates used to calculate the mean; two replicate tows were collected on each date). able to larvae have affected the seasonality of reproduction in queen conch; however, recent measurements of chlorophyll a near LSI and throughout the Exuma Sound, in November 1993 and June 1994. have shown that concentrations of chlorophyll change little with season (A. Stoner. unpubl. data). The midsummer reproduc- tive strategy in queen conch appears to be linked primarily to physical cycles in the environment, especially temperature and photoperiod, as suggested by Stoner et al. ( 1992), and not variation in phytoplankton biomass. Because adult conch are abundant in 10- to 20-m depth all TABLE 2. Mean density of midsize (500-9(10 um SL) and late-stage (>900 urn SL) veligers of S. gigas collected during three reproductive seasons in the Shark Rock flow field at nursery site SR2*. Density of Veligers (no. • 10 m~ 3 ) 1992 1993 1994 Date Mid Late Date Mid Late Date Mid Late 5/20 6/1 6/9 6/18 7/1 7/8 7/20 7/31 8/5 8/18 8/29 9/4 9/16 0.047 0.029 0.768 0.076 0.031 0.530 0.025 0.062 0.025 0.027 5/28 6/4 6/10 6/21 6/27 7/6 7/14 7/23 7/31 8/10 8/19 8/26 9/9 9/16 0.024 o 0.014 0.058 0.015 0.019 0.019 (I 0.041 0.039 0.038 0.016 I) 5/27 6/2 6/10 6/19 6/30 7/9 7/15 7/25 8/1 8/12 8/22 8/29 9/7 9/14 0.020 0.016 0.020 0.022 0.053 0.295 0.017 along the Exuma Island chain (Stoner and Schwarte 1994, Stoner and Ray 1996) and because of the typically southeast to northwest alongshore drift (N. Smith, unpubl. data), it is likely that LSI receives larvae from spawning stocks to the south. Densities of midsize and late-stage larvae, therefore, are subject to events that are kilometers to tens of kilometers upstream in the Exuma Cays, whereas newly hatched veligers represent the local spawning stock. The significance of alongshore larval drift for the recruit- ment of dungeness crabs on the Pacific Coast of Washington was reported by McConnaughey et al. (1992). A large literature has developed with respect to how larval fishes (Rowe and Epifanio 1994) and invertebrates (Heron et al. 1994) enter estuarine nursery areas by selecting particular strata on the flood tides (i.e., selective tidal stream transport, sensu Boehlert and Mundy 1988); however, there are very few data for molluscs (Mann 1988). and the mechanisms in nonestuarine systems may be quite different. Most likely, in the Exuma Cays, conch veligers are TABLE 3. Mean density of midsize (500-900 um SL) and late-stage (>900 um SL) veligers of S. gigas collected during two reproductive seasons in the Shark Rock flow field at the small nursery site SR6*. Density of Veligers (no. • 10 m "') 1992 1993 Date Mid Late Date Mid Late 7/1 7/8 9/16 0.080 0.135 6/21 0.050 0.030 7/23 0.025 7/31 8/19 8/26 0.016 0.100 0.019 0.038 0.084 0.084 0.018 0.028 0.048 Overall abundance data are shown in Figure 7. Sampling dates with only newly hatched veligers were not included. Over- all abundance data are shown in Figure 7. Horizontal Distribution of Queen Conch Larvae 15 nfl dOILD 1992 rnn^n nnJI nlnfl Ytttttttttttt fl 1993 nD« Ol ' ■■ ■ rp 'i' T - ■ ■■■■■■■■■■■ i ■■■■ 1994 I I newly-hatched mid-size I late-stage Figure 8. Mean density of queen conch veligers collected in the Children's Bay Cay nursery area (CBC2*) during three spawning seasons (1992, 1993, and 1994). Densities are reported by size category. Each column represents the mean for two plankton collections made on 13-14 sampling dates between May and September. Newly hatched veligers were 3(10-500 uni SL, midsize veligers were 500-900 urn SL, and late-stage veligers were >900 um SL. transported northwest on the alongshore current and then through passes between the islands on flood tide. Whether or not the larvae use behavioral processes to enter the inlets or remain on the bank is unknown; however, queen conch larvae migrate vertically over a few meters or tens of meters on a diurnal periodicity (Barile et al. 1994, Stoner and Davis 19971. and they may respond to salinity or temperature gradients that occur in the inlets. The most parsimo- nious explanation for the presence of large queen conch nurseries along the Exuma Cays island chain is the net bankward flow of shelf water (Smith and Stoner 1993). Data from surface drifters indicate that once the veligers are drawn through the numerous inlets, they will be transported to nursery sites on the shallow bank (Stoner et al. 1996b). Plankton data presented in this study confirm the hypothesis that larval densities were high in the primary tidal streams, low in secondary branches, and near zero in areas that do not receive regular incur- sions of water from the Exuma Sound. These new data provide an explanation for the observation that queen conch juveniles are absent from large, seemingly appropriate benthic habitats of sea- grass outside major tidal currents in the Exuma Cays (Stoner et al. 1994). In fact, the highest larval densities occurred not only in association with the flow fields, but directly over the primary nursery ground at Shark Rock. Although significant densities of conch larvae were collected at sites farthest from the Exuma Sound (e.g., 5 km beyond the nursery), no midstage or late-stage larvae were ever found at locations beyond the nurseries. Comparable to our observations with queen conch. Field and Butler (1994) found that the postlarvae of spiny lobster (Panulirus argus) rarely settled beyond the emergent banks that ring Florida Bay. They concluded that the postlarvae were not regularly transported to the interior of E o o 05 a> > o & in c a> Q NBC* r-ir-iri ,n I I I I I I I I I I I I I I CO y- CM 4 r CHB* n l' I 'I'T I 'I p i- a> eo cm nn , . , _ EL NBC* XL -n n I T I T I YY I Y I Y I 1993 newly-hatched mid-size late-stage 1992 Figure 9. Mean density of queen conch veligers collected at the ephemeral nursery sites. Neighbor Cay (NBC*) and Charlie's Beach (CHB*), during two spawning seasons (1992 and 1993). Densities are reported by size category. Each column represents the mean for two plankton collections made on 13-14 sampling dates between May and September. Newly hatched veligers were 300-500 um SE, midsize veligers were 500-900 um SL, and late-stage veligers were >900 um SL. 16 Stoner and Davis TABLE 4. Mean density of midsize (500-900 um SL) and late-stage (>900 urn SL) veligers of S. gigas collected during the three reproductive seasons in the Children's Bay Cay flow field at nursery site CBC2*. Density of Veligers (no.- 10 m- 3 ) 1992 1993 1994 Date Mid Late Date Mid Late Date Mid Late 5/2(1 5/28 5/27 0.019 6/1 6/4 0.021 6/2 6/9 0.024 6/10 6/10 0.609 6/1 S 6/21 0.103 0.03 1 6/19 0.081 0.181 7/1 0.073 0.176 6/27 0.015 6/30 7/8 0.105 7/6 0.016 7/9 7/20 0.019 7/14 7/15 0.022 7/31 0.022 7/23 0.206 0.018 7/25 0.022 8/5 7/31 0.176 0.059 8/1 8/18 8/10 8/12 8/29 0.337 8/19 0.411 0.629 8/22 9/4 8/26 0.046 8/29 0.020 9/16 9/9 0.018 0.036 9/7 0.023 0.045 9/16 0.040 9/14 Overall abundance data are shown in Figure 9. the bay, analogous to our findings for late-stage conch larvae on the Great Bahama Bank. There is now considerable literature indicating that spatial and interannual heterogeneity in larval supply has an important influ- ence on the settlement and recruitment of invertebrates ( Yoshioka 1982, Gaines et al. 1985, Olmi et al. 1990, Bertness et al. 1992. Peterson and Summerson 1992. Martel et al. 1994) and fishes (Milicich et al. 1992). Sites with ephemeral juvenile conch popu- lations had more sporadic densities of larvae than did the larger and more stable nursery sites near Shark Rock and Children's Bay Cay. For example, the ephemeral Tugboat Rock population (SR6*) had a lower mean density of veligers than the Shark Rock nursery (SR2*). SR6* lies in a secondary tidal branch associated with the inlet north of LSI and probably does not receive oceanic water on every tide, as does station SR2*. Low tidal current velocities would also reduce the flux of larvae to a site. Densities of larvae observed at NBC* and CHB* were typically much lower than TABLE 5. Mean density of midsize (500-900 um SL) and late-stage (>900 um SL) veligers of S. gigas collected during two reproductive seasons at the ephemeral nursery site NBC*. Density of Veligers (no. • 10 m 'l 1992 1993 Date Mid Late Date Mid Late 7/1 0.025 0.025 6/27 0.016 7/8 0.025 0.025 7/31 0.030 0.058 7/20 0.040 8/10 0.016 0.033 8/26 0.016 0.017 Sampling dates with only newly hatched veligers were not included. those at the more stable nursery sites, and densities of midstage and late-stage larvae were very low and erratic. It is likely, there- fore, that population size and stability are related to the quantity and regularity of larvae arriving at a nursery. These results indicate that the importance of presettlement processes should be consid- ered in the distributional ecology and management of queen conch populations. Transplant experiments with juvenile conch near LSI (Stoner and Sandt 1992, Ray and Stoner 1994, Stoner et al. 1994) have shown that: ( 1 ) some habitats without resident conch can support juveniles, (2) conch nursery grounds are probably not saturated with juveniles in most years, and (3) recruitment probably limits the number of individuals in a nursery ground. We have also concluded that the settlement of conch larvae is not random. The Shark Rock nursery area possesses specific biological cues that induce a higher settlement rate than areas with seemingly similar general features both upstream and downstream from the nursery (Davis and Stoner 1994). Therefore, long-term queen conch nurs- eries, whether supporting large, stable populations of juveniles or small and ephemeral populations, are associated with a combina- tion of important attributes: ( I ) hydrodynamic properties that sup- ply and retain larvae, and (2) unique benthic characteristics that attract settlement of competent larvae and provide food and shelter for juveniles. Although densities of larvae, particularly late stages, probably affect the distribution and abundance of juvenile conch in the nursery habitats, the abundance of early-stage larvae is undoubt- edly influenced by the density or abundance of nearby spawners. Although few data exist for densities of queen conch larvae, some comparisons can be made on a regional scale. Densities of queen conch veligers were typically 1-2 larvae/ 10 nr during the primary reproductive season near LSI, with some densities as high as 10 veligers/10 m 3 . In the Florida Keys, densities rarely exceeded 0.5 veligers/10 nr between 1992 and 1994, even near important spawning sites (Stoner et al. 1996a. Stoner et al. 1997). The mean density of queen conch veligers in surface tows made by Posada and Appeldoorn (1994) in a July 1989 cruise along the islands of the eastern Caribbean Sea from Martinique to the Grenadines was 0.18 larvae/ 10 irr (SD = 0.33. n = 19). The highest value in a single tow was 1.22 veligers/10 nr. found downcurrent from the important conch-producing banks of the Grenadines. Densities higher than those near LSI have been found only in the northern Exuma Sound. Bahamas. Stoner and Ray (1996) reported values commonly between 25 and 50 queen conch veligers/10 m in repeated samplings in the Exuma Cays Land and Sea Park during 1993 and 1994. Most larvae in collections just described have been newly hatched individuals: therefore, regional variation in ob- served densities is probably a direct function of spawning stock size or density in the surrounding areas. The density of adult conch in the Florida Keys was <3 individuals/ha in the primary habitats in 1989 (Glazer and Berg 1994), compared with 60-90 adults/ha in 10- to 20-m depth off LSI (Stoner and Schwarte 1994) and 100- 250 adults/ha in the Exuma Cays Land and Sea Park (Stoner and Ray 1996). The presence of queen conch in offshore waters of the Carib- bean Sea (Posada and Appeldoorn 1994) and the genetic similarity of populations throughout the species' geographic range (Mitton et al. 1989) indicate that dispersal potential is high. Therefore, future research should emphasize the transport and supply of veligers on a local and regional scale. It will also be important to determine the relative importance of larval production and planktonic processes Horizontal Distribution of Queen Conch Larvae 17 in determining larval supply to nursery grounds (Meekan et al. 1993). Basic information — such as growth rates and length of larval life, survivorship in the water column, details of larval be- havior, and physical oceanography — is needed to predict transport and to understand the interdependence of local populations. Ex- periments with natural phytoplankton foods (Olson and Olson 1989) and analysis of statolith rings, which represent daily growth rates in gastropod veligers (Grana-Raffucci 1989, Bell 1993, M. Davis, unpubl. data), should be particularly valuable. Larval trans- port studies will be key to the preservation of the most important spawning populations and to the rehabilitation of this, and other, overexploited species. ACKNOWLEDGMENTS This research was supported by grants from the National Un- dersea Research Program of NOAA (U.S. Department of Com- merce) and the Shearwater Foundation (New York). A large num- ber of persons assisted in the collection, sorting, and identification of veligers over the 7 y of research; these include I. Boidron- Metairon, B. Bower-Dennis, J. Chaplin, R. Gomez, L. Hambrick. R. Jones, C. Kelso, J. Lally, E. Martin, K. McCarthy, N. Mehta, S. O'Connell, M. Ray. V. Sandt. K. Schwarte. and E. Wishinski. N. Mehta, M. Ray. and anonymous reviewers helped to improve the manuscript. LITERATURE CITED Appeldoorn. R. S. 1994. Queen conch management and research: status, needs and priorities, pp. 301-319. In: R. S. Appeldoorn and B. Ro- driquez (eds.). Queen Conch Biology, Fisheries and Mariculture. Fun- dacion Cientifica Los Roques, Caracas, Venezuela. Appeldoorn, R. S., G. D. Dennis & O. Monterrosa L. 1987. Review of shared demersal resource of Puerto Rico and the Lesser Antilles region. FAO Fish. Rept. 383:36-106. Barile, P. J.. A. W. Stoner & C. M. Young. 1994. Phototaxis and vertical migration of the queen conch 'Strombus gigas Linne) veliger larvae. J. Exp. Mar. Biol. Ecol 183:147-162. Bell. J. L. 1993. Feeding and growth of prosobranch veligers. Ph.D. Dis- sertation, Department of Zoology. University of Hawaii, Oahu, Hawaii. 172 pp. Berg. C. J.. Jr. & D. A. Olsen. 1989. Conservation and management of queen conch (Strombus gigas) fisheries in the Caribbean, pp. 421-442. In: J. F. Caddy (ed.). Manne Invertebrate Fisheries; Their Assessment and Management. John Wiley and Sons. New York. Bertness, M. D , S. D. Gaines. E. G. Stephens & P. O. Yund. 1992. Com- ponents of recruitment in populations of the acorn barnacle Semibala- nus balanoides (Linnaeus). J. Exp. Mar. Biol. Ecol. 156:199-215. Boehlert. G. W. & B. C. Mundy. 1988. Roles of behavioral and physical factors in larval and juvenile fish recruitment to estuarine nursery areas. Am. Fish. Soc. Symp. 3:51-67. Chaplin. J. 1989. The vertical and geographic distribution of queen conch (Strombus gigas) larvae in the Exuma Cays. Bahamas. M.S. Thesis. Department of Oceanography, University of Southampton. Southamp- ton. United Kingdom. 40 pp. Chaplin. J. & V. J. Sandt. 1992. Vertical migration and distribution of queen conch veligers. Proc. Gulf Caribb. Fish. Inst. 42:158-160. Creswell, R. L. 1994. An historical overview of queen conch mariculture. pp. 223-230. In: R. S. Appeldoorn and B. Rodriquez (eds.). Queen Conch Biology, Fisheries and Mariculture. Fundacion Cientifica Los Roques. Caracas. Venezuela. D'Asaro. C. N. 1965. Organogenesis, development and metamorphosis in the queen conch, Strombus gigas. with notes on breeding habits. Bull. Mar. Sci. 15:359—416. Davis, M. 1994a. Mariculture techniques for queen conch (Strombus gigas L.): egg mass to juvenile stage, pp. 231-252. In: R. S. Appeldoorn and B. Rodriquez (eds.). Queen Conch Biology. Fisheries and Mariculture. Fundacion Cientifica Los Roques. Caracas, Venezuela. Davis. M. 1994b. Short-term competence in larvae of queen conch Strom- bus gigas: shifts in behavior, morphology and metamorphic response. Mar. Ecol. Prog. Ser. 104:101-108. Davis, M„ C. A. Bolton & A. W. Stoner. 1993. A comparison of larval development, growth, and shell morphology in the three Caribbean Strombus species. Veliger 36:236-244. Davis, M. & A. W. Stoner. 1994. Trophic cues induce metamorphosis of queen conch larvae (Strombus gigas Linnaeus). J. Exp. Mar. Biol. Ecol. 180:83-102. Field. J. M. & M. J. Butler IV. 1994. The influence of temperature, salinity. and postlarval transport on the distribution of juvenile spiny lobsters Panulirus argus (Latreille. 1804), in Florida Bay. Crustaceana 67:32- 45. Gaines. S., S. Brown & J. Roughgarden. 1985. Spatial variation in larval concentrations as a cause of spatial variation in settlement for the barnacle. Balanus glandula. Oceologia (Berlin) 67:267-272. Glazer, R. A. & C. J. Berg, Jr. 1994. Queen conch research in Florida: an overview, pp. 79-95. In: R. S. Appeldoorn and B. Rodriquez (eds.). Queen Conch Biology. Fisheries and Mariculture. Fundacion Cientifica Los Roques. Caracas, Venezuela. Grana-Raffucci. F. A. 1989. Age determination of two Strombid gastro- pods by statolith microstructure interpretation. M. S. Thesis. Depart- ment of Marine Sciences, University of Puerto Rico. Mayaquez, Puerto Rico. 47 pp. Heron, M. L.. H. X. Wang & D. J. Staples. 1994. Transport processes affecting banana prawn postlarvae in the estuaries of the Gulf of Car- pentaria, pp. 253-277. In: P. W. Sammarco and M. L. Heron (eds.). The Bio-Physics of Marine Larval Dispersal. American Geophysical Union. Washington. D.C. Laughlin. R. A. & E. Weil. 1983. Queen conch mariculture and restoration in the Archipielago de Los Roques: preliminary results. Proc. Gulf Caribb. Fish. Inst. 35:64-72. Mann. R. 1988. Distribution of bivalve larvae at a frontal system in the James River. Virginia. Mar. Ecol. Prog. Ser. 50:29—14. Martel. A.. A. F. Mathieu. C. S. Findlay, S. J. Nepszy & J. H. Leach. 1994. Daily settlement rates of the zebra mussel, Dreissena polymorpha. on an artificial substrate correlate with veliger abundance. Can. J. Fish. Aquat. Sci. 51:856-861. McConnaughey, R. A., D. A. Armstrong. B. M. Hickey & D. R. Gunder- son. 1992. Juvenile dungennes crab (Cancer magister) recruitment variability and oceanic transport during the pelagic larval phase. Can. J. Fish. Aquat. Sci. 49:2028-2044. Meekan. M. G., M. J. Milicich & P. J. Doherty. 1993. Larval production drives temporal patterns of larval supply and recruitment of a coral reef damselfish. Mar. Ecol. Prog. Ser. 93:217-225. Milicich. M. J., M. G. Meekan & P. J. Doherty. 1992. Larval supply: a good predictor of recruitment of three species of reef fish (Pomacen- tndae). Mar. Ecol. Prog. Ser. 86:153-166. Mitton, J. B.. C. J. Berg, Jr. & K. S. Orr. 1989. Population structure, larval dispersal, and gene flow in the queen conch, Strombus gigas. of the Caribbean. Biol. Bull. 177:256-362. Olmi, E. J.. III. J. van Montfrans. R. N. Lipcius, R. J. Orth & P. W. Sadler. 1990. Variation in planktonic availability and settlement of blue crab megalopae in the York River, Virginia. Bull. Mar. Sci. 46:230- 243. Olson, R. R. & M. H. Olson. 1989. Food limitation of planktotrophic ma- rine invertebrate larvae: does it control recruitment success' 1 Annu. Rev. Ecol. Syst. 20:225-247. Peterson. C. H. & H. C. Summerson. 1992. Basin-scale coherence of popu- 18 Stoner and Davis lation dynamics of an exploited marine invertebrate, the bay scallop: implications of recruitment limitation. Mar. Ecol. Prog. Ser. 90:257- 272. Posada. J. & R. S. Appeldoom. 1994. Preliminary observations on the distribution of Strombus larvae in the eastern Caribbean, pp. 191-200. In: R. S. Appeldoom and B. Rodnquez (eds.). Queen Conch Biology. Fisheries and Mariculture. Fundaci'on Cientiffca Los Roques, Caracas, Venezuela. Ray. M. & A. W. Stoner. 1994. Experimental analysis of growth and survivorship in a marine gastropod aggregation: balancing growth with safety in numbers. Mar. Ecol. Prog. Ser. 105:47-59. Rowe, P. M. & C. E. Epifanio. 1994. Tidal stream transport of weakfish larvae in Delaware Bay, USA. Mar. Ecol. Prog. Ser. 1 10:105-1 14. Sandt. V. J. & A. W. Stoner. 1993. Ontogenetic shift in habitat by early juvenile queen conch, Sirombus gigas: patterns and potential mecha- nisms. Fish. Bull. U.S. 91:516-525. Scheltema. R. 1986. On dispersal and planktonic larvae of benthic inver- tebrates: an eclectic overview and summary of problems. Bull. Mar. Sci. 39:257-268. Siddall. S. E. 1983. Biological and economic outlook for hatchery produc- tion of juvenile queen conch. Proc. Gulf Caribb. Fish. Inst. 35:46-52. Smith N. P. & A. W. Stoner. 1993. Computer simulation of larval transport through tidal channels: role of vertical migration. Est. Coast. Shelf. Sci. 37:43-58. Stoner, A. W. & M. Davis. 1997. Abundance and distribution of queen conch veligers (Strombus gigas Linne) in central Bahamas: II. Vertical patterns in nearshore and deep-water habitats. J. Shellfish Res. 16:19- 29. Stoner, A W., R. A. Glazer & P. J. Barile. 1996a. Larval supply to queen conch nurseries: relationships with recruitment process and population size. J. Shellfish Res. 15:407-420. Stoner, A W., M D. Hanisak, N. P. Smith & R. A. Armstrong. 1994. Large- scale distribution of queen conch nursery habitats: implications for stock enhancement, pp. 169-189. In: R.S. Appeldoom and B. Rod- riguez (eds.). Queen Conch Biology. Fisheries and Mariculture. Fun- daci'on Cientiffca Los Roques, Caracas. Venezuela. Stoner. A. W.. N. Mehta & T.N. Lee. 1997. Recruitment of Strombus veligers to the Florida keys reef tract: relation to hydrographic events. J. Shellfish Res. 16:1-6. Stoner. A. W.. P. A. Pitts & R. A. Armstrong. 1996b. The interaction of physical and biological factors in the large-scale distribution of juvenile queen conch in seagrass meadows. Bull. Mar. Sci. 58:217-233. Stoner. A. W. & M. Ray. 1996. Queen conch. Strombus gigas. in fished and unfished locations of the Bahamas: effects of a marine fishery reserve on adults, juveniles, and larval production. Fish. Bull. U.S. 94:551-565. Stoner, A. W.. V. J. Sandt & I. F. Boidron-Metairon. 1992. Seasonality in reproductive activity and larval abundance of queen conch Strombus gigas. Fish. Bull. U.S. 90:161-170. Stoner. A. W. & V.J. Sandt. 1992. Population structure, seasonal move- ments and feeding of queen conch, Strombus gigas, in deep-water habitats of the Bahamas. Bull. Mar. Sci. 51:287-300. Stoner, A. W. & K. C. Schwarte. 1994. Queen conch. Strombus gigas, reproductive stocks in the central Bahamas: distribution and probable sources. Fish. Bull. U.S. 92:171-179. Yoshioka. P. M. 1982. Role of planktonic and benthic factors in the popu- lation dynamics of the bryozoan Membranipora membranacea. Ecol- ogy 63:457^68. Journal of Shellfish Research. Vol. 16. No. 1. 19-29, 1997. ABUNDANCE AND DISTRIBUTION OF QUEEN CONCH VELIGERS (STROMBUS GIGAS LINNE) IN THE CENTRAL BAHAMAS. II. VERTICAL PATTERNS IN NEARSHORE AND DEEP-WATER HABITATS ALLAN W. STONER* AND MEGAN DAVIS f Caribbean Marine Research Center 805 E. 46th Place Vero Beach, Florida 32963 ABSTRACT The vertical distribution of queen conch veligers was investigated in three different habitats in the central Bahamas: ( 1 ) over an island shelf reproductive site ( 18 m deep) near Lee Stocking Island. Exuma Cays. Bahamas; (2) in a tidal channel leading from the reproductive ground to nursery grounds on the bank (8 m deep); and (3) in the deep-water basin of Exuma Sound. At the reproductive site, early-stage larvae were most abundant near the surface (0- to 1-m depth) during calm periods. Concentrations were highest in the deepest layer sampled (16 m) during periods of moderate swell and wave height. Day/night variation was not consistent, and vertical distribution appeared to be more closely associated with surface conditions than time of day. In the tidal channel, conch veligers were most abundant in the upper 1 m during day. night, and crepuscular hours. Lower concentrations of veligers were found in the neuston and at 3- and 6-m depths. Midstage (500-900 u.m shell length) and late-stage (>900 p_m) larvae were relatively rare at the nearshore reproductive and tidal channel sites. At the deep-water site, offshore in the Exuma Sound, queen conch veligers of all stages (mostly midstage and late stage) were collected at all depths sampled, from the surface to 100 m. However, the larvae were concentrated in the upper mixed layer, which was 25-30 m deep. During the day. 83% of the larvae were found in the 0.5- to 5.0-m-deep layer, well above the pycnocline. Low concentrations were found at all other depths, including the upper 0.5 m. At night, the larvae were evenly distributed between 0.5- and 30-m depths. These data corroborate laboratory experiments showing positive phototaxis in queen conch larvae up to very high light levels and disruption of phototaxis at night when the light cue is weak. The fact that few larvae were found below the thermocline. even at night, suggests that they are adapted to remain in warm surface waters, where growth is maximum. KEY WORDS: Bahamas, oceanography. Strombus gigas, vertical migration INTRODUCTION Vertical distribution and diel migration are common phenom- ena in marine zooplankton. including invertebrate larvae. Depth regulation is influenced by exogenous factors such as presence of predators (Bollens and Frost 1989, Forward 1988, Neill 1990), distribution of food (Enright 1977. Dagg 1985. Daro 1988). changes in salinity (Sulkin 1984. Mann et al. 1991). and intensity of light (Forward 1976. Kaartvedt et al. 1987, Swift and Forward 1988). Physical parameters, such as tidal currents, may also affect the vertical position of larvae in the water column and can be responsible for the retention and transport of larvae to favorable settlement habitats (Hill 1991, Olmi 1994). The queen conch, Strombus gigas, is an important fisheries' species in the Caribbean region, and many populations have suf- fered overexploitation during the past two decades (Berg and Olsen 1989, Appeldoorn 1994). For effective fisheries and con- servation management of the species, it is necessary to determine larval sources and their dispersal mechanisms to preserve key spawning populations and to predict juvenile recruitment. The ba- sic life history of juvenile and adult queen conch is well studied (Randall 1964. Brownell and Stevely 1981. Stoner et al. 1994), but the natural history of the planktotrophic conch veliger is poorly known. Data on the geographic distribution and abundance of veligers have been published only recently (Stoner et al. 1992. Posada and Appeldoorn 1994. Stoner et al. 1994). Although pre- liminary analyses of vertical distribution revealed that queen conch veligers were most abundant at the surface ( 1 m deep) during the *Present address: Northeast Fisheries Science Center. National Marine Fisheries Service. 74 Magruder Road, Highlands, NJ 07732. Present address: Harbor Branch Oceanographic Institution. 5600 Old Dixie Highway. Fort Pierce. FL 34946. day (Chaplin and Sandt 1992, Stoner and Davis 1997), they have been collected from depths as great as 30 m in deep-water habitats of the eastern Caribbean Sea (Posada and Appeldoorn 1994). Recent laboratory and field mesocosm experiments showed that queen conch veligers are positively phototactic up to very high light levels and that this taxis decreases with age (Barile et al. 1994). Conch veligers swam toward the water surface in both light and dark conditions, suggesting that negative geotaxis is also im- portant for orientation in this species. A detailed analysis of the vertical distribution of queen conch veligers in the field is reported for the first time in this study. Collections were made in waters near Lee Stocking Island (LSI). Exuma Cays, Bahamas, an area characterized by large spawning populations (Stoner and Schwarte 1994, Stoner and Ray 1996). We provide data on day and night vertical distribution and abundance of veligers at three different locations — a spawning site on the island shelf, a tidal channel between reproductive and nursery habitats, and deep oceanic water. These field studies contribute information on stimuli that control vertical distribution in the ve- ligers and will be critical in modeling larval transport and recruit- ment potential. MATERIALS AND METHODS Vertical Sampling in the Shelf and Channel Locations Stratified vertical sampling for veligers was conducted at a well-studied reproductive site (RS) and a tidal channel station (SRI), both near LSI in the central Bahamas (Fig. 1). Detailed descriptions of the study site and tidal circulation were provided in earlier publications (Stoner et al. 1994. Stoner et al. 1996. Stoner and Davis 1997). RS was approximately 1 km to the east of LSI. on the island shelf in Exuma Sound (ES), where a spawning popu- 19 20 Stoner and Davis ELEUTHERA \ \ GREAT BAHAMA BANK TONGUE ot the OCEAN ~r~ -23° 9 ES V CAT ISLAND \ \ Exuma Sound ^^m f Bahama Florida^* i. *v « y. * .■ ^>' SAN SALVADOR, .*;-. SR1 >^® RS Lee Stocking ^V - . c' r : :%>. \ '1— "200 m \ j :' >,' £•0 t Figure 1. Map showing three sites where vertical plankton collections were made in the ES area of the central Bahamas. RS was the reproductive site on the island shelf immediately offshore from LSI. SRI was in the tidal channel north of LSI. ES was a deep-water site in the northern sound. lation occupies a sand- and algae-covered platform 18 m deep (Stoner and Sandt 1992). In 1990, collections were made at four depths at RS: at the air-sea interface (neuston). in the upper 1 m (surface), at 8-m depth (midwater). and at 16-m depth (near- bottom). A simple conical net (0.5 m in diameter, 2.5 m in length) with 202-u.m mesh was used to collect all larval stages, including newly hatched veligers, which have a maximum shell dimension of ap- proximately 300 p.m. Nets were generally towed with a small boat (6 m) at 1.0 m/sec for 10-15 min in the downwind and alongshore direction (usually northwest). A calibrated General Oceanics flow- meter suspended off-center in the mouth of the net was used to estimate water volume sampled, typically 200-250 nr\ The neus- ton layer was sampled by towing the net with the middle of the ring at the air-sea interface. Tows at depth did not use an opening/ closing mechanism, but the nets were lowered and raised through the water column while the boat was stationary. Lead weights suspended from the net ring allowed for towing at 8- and 16-m depth and for quick lowering to avoid depth contamination. The position of the near-bottom net was maintained by permitting the weight, suspended 2 m below the ring, to touch the bottom peri- odically. The depth of the midwater net was estimated by wire angle and by eye in the very clear water. Replicate tows (n = 2) were made for each depth on each sample date. Samples were collected in June, July, and August, with seven vertical series collected during the day ( 1 100-1400 h) and two complete series collected for day and night (2300-0300 h) comparisons. Wind speed and direction, wave height, and cloud coverage were esti- mated and recorded at the time of each collection. The second site, SRI, was located on the bank in the tidal channel north of LSI (Fig. 1). Conch larvae are carried through this inlet to a large, well-studied nursery ground west of LSI (Stoner et al. 1996, Stoner and Davis 1997). High current velocities on flood tide (to 2.0 m/sec) made it possible to collect plankton with fixed nets moored at three different depths. In 1989 and 1990, a taut mooring was rigged with a large concrete block and surface bouys. Nets identical to those described above were attached to the moor- ing at 3- and 6-m depth using SCUBA and were retrieved 30 min later. The nets were kept closed by the diver during deployment and retrieval. At the same time, surface water (upper 1 m) was sampled using a net suspended from an outrigger on a boat that was anchored adjacent to the mooring. In 1990, the neuston layer was also sampled using the outrigger mechanism. Because of high current velocities at the surface, the duration of sampling was reduced to 20 min. The average volume of water sampled at each depth was 160 m\ Collections were made during the middle of the flood tide, when the tidal current was strongest and when veliger abundance was highest (Stoner and Davis 1997). In 1989, vertical collections for veligers at SRI were made for three day /night series in July and August; one net was set at each depth on the mooring on each sampling date. In 1990, 14 vertical collections were made during the day (0700-1900 h), during the night (2300-0300 h), and during crepuscular hours (approximately 0500 and 2100 h) be- tween June and August. Collections were replicated (n = 2) for each of the four depths on each sampling date. Vertical Sampling al the Deep-Water Site Vertical distribution of conch veligers was examined to depths of 100 m in the deep-water basin of ES during June 1994. The 18-m R/V Shadow was maintained on station 90 km northwest of LSI and 18 km east of the Exuma Cays (24°30'N. 76°34.5'W) (Fig. 1). This site is approximately 1,600 m deep and was chosen because sampling at the site in 1993 yielded high concentrations of conch larvae (primarily midstage and late stage) (Stoner, unpubl. data). Vertical Distribution of Queen Conch Larvae 21 % of Veligers by Depth 20 40 60 80 100 1 8 16 1 r I i i 1 84/56 -- 8/20 "'_ H 2/ 7 12 Jun 1990 20 40 1 1 — 60 80 100 a. o a o 1 8 16 i r 6 Sep 1990 46/18 33/20 20 40 60 80 100 1 1 H 6/1 1 19/35 1 ■ n 18 Sep 1990 -) 69/62 16 Figure 2. Percentage of veligers at each of four depths during daytime collections made at the RS on two dates in 1990 {mean ± SE, n = 2 tows at each depth). Values next to the error bars represent the actual number of veligers collected in Tow 1 and Tow 2. Collections were made during the day (1 100-1500 h) on June 23 and the following night (2200-0300 h). Nets were similar to those described earlier, except that they were equipped with Gen- eral Oceanics double-trip, opening-closing mechanisms for de- ployment on hydrographic wire. While maintaining a vessel speed of 1 .0 m/sec, nets were set over the side from a long boom where they were unaffected by the ship's wake. Oblique tows were made TABLE I. Size and density of veligers collected at the RS in 199(1 during three daytime sampling dates at three depths. Veliger Stage Location Newly Hatched Mid Late June 12 Surface (1 m) 4. 17 ±0.71 0.14 ± 0.14 Midwater (8 m) 1.05 + 0.31 Near-bottom ( 16 m) 0.3410.13 September 6 Surface (1 m) Midwater (8 m) 2.20 ±0.68 Near-bottom (16 m ) 1.81 ±0.29 September 18 Surface (1 m) 0.04 ± 0.04 0.26 + 0.09 Midwater (8 m) 1.93 ±0.41 Near-bottom (16 m) 4.81 +0.13 0.04 ± 0.04 No veligers were collected in the neuston. Newly hatched veligers were 300-500 u.m SL. midsize veligers were 500-900 |xm SL, and late-stage veligers were >900 u.m SL. Values are mean number of veligers/10 m (+SE) (n = 2 tows). over depth intervals of 0-0.5, 0.5-5. 5-10, 10-30, 30-60. and 60-100 m. The nets were sent to the greatest depth of the desired sampling interval (calculated from wire angle), opened at depth, slowly raised through the water column to the upper depth limit over a period of approximately 15 min, and then closed and re- trieved. Nets set for the two shallowest depths were left open. To maintain precise control of sampling intervals, nets were set one at a time on the wire. All tows were replicated twice in random order for each time period. Profiles of temperature and salinity were made to 150-m depth with a SeaBird CTD at the beginning and end of both day and night collections. Identification and Staging of Veligers In 1989 and 1990, plankton samples were sorted live within 4 h after collection. In 1994, samples were preserved in 5% buffered formalin immediately after collection and were sorted within 5 mo. Veligers of S. gigas were identified by comparison with labora- tory-reared specimens and by using shell features described by Davis et al. (1993). Sorting and measuring techniques were sum- marized by Stoner and Davis (1997). Conch larvae were classified as newly hatched (300-500 p. shell length [SL] ). midsize (500-900 (xm), or late stage (>900 p,m), most of which were competent for metamorphosis. Data Analysis Larval densities were standardized by converting counts to number of veligers/10 m 3 . Means of replicate tows were calculated and. in some cases, were used to calculate the mean density of veligers for a particular time of day or depth. Because of temporal variation in absolute larval densities, vertical abundance data were converted to percentages of larvae at each depth. Veliger densities were low at SRI in 1990. When fewer than four veligers were collected on a particular date, these samples were not considered in the analysis. The x 2 goodness of fit test was used to analyze percent distribution for the 1990 data from SRI. with the null hypothesis that veliger distribution was homogeneous over depth. 22 Stoner and Davis One-way analysis of variance, followed by Tukey's multiple com- parison test, was used to determine if the depth distribution and abundance of veligers in the deep-water ES site were significantly different between day and night. The proportional data were arc- sine transformed to reduce heteroscedasticity. RESULTS RS The daytime vertical distribution of veligers was examined to 16-m depth (near-bottom) at the offshore RS in 1990 (Fig. 2). Large temporal variation in the vertical distribution of veligers was associated with changing weather and sea conditions. On June 12, a calm, overcast day (3.6 m/sec wind; 0.5-m sea height; no swell), conch veligers were concentrated (74%) in the upper 1 m of the water column. On September 6 and 18, in association with turbu- lent sea conditions, the depth distributions were distinctly deeper. The majority of veligers were collected at 8 and 16 m. with few to none collected in the neuston and at 1-m depth (Fig. 2). Fewer than 10 veligers were collected on each of the four other sampling dates (June 7. July 12, August 8. and September 3), and results are not shown. On these dates, winds were 6-10 m/sec and seas were 1-2 m. The percentage of veligers at different depths showed no direct correlation with wind speed (r < 0.2; p > 0.5) but may have been inversely related to sea height, which was not measured precisely. Conch veligers were never collected in the upper 1 in of the water column when waves were breaking into white caps in winds >8 m/sec. The majority of the veligers found at RS were newly hatched. Midsized veligers were found only on June 12. and September 18, 1990, just four veligers in the surface water and one near the bottom, respectively (Table 1 ). Late in the reproductive season (September 18, 1990), seven late-stage veligers (0.26 veligers/10 nv ) were found in the surface water. No clear diurnal vertical pattern of distribution was discernible from the day/night vertical collections of veligers made during two relatively calm 24-h periods at RS (<3 m/sec wind; <0.3-m sea height; no swell) (Fig. 3). At night in June, most veligers (78%) were collected in surface water (1-m depth). During midday, the abundances of veligers were relatively similar at 1- and 8-m depth. % of Veligers by Depth 100 80 60 40 20 20 40 60 80 100 ( 1 1 1 i 1 8 16 22/10 60/145 H 58/106 □ Day ■ Night a. s 100 80 60 40 20 20 40 60 80 100 i 1 i 1 1 1 1 1 1 1 1 14/20 HTOH^^H 16 34/36 12/22 46/30 1 Aug 1990 Figure 3. Percentage of veligers at each of four depths during the day and night at the RS on three dates in 1990 (mean ± SE, n = 2 tows at each depth). Values next to the error bars represent the actual number of veligers collected in Tow 1 and Tow 2 and do not necessarily parallel the results in percentages, which were standardized per unit of water volume sampled. Vertical Distribution of Queen Conch Larvae 23 In August, a high percentage (49%) of veligers was collected in the neuston at night, whereas in the daytime, the majority of veligers (61%) were near the bottom. The overall density of veligers was higher in the daytime than at night, with the highest density (7.44 veligers/10 m 3 ) estimated for daytime surface waters on June 1 1. 1990 (Table 2). Most veligers collected at RS in day/night series were newly hatched (Table 2). Only one midsized veliger was collected, and this individual was in surface water during the day on August I, 1990. Slightly more abundant late-stage veligers were collected on several occasions during the day and night in the neuston, surface, and near-bottom waters. Relatively low densities of midstage and late stage (<0.17 veligers/10 m 3 ) (Table 2) precluded conclusions about vertical or day/night distribution in these stages in the island shelf habitat. Channel Site (SRI) Veligers collected in the inlet in 1989 were distributed through- out the water column, apparently the result of surface conditions. In the daytime on July 7 and August 7, light wind (2.5-3.6 m/sec) and mild sea conditions (<0.5-m height) resulted in the highest percentages of veligers in surface waters (Fig. 4), approximately twice the densities estimated for 3- and 6-m depths (Table 3). In contrast, in the daytime on July 24. more turbulent surface condi- tions (5-7 m/sec winds. 0.6- to 1.3-m sea height) resulted in a lower percentage of veligers in the upper water column than at 3- and 6-m depths. At night, the numbers of veligers collected were very low, and vertical patterns were inconsistent (Fig. 4). On July 6, 1989, veligers were found exclusively at 3 m; on July 23, they were found at the surface and at 3 m; and on August 6, the dis- tribution was relatively uniform. The higher percentage of veligers near the surface on the night of July 23, compared with the low percentage of veligers during the following day, probably resulted from weather conditions that were calmer at night than during the day. On all three sample dates in 1989, the density of veligers was consistently higher in the day (0.93-5.46 veligers/10 m 3 ) than at night (0-0.25 veligers/10 m 3 ) at all three depths (Table 3). SL did not vary from day to night or by depth. All veligers were newly hatched. 417-450 |xm. More intensive sampling resulted in a clearer depth distribution of veligers in 1990 (Fig. 5) than in 1989. The percentages of veligers were always highest in the surface 1 .0-m depth during all three time periods — day, crepuscular, and night. Densities in the neuston layer were highest at night but were always lower than in the water column immediately below (Fig. 5b). The depth distri- butions were significantly different from homogeneous in collec- tions made both with and without neuston samples (x"ooi 3 > 40.0, and x"om. 2 > - 4 -6- respectively). The majority of the 727 veligers collected in the 1990 surveys were newly hatched. Three late-stage veligers were collected at 1-m depth on July 29, and two midsize and one late-stage veliger were collected in the same upper layer on August 27 (Tables 4 and 5). Differences in absolute concen- trations of veligers at SRI between day and night observed in 1989 were not apparent in 1990. Deep-Water Site A distinct vertical distribution of queen conch veligers and an indication of diel variation were evident in the oceanic water of the ES during the relatively calm sampling period (wind < 5 m/sec; sea < 1.0 m) (Fig. 6). As observed in the nearshore habitats, only a small percentage of veligers were located in the neuston during both day and night: however, night concentrations in the neuston were higher than during the day (Fig. 6). Depth differences in the distribution of veligers were significant in the daytime [F (5 6) = 27.34, p < 0.001]. The highest percentage (83%) of veligers was observed in the surface layer (0.5-5 m) (Tukey's test, p < 0.002); abundances at all other depths were low and not different (p > 0.646) (Fig. 6). Even though the majority of veligers occurred between 0.5- and 30-m depth at night, the abundance of veligers did not vary significantly with depth [F c5-6) = 1.20, p = 0.407], probably because of the low number of replicate samples (n = 2 at each depth). Veligers were collected as deep as 100 m during both day and night; however, >95% of all veligers were collected above 30 m. Very few early-stage larvae were collected at the deep-water site in ES (Fig. 7). in contrast to the mostly early-stage larvae collected at RS and SRI. The highest recorded mean density of late-stage veligers (1.56 veligers/10 m 3 ) was found in night tows at 5- to 1 0-m depth; this was more than twice the maximum value TABLE 2. Size and density of veligers collected at the RS in 1990 during two day/night sampling series at four depths (neuston = 0-0.5 m; surface m; mid-water = 8 m; and near-bottom = 16 m). Day Night Veliger Stage Veliger Stage Location Newly Hatched Mid Late Newly Hatched Mid Late June 1 Surface Mid Bottom August 1 7.44 ±2.38 5.83 ± 1.27 0.61 ±0.05 1.00 ±0.37 0.29 ±0.21 0.10 ±0.02 0.03 ± 0.03 Neuston Surface Mid Bottom 0.70 ±0.40 0.79 ±0.29 2.5 ±0.36 0.02 ± 0.02 0.17 + 0.17 0.11 ±0.06 2.38 ±0.29 1.46 ±0.01 0.52 0.40 ±0.10 0.05 ± 0.05 0.04 ± 0.04 Newly hatched veligers were 300-500 urn SL, midsize veligers were 500-900 u.m SL, and late-stage veligers were >900 \xm SL. Values are mean number of veligers/10 m 1 (± SE) (n = 2 tows). 24 Stoner and Davis g .c «-> a. % of Veligers by Depth 100 80 60 40 20 20 40 60 80 100 1 r 1 1 1 1 — i — i — I 1 1 40 HI 20 18 □ Day ■ Night 6-7 Jul 1989 100 80 60 40 20 20 40 60 80 100 i 1 1 1 1 1 1 1 1 1 1 f 3 23-24 Jul 1989 100 80 60 40 20 20 40 60 80 100 6-7 Aug 1989 Figure 4. Percentage of veligers at each of three depths during the day and night in the tidal channel (SRI I on three dates in 1989 (n at each depth). Values next to the hars represent actual number of veligers collected in the tow. 1 tow estimated for day collections (0.68 veligers/10 m 3 at 0.5-5 m) (Fig. 7) and six times the highest value found at the RS (0.26 veligers/10 m 3 ) (Table 1). The vertical and diurnal patterns of density in midsize veligers were similar to those of late-stage larvae (Fig. 7). Midsize larvae were very abundant (2.0 veligers/10 m 3 ) during the day in 0.5- to 5-m depth. Results from day and night CTD casts at the deep-water site were virtually identical: therefore, only one representative is shown (Fig. 8 1. A well-defined upper mixed layer to 25- to 30-m depth was clearly indicated by the strong discontinuities in tem- perature, salinity, and density. A lens of water with slightly lower salinity than the surface layer was centered at approximately 35 m. The surface layer where most queen conch veligers were found had a temperature of 29°C and a salinity of approximately 37 ppt. DISCUSSION In a preliminary investigation of vertical distribution in queen conch veligers, Chaplin and Sandt (1992) concluded that the ve- ligers move upward during the day and downward at night, a reverse diumal vertical migration. This conclusion was based pri- marily on the low abundance of veligers detected in surface waters during the night. In conflict with this reverse migration hypothesis. Barile et al. (1994) found that queen conch larvae of all stages Vertical Distribution of Queen Conch Larvae 25 TABLE 3. Density of veligers collected in the tidal channel (SRI) in 1989 during three day/night sampling series at three depths. Density of Veligers (no./IO m l ) Location Day Night July 6-7 Surface (1 m ) 5.46 Midwater (3 m) 1.93 Near-bottom (6 nil 1.70 Julv 23-24 Surface (1 ml 0.93 Midwater (3 m) 1.14 Near-bottom (6 ml 1.71 August 6-7 Surface (1 ml 5.32 Midwater (3 m) 2.32 Near-bottom (6 m) 2.4(1 o.os 0.23 0.09 0.17 0.25 0.23 All veligers were newly hatched (300-500) u.m SL). Values are for one net set at each sampling depth. migrated upward at night in 3-m-deep field mesocosms. On the basis of several laboratory experiments, they concluded that depth distribution in this species is affected by positive phototaxis. nega- tive geotaxis. and negative phototaxis at very high light intensity. Similar behavior has been observed for larvae of the gastropod Phestilla sibogae (Miller and Hadfield 1986). Whether an endog- enous rhythm is entrained in the vertical movements of queen conch veligers is still unresolved. Barile et al. (1994) suggested that vertical distribution in the species is associated with particular light levels rather than specific rhythms. Variation in vertical distribution patterns in the field (this study) shows that observations from the laboratory should be ex- trapolated to the field with care. Under calm surface conditions, the majority of larvae (all stages) were found in near-surface wa- ters during the daytime in all of the habitats sampled (e.g.. upper 1.0 m in shallow water; 0.5- to 5-m layer in deep water). This was congruent with laboratory and mesocosm indications of positive phototaxis and upward movement during the day (Barile et al. 1994). However, the clear pattern of diel vertical migration ex- pressed in mesocosms was not always observed in the field. For example, day. night, and crepuscular distribution patterns were not different in intensive sampling of the tidal inlet in 1990. Con- versely, the depth of the veligers appeared to be directly related to wind velocity; this was particularly evident in the open-water en- vironment over the RS. A similar effect of wind-induced turbu- lence on vertical distribution has been observed for bivalve ve- ligers (Raby et al. 1994). Although some conch veligers moved to the upper 0.5 m at night, as would be predicted by the laboratory and mesocosm experiments (i.e.. with negative geotaxis), a large number also dispersed to greater depths (to 30 m) in the deep- water habitat. This appears to be associated with the absence of a light cue and a relatively weak negative geotaxis. Veligers were associated with surface waters, but very few were found in the neuston during the day. This could be a simple response to veliger preference for specific light levels or avoidance of particular wavelengths. Many marine zooplankton. including echinoid plutei (Pennington and Emlet 1986), avoid potentially harmful wavelengths in the upper water column (Damker et al. 1980). Barile et al. (1994) found that queen conch larvae were higher in laboratory water columns and field mesocosms when ultraviolet wavelengths were filtered out; however, the differences with and without filtration were not statistically significant, and the exact reasons for avoiding the brightly lit surface remain unknown. Although queen conch veligers were found as deep as 100 m in the ES, densities below the pycnocline (30 m) were very low. Similar surface-layer associations have been observed for a variety % of Veligers by Depth 20 40 60 80 100 % of Veligers by Depth 20 40 60 80 100 Q. a> a - D Day (n=10) □ Crepuscular (n=2) ■ Night (n=6) a. a 6 □ Dav(n=14) D Crepuscular (n=7) ■ Night (n=2) Figure 5. Percentage of veligers collected over depth during day, crepuscular, and night periods in the tidal channel (SKI 1 in 1490 (mean ± SE). Samples with fewer than four veligers were eliminated from the analysis, (a) Dates with three depths sampled, (b) Dates with four depths sampled, including the neuston layer. 26 Stoner and Davis TABLE 4. Density of veligers collected in the tidal channel (SRI I in 1990 during multiple day, night, and crepuscular samplings at three depths. TABLE 5. Density of veligers collected in the tidal channel (SRI) in 1990 during multiple day, night, and crepuscular samplings at four depths. Density of Veligers (no./10 m'l Sampling Period Average (± SE) Minimum Maximum Day (n = 10 tows) Surface (1ml 1.19±0.52 0.15 5.69 Midwater (3 m) 0.28 ± 0.09 0.88 Near-bottom (6 m) 0.2510.10 0.92 Crepuscular (n = 2 tows) Surface (1 m ) 1.51 10.97 0.54 2.47 Midwater (3 m) 0.1810.12 0.06 0.29 Near-bottom (6 m) 0.40 1 0.24 0.16 0.64 Night (n = 6 tows) Surface (1 mf 1.2910.45 3.17 Midwater (3 m) 0.74 i 0.40 0.12 2.69 Near-bottom (6 m) 0.45 10.22 1.30 Sampling Period Density of Veligers (no./lO m 3 ) Average (±SE) Minimum Maximum All veligers were newly hatched (300-500 p.m SL). except in one night surface tow. a J On July 29. three late-stage veligers were found in one two (0.33 veligers/ 10 m 3 ). of invertebrate larvae (Young and Chia 1987) including other mol- luscs (Tremblay and Sinclair 1992). Many marine zooplankton (Dagg 1985, Daro 1988). including bivalve larvae (Raby et al. 1994). are known to migrate vertically according to distribution of food. However, concentrations of chlorophyll in ES were essen- tially uniform to 250 m during June 1994, and there was no indi- cation of a deep chlorophyll maximum (unpubl. data). Of course, it is possible that there were depth-related differences in the phy- toplankton community or in their nutritional value to queen conch larvae. Investigations on the relationship between natural phy- toplankton foods and the growth and nutrition of conch larvae have just begun (Davis, in press, Davis et al. 1996). Day (n = 14 tows) Neuston (0-0.5 m) 0.1710.44 6.3 Surface (1 m) a 1.5110.37 0.06 5.48 Midwater (3 m) 0.4210.13 1.83 Near-bottom (6 m) 0.4610.12 1.32 Crepuscular (n = 7 tows) Neuston (0-0.5 in) 0.1310.06 0.40 Surface (1 m) 0.61 10.16 0.11 1.4 Midwater (3 m) 0.1610.06 0.35 Near-bottom (6 m) 0.13 10.04 0.22 Night (n = 2 tows) Neuston (0-0.5 m) 0.25 10.25 0.50 Surface (1 m) 0.8510.60 0.25 1.44 Midwater (3 m) Near-bottom (6 m) All veligers were newly-hatched (300-500 p.m SL), except in one daytime surface tow*. J On August 27, two midsize veligers (0.13 veligers/ 10 m') and one late- stage veliger (0.10 veligers/ 10 m 1 ) were found in one tow. One of the most important adaptive advantages conferred on queen conch larvae living in the upper mixed layer is the high growth rate associated with high temperature. In the laboratory, growth and survivorship in conch larvae are maximum at 28-29°C (Davis 1994). Mixed layer temperatures in the ES during the sum- mer spawning season are typically 28-30°C. with temperature de- creasing rapidly to approximately 25°C at 100-m depth; therefore, larvae remaining above the thermocline probably have the highest growth rates and the shortest time exposed to pelagic predators. 100 80 60 0-0.5 0.5-5 E 5-10 JZ *■> Q. 0) Q 10-30 30-60 60-100 % of Veligers by Depth 40 20 20 40 60 80 100 24/57 0/62 1 5/5 i I 9/2 0/3 2/2 H 72/27 -H 0/5 ~ D Day 1 1/0 | Night Figure 6. Percentage of veligers at each of six depths during the day and night at the deep-water site in ES in June 1994 (mean ± SE, n : at each depth). Values next to the error bars represent the actual number of veligers collected in Tow 1 and Tow 2. Vertical Distribution of Queen Conch Larvae 27 Density of Veligers (no. 10 / m ) a. & 12 3 0-05 - 0.5 - 5 - ^■1 10-30-1 j I Day -^ 5-10 Density of Veligers (no. 10 / m ) 1 2 3 Figure 7. Mean density and size composition of veligers collected at the deep-water site in ES in June 1994 (n = 2 tows per sampling date). Newly hatched veligers were 300-5(10 urn SL, midsize veligers were 500-900 urn SL, and late-stage veligers were >900 um SL. Ontogenetic shifts in vertical position and migration are com- monly observed in marine invertebrate larvae (Forward and Cost- low 1974. Chia et al. 1984, Power 1989), but no obvious age- related changes in vertical distribution were detected in queen conch larvae. Similarly, larvae of sea scallops, Placopecten ma- gellanicus, also demonstrate no ontogenetic shift in vertical dis- tribution (Tremblay and Sinclair 1990). Larval supply can be an important variable in determining the distribution, abundance, and year-class strength of queen conch (Stoner et al. 1994, Stoner and Davis 1997). Because this study shows that estimates for veliger abundance are strongly influenced by vertical movements, we provide the following guidelines for routine surveys. (1) Sampling can be made with minimal sampling gear because conch larvae are near the surface (0.5-1 m) during the daytime in relatively calm weather. However, because they disperse to the greatest range of depth at night and sink to greater depths during rough weather, we recommend that sampling be restricted to daylight hours when wind is <6 m/sec (10-12 knots) and seas are <1 m. Where depth permits, we recommend making oblique tows to 5-m depth. (2) Sampling in tidal areas is particu- larly problematic. Our studies in the tidal inlet north of LSI have shown that veliger abundance is strongly influenced by tidal period (Stoner and Davis 1997) as well as by time of day. Vertical tur- bulence in a tidal channel is also dependent on current velocity: therefore, the vertical position of zooplankton in a tidal field and their associated transport potential will depend on sampling time and the locomotory capacities of the organisms (Smith and Stoner 1993). Preliminary analyses for sampling strategy will be critical in tide-influenced areas. (3) Net mesh size should be chosen ac- cording to the purpose of sampling. A 202-p.m mesh must be used to collect newly hatched queen conch veligers, whereas a mesh size of 333 u.m is appropriate for the collection of late-stage ve- 20 Temperature (°C) 30 t 75 o Q 150 36 l — Salinity (ppt) 38 _j 23 Sigma-t 26 Figure 8. Temperature (T), salinity (S), and density (D) profiles during plankton sampling at the deep-water site in ES in June 1994. Four CTD profiles taken over the 16-h sampling period were identical. 28 Stoner and Davis ligers. A larger mesh size allows for towing through a larger water volume without clogging the net mesh and for easier sorting. (4) The size of the net mouth will depend on the tow vessel and the density of the target species. We have found that 0.5-m-diameter nets are useful in general surveys for conch larvae, even in sites with low larval density. Larger nets (0.75 or 1.0 ml are particularly valuable when the less abundant late-stage veligers are being sampled with larger mesh. (5) Conch veligers should be preserved in a buffered (pH 7.5-8.5) 5% formalin-seawater mixture, and if the sample contains a large amount of organic material, its pH should be checked and maintained every month. The shells of early-stage conch veligers are damaged quickly in low-pH solu- tions, making positive identification to species difficult if not im- possible. Knowledge of larval transport and supply is critical to the un- derstanding of recruitment processes and sound fisheries" manage- ment in species with pelagic larvae. The development of realistic models for transport will depend on good information on the ver- tical distribution and movements of larvae relative to both physical and biological characteristics of the environment. It is clear from this study that vertical distribution in queen conch veligers is in- fluenced by the basic behavioral traits of the larvae, their responses to physical gradients, which vary over both time and space, and their swimming capabilities in the upper water column, which is frequently turbulent. ACKNOWLEDGMENTS This study was sponsored by grants from the Shearwater Foun- dation (New York) and the National Undersea Research Program, NOAA (U.S. Department of Commerce). We thank J. Chaplin, R. Gomez. L. Hambrick. C. Kelso. J. Lally. E. Martin. K. McCarthy. N. Mehta. S. O'Connell. M. Ray. V. Sandt. K. Schwarte. and E. Wishinski, who assisted with plankton collections and the sorting and identification of veligers. H. Proft assisted in making the deep- water collections and provided CTD data from the cruise. Captain M. Laudicina assisted with patient and precise maneuvering of the R/V Shadow. We are grateful to Dr. W. J. Richards for loaning the opening-closing gear. M. Ray and anonymous reviewers provided helpful criticism of the manuscript. Appeldoorn. R. S. 1994. Queen conch management and research: status. needs and priorities, pp. 301-319. In: R. S. Appeldoorn and B. Rodriguez (eds). Queen Conch Biology. Fisheries and Manculture. Fundacfon Cientifica Los Roques, Caracas. Venezuela. Barile. P. J.. A. W. Stoner & C. M. Young. 1994. Phototaxis and vertical migration of the queen conch (Strombus gigas Linne) veliger larvae. J. Exp. Mar. Biol. Ecol. 183:147-162. Berg, C. J.. Jr. & D. A. Olsen. 1989. Conservation and management of queen conch (Strombus gigas) fisheries in the Caribbean, pp. 421—142. In: J. F. Caddy (ed.). Marine Invertebrate Fisheries: Their Assessment and Management. John Wiley and Sons. New York. Bollens. S. M. & B. W. Frost. 1989. Predator-induced diel vertical migra- tion in a marine planktonic copepod. /. Plankton Res. 1 1:1047-1065. Brownell, W. N. & J. M. Stevely. 1981. The biology, fisheries, and man- agement of the queen conch, Strombus gigas. Mar. Fish. Rev. 43:1-12. Chaplin. J. & V.J. Sandt. 1992. Vertical migration and distribution of queen conch veligers. Proc. Gulf Caribb. Fish. Inst. 42:158-160. Chia. F. S.. J. Buckland-Nicks & C. M. Young. 1984. Locomotion of ma- rine invertebrate larvae: a review. Can. J. Zool. 62:1205-1222. Dagg. M. J. 1985. The effects of food limitation on diel migratory behavior in marine zooplankton. Arch. Hydrobiol Beih. Ergebn. Limnol. 21: 247-255. Damker. D. M.. D. B. Dey. G. A. Heron & E. F. Prentice. 1980. Effects of UV-B radiation on near-surface zooplankton of Puget Sound. Oecolo- gia (Berlin). 44:149-158. Daro. M. H. 1988. Migratory and grazing behavior of copepods and ver- lical distribution of phytoplankton. Bull. Mar. Sci. 43:710-729. Davis. M. In press. Differential growth rate influences dispersal potential of queen conch larvae. Proc. Gulf Caribb. Fish. Inst. Davis, M. 1994. Mariculture techniques for queen conch {Strombus gigas L): egg mass to juvenile stage, pp. 231-252. In: R. S. Appeldoorn and B. Rodriguez (eds.). Queen Conch Biology. Fisheries and Mariculture. Fundacfon Cientifica Los Roques. Caracas. Venezuela. Davis. M.. C. A. Bolton & A. W. Stoner. 1993. A comparison of larval development, growth, and shell morphology in the three Caribbean Strombus species. Veliger 36:236-244. Davis. M., G. A. Hodgkins & A. W. Stoner. 1996. A mesocosm system for ecological research with marine invertebrate larvae. Mar. Ecol. Prog. Ser. 130:97-104. Enright, J. T. 1977. Diurnal vertical migration: adaptive significance and LITERATURE CITED timing. Part 1. Selective advantage: metabolic model. Limnol. Ocean- ogr. 22:856-872. Forward. R. B.. Jr. 1976. Light and diurnal vertical migration: photobe- havior and physiology of plankton, pp. 457-509. In: K. C. Smith (ed). Photochemical and Photobiological Reviews. Vol. 1. Plenum Press, New York. Forward. R. B.. Jr. 1988. Diel vertical migration: zooplankton photobiol- ogy and behavior. Oceanogr. Mar. Biol. Annu. Rev. 26:361-393. Forward, R. B.. Jr. & J. D. Costiow. Jr. 1974. The ontongeny of phototaxis by larvae of the crab Rhithropanopeus harrisii. Mar. Biol 26:27-33. Hill. A. E. 1991. Vertical migration in tidal currents. Mar. Ecol. Prog. Ser. 75:39-54. Kaartvedt. S.. D. L. Aksnes & J. K. Egge. 1987. Effect of light on the vertical distribution of Pecten maximus larvae. Mar. Ecol. Prog. Ser. 40:195-197. Mann. R.. M. C. Bemardita & M. W. Luckenbach. 1991. Swimming rate and responses of larvae of three mactrid bivalves to salinity disconti- nuities. Mar. Ecol. Prog. Ser. 68:257-269. Miller. S. E. & M. G. Hadfield. 1986. Ontogeny of phototaxis and meta- morphic competence in larvae of the nudibrach Phestilla sibogae Bergh (Gastropoda: Opisthobranchia). J. Exp. Mar. Biol. Ecol. 96:1-18. Neill. W. E. 1990. Induced vertical migration in copepods as a defense against invertebrate predation. Nature 345:524-526. Olmi, E. J., III. 1994. Vertical migration of blue crab Callinectes sapidus megalopae: implications for transport in estuaries. Mar. Ecol. Prog. Ser. 113:39-54. Pennington. J. T. & R. B. Emlet. 1986. Ontogenetic and diel vertical mi- gration of a planktonic echinoid larvae Dendraster excentricus (Eschscholtz): occurrence, causes, and probable consequences. J. Exp. Mar. Biol. Ecol. 104:69-95. Posada. J. & R. S. Appeldoorn. 1994. Preliminary observations on the distribution of Strombus larvae in the eastern Caribbean, pp. 191-200 In: R. S. Appeldoorn and B. Rodriguez (eds.). Queen Conch Biology. Fisheries and Mariculture. Fundacfon Cientifica Los Roques. Caracas. Venezuela. Power, J. H. 1989. Sink or swim: growth dynamics and zooplankton hy- dromechanics. Am. Nat. 133:706-721. Raby. D.. Y. Lagadeuc. J. J. Dodson & M. Mingelbier. 1994. Relationship between feeding and vertical distribution of bivalve larvae in stratified and mixed waters. Mar. Ecol. Prog. Ser. 103:275-284. Vertical Distribution of Queen Conch Larvae 29 Randall. J.K.I 964. Contributions to the biology of the queen conch Strom- bus gigas. Bull. Mar. Sci. GulJ Caribb. 14:246-295. Smith, N. P. & A. W. Stoner. 1993. Computer simulation of larval trans- port through tidal chanels: role of vertical migration. £.v/. Coast. Shelf Sci. 37:43-58. Stoner. A. W. & M. Davis. 1997. Abundance and distribution of queen conch veligers {Strombus gigas Linne) in the central Bahamas. I. Hori- zontal patterns in relation to reproductive and nursery grounds. J. Shell- fish Res. 16:7-18. Stoner. A. W.. M. D. Hanisak. N. P. Smith & R. A. Armstrong. 1994. Large-scale distribution of queen conch nursery habitats: implications for stock enhancement, pp. 169-189. In: R. S. Appeldoorn and B. Rodriguez (eds.). Queen Conch Biology. Fisheries and Mariculture. Fundacion Cientiffca Los Roques. Caracas. Venezuela. Stoner. A. W.. P. A. Pitts & R. A. Armstrong. 1996. The interaction of physical and biological factors in the large-scale distribution of juvenile queen conch in seagrass meadows. Bull. Mar. Sci. 58:217-233. Stoner, A. W. & M. Ray. 1996. Queen conch (Strombus gigas) in fished and unfished locations of the Bahamas: effects of a marine fishery reserve on adults, juveniles, and larval production. Fish. Bull. U.S. 94:551-565. Stoner, A. W. & V.J. Sandt. 1992. Population structure, seasonal move- ments and feeding of queen conch. Strombus gigas, in deep-water habitats of the Bahamas. Bull. Mar. Sci. 51:287-300. Stoner, A. W. & K. C. Schwarte. 1994. Queen conch, Strombus gigas, reproductive stocks in the central Bahamas: distribution and probable sources. Fish. Bull. U.S. 92:171-179. Stoner, A. W.. V. J. Sandt & I. F. Boidron-Metairon. 1992. Seasonality in reproductive activity and larval abundance of queen conch Strombus gigas. Fish. Bull. U.S. 90:161-170. Sulkin. S. D. 1984. Behavioral basis of depth regulation in the larvae of brachyuran crabs. Mar. Ecol. Prog. Ser. 15:181-205. Swift. M. C. & R. B. Forward. Jr. 1988. Absolute light intensity vs. rale of relative change in light intensity: the role of light in the vertical mi- gration of Chaoborus punclipcnius larvae. Bull. Mar. Sci. 43:604-619. Tremblay, M. J. & M. Sinclair. 1990. Diel vertical migration of sea scallop larvae Placopecten majellanicus in a shallow embayment. Mar. Ecol. Prog. Ser. 67:19-25. Tremblay, M. J. & M. M. Sinclair. 1992. Planktonic sea scallop larvae Placopecten majellanicus in the Georges Bank region: broadscale dis- tribution in relation to physical oceanography. Can. J. Fish. Acpiat. Sci. 49:1597-1615. Young, C. M. & F. S. Chia. 1987. Abundance and distribution of pelagic larvae as influenced by predation, behavior, hydrographic factors, pp. 385-462. In: A. C. Giese. J. S. Pearse, and V. B. Pearse (eds.). Repro- duction of Marine Invertebrates. Blackwell Scientific Publications. Palo Alto, California. Journal oj Shellfish Research. Vol. 16, No. 1, 31-37. 1997. LABORATORY SPAWNING AND JUVENILE REARING OF THE MARINE GASTROPOD: SPOTTED BABYLON, BABYLONIA AREOLATA LINK 1807 (NEOGASTROPODA: BUCCINIDAE), IN THAILAND NILNAJ CHAITANAWISUTI 1 AND ANUTR KRITSANAPUNTU' Fishery Resources Research Unit Aquatic Resources Research Institute Chulalongkorn University Phya Thai Road Bangkok, Thailand 10330 ABSTRACT Laboratory spawning and juvenile rearing of the marine gastropod. Babylonia areolata L., is described. Adults were collected from wild stock in the inner eastern Gulf of Thailand. B. areolata laid individual, moderately transparent, vasiform egg capsules and firmly attached them to the sand substratum by a long narrow stalk. The egg capsules average 21.43 mm long and 9.57 mm wide. Each adult female spawned 47 capsules with 85 1 eggs per capsule. Average eggs were 425.70 u.m in diameter. Embryonic development occurred inside the egg capsules for 7 days. On Day 8. larvae (720.40 p.m) hatched through the apical opening of the egg capsule into water. Metamorphosis from the free swimming planktotrophic larvae to benthic juveniles took 18 days. Newly settled juveniles were 1.52 and 1.16 mm in shell length and width, respectively. The monthly average growth increments of juveniles were 4.26 per month in length and 2.28 g/mo in weight. Average survival rate was 94.08%. AT;)' WORDS: Babylonia areolata, growth, survival, spawning, rearing, egg capsules INTRODUCTION The spotted babylon. Babylonia areolata, commonly known as Hoy wan in Thailand, is a carnivorous marine benthic gastropod of the order Neogastropoda. family Buccinidae (Habe 1965). The shell is thick with a high, pointed apex, and the body whorl is patterned, with round brownish patches on the white shell back- ground. The species is. abundant all year and inhabits the muddy sand bottom, usually less than 10-20 m in depth. Literature on B. areolata is limited (Munprasit and Wudthisin 1988, Singhagrai- wan et al. 1989). Other studies on the biology of Babylonia spi- rala. Babylonia zeylonica, and Babylonia lutos have been recently done in Hong Kong and India (Thirumavalavan 1987, Morton 1990. Shanmugaraj et al. 1994. Ayyakkannu 1994. Raghunathan et al. 1994, Patterson et al. 1994. Raghunathan and Ayyakkannu 1995). Babylonia is an important marine resource harvested from the natural local beds. The fishery has developed as a by-product of sand crab (Portunus pelagecus) harvests. Babylonia harvest has recently declined in traditional areas, particularly in the larger size classes. In response to the decreased production was a resulting increase in both demand and price. The price of babylonia of 5.0-6.5 cm shell length is about 7.2 and 10.0 US$ per kilogram in seafood markets and restaurants, respectively. In recent years, babylonia mariculture has been proposed as a means of increasing supply. Continuous yearly exploitation of babylonia may result in the depletion of local stocks. The declining stocks and interest in aquaculture prompted this study on spawning, larval development, and juvenile rearing of B. areolata. Thailand (Fig. 2). by local fishermen. These broodstocks were then transported to Sichang Marine Science Research and Training Sta- tion, about 40 km from the collection site. The animals were held in 2.0 x 1.0 x 0.8 m spawning tanks supplied with flow-through seawater (10 L/min). Salinity and temperature ranged from 26 to 29 ppt and 28 to 29°C. respectively. A 10-cm layer of fine sand was provided as substratum. The animals were fed twice daily with fresh meat of carangid fish, Selaroides leptolepis. The adult snails were acclimatized for 5-10 days to spawn naturally in the labora- tory. Hatching After spawning, egg capsules were collected and rinsed with 1-u.m filtered seawater. In order to remove the foulings contami- nating the surface of egg capsules, the capsules were soaked in wellwater for 30-60 sec. The capsules were then placed in plastic- baskets of 1-cm mesh size and submerged in 1.5 x 0.5 x 0.3 m hatching tanks containing 1-u.m filtered and gently aerated seawa- ter. Water was replenished daily until hatching. The egg capsules per snail and eggs per capsule were counted and measured for morphology studies. Egg capsules containing eggs and embryos in different stages of development were sampled daily, preserved in 5% neutral formalin, and examined microscopically to evaluate intracapsular development. Larval Development MATERIALS AND METHODS Broodstock Preparation One hundred adult B. areolata (Fig. 1) were obtained from the littoral region of Rayong Province, in the inner eastern Gulf of After hatching, the newly hatched planktonic veligers were collected with a 200-u.m nylon mesh sieve and rinsed with 1-u.m filtered, ambient seawater three times. These veligers were trans- ferred to 1.5 x 0.5 x 0.3 m rearing tanks containing 1-u.m filtered, ambient, continuously aerated seawater. The initial stocking den- sity was 10,000 larvae per liter. Larvae were fed twice daily with 31 32 Chaitanawisuti and Kritsanapuntu Figure 1. Breeders' B. areulata (5.7 cm shell length! used in this stud). 20 x 10^ cells mL" 1 of mixed unicellular microalgae consisting of a 1:1 ratio of Chlorella spp. and Tetraselmis spp. Water was changed every 2 days, and the rearing tank was cleaned with 3 ppt chlorine concentration for 10 min and rinsed with well water Figure 2. Fishing ground of Rayong province, in the eastern Gulf of Thailand. two to three times. The development of larvae was monitored during the first hour; larvae were then sampled at 1.2. 3, 4. 8, 12. 14. 16. and 24 h after hatching and at 24-h intervals during the following 2 wk. Growth of at least 20 larvae per sample was per- iodically examined under microscope over the 4 wk. The larvae were also sampled and counted every 3 days to calculate survival rate. Juvenile Rearing Larvae set on the bottom of the larval rearing tanks. The settled juveniles were transferred into 1.5 x 0.5 x 0.5 m rearing tanks. The tanks were supplied with flow-through seawater (10 L/minl and gently aerated. At an average shell length of 16.50 x 2.58 mm. the juveniles were transferred to duplicate 1.0 x 1.0 x 0.5 m rearing tanks containing a 10-cm layer of fine sand. The initial stocking density was 100 individuals per m 2 . After transfer, the food was changed from unicellular microalgae to chopped carangid fish (5. leptolepis). Snails were fed twice daily at 9:00 PM and 17.00 AM. Food was offered until the animals stopped eating. The length and weight were measured at monthly intervals over a 6-mo period. The absolute growth rates and their standard deviation were cal- culated from average increments in shell size and total weight per month. The number of dead individuals was recorded in each tank at monthly intervals, and an average monthly survival rate was calculated. Geometric mean regression analyses of shell dimensions (length and width) were calculated to determine morphological relationships (Wolff and Garrido 1991). TL = a + bWi. Laboratory Spawning of B. areolata 33 Figure 3. Egg capsules of B. areolata before (right) and after (left) hatching. where TL and Wi are any two shell dimensions of shell length and width (mm), and a is the intercept and b is the slope of the regres- sion line. Similarly, the length-weight relationship was determined using the logarithmically transformal allometrie equation: log Wt = log a + b log TL where W represents body weight (g) and Lx stands for any one shell dimension (cm). RESULTS Spawning Behavior Adult spawning stocks of B. areolata, with average shell length of 5.69 ± 0.3 cm (SD; n = 35), spawned naturally during March and April 1996. Spawning took place during early morning. Most egg capsules were individually attached to the sand substratum by a long, narrow stalk. Newly laid egg capsules were moderately TABLE 1. Morphology of broodstock, egg capsules, and number of eggs per capsule of B. areolata under hatcher) conditions. Broodstock Sizes Egg Capsules Spawned Capsules Containing ; Eggs Length" Width" Weight 1 Number Length" Width 1 Egg No. Egg Diameter No. (cm) (cm) (gm) (capsule) (mm) (mm) per Capsule (um) 1 6.13 3.57 40.70 56 23.30 10.00 1133 420 2 5.83 3.36 40.10 54 21.10 11.20 1041 415 3 6.05 3.36 42.10 46 20.0 11.10 933 446 4 5.94 3.46 40.50 57 21.10 10.00 940 465 5 5.46 3.33 38.80 44 22.20 9.20 807 438 6 5.81 3.35 44.10 42 20.00 8.20 493 422 7 5.52 3.24 33.20 53 22.20 9.00 733 389 8 5.11 3.94 32.50 22 23.30 10.00 1 253 428 9 5.38 2.41 38.30 37 19.59 8.00 507 394 10 5.74 3.33 30.90 56 21.60 9.00 673 440 X±SD 5.69 ±0.32 3.67 ± 0.38 38.12 ±4.42 46.70 ± 11.08 2 1 .43 ± 1 .32 9.57 ± 1.08 851.30 ±255.15 425.70 ±23.16 1 Measured from the maximum distance between the tip of the spire and the siphonal canal. ' Measured from the greatest distance to the opposing side of the body whorl. Whole weight. ' Measured from the distance between the beginning of the stalk and the tip of capsule. ' Measured from the greatest distance to the opposing side of the capsule. 34 Chaitanawisuti and Kritsanapuntu 3 X o 111 5 LU o T 3 T 4 Shell length -■- Shell width T 5 T 6 CULTURE PERIOD ( months) Figure 4. Average monthly growth in length and weight of juvenile B. areolata reared under hatchery conditions over a 6-mo period. Laboratory Spawning of B. areolata 35 transparent and vasiform in shape. The capsules were broad at the apex and narrower toward the base, and each capsule possesses a short stalk (peduncle) that is cemented to the substrate (Fiji. 3). The eggs are visible and suspended in albuminous fluid inside the capsule. Egg capsules averaged 21.43 ± 1.32 mm in width. 9.57 ± 1.08 mm in length. 1 1.40 ±0.85 mm in peduncle length, and 10.54 ± 0.52 in escape aperture length (SD; n = 20). An average female babylonia (5.69 cm long) spawned 46.7 ± 1 1 .08 egg capsules (SD; n — 12; range = 22-57). The average egg number per capsule was 851.30 ± 255.15 (SD; n = 15; range = 493-1.133). and the average egg diameter was 425.70 ± 23.16 u,m (SD; n = 20). B. areolata fecundity averaged 39.146 eggs per individual. The mor- phometric comparisons of broodstock, egg capsules, and egg num- bers of B. areolata are presented in Table I . Larval Development The larvae developed from single cell to early veliger stage inside egg capsules during the first 7 days. The veliger larvae hatched through the apical opening into the water column within 8 days after spawning. Veliger larvae were hatched at 28-30 ppt salinity and 26-28°C water temperature. The hatching rates were 89.00%. The newly hatched veligers had a transparent, thin shell and two large, lobed velums. The average shell length of veligers was 720.40 ± 1.52 u.m (SD; n = 20). After hatching, veliger larvae were positively phototactic and planktotrophic. At Day 9, the velar lobes became enlarged, with shells visible, and the larvae were about 870 \x.m long. The velar lobes degenerated, and the foot became obvious on about Day 13. At this stage, the larvae were about 1 ,450 u.m long. By Day 16, the siphonal canal, tentacles, and eyes had become visible and the velar lobes were almost disinte- grated. The larvae settled to the bottom at about 1,540 p.m. On Day 16, the presence of a foot and swimming near the bottom were the first indications that the larvae were competent to settle. Metamor- phosis was completed by Days 18-20, and the juveniles averaged 1.520 ± 1.64 u.m long and 1,160 ± 1.36 u.m wide (SD; n = 20). Larvae metamorphosed and settled in the absence of substratum. Average growth increment was 84.44 u.m in shell length per day. and survival was 2.4%. During the period of settlement, heavy mortality occurred because the newly settled juveniles continually crawled out of the water and died as a result of dessication. Ju penile Rearing The average growth and survival rates of juvenile B. areolata reared under hatchery conditions is represented in Figure 4. Av- erage monthly growth of juvenile B. areolata (both length and weight) rapidly increased over the first 4 mo and. thereafter, gradu- ally decreased. The average monthly growth increments were 4.26 mm/mo in length. 2.80 mm/mo in width, and 2.28 g/mo in weight. At the end of the experiment, juveniles had reached an average total length, width, and body weight of 42.10 ± 8.97 mm, 26.18 ± 5.90 mm. and 14.24 ± 4.13 g (SD; n = 50). respectively (Fig- ure 5). Monthly survival rate increased during the first 3 mo, and then no further mortality was observed. The average monthly survival rate was 94.08 over the 6 mo of juvenile culture (Fig. 6). The equations describing the relationships between shell length, width, and weight were as follows: TL = 3.1200 + 1.5073 Wi (r = 0.9660) log W = -7.0923 + 0.4005 log TL (r = 0.8498) Figure 5. Juvenile B. areolata of 40 mm shell length at the end of the study. 36 Chaitanawisuti and Kritsanapuntu CULTURE PERIOD (months) Figure 6. Average monthly survival of juvenile B. areolata reared under hatchery conditions over a 6-mo period. DISCUSSION B. areolata was spawned under hatchery conditions during March and April 1996. Egg capsules were individually attached to sand substratum with a long, narrow stalk. Each female spawned an average of 41 egg capsules, with 1.052 eggs per capsule. Spawning of B. areolata was similar to that of the spiral babylonia. B. spirata, but differed from that of the muricid gastropod. Chicoreus ramosus. Shanmugaraj et al. (1994) reported that B. spirata laid 24-35 transparent, vasiform egg capsules attached to the substratum. Each capsule contained about 900 eggs in a jelly- like fluid. The veliger larvae hatched and metamorphosed within 10 and 19 days after hatching, respectively. Kannapiran (1994) found that B. spirata had between 28 and 41 egg capsules. The highest fecundity was 36.900 eggs per snail per year. Bussarawit and Ruangchua (1991) and Nugranad (1992) reported that egg capsules of the muricid gastropod, C. ramosus, are laid in a com- pact mass of multiple capsules firmly attached to the substratum. The egg capsules were moderately translucent, vase shaped, tough, and creamy-white in color, and they measured about 16.0 mm in height and 3.7 mm wide. The escape aperture is placed centrally on top of the egg capsules, and the mean number of larvae per capsule was 341. Larval development of B. areolata was similar to that of B. spirata, but differed from that of C. ramosus. Shanmugaraj et al. (1994) reported that larvae of B. spirata hatched through apical openings within 10 days after spawning, and they completely metamorphosed into juveniles 1.9 mm long within 19 days. In C. ramosus, hatching occurred 25-28 days after spawning: the newly hatched larvae were about 580 u,m and larvae 1.4 mm long meta- morphosed within 3 wk. Survival ranged from 1.75 to 99.5% after hatching (Nugranad 1992, Bussarawit and Ruangchua 1991, Nugranad et al. 1994). Heavy postset mortality of B. areolata took place because the newly settled juveniles crawled out of the water, desiccated, and died. Similar phenomena were observed in B. spi- rata (Shanmugaraj et al. 1994) and busyconine whelk, Busycon carica, (Kraeuter et al. 1989, Castagna and Kraeuter 1994). Average monthly growth rate of juvenile B. areolata was 4.06 mm/mo in length and 0.97 g/mo in weight. The growth rate of B. areolata was greater than that of B. spirata or C. ramosus. Raghunathan et al. (1994) reported that the average growth of B. spirata showed a gradual increase from 2.95 to 3.00 to 3.55 to 3.86 cm in shell length and 6.4 to 7.8 to 11.10 to 14.10 g in total weight over a 10-mo period. Patterson et al. (1995) reported av- erage growth rates of 1.22 mm and 0.05 g/day for B. spirata fed oyster and crab. In contrast, juveniles of C. ramosus showed av- erage growth increments of 2.60, 9.26. 4.27. and 1.01 mm/mo in shell length at 2, 5, 8. and 12 mo. respectively (Nugranad et al. 1994). Kraeuter et al. ( 1989) reported that average growth rate of knobbed whelk. B. caricsa, was 14.40 mm/y for the first 10 y of life in laboratory conditions. This study showed that B. areolata has characteristics that may make it a potentially valuable aquaculture species. It exhibited a fast growth, market size being reached in 10 mo, using relatively simple hatchery techniques. Additional studies to manipulate go- nadal development and culture technology of this species are nec- essary. These should include methods to improve growth and sur- vival of larvae and juveniles in both nursery and growout systems and to evaluate the costs of scaling up the culture conditions to commercial levels. ACKNOWLEDGMENTS This research was a part of "Research on Cultivation Tech- niques of the Areola Babylon (Babylonia areolata) for Commer- cial Purposes." We thank the National Research Council of Thai- land (NRCT). who provided funds for this research in fiscal year 1995 We are especially grateful to Professor Dr. Piamsak Mensveta. Director of Aquatic Research Research Institute (ARRI), Chulalongkorn University, for his encouragement and Laboratory Spawning of B. areolata 37 suggestions. We thank Dr. Porchum Aranyaganon for providing facilities and research assistance and Dr. J. K. Patterson Edward who provided access to the literature. Last, we thank Associate Dr. Somkiat Piyatiratitivorakul for statistical analyses and Dr. Maria Zteresa Viana for review and suggestions that improved the manu- script. LITERATURE CITED Ayyakkannu. K. 1994. Fishery status of Babylonia spirata at Porto Novo, southeast coast of India. Phuket. Mar. Biol. Cent. Spec. Publ. 13:53-56. Bussarawit, N. & T. Ruangchua. 1991. The production and morphology of egg capsules and veliger larvae of Chicoreus ramosa. Phuket Mar. Biol. Cent. Spec. Publ. 9:70-74. Castagna. M. & J. N. Kraeuter. 1994. Age. growth rate, sexual dimorphism and fecundity of the knobbed whelk Busycon carica in a western mid- Atlantic lagoon system. Virginia. J. Shellfish Res. 13:581-585. Habe. T. 1965. Notes on the ivory shell genus Babylonia Schluter (Mol- lusca). Bull. Nat. Sci. Mus. Tokyo 8:115-125. Kannapiran. E. 1994. Breeding biology of Babylonia spirata (Mollusca: Neogastropoda: Buccinidae). Masters Thesis. Annamalai University. Parangipettai. India. 28 pp. Kraeuter. J. N.. M. Castagna & R. Bisker. 1989. Growth rate estimates for Busycon carica in Virginia. J. Shellfish Res. 8:219-225. Morton. B. 1990. The physiology and feeding behaviour of two marine scavenging gastropods in Hong Kong: the subtidal Babylonia lutosa (Larmarck) and the intertidal Nassarius festivus (Powys). J. Moll. Stud. 56:275-288. Munprasit, R. & P. Wudthisin. 1988. Preliminary study on breeding and rearing of areolata babylonia. Babylonia areolata. Technical Report No. 8/1988. Eastern Marine Fisheries Development Center. Depart- ment of Fisheries, Bangkok. Thailand. 14 pp. Nugranad. J. 1992. Experimental rearing of Chicoreus ramosus larvae at the Prachuap Khiri Khan Hatchery. Phuket. Mar. Biol. Cent. Spec. Publ. 10:65-71, Nugranad. J.. T. Poomtong & K. Promchinda. 1994. Mass culture of Chicoreus ramosus (Gastropoda: Muricidae). Phuket. Mar. Biol. Cent. Spec. Publ. 13:67-70.' Patterson, J. K., C. Raghunathan & K. Ayyakkannu. 1995. Food prefer- ence, consumption and feeding behaviour of the scavenging gastropod. Babylonia spirata (Neogastropoda: Buccinidea). Indian J. Mar. Sci. 24:104-106. Patterson. J., T. Shanmugaraj & K. Ayyakkannu. 1994. Salinity tolerance of Babylonia spirata (Neogastropoda: Buccinidea). Phuket. Mar. Biol. Cent. Spec. Publ. 13:185-187. Raghunathan. C. & K. Ayyakkannu. 1995. Chemoreception in the buccinid gastropods. Babylonia spirata and Babylonia zexlonica (Neogas- tropoda: Buccinidea). Phuket. Mar. Biol. Cent. Spec. Publ. 15:199- 204 Raghunathan. C. J. K. Patterson & K. Ayyakkannu. 1994. Long-term study on food consumption and growth rate of Babylonia spirata (Neo- gastropoda: Buccinidea). Phuket. Mar. Biol. Cent. Spec. Publ. 13:207- 210. Shanmugaraj. T.. A. Murugan & K. Ayyakkannu. 1994. Laboratory spawn- ing and larval development of Babylonia spirata (Neogastropoda: Buc- cinidea). Phuket. Mar. Biol. Cent. Spec. Publ. 13:95-97. Singhagraiwan, T., S. Singhagraiwan & M. Sasaki. 1989. Effect of irradi- ated seawater with ultraviolet rays on inducing to spawn of the areolata babylonia. Babylonia areolata. Technical Report No. 12/1989. Eastern Marine Fisheries Development Center. Department of Fisheries. 14 pp. Thirumavalavan, R. 1987. Studies on Babylonia spirata Mollusca (Gas- tropoda: Buccinidae) from Porto Novo waters. M. Phil. Thesis, Anna- malai University. Parangipettai. India. Wolff. M., & J. Garrido. 1991. Comparative study of growth and survival of two colour morphs of the Chilean scallop Argopecten purpuratus (Lamarck 1819) in suspended culture. J. Shellfish Res. 10:47-53. Journal of Shellfish Research. Vol. 16, No. 1. 39-45, 1997. SEASONAL STUDIES OF FILTRATION RATE AND ABSORPTION EFFICIENCY IN THE SCALLOP CHLAMYS FARRERI SHIHUAN KUANG, JIANGUANG FANG, HUILING SUN, AND FENG LI Yellow Sea Fisheries Research Institute Qingdao 266071, China ABSTRACT Seasonal studies of filtration rate, retention efficiency, and absorption efficiency in the native scallop Chlamys farreri, a major component of shellfisheries and aquaculture species in northern China since the 1970s, were carried out four times between September 1993 and May 1995 in Sungo Bay. Shandong. China. This is the first time that the feeding physiology of C. farreri has been studied. The experiments were carried out semi-in situ using a running seawater system into which natural seawater was pumped directly from the nearshore off-bottom of the experimental site and no other food was added. The variation of particle organic matter, total particle matter, and chlorophyll a (chl a) in the natural seawater of the experimental site was 1.09-4.40 mg/L, 3.79-17.66 mg/L, and 1.98— \. 89 u.g/L. respectively. The exponents (/>) in the allometric equation (FR = aW 1 ') of filtration rate as a function of dry tissue weight in different seasons varied in a narrower range and averaged 0.43, whereas the elevations (a) varied in a wider range between 1.33 and 4.35, and the order of elevations from the highest to the lowest was September. May. April, and November. This indicated that the seasonal patterns in the filtration rate of this scallop were correlated with seawater temperature. Measurement of the absorption efficiency showed no differences among individuals of different sizes, but there were differences among different seasons. November (63.1%) and April (60.7%) had higher mean absorption efficiencies than did September (44.6%). It seemed that the higher the chl a content of the seston and the more favorable the environmental condition, the higher the scallop's absorption efficiency. KEY WORDS: Chlamys farreri, filtration, absorption, retention, semi-in silu INTRODUCTION As one of the most important shellfish species in northern China, the annual production of the native scallop, Chlamys farreri Jones & Preston, amounted to 400,000 metric tons in 1994. The initiation of intensive aquaculture of this species in recent years has led to negative effects on the scallops such as depressed growth rates and increased mortality, mainly caused by over- crowding, which ultimately reduced the food available to the scal- lops. One way to resolve this problem is to estimate the carrying capacity for scallop culture in the culture area and adjust the cul- turing density accordingly. In order to model carrying capacity, it is necessary to determine the feeding physiology and the energy budget of the scallops in the aquaculture ecosystem. Furthermore, feeding physiological indices such as filtration rate and absorption efficiency are fundamental parameters in the bioenergetic studies of suspension feeding bivalves (Riisgard 1991 ). There are studies on the reproduction, spat collection, growth, and aquaculture of this scallop species (Zhang et al. 1956, Wang et al. 1987. Zhang 1992). but little is known about its feeding ecology and physiol- ogy- The ecological and physiological aspects of feeding in bivalve molluscs have long been studied. However, many previous experi- ments were carried out in the laboratory at various artificially controlled conditions. These laboratory studies on growth and en- ergetics in suspension feeding bivalves achieved peak growth rates that are usually less than the maximal growth rates observed in nature (Kiorboe et al. 1981. Jorgensen 1990). This is because suspension feeding bivalves are exposed to a food supply that consists of a complex mixture of organic and inorganic particles and that fluctuates unpredictably both in quantity and quality in the field, and it has been known that both food quality and quantity are important factors mediating the feeding behavior and physiology of suspension feeders (Bayne and Hawkins 1992). Besides food condition, many other physical and chemical factors such as tem- perature, salinity, and water flow can also affect the feeding of bivalves. However, it is very difficult to mimic the natural food regimen as well as the above-mentioned physical and chemical conditions in the laboratory. Growth rates comparable with those measured in the field may only be obtained in laboratory experi- ments that are carried out under simulated natural conditions. Moreover, there have always been technical and conceptual difficulties in predicting an organism's response in the natural environment from data measured in the laboratory (MacDonald and Ward 1994). This largely restricted the application of labora- tory data. Despite these differences, few workers have conducted studies on the feeding strategies of bivalves using natural seston and under simulated natural conditions, and even fewer have evaluated feeding activity when food and other conditions such as temperature varies temporally and seasonally (MacDonald and Ward 1994). However, information regarding changes in feeding behavior under natural conditions is critical to the analysis of bivalve energetics (Bayne et al. 1988). Only those data measured in situ or in the laboratory under simulated natural conditions can be used to readily predict the organism's feeding pattern in nature. Sungo Bay is located at 37°01'-09'N and 122°24'-35'E (Fig. 1 ) and is one of the most intensive aquaculture areas in northern China. The main culture species in this bay are scallops. C. farreri, and kelp. Laminaria japonica. The quantity of cultivated scallops was 2 billion individuals in 1994. At such high densities, scallops may deplete the seston from the water column. It is thus critical to measure the filtration rate and absorption efficiency of scallops in this bay. In this study, the filtration rate and absorption efficiency of the scallop C. farreri were measured in a novel semi-in situ running seawater system using natural seawater (Fig. 2). This is the first time that the filtration rate and absorption efficiency of C. farreri were measured. Further, the experiments were carried out four times at the same site in spring and autumn, in order to understand the scallops' feeding regimens under different water conditions. The goal of this research was to understand the cultivated scallops' 39 40 KUANG ET AL. 12242 122-44 12246 12248 12250 12252 12254 12256 1225 E Figure 1. Sungo Bay and the experimental site for measuring I ill ration rate and absorption efficiency of the scallop C. farreri. feeding strategies in natural conditions by determining the feeding parameters of this species in a simulated natural environment. MATERIALS AND METHODS Seawater and Experimental System Seawater was pumped 50 m from the nearshore off-bottom in Sungo Bay into two large precipitating tanks (500 nr ). After pre- cipitating for about 24 h (because of the stirring of the pump, it is better to precipitate the seawater for a period of time to match the seston concentration in nature), seawater was siphoned via a rub- ber hose with a diameter of 2 cm into a fiberglass tank (0.4 m ) in the hatchery room. The inflowing end of the hose was masked by a screen with pore diameter of 1 mm to prevent large-sized par- ticles from entering the experimental flumes. The fiberglass tank acted as the header tank for the running seawater system in order to maintain a constant water level and flow. Seawater was then siphoned from the fiberglass tank via a 1.5-cm-diameter plastic tube with a multipipe joint at the outflowing end. by which sea- water flowed at a controlled and constant flow rate into the ex- perimental flume tanks (Fig. 2). The flume tanks were made of perspex and measured 20 x 20 x 60 cm in length, width, and height. The inflowing hole was 1 cm above the bottom of one end. and the outflowing hole was I cm under the surface of the other end (Fig. 2). No extra unicellular algae or seston was added to the precipitation tank Sungo Bay pump /^outflowing hole\ experimental flume tank Figure 2. Running seawater system for the measurement of filtration rate and absorption efficiency of C. farreri. experimental seawater. Table 1 lists the experimental seawater conditions at the different seasons. Scallops Experimental scallops were collected from a scallop aquacul- ture site in Sungo Bay, city of Rongcheng, province of Shandong. These randomly selected scallops of different sizes were then car- ried to the hatchery room of Aitou Farm for the experiments. After the epibiota were cleaned from the shell surface, scallops were grouped according to their size and placed in the running seawater system to acclimate for at least 2 days. Scallops collected in Sep- tember and November 1993 were divided into six groups, as well as one to two control groups (Table 2). The six experimental groups were marked as S2, M2. B2. S4, M4, and B4; the capital letters S, M, and B represented small-, middle-, and big-sized scallops, and the numbers 2 and 4 indicated the number of indi- viduals in each experimental flume tank. M4. for example, means that there were four middle-sized scallops in this group. The group of scallops measured in May 1994 were divided into three groups marked as S4, M4, and B4. Scallops measured in April 1995 were divided into seven size groups with four individuals per flume tank (Table 2). There were no scallops in the control tanks, but seawater ran through these tanks as in the experimental groups. After the completion of each set of experiments, the shell height and dry tissue weight (60°C for 24 h) of each scallop were recorded. The physical measurements of experimental scallops in different sea- sons are listed in Table 2. Experiment Procedures After 2-5 days of acclimation, individual experimental flume tanks were cleaned of feces and other sediments. A 1,000-mL sample of outflowing seawater from the control and each experi- mental flume tanks as well as from the fiberglass tank was then TABLE 1. Experimental seawater conditions in different seasons. Dates Temperature Range ( C) pH Salinity (mL/minl POM (mg/L) TPM (mg/L) POM/TPM PCC (Mg/L) Sept. 9-18, 1993 Nov. 15-24. 1993 May 23-28. 1994 Apr. 21-29. 1995 24.5 ± 0.8 8.12 31.2 520-630 4.40 10.61 0.415 1.98 10.1 ±0.5 8.05 32.0 405-522 4.21 9.28 0.454 4.89 17.9+ 1.1 8.12 32.1 470-700 3.87 17.66 0.219 3.52 13.7 ± 1.1 8.02 31.6 309-189 1.09 3.79 0.288 4.45 Abbreviations: F. flow rate at different experimental flumes; PCC. particulate concentration of chl a. Seasonal Studies on C. Farrfri 41 TABLE 2. Characteristics of experimental scallops in different seasons during September 1993 and April 1995. Date Group SH (cm) DTW (g) TI* tg/cm) FR (E/ht RE (%) AE | % ) S2 3.28 ±0.19 0.17 ±0.03 0.052 3.90 ±0.60 24.35 ± 8.05 43.91 S4 3.13 ±0.18 0.17 ± 0.02 0.054 1.91 ±0.14 21.70+ 10.11 49.89 Sept. M2 4.15 ±0.05 0.48 ± 0.08 0.116 5.70 ±0.13 38.00 + 18.25 52.60 1993 M4 3.95 ± 0.07 0.28 ± 0.06 0.071 2.74 ±0.1 8 36.35 ± 11.55 35.23 B2 5.55 ± 0.05 1.20 ±0.13 0.216 6.17 ±0.05 38.92+ 18.66 43.90 B4 5.64 ±0.11 1.11 ± 0.09 0.197 4.45 ± 0.93 52.93 ±21.32 42.03 S2 3.03 ±0.15 0.11 ±0.02 0.036 0.43 4.33 73.39 S4 3.45 ±0.10 0.16 ±0.06 0.046 0.47 6.21 60 26 Nov. M2 5.05 ±0.15 0.56 ± 0.08 0.102 0.68 4.98 62.11 1993 M4 5.05 ± 0.34 0.54 ±0.11 0.107 0.76 12.53 64.79 B2 6.90 ±0.10 1.37 ±0.13 0.199 2.76 17.64 58.86 B4 6.65 ± 0.60 1 .43 ± 0.26 0.215 1.83 29.26 59.01 May S4 3.58 ±0.1 5 0.17 ±0.03 0.047 2.10 ±0.37 27.60+ 1.95 1994 M4 5.13 ±0.17 0.72 ±0.11 0.140 2.90 ± 0.43 33.31 ±4.56 B4 6.28 ±0.29 1 .77 ± 0.23 0.282 4.65 ± 1.77 42.87 ± 16.00 1 3.05 ±0.19 0.16 ±0.02 0.052 1.35 ±0.77 22.51 ± 18.05 61.60 2 3.65 ± 0.06 0.27 ± 0.03 0.074 1.84 ±0.86 33.77 ±9.72 62.27 Apr. 3 4.10 ±0.26 0.41 ±0.01 0.100 2.17 ± 1.61 32.84+ 19.95 56.64 1995 4 4.48 ±0.21 0.47 ± 0.08 0.105 2.44 ±0.79 41.71 ±6.16 60.59 5 5.38 ± 0.05 1.01 ±0.10 0.188 2.58+ 1.26 47.00 ± 14.29 62.80 6 5.73 ± 0.05 1.I6±0.14 0.202 2.95 ± 1.48 46.39 ± 20.85 60.26 7 6.63 ± 0.06 1.44 ±0.03 0.217 3.31 ± 1.36 47.82 ± 14.63 60.72 Abbreviations: SH. shell height: DTW, dry tissue weight; TI. tissue index: FR. filtration rate: RE. retention efficiency; AE. absorption efficiency. * TI = DTW/SH x 100. collected every 6 h — 500 mL for analysis of chlorophyll a (chl a) concentration, and another 500 mL for analysis of particle organic- matter (POM) and total particle matter (TPM). Each experiment lasted for 24 h; thus, we sampled five times for each experiment. Water flow rates were calculated from the volume of collected seawater over a known time period. Preliminary analysis of filtra- tion rates determined by the differences of chl a and POM in the inflowing and outflowing seawater. respectively, indicated that filtration rate determined by chl a (difference) was more stable and reliable than that determined by POM (Kuang et al. 1996a). There- fore, differences of chl a concentrations between the inflowing and outflowing seawater were only used to calculate the filtration rate in the following experiments, whereas differences of POM/TPM between the seawater and feces were used to calculate the absorp- tion efficiency. After precipitating for 24 h, the seston concentra- tions in the fiberglass tank and in the outflowing water of the control group were comparable, and the seston concentration in the outflowing seawater of the control tanks was used to estimate the seston supply in the experimental groups. Feces were collected gently with a pipette at the end of each experiment (24-h period). Because of electrical power failure, the feces collected in May 1994 cannot be processed: thus, the related absorption efficiencies were omitted. Analysis of Chl a The concentrations of chl a were determined according to Par- sons et al. (1992). The 500-mL seawater samples were filtered through acetate fiber membranes (pore size. 0.45 p.m) under 0.5- MPa atmospheric pressure vacuum. As the seawater was being filtered, a few drops of a suspension of magnesium carbonate (MgCO,) were added to prevent acidity on the filter. After filtra- tion, filters were placed in 10-mL centrifuge tubes and chl a was extracted in 90% acetone overnight in a cold (4°C). dark place. The contents of each tube were centrifuged for 10 min at 4.000 revolutions/min after extraction. The extinctions of the superna- tants were then measured immediately at 750-. 664-. 647-. and 630-nm wavelength using a 2-cm path length model 7230 spec- trophotometer. The chl a was calculated as follows: chl a (|ULg/L) = (11.85£ 664 - 1.54 E,, 0.08 £ 630 - 10.23 E 750 ) x v/(V x Q, -«,4- ^647' ^630' 3n " W50 ' 647-. and 630-nm wavelength, respectively, v is the volume (mL) of acetone: Vis the volume (L) of the seawater sample: / is the path length (cm) of the spectrophotometer. Analysis of POM and TPM The 500-mL seawater samples were filtered onto preashed and preweighted (W lt ) 40-mm GF/C-grade filters (nominal pore size, 1.2 Luri) under a 0.03-MPa vacuum. Approximately 10 mL of distilled water was added to the last few milliliters of the sample to remove residual salts. Filters with seston were then frozen at -20°C in a desiccator in the dark until they were transported back to the laboratory for processing. The TPM concentrations were determined after drying filters at 60°C for 48 h and weighting (W 6() ) (TPM = W H) - VV„). These filters were then ashed in a muffle furnace at 450°C for 5 h and weighted (,W 450 ), POM con- tents were calculated as follows: POM = W h0 - W 450 . A Sartorius Research R 200D electronic semimicrobalance was used for weighting. The total weight and organic portion of feces were determined by the same method as seawater samples above. Calculations Filtration rate (FR). also termed as clearance rate (CR). was measured as the volume of water cleared of chl a per unit time; it KUANG ET AL. was calculated as follows: FR = F x (C, - C 2 )/C l , where F is the flow rate of water through the experimental flumes (L/h). C, is the chl a concentration in the inflowing seawater, and C, is the chl a concentration in the outflowing seawater after it has been pro- cessed by the scallops. FR values presented in this study were the mean values per individual. Retention efficiency (RE) was calculated as follows: RE = ( 1 - C 2 /C,) x 100. RE values presented in this study were the group values. Absorption efficiency (AE) for ingested food is estimated by measuring the dry weight (DW, = TPM) and ash-free dry weight (AFDW. = POM) of seston in inflowing seawater and fecal pellets from scallops. The POM/TPM of food and feces was then used to calculate the absorption efficiency: AE = ( 1 - E/F)H 1 - E)x 100 (Conover 1966a), where F = POM/TPM of inflowing seawater and E = POM/TPM of scallop feces. Data analyses and statistical tests were performed with Microsoft EXCEL 5.0. RESULTS Summary of Experimental Conditions Table 1 reported the environment conditions of experimental seawater. and Table 2 reported the shell height, body weight, and tissue index of scallops. Data in Table 1 represent the natural seawater conditions in the experimental site and were comparable to the results of a synchronous biological and hydrochemical sur- vey of the bay (Kuang et al. 1996b). Tissue indices of scallops varied with body size and season. This was related to the scallop's gametogenesis and reproductive cycle (Zhang et al. 1956). C. far- reri has two reproductive peaks per year; the minor one occurs during middle May and early June, and the major one occurs in October (Zhang et al. 1956). In our study, the scallops' highest tissue indices occurred in September, followed by May. April, and November (Fig. 3). This is because scallops in September have not yet spawned, so they have the highest tissue indices, whereas scallops in November were spent, so they have the lowest tissue indices. Filtration Rate Filtration rates of scallops are listed in Table 2. Data analysis indicated that the filtration rate of scallops increased with size 3.5 3 r 2.5 A M 4) 5 u 3 15 n H Q 0.5 2.5 3.5 4.5 5.5 Shell Height (cm) 6.5 7.5 ♦ Sept. ONov. A May X April Figure 3. Dry tissue weight (W0 of the scallop C. farreri as a function of shell height (H) in different seasons. Regression equations for dif- ferent seasons are as follows: a (Sept.): W = 0.013V" 1011 , R 2 = 0.974; b (May): W = 0.0077e 087H , R 2 = 0.998; c (Apr.): W = 0.027? ° MH , R 2 = 0.5 1.5 2.5 W(g) ♦ Sep-93 ONov-93 AMay-94 XApr-May-95 0.962; d (Nov.): W = 0.0164r" R- = 0.984. Figure 4. Filtration rate {FR I of the scallop C. farreri as a function of dry tissue weight (W) in different seasons. Regression equations for different seasons are as follows: a (Sept.): FR = 4.35W 043 , R 2 = 0.97; b (May): FR = 3.601**", R 2 = 0.94; c (Apr.): FR = 2.85W - 56 , R 2 = 0.93; d (Nov.): FR = 1.33W 1 "" 1 , R 2 = 0.95. (weight), and the relationship between these two variables can be represented as FR = aW 1 '. Our results showed that the exponent b of filtration rate as a function of dry tissue weight in different seasons ranged from 0.33 to 0.61 with an average of 0.43, whereas the elevations a varied more widely between 1.33 and 4.35 (Fig. 4). Figure 4 showed that filtration rate varied among seasons and decreased in the order of September. May. April, and November; this was consistent with the order of seawater temperature. Analy- sis showed that filtration rate was correlated with the seawater temperature — the higher the seawater temperature, the higher the scallop's filtration rate (Fig. 5). Results in September and Novem- ber 1993 showed that the filtration rate of scallops was also related to their densities — the higher the scallop densities, the lower the individual filtration rate (Table 2). Although seston quantity and quality (POM ratio in TPM) also varied among seasons, they did not seem to be the dominant factors that directly influenced the filtration rate of scallops in natural seawater conditions (Table 1). Retention Efficiency Similar to filtration rate, retention efficiency of the different groups increased in relation to body size (Table 2) and dry tissue weight (Fig. 6). The highest retention efficiency occurred in Sep- tember, followed by April and May, whereas the lowest value occurred in November (Fig. 6). Unlike the filtration rate, both the exponent (range, 0.18-0.70; average, 0.41) and the elevation (range. 21.47-52.98; average, 39.37) of retention efficiency varied widely as a function of dry tissue weight among the different seasons. Absorption Efficiency Unlike the filtration rate and retention efficiency, the absorp- tion efficiency of the scallops had no relationship to body size/ weight, and the absorption efficiency values of scallops of differ- ent body sizes varied in a narrow range in the same month (Table 2; also see SD of absorption efficiency below). However, scallops in different months had different mean absorption efficiency val- ues. Although scallop absorption efficiencies in November (aver- age ± SD. 63.07 ± 5.52%) and April (average ± SD. 60.70 ± 2.02%) were not significantly different (analysis of variance [ANOVA], F = 1.12, df = 12, p = 0.31). the absorption effi- Seasonal Studies on C. Farreri 43 35 r 3.0 2.5 2 2 1.0 0.5 00 FR = 1655T- 0.9525 R 2 = 0.9278. n= 10. p = 0.012 10 15 20 25 30 Temperature (°C) Figure 5. Filtration rate {FR) of C. farreri as a function of seawater temperature (T). Data are not based on the same batch of scallops, but on different scallops of the same dry tissue weight (0.16-0.17 g/indi- viduall in different seasons. ciencies of these two months were higher than that of September (average ± SD, 44.59 ± 6.12%) (ANOVA. between September and November. F = 30.14, df = 1 1. p = 0.00026; between September and April. F = 43.56, df = 12, p = 0.00004). It seemed that the higher the chl a concentration (seston quality) and the more suit- able the environmental condition (temperature), the higher the scallop's absorption efficiency. In this case, the food quality (higher chl a concentration: Table 1 ) and environmental conditions (e.g., temperature) in November and April resulted in higher ab- sorption efficiency for this species, compared with September, which had high seawater temperatures and low chl a concentra- tions (Tables 1 and 2). DISCUSSION Filtration rate is a dynamic index that reflects the physiological feeding state of suspension feeding animals. It is reported that temperature (e.g.. Newell et al. 1977, Buxton et al. 1981), salinity (e.g.. Loosanoff 1953, Navarro 1988), variation in food quantity and quality (e.g., Higgins 1980a, Higgins 1980b, Riisgard 1988, Riisgard 1991, MacDonald and Ward 1994). and other environ- mental parameters (e.g., Walne 1972, Shumway et al. 1983, Bricelj W(g) ♦ Sep-93 ONov-93 AMay-94 XApr-May-95 Figure 6. Retention efficiency {RE) of the scallop C. farreri as a func- tion of dry tissue weight (W) in different seasons. Regression equations for different seasons are as follows, a (Sept.): RE = 52.98W*' I \ R 2 = 0.88: b (Apr.): RE = 45.69VT ", R 2 = 0.87; c (May): RE = 37.35W ,8 , R 2 = 0.95; d (Nov.): RE = 21.46W 070 , R 2 = 0.98. and Malouf 1984. Bayne et al. 1987. Cranford and Grant 1990) can all influence the filtration rate of bivalves. It seemed that the seasonal variation in the filtration rate of the scallop C. farreri was mainly decided by seawater temperature. Our results show that the changes of the scallop's filtration rate in different seasons were consistent with the changes of natural seawater temperature. This trend was comparable to that of the sea scallop Placopecten ma- gellanicus, measured by MacDonald and Thompson ( 1986) under ambient temperatures and natural seston levels. However. Thomp- son ( 1984) reported that there were no seasonal patterns in clear- ance rate for the mussel Mytihis edidis. The filtration rate of C. farreri reported in this article is lower than that of the Pacific oyster Crassostrea gigas, measured at the same site and at the same time (Kuang et al. 1996d). Although food quantity (TPM content) and quality (POM percent in TPM. or chl a content) may affect the filtration rate of C. farreri in single-factor experiments, they did not seem to be major factors influencing the scallop's seasonal filtration rate patterns in the natural environment. For example, the highest TPM content occurred in May and the highest percent POM occurred in November, but the highest filtration rate was in September. Water flow in this experiment also did not affect filtration rate, even though our experimental seawater flows were faster than those used in previous studies (e.g., MacDonald and Ward 1994). C. farreri was most often found under conditions of fast flow rates in nature (Zhang et al. 1962). In this study, for example, scallops often assembled themselves near the inflowing hole in the experimental flume tank, where the water flow was relatively faster. Other studies have reported that the filtration rate of C. farreri remained constant in the water flow range of 300-600 mL/min (Kuang et al. 1996c). Hildreth (1976) reported that the filtration rate of blue mussel was unresponsive to changes in flow rate of 2-41 L/h. Therefore, in C. farreri, the seasonal filtration rate pattern was related more to temperature than to food or other aspects of the seawater. An allometric relationship of group retention efficiency as a function of dry tissue weight was observed in this study. The retention efficiency of bivalves was reported, by many authors, to vary with particle sizes (e.g.. Shumway et al. 1983, Cranford and Grant 1990, MacDonald and Ward 1994). It is possible that the seasonal patterns of retention efficiency in C. farreri may be af- fected by several factors. The measurement of absorption efficiency in this article indi- cated that there were no differences among individuals of different sizes, but that there were differences between seasons. It has long been believed that food quality, rather than temperature, food quantity, or other variables, is the major factor affecting the ab- sorption efficiency (Conover 1966b, Vahl 1980, Bayne et al. 1988, Navarro et al. 1991, Navarro et al. 1992, Iglesias et al. 1992, Navarro and Iglesias 1993. Cranford 1995). Cranford (1995) re- ported that diet quality, expressed as POM. POC (particle organic- carbon), or PN (particle nitrogen) content per unit of the particulate matter, explained between 74 and 84% of the variance in sea scallop absorption efficiency measurements. Our results suggest that food quality (chl a concentration) may explain the variation of absorption efficiency between different months. The seasonal ab- sorption efficiency patterns may also be related to several other factors. Despite the higher POM proportion in September, the low chl a concentration and less "comfortable" environmental condi- tion (e.g., elevated seawater temperature — the most suitable tem- perature for the growth of C. farreri is 15-20°C) resulted in a lowered absorption efficiency. The absorption efficiencies mea- 44 KUANG ET AL. sured in this study were within the range of other bivalves: Powell and Stanton ( 1985) found an average absorption efficiency of 0.54 for bivalves. However, compared with that of the Pacific oyster C. gigas (Kuang et al. 1996d). the absorption efficiency of C. farreri was relatively low. The average absorption efficiency of the Pa- cific oyster in the same environmental conditions is 80%. Oysters can increase their absorption efficiency by selective ingestion and then produce a large amount of pseudofeces. However, the scallop C. farreri produced very few pseudofeces during our experiment, although they sometimes produced large amounts of pseudofeces during the artificial cultivation of broodstocking when food con- centration was very high. Similar results also have been reported for P. magellanicus, M. edulis, and Cardium edule when feeding on natural seston (MacDonald and Thompson 1986. Newell and Bayne 1980). In this study, scallops were exposed to natural seston and ambient temperatures, and the results may be more environ- mentally representative. The exponents of the allometric equation relating filtration rate to dry tissue weight varied in a narrower range than that of reten- tion efficiency (Figs. 4 and 6). However, the averaged exponents of 0.43 for the filtration rate and 0.41 for the retention efficiency were very similar. These slopes were relatively low compared with those of other bivalve species. The common exponents of 0.70 for P. magellanicus, 0.60 for Chlamys islandica (MacDonald and Thompson 1986). 0.58 for Argopecten irradians (Kirby-Smith 1972), 0.82 for Pecten irradians (Chipman and Hopkins 1954), and 0.66 for M. edulis (Mohlenberg and Riisgard 1979), and an overall range of 0.4-0.6 for bivalves (Officer et al. 1982. Winter 1978). have been reported. Many previous studies have determined filtration rates and ab- sorption efficiency of bivalves under conditions of one individual per experimental chamber (e.g.. MacDonald and Thompson 1986). Our experiments, however, have been adopted a typical density of four individuals per flume tank. Theoretically speaking, this may underestimate the results because seston may be depleted by scal- lops in high density. However, the purpose of this study is to predict the feeding regimens of C. farreri cultivated in Sungo Bay. and the densities of scallop-intensive culture in the bay were very high (50 individuals/nr). The densities adopted by this study are comparable to the scallop aquaculture density in the bay. Further- more, water flow rate in this study was higher than in previous studies, and the volume of the experimental flume tank is large enough to accommodate four scallops; in such conditions, it is not possible for the scallops to deplete the seston. Although the sea- sonal patterns of the filtration and absorption in C. farreri have been recorded, the mechanisms that scallops use to regulate their feeding physiological change and the feeding physiology of larval C. farreri have yet to be determined. ACKNOWLEDGMENT We give our most sincere thanks to Dr. Bruce A. MacDonald at the University of New Brunswick for his careful review of the manuscript and his very good suggestions about the article. This work was supported by the International Development and Re- search Center (IDRC) of Canada. National Sciences and Technol- ogy Committee of China, and Natural Science Foundation of Shan- dong Province (Q94D0121 1. LITERATURE CITED Bayne. B. L. & A. J. S. Hawkins. 1992. Ecological and physiological as- pects of herbivory in benthic suspension-feeding molluscs, pp. 265- 288. In: D. M. John. S. J. Hawkins, and J. H. Price (eds.). Plant-Animal Interactions in the Marine Bethos. Systematics Association Special Vol. No. 46. Clarendon Press. Oxford. Bayne, B. L., A. J. S. Hawkins & E. Navarro. 1987. Feeding and digestion by the mussel Mytilus eduiis L. (Bivalvia: Mollusca) in mixtures of silt and algal cell at low concentrations. J. Mar. Biol. Ecol. 111:1-22. Bayne. B. L.. A. J. S. Hawkins & E. Navarro. 1988. Feeding and digestion in suspension-feeding bivalve molluscs: the relevance of physiological compensations. Am. Zool. 28:147-159. Bricelj. V. M. & R. E. Malouf. 1984. Influence of algae and suspended sediment concentrations on the feeding physiology of the hard clam Mercenaria mercenaria. Mar. Biol. 84:155-165. Buxton. C. D., R. C. Newell & J. G. Field. 1981. Response-surface analysis of the combined effects of exposure and acclimation temperatures on filtration, oxygen consumption and scope for growth in the oyster Oslrea edulis. Mar. Ecol. Prog. Ser. 6:73-82. Chipman, W. A. & J. G. Hopkins. 1954. Water filtration by the bay scallop. Pecten irradians. as observed with use of radioactive plankton. Biol. Bull. Mar. Lab. Woods Hole 107:80-97. Conover. R. J. 1966a. Absorption of organic matter by zooplankton. Lim- nol. Oceanogr. 11:338-345. Conover, R. J. 1966b. Factors affecting the absorption of organic matter by zooplankton and the question of superfluous feeding. Limnol. Ocean- ogr. 11:346-354. Cranford. P. J. 1995. Relationships between food quantity and quality and absorption efficiency in sea scallops Placopecten magellanicus (Gme- lin). J. Exp. Mar. Biol. Ecol. 189:123-142. Cranford, P.J. & J. Grant. 1990. Particle clearance and absorption of phytoplankton and detritus by the sea scallop Placopecten magellani- cus (Gmelin). J. Exp. Mar. Biol. Ecol. 137: 105-121. Higgins. P. J. 1980a. Effects of food availability on valve movements and feeding behavior of juvenile Crassostreas virginica (Gmelin). I. Valve movements and periodic activity. J. Exp. Mar. Biol. Ecol. 45:229-244. Higgins. P. J. 1480b. Effects of food availability on valve movements and feeding behavior of juvenile Crassostreas virginica (Gmelin). II. Feed- ing rates and behavior. J. Exp. Mar. Biol. Ecol. 46:17-27. Hildreth. D. I. 1976. The influence of water flow on pumping activity in Mytilus edulus using a refined direct measurement apparatus. J. Mar. Biol. Assoc. U. K. 56:311-319. Iglesias. J. I. P.. E. Navarro, P. A. Jorna & I. Armentia 1992. Feeding, particle selection and absorption in cockles Cerastoderma edule (L.) exposed to variable conditions of food concentration and quality. J. Exp. Mar. Biol. Ecol. 162:177-198. Jorgensen. C. B. 1990. Bivalve Filter Feeding: Hydrodynamics, Bioener- getics. Physiology and Ecology. Olsen & Olsen, Fredensborg. Den- mark. Kiorboe. T., F. Mohlenberg & O. Nphr. 1981. Effect of suspended bottom material on growth and energetics in Mytilus edulis. Mar. Biol. 61: 283-288. Kirby-Smith. W. W. 1972. Growth of the bay scallop: the influence of experimental water currents. J. Exp. Mar. Biol. Ecol. 8:7-18. Kuang. S.. J. Fang. H. Sun & F. Li. 1996a. Seasonal variation of filtration rate and absorption efficiency of scallop Chlamys farreri in Sungo bay. Oceanol. Limnol. Sinica 27:194-199 (in Chinese with English ab- stract). Kuang. S., J. Fang. H. Sun & F. Li. 1996b. Seston dynamics in Sungo Bay. Mar. Fish. Res. 17(2):60-67. (in Chinese with English abstract). Kuang, S., H. Sun, F. Li & J. Fang. 1996c. Feeding and growth of scallop Seasonal Studies on C. Farreri 45 Chlamys farreri before and after spawning. Mar. Fish. Res. 17(2):80- 86. (in Chinese with English abstraet). Kuang. S., H. Sun, F. Li & J. Fang. 19%d. Comparative feeding physiol- ogy of wild and eultured populations of Pacific oyster Crassostrea gigas. Mar. Fish. Res. 17(2):87-94. (in Chinese with English abstract). Loosanoff. V. L. 1953. Behavior of oysters in water of low salinities. Proc. Natl. Shellfish. Assoc. 43:135-151. MacDonald. B. A. & R. J. Thompson. 1986. Influence of temperature and food availability on the ecological energetics of the giant scallop Pla- copecten magellanicus III. Physiological ecology, the gametogenic cycle and scope for growth. Mar. Biol. 93:37—48. MacDonald. B. A. & J. E. Ward. 1994. Variation in food quality and par- ticle selectivity in the sea scallop Placopecten magellanicus (Mollusca: Bivalvia). Mar. Ecol. Prog. Ser. 108:251-264. Mohlenberg. F. & H. U. Riisgard. 1979. Filtration rate, using a new indi- rect technique, in thirteen species of suspension-feeding bivalves. Mar. Biol. (Berl.) 54:143-148. Navarro. J. M. 1988. The effect of salinity on the physiological ecology of Choromytilus chorus (Molifia, 1782) (Bivalvia:Mytilidae). J. Exp. Mar. Biol. Ecol. 122:19-33. Navarro. E. & J. I. P. Iglesias. 1993. Infaunal filter-feeding bivalves and the physiological response to short-term fluctuations in food availabil- ity and composition, pp. 25-26. In: R. F. Dame (ed.). Bivalve Filter Feeders in Estuarine and Coastal Ecosystem Processes. NATO ASI series. Vol. G 33. Springer- Verlag. Berlin. Navarro. E.. J. I. P. Iglesias. A. P. Camacho. U. Labarta & R. Beiras. 1991. The physiological energetics of mussels {Mytilus galloprovincialis Lmk) from different cultivation rafts in the Ria de Arosa (Galicia, N. W. Spain). Aquaculture 94:197-212. Navarro. E.. J. I. P. Iglesias & M. Ortega. 1992. Natural sediment as a food source for the cockle Cerastodenna edule (L. ): effect of variable par- ticle concentration on feeding, digestion and the scope for growth. J. Exp. Mar. Biol. Ecol. 156:69-87. Newell. R. I. E. & B. L. Bayne. 1980. Seasonal changes in the physiology, reproductive condition and carbohydrate content of the cockle Cardium ( = Cerastodenna) edule (Bivalvia: Cardiidae). Mar. Biol. 56:1 1-19. Newell. C. R.. L. G. Johnson & L. H. Kofoed. 1977. Adjustment of the components of energy balance in response to temperature change in Ostrea edulis. Oceanologia (Berl.) 30:97-110. Officer. C. B.. T.J. Smayda & R. Mann. 1982. Benthic filter feeding: a natural eutrophication control. Mar. Ecol. Prog. Ser. 9:203-210. Parsons. T. R., Y. Maita & C. M. Lalli. 1992. A Manual of Chemical and Biological Methods for Seawater Analysis. Pergamon Press. Oxford. Powell. E.N. & R.J. Stanton. Jr. 1985. Estimating biomass and energy flow of molluscs in palaeocommunities. Palaeontol. (Lond.) 28:1-34. Riisgard. H. U. 1988. Feeding rates in hard clam iMercenaria mercenaria) veliger larvae as a function of algal (Isochrysis galband) concentration. J. Shellfish Res. 7:377-380. Riisgard. H. U. 1991. Filtration rate and growth in the blue mussel Mytilm edulis Linneaus. 1758: dependence on algal concentration. J. Shellfish Res. 10:29-35. Shumway. S. E.. T. M. Scott & J. M. Hick. 1983. The effects of anoxia and hydrogen sulphide on survival, activity and metabolic rate in the coot clam. Mulinia lateralis (Say). J. Exp. Mar. Biol. Ecol. 71:135-146. Thompson. R. J. 1984. The reproductive cycle and physiological ecology of the mussel Mytilus edulis in a subarctic, non-estuarine environment. Mar. Biol. 79:277-288. Vahl. O. 1980. Seasonal variations in seston and in the growth rate of the Iceland scallop. Chlamys islandica (O. F. Mullen from Balsfjord. 70°N. J. Exp. Mar. Biol. Ecol. 48:195-204. Walne. P. R. 1972. The influence of current speed, body size and water temperature on the filtration rate of five species of bivalves. J. Mar. Biol. Assoc. U.K. 52:345-374. Wang, R„ J. Gao. L. Zhang & Z. Wang. 1987. Study on the seedling catching of Chlamys farreri from natural sea areas. J. Shandong Ocean College 17:93-100 (in Chinese with English abstract). Winter. J. E. 1978. A review of the knowledge of suspension-feeding in lamellibranchiate bivalves, with special reference to artificial aquacul- ture systems. Aquaculture 13:1-33. Zhang. F. 1992. Aquaculture of Chlamys farreri in Sungo Bay. Yellow Sea and Bohai Sea 2:52-58 (in Chinese). Zhang. X.. Z. Qi & J. Li. 1956. Reproduction and growth of Chlamys farreri. Acta Zool. Sinica 8:235-253 (in Chinese with English abstract). Zhang. X.. Z. Qi & J. Li. 1962. Record of Commercial Animals in China — Marine Molluscs. Science Press. Beijing. Journal of Shellfish Research, Vol. 16. No. 1. 47-54. 1997. CLEARANCE AND INGESTION RATES OF ISOCHRYSIS GALBANA BY LARVAL AND JUVENILE BAY SCALLOPS, ARGOPECTEN IRRADIANS CONCENTRICUS (SAY) Y.T. LU AND N.J.BLAKE Department of Marine Science University of South Florida St. Petersburg, Florida 33701 ABSTRACT The clearance and ingestion rates of larval and juvenile bay scallops Argopecten irradians concentricus were investi- gated using algal suspensions of Isochrysis galbana. An inverse relationship existed between clearance rate and algal cell concentration. Mean clearance rate ranged from 0.0034 to 0.0385 mL h~' for larvae and 0.14 to 0.41 mL lr' for juveniles of 0.5 mm shell height and increased to 248—420 mL h"' for juveniles of 10 mm shell height. Clearance rate increased with shell size allometrically. with a mean exponent of 2.460 ± 0.087. Weight-specific clearance rate was independent of shell size at >20 cells (jlL"'. but slightly decreased with increasing shell size at lower cell concentrations. Larvae and juveniles showed higher ingestion rates at higher algal concentrations, and the hyperbolic relationship was described using an enzymatic kinetic equation. Maximum ingestion rate occurred at approximately 20 cells u.L~' in larvae and at >20 cells u.L"' in juveniles. As temperature increased from 10 to 30°C. larvae and juveniles became more active in feeding. Relative rates of clearance and ingestion at 10. 15. 20. and 25°C were 8.4, 22.0. 52.8. and 88.2%. respectively, of the mean rate at 30°C. A'£Y WORDS: Bay scallops. Argopecten irradians, feeding, food uptake, clearance, ingestion INTRODUCTION The growth of larval and juvenile bay scallops over several developmental stages at various Isochrysis galbana concentrations has been reported (Lu and Blake 1996). Larval and juvenile bay scallops have been shown to reach optimal growth at 10-20 cells u.L~' of /. galbana. Although growth continued to increase at higher /. galbana concentrations, the increase was statistically in- significant. A complete understanding of the interaction between growth and food supply needs the knowledge of feeding response of those stages to changes in food supplies. The ability of bivalves to maintain a positive energy balance depends primarily on food ingestion. Therefore, a great deal of work has been done on feedings of adult bivalves, including the bay scallop. Previous studies have shown that the clearance rate of the bay scallop was related to shell height following a general allometric equation (Chipman and Hopkins 1954. calculated by Winter 1978. Kirby-Smith 1970), but it was inversely related to cell concentration (Palmer 1980, Kuenstner 1988). Very limited data are available, however, on the feeding physi- ology of larval and juvenile bay scallops. The only data available on the feeding of larval bay scallops appear to be that of Gallager et al. (1989). However, extrapolation of their results is difficult because they determined ingestion and clearance at only one cell concentration each of /. galbana and Aureococcus anophageffer- ens. Feedings of several size groups of juvenile bay scallops were studied using cultured algae and natural assemblages of particular organic matter (Bricelj and Kuenstner 1989. Cahalan et al. 1989, Shumway et al. 19961. but a clear relationship between feeding rate and body size is still not available. In this study, clearance and ingestion rates of larval and juve- nile bay scallops, Argopecten irradians concentricus (Say), were determined. To investigate the effect of body size and cell con- centration on feeding activity, three larval classes and seven juve- nile classes (0.5-10 mm in shell height) were tested at six cell concentrations of /. galbana. The effect of temperature on clear- ance rate and ingestion rate was determined using juveniles ap- proximately 5 mm in shell height. MATERIALS AND METHODS Adult bay scallops were collected from Homosassa, FL. and experiments were carried out at the Department of Marine Science, University of South Florida at St. Petersburg, FL. Ripe scallops were allowed to spawn in seawater of 25-28%c salinity at 24- 26°C. Fertilized eggs were allowed to develop for 20-30 h until they developed to D-shaped larvae. Larvae were collected using a 35-p.m screen and were transferred to 500-L stock tanks. They were cultured at a density of 4-8 mL -1 and fed twice daily in the total amount of 10-30 cells u,L~' of /. galbana. which was grown in f/2 media (Guillard and Ryther 1975). At the end of the plank- tonic stage, black plastic Thalassia mimics were added to the larval tank as substrate for larvae settlement. The daily food ration for juveniles was increased gradually from 30 to 100 cells p.L _l of /. galbana. Larvae used for feeding experiments were collected from the 500-L stock tanks. They were not fed 1 2 h before the start of each experiment. Larvae were filtered onto a 35-p.m screen, washed with filtered seawater (Whatman G4, 1.2 p.m). and released into 1,000 mL of filtered seawater. Larval density was determined by counting five 2-mL samples. Aliquots of the larval stock were placed in 1-L glass beakers, and filtered seawater was added to a final volume of 500 mL. Larval density was maintained at about 3-10 mL~'. Clearance rate was determined at 25°C under six /. galbana concentrations (1-60 cells p-L -1 ) for 4-6 h. All experi- mental cultures were duplicated. A set of beakers with no larvae were used as controls to determine the changes of cell concentra- tion unrelated to larval feeding, such as cell division and cell sinking. Gentle aeration was supplied to each test beaker to keep the algae in suspension. Samples of 20 mL were drawn from each beaker with a pipette at the beginning of the experiment and every 60 min thereafter. Samples were passed through a 35-p.m screen to remove larvae. /. galbana cells were counted twice for each sample with a Coulter Counter Model FN fitted with a 100-p.m aperture tube. Clearance rate (CR) (modified from Coughlan 1969) and weight-specific clearance rate (CRJ were calculated using the fol- lowing equations: 47 48 Lu and Blake CR = [ln(C,/C,)]/(td) CR, = CR/AFDW where C, and C f are the initial and final cell concentrations; t is the time interval; d is the density of larvae; and AFDW is ash free dry weight of larvae or juveniles (Lu and Blake 1996). At any cell concentration (C). ingestion rate (IR) was obtained by the equa- tion: IR CR xC Similar experimental procedures were used to determine clear- ance and ingestion rates for juveniles (0.5-10 mm shell heightl. except that juveniles were kept at lower densities to avoid a sharp reduction of cell concentrations in the experimental beakers by feeding. Shell height of juvenile scallops was measured under a compound microscope fitted with a micrometer. Only those within ±5% of the proposed shell size were picked out for each study. Juvenile densities used were adjusted according to size, ranging from one 0.5-mm juvenile per 10 mL of medium to one 10-mm juvenile per 2.000 mL of medium. Experiments lasted 3-5 h. Juveniles of approximately 5 mm shell height were used to determine the effect of temperature on the feeding activity of ju- venile bay scallops. Experimental temperatures ranged from 10 to 30°C at 5°C intervals. Before each measurement, juveniles were kept at ±0.5°C of the experimental temperature for 48 h. Two juveniles were placed in each of the 1-L beakers containing 600 mL of experimental medium. All feeding regimens were dupli- cated. Six beakers (one for each cell concentration) with no juve- niles were set up as controls. RESULTS Measured clearance rates were plotted against /. galbana (3-6 pirn) concentration for each of the 10 size classes of larval and juvenile bay scallops, and results are shown in Figure 1. Clearance rate at each of the six standardized cell concentrations (1,5, 10, 20, 30, and 50 cells u.L _1 ) was obtained by weight averaging of the measured rates, and results are summarized in Table 1 and illus- trated in Figure 1 . In all cases, clearance rate was high at low cell concentrations but decreased as cell concentrations increased. There was a 60-90% decline in clearance rate as cell concentration increased from 1 to 50 cells u,L~'. Such declines were more ob- vious for larval and small juvenile classes. At the lowest cell concentration of 1 cell uX _1 , a slight decline in clearance rate compared with that at 5 cells p.L~' was observed in the 1.1-. 2.1-. and 3.1 -mm juvenile classes, but not in the larval classes or the rest of juvenile classes. Clearance rates increased with increasing larval and juvenile size at all cell concentrations. Allometric equations were fitted to the clearance rate-shell size and clearance rate-AFDW relation- ships, and the fitted parameters are given in Table 2. Fitted b- values for the clearance rate-shell height relationships were very close to each other and ranged from 2.312 to 2.546 (mean = 2.460). Two representative curves of the relationships are shown in Figure 2. The ^-values for the clearance rate-AFDW relationships ranged from 0.868 to 0.956, with a mean of 0.923. Weight-specific clearance rates were similar throughout the animal size range tested at >20 cells p.L~' cell concentrations (Table 1). At cell concentrations <10 cells u.L"'. however, there was a general decline in weight-specific clearance rate with in- creasing shell size. Cell ingestion rate-body size relationships were converted from the clearance rate-body size relationships. Therefore, inges- tion rate showed the same pattern as clearance rate in relation to body size. Both rates had the same /^-values at corresponding cell concentrations (Table 2). Figure 3 shows cell ingestion rate-cell concentration relation- ships of two size classes of the bay scallop. Ingestion rate in- creased rapidly with the increase of cell concentration at low con- centrations and leveled off at higher concentrations. A 10- to 20- fold increase in ingestion rate was observed over the cell concentration range of 1-50 cells u,L"' in all size classes. Inges- tion rates of each size class were fitted to an enzymatic kinetic equation: IR = 1R„,. 1X x C/(K„ + C> where IR max is the estimated maximum ingestion rate. K s is the half-saturation concentration, and C is the cell concentration. Fig- ure 4 shows the two parameters fitted, IR max and K„, in relation to juvenile shell height. The relationship of IR max versus size is ex- ponential, whereas that of K s versus size is hyperbolic. Production of pseudofeces was observed at 30 and 50 cells fj.L~' in the experiment with 2-mm juveniles. However, no pseu- dofeces were produced in an experiment done a day later using the same juveniles under the same experimental conditions. Pseudofe- ces were also observed in the experiment with 3-mm juveniles at >10 cells (jlL^ 1 . but not in the rest of the experiments with juve- niles of other size classes. The highest clearance rates were found at high temperatures and low cell concentrations (Table 3). The data in Table 3 were regressed against temperature and cell concentration, and results were plotted in Figure 5. When clearance rates were expressed as a percentage of maximum rate, a sharp increase in clearance rate with temperature was obvious at all of the six cell concentrations (Fig. 6). Mean relative clearance rates were 88.2. 52.8, 22.0, and 8.4% at temperatures of 25. 20. 15, and 10°C, respectively, with respect to that at 30°C. Two-way analysis of variance showed that clearance rates and ingestion rates were significantly affected by temperature and cell concentration (Table 4). Multiple range tests demonstrated that clearance rate and ingestion rate were significantly affected by temperature in the range of 15-25°C, whereas they were not sig- nificantly different between 10 and 15°C and between 25 and 30°C (Table 5). Ingestion rates in relation to temperature are shown in Figure 7. Hyperbolic relationships were found between ingestion rates and temperatures between 15 and 30°C. The relationships broke down at temperature below 15°C. where the determined ingestion rates were much higher than that predicted by the hyperbolic curves. DISCUSSION Results of this study show that larvae of 150 p.m shell length had the highest clearance rate among the three larval classes: 120, 150. and 180 u.m. This agrees with results of the growth studies for this species (Lu and Blake 1996), in which it was found that the maximum larval growth occurred at shell lengths of 150-170 p.m. In the feeding experiment with the 180-p.m class, all of the larvae used had developed eye-spots, a sign indicating that they were ready to settle and metamorphose. This may have led to the slightly lower clearance rates observed for this size class, because Clearance and Ingestion Rates of /. galbana 49 Larvae (L=122±3. 8 urn) Larvae(L=151.0±7.2 urn) 0.03 00 0.04 0.03 0.02 0.01 10 20 30 40 50 0.00 10 20 30 40 50 Larvae (L=183.4±3.7 urn) 0.06 A g 0.04 O O § g 0.02 u ▲ ^A A A • Aa Juveniles (H=584±34 urn) 10 20 30 40 50 60 10 20 30 40 50 60 Juveniles (H=l.l 1±0 03 mm) 2.0 A 1.5 \ A 1.0 0.5 ^1— ^ 'a * Juveniles (H=2.08±0. 1 5 mm) 10 20 30 40 50 60 10 20 30 Cell cone (cells/u,l) Cell cone (cells/ul) Figure 1. .4. i. concentricus. Clearance rate of larvae and juveniles of various size classes in relation to /. galbana concentration (cone). L, length; H. height. during metamorphosis, larvae lose their vela and are unable to feed, relying entirely on energy reserves accumulated during their planktonic stage (Yonge 1947. Sastry 1965. Bayne 1965). Clearance rates of 1.2-8.2 u.L h~' were reported for Ar- gopecten irradians irradians larvae at 50 cells u,L~' of/, galbana 3.4—7.8 (iL fT 1 determined for A. i. concentricus at the same /. galbana concentration in this study. Our values (3.4—38.5 u.L h _1 at 1-50 cells p,L _l ) for the larvae of A. i. concentricus are also comparable to that of other bivalve larvae: 4-52 p.L fT 1 for Myti- lus edulis (Sprung 1984), about 4—55 u.L h"' for Mercenaria mer- (Gallager et al, 1989). Those values are similar to the rates of cenaria (Riisgard 1988). and about 10-45 p.L h for Pati- 50 Lu and Blake Juveniles (H=3.12±0. 09 mm) Juveniles (H=5.20±0. 11 mm) 10 20 30 40 50 60 10 20 30 40 50 Juveniles (H=7.34±0.21 mm) Juveniles (H=9.78±0. 11 mm) 600 ▲ 500 ▲ ▲ ▲ 400 A m A A ^^~~*^^Jl 300 ^^*: A. A ^~~~^^ 20 30 40 Cell cone (cells/ul) Observed •- Calculated 60 Figure 1. Continued. 10 20 30 40 Cell cone (cells/uj) 50 60 Observed ••- Calculated nopecten yessoensis (MacDonald 1988) over similar /. galbana concentrations. Higher clearance rates of 8.2-106 p,L h~' were reported for M. edulis in another study (Jespersen and Olsen 1982). and slightly lower rates (2-17 p.L IT 1 ) were documented for Os- trea edulis larvae (Beiras and Camacho 1994). A summary on clearance rates of bivalve larvae can be found in Sprung ( 1984). At >20 cells p-LT 1 of /. galbana, larvae and juveniles in this study had similar weight-specific clearance rates. At lower cell concentrations, however, larvae had higher weight-specific clear- ance rates than juveniles, indicating that larvae are more efficient at obtaining food at low cell concentrations than juveniles. Such an adaptation may enable larvae to exploit more efficiently the food TABLE 1. A. i. concentricus: clearance rate of larvae and juveniles of various sizes at the six standard cell concentrations (cuL -1 ) of /. galbana concentrations. Height (mm) 1 cuL" 1 Clea 5 c.ulr 1 ranee Rate (ml. ind ' lOc.ul.- 1 2uc.uL-' per h( 30 culr' 50 cuL 1 Weight-Specific Clearance Rate (mL mg 1 [AFDVV] per h) 1 c.ulr 1 5 c.ulT 1 10 c.uIT 1 20 cuL 1 30c.(iL ' 50 cpL 1 0.122 0.0230 0.0154 0.0101 0.005 1 0.0039 0.0034 266.82 178.65 116.01 59.16 45.24 25.52 0.151 0.0364 0.0284 0.0235 0.0146 0.0093 0.0078 223.47 174.36 144.27 100.68 57.10 40.52 0.183 0.0385 0.0194 0.0183 0.0178 0.0107 0.0075 128.95 64.98 61.29 59.62 35.84 28.47 0.548 0.41 0.30 0.23 0.19 0.18 0.14 147.50 106.49 84.18 66.91 64.04 53.96 1.112 1.43 1.63 1.10 0.64 0.44 0.32 78.10 89.02 59.97 34.95 24.03 17.48 2.080 11.60 14.90 9.85 6.24 3.98 3.40 IP).47 153.46 101.45 64.27 40.99 35.02 3.100 33.50 44.50 3 1 .64 20.59 14.00 8.10 119.17 158.30 112.56 69.37 46.25 28.81 5.500 145.96 127.14 118.16 92.38 82.14 52.45 112.73 98.19 91.26 71.35 63.44 40.51 7.300 253.50 241.00 192.10 144.71 126.46 87.90 92.09 87.55 69.78 52.57 45.94 31.93 9.800 420.00 394.48 387.90 350.00 325.88 247.88 69.62 65.39 64.30 58.02 54.02 41.09 Clearance and Ingestion Rates of /. galbana 51 TABLE 2. A. i. concentricus: fitted parameters for allometric relationships between clearance rate (CR, uL h"') or ingestion rate (IR, cells h" 1 ) and shell size (H. mm) or body weight (AKDW, nig). Relationship Parameter 1 cuL ' 5 cuL" 1 10 cuL- 1 20 e.uL 1 30 cuL ' 50 CfiL 1 Mean CR = a x H h a 2.1 S3 1.889 1.483 1.051 0.775 0.554 b 2.312 2.429 2.451 2.479 2.542 2.546 2.460 r 0.986 0.950 0.996 0.976 0.945 0.928 CR = a x AFDW h a 89.824 93.814 76.302 56.565 46. 1 57 33.208 b 0.868 0.912 0.920 0.931 0.954 0.956 0.923 IR = a x H h a 2183 9445 14830 21020 23250 27700 b 2.312 2.429 2.451 2.479 2.542 2.546 2.460 IR = a x AFDW h a 89824 469068 763020 1131302 1384708 1660382 b 0.868 0.912 0.920 0.931 0.954 0.956 0.923 resources to meet their high metabolic demand during periods of low food supply. Larvae cannot afford to rely on their limited energy storage for as long as larger individuals do under unfavor- able conditions. The clearance rates of juvenile bay scallops determined in this study are comparable to the 42-96 mL mg~' (dry weight |DW1 per hour found for juveniles of A. i. irradians (calculated from Ca- halan et al. 1989. assuming 30% is wet tissue. 80% of wet tissue is water) at comparable /. galbana concentrations and temperature. These values are much higher than that determined for juvenile A. i. irradians: 1.38-10.3 mL mg" 1 (DW) per hour by Kuenstner (1988) (feeding on Thalassiosira weissflogii), and for adult bay scallops: 0.31-11.90 mL mg" 1 (DW) per hour by Palmer (1980) and 1.3-8.9 mL mg~' (DW) per hour (calculated, assuming 80% water content of tissue) by Chipman and Hopkins (1954). The higher weight-specific clearance rates determined for young juve- niles in this study are in good agreement with the higher growth rate and metabolic rate (Lu 1996) measured for these early stages. In the clearance rate-AFDW allometric relationships, the b- value is 0.923 ± 0.029 for bay scallop larvae and juveniles, which is much higher than the 0.584 (Kirby-Smith 1970) and the 0.82 (Chipman and Hopkins 1954. calculated by Winter 1978) deter- mined for adult bay scallops, indicating that the clearance rate of juveniles increases much faster with increasing body size than that 1000 001 01 1 10 Shell height (mm) Figure 2. A. i. concentricus. Clearance rate of larvae and juveniles at two cell concentrations of /. galbana in relation to shell height. of larger individuals. A similar trend was found in M. edulis, where the b-value was 1.03 for juveniles (Riisgard et al. 1980). but 0.66 (Mohlenberg and Riisgard 1979) and 0.72 (Riisgard and Mohlen- berg 1979) for larger mussels. Mussel larvae often demonstrate b-values close to 0.8 (Jespersen and Olsen 1982, Sprung 1984). In oyster larvae, b-values were also found close to 1. e.g., 0.97 in Crassostrea gigas (Gerdes 1983) and 1.02 (Beiras et al. 1990) and 0.98 (Beiras and Camacho 1994) in O. edulis. Clearance rate as a function of cell concentration determined in this study agrees well with those reported in other studies: it is high at low cell concentrations and decreases with increasing cell con- centrations. In some experimental runs, clearance rates of juvenile bay scallops showed a reduction at I cell u,L _1 of/, galbana. Such a reduction at very low particle concentrations was also present for larvae of M. edulis (Sprung 1984), M. mercenaria (Riisgard 1988). and O. edulis (Beiras and Camacho 1994). A decrease of clearance rate at very low particle concentrations may serve to reduce energy consumption for the filtration process (Lam and Frost 1976, Leh- man 1976). However. Sprung ( 1984) argued that the cilia of larvae have to move for larvae to swim, and thus, reduction in feeding cannot make a significant saving of energy. He postulated that the reduction in filtration activity was probably caused by errors re- sulting from processes such as contamination by dust, air bubble formation, and feces and mucus production of experimental ani- mals. 250 200 S 150 100 Si5 20 30 Cell cone (cells/ul) Figure 3. A. i. concentricus. Ingestion rate of 150-pm larvae and 10- mm juveniles in relation to /. galbana concentration (cone). 2 4 6 8 10 Shell height (mm) Figure 4. A. i. concentricus. Maximum ingestion rate (IR nuix ) and half- saturation concentrations (K s ) in relation to shell height. Ingestion rate shows a completely different pattern from that of clearance rate: a rapid increase in ingestion rate occurs at low cell concentrations, whereas the increase slows as cell concentration increases. Ingestion rate at high cell concentrations becomes rela- tively constant, indicating a maximum rate (the ingestion capacity) has been approached. The maximum rate is probably limited by the passage of food through the gut (Sprung 1984, Crisp et al. 1985). Winter (1978) stated that as the maximum ingestion rate was reached, the filtration rate decreased continuously in such a way that the amount of food ingested was kept constant and this pattern remained unchanged up to the food concentration at which animals began to produce pseudofeces. Such a plateau in ingestion was also observed for bay scallop larvae and juveniles in this study. The satiation points are at /. galbana concentration of 20 cells p.L~' for larvae and at >50 cells p.L~' for juveniles. The latter is in accor- dance with the >57 cells p.L~' determined for larger bay scallops (39.8-49.3 mm shell height) (Palmer 1980). It seems that larvae and small juveniles reach maximum ingestion rates at a lower cell concentration than juveniles of larger sizes. Such a trend can also be reflected in the values of K s determined for the fitted ingestion rate-cell concentration kinetic curve, where K s is an increasing function of juvenile size. The fact that larger bay scallops have higher saturation points than larvae and juveniles demonstrates that larger individuals are more capable of handling dense particle concentrations than are the early developmental stages. In addition to the ingestion by the experimental animals, the production of pseudofeces may contribute to the loss of algal cells from the experimental media. This issue is rarely addressed in the literature on bivalve larvae, probably because the production of TABLE 3. A. I. concentricus: clearance rates of 5-mm juveniles at different temperatures and cell concentrations (c.pL -1 ). Clearance rate (mL ind per h) Temperature 1 5 10 20 30 50 ( J C) c.uL- 1 CUIT 1 cjiL" 1 c.uL- 1 c.uL- 1 c.uL"' 30 1 3 1 .0 142.3 135.0 94.1 90.6 63.5 25 120.0 120.0 112.3 89.8 77.2 51.9 20 76.3 66.6 61.1 52.5 53.1 33.1 15 40.7 25.0 23.3 20.5 16.7 16.5 10 6.9 12.2 12.4 8.1 9.4 5.5 Cell cone (cells/ul) Figure 5. A. i. concentricus. Response surface plot of clearance rate (CR) of 5-mm juveniles to temperature (Temp., T) and cell concen- tration (cone, C) CR = -58.6040 + 0.9810 x C + 6.3150 x T - 0.0030 x C 2 + 0.0218 x T 2 - 0.0853 x C x T. r = 0.623. pseudofeces in larvae is difficult to detect. It is also possible that the larvae may reject panicles using their cilia (Strathmann et al. 1972) rather than producing pseudofeces. The relative importance of both mechanisms remains to be determined. The production of pseudofeces was not quantified and was variable between repli- cates in this study. The findings that pseudofeces were only ob- served in some of the tests, and more importantly, that pseudofeces were observed in one test and not in a follow-up test using the same animals under the same experimental conditions, demon- strate that the production of pseudofeces is not only just a function of the ambient cell concentration, but is also a process that may be related to the physiological condition of both the experimental animals and the algal cells. The production of pseudofeces by larvae and juveniles over the range of /. galbana concentrations ( 1-50 cells U.L.-' ) adopted in this study appears to be an exception. Cell reduction measured in the experimental media over the range of 1-50 cells mL^ 1 of/, galbana can be considered predominately as the result of ingestion, although it has been reported that the production of pseudofeces at much higher cell concentrations of another algae (0.55 x 10 6 to 1.46 x 10 6 cells mL" 1 of chrysophyte Aureococcus anorexefferns) reached 25-35% of the algal cells filtered in bay scallops (Kuenstner 1988). 100% 80% 60% 40% 20% 25 20 15 10 Temperature (°C) Figure 6. A. i. concentricus. Relative clearance rate of 5-mm juveniles vs. temperature at various /. galbana concentrations. Clearance and Ingestion Rates of /. calb i \ \ 53 TABLE 4. A. i. concentricus: analysis of variance for clearance rate and ingestion rate of larvae and juveniles. Parameter Source of Variation Sum of Squares d.f. Mean Square f-Ratio Significance Level Clearance rate (mL h ') Ingestion rate (million cells h ') Main effects Temperature 45535.40 4 11383.90 62.14 0.0000 Cell concentration 6514.22 5 1302.84 7.11 0.0006 Residual 3663.75 20 183.19 Total (corrected) 55713.5 29 Main Effects Temperature 10.27 4 2.56 12.55 0.0000 Cell concentration 10.11 5 2.02 9.88 0.0001 Residual 4.09 20 0.20 Total (corrected) 24.47 29 TABLE 5. A. i. concentricus: multiple range analysis (95% LSD) for clearance rate and ingestion rate by temperature (homogeneous groups are marked by vertically aligned Xs). Clearance Rate vs. Temperature Ingestion Rate vs Temperature Temperature Least Squares Homogeneous Least Squares Homogeneous 15°C. Therefore. 15 C C can be regarded as a critical temperature at which bay scal- lops change from a less active state to a more active state with respect to feeding activities. This is in accordance with the fact that in the natural habitat of bay scallops in central Florida, water temperature seldom drops below 15°C. For example, temperature ranges from 14 to 32.5°C in the Bayboro Harbor of Tampa Bay iLu 1996) and from 12 to 30°C at Anclote Key, FL (Barber and Blake 1983). On the other hand, a rapid increase in clearance rate at >15°C probably reflects the adaptive strategy of the Florida bay scallop to the higher temperatures it generally experiences. ACKNOWLEDGMENTS The authors thank Dr. Joseph Torres and Dr. Dan Marelli for their reviewing and their valuable comments on the manuscript. 54 Lu and Blake Barber. B. J. & N. J. Blake. 1983. Growth and reproduction of the hay scallop. Argopecten irradians (Lamarck) at its southern distributional limit. /. Exp. Mar. Biol. Ecol. 66:247-256. Bayne. B. L. 1965. Growth and delay of metamorphosis of the larvae of Mytilus edulis (L.). Ophelia 2:419-443. Bayne. B. L.. R.J. Thompson & J. Widdows. 1976. Physiology: I. pp. 121-206. In: B. L. Bayne (ed.). Marine Mussels: Their Ecology and Physiology. Cambridge University Press. Cambridge. Beiras. R. & A. P. Camacho. 1994. Influence of food concentration on the physiological energetics and growth of Oslrea edulis larvae. Mar. Biol. 120:427-435. Beiras. R . A. P. Camacho & M. Albentosa. 1990. Tasas de filtration ed larva de ostra. (Ostrea edulis L.). In: A. Landin and A. Cervifio (eds.). Proceedings of the Third National Congress on Aquaculture. Experi- mental Centre of Vilaxoan, Apd. 208. Vilagarcia. Pontevedra. Spain. Bricelj. V. M. & S. H. Kuenstner. 1989. Effects of the "brown tide" on the feeding physiology and growth of bay scallops and mussels, pp. 491- 509. In: E. M. Cosper, E. J. Carpenter and V. M. Bricelj (eds.). Novel Phxtoplankton Blooms: Causes and Impacts of Recurrent Brown Tides and Other Unusual Blooms. Lecture Notes on Coastal and Estuarine Studies. Springer- Verlag, Berlin. Cahalan. J. A.. S. E. Siddall & M. W. Luckenbach. 1989. Effects of flow velocity, food concentration and particle flux on growth rates of juve- nile bay scallops Argopecten irradians. J. Exp. Mar. Biol. Ecol. 129: 45-60. Chipman, W. A. & J. G. Hopkins. 1954. Water filtration by the bay scallop, Aequipecten irradians as observed with the use of radioactive plankton. Biol. Bull. Mar. Biol. Lab. Woods Hole 107:80-91. Coughlan. J. 1969. The estimation of filtration rate from the clearance of suspensions. Mar. Biol. 2:356-358. Crisp, D. J.. A. B. Yule & K. N. Whyte. 1985. Feeding by oyster larvae: the functional response, energy budget and a comparison with mussel lar- vae. J. Mar. Biol. Assoc. U.K. 65:759-783. Gallager. S. M. V. M. Bricelj & D. K. Stoecker. 1989. Effects of the brown tide alga on growth, feeding physiology and locomotory behav- ior of scallop larvae {Argopecten irradians). pp. 511-541. In: E. M. Cosper, E. J. Carpenter and V. M. Bricelj (eds.). Novel Phytoplankton Blooms: Causes and Impacts of Recurrent Brown Tides and Other Unusual Blooms. Lecture Notes on Coastal and Estuarine Studies. Springer- Verlag. Berlin. Gerdes. D. 1983. The Pacific oyster Crassostrea gigas. Part 1. Feeding behaviour of larvae and adults. Aquaculture 31:221-231. Guillard. R. & L. Ryther. 1975. Culture of phytoplankton for feeding marine invertebrates, pp. 29-60. In: W. L. Smith and M. H. Chanley (eds.). Culture of Marine Invertebrate Animals. Plenum Press. New York. Jespersen, H. & K. Olsen. 1982. Bioenergetics in veliger larvae of Mytilus edulis L. Ophelia 21:101-113. Kirby-Smith. W. W. 1970. Growth of the scallops, Argopecten irradians concentricus (Say) and Argopecten gibbus (Linne). as influenced by food and temperature. Ph.D. Thesis, Duke University, Durham. NC. 126 pp. Kuenstner, S. H. 1988. The effects of the "Brown Tide" alga on the feeding physiology of Argopecten irradians and Mytilus edulis. M.S. Thesis. State University of New York at Stony Brook. 84 pp. LITERATURE CITED Lam. R. K. & B. W. Frost. 1976. Model of copepod filtering response to changes in size and concentration of food. Limnol. Oceanogr. 21:490- 500. Lehman, J. T. 1976. The filter- feeder as an optimal forager, and the pre- dicted shapes of feeding curves. Limnol. Ocenanogr. 21:501-516. Lu, Y. T. 1996. Physiological energetics of larvae and juveniles of the bay scallop Argopecten. irradians concentricus (Say), Ph.D. Dissertation. University of South Florida. St. Petersburg. FL. 160 pp. Lu. Y. T. & N. J. Blake, 1996. Optimum concentrations of Isochrysis gal- bana for growth of larval and juvenile bay scallop. Argopecten irra- dians concentricus (Say). J. Shellfish Res. 15:635-644. MacDonald. B. A. 1988. Physiological energetics of Japanese scallop Pa- tinopecten vessoensis larvae. J. Exp. Mar. Biol. Ecol. 120:155-170. Mohlenberg. F. & H. U. Riisgard. 1979. Filtration rate, using a new indi- rect technique, in thirteen species of suspension-feeding bivalves. Mar. Biol. 54:143-147. Palmer. R. E. 1980. Behavioral and rhythmic aspects of filtration in the bay scallop. Argopecten irradians concentricus (Say), and the oyster, Cras- sostrea virginica (Gmelin). J. Exp. Mar. Biol. Ecol. 45:273-295. Riisgard. H. U. 1988. Feeding rates in hard clam {Mercenaria mercenaria) veliger larvae as a function of algal {Isochrysis galbana) concentration. J. Shellfish Res. 7:377-380. Riisgard. H. U. & F. Mohlenberg. 1979. An improved automatic recording apparatus for determining the filtration rate of Mytilus edulis as a function of size and algal concentration. Mar. Biol. 52:61-67. Riisgard. H. U„ A. Randlov & P. S. Kristensen. 1980. Rates of water processing, oxygen consumption and efficiency of particle retention in veligers and young post-metamorphic Mytilus edulis. Ophelia 19:37- 47. Sastry. A. N. 1961. Studies of the bay scallop, Aequipecten irradians con- centricus Say, in Alligator Harbor, Florida. M.S. thesis. Florida State University. Tallahassee, FL. 118 pp. Sastry, A. N. 1965. The development and external morphology of pelagic larval and post-larval stages of the bay scallop, Aequipecten irradians concentricus Say. reared in the laboratory. Bull. Mar. Sci. 15:417^435. Schulte. E. H. 1975. Influence of algal concentration and temperature on the filtration rate of Mytilus edulis. Mar. Biol. 30:331-341. Shumway, S. E.. T. L. Cucci. M. P. Lesser. N. Bourne & B. Bunting. 1996. Particle clearance and selection in three species of juvenile scallops. Aqualcult. Int. 5:89-99. Sprung. M. 1984. Physiological energetics of mussel larvae (Mytilus edu- lis). II. Food uptake. Mar. Ecol. Prog. Set: 17:295-305. Strathmann. R. R.. T. L. Jahn & J. R. C. Fonseca. 1972. Suspension feed- ing by marine invertebrate larvae: clearance of particles by ciliated bands of a rotifer, pluteus. and trochophore. Biol. Bull. Mar. Biol. Lab. Woods Hole 142:505-519. Theede. H. 1963. Experimnetelle unterssuchimgen iiber die filtrationslei- stung der miesmuschel Mytilus edulis L. Kiel Meereoforsch. 19:20^41 . Winter, J. E. 1978. A review on the knowledge of suspension-feeding in lamellibranchiata bivalves, with special reference to artificial aquacul- ture system. Aquaculture 13:1-33. Yonge. C. M. 1947. The pallial organs in the aspidobranch gastropoda and their evolution throughout the mollusca. Philos. Trans. R. Soc. Land. B Biol. Sci. 232:443-518. Journal of Shellfish Research. Vol. 16. No. I. 55-58. 1997. GENETIC DIVERGENCE AND LOSS OF DIVERSITY IN TWO CULTURED POPULATIONS OF THE BAY SCALLOP, ARGOPECTEN IRRADIANS (LAMARCK, 1819) SANDRA G. BLAKE, 1 NORMAN J. BLAKE, 2 MICHAEL J. OESTERLING, 1 AND JOHN E. GRAVES 1 * 'School of Marine Science Virginia Institute of Marine Science College of William and Man- Gloucester Point. Virginia 23062 'Department of Marine Science University of South Florida 140 7th Avenue South St. Petersburg. Florida 33701 ABSTRACT Researchers at the Virginia Institute of Marine Science (VIMS) have been maintaining a small-scale bay scallop (Argopecten irradians) culturing operation since the late 1960s. The cultured line was originally established with broodstock collected from the coasts of Virginia and North Carolina, but it has since been augmented with a "grab bag" of introductions from other source populations. A large bay scallop-culturing operation was reportedly founded in China in the early 1980s, with 26 individuals provided by the VIMS researchers. The degree of genetic divergence between these two populations since the founding of the Chinese operation is unknown, as are the relative amounts of genetic diversity that may have been maintained under the selective pressures of the hatchery. Samples of cultured bay scallops were obtained from culturing operations in Wachapreague. VA, in 1993 and 1995, and from the Shandong Province of China in 1993. Mitochondrial DNA (mtDNA) was isolated from individual scallops, digested with a battery of eight restriction enzymes, and analysed by restriction fragment length polymorphism analysis. Measures of haplotype diversity and divergence were calculated for the samples to reveal genetic differences between the cultured populations and to allow comparison of the levels of genetic variation maintained in the cultured populations relative to those observed in several natural populations of bay scallops. A sample of 55 Virginia cultured bay scallops was found to be monotypic, represented by a single haplotype. and three haplotypes were observed in 36 individuals sampled from China. No haplotypes were shared between the samples, indicating that significant divergence has occurred between the populations. The single haplotype from Virginia was observed in a sample of bay scallops from New England, and the least common haplotype from the Chinese sample was also found in samples from New England, North Carolina, and Crystal River. FL. Haplotype diversity and genotypic divergence values for the cultured samples indicate that mtDNA variation may be lost in the culturing process and that a bottleneck effect and/or genetic drift has affected the levels of variation in these populations differently. Assuming that the Chinese culturing operation was founded exclusively with individuals from the Virginia population, it can be concluded that the latter has lost a greater proportion of the original variation in the intervening generations of hatchery breeding. KEY WORDS: Bay scallop, aquaculture. genetics. mtDNA variation, inbreeding INTRODUCTION scallops [A. i. concentricus) collected from bays along the Eastern Shore of Virginia and from Bogue Sound, NC. In the years since The bay scallop, Argopecten irradians (Lamarck), is endemic the establishment of this original cultured line, the broodstock has to shallow estuarine habitats along the East Coast of the United been supplemented with individuals from Massachusetts (A. i. ir- States, from Massachusetts to Texas (Clarke 1965). Four subspe- radians) and Texas {A. i. amplicostatus) (M. Castagna, Virginia cies have been described based on shell morphometries (Waller Institute of Marine Science [VIMS] 1993. pers. comm.). These 1969, Petuch 1987), although both restriction fragment length additions served to contaminate the line so that the exact subspe- polymorphism (RFLP) analysis of mitochondrial (mt) DNA (Blake rifle composition of the current broodstock is unknown. This bay and Graves 1995) and allozyme studies (Marelli et al. 1997) have scallop-culturing effort has continued to the present day, coordi- indicated that individuals described as the subspecies A. i. taylorae nated by researchers from VIMS. A typical spawning protocol in- are genetically indistinguishable from A. i. concentricus. The bay volves a mass, induced spawning of 100-200 broodstock animals, to scallop has been fished commercially since the mid- 1800s. and it produce an estimated 50-150 million eggs. Several such spawns may also supports a large recreational fishery (Shum way and Castagna be performed and the resultant eggs pooled. A small commercial 1994). Because populations appear to be recruitment limited market has developed for the Virginia cultured product. (Peterson and Summerson 1992) and highly variable in size, the In 1982, 128 scallops from the VIMS Eastern Shore culturing potential for aquaculture of the species has received considerable operation were transported to laboratories in Qingdao. China, with attention. the intent of establishing a bay scallop-culturing effort in the wa- The first significant attempt to rear cultured bay scallops to ters of China's eastern bays. Twenty-six of the transported indi- market size was undertaken by Castagna and Duggan (1971 ) in the viduals survived the journey to spawn in January 1983 (Chew late 1960s. The initial stock for the study consisted of 66 adult bay 1990). By 1989. Chinese production of the "Virginia" American bay scallop exceeded 50,000 metric tons in-shell live weight * Author to whom all correspondence should be addressed. (Chew 1990). 55 56 Blake et al. The potential for loss of genetic variability due to inbreeding seems great for both the Virginia cultured line and that maintained in China. The relative degree and possible consequences of this loss are unknown. A. irradians is a functional hermaphrodite, and many of the larvae produced in the hatchery may be the result of facultative selfing. Inbreeding depression has been observed in self-fertilized larvae of the catarina scallop, Argopecten circularis, manifest as decreased larval growth and lower rates of survival (Ibarra 1995). Such inbreeding effects are thought to be a general danger for cultured species with very high fecundities, in which few individuals may produce large numbers of offspring (Newkirk 1978). The effective population size (NJ, or the number of brood- stock individuals contributing gametes to the subsequent genera- tion, may in fact be much smaller than the census number (N) of individuals used as broodstock in a hatchery (Gaffney et al. 1992). Culturing techniques in which parental individuals are mass spawned may exacerbate inbreeding problems, even when the number of resultant progeny is satisfactory. By 1993. the Virginia and Chinese broodstocks had been iso- lated for 10 y and at least 10 generations, a period that should have permitted effects of the founding event and genetic drift to become apparent. It has been shown that RFLP analysis of the mtDNA reveals considerable genetic variation in natural populations of the bay scallop and that geographically isolated populations are ge- netically distinct (Blake and Graves 1995). In this study, a com- parison of mtDNA variation in the Chinese and Virginia cultured populations, and in samples from bay scallop populations in their natural range, was undertaken to determine the genetic divergence between the Chinese and Virginia populations and the level of genetic variation maintained in the cultured populations relative to that in natural populations. MATERIALS AND METHODS A sample of 27 cultured bay scallops was provided by the VIMS laboratory at Wachapreague in March 1993. These were year-old individuals, progeny of broodstock spawned in April 1992. An additional sample of 28 individuals was obtained from the facility in March 1995. These were products of the April 1994 spawn and permitted comparison of temporally isolated samples of the Virginia cultured scallops. Fresh tissue from 36 cultured bay scallops — 18 from a northern growout site (Laizhou) and 18 from a growout site in Qingdao (Tiaonan) — was obtained from Chinese culturing facilities in Oc- tober 1993. Although these individuals had been reared at the different sites, the seed originated from the same broodstock (X. Qinzhao. Institute of Oceanology. Academia Sinica. 1993. pers. comm.). Dissection of tissues from Chinese bay scallops was per- formed by the investigator (SGB), and confirmation that all were A. irradians was made at this time. The indigenous Chinese scal- lop, Chlamys farreri, is easily distinguished from the bay scallop, and none was included among the sampled individuals. MtDNA was purified from scallop gonad, mantle, and gill tis- sue by cesium chloride density-gradient ultracentrifugation, as de- scribed in Blake and Graves ( 1995). The difficulties of transport- ing usable tissue from China to the United States made it necessary for initial preparative steps to be taken in the laboratories of the Institute of Oceanology in Qingdao, China. DNA isolation was initiated in China in early October 1993. but because equipment for ultracentrifugation was not available at this facility, the samples were maintained on ice (or orange-flavored ice pops. when ice was unavailable) after the addition of CsCl-saturated water to the tissue preparations (see Blake and Graves 1995). The samples were then transported to VIMS, where mtDNA was pu- rified by cesium chloride density-gradient ultracentrifugation. Purified mtDNA was digested with a battery of eight restriction enzymes for all individuals: Aval, Banl. Banll. BglU, B.vfEII, EcoRl. Haell. and Himill. Restriction fragments were end-labeled with the Klenow fragment of DNA polymerase I and 1:, S-labeled nucleotides, electrophoresed at 1 V/cm in 1% agarose gels over- night, and visualized by autoradiography (Sambrook et al. 1989). 15 S-labeled l-kilobase ladder DNA (BRL) provided a molecular- weight size standard. Sizes of mtDNA fragments were estimated by fitting band mi- gration distances to those of the standard by the local reciprocal method of Elder and Southern (1983) by use of the program Gel Frag Sizer (Gilbert 1989). Restriction sites were inferred from completely additive fragment patterns, and letter designations were assigned to the different patterns. Eight-letter composite haplo- types were compiled for the series of enzymes and analyzed for site changes, following Blake and Graves (1995). Statistical analyses were performed with the Restriction En- zyme Analysis Package (REAP) (McElroy et al. 1991). For each sample, haplotype and nucleotide diversities were calculated fol- lowing the methods of Nei (1987) and Nei and Miller (1990), respectively. Mean nucleotide sequence divergence between samples was calculated following Nei and Miller (1990) and was corrected for within-population polymorphism by subtracting the average of within-sample diversities. Because several of the hap- lotypes observed were rare, a Monte Carlo simulation (Roff and Bentzen 1989) was performed to estimate heterogeneity and assess the likelihood that the sampled populations shared a common gene pool. Data from natural bay scallop populations (Blake and Graves 1995) were also used in comparative analyses with the Chinese and Virginia cultured bay scallop samples, to assess changes in diversity and divergence under culturing conditions. RESULTS DNA from a total of 9 1 cultured bay scallops was analyzed with eight restriction endonucleases, revealing four distinct mtDNA haplotypes (Table 1). The 1993 and 1995 samples of cultured bay scallops from Virginia were both monotypic, charac- terized by a single haplotype. AABAAAAE, and the two were combined into a single pooled sample (VA) of 55 monotypic in- dividuals for further analysis. The haplotype diversity and mean nucleotide sequence diversity for the Virginia population were both calculated to be zero. The combined cultured Chinese sample (Q) comprised three haplotypes, none of which was identical to that observed in the Virginia sample. A Monte Carlo test for heterogeneity was performed (Roff and Bentzen 1989) on the two subpopulations of bay scallops from China (Laizhou and Tiaonan). to determine whether these shared a common gene pool (originated from a common broodstock) and could be treated in subsequent analyses as one population. One thousand Monte Carlo randomizations yielded 126 \ 2 values ex- ceeding the value from the original data, indicating that at p = 0.126. the populations are not significantly heterogeneous. The Chinese sample is hereafter discussed as a single population. For the Chinese sample, haplotype diversity was 0.55 and mean nucleotide sequence diversity was 0.33%. The two less common haplotypes in the Chinese sample, ACCAAAAA and Genetics of Cultured Bay Scallops 57 TABLE 1. A. irradians: composite haplotypes from two populations of cultured bay scallops and numbers of these baplotypes observed in live samples representing natural bay scallop populations. A Cultured "Natural" Haplotype Q VA MA NC FL RK AABAAAAE 55 5 AACAAAAA 2 4(4) 7 26 2 AABAAAEE 1 22(9) ACCAAAAA 3 10(5) Total n 36 55 26 4S 27 34 Q, Qingdao, China; VA. Wachapreague. VA; MA. New England; NC, Harker's Island. NC; FL. Crystal River. FL; RK. Rabbit Key, FL. Restric- tion enzymes used: Aval, Ban\. BanW. Bt>ll\. BstEll. EcoRl, HaeU, and HiniM. Values in parentheses are totals from the Laizhou. China, growout facility. A is the number of site changes between the haplotype and the arbitrary standard of the single Virginia haplotype. A complete list of the haplotypes from the '"natural" samples is provided in Blake and Graves (1995). AACAAAAA. differed from each other by a single site change, whereas the third and most common. AABAAAEE. differed from these by several site changes. The latter, however, differed from the Virginia haplotype (AABAAAAE) by only one site change. The corrected mean nucleotide sequence divergence between the Virginia and Chinese samples was 0.13%. Because there were no shared haplotypes between the Virginia and Chinese samples, it was not necessary to apply a rigorous test for heterogeneity to these two populations. Data from other bay scallop populations (Blake and Graves 1995) were used for com- parison with the cultured samples of this study. Included in analy- ses were samples of natural populations from Harker's Island. NC, and Rabbit Key. FL. Hatchery-reared scallops from Woods Hole. MA (New England), and Crystal River. FL (Florida Gulf), were used to approximate genotype distributions for their regions of origin. These were not used in comparisons of genetic diversity. An abbreviated list of the genotypes found in the "natural'" popu- lations, including those found also in one of the cultured samples, is presented in Table 1 . Tests for heterogeneity were performed between the cultured samples and the sample from New England (Blake and Graves 1995), with which each shared a single haplotype. In both tests (MA and VA. and MA and Q). the 1,000 randomizations produced no x 2 values higher than the observed, indicating that significant heterogeneity exists between the tested pairs. The least common haplotype from the Chinese sample (AACAAAA), represented by four individuals, was also present in the New England. North Carolina, and Crystal River samples. No other haplotypes were shared between the cultured samples and those representing natu- ral bay scallop populations. Two of the three Chinese haplotypes were unique to that sample. Corrected mean nucleotide sequence divergences between the cultured samples evaluated in this study and the natural popula- tions previously described (Blake and Graves 1995) ranged be- tween 0.04% (Q vs. MA) and 0.24% (Q vs. FL) (Table 2). Al- though still not sharing a common gene pool, the Chinese sample was found to be less divergent from the New England sample (0.04%). with which it shared a single common haplotype. than from the cultured Virginia population (0.13%). DISCUSSION The genetic aspects of hatchery rearing of bay scallops are of great interest to culturists in the United States, where the bay scallop is native, and China, where culture of this scallop is being undertaken on a very large scale. A loss of genetic variation due to drift is apparent in both the Chinese and the Virginia cultured lines, although most notably so in the latter, which has apparently be- come fixed for a single mtDNA haplotype. The Chinese population represented by the Qingdao sample also possessed a lower haplo- type diversity (0.55) than natural bay scallop populations from Rabbit Key (0.91) and North Carolina (0.69. pooled) (Blake and Graves 1995). The strategy of broadcast spawning appears to be very effective at maintaining genetic diversity for the bay scallop in nature. It can be stated with confidence that at the founding of the Virginia line, there was a greater level of genetic diversity present in the broodstock than was measured in this study. MtDNA analy- ses of a natural population from North Carolina, one of the putative sources of the Virginia line, revealed considerably higher haplo- type diversities (Blake and Graves 1995). In two samples from North Carolina, the haplotype diversity, or probability of encoun- tering different haplotypes when two individuals are sampled from a population, ranged between 0.63 and 0.74. Even if hatchery rearing has reduced this level in the Virginia cultured line, periodic introductions from other source populations should have served to reintroduce genetic variability to the population. The samples obtained from the Virginia culturing facility orig- inated from mass spawnings of 100-200 animals. The numbers of contributing parent individuals are not precisely known, but it is apparent from the current monotypic state of the population that differential reproductive success has occurred during one or more of these spawning events. A single cataclysmic loss may have occurred in which one or very few maternal individuals contrib- uted to the subsequent generation. Similarly, a series of less dra- matic losses may have occurred, to bring the population to fixation over a period of generations. The initial bottleneck of no more than 26 breeding individuals that established the Chinese culturing operation in 1983 was not sustained, because the production of bay scallops in China grew extremely rapidly. If the 26 transplanted individuals reflected all of the genetic variation present in the Virginia source population at the time, it would appear that Chinese culturing methods have been more conducive to a maintenance of that variation. This tendency for loss under the Virginia hatchery regimen may be even more pronounced, if subsequent additions have been made to the Vir- TABLE 2. A. irradians: matrix of nucleotide sequence divergences among populations, in percents, corrected for within-sample variation. VA Q MA NC FL Q 0.13 MA 0.06 0.04 NC 0.32 0.18 0.12 FL 0.31 0.24 0.18 0.14 RK 0.21 0.15 0.12 0.19 04 1 VA, Wachapreague, VA; Q. Qingdao, China; MA, New England: NC. Harker's Island. NC: FL, Crystal River, FL; RK, Rabbit Key. FL. Values in boldface represent samples from this study. 58 Blake et al. ginia broodstock since [he founding of the Chinese line. The num- ber of individuals spawned to produce the sampled population from China is not known but, based on the relative magnitude of the operation, is presumed to be higher than that used in Virginia. It may be that it is simply the scale of the operation that leads to a greater maintenance of diversity. That is, the Chinese may be maintaining multiple lines with low or no diversity, rather than one. Conversely, it may be that instead of fewer mass spawnings, progeny from many, relatively small spawning events are pooled, as recommended by Gaffney et al. (1992). to prevent loss of varia- tion by genetic drift. Genetic divergence between the two cultured populations is difficult to assess, particularly given the monotypic character of the Virginia samples. Cultured bay scallops from China were not found to share a common gene pool with the putative Virginia source population, and given the lack of any shared mtDNA hap- lotypes, this finding is not surprising. Corrected mean nucleotide sequence divergences (Table 2) indicate that the bay scallops in the Chinese sample were least divergent from those in the New En- gland sample. It is likely that at the time the scallops were sent to China, the Virginia population also contained a genetic component resembling that found in New England. The introduction of New England bay scallops into the Virginia broodstock is known to have taken place, although it has generally been assumed that the majority of the stock originated from animals that set naturally off North Carolina and Virginia's Eastern Shore. The presence in the Chinese sample of two haplotypes not found in the other sampled populations may also indicate that some rare genotypes, missed in the sampling of the natural populations, were present in the brood- stock sent to China but disappeared in the Virginia line. The lack of a sample from the putative Texas source population (from which introductions were made to the Virginia line) makes the possible presence of Texas genotypes in the Chinese sample im- possible to evaluate. It is difficult, in conclusion, to say much about mtDNA varia- tion in either of these cultured lines beyond what can be measured in the current population, because there are periods in the devel- opment of both in which the origin of broodstock or the methods of breeding are unclear. This lack of information underscores the importance of good hatchery recordkeeping, and the loss of diver- sity in both cases highlights the need for a careful breeding regi- men that maximizes the number of parental contributors in a broodstock. If one or few spawnings are used to replace the Vir- ginia broodstock for the subsequent generations, then a monotypic lineage will likely persist and problems associated with inbreeding depression may become more apparent. This may also be a prob- lem, more slow to develop but likely more catastrophic to the industry if it does occur, in the Chinese culturing operation. ACKNOWLEDGMENTS We thank the many individuals without whose assistance and advice this project would not have been possible. Dr. Fusui Zhang. Dr. Qinzhao Xue. and Mr. Chunde Wang were gracious hosts and translators during the sampling trip to China, and their help in obtaining and processing the Chinese scallops for transport to the United States was invaluable. Financial assistance for the trip was provided in part by Drs. John Milliman and Roger Mann, and Dr. Mann also served as a primary source of advice and encourage- ment throughout the duration of the study. We are also very grate- ful for the information provided by Michael Castagna of the VIMS Wachapreague Laboratory, regarding the early practices of the shellfish-rearing facility there. Virginia Institute of Marine Science Publication Number 2049. LITERATURE CITED Blake. S. G. & J. E. Graves. 1995. Mitochondrial DNA variation in the bay scallop. Argopecten irradians (Lamarck), and the calico scallop. Ar- gopeaen gibbus (Dall). J. Shellfish Res. 14:79-85. Castagna. M. & W. Duggan. 1971. Rearing the bay scallop. Aequipecten irradians. Proc. Natl. Shellfish. Assoc: 61:80-85. Chew. K. K. 1990. Global bivalve introductions. World Aquacult. 21:9-22. Clarke. A. H.. Jr. 1965. The scallop superspecies Aequipecten irradians (Lamarck). Malacologia 2:161-188. Elder, J. K. & E. M. Southern. 1983. Measurement of DNA length by gel electrophoresis II: comparison of methods for relating mobility to frag- ment length. Anal. Biochem. 128:227. Gaffney. P. M.. C. V. Davis & R. O. Hawes. 1992. Assessment of drift and selection in hatchery populations of oysters (Crossostrea virginica). Aquaculture 105:1-20. Gilbert. D. G. 1989. Gel Frag Sizer. dogStar Software. Bloomington, IN. Ibarra, A. M., P. Cruz & B. A. Romero. 1995. Effects of inbreeding on growth and survival of self-fertilized catarina scallop larvae, Ar- gopecten circularis. Aquaculture 134:37—17. Marelli. D. C, W. G. Lyons. W. S. Arnold & M. K. Krause. 1997. Sub- specific status of Argopecten irradians concentricus (Say. 1822) and of the bay scallops of Florida. The Nautilus I 10:42—14. McElroy, D„ P. Moran. E. Bermingham & I. Komfield. 1991. REAP: The Restriction Enzyme Analysis Package. University of Maine. Orono. ME. Nei, M. 1987. Molecular Evolutionary Genetics. Columbia University Press, New York. 5 1 2 pp. Nei. M. & J. C. Miller. 1990. A simple method for estimating average number of nucleotide substitutions within and between populations from restriction data. Genetics 125:873-879. Newkirk, G. F. 1978. A discussion of possible sources of inbreeding in hatchery stock and associated problems, pp. 93-100. In: J. W. Avault. Jr. (ed.). Proceedings of the Ninth Annual Meeting of the World Mari- culture Society. Louisiana State University. Baton Rouge, LA. Peterson, C. H. & H. C. Summerson. 1992. Basm-scale coherence of popu- lation dynamics of an exploited marine invertebrate, the bay scallop: implications of recruitment limitation. Mar. Ecol. Prog. Ser. 90:257- 272. Petuch, E. J. 1987. New Caribbean Molluscan Fauna. The Coastal Educa- tion and Research Foundation, Charlottesville. VA. 158 pp. Roff. D. A. & P. Bentzen. 1989. The statistical analysis of mitochondrial DNA polymorphisms: x : and the problem of small samples. Mol. Biol. Evol. 6:539-545. Sambrook. J.. E. F. Fritsch & T. Maniatis. 1989. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor. NY. Shumway, S. E. & M. Castagna. 1994. Scallop fisheries, culture and en- hancement in the United States. Memoirs Queensland Museum 36:283- 298. Waller. T. R. 1969. The evolution of the Argopecten gibbus stock (Mol- lusc: Bivalvia). with emphasis on the Tertiary and Quaternary species of eastern North America. J. Paleontol. 43(suppl. to no. 5): 1-125. Journal of Shellfish Research, Vol. 16. No. 1, 59-62, 1997. SOUTHWESTERN ATLANTIC SCALLOP (ZYGOCHLAMYS PATAGONICA) FISHERY: ASSESSMENT OF GEAR EFFICIENCY THROUGH A DEPLETION EXPERIMENT MARIO L. LASTA 1 AND OSCAR O. IRIBARNE 2 Instituto National de Investigation v Desarrollo Pesquero Victoria Ocampo N° 1. CC 175. {7600) Mar del Plata, Argentina Departamento de Biologia (FCEyN) Universidad National de Mar del Plata Funes 3250 (7600) Mar del Plata, Argentina ABSTRACT Fishing gear (demersal otter trawl) efficiency in a scallop fishery (Zygochlamys patagonica) of the southwestern Atlantic was estimated through a depletion experiment. During the autumn of 1995, a plot (1.257 km 2 ) was delimited (using satellite-derived coordinates) within a scallop bed (42°15'S. 58°33'W), and a 56-m scallop boat performed 89 tows (in a 3-day period), sweeping an estimated area of 2.796 km" (2.2 times the size of the experimental plot). During the experiment, the catch per unit of effort dropped to 52% of its initial value. Catchability (q,,) and initial biomass in the experimental bed were estimated by the use of depletion methods. Estimates of mean biomass density ranged between 0.23 (Leslie method) and 0.22 kg of scallops • nr ; (DeLury method). On the basis of the depletion experiment, and defining F = q h -f = (a/A) ■ e -f(a. area swept by a unit of fishing effort; A, total fishable area of the experimental plot; e, gear efficiency;/ effort) for the experimental bed, e could be calculated when the area swept by a unit of effort is known. The estimated efficiencies of the net using this relationship were in the range of 21-31%, which is in the upper range of values estimated for most scallop fisheries using dredges, but less than that recorded for scallop fisheries that use trawl gear. KEY WORDS: Gear efficiency, scallop. Zygochlamys paiagonica. fishery INTRODUCTION The scallop Zygochlamys patagonica is an abundant species distributed along the Magellanic province: on the Chilean coast from about 42°S to Tierra del Fuego, and on the Argentinean shelf north to the estuary of the La Plata River (35°S), at depths ranging between 60 and 1 75 m ( Waloszek 1 99 1 . Lasta and Zampatti 1 98 1 ). Maximum shell height is 79 mm, and sexes are separated (Waloszek and Waloszek 1986; but see Orensanz et al. 1991 for contradictory evidence). Sexual maturity is reached at 45 mm shell height (=2 y old), and spawning takes place during spring and late summer to early fall (Orensanz et al. 1991 ). Previous evidence suggests that stocks of this species may be very large (Waloszek 1991. Orensanz et al. 1991), although be- cause of the small size of the adductor muscle, hand processing onboard was thought to be impractical (Orensanz et al. 1991). Furthermore, potential fishing grounds are located in offshore wa- ters, which are generally beyond the working range of the small inshore fleet that operate on the tehuelche scallop (Aequipecten tehuelcha; Orensanz et al. 1991). The incorporation of onboard automatic processing techniques to scallop fisheries (i.e.. Iceland scallop. Chlamys islandica) and the incorporation of larger boats in the Argentinean fleet have generated renewed interest in develop- ing a fishery for this species. One fishing vessel landed 1,313 metric tons of meat (pers. obs.) during an experimental fishing season in 1995. However, major gaps in the knowledge of this species exist. Among these omissions are estimates of absolute abundance. To estimate abundance based on sweep area methods, it is important to evaluate gear efficiency (e, the fraction of the animals present in the path of the gear that are actually captured). The efficiency of scallop fishing gears has been assessed by several means (quadrat samples of dredged and undredged areas: Caddy 1968. Shafee 1979: underwater TV surveys: Mason et al. 1982. Giguere and Brulotte 1994; tagging experiments: Dickie 1955, Gruffydd 1972). Alternatively, gear efficiency estimation based on depletion methods can be made more general by encompassing larger areas, therefore avoiding variability in the spatial distribu- tion of scallops within the bed (Joll and Penn 1990). These meth- ods have been successfully used in the tehuelche scallop Aequipecten (Chlamys) tehuelcha dredge fishery (Iribarne et al. 1991). The primary objective of this work is to estimate gear effi- ciency of the bottom trawl net used in this developing scallop fishery. The study is based on a depletion experiment conducted in the southwestern Atlantic (42°15'S. 58°33'W; 180 miles off the Argentinean coast). MATERIALS AND METHODS The fishing gear used in this experiment (also used in the developing fishery) was a bottom otter trawl similar to the gear used in the calico scallop fishery (Argopecten gibhus). This gear had a total length of 13 m. The otter boards were conventional rectangular, steel-framed doors with timber panels. 1 m in height. 3.4 m long, and weighing 490 kg each. Doors were attached to a single tow wire (or trawl warp) by a 26-m-long bridle. The head rope and the foot rope (made of 1.9-cm-diameter rope) were 15 m long and directly attached to the doors (otter boards). There are two tickler chains (4.3 kg • m~' each) attached to the foot rope. The net was constructed of 6-mra polypropylene twine with 10-cm mesh size. The cod-end was made of 8-mm nylon twine with a 10-mm mesh size. The top and bottom panels were identical, but the bottom had attached pulley chaffing gear to protect the net. The net's path width was estimated by Garcia and Ercoli (1996), fol- lowing the procedures suggested by Tauti ( 1963). By this proce- dure, the net drag under an average velocity of 3.85 knots was Rr = 1,040 kg, the door's resistance was rx = 465 ka. and the 59 60 Lasta and Iribarne spreading force was ry = 530 kg. Using an iterative procedure, Garcia and Ercoli (1996) showed that the net mouth opening is 12.6 m {Ah = 0.S4 K = 12.6 m; Ah. net path width; \. headline length) when the net was empty and 10.6 m with a load of 1.200 kg, which was the average load found in our study (half of the average load gives an estimation of 1 1.5 m). The scallop boat (56 m long, two 1,120-horsepower engines) normally operated with two nets, towed by a cable and using a length-to-depth ratio of 3:1. The null hypothesis of no difference in the capture per haul be- tween the two nets was evaluated by use of a paired /-test (Zar 1984). A 1.257-km 2 -area (parallelogram-shaped) experimental plot was located by satellite coordinates (precision of ±40 m) by a procedure similar to that of Joll and Penn (1990). The area was selected because of high scallop density, after an initial exploration of the fishing ground. The bottom was homogeneous, composed of fine sand with a depth ranging between 90 and 105 m. The ex- periment lasted 3 days (March 5-7, 1995). During the fishing operation, we recorded the location of each trawl start and end point, towing time and speed, unsorted catch per tow, and direction of the tows. To randomize fishing hauls within the experimental plot, the starting point of each tow was chosen randomly before the commencement of the experiment. The unsorted catch (scallop and bycatch) per tow was calcu- lated by visually estimating the extent to which the cod-end was filled, based on categories of 10% before it was opened on deck. The relationship between catch weight and extent of cod-end fullness was investigated before field trials commenced, and a linear relationship between the extents of cod-end volume (or fill- ing) was assumed. The catch weight for different proportions of cod-end fullness (20. 50. 80. or 100%) was evaluated before the experiment and provided the following relationship: a = 22.98 (SE = 1.37). h = 4.28 (SE = 98.89). r = 0.95, n = 17. Then, on the basis of these data, the proportion of cod-end filling was transformed to weight on the basis of a linear relationship between cod-end percentage fullness and capture weight. A full net was estimated to contain 2,298 kg (SD = 294 kg, n = 17) of scallop + bycatch. Then, the amount of scallop catch (in kg) was estimated by randomly taking a 10-kg sample from 72 tows (80.1% of the tows during the experiment). This sample size was decided on the basis of the time and space available onboard for sampling. This sample was weighed (0.1 kg accuracy). Catch per unit of effort (CPUE) was expressed as scallop weight (in kg) captured by towing time (in hours) per net. Fishing distance was estimated on the basis of towing time and speed, calculated with the help of the General Positioning System equipment. The size of scallops captured during the experiment ranged from 28 to 80 mm shell height. The initial abundance of scallops in the experimental plots and the catchability coefficients (q) were estimated by the use of re- moval methods (Ricker 1975). These methods assume a closed population and constant q; thus, abundance on a given fishing ground declines only as a consequence of fishing. The data were fitted to both Leslie and DeLury models by the use of linear regression. The models are as follows: Leslie model: C,lf, = q-N,-q (K, + C,/2) DeLury model: In (C,//,) = In (q ■ N,) - q (E, +/,/2) where CJf, is CPUE over time period /. N, is initial popula- tion abundance, N, is abundance at the beginning of time period t. K, is cumulative catch taken before that time period, E, is total fishing effort applied before time period t, and In is natural loga- rithm. Given that abundance was expressed in terms of biomass, we assume either that growth and natural mortality was insignificant over the duration of the experiment, or that growth and natural mortality balanced each other out. This is a reasonable assumption, given the short experimental time (3 days). All evidence suggests that the other assumptions implicit in the use of these models (closed population: no emigration, immigration, mortality, or re- cruitment) were satisfied. The parameters of both models were estimated by means of linear regression. The assumptions of the two methods differ: CPUE is normally distributed in the Leslie model and log normally distributed in the DeLury model. Errors in the observation of the independent variable (Leslie, cumulative catches; DeLury. cumulative efforts) are assumed to be negligible. According to Caddy ( 1979). fishing mortality (F) in this type of fishery can be expressed as F = q ■ f = (.a/A) • e \f(a, area swept by a unit of fishing effort: A. total fishable area of the experimental plot; e, gear efficiency;/ effort). Given that we refer to fishing mortality and catchability within an experimental patch of scal- lops, rather than using the entire stock, we will name these values as F h and q h . Because a and A are known, e can be calculated using q h estimated from the depletion experiment. The value a is as- sumed to be constant, and its estimation is based on a constant door-to-door distance during the fishing operation. The variance of e was estimated by assuming that a and q are random variables (Iribarne et al. 1991). Nonparametric bootstrap techniques (Efron 1982) were used to estimate median values, standard errors, and confidence intervals (CI) for the three estimated parameters (N x , q h , and e). In each bootstrap sample, the value of the dependent variable ( v) was recomputed for each value of .v. by adding to the predicted y value a residual obtained with replacement from the set of residuals produced by the original regression (see Iribarne et al. 1991 for similar application). Then, using linear regression, a com- bination of 0V|, q h ) estimates were then produced using these recomputed y values. An estimate of e was obtained from this set of (A',. q h ) estimates using a value of a randomly selected from a normal distribution, with mean and variance fixed at their esti- mated values. Median and 95% confidence limits for the three estimates were obtained from the distribution of (A',, q h . e) that resulted from 1.000 bootstrap replications. This nonparametric procedure avoids the assumptions of normality of residuals made in conventional linear regression. RESULTS AND DISCUSSION The scallop catch from the experimental plot was 144.2 metric tons, obtained from 89 fishing operations (79% of them with two nets) in three fishing days. There was no significant difference in the catch of the two fishing nets (average difference = 1 39 kg, SE = 189 kg. n = 76) when operating simultaneously (i pjncd = 0.735. df = 75; f a05(2) = 1.993, P > 0.05). Therefore, the catch rate from both nets was used to calculate CPUE. The proportion of cod-end filling averaged 51% (SE = 13%', n = 89). Total fishing effort was 32 h and 40 min of towing time (per net). Estimated mean fishing speed was 8.0377 km ■ h _I (SE = 1.0927 km -h" 1 , n = 89). Assuming a net opening of 10.6 m, the estimated area swept by one unit of effort (one net during 1 h) was a = 0.0856 knr-h" 1 (SE = 0.0102 knr-lT 1 ). Total area swept over the whole experiment was 2.796 km 2 (2.2 times the size of the ex- Assessment of Scallop Gear Efficiency perimenta] plot). However, if the estimation were performed with half of the average eateh (600 kg), the net opening would be 1 1 .5 m, and the estimated area swept by one unit of effort is a = 0.0928 knr-fT 1 (SE = 0.0113 kirr-h -1 ). In this case, the total area swept over the whole experiment is 3.032 km 2 (2.4 times the size of the experimental plot). The assumption of a constant net open- ing is a matter of discussion, but several lines of evidences suggest that it is a reasonable assumption in our study. Although this is a value with high variability that is difficult to estimate in trawl bottom fishing gears (see Gunderson 1993), it has been shown that when bridles are strapped together in front of the doors, there is a remarkable increase in the constancy of the doorspread at any depth (see Engas and Ona 1991). Our gear should work in similar way, because bridles are short and merge into one tow wire. Thus, we believe that it is safe to assume a constant opening. Furthermore, because our experiment was restricted to an area where depth was constant, the variability produced by the warp/ depth ratio on the door opening (see Koeller 1991 ) was avoided. In any case, we believe that it will be constructive to study the be- havior of this type of nets, under different loads and depth regi- mens. During the experiment, CPUE (yield of scallops per hour) fell to 52% of its initial value (Fig. 1 ). Estimates of catchability and initial biomass (IB) obtained by the Leslie model u/,, = 0.01992, SE = 0.00333; IB = 297 tons. 95% CI-254-365 tons) and the DeLury model ( 0.05). These figures correspond to 0.19-0.28 kg of scallops ■ m . Estimated efficiency, depending on the load as- sumed, was in the range of 23-34% (Leslie: e = 29.2, 95% CI = 23.8-33.8; DeLury; e = 29.4, 95% CI = 23.6-34.1 ) when aver- age load was assumed and 21-31% (Leslie: e = 26.9, % CI = 21.9-31.1; DeLury: e = 27.1, 95% CI = 21.8-31.5) when we used one-half of the average load. The catchability coefficients estimated in our study by two methods are not statistically different. Thus, the estimated values could be taken as robust estimates. These values of gear efficiency are rather high (e, 21-31%) when compared with that estimated for the scallop dredge commonly used in the southwestern Atlantic (A. tehuelcha 15-21%; Iribarne et al. 1991 ) and lower when compared with trawl efficiency (60-64%) estimated by Joll and Perm ( 1990) for the scallop Amusium balloti, but it is in the same range of values of many scallop dredges used worldwide. These values range from 2 to 78% (i.e., Placopecten magellanicus: 5-20% Digby bay dredge, Dickie 1955; 8-78%, Giguere and Brulotte 1994: 2.1% 8-foot dredge, Jamieson 1978: 0.6-8.3%, offshore dredge; Pecten maximus: 13.4-35%. Gruffydd 1972. Mason et al. 1982, Mason et al. 1979; Pecten Jumata: 26-62%, Gwyther et al. 1986, Gwyther and McShane 1984, Butcher et al. 1981 ; Chlamys varia: 6.7-28.3%. Shafee 1979: and Chlamys opercularis: 17%, Dupouy and Latrouite 1976). However, the values are still low and S o I til D o 20 40 60 80 100 120 140 CUMULATIVE CATCH (thousands) 5 10 15 20 25 30 35 CUMULATIVE EFFORT Figure 1. (a) Leslie method: (CPUE kg ■ h"'l against cumulative catch (in kg), (b) DeLury method: natural logarithm of CPUE (kg-h -1 ) against cumulative effort (towing time in hours). Each point represents one haul. biomass estimation based on fishery surveys should take into ac- count these low gear efficiencies. ACKNOWLEDGMENTS We greatly appreciate the collaboration of the skipper Malcolm "Apple" Daniels and crew of the scalloper "Erin Bruce." The study was supported by INIDEP-Argentina. and one of us (O.I.) was supported by the Universidad Nacional de Mar del Plata. We are also grateful to two anonymous reviewers for many valuable suggestions. LITERATURE CITED Butcher. T.. J. Matthews. J. Glaister & G. Hamer. 1981. Study suggests dredges causing few problems in Jervis Bay. Aust. Fish. September:9-12. Caddy, J. F. 1968. Underwater observations on scallop [Placopecten ma- gellanicus) behavior and drag efficiency. J. Fish. Res. Bd. Can. 25: 2113-2114. Caddy. J. F. 1979. Some Considerations Underlying Definitions of Catch- ability and Fishing Effort in Shellfish Fisheries, and Their Relevance for Stock Assessment Purposes. Fish. Mar. Sen: (Canada). MS Rep.. No. 1489, 19 pp. Dickie, L. M. 1955. Fluctuations in abundance of the giant scallop. Pla- copecten magellanicus (Gmelin), in the Digby area of the Bay of Fundy. J. Fish. Res. Bd. Can. 12:797-857. Dupouy. H. & D. Latrouite. 1976. Scallop fisheries in France. Scallop Workshop. Baltimore. Ireland. May 11-16. 1976. 62 Lasta and Iribarne Efron. B. 1982. The Jackknife. the Bootstrap and Other Resampling Plans. Society for Industrial and Applied Mathematics. Philadelphia. Engas. A. & E. Ona. 1991. A method to reduce survey bottom trawl variability. International Council tor the Exploration of the Sea. CM. 1991/B: 39. 6 pp. (mimeo). Garcia. J. & R. Ercoli. 1996. Analisis dinamico-teorico aproximado del funcionamiento de una red de arrastre para vieira y estimacion de su abertura horizontal para distintos niveles de carga. Technical Report N 102 INIDEP (Argentina). 6 pp. Giguere, M. & S. Brulotte. 1994. Comparison of sampling techniques, video and dredge, in estimating sea scallop (Placopecten magellanicus, Gmelin) populations. J. Shellfish Res. 13:25-30. Gruffydd. L. L. D. 1972. Mortality of scallops on a Manx scallop bed due to fishing. J. Mar. Biol. Assoc. U.K. 52:449-455. Gunderson. D. R. 1993. Surveys of Fisheries Resources. John Wiley & Sons. Inc., New York. Gwyther. D. & P. F. McShane. 1984. Port Philip scallop prediction opti- mistic — but warning sounded for future. Aust. Fish. May:12-14. Gwyther. D.. B. Sause & D. C. Burgess. 1986. Scallop stocks low in Victoria. Aus. Fish. 45(10):14-17. Iribarne. O., M. Lasta. H. Vacas. A. Parma & M. Pascual. 1991. Assess- ment of abundance, gear efficiency and disturbance in a scallop dredge fishery: results of a depletion experiment, pp. 242-248. In: S. E. Shum- way and P. A. Sandifer (eds.). "An International Compendium of Scal- lop Biology and Culture." A tribute to James Mason. Selected papers from the '7th International Pectinid Workshop.' National Shellfisheries Association. The World Aquaculture Society. Parker Coliseum, Loui- siana State University, Baton Rouge, USA. Jamieson, G. S. 1978. Identification of offshore scallop (Placopecten ma- gellanicus) concentrations and the importance of such procedure in stock assessment and population dynamics. Pectinid Workshop. Brest, France. 7 pp. ill. (mimeo). Joll, L. M. & J. W. Penn. 1990. The application of high-resolution navi- gation systems to Leslie-DeLury depletion experiments for the mea- surement of trawl efficiency under open-sea conditions. Fish. Res. 9:41-55. Koeller, P. A. 1991. Approaches to improving groundfish survey abun- dances estimates by controlling the variability of survey gear geometry and performance. J. Northwest Atl. Fish. Sci. 11:51-58. Lasta. M. L. & E. Zampatti. 1981. Distnbucion de capturas de moluscos bivalvos de importancia comercial en el mar Argentino. Resultados de las campanas de los B/I "Walter Herwig" y "Shinkai Maru." anos 1978 y 1979, INIDEP (Argentina). Contribucion 383:128-135. Mason. J.. C. J. Chapman & J. A. M. Kinnear. 1979. Population abundance and dredge efficiency studies on the scallop, Pecten maximus (L.). Rapport Proces-Verbaux Reunions Cons. Int. I'Explor. Mer 175:91-96. Mason. J.. J. Drinkwater. T. Howell & D. Fraser. 1982. A comparison of methods of determining the distribution and density of the scallop. Pecten maximus lL). International Council for the Exploration of the Sea. CM 1982/K: 24. 5 figs, (mimeo). Orensanz, J. M., M. S. Pascual & M. Fernandez. 1991. Scallop resources from the Southwestern Atlantic (Argentina), pp. 981-999. In: S. E. Shumway (ed.). Scallops: Biology, Ecology and Aquaculture. Elsevier, Amsterdam. Ricker, W. E. 1975. Computation and interpretation of biological statistics offish populations. J. Fish. Res. Bd. Can. Bull. 191. 382 pp. Schafee, M. S. 1979. Underwater observations to estimate the density and spatial distribution of the black scallop. Chlamys varia (L.) in Lanveoc (Bay of Brest). Bull, f Office Natl. Peches Tunisie 3:143-156. Tauti, M. 1963. Fishing Physics. Koseisha Koshikaku, Tokyo. Japan. 116 pp. Waloszek. D. 1991. Chamys patagonica (King & Broderip 1832). a long "neglected" species from the shelf off the Patagonia Coast, pp. 256- 263. In: S. E. Shumway and P. A. Sandifer (eds.). "An International Compendium of Scallop Biology and Culture." A tribute to James Mason. Selected papers from the '7th International Pectinid Work- shop.' National Shelfisheries Association. The World Aquaculture So- ciety. Parker Coliseum, Louisiana State University, Baton Rouge, USA. Waloszek, D. & G. Waloszek. 1986. Ergebmsse der Forschungsreisen des FFS 'Walther Herwig' nach Sudamerika, LXV. Vorkommen, Re- produktion, Wachstum und mogliche Nutzbarkeit von Chlamys pata- gonica (King & Broderip. 1832) (Bivalvia, Pectinidae) auf dem Schelf von Argentinien. Arch. Fish. Wiss. 37:69-99. Zar, J. H. 1984. Biostatistical Analysis. Prentice-Hall Inc.. Englewood Cliffs, NJ. 620 pp. Journal of Shellfish Research. Vol. 16. No. I, 63-66. 1997. SURVIVAL OF SAUCER SCALLOPS, AMUSIUM JAPONICUM BALLOTI, AS A FUNCTION OF EXPOSURE TIME M. C. L. DREDGE Queensland Department of Primary Industries Southern Fisheries Centre Beach Road Deception Bay 4508 Queensland, Australia ABSTRACT The use of size limits for saucer scallops, Amusium japonicum balloti, presupposes that undersized scallops returned to the water survive the stress of capture and subsequent exposure to air while being sorted and graded. A tagging experiment, conducted to assess the survival of saucer scallops exposed to air for varying periods of time, showed that scallops exposed to air for periods of up to 120 min (2 h) were recaptured at the same rate as controls exposed to air for less than 2 min. Scallops exposed for periods longer than 150 min were recaptured at significantly reduced rates. These results suggest that saucer scallops can survive exposure to air for up to 2 h without suffering significant mortality and that the current use of size limits is justified in the context of maximizing value per recruit. KEY WORDS: Amusium, scallop, survival, exposure INTRODUCTION Saucer scallops. Amusium japonicum balloti, support a trawl fishery off of the Queensland (Australia) eastern coast from which annual landings average approximately 1.200 tons of adductor meat. The Queensland saucer scallop stock appears to be exploited heavily (Dredge 1988). although a stringent management regime appears to have reduced the risk of recruitment overfishing (Dredge 1992). The fishery is regulated through a range of input controls, as a means of both maintaining breeding stock levels and maximizing value per recruit. Size limits are one of the range of management measures used in the fishery's management (Dredge 1992). Saucer scallops are normally landed whole, for shucking in shore-based factories or in designated near-shore shucking zones where size limits can be policed. They are subject to a size limit of 90 mm (shell height [SH]) in summer and early autumn (November to April inclusive) and 95 mm SH in late autumn, winter, and spring (May to Octo- ber). The differences relate to variations in spawning and the ad- ductor muscle condition of the scallop (Dredge 1981 ). By having a larger size limit in autumn and winter months, fishery managers planned to reduce exploitation during the winter spawning season, when adductor condition is at its poorest, and thus maximize value per recruit (Dredge 1994). Participants in the fishing industry have expressed concerns about the effectiveness of these measures, particularly in relation to the process of holding scallops while they are accurately mea- sured. Some fishermen have suggested that there may be incidental mortality as a consequence of exposing scallops to air before they are graded or measured. Joll (1988) noted the presence of trawl- induced check marks on saucer scallop shells. Such check marks indicate that trawling has a physiological effect on scallops. Saucer scallops taken by Queensland scallop trawlers can be taken in large numbers: up to 30.000 scallops per hour have been taken in opti- mal conditions, although average catches are more typically about 500-1,000 scallops per hour (Trainor pers. comm., Dredge 1992). The time taken to sort and grade such catches varies with catch and by-catch volume, but may be as long as an hour. Should scallops die as a consequence of being exposed to air. the incidental mor- tality of undersized scallops could lead to appreciable wastage and negate the value of size limits. A short study was conducted to determine the survival of cap- tured scallops left out of water, on the sorting tray of a trawler, over varying lengths of time. This study was designed to determine how long saucer scallops could be left out of water, in conditions similar to those encountered on a commercial trawler, before suf- fering mortality in excess of that observed in a control group that underwent minimum exposure to air. METHODS A field trial was commenced in September 1991 at an ex- perimental site off Yeppoon (22 J 47'S. 151°40'E) (Fig. 1). The experiment was commenced in mid-afternoon, at which time weather conditions were sunny and warm (25°C). with wind speeds of less than 10 knots. Approximately 2,000 scallops were captured by otter trawl, with conventional paired. 1 1-m head rope trawls towed from a research vessel in a trawl shot of about 1.5-h duration. This is within the range of trawl shot duration normally used in the Queensland scallop fishery (Dredge and Trainor 1994). Scallops were released from the trawl nets onto a sorting tray, sorted, and treated to according to predetermined procedures. Approximately 350 scallops were removed from the sorting tray and placed in a holding tank, through which fresh seawater was exchanged continuously. Forty-one of these were immediately measured and double tagged with individually identifiable tags (numbered Dymo-tape tags glued to shells with cyano-acrylate adhesive, after Heald 1978: Williams and Dredge 1981 ). A further 300 were similarly measured and tagged after being kept in the holding tank for 3 h. This group of 341 tagged scallops, all of which had been exposed to air for less than 2 min. were used as an experimental control. Tagged scallops were placed into a second deck tank of changing seawater. All remaining scallops were left on the sorting tray, simulating conditions experienced on commer- cial trawlers. At 30-min intervals, 41—45 scallops were taken from the sorting tray, placed in a holding tank, and then individually measured, tagged, and placed in the second holding tank. This procedure was continued for a total time of 3 h. Thus, the final 63 64 Dredge 146° 151° 152° 22°47'S, 151° 40' E Capture location "T 153° 23! Release location * 23°57'S, 151° 58' E Bustard Head 6°_ Bundab nset - Yeppoon to Hervey Bay Figure 1. Location of capture and release positions off of the Queensland coast. Survival of Saucer Scallops 65 batch of tagged scallops had been exposed to air and prevailing weather conditions for 3 h. without being hosed, washed, or at- tended to in any way. The minor variation in numbers of scallops tagged in each 30-min treatment resulted from the field staffs' ability to tag scal- lops at a consistent rate through the duration of the tagging opera- tions (Table 1 ). A total of 597 scallops were tagged during this part of the experiment. At the conclusion of the tagging phase, scallops were checked to find dead animals (none noted) or those that had lost a tag. Such animals were noted, but were released. Tagged scallops were re- leased at a major fishing ground off of Bustard Head, at 23°57'S, 151°58'E. approximately 65 nautical miles south of where they were captured. Fishermen were notified of the experiment (but were not given data on the markings of scallops subjected to dif- ferent exposures) and were asked to return tagged scallops to re- search staff. RESULTS One hundred fifty-one tagged scallops (25.3*) were recaptured and returned. Comparison of recapture rates as a function of size at release (Fig. 2) indicated that recapture rates did not signifi- cantly vary as a function of size at release (Kolmogorov-Smirnov test. D m „ = 0.103. p < 0.05. n = 151). Recapture rates as a function of time exposed to air and pre- vailing weather conditions are given in Table 1. The recovery rate from each treatment has been compared with that from the control group (minimal exposure to air), by the use of x 2 tests. Recovery rates between the control and other treatments were not signifi- cantly different for all treatments exposed to air and prevailing weather conditions up. to and including an exposure time of 150 min. although the recovery rate for scallops exposed for 150 min was lower than that of scallops exposed for 0-120 min. Scallops that had been exposed for longer than 150 min had a significantly reduced recovery rate. DISCUSSION The concept of discarding undersized animals in a fishery is based on the premise that the fishery will be enhanced, either by having discards grow to a size that will increase their market value when subsequently captured or by having them contribute to an enhanced spawning stock. Saucer scallop size limits currently ap- TABLE 1. Releases and recaptures of tagged saucer scallops. Time Exposed Number Number Proportion {% ) (min) Released Recaptured Recaptured x 2 1 341 94 27.6 30 41 13 31.7 0.25 60 43 16 37.2 1.44 90 45 10 22 2 0.47 120 43 13 30.2 0. 1 1 150 42 5 11.9 3.75 180 42 11.59* Total 597 151 * Significant variation between return rate and that of control at 0.01 prob- ability level. Shell height of tagged scallops (mm) Figure 2. Size frequency composition of all tagged scallops versus size frequency composition of recaptured scallops at release. plied to the Queensland trawl fishery are largely based around the concept of maximizing value per recruit. Implicit in this assump- tion is that undersized scallops survive the process of being graded and discarded. Although there is an extensive literature on the survival of intertidal molluscs as a function of exposure to air (see. for ex- ample. McMahon 1988, Gudereley et al. 1994), few if any studies on the survival of scallops in air appear to have been undertaken or documented. Naidu and Cahill (1985) estimated tagging-induced mortality for sea scallops (Placopecten magellicanicus) and con- cluded that the tagging process (which included having live, healthy animals exposed to air for short but undocumented peri- ods) did not induce significant mortality. Brand and Murphy ( 1992) and Allison and Brand ( 1995) undertook extensive tagging programs on scallops (Pecten maximus and Aequipecten opercu- laris, respectively) in the Irish Sea and observed measurable mor- tality induced by tagging. Their studies, however, did not extend to an examination of scallop mortality rates as a function of exposure duration. Return rates (>25% overall) give some indication of how heavily saucer scallops are being fished, at least within local areas. It is interesting to note that a replicate of this experiment, con- ducted some 200 nautical miles north of the release site reported here, at approximately the same time, gave tag returns of <1% (Dredge unpub. data). Such a return rate gave insufficient data to allow statistically meaningful analysis and interpretation. The importance of ensuring that tagged, sedentary animals are released in fishing grounds from which there is a high probability of recapture is emphasized by these results. The results also suggest that there has been no bias induced by the differential release or recapture of different sized scallops through the experiment. There may be the potential for field staff to unconsciously se- lect particular size classes of scallop for tagging. If this had been the case, there may have been potential for confounding errors between size-selective mortality and mortality attributable to ex- posure. The data given in Figure 2 suggest that this was not the case. The results from this study suggest that saucer scallops can withstand extended exposure to air — for upwards of 2 h — before suffering appreciable mortality. This finding is significant in the context of justifying the existing use of size limits as a yield- maximizing tool in the Queensland scallop fishery. The sorting and grading of scallop catches are not prolonged activities during nor- mal fishing operations, with trawl catches typically being cleaned up within 30—15 min of being released onto the deck for sorting. Commercial scallop trawling is now restricted to night time as a means of reducing fishing effort from the previous 24-hour-a-day 66 Dredge fishery. Therefore, the stress suffered by scallops taken in the fishery is less severe than that experienced by animals in this experiment, because they are exposed when temperatures and evaporation rates are reduced. As a consequence, the utilization of size limits as a yield-maximizing technique appears to be based on sound premises. ACKNOWLEDGMENTS The contribution of Dave Trama and Peter Pardee, the crew of the R.V. "Deep Tempest," is gratefully acknowledged. Thanks are also due to Julie Robins for assistance with field operations and subsequent data collation and entry. LITERATURE CITED Allison, E. H. & A. R. Brand. 1995. A mark-recapture experiment on queen scallops. Aequipecten opercularis. on a north Irish Sea fishing ground. J. Mar. Biol. Assoc. U.K. 75:323-335. Brand, A. R. & E. J. Murphy. 1992. A tagging study of north Irish sea scallop (Pecten maximus) populations: comparisons of an inshore and an offshore fishing ground. J. Med. Appl. Malacol. 4:153-164. Dredge. M. C. L. 1981. Reproductive biology of the saucer scallop Amu- sium japonicum balloti (Bernardi) in central Queensland waters. Aust. J. Mar. Freshwater Res. 32:775-787. Dredge, M. C. L. 1988. Recruitment overfishing in a tropical scallop popu- lation? J. Shellfish Res. 7:233-239. Dredge. M. C. L. 1992. Using size limits to maintain scallop stocks in Queensland. In: D. A. Hancock (ed.). Legal Sizes and Their Use in Fisheries Management. Bureau of Rural Resources Proceedings No. 13.. A.G.P.S., Canberra. Australia. Dredge. M. C. L. 1994. Modelling management measures in the Queens- land scallop fishery. Mem. Qld. Mas. 36:277-282. Dredge. M. C. L. & N. Trainor. 1994. The potential for interaction between trawling and turtles in the Queensland east coast fishery. In: R. James (ed.). Proceedings of the Australian Turtle Conservation Workshop. Australian Nature Conservation Agency, Canberra, Australia. Gudereley. H.. A. Demers & P. Couture. 1994. Acclimatization of blue mussel (Mytilus polymorpha, Linnaeus, 1758) to intertidal conditions: effects on mortality and gaping during air exposure. J. Shellfish Res. 13:379-385. Heald, D. I. 1978. A successful marking method for the saucer scallop Amu- sium balloti (Bernardi). Aust. J. Mar. Freshwater Res. 29:845-851. Joll. L. M. 1988. Daily growth rings in juvenile saucer scallops. Amusium balloti (Bernardi). J. Shellfish Res. 7:73-76. McMahon, B. 1988. Respiratory response to periodic emergence in inter- tidal molluscs. Am. Zool. 28:97-114. Naidu. K. S. & F. M. Cahill. 1985. Mortality Associated With Tagging in the Sea Scallop. Placopecten magellanicus (Gmelini. Canadian Atlantic Fish- eries Scientific Advisory Committee Research Document 85/21. Ottawa, Ontario, Canada. Williams, M.J. & M. C. L. Dredge. 1981. Growth of the saucer scallop Amusium japonicum balloti Habe in central eastern Queensland. Aust. J. Mar. Freshwater Res. 32:657-666. Journal of Shellfish Research, Vol. 16. No. 1, 67-70. 1997. REPRODUCTIVE MATURITY AND SPAWNING INDUCTION IN THE CATARINA SCALLOP ARGOPECTEN VENTRICOSUS (=CIRCUEARIS) (SOWERBY II, 1842) P. MONSALVO-SPENCER, A. N. MAEDA-MARTINEZ, AND T. REYNOSO-GRANADOS Division de Biologia Marina Centro de Investigaciones Biologicas del Noroeste, S.C. P.O. Box 128 La Pa:, B.C.S. Mexico, 23,000 ABSTRACT Reproductive maturity and spawning induction were studied in the hermaphroditic catanna scallop Argopecten ven- tricosus. A closed system with seawater recirculation (3 L/min), constant temperature (23 ± 1°C), and salinity (37 ppt) was used. The scallops were fed 3.9 x 10 g cells/animal per day of a 6:3:1 mixture of the microalgae Isochnsis galbana, Chaetoceros sp., and Tetraselmis suecica. Ninety-five percent of the scallops reached reproductive maturity in 27 days. For spawning induction, several methods were used. Thermostimulation combined with the addition of sexual products produced spawning in 50<7t of the animals and was the only method from which both gametes were obtained. Male spawning was initiated in a higher proportion than female spawning. Serotonin (5-hydroxytryptamine) was very effective, inducing sperm spawning only. The rest of the methods (electnc shocks and KC1 injections) failed as spawning inducers. KEY WORDS: Reproductive maturation, spawning induction. Argopeclen ventricosus ( = circularis), scallops INTRODUCTION The catarina scallop Argopecten ventricosus is commercially exploited along the Pacific Coast from Mexico to Peru (Keen 1971 ). This species has a great potential for intensive culture, as in Mexico where some companies are successfully producing it com- mercially by specific methods for hatchery (Maeda-Martinez et al. 1995) and growout (Maeda-Martinez and Ormart-Castro 1995) phases. However, key factors of hatchery seed production, such as the capability to mature broodstock and to induce them to spawn throughout the year, need further attention. Sastry (1963) has sug- gested that gamete development to maturity in Argopecten irradi- ans can be accelerated after gametogenesis has been initiated and that the rate of development to maturation is dependent on tem- perature. In Mercenaria mercenaria and A. irradians, gametoge- nesis has been induced several times in a year by controlling environmental conditions, providing the animals can recuperate from each of the postspawning activities (Loosanoff and Davis 1963, Sastry 1966). The Alligator Harbor population of A. irra- dians that spawns in late summer and autumn has been induced to maturation and stimulated to spawn throughout most of the year (Sastry 1963). Oocyte growth and spawning have been advanced by exposing animals with developing oocytes to 25°C and to 30°C (Sastry 1966). Sastry (1963) developed a reproductive maturity scale for A. irradians, based on the morphoehromatic appearance of the gonads. Stages I to III are immature. IV is mature, and V and VI are partially spent and spent conditions. A specific five- stage scale was proposed by Villalejo-Fuerte and Ochoa-Baez (1993) for A. ventricosus, based on histologic observations. Some aspects of the reproductive biology of A. ventricosus. such as gonad index variation and gonad maturation by histologic meth- ods, have been studied by Villalaz (1992. 1993. 1994, 1996), Villalejo-Fuerte and Ochoa-Baez (1993). and Felix-Pico et al. (1991). Artificial reproductive maturation and spawning induction in A. ventricosus were studied by Aviles-Quevedo and Mucino- Diaz (1988). Those authors found that adult scallops with undif- ferentiated gonads mature in only 20 days at 18°C and 35 ppt salinity and on a diet of 4.0 x 10 9 to 5.0 x I0 9 cells/scallop per day of Isochrysis galbana. The factors inducing spawning in pelecypods have been dis- cussed in reviews by Giese (1959), Loosanoff and Davis (1963), Galtsoff (1961, 1969), Fretter and Graham (1964), Loosanoff (1954, 1971), Giese and Pearse (1974). and Sastry (1979). Tem- perature changes, salinity, light, mechanical shock, and chemicals have been reported to induce spawning. Temperature has been considered one of the important factors in stimulating spawning in a number of pelecypods. It has been reported that serotonin- creatinine-sulfate complex induces sperm spawning in Argopecten irradians (Gibbons and Castagna 1984). Pecten albicans (Tanaka and Murakoshi 1985), Pecten ziczac (Velez et al. 1990), and Ar- gopecten purpuratus (Martinez et al. 1996). In the dioecious scal- lop Patinopecten yessoensis, this neurotransmitter is effective in both males and females (Matsutani and Nomura 1982). The pre- cise role of this complex still remains unknown. In this article, we report a method for artificial reproductive maturation for A. ven- tricosus and the efficacy of different spawning methods. MATERIALS AND METHODS Ripe. Stage IV (Sastry 1963) A. ventricosus were collected by diving at 2 to 3-m depth in Ensenada de La Paz (24°07'N- 110°24'W), Mexico, with only those measuring over 45 mm in shell length selected for the reproductive maturity and spawning induction experiments. At the time of collection, temperature and salinity were 28°C and 37 ppt. Each individual was cleaned and tagged with a plastic label tied to the dorsal auricular lobe of the shell and then left undisturbed in 1,100-L tanks with filtered sea- water at 23 ± 1°C and 37 ppt. This procedure induced massive spawning within the following 5 h. Once spawning stopped, 300 completely spent scallops (pale brown gonads with no differenti- ation between testicular and ovarian regions) were selected for the reproductive conditioning experiments. Reproductive conditioning was done in a closed system with a constant seawater flow (3 67 68 Monsalvo-Spencer et al. L/min). Temperature and salinity remained constant at 23 ± 1 C C and 37 ppt. The scallops were fed with a mixture of Isochrysis galbana, Chaetoceros sp., and Tetraselmis suecica (6:3:1). The amount of food provided was 3.9 x 10 9 cells/scallop per day. Every 4 days, gonadal condition was visually checked. When reproductive maturity was again reached, the following spawning induction methods were tested in 20 individuals for each method: sudden 12°C thermal shock (18-30°C); fast thermal change from 18 to 30°C over 4 min (3°C/min); gradual thermal change over 12 min (l°C/min); gradual thermal change (l°C/min) with sexual product addition (Loosanoff and Davis 1963); 0.025, 0.25, and 2.5 mM intragonadal serotonin (5-hydroxytryptamine) injections (Tanaka and Murakoshi 1985); 0.5. 1 .0. and 2.0 mM intragonadal KG injections (Young 1945); and electric shocks (20 V for 1 sec) (Iwata 1951). Thermal shock experiments were done in 70-L tanks. Initially, each tank contained seawater at the same temper- ature and salinity as in the reproductive maturation experiment. Then, warm seawater was siphoned into the tanks at a rate that produced the desired temperature change until 30°C was reached. Spawning response was considered fast, medium, or slow if ga- mete release began in <3 h, from 3 to 5 h, or >5 h from the stimulus application. RESULTS Tagging techniques allow an exact observation of the gonad behavior and good control to avoid using the same animal in different experiments, although there is no damage. In Figure 1, the development of the A . ventricosus gonad is shown during the experiment. In Stage I or the indifferent stage, gonadal tissue was transparent and it was not possible to distinguish the portion cor- responding to each sex. On the second conditioning day, a few follicles of the gonads on 5% of the animals had developed sper- matogonia and oogonia, as seen by microscopic examination (Stage II). Between days 7 and 9, 85% of the individuals were Stage II. After day 18, 85% of the animals were Stage III, char- 100 TABLE 1. Spawning response of ripe (Stage IV; Sastry 1963) A. ventricosus to different stimuli (n = 20 individuals per treatment). 3 6 9 12 15 18 [ZD Stage I Days EZ23 Stage II 21 24 ■■ Stage Ml ED Stage IV Figure 1. Temporal change in gonad maturation of 45-mm-shell- length A. ventricosus, fed with 3.9 x 10* cells/scallop per day of a mixture of/, galbana, Chaetoceros sp., and T. suecica (6:3:1) at 23 ± 1°C and 37 ppt salinity, n = 300 scallops. Response (% of Group Spawning) % Initial Spawn Fast Medium Slow Method <3h 3-5 h >5h Female Male Sudden thermal 20 25 75 shock, 18-30°C in 1 sec Fast thermal change. 10 50 50 3°C/min increase (18-30°) Gradual thermal 30 35 65 change, l°C/min increase (18-30°C) Gradual thermal 50 40 60 change (as above) plus gamete addition Average % spawn 37.5 62.5 using thermal stimuli Serotonin injection 100 100 (0.025. 0.25, or 2.5 mM) acterized by a uniform pigmentation of the cream-colored testicle and the orange ovary. Gonadal volume was considerably increased at this time. Stage IV. the mature stage, showed brilliant colors in both gonadal portions, dark cream for male and red-orange for female. Pigmentation was very smooth, and gonadal volume in- creased as compared with somatic tissue. On day 27, 95% of the animals were in Stage IV. Table 1 presents the results of the spawning induction experi- ments. Sudden thermal shock ( 18-30 D C) induced spawning in only 20% of the individuals. With this treatment, a medium response was obtained (between 3 and 5 hours from the stimulus applica- tion), and in most cases (75%). sperm was released first. In the fast and gradual thermal change treatments, (3 and l°C/min from 18 to 30°C), only 10 and 30% of the individuals spawned. Re- sponse in these treatments was slow and medium, respectively. When the latter treatment was applied with the addition of sexual products from another scallop, response improved to 50%. Time of response remained at a medium level. An average of female and male initial-spawn percentages from our thermal treatments (Table 1) showed that only 37% of spawning began with ova release whereas 63% were male spawnings. Those scallops that spawned continued to do so, switching from one sexual product to the other, following a random pattern. Serotonin induced sperm spawning in 100% of scallops injected. For the three concentrations tested, the response was the same and fast. Sperm was released 9 ± 1 min after injection. However, this method seemed to be very stressful because the animals opened and closed their valves violently. Be- cause of this movement, there was even a loss of gill fragments. Ejaculation ceased at about the third hour after injection. Ova release after serotonin injections was not observed during this time. There was no response to electric discharge or intragonadal K.C1 injection at any of the concentrations tested. Maturity and Spawning Induction in A. ventricosus 69 DISCUSSION In the Pectinidae, visual examination of the gonad is a direet method of distinguishing the sex and maturity stage on the basis of morphochromatic appearance. Our results show that Sastry's scale developed for A. irradians could be applied in A. ventricosus. In a production hatchery, the morphochromatic method is a fast and reliable alternative in selecting the broodstock for spawning. The reproductive maturity method described in this article pro- duces ripe broodstock in only 27 days at 25°C and 3.9 x 10 9 cells/scallop per day of a mixture of microalgae. It is not known, however, if this treatment accelerates reproductive maturation when compared with the wild. Aviles-Quevedo and Muciho-Diaz (1988) achieved full maturation in this species in less time (20 days), in colder ( 18°C) conditions, and with a higher ration (4.0 x 10 9 to 5.0 x 10 9 cells/scallop per day). The 7-day difference with our results could be caused by a lack of precision in the selection of the broodstock by Aviles-Quevedo and Mucino-Diaz (1988). Those authors used undifferentiated scallops, meaning that Stage II animals could have been used. With the naked eye, it is not possible to distinguish the portion corresponding to each sex in this stage, but maturation is already well advanced. In contrast, in our experiments, the whole cycle from spawning to spawning was considered. If those authors actually matured Stage VI scallops, two alternatives could explain the 7-day difference: (a) scope for activity measurements indicate that in A. ventricosus, there is higher energy available for growth and reproduction at 19°C than at 25°C (M. T. Sicard pers. comm.) and (b) Aviles-Quevedo and Mucino-Diaz used a higher food ration, which probably promoted faster gonad maturation. In a similar scallop (A. irradians), ma- turity under laboratory conditions is reached in 26-30 days at 1 8°C (Castagna and Duggan 1 97 1 ) and in 35 days at 29 ± 1 .0°C ( Sastry 1963), when held in running raw marine water. Thermostimulation is one of the common methods to induce spawning in molluscan species (Loosanoff and Davis 1963). Of the methods used for spawning induction, temperature shock seems to be the alternative to obtain sperm and ova from ripe A. ventricosus. The efficiency of this inducer is improved if sexual products from another scallop is added to the spawning tank. En- hanced effectiveness of thermal stimulation combined with the addition of gametes of the opposite sex has been reported for a number of species (Loosanoff and Davis 1963, Bayne 1965). Wada (1954) has reported that the addition of an egg water sus- pension or a sperm suspension stimulates spawning in Tridacna. Even if our temperature treatments seem to be stressful, there appears to be no negative effect on development because the re- sultant larvae were cultured in the laboratory. Another spawning alternative in this work proved to be the combination of handling (mechanical shock) with a temperature change. Although this method was not evaluated, it produced an unwanted spawning shortly after scallop collection. In A. ventricosus . male spawning was initiated in a higher proportion than female spawning. This also occurs in A. irradians, where spermatozoa are released more readily than are ova (Sastry 1966). No explanation was found for this. In A. ventricosus, serotonin is a very effective inducer of sperm spawning, as it is in A. irradians (Gibbons and Castagna 1984). P. albicans (Tanaka and Murakoshi 1985), P. ziczac (Velez et al. 1990), and A. purpuratus (Martinez et al. 1996). Serotonin, how- ever, fails to induce ova spawning in these hermaphroditic species, whereas in a dioecious species such as P. yessoensis (Matsutani and Nomura 1982), it induces spawning in females. Martinez et al. (1996) have recently found that injections of dopamine and pros- taglandin E 2 , with a 30-min lapse between them, successfully induced ova and sperm spawning in A. purpuratus. This is a promising alternative to be tested in A. ventricosus and other her- maphroditic scallops. ACKNOWLEDGMENTS The authors thank Francisco Cardoza-Velasco for his critical review of the manuscript and Dr. Ellis Glazier for his editing of the English language manuscript. LITERATURE CITED Aviles-Quevedo, M. A. & M. O. Mucino-Diaz. 1988. Gonad condition- ing and spawning of Argopecten circular!! (Sowerby, 1835) under laboratory conditions. Rev. Lalinoam. Acuicult. 38:13-21. Bayne, B. L. 1965. Growth and the delay of metamorphosis of the larvae of Mytilus edulis (L.). Ophelia 2:1-47. Castagna, M. & W. Duggan. 1971. Rearing of the bay scallop, Ae- quipecten irradians. Proc. Natl. Shellfish. Assoc. 61:86-92. Felix-Pico, E. F., M. T. Ibarra-Cruz, R. E. Merino-Marquez. V. A. Levy-Perez, F. A. Garcia-Domiguiez & R. Morales-Hernandez. 1991. Reproductive cycle of Argopecten circularis in Magdalena Bay. B.C.S., Mexico. IFREMER. Actes Collogues. 17:151-155. Fretter, V. & A Graham. 1964. Reproduction, pp. 127-164. In: K. M. Wilbur and C. M. Yonge (eds.). Physiology of Mollusca. vol. 1. Ac- ademic Press, New York. Galtsoff, P. S. 1961 Physiology of reproduction in molluscs. Am. Zool. 1:273-289. Galtsoff, P. S. 1969 The American oyster Crassolrea virginica Gmelin. U.S. Fish. Wildl. Sen-.. Fish. Bull. 64:1^80. Gibbons, M. C. & M. Castagna. 1984. Serotonin as an inducer of spawn- ing in six bivalve species. Aquaculture 40:189-191. Giese, A. C. 1959. Comparative physiology: annual reproductive cycles of marine invertebrates. Annu. Rev. Phxsiol. 21:547-576. Giese. A. C. & J. S. Pearse. 1974. Introduction: general principles, pp. 1^9. In: A C. Giese and J. S. Pearse (eds). Reproduction of Marine Invertebrates, vol. 1. Academic Press, New York. Iwata, K. S. 1951. Gonad development, spawning and rearing of Mytilus sp. larvae in the laboratory. Stud. Rev. GFCM 52:53-65. Keen, A. M. 1971. Sea shells of Tropical West American Marine Mol- luscs from Baja California to Peru. California Stanford Press. Stanford. CA. 1025 pp. Loosanoff, V. L. 1954. New advances in the study of bivalve larvae. Am. Sci. 42:607-624. Loosanoff, V. L. 1971. Development of shellfish culture techniques Proc. Conf. Artif. Propag. Commer. Valuable Shellfish-Oysters. Coll. Mar. Stud., Univ. Delaware 9-40. Loosanoff, V. L. & H. C. Davis. 1963. Rearing of bivalve molluscs. Adv. Mar. Biol. 1:1-136. Maeda-Martinez, A. N. & P. Ormart-Castro. 1995. Sistema marino para el crecimiento y engorda hasta la fase adulta de almeja catarina. Patent No. 180211. I. N.P.I. Mexico. Maeda-Martinez, A. N., P. Monsalvo-Spencer & T. Reynoso-Granados. 1995. Sistema para la crianza intensiva en su etapa juvenil de almeja catarina. Patent No. 180212. I. N.P.I. Mexico. Martinez. G., C. Garrote. L. Mettifogo, H. Perez & E. Uribe. 1996. Monoamines and prostaglandin E, as inducers of the spawning of 70 Monsalvo-Spencer et al. the scallop, Argopeclen purpuralus Lamark. J. Shellfish Res. 15:245— 249. Matsutam, T. & T. Nomura. 1982. Induction of spawning by serotonin in the scallop, Patinopecten yessoensis (Jay). Mar. Biol. Leu. 3:353- 358. Sastry, A. N. 1963. Reproduction of the bay scallop, Aequipecten irradi- ans Lamarck. Influence of temperature on maturation and spawning. Biol. Bull. 125:146-153. Sastry, A. N. 1966. Temperature effects in reproduction of the bay scal- lop. Aequipecten irradians Lamarck Biol. Bull. 130:118-134. Sastry. A. N. 1979. Pelecypoda (excluding Ostreidael. pp. 113-292. In: AC. Giese and J. S. Pearse (eds.). Reproduction of Marine Inverte- brates. Academic Press, New York. Tanaka, Y. & M. Murakoshi. 1985. Spawning induction of the hermaph- roditic scallop Peaen albicans, by injection with serotonin. Bull. Natl. Res. Inst. Aquaculture 7:9-12. Velez, A.. A. Alifa & O. Azuaje. 1990. Induction of spawning by tem- perature and serotonin in the hermaphroditic tropical scallop Peaen ziczac. Aquaculture 84:307-313. Villalaz, J. R. 1992. Laboratory study of reproduction in Argopecten cir- cularis. Natl. Shellfish. Assoc. Abstracts, p. 208. Orlando FI. Villalaz, J. R. 1993. Laboratory study of reproduction in Argopecten ven- tricosus. Natl. Shellfish. Assoc. Abstracts, pp. 134-135. Portland, Oregon. Villalaz, J. R. 1994. Laboratory study of food concentration and temper- ature effect on the reproductive cycle of Argopecten ventricosus . J. Shellfish Res. 13:513-519. Villalaz, J. R. 1996. Histological study of reproduction in Argopecten ventricosus. Natl. Shellfish. Assoc. Abstracts, p. 510. Baltimore, Maryland. Villalejo-Fuerte, M. & R. I. Ochoa-Baez. 1993. The reproductive cycle of the scallop Argopecten circularis (Sowerby, 18351 in relation to tem- perature and photoperiod, in Bahia Concepcion, B.C.S.. Mexico. Cienc. Mar. 19:181-202. Wada, S. K. 1954. Spawning in the tridacnid clams. Jpn. J. Zool. 11: 273-278. Young, R. T. 1945. Stimulation of spawning in the mussel Mytilus cali- fornianus. Ecology 26:50-69. Journal of Shellfish Research, Vol. 16. No. 1. 71-76. 1997. PRIMARY AND SECONDARY SETTLEMENT BY THE GREENSHELL MUSSEL PERNA CANALICULUS S. BUCHANAN* AND R. BABCOCK Leigh Marine Laboratory School of Biological Sciences University of Auckland Auckland. New Zealand ABSTRACT The settlement and postsettlemenl dispersal behavior of the mussel Pema canaliculus was investigated in laboratory- and field-based experiments to determine the role of primary and secondary settlement in its early life history. Field survey data showed size-specific patterns of mussel residency on a variety of rocky shore floral and faunal substrata. Primary settlement (<0.5 mm), both in the laboratory and in the field, was largely on the hydroid Amphisbetia bispinosa and the turfing algae Corallina officinalis, Champia laingii. and Laurencia thyrsifera. In the field, postlarvae between 0.5 and 5.5 mm occurred mostly on the algae C. officinalis, C. laingii. L. thyrsifera, Melanthalia abscissa. Pterocladia lucida, Gigartina cranwellae, and Gigariina alveata. Juvenile and adult mussels (>5.5 mm) were resident predominantly on P. lucida, G. alveata, and Pachymenia himantophora. Size-frequency data from established mussel beds indicated low levels of primary settlement, with the majority of recruitment coming from secondary settlement of larger individuals. Recruitment patterns were consistent with Bayne's primary-secondary settlement model. Substrata deployed onto and near the shore for 21 days recruited both a primary settlement cohort and a secondary settlement cohort of mussels too large to have originated from primary settlement. Differential residency patterns in the field and settlement/recruitment experiments suggested a change in substratum preference by juveniles as a function of size and age. It is suggested that mucus drifting was the likely means of movement for young postlarvae among habitats. Mucus-drifting experiments demonstrated that postlarvae <6 mm in length were able to slow their rate of descent in the water column to 30% of the passive sinking speed using long mucus threads. The size at which P. canaliculus were able to use this method of dispersal greatly exceeds that seen in Mytilus edulis. KEY WORDS: Mussel, Pema canaliculus, settlement, dispersal, mucus drifting, byssopelagic migration INTRODUCTION Many marine mussel species settle preferentially on fine fila- mentous substrata and seldom in the main adult bed (de Blok and Geelen 1958, Bayne 1964, Seed 1969. King et al. 1989, King et al. 1990, Caceres-Martinez et al. 1993). Subsequent dispersal of post- larvae is thought to be the major means of recruitment of juvenile mussels into adult beds. Juveniles of Mytilus edulis. once detached from the site of primary settlement, may "drift" in the sea, sus- pended under a thread of mucus, a behavior termed byssopelagic migration (Sigurdsson 1976. de Blok and Tan-Mass 1977, Lane et al. 1985). The mussel is thus able to move over significant dis- tances and then reattach at some different site. This detachment and reattachment behavior may occur numerous times before re- cruitment into the adult bed occurs. Bayne (1964) first linked the depletion of juvenile mussels attached to seaweed with the con- current recruitment of these same size classes into the adult bed, a pattern later recognized by others (Dare 1976. Seed 1976). It has been suggested that this primary settlement-dispersal-secondary settlement and later recruitment into the adult bed represent an adaptive trait to remove settlement of the larvae from unfavorable conditions within the adult bed (Bayne 1964. Lane et al. 1985.) Evidence for primary settlement directly into the adult bed has raised some controversy as to the general applicability of the pri- mary-secondary settlement model (Petersen 1984a. Petersen 1984b, McGrath et al. 1988, King et al. 1990, Lasiak and Barnard 1995, Caceres-Martinez et al. 1994). Pema canaliculus or the Greenshell mussel is a common bi- valve in New Zealand, growing naturally on lower intertidal and sublittoral coastal shores and on sandy bottoms in deeper water *Present address: Cawthron Institute, Private Bag 2, Nelson. New Zealand. (Paine 1971). Greenshell mussels are intensively farmed in parts of New Zealand, making this shellfish the most important aquaculture species in the country. The larvae of P. canaliculus settle in large numbers on a variety of coastal and drift substrata, such as sea- weeds and hydroids (Hickman 1976). However, abundant primary settlement on commercial spat catching ropes is often followed by virtually complete loss over the following weeks, causing spat supply problems for the Greenshell mussel industry. Spat loss may, in part, be due to juvenile dispersal. This study aimed to identify patterns of settlement and substratum preferences in P. canaliculus, as well as to assess the role of byssopelagic dispersal and the applicability of the primary-secondary settlement model in the recruitment of this species. MATERIALS AND METHODS Mussel Habitat Survey Two study sites were chosen at either end of Piha beach on the West Coast of the North Island, New Zealand (36°S. 174°E). This coast is highly exposed to almost continuous wave action and has a tidal range of approximately 3 m. The volcanic conglomerate rocky shore supports large beds of adult P. canaliculus, extending from approximately 2 m above chart datum to a lower level of about 0.5 m below chart datum. The shoreline is inhabited by a large variety of algal and hydroid substrata on which young mus- sels are abundant. Nine species of substratum were chosen for the survey: Corallina officinalis (COR). Champia laingii (CHA). Melanthalia abscissa (MEL), Pterocladia lucida (PTE), Gigartina alveata (G.A.). Gigartina cranwellae (G.C.). Pachymenia himan- tophora (PAC). Laurencia thyrsifera (L.T.), and Amphisbetia bis- pinosa (AMP). All species occurred at both sites, except for L.T., which was found at Site 1 only, and AMP, which was found at Site 2 only. Three replicate whole-algal samples of each substratum 71 72 Buchanan and Babcock species were collected by hand, by separating the substrata hold- fast from the rock to which they were attached within an area between 0. 1 and 0.4 m above chart datum. Samples were collected during spring tide over 10 sample collections spanning the January to November period of 1993. In addition to the above substrata, three circular 15-cm* core samples of adult mussel bed were col- lected by scraping off the animals within the core. Individual samples were placed in plastic bags and frozen for later analysis. Samples were thawed, and small mussels were removed by vigorous agitation in a domestic bleach solution [1.2% (v/v) hy- pochlorite], which removed the majority of smaller animals. Any remaining individuals were removed with tweezers. Mussels re- moved from substratum samples were recorded on video under appropriate magnification, and the resultant image was captured by JAVA image analysis software (Jandel Scientific). Maximum mus- sel lengths from umbo to ventral margin were then obtained, and length measurements were arranged into size-frequency data. Size-frequency data were further classified into three groups representing different life history stages, as identified in the dis- persal experiments. The first category was the "settlement co- hort": mussels of <0.4999 mm in length. P. canaliculus larvae typically settle at sizes between 250 and 300 u.m (Redfearn et al. 1986. King et al. 1989. Caceres-Martmez et al. 1993). The "dis- persal cohort" contained animals that had the ability to disperse by mucus drifting. These animals were within the size range of 0.5- 5.4999 mm in length. The third category was the "stable cohort": animals >5 mm. representing the proportion of the population that was unable to use mucus drifting as a means of dispersal. All size-frequency data were expressed as proportions and normalized by arcsine square root transformation before statistical analyses were performed. Data were analyzed using single-factor Kruskal- Wallis analysis of variance (at a = 0.05, k = 10, n = 30). after Bartletfs tests for homogeneity of variance (at a = 0.05) identi- fied strong heteroscedasticity in the data. A posteriori Tukey-type Nemenyi multicomparison testing at a = 0.05 was performed for all substrata at each size class for both sites (Zar 1996). This test was used to identify significant differences in the proportional mussel occupancy between substrata in each of the three size co- horts. Transplant Experiment Samples of substrata were collected at low tide from the rocky shore at Site 2. The substrata used were AMP. COR. PTE. MEL. and PAC. Samples were removed with a cold chisel, together with the rock to which they were attached, and transported to the labo- ratory. Mussels were removed from the substrata with tweezers under magnification to ensure that no mussels remained. The rocks to which the substrata were attached were then fixed to tiles (200 x 80 mm) with a fast-curing cement and transported back to the field site. Tiles were cemented directly onto the rocky shore or attached to steel frames positioned approximately 2 m away from the rocky shore in adjacent sand banks. The shore-based and frame-based transplants were composed of 6 and 15 replicate samples, respectively, and were located at 0.2 m above chart da- tum. After a period of 21 days, the remaining substrata were col- lected and returned to the laboratory. Because of storm conditions, some samples were lost from both sites. All mussels on the sub- strata were removed and measured. These measurements were then transformed into size-frequency data for each substratum type as described above. Proportional occupancy in the size classes of resident mussels that could have primarily settled on the substrata (<0.5— 1 .5 mm) and the proportion that colonized via dispersal (size class >1.5 mm) for each sample were normalized by the arcsine square root transformation. Single-factor analysis of vari- ance (a = 0.05) for each size range (settlement and recoloniza- tion) was performed for both transplant types (shore and remote). Tukey multiple comparison (a = 0.05) was used to identify sig- nificant differences among the substrata in both the settlement and the recolonization size ranges for the two transplant types. Sinking Velocity A 1.5-m-tall vertical glass column of 2-cm internal diameter, based on the design of Lane et al. ( 1985). was used to determine the sinking rates of freefalling and mucus-drifting juvenile P. canaliculus. The column was marked at 5-cm intervals and filled with freshly filtered (10-u.m-pore-size filter) seawater at ambient temperatures (16 ± 1°C). Individual mussels collected off of sea- weed from the experimental site were placed in the top of the column and allowed to descend freely. The time interval between consecutive 5-cm intervals was measured with a computer-based timer, and the rate of descent was calculated. The experiment used individuals between 1.5 and 2 mm in length, both live and dead (fixed in 5% formalin). These experiments demonstrated the abil- ity of mussels to reduce sinking rate but were not suitable to describe the slowest sinking velocities of mucus drifters. To mea- sure the slowest sinking velocity, a variable flow chamber was constructed such that prolonged mucus production was possible and sinking rates could stabilize. A longitudinal half-section of 65-mm polyvinyl chloride (PVC) pipe was attached to the inside of a tall (850-mm) glass aquarium. Vertical water flow up the pipe was controlled by varying the water supply through a diffuser at the base of the column. Fine chalk powder could be added to the column, making the invisible mucus thread produced by mussels visible. Animals in the size range of 0.6-8 mm in length were introduced into the top of the flow chamber and allowed to fall against the water flow. As mucus secretion progressed, water ve- locity could be adjusted to prevent the animal from being washed out of the chamber. Once the sinking rate of the mussels stabilized, the water flow was stopped and three to five replicate measure- ments of slowest sinking velocity for each individual were taken. The experiment compared both actively mucus-drifting and non- active live mussels. Settlement Preference in the Laboratory Sexually mature mussels were collected from the field site in midspring and transported to the laboratory, where they were im- mediately induced to spawn using temperature shock (+8°C from ambient) in conjunction with stripped sperm. Spawned eggs were collected and suspended in 200-L tanks with 10 filtered (10-u.m- pore-size filter) seawater at 21-23°C. into which a small amount of sperm was added. Resultant larvae were fed on Isochrysis galbana and Pavlova lutheri for the first 14 days. From Day 15 onward, Chaetoceros gracilis and Chaetoceros calcitrans were added to the diet to give a final algal concentration of approximately 25-35 cells/p,L. Larvae were competent to settle after 25 days. Larvae were placed in an experimental environment containing four ran- domly placed replicates of five mussel-free substratum types, among which they could move freely, and were allowed to settle over 4 days. The apparatus consisted of 20 PVC tubes of 5 cm in diameter and 3 cm in length, covered on the bottom end with Settlement by P. canaliculus 73 160-p.m pore size mesh. The tubes were set with the top end Hush with the bottom of a rectangular PVC open tray. The whole ap- paratus was suspended in a 250-L tank (Fig. 1). The apparatus received water input via two water uplifters at approximately 100 mL/min (=15% total volume). At the end of the experiment, the substrata were gently removed and all settled and metamorphosed mussels were counted for each substratum replicate. The substrata offered were hydroid AMP. COR, Laurencia botryoides (L.B.), MEL. and PAC. Adult mussels were not offered as a substrata choice because observations showed that adult ventilation often binds larvae in pseudofeces. usually killing them. Settlement pref- erence was calculated in two ways: first, in terms of mussel num- bers per gram wet weight, and second, as specific surface area (cnr/gm wet weight) of the substratum offered. Surface area was estimated using JAVA image analysis measurements of a known wet weight of substrata and replicated until variance was <15% of the mean. RESULTS Mussel Residency Survey All of the substrata collected were used by mussels as sites of attachment, and settlement occurred throughout the year, with a peak period in the spring to summer season. Significant differences (p < 0.001 ) in substratum preference were observed among mus- sels of different size classes (Table 1 ). Three general patterns of occupancy among the different substrata could be distinguished: (1) AMP. L.T.. CHA. and COR had high proportions of primary settlers, many residents in the "dispersal" cohort, and an ex- tremely low proportion of "stable" mussels. (2) G.A.. G.C.. and PTE had an intermediate pattern with substantial proportions in each cohort. (3) MEL. PAC. and mussel bed had a low proportion of settlers and increasing proportions in the "dispersal" cohort, with the majority of occupants in the "stable" cohort. Settlement tray & substrata choices , Water uplifter tank I uumn V V V water; flow A movement — larvae substrata choice PVC pipe mesr Figure 1. Diagram showing design of the settlement preference appa- ratus. Substrata were contained within PVC tubes; larvae were able to move freely between substrata choices but were contained within the tray. The apparatus was suspended in a 250-L tank and received water and nutrient flow via a water uplifter. Transplant Experiment Growth of juvenile P. canaliculus at 22°C in the laboratory indicated a growth rate of =21 p.m/day (S. Buchanan, unpubl. data), results similar to those of Bayne (1965). who reported a growth rate of 25 p.m/day in M. edulis. At this growth rate, an) primary settlers (at 300 p.m) on the substrata presented in the transplant experiment should not have exceed 740-825 u.m in length after 21 days. The cutoff between primary settlers and po- tential migrant residents on the transplant substrata was extended to 1.5 mm (between the second and third size classes) to ensure that no primary settlers were misrepresented as postdispersal resi- dents. The transplant experiment demonstrated that postlarval mussels colonized the new substrata. For both the shore transplants and the remote transplants, approximately 55% of the mussels on the transplants were too large (>1.5 mm) to have originated from the primary settlement and growth within the 21 -day experimental period during which the test substrata were deployed (Fig. 2). A significantly differential pattern of secondary settlement specific- ity was found that reflected results of field survey and laboratory settlement experiments. At both transplant sites, differences among the substrata in both the potential primary settler size range and the recolonization size range were significant. On the finely branched substrata COR (settlement: shore, 86%; remote, 89%) and AMP (settlement: shore, 84%; remote, 87%), the majority of recruits were within a size range that could have come via primary settlement, with few animals having been recruited by the second- ary settlement pathway. In contrast, the coarsely branched sub- strata PTE (settlement: shore, 49%; remote. 59%). MEL (settle- ment: shore. 36%; remote, 36%). and PAC (settlement: shore. 0%; remote, 20%) recruited a greater proportion of resident mussels via a secondary settlement pathway than from primary settlement. These results indicated that movement among substrata, rather than differential mortality alone, was a major source of differential distribution of juvenile P. canaliculus on various substrata. Sinking Velocity Over a distance of 1 .5 m. many sinking mussels were able to slow their rate of descent from 5 to around 2 cm s" 1 (Fig. 3). These animals produced mucus threads from the extended pedal organ through the course of their descent. In comparison, dead or inac- tive animals maintained a sinking rate close to 5 cm s _1 through- out. Results from the variable flow chamber demonstrated the effectiveness of the mucus production as a means to reduce the sinking rate (Fig. 4). In all size classes, a velocity reduction to at least 50% of the non-mucus-drifting speed occurred once mucus production reached its peak and slowest sinking rate was estab- lished. Slowest sinking rate was highly correlated with mussel size for both active and nonactive animals. In smaller animals (<3 mm ). a greater reduction to 30% of maximum velocity was usual. The largest animal to produce mucus and effectively reduce its sinking rate was 6 mm in length. The mucus thread, once marked with chalk powder, showed that although mucus was secreted as a fine thread, it often became entwined, folded on itself in the current, and appeared much like a tangled parachute. The mucus thread often exceeded 20-25 cm in length, over 100 times the length of the mussel itself. Settlement ' 'Preference ' ' in the iMboratory Various substrata attracted significantly different levels of mus- sel settlement [Tukey's HSD performed on log (\) transformed data]. Larvae showed a significantly higher settlement on the 74 Buchanan and Babcock TABLE 1. Mean occupancy of P. canaliculus postlarvae in settlement, dispersal, and stable size classes on various substrata at Sites 1 and 2. Site 1 Site 2 Size Class Settlement Kruskal- Wallis 51 Statistic & proba- bility 0.001 Dispersal Stable Size Class Settlement Kruskal- Wallis 114 Statistic & proba- bility 0.001 Dispersal Stable 50 0.001 98 0.001 78 o.ooi 172 0.001 Mean Nemenyi Mean Nemenyi Mean Nemenyi Mean Nemenyi Mean Nemenyi Mean Nemenyi Sub- Occupancy Test Occupancy Test Occupancy Test Sub- Occupancy Test Occupancy Test Occupancy Test stratum Proportion Ranking Proportion Ranking Proportion Ranking stratum Proportion Ranking Proportion Ranking Proportion Ranking COR 0.24 A 0.72 ABC 0.04 DE AMP 0.47 A 0.5 1 B 0.02 DE G.A. 0.17 B 0.50 CD 0.32 BC CHA 0.34 AB 0.54 B 0.11 DE L.T. 0.17 AB 0.68 AB 0.16 CD COR 0.30 BC 0.67 AB 0.03 EF CHA 0. 1 1 AB 0.72 A 0.17 CD PTE 0.20 BC 0.52 B 0.28 BC G.C. 1 1 B 0.74 AB 0.15 CD G.C. 0.14 D 0.66 AB 0.20 CD MEL 0.08 B 0.60 ABC 0.32 B G.A. 0.13 CD 0.29 C 0.58 A PTE 0.03 B 0.61 BC 0.36 B MEL 0.09 D 0.77 A 0.14 CD MUS 0.03 B 0.32 DE 0.64 A MUS 0.02 DE 0.27 C 0.71 A PAC o.oo C 0.46 E 0.54 C PAC 0.02 E 0.36 c 0.62 AB Mean occupancy frequency of resident mussels in the settlement, dispersal, and stable size classes for each substratum over the survey penod for Sites 1 and 2. Kruskal-Wallis single-factor analysis of variance statistics are shown. All size classes produced highly significant results (p < 0.001 at a = 0.05) at both sites. Tukey type a posteriori Nemenyi test results are represented by alphabetical characters. Substrata marked with the same character are not significantly different from one another. Substratum: Corallina officinalis (COR). Champia laingii (CHA). Melanthalia abscissa (MEL). Ptcrocladia lucida (PTE). Giganina alveta (G.A.), Gigartina cranwellae (G.C), Pachymenia himantophora (PAC), Laurencia thyrsifera (L.T.), Amphisbetia bispinosa (AMP), and adult mussel bed (MUS). finely branched substrata AMP and COR in comparison to the substrata MEL and PAC, which were the most coarsely branched. L. botryoides attracted an intermediate settlement density (see Fig. 5). This relationship was seen for settlement expressed as a func- tion of both substratum weight and surface area. On the finely branched substrata, postlarvae were found attached both to branches and at bifurcations within the branching structure. DISCUSSION P. canaliculus is able to disperse using the mucus-drifting mecha- nism also seen in other species of mussels (Sigurdsson 1976, de Blok and Tan-Mass 1977, Lane et al. 1985). Juveniles at lengths of <1 mm exhibited the slowest sinking velocities. <0.5 cms' 1 , results similar to those of Lane et al. ( 1 985 ). who recorded velocities in the order of 0.3 down to 0.03 cm s" 1 for 500- to 700-(xm M. eclulis. Although Lane et al.( 1985) did collect data for M. edulis indicating near-slowest sinking velocity at lengths =1.7 mm, the maximal size at which the postlarvae of this species are able to mucus drift is unclear. Data presented here demonstrated a maximal size limit of 5-6 mm for mucus drifting of P. canaliculus postlarvae, a size significantly larger than the range of = 1-2.5 mm suggested for M. edulis (Bayne 1964. de Blok and Tan- Mass 1977, Lane et al. 1985, King et al. 1990). It has been suggested that the upper size limit for mucus drifting is set more by anatomical changes of pedal glands (Lane et al. 1982) rather than by an inability to cut byssus anchors (Board 1983). If true, this would suggest a prolonged retention of these anatomical features in P. canaliculus. In the wave surges common at the experimental sites used in the survey, upwelling velocities in excess of typical terminal sinking velocities would be expected to be common. Additionally, small air bubbles generated in the surf and bound to mucus would give sig- nificant buoyancy to mucus-drifting mussels in this environment. Un- fortunately, observations of mucus drifting in the field are nearly impossible, and there are no reported observations of such activity in mussels. Postlarvae that are at least neutrally buoyant as the result of mucus may drift over considerable distances, certainly in the order of meters, from the site of original detachment. Observations in this study, of mussels crawling up mucus threads that had attached to the overflow of the variable flow chamber, suggest that mucus thread dispersal serves not only as a means of dispersal but also as an extension of the body that can come in contact with attachment sites, also described by de Blok and Tan-Mass ( 1977). The transplant ex- periments demonstrated that resettlement occurred at the experimen- tal site. Dispersal size class mussels on the shore transplants may have colonized these substrata using byssus reattachment or foot walking; however, this mode of movement was not available at the remote transplant destination where resuspension of the postlarvae is neces- sary. Postdispersal residents on the remote transplants, located adja- cent to. but not continuous with, the rocky shore could have only arrived there through transport in the water column, most likely via mucus drifting. The data demonstrated the important contribution that secondary settlement made to the resident mussel population on these experimental substrata, particularly to the thicker branched species PAC (70%). MEL (60%), and PTE (50%) (Fig. 5). In contrast, mus- sels arriving as primary settlers dominated the population on finer substrata such as COR (85%) and AMP (85%). It was evident from the field survey that a large range of sub- strata were accepted as settlement sites for P. canaliculus pedive- ligers; however, there was a higher settlement rate on the filamen- tous substrata COR. CHA. L. botryoides. L.T., and the hydroid AMP in the field (Table 1 ). This pattern was supported by results of laboratory experiments (Fig. 5). These filamentous species play an important role in community structuring and can be viewed as focal points of intense P. canaliculus settlement. Low settlement specificity to substrata of similar form suggests that the attraction Settlement by P. canaliculus 75 Remote transplants Combined average 08 06 04 02 00 nil ii 1 M TT l I 2 3 4 5 6 7 08 06 04 02 00 Shore transplants Pom, ,, T 08 06 04 02 00 A Coralltna officinalis W+- -1 — I — I — I — I 08 06 04 02 00 012345678 A Hn N=12 12 3 4 5 6 7 08 -i 04 -H _ N=175 oq I I,M | I Ttt i i i i Amphisbetia bispinosa 08 06 04 02 00 12 3 4 5 6 7 AB 12 3 4 5 6 7 TT 012345678 08 or 04 02 00 u$$ *T^ Melanthalia abscissa 08 06 D 04 02 1 00 12 3 4 5 6 7 08 06 04 02 00 08 06 04 02 00 41 A D Pachymenia himantopora 08 06 N=38 4 *T I T I 00 2 3 4 C 12 3 4 5 6 7 C 12 3 4 5 6 7 Pterocladia lucida nfl *#- 08 06 04 02 00 T — I — I 012345678 iOMf . 12 3 4 5 6 7 Size Class (mm) 12 3 4 5 6 7 Size Class (mm) Figure 2. Average frequency as a proportion of the total (±SE) of resident mussels at eight size classes on five substrata from shore- and remote-based transplants after a period of 21 days. Open columns represent size classes of residents that may have been primary settlers; filled columns represent the size classes of mussels that could not have originated from primary settlement within the 21 days of deployment of test substrata. Substrata on which the proportion of recolonization occupancy was not significantly different at a = 0.05, analyzed by Tukey's HSD, have the same alphabetical character above the histo- gram. 6 -i LU CO -H E o a> TO IT D) C 'SZ c 5 - 4 - 3 - 2 - t i I 1 1 1 1 1 20 40 60 80 100 120 140 Distance from top (cm) Figure 3. Change in average sinking velocity ± SE per 5-cm interval of nine active mucus-drifting juvenile P. canaliculus (1.5-2 mm) during descent in a 1.5-m glass column. 12 10 - 8 - 6 - 2 - Non-Active Active 10 Mussel Size (mm) Figure 4. Average sinking velocity ± SE of non-mucus-drifting P. canaliculus (•) and active mucus-drifting P. canaliculus (■). deter- mined in a variable flow chamber. Non-mucus drifters, R 2 = 0.926: active mucus drifters, R 2 = 0.809. of mussel pediveligers to a particular settlement substratum is related more to the general morphology than to chemical compo- sition (Seed 1976), although chemical (Cooper 1982, Eyster and Pechenik 1987), biological (Falmange 1982), and hydrodynamic processes (Taylor and Beattie 1984. Martel et al. 1994) may also play roles in the selection of settlement sites. A A AB BC C weight comarison A A B BC C area comparison Substratum type Figure 5. Primary settlement preference of P. canaliculus larvae in the laboratory. Average numbers ± SE per gram of substrata (black bars) and per cm' of substrata (white bars). Substrata on which settlement preference was not significantly different at a = 0.05, analyzed by Tukey's HSD, have the same alphabetical character above the bar, for numbers ± SE per gram of substrata (weight comparison) and for numbers ± SE per cm 2 of substrata (area comparison!. Substratum are as given for Table 1. 76 Buchanan and Babcock Size-specific distribution patterns observed in the field survey can largely be explained by an active ontogenetic change in post- larval substratum preference with increasing size (and age) facili- tated by byssopelagic dispersal. Differential mortality particular to different substratum types may also account for these patterns; however, the behavioral, demographic, and experimental evidence showed that significant levels of recolonization of natural substrata occurred by >1.5-mm-plus juveniles (even on substrata discontinu- ous with the normal habitat), indicating that dispersal was likely to be the more important of these two processes on many substrata. Substrata such as AMP appeared to have rapidly lost mussel resi- dents of more than a few millimeters in length, whereas others such as PAC and the adult mussel bed appeared to receive indi- viduals by means other than primary settlement. In contrast, in other substrata such as PTE. primary settlement alone could ac- count for the mussels resident. The comparatively low levels of primary settlement in the main adult bed suggested that secondary settlement forms the most sig- nificant mode of recruitment there. The results appear largely con- sistent with Bayne's (1964) primary-secondary settlement model, in which primary settlement and growth are followed by a dis- persal phase that accounts for the juvenile recruitment into the adult bed. This infers that, as an adaptive mechanism to avoid competition with adults, veligers choose not to settle into the adult bed, despite byssal threads providing ample suitable filamentous substrata there (Ester and Pechenick 1987). High survival of pri- mary settling juveniles on filamentous material outside the adult bed, leading to an accumulation of dispersed animals, coupled with low levels of survival of primary settlers within the adult bed. may also account for the common pattern that is the basis of the primary- secondary settlement model. Bayne ( 1964) observed that inhalation of M. edulis larvae by conspecifics will often lead to the death of the larvae. The effect of inhalation of plantigrade P. canaliculus by adults does indeed cause some larval mortality (S. Buchanan, unpubl. data). There is growing acknowledgment that typical mussel recruitment patterns can be more a consequence of primary-settler mortality and postlarval dispersal than an active avoidance by settling larvae. In- creasing evidence of primary settlement directly into the adult bed reinforces the notion that the settlement of larvae is highly variable and does not conform to any one particular pattern (Petersen 1984a. Petersen 1984b. McGrath et al. 1988. King et al. 1990. Caceres- Martinez et al. 1994. Lasiak and Bernard 1995). Larvae may. how- ever, be attracted to the adult bed. despite the inherent risks that this may entail: the attraction of byssus threads and adult shells as sites of settlement has been demonstrated (Eyster and Pechenick 1987, Cac- eres-Martfnez et al. 1994). LITERATURE CITED Bayne. B. L. 1964. Primary and secondary settlement in Mytilus edulis L. J. Anim. Ecol. 33:513-523. Board. P. 1983. The settlement of post larval Mytilus edulis (settlement of post larval mussels). J. Moll. Stud. 49:53-60. Caceres-Martinez, J.. J. A. F. Robledo & A. Figueras. 1993. Settlement of mussels Mytilus galloprovincialis on an exposed rocky shore in Ria de Vigo, NW Spain. J. Shellfish Res. 93:195-198. Caceres-Martinez, J.. J. A. F. Robledo & A. Figueras. 1994. Settlement and postlarvae behaviour of Mytilus galloprovincialis; field and laboratory experiments. Mar. Ecol. Prog. Ser. 112:107-117. Cooper. K. 1982. Potential for application of the chemical DOPA to com- mercial bivalve setting systems. J. Shellfish Res. 3:110-111. Dare, P. J. 1976. Settlement, growth and production of the mussel Mytilus edulis L.. in Morecambe Bay. England. Fish In\'est. 28:1-25. de Blok, J. W. & H. J. F. M. Geelen. 1958. The substratum required for the settling of mussels (Mytilus edulis L.). Arch. Neerlandaises 13:446- 460. de Blok. J. W. & M. Tan-Mass. 1977. Function of byssus threads in young postlarval Mytilus. Nature 267:558. Eyster, L. S. & J. A. Pechenik. 1987. Attachment of Mytilus edulis L. on algal and byssal filaments is enhanced by water agitation. J. Exp. Mar. Biol. Ecol. 114:99-110. Falmagne, C. 1982. Problems associated with the rearing and setting of larvae of the Californian mussel Mytilus califomianus Conrad in a hatchery. Abstracts, National Shellfish Association. West Coast section meeting. 3:1 12. Hickman, R. W. 1976. Potential for the use of stranded seed mussels in mussel farming. Aquaculture 9:287-293. King. P. A.. D. McGrath & W. Britton. 1990. The use of artificial sub- strates in monitoring mussel (Mytilus edulis) settlement on an exposed rocky shore in the west of Ireland. J. Mar. Biol. Assoc. U.K. 70:371- 380. King, P. A., D. McGrath & E. M. Gosling. 1989. Reproduction and settle- ment of Mytilus edulis on an exposed rocky shore in Galway Bay, west coast of Ireland. J. Mar. Biol. Assoc. U.K. 69:355-365. Lane D. J. W.. A. R. Beaumont & J. R. Hunter. 19S5. Byssus drifting and the drifting threads of the young post larval mussel Mytilus edulis. Mar. Biol. 84:301-308. Lane, D. J. W.. J. A. Nott & D. J. Crisp. 1982. Enlarged stem glands in the foot of the post larval mussel Mytilus edulis: adaptation for bysso- pelagic migration. J. Mar. Biol. Assoc. U.K. 62:809-818. Lasiak, T. A. & T. C. E. Barnard. 1995. Recruitment of the brown mussel Perna perna onto natural substrate: a refutation of the primary/ secondary settlement hypothesis. Mar. Ecol. Prog. Ser. 120:147-153. Martel. A.. C. Mathieu, S. Findlay. S. J. Nepszy & J. H. Leach. 1994. Daily settlement rates of the Zebra mussel Dreissena polymorpha. on an artificial substrate correlate with veliger abundance Can. J. Fish Aquat. Sci. 51:856-861. McGrath. D . P. A. King & E. M. Gosling. 1988. Evidence for the direct settlement of Mytilus edulis larvae on adult mussel beds. Mar. Ecol. Prog. Ser. 47:103-106. Paine, R. T. 1971. A short term experimental investigation of resource partitioning in a New Zealand Rocky intertidal habitat. Ecology 52: 1096-1 106. Petersen. J. H. 1984a. Larval settlement behaviour in competing species: Mytilus califomianus (Conrad) and M. edulis L. J. Exp. Mar. Biol. Ecol. 82:147-159. Petersen, J. H. 1984b. Establishment of mussel beds: attachment behaviour and distribution of recently settled mussels {Mytilus califomianus). The Veliger 27:7-13. Redfearn, P., P. Chanley & M. Chanley. 1986. Larval shell development of four species of New Zealand mussels: (Bivalvia. Mytilacea). N.Z. J. Mar. Fresh. Res. 20:152-172. Seed. R. 1969. The ecology of Mytilus edulis L. (Lamellibranchiata) on exposed rocky shores. 1. Breeding and settlement. Oecologia 3:277- 316. Seed. R. 1976. Marine mussels: their ecology and physiology [edited by B. L. Bayne]. Cambridge University Press, Cambridge. Chap. 2, pp. 31-33. Sigurdsson, J. B. 1976. The dispersal of young post-larval bivalve molluscs by byssus threads. Nature 262:386-387. Taylor. R. E. & J. H. Beattie. 1984. Metamorphosis of larvae of the Cali- fornian mussel Mytilus califomianus Conrad. Abstracts, National Shellfish Association. West Coast section meeting. 3:54. Zar, J. H. 1996. Biostatistical Analysis. 3rd ed. Prentice-Hall. Inc., Engle- wood Cliffs, New Jersey. Chap. 1 1, pp. 197-205. Journal of Shellfish Research. Vol. 16. No. 1. 77-82, 1447, ABSORPTION EFFICIENCY AND CONDITION OF CULTURED MUSSELS (MYTILUS EDULIS GALLOPROVINCIALIS LINNAEUS) OF GALICIA (NW SPAIN) INFECTED BY PARASITES MARTEILIA REFRINGENS GRIZEL ET AL. AND MYTIEICOLA INTESTINALIS STEUER ALEJANDRO PEREZ CAMACHO,'* ANTONIO VILLALBA, 2 RICARDO BEIRAS, 3 AND UXIO LABARTA 4 1 Instituto Espanol de Oceanografia Mnelle de Animas Aptdo. 130, E- 15080 A Coruna, Spain Centra de Investigations Marinas Ximta de Galicia Aptdo. 208, E-36600 Vilagarcia de Arousa, Spain Universidade de Vigo Facidtade de Ciencias Departamento de Recursos Naturals e Medio Ambiente Marcosende. E-36200 Vigo, Spain Consejo Superior de Investigations Cientificas Instituto de Investigations Marinas Eduardo Cabello 6, E-36208 Vigo, Spain ABSTRACT An experiment was performed with cultured mussels (Mytilus edulis galloprovincialis) in the Ria de Arousa. NW Spain, under environmental conditions of temperature, salinity, and food availability, in order to determine the effects of Marteilia refringens and Mytilicola intestinalis on the absorption rate, absorption efficiency, and condition of the mussel. M. refrigens significantly reduced absorption of organic matter only when the infection was spread throughout the digestive diverticula of the mussel (heavy infection). Moreover, heavy infections by M. refringens caused a significant loss of the condition of the mussels, probably as a consequence of reduced energy acquisition. The occurrence of M. intestinalis was not associated with reduction of either absorption efficiency or ingestion rate, but infected mussels showed a significantly worse condition. KEY WORDS: Mussel. Mytilus. energetics, absorption, parasites. Marteilia. Mytilicola INTRODUCTION parasites and pathological conditions (Gilek et al. 1992) have been studied. The mussel cultured in the Galician Ri'as (northwestern [NW] Spain) was traditionally referred to as Mytilus edulis (Perez Ca- macho et al. 1991 ). Sanjuan et al. (1990) and Crespo et al. (1990) pointed out that this mussel corresponds to the form Mytilus ?o/- Bayne et al. 1993). Sublethal effects of parasitic infections are , ...... , . r .. , r , „ , . loprmincialis. However, the taxonomic status of this mussel torm Parameters controlling energy acquisition (rates of feeding and absorption efficiency) in bivalve molluscs show a high variability in response to many factors, both exogenous and endogenous (Bayne and Newell 1983. Bayne et al. 1987, Navarro et al. 1991. among those factors (Newell and Barber 1988). Interference by parasitism with physiological mechanisms controlling the energy budget of bivalves has mainly been studied in the Eastern oyster, Crassostrea virginica (Gmelin). parasitized by the endoparasites Haplosporidium nelsoni (Haskin. Stauber & Mackin) (Newell 1985. Barber et al. 1988a, Barber etal. 1988b, Littlewood and Ford 1990. Barber et al. 1991) and Perkinsus marinus (Mackin. Owen & Collier) (Choi et al. 1989) and the ectoparasitic gastropod Boo- nea impressa (Say) (Ward and Langdon 1986. White et al. 1988, Gale et al. 1991 ). In the case of the blue mussel. Mytilus edulis. effects of marine vibrios (McHenery and Birkbeck 1986). the copepod Mytilicola intestinalis (Bayne et al. 1978). and various is still the subject of discussion, because it is considered to be a true species by some authors (Koehn 1991, Sanjuan et al. 1990) and to be the subspecies M. edulis galloprovincialis by others (Gardner 1992. Gosling 1992). Infections by the protistan Mar- teilia refringens and the copepod M. intestinalis are the most sig- nificant pathological conditions affecting cultured mussels of the Galician Rias. with regard to prevalence and pathogenicity (Paul 1983. Figueras et al. 1991. Villalba et al. 1993b). Both parasites inhabit the digestive system of their host. M. refringens multiplies through the digestive epithelia of mussels, and a wide surface of the digestive diverticula epithelium of the host becomes occupied by parasites in heavy infections (Villalba et al. 1993b). The inhi- bition of both gonad and storage tissue development of mussels as *Address for correspondence: Alejandro Perez Camacho. Instituto Espanol a consequence of infection by M. refringens (Villalba et al. 1993a) de Oceanografia. Muelle de Animas. Apdo. 130. E-15080. A Coruna. can be considered as a sign of broad impairment of mussel physi- Spain. FAX: (34) 81229077. ology. Thus, interference by M. refringens with the digestive 77 78 Perez Camacho et al. physiology of the mussel should be expected. M. intestinalis in- habits mainly the intestinal lumen of the host, and its effects on mussel physiology are controversial (Bayne et al. 1978, Theisen 1987, Davey and Gee 1988). According to the review by Morton ( 1983). absorption and intracellular digestion of most of the food ingested by bivalves occur in the digestive tubules in the digestive diverticula, whereas the function of the intestine was not well known at that time. Subsequently. Hawkins et al. (1986) demon- strated that substantial absorption of nonchlorophyll organics takes place in the intestine. An experiment was performed to determine whether infections by M. refringens and M. intestinalis influence the absorption efficiency of the host, and consequently, whether they have an effect on mussel condition. MATERIALS AND METHODS The experiment was carried out with cultured mussels on a raft located at the inner part of the Ri'a de Arousa (Galicia. NW Spain) in late July 1991, under natural environmental conditions of tem- perature, salinity, and food availability. Forty-six mussels of 78.7 ± 0.8 mm (mean ± standard error) in length were taken from a culture rope and arranged within individual trays on the raft itself, with continuous seawater flow. Mussels were permitted to acclimatize for 2 h. and feces pro- duced during this period were removed immediately before the start of the experiment. Subsequently, samples of seawater were taken every 20 min over 3 h for analysis of seston availability. Mussels did not produce pseudofeces during the experiment. Feces produced by each mussel were collected after 1 .5 and 3 h, to determine organic and inorganic contents. Samples of both sea- water and feces were filtered onto preashed (450°C) and weighed in Whatman glass filters type C and rinsed with isotonic ammo- nium formate. Total dry matter (DW) was established as the weight increment determined after drying the filters to constant weight at 90°C. Organic matter (OM) corresponded to the weight loss after ignition at 450°C in a muffle furnace for 24 h. Egestion rates (ER = mg of DW/Ti) were estimated from the total DW content of the feces. Considering that absorption of inorganic matter (IM) through the digestive system is negligible. ER and ingestion rates (IR) of inorganics were assumed to be identical. Thus. IR were calculated from egestion and organic con- tent of the seston (Navarro et al. 1991, Iglesias et al. 1992). IR = ER x %IM(f)/%IM(s) where % IM(f) is the fecal inorganic matter content and % IM(s) is the inorganic matter content of the seston. Absorption rates (A) were estimated as the difference between organic ingestion rates (OIR) and organic egestion rates (OER). Absorption efficiencies (AE) corresponded to the ratio A/OIR. After the completion of physiological determinations, soft tis- sues of each mussel were excised and a cross-section of tissue was removed and processed for histology (Villalba et al. 1993b). The remaining tissue was weighed wet and then dried ( 100°C) so that a dry weight/wet weight ratio could be obtained and used to cal- culate the total dry weight (TDW) of soft tissues of the entire animal (Barber et al. 1988a). The weight of valves was also cal- culated (SW). A condition index (CI) was calculated as follows (Davenport and Chen 1987): CI = 100(TDW/SW) A histological section of each mussel was examined under light microscopy to assess the occurrence of M. intestinalis and to quan- tify the intensity of infection by M. refringens using the scale of Villalba et al. (1993b). The mussels were distributed within the following classes: noninfected mussels (NI), when neither of the parasites was detected; AfyftVzcoZa-infected mussels (Myl) when M. intestinalis was detected; mussels with light infection (LI) by M. refringens, when cells of this parasite were confined to the stom- ach epithelium or even reached primary ducts; heavily infected mussels (HI) by M. refringens, when this parasite was spread through the digestive diverticula and mussels with a mixed infec- tion by both parasites (MIX). No case of moderate infection by M. refringens was found among the mussels used in the experiment. Differences in the percentage of organic content of feces, ER, and absorption among mussels with different intensities of infection by M. refringens and M. intestinalis were analyzed by means of analysis of variance ( ANOVA) of nested samples (Lison 1968. Snedecor and Cochran 1971). Percent data describing or- ganic content were normalized by arcsin transformation. Homo- geneity of variances was checked by Barletf s test. Regressions of In OM Versus In IR for the different infection groups were com- pared by analysis of covariance (ANCOVA). Differences among CI were analyzed by comparing regressions of DW on SW using ANCOVA (Lison 1968. Snedecor and Cochran 1971). All statis- tical procedures were performed with STATGRAPHICS software. In order to use the highest possible number of individuals from the groups with a lower representation, a comparison was made among the fecal organic matter percentage, ER. and IR of the groups NI and Myl; ANOVA did not show significant differences between both groups (p > 0.05). Thus, mussels with mixed infec- tion of M. refringens and M. intestinalis were added to the group of M. refringens-infecled mussels. Regarding CI. significant dif- ferences among NI and Myl were found; therefore, the data for mussels with mixed infection were discarded. RESULTS Infection Levels The distribution of mussels between classes of intensity of infection was as follows: 22 NI, 6 LI, 5 HI, and 9 Myl mussels. The remaining four mussels showed mixed infections. Characteristics of Seston Seston characteristics were 0.56 ± 0.08 mg L~' of particulate matter. 0.38 ± 0.05 mg L' of particulate OM. and an average of 67.5% OM (n = 8). Organic Content of Feces HI mussels had the highest fecal organic percentage, being very similar in all of the groups (Table 1). ANOVA of nested samples showed significant differences in fecal OM* between HI mussels and all of the other groups (p < 0.01 ). Nonsignificant differences were found between NI-Myl, NI-LI. and Myl -LI. Ingestion IR calculated for every infection intensity class are shown in Table 1 . Infection with Marteilia showed a significant effect on IR. Infected mussels (LI and HI) ingested significantly less food than noninfected (NI and Myl) (p < 0.05). In NI mussels, the percentage of organic matter of the feces Mussel Physiology and Parasitic Infections 79 TABLE 1. Percentage of OMF, IR (mg of DW/h), AE. and absorption rate (A = mg OM/h) of each infection intensity class. Parameter NI LI HI Mvl OMF 49.5 ±1.2 IR 1.66 ±0.20 AE 0.51 ±0.03 48.6+1.3 57.7 ±3.0 47.8 ±1.5 1.25 + 0.12 1.04±0.14 I.62±0.31 0.54 ± 0.03 0.31 ± 0.07 0.55 ± 0.04 A 0.65 ±0.09 0.48 ±0.06 0.22 ±0.06 0.68 ±0.15 Presented values correspond to the means of records at 1.5 and 3 h of sampling ± standard error. See text for explanation of abbreviations. (OMF) was related to IR (mg of DW/h). according to the double- logarithmic regression model (means ± standard error): In OMF = 3.919 ± 0.017 - 0.143 ± 0.021 In IR ( r = -0.73. n = 44, p < 0.001) The same model was fitted for the remaining infection classes independently, and the resulting equations were compared by ANCOVA. Significant differences with the MyL and LI infection classes were not found (p > 0.05). In contrast, the OMF was independent of IR for the HI mussels (r = -0.04. p > 0.05). Absorption HI mussels showed a notable decrease in the absorption rate (A; mg/h OM). but the rate was similar between the remaining infection-intensity classes (Table 1 ). Absorption rate data, com- pared by ANOVA after logarithmical transformation in order to homogenize HI variances, showed significant differences between the groups HI-NI (p < 0.05; 1.54 df) and HI-Myl (p < 0.01; 1,28 df). No significant differences were found between NI-Myl, NI-LI, and LI-Myl (p > 0.05). The AE of HI mussels was considerably lower than those of the other groups (Table 1). The AE of NI, Myl. and LI mussels increased as the ingested food content increased, according to y = a • e (Fig. 1). a , whereas it was independent of the IR in HI mussels NI and LI mussels showed the highest CI. The lowest one corresponded to the HI class, whereas Myl mussels showed an intermediate value (Table 2). When regressions of TDW on SW were compared by ANCOVA, significant differences in intercept were found between the groups HI and all other groups and also between NI and Myl (Table 2). DISCUSSION Mean values of absorption rates and AE indicate that heavy infections by M. refringens significantly reduce the absorption of OM in mussels. Light infections by M. refringens did not seem to have marked effects on food absorption, with similar A and AE values for noninfected and lightly infected mussels. In noninfected mussels, AE increased with IR and became asymptomatic at ingestion values of 2-6 mg/h of DW (Fig. 1), reaching a maximum of about 0.65. A very similar relation was observed for Myl and LI mussels, whereas AE was significantly reduced and independent of food ingestion in HI mussels. This pattern does not agree with studies that showed an inverse rela- tionship between AE and IR (e.g., Thompson and Bayne 1972. Navarro and Winter 1982). Nevertheless, it has been pointed out that negative correlations derive from laboratory studies carried out with pure phytoplankton. but not from the heterogeneous sus- pensions occurring in the natural environment (Bayne and Newell 1983). Griffiths (1980) showed that the AE of black mussels did not decrease for seston charges up to 20 mg/L. when feeding on natural detritus. Their result agrees with those of this study. It has been established that AE is directly related to food or- ganic content, and that for seston concentrations lower than the pseudofeces threshold, AE rises with increasing food quality (Bayne et al. 1987, Navarro et al. 1991). At high particle concen- trations (above the pseudofeces threshold), AE can increase with filtration rate as the result of an enrichment of the ingested ration as a consequence of a preingestive selection of organically rich particles (Navarro et al. 1992, Iglesias et al. 1992). However, production of pseudofeces was not observed in this study, and the seston characteristics (both quantity and quality) were the same for all of the mussels, regardless of the quantity of food ingested. Alternatively, the decrease in AE found at decreasing IR (Fig. 1) may be explained as a consequence of the metabolic fecal loss, defined as endogenous material lost from secretion and/or abrasion in the gut (Hawkins et al. 1990). The minimum value of the meta- bolic fecal loss is the absorption value corresponding to ingestion = 0. i.e., the intercept in the equation relating absorption rate (A) to IR in noninfected mussels: A = -0.126 ± 0.021 + 0.471 ± 0.010IR (r = 0.991, n = 44, p < 0.001). We can now recalculate the actual efficiency of food absorption, disregarding losses of endogenous material, by subtracting this value, 0.126, from the organic content of the feces. This would transform the asymptotic curve of Figure 1 into a straight line parallel to the abscissa axis. Y = 0.699. which approximately corresponds to the asymptote of Figure 1. This value is similar to the maximum AE found for mussels from the Ria de Arousa by Navarro et al. (1991). Intracellular digestion and absorption occur in digestive diver- ticula (Bayne et al. 1976. Morton 1983). Observation of histologi- cal sections of HI mussels under light microscopy has shown that wide areas of the digestive diverticula epithelium are occupied by parasites, whereas the surface of digestive tubules is not increased (Villalba et al. 1993b). Therefore, even assuming that the func- tionality of the remaining cells in digestive tubules is intact, the functional surface for intracellular digestion and absorption is sig- nificantly reduced. In addition, the absorption of food material from tubule lumen or neighbor digestive cells by parasites is likely to occur. That would explain why AE is significantly reduced in HI mussels, but not in LI mussels, in which parasites are confined to stomach epithelium or reach primary ducts at most. Because energy acquisition is reduced in heavily infected mus- sels, significant disturbance of mussel physiology can be expected. This could explain the inhibition of both development of storage tissue and gametogenesis in mussels heavily infected by this para- site (Villalba et al. 1993a). Our results show that mean IR was significantly depressed in mussels heavily and lightly infected by M. refringens. Newell (1985) detected significantly lower clearance rates in oysters, C. virginica, infected by H. nelsoni than in noninfected controls. That author suggested that because H. nelsoni multiplies through gill and labial palps, causing a sloughing of their epithelia. disturbance of ciliary function was the most likely cause of clearance rate reduction. However. Barber et al. ( 1991 ) did not detect significant differences in clearance rates between oysters infected and nonin- 80 Perez Camacho et al. > o z III u. 1L UJ z 1 1 1 " a I I 1 1 T 3 - □ - a a a * a 6 a o <*° / fJb a a / a a a 4 2 a / / a / a • a ° a a ■ 9 l 1 1 ' 1 I I >■ u z 1U 1 " b 1 I 1 a 1 1 1 8 □ a ^^*"^a a 6 a a/a a □ a 4 a " 2 a ■ e 1 i 1 ' , ' i. 1 2 3 4 S INGESTION RATE (mg DU/H) INGESTION RATE (mg OU/h) u. 1L UJ z o a. 2 1 n — "d i -1 1 1 i r 8.3 - a - a. 6 - □ ■ a. 4 a a a - e.2 a a a ■ -e.2 a - -e.4 i I 1 1 1 1 1 INGESTION RATE (mg DU/h) INGESTION RATE (mg DU/h) Figure 1. Relationship between AE and IR (mg/h I)W) in: (a) noninfected mussels (AE = 0.7067 ± 0.03V" , ' 6 ± " 05S " R| , r = 0.73, n = 44); (b) M. intestinalis-infected mussels (AE = 0.731 ± 0.066<' 1 - 285 * ° ° go " R ', r = 0.65, n = 18); (c) mussels lightly infected by M. refrigens (AE = 0.774 ± 0.460f'- n385 * 065 »' IRI , r = 0.86, n = 18); (d) mussels heavily infected by M. refrigens (AE = 0.991 fected by H. nelsoni. McHenery and Birkheck ( 1986) also reported inhibition of filtration in M. edulis by marine vibrios, suggesting that some bacterial ciliostatic toxin could be the cause. M. refrin- gens does not proliferate through either gill or labial palp tissues; therefore, a physical interference with the normal function of the filtration system by M. refringens should not be expected. Never- theless, the reduction of IR detected in this experiment concomi- tantly with infection could be considered as a compensatory re- sponse induced by the limited functional capability of the digestive system (Bayne et al. 1987, Navarro et al. 1994). The significant decline in CI in HI mussels could also be ex- plained by the reduction of CI in the rate of energy acquisition. Loss of condition caused by protistan parasites has also been de- scribed in oysters. C. Virginia/, (Newell 1985. Barber et al. 1988a, Choi et al. 1989) and Ostrea edulis (Robert et al. 1991). Similar effects are caused by metazoan parasites in bivalves (reviewed by Lauckner 1983). The effects of M. intestinalis on mussels are controversial. This copepod was blamed for mass mortalities in European mussel beds before 1 970 and. consequently, was considered as a pest for mus- sel populations. However, more recent studies, based on a more detailed knowledge of mussel physiology, have led others to at- tribute minimum detrimental effects to this parasite (reviewed by Davey and Gee 1988). Gee et al. (1977) found some effect of M. Mussel Physiology and Parasitic Infections 81 TABLE 2. DW-SW linear regression parameters, CI ± SE, and ANCOVA between l)W (dependent variable) and SVV (independent variable) of mussels with different types of infection. Regression DW-SW ANCOVA NI LI HI Myl HI-NI HI-Myl HI-Li Myl-NI Myl-LI LI-NI Intercept Slope -0.033 0.264 -0. 1 5 1 -0.307 0.261 0.161 3.540 -0.123 CI 26.0±0.99 24.7±3.60 13.0+2 33 21.6+1.62 Intercept F 29.752 8.484 6.751 5.971 'if S.l. 1.24 0.000 1.11 0.014 1,8 0.032 1.28 0.02 1 NS NS Slope S.l. NS NS NS NS NS NS SW. shell weight; S.l., significance level; df. degrees of freedom. F = F ratio; NS. S.l. > 0.05. intestinalis on the CI of mussels, but only in winter months, when the mean number of parasites per host was over 25. Bayne et al. (1978) reported inhibition of feeding rate by this parasite only when M. edulis had more than 10 parasites, under conditions of high temperature and low ration. Results from our experiment indicated that infection by M. intestinalis did not reduce either IR or AE. but CI was significantly reduced in infected mussels at the experimental conditions, probably because of competition for food energy. Theisen ( 1987) concluded that M. intestinalis has a strong adverse effect on the condition of its host, M. edulis. That author stated that this effect is masked in individual samples with large variation in the condition of the host by the fact that mussels with higher CI are able to lodge more copepods than mussels in poorer condition. In the case of mussels cultured in Galician Rias. Paul ( 1983) found significant effects on host CI by the occurrence of the copepod in only a few samples, mainly in spring and summer, after the main spawning period. That author suggested that heavy bur- dens of copepods may affect the ability of mussels to recover from spawning. The prevalence of this parasite estimated from our mus- sels was much lower than that 80-100% found by Paul ( 1983) in cultured mussels from the same location. The usual method for detection of this parasite involves the dissection of the whole mus- sel. We examined a 6-p.m-thick histological section instead, and thus, the occurrence of M. intestinalis was probably only detected in mussels with a high burden of parasites. The method used for parasite detection and the fact that our experiment was accom- plished in late July, when the mussel reproductive season ends (Villalba et al. 1993a), could explain the detection of a significant effect on mussel condition by M. intestinalis. ACKNOWLEDGMENTS The authors are indebted to Mr. Alfredo Padfn and OPMAR (Organizacion de Productores de Mejillon de Galicia) for its col- laboration in this study. This work was funded by the Conselleria de Pesca, Marisqueo e Acuicultura of the Galician Government. LITERATURE CITED Barber, B. J.. S. E. Ford & H. H. Haskin. 1988a. Effects of the parasite MSX {Haplosporidium nelsoni) on oyster {Crassostrea virginica) en- ergy metabolism. I. Condition index and relative fecundity. J. Shellfish Res. 7:25-31. Barber. B. J.. S. E. Ford & H. H. Haskin. 1988b. Effects of the parasite MSX {Haplosporidium nelsoni) on oyster (Crassotrea virginica) en- ergy metabolism. II. Tissue biochemical composition. Comp. Biochem. Physiol. 91A:603-608. Barber, B. J.. S. E. Ford & D.T.J. Littlewood. 1991. A physiological comparison of resistant and susceptible oysters Crassostrea virginica (Gmelin) exposed to the endoparasite Haplosporidium nelsoni (Haskin. Stauber & Mackin). /. Exp. Mar. Biol. Ecol. 146:101-112. Bayne. B. L., J. M. Gee, J. T. Davey & C. Scullard. 1978. Physiological responses of Mytilus edulis L. to parasitic infestation by Mytilicola intestinalis. J Cons hit Explor. Mer 38:12-17. Bayne, B. L., A. J. S. Hawkins & E. Navarro. 1987. Feeding and digestion by the mussel Mytilus educlis L. (Bivalvia: Mollusca) in mixtures of silt and algal cells at low concentrations. J. Exp. Mar. Biol. Ecol. 111:1-22. Bayne. B. L.. J. 1. P. Iglesias. A. J. .S. E. Navarro. H. Heral & J. M. Deslous-Paoli. 1993. Feeding behaviour of the mussel, Mytilus edulis: responses to variations in quantity and organic content of the seston. /. Mar. Biol. Assoc. U.K. 73:813-829. Bayne, B. L. & R. C. Newell. 1983. Physiological energetics of marine molluscs, pp. 407-5 15. In: K. M. Wilbur and A. S. M. Saleuddin (eds.). The Mollusca. 4. Physiology. Part I . Academic Press, London. Bayne, B. L., R.J. Thompson & J. Widdows. 1976. Physiology: I. pp. 121-206. In: B. L. Bayne (ed.). Marine Mussels: Their Ecology and Physiology. Cambridge University Press. New York. Choi. K. S„ E. A. Wilson, D. H. Lewis. E. N. Powell & S. M. Ray. 1989. The energetic cost of Perkinsus marinus parasitism in oysters: Quan- tification of the thyoglicollate method. J. Shellfish Res. 8:125-131. Crespo. C. A.. T. Garcia Caballero, A. Beiras & J. Espinosa. 1990. Evi- dence from sperm ultrastructure that the mussel of Galician estuaries is Mytilus galloprovincialis Lamarck. J. Moll. Stud. 56:127-128. Davenport, J. & X. Chen. 1987. A comparison of methods for the assess- ment of condition in the mussel (Mytilus edulis L.). J. Moll. Stud. 53:293-297. Davey, J. T. & J. M. Gee. 1988. Mytilicola intestinalis. a copepod parasite of blue mussel. Am. Fish. Sac. Spec. Publ. 18:64—73. Figueras. A. J., C. F. Jardon & J. R. Caldas. 1991. Diseases and parasites of rafted mussels (Mytilus galloprovincialis Lmk.): preliminary results. Aquaculture 99:17-33. 82 Perez Camacho et al. Gale, L. D.. I.J. Manzi & M. P. Crosby. 1991. Energetic costs to the Eastern oyster Crassostrea virginica due to recent parasitism by the ectoparasitic gastropod Boonea impressa. Mar. Ecol. Prog. Ser. 79: 89-98. Gardner. J. P. A. 1992. Mytilus galloprovincialis (Lmk) (Bivalvia. Mol- lusca): the taxonomic status of the Mediterranean mussel. Ophelia 35:219-243. Gee, J. M. L. Maddock. J. T. Davey. 1977. The relationship between in- festation by Mytilicola intestinalis, Steuer (Copepoda. Cyclopoidea) and the condition index of Mytilus edulis in southwest England. J. Cons. Int. Explor. Mer. 37:300-308. Gilek. M, M. Tedengren & N. Kautsky. 1992. Physiological performance and general histology of the blue mussel, Mytilus edulis L.. from the Baltic and North Seas. Neth. J. Sea Res. 30:1 1-21. Gosling, E. M. 1992. Genetics of Mytilus. pp. 309-372. In: E. Gosling (ed.). The Mussel Mytilus: Ecology, Physiology, Genetics and Culture. Elsevier. Amsterdam. Griffiths, R. J. 1980. Natural food availability and assimilation in the black mussel Choromytilus meridionalis. Mar. Ecol. Prog. Serg. 3:151-156. Hawkins. A. J. S.. B. L. Bayne. R. F. C. Mantoura. C. A. Llewellyn & E. Navarro. 1986. Chlorophyll degradation and absorption throughout the digestive system of the blue mussel Mytilus edulis L. J. Exp. Mar. Biol. Ecol. 96:213-223. Hawkins. A. J. S.. E. Navarro, J. I. P. Iglesias. 1990. Comparative allom- etries of gut-passage time, gut content and metabolic faecal loss in Mytilus edulis and Cerastoderma edule. Mar. Biol. 105:197-204. Iglesias. J. I. P., E. Navarro. P. Alvarez Joma & I. Armentia. 1992. Feed- ing, particle selection and absorption in cockles Cerastoderma edule (L.) exposed to variable conditions of food concentration and quality. J. Exp. Mar. Biol. Ecol. 162:177-198. Koehn. R. K. 1991. The genetics and taxonomy of species in the genus Mytilus. Aquaculture 94:125-145. Lauckner, G. 1983. Diseases of Mollusca: Bivalvia. In: O. Kmne (ed.). Diseases of Marine Animals. 2. Biologische Anstalt Helgoland, Ham- burg. Lison. L. 1968. Statistique Appliquee a la Biologic Experimentale. Gauthier-Villars. Paris. Littlewood, D. T. J. & S. E. Ford. 1990. Physiological responses to acute temperature elevation in oysters. Crassostrea virginica (Gmelin. 1971 ). parasitized by Haplosporidium nelsoni (MSX) (Haskin. Stauber. and Mackin, 1966). J. Shellfish Res. 9:159-163. McHenery, J. G. & T. H. Birkbeck. 1986. Inhibition of filtration in Mytilus edulis L. by marine vibrios. J. Fish Dis. 9:257-261. Morton. B. 1983. Feeding and digestion in Bivalvia. pp. 65-147. In: K. M. Wilbur and A. S. M. Saleuddin (eds.). The Mollusca. 5. Physiology. Part 2. Academic Press. London. Navarro, E., J. I. P. Iglesias & M. Ortega. 1992. Natural sediment as a food source for the cockle Cerastoderma edule (L.): effect of variable par- ticle concentration on feeding, digestion and the scope for growth. J. Exp. Mar. Biol. Ecol. 156:69-87. Navarro. E.. J. I. P. Iglesias. M. Ortega & X. Larretxea. 1994. The basis for a functional response to variable food quantity and quality in cockles Cerastoderma edule (L.) (Bivalvia. Cardiidae). Physiol. Zool. 67:468- 496. Navarro. E., J. I. P. Iglesias, A. Perez Camacho. U. Labarta & R. Beiras. 1991. The physiological energetics of mussels (Mytilus galloprovin- cialis Lmk) from different cultivation rafts in the Ria de Arosa (Galicia, N. W. Spain). Aquaculture 94:197-212. Navarro. J. M. & J. E. Winter. 1982. Ingestuion rate, assimilation effi- ciency and energy balance in Mytilus chilensis in relation with body size and different algal concentrations. Mar. Biol. 67:255-266. Newell, R. I. E. 1985. Physiological effects of the MSX parasite Haplospo- ridium nelsoni (Haskin. Stauber & Mackin) on the American oyster Crassostrea virginica (Gmelin). J. Shellfish Res. 5:91-95. Newell. R. I. E. & B. J. Barber. 1988. A physiological approach to the study of bivalve molluscan diseases. Am. Fish. Soc. Spec. Publ. 18: 269-280. Paul, J. D. 1983. The incidence and effects of Mytilicola intestinalis in Mytilus edulis from the Ri'as of Galicia, North West Spain. Aquaculture 31:1-10. Perez Camacho, A. R. Gonzalez & J. Fuentes. 1991. Mussel culture in Galicia (N. W. Spain). Aquaculture 94:263-278. Robert, R„ M. Borel. Y. Pichot & G. Trut. 1991. Growth and mortality of the European oyster Ostrea edulis in the Bay of Arcachon (France). Aquat. Living. Resour. 4:265-274. Sanjuan. A.. H. Quesada. C. Zapata & G. Alvarez. 1990. On the occurrence of Mytilus galloprovincialis Lmk. on the NW coast of the Iberian Peninsula. J. Exp. Mar. Biol. Ecol. 143:1-14. Snedecor, G. W. & W. G. Cochran. 1980. Statistical Methods. Iowa State University Press, Ames. Theisen, B. F. 1987. Mytilicola intestinalis Steuer and the condition of the host Mytilus edulis L. Ophelia 27:77-86. Thompson. R. J. & B. L. Bayne. 1972. Active metabolism associated with feeding in the mussel Mytilus edulis L. Kiel. Meeresforsch. 19:20-41. Villalba, A., S. G. Mourelle. N. J. Carballal & M. C. Lopez. 1993a. Effects of infection by the protistan parasite Marteilia refringens on the repro- duction of cultured mussels Mytilus galloprovincialis in Galicia (NW Spain). Dis. Aquat. Org. 17:205-213. Villalba. A.. S. G. Mourelle. M. C. Lopez, M. J. Carballal & C. Azevedo. 1993b. Study of marteiliasis affecting cultured mussels {Mytilus gal- loprovincialis) of Galicia (NW of Spain). I. Etiology, phases of the infection and temporal and spatial variability in prevalence. Dis. Aquat. Org. 16:61-72. Ward. J. E. & C. J. Langdon. 1986. Effects of the ectoparasitic Boonea ( = Odostomia) impressa (Say) (Gastropoda: Pyramidellidae) on the growth rate, filtration rate, and valve movements of the host (Crassos- trea virginica) (Gmelin). J. Exp. Mar. Biol. Ecol. 99:163-180. White, M. E„ E. N. Powell & S. M. Ray. 1988. Effect of parasitism by the Pyramidellid Gastropod Boonea impressa on the net productivity of oysters (Crassostrea virginica). Estuar. Coast. Shelf Sci. 26:359-377. Journal of Shellfish Research. Vol. 16. No. 1, 83-85. 1997. MUSSEL (MYTILUS GALLOPROVINCIALIS LAMARCK) SETTLEMENT IN THE RIA DE VIGO (NW SPAIN) DURING A TIDAL CYCLE JORGE CACERES-MARTINEZ* AND ANTONIO FIGUERASt Institute de Investigaciones Marinas CSIC Eduardo Cabello 6. 36208 Vigo. Spain ABSTRACT The settlement of mussel was determined during a tidal cycle in an exposed rocky shore in the Ria de Vigo (north- western Spain) and 300 m away from it. In the exposed rocky shore, mussel settlement was recorded throughout the intertidal profile during the tidal cycle. Settlement was more abundant in the lower than in the upper intertidal zone. The size of settled mussels varied from 0.250 to 1 1 mm. The largest mussels were found in the lower intertidal zone. Maximum densities were recorded during the high tide, and the minimum were recorded during the low tide. Three hundred meters away from the mussel bed, settlement occurred during the complete tidal cycle. A light increase in the number of settled mussels during high tide was recorded. Settlement was more abundant at 2- than at 5- and at 8-m depth. The size of settled mussels varied from 0.225 to 0.375 mm and was similar at all depths studied. KEY WORDS: Mytilus galloprovincialh, settlement, tidal cycle INTRODUCTION Mussel settlement has been widely studied under very different conditions (Maas-Geesteranus 1942. de Blok and Geelen 1958. Bohle 1971. Dare 1973. Dare 1976. Hrs-Brenko 1973, Dare et al. 1983. Sigurdsson et al. 1976. Petersen 1984, King et al. 1989. Newell et al. 1991, Caceres-Martt'nez et al. 1993, Caceres- Martt'nez et al. 1994. McGrath et al. 1994). Mussel larvae distri- bution during tidal cycles have been studied by Newell et al. (1991). However, to our knowledge, no studies on mussel settle- ment during a tidal cycle have been done. The aim of this work was to determine variation in the number of settling mussels (Mytillus galloprovincialis) during a tidal cycle through the inter- tidal zone in a mussel bed and in a sampling location 300 m away from it. MATERIALS AND METHODS In the summer of 1993, when major settlement of mussels occurs in the area (Caceres-Martinez et al. 1993. Caceres-Martt'nez et al. 1994), sampling of the mussel settlement during a tidal cycle, from July 2 1 to 22, in the exposed rocky shore of Cabo Home, on the oceanic side of the Ria de Vigo (42°5'N, 8°52'W). was carried out. Pieces of 19 x 16 x 0.8 cm of synthetic fibrous material (Commercial Scotch Brite®) were used as collectors. A pulley system was placed in the exposed rocky shore, from the low-water spring tide level mark to the high-water spring tide level mark. A series of six collectors by duplicate were hung from the polyeth- ylene ropes (0.5 cm in diameter) every 5 m: additional weight for collectors was not required. These collectors were replaced three times: 6 h after the first low tide, 6 h after the high tide, and 6 h after the following low tide. Simultaneously, a series of three collectors were submerged at 2-. 5-, and 8-m depth from a boat anchored at 300 m in front of the rocky shore (sublittoral site), and these were replaced every 2 h. Two replicates were made for each collector. The tidal heights in meters were estimated from the lower water tidal height mark recorded for June 21. 1993, in the *Present address: Centro de Investigacion Cienti'fica y de Educacidn Su- perior de Ensenada. Depaitamento de Acuicultura, Apartado. Postal 2732. 2800. Ensenada, Baja California, Mexico. tAuthor to whom any correspondence should be sent. tide tables (Junta del Puerto y Ria de Vigo 1993). The low-water tide level is defined as zero in tidal data. To separate spat, each collector was immersed for 5 min in a 10% solution of commercial sodium hypochlorite (Na CIO) and rinsed in tap water onto a series of 0.09- to 4.0-mm sieves. The resulting fractions were dried in an oven at 80°C for 24 h. Mussels were separated with a brush for study under a stereoscopic micro- scope, and all mussels in a fraction were counted. The mussels obtained in the sieves under 3-mm mesh were measured with a micrometer (total shell length). Larger mussels were measured with a caliper to the nearest 0.5 mm. Results are presented as the number of individuals per collector. STATISTICAL ANALYSIS A Kruskal-Wallis test followed by Tukey-type multiple com- parisons (Zar 1984) was used for comparisons in the settlement in different localities. A Mest and one-way analysis of variance (ANOVA) to Wilcoxon signed rank test were used to compare size composition in different localities (Sokal and Rohlf 1981). RESULTS AND DISCUSSION Mussel Settlement in the Rocky Intertidal Zone Over a Tidal Cycle Mussel settlement occurred at each sampling site in the inter- tidal zone. The abundance was higher after the flood of the tide than after the ebb, suggesting that tide and waves may carry mus- sels to the shore and transport them again to the open sea (Fig. 1 ). This is similar to barnacle dispersion during their settlement pro- cess (de Wolf 1973). Cyprids are transported by tidal currents, sinking at that time to the bottom during periods of low current speed and then being dispersed again in the water column when water speed increases. Newell et al. (1991) found that mussel larvae are more abundant on the flood tides, indicating inshore and estuarine retention. In this study, the minimum and maximum mussel sizes recorded were, respectively, 0.225 and 11.0 mm. Interestingly, mussels larger than 10.0 mm were found attached to collectors. Their presence may be explained by the capacity of bivalve and gastropod postlarvae stages up to 2 mm to produce the contact mucous threads, also named byssus threads (Sigurdsson et al. 1976. de Blok and Tan Mass 1977). that are used to make contact with the substrate, allowing settlement and providing ad- 83 84 Caceres-Marti'nez and Figueras + 3.6 intertidal height (ml Figure 1. Mean number of mussels settled (4-standard deviation) on collectors placed at different heights of the intertidal profile of an exposed rocky shore during a tide cycle in Ria de Vigo, Spain. ditional buoyancy for dispersion. (Beukema and de Vlas 1989, Martel and Chia 1991. Caceres-Marti'nez et al. 1994). In mussels, the superior limit in the size of individuals with the capacity to produce these mucous threads has not been established; our results suggest that this size is around 10 mm. Similarly. Beukema and de Vlas (1989) found Macoma balthica of a shell length of 10 mm with these mucous threads. The range of mussel sizes found on the lower intertidal zone was wider than that on the upper zone (Fig. 2). The number of mussels settled on collectors increased from the upper to the lower intertidal zone. This was corroborated by a regression between tidal height and total number of settled mus- sels, which was significant (v = 95.047 - 8.9.v. R 2 = 0.7. p < 0.01). If settlement depends on a chance encounter between the mussel and the appropriate substrate, one reason for higher mussel abundance in collectors placed at the lower intertidal zone than at the upper one could be the longer immersion time period of these collectors. On the other hand, the buoyancy of pediveliger stages, marked by the coexistence of velum, foot (Widdows 1991). and contact mucous thread (Sigurdsson et al. 1976. de Blok and Tan Mass 1977. Caceres-Marti'nez et al. 1994), seems to be greater than the buoyancy of postlarvae and larger mussels (>0.470— 1 1 mm) (Bayne 1971 ). This could explain why postlarvae and larger mus- sels are more abundant in the lower intertidal zone than in the upper one. Spat Abundance al Different Depths During the Tidal Cycle Spat abundance recorded during the tidal cycle showed an ir- regular pattern at the sublittoral site. However, an increase in spat after high tide was detected, especially in the collector placed at 2 m (Fig. 3). Spat were more abundant at 2- than at 8-m depth, and this was statistically significant (Kruskal-Wallis test. H = 8.502, p < 0.01. followed by Tukey-type multiple comparisons). This could be explained by two hypotheses: ( 1 ) collectors placed at 8-m depth were dragged by the waves on the sandy bottom, limiting 30 20 20- 10- 20 10 30- 20 n=49 -3.6 n = 51 2.4 :52 0651 0851 I- 051 0750 0950 1 150 551 751 951 1 151 650 0850 1050 1250 Shell length (mm) 1251 1 550 1451 ' 1550 1.351 1450 Figure 2. Size distribution of mussels settled on collectors placed at different heights of the intertidal profile of an exposed rocky shore during a tide cycle in Ria de Vigo, Spain. (+standard deviation). attachment as the result of friction against the substrate and/or (2) ascending (tidal) currents occurred at that moment. It is known that bivalve larvae vertical distribution may respond to tidally induced cues such as changes in pressure, temperature, or salinity. How- Figure 3. Mean number of mussels settled (+standard deviation) on collectors placed at 2- (black bars), 5- (gray bars) and 8-m (white bars) depth, during a 24-h cycle in Ria de Vigo, Spain. The black line indi- cates the tidal fluctuation in depth (m). Mussel Settlement During a Tidal Cycle 85 ever, this responses may be overridden by the energy of the sys- tem, and the larvae behave as inanimate particles in their distri- bution (Newell et al. 1991). Interestingly, the minimum and maximum sizes recorded on- collectors placed at the sublittoral site were 0.225 and 0.375 mm (corresponding to pediveliger larvae stages), respectively. No sta- tistically significant differences were detected among the sizes of mussels settling at different depths (one-way ANOVA. F = 0.064. p > 0.05). However, there were statistically significant differences between the sizes of the mussel spat from the sublittoral site (0.225 and 0.375 mm) and those placed in the intertidal zone (0.225-1 1.0 mm) (/-test, p < 0.001). The phenomenon of mussel postlarvae dispersion is not entirely understood. Several authors (Bohle 1971. Hrs-Brenko 1973, Rees 1954, Kautsky 1982) found very few mus- sel larvae larger than 300 p.m in plankton hauls in several studies carried out in very different areas, concluding that postlarvae were absent or scarcely present in plankton and disregarding the occur- rence in this species of planktonic postlarvae dispersion. Our re- sults suggest that postlarvae dispersion occurs mainly at a local level, especially in the lower intertidal zone (see above), where the mussel bed is dense and postlarvae attached to suboptimal sub- strates may be continuously detached (Caceres-Martmez et al. 1994). The early life strategy of the mussel, with its planktotrophic existence, accounts for the high dispersal capability of most spe- cies within the genus Mytilus (Lutz and Kenish 1992). Additional dispersion of postlarvae stages provides the species with a more or less local redistribution possibility. Further research on the factors that control the postlarvae dispersion process and its ecological significance, among them, settlement studies during tidal cycles, is needed. ACKNOWLEDGMENTS The authors thank J. A. F. Robledo. I. Sanchez, G. Fernandez, and R. Casal for their help during the field study and H. Alvarez and C. Feijoo for their help in the sampling process. Thanks also to the mussel farmers A. Acufia and R. Curras. J. C.-M. was sup- ported by a grant from the Consejo Nacional de Ciencia y Tech- nologi'a (CONACyT) from Mexico and by the Consejo Superior de Investigacion Cientffica (CSIC) from Spain. LITERATURE CITED Bayne. B. L. 1971. Some morphological changes that occur at the meta- morphosis of the larvae of Mytilus edulis. pp. 259-280. In: D. J. Crisp (ed.). Proceedings of the 4th European Marine Biological Symposium, Bangor, U.K.. 1969. Cambridge University Press. London. Beukema, J. J. & J. de Vlas. 1989. Tidal-current transport of thread-drifting postlarval juveniles of the bivalve Macoma balthica from the Wadden Sea to the North Sea. Mar. Ecol. Prog. Ser. 52:193:200. Bohle. B. 1971. Settlement of mussel larvae Mytilus edulis on suspended collectors in Norwegian waters, pp. 63-69. In: D. J. Crisp (ed.). Pro- ceedings of the Fourth European Marine Biological Symposium. Bangor. Cambridge University Press. London. Caceres-Marti'nez. J., J. A. F. Robledo & A. Figueras. 1993. Settlement of mussels Mytilus galloprovincialis on an exposed rocky shore in Ria de Vigo. N W Spain. Mar. Ecol. Prog. Ser. 93:195-198. Caceres-Marti'nez. J.. J. A. F. Robledo & A. Figueras. 1994. Settlement and post-larvae behaviour of Mytilus galloprovincialis: field and laboratory experiments. Mar. Ecol. Prog. Ser. 112:107-117. Dare. P. J. 1973. The stocks of young mussels in Morecambe Bay. Lan- cashire. Shellfish information leaflet. Minist. Agric. Fish. Food bond. No. 28:1-14. Dare. P. J. 1976. Settlement, growth and production of the mussel. Mytilus edulis L.. in Morecambe Bay. England. Fish. Invest. Minist. Agric. Fish. Food Lond.. Ser II.. 28:1-25. Dare, P. J., D. B. Edwards & G. Davies. 1983. Experimental collection and handling of spat mussels (Mytilus edulis L.) on ropes for intertidal cultivation. MAFF (Lowestoft) Fish. Res. Tech. Rep. No. 74:1-23. de Blok. J. W. & H.J. Geelen. 1958. The substratum required for the settling of mussels (Mytilus edulis L.). Arch Neerl. Zool. Vol. Jubilaire, 13(11:446-460. de Blok. J. W. & M. Tan Maas. 1977. Function of byssus threads in young post-larval Mytilus. Nature. 267:558. de Wolf. P. 1973. Ecological observations on the mechanisms of dispersal of barnacle larvae during planktonic life and settling. Neth. J. Sea Res. 6:1-129. Hrs-Brenko. M. 1973. The study of mussel larvae and their settlement in Vela Draga Bay (Pula, the northern Adriatic sea). Aquaculture 2:173- 182. Junta del Puerto y Ria de Vigo. 1993. Tablas deMareas. Servicio de pub- licaciones. Secretaria General Tecnica del Ministerio de Obras Piiblicas y Urbanismo. Espana. Kautsky. N. 1982. Quantitative studies on gonad cycle, fecundity, repro- ductive output and recruitment in a Baltic Mytilus edulis population. Mar. Biol. 68:143-160. King, P. A., D. McGrath & E. M. Gosling. 1989. Reproduction and settle- ment of Mytilus edulis on an exposed rocky shore in Galway Bay, West coast of Ireland. J. Mar. Biol. Assoc. U.K. 69:355-365. Lutz, R. A. & J. M. Kennish. 1992. Ecology and morphology of larval and early postlarval mussels pp. 53-85. In: E. Gosling (ed.). The Mussel Mytilus: Ecology. Physiology, Genetics and Culture. Elsevier. Amster- dam. Maas-Geesteranus. R. A. 1942. On the formation of banks of Mytilus edu- lis. Arch. Neerl. Zool. 6:283-325. Martel. A. & F. S. Chia. 1991. Foot-raising behaviour and active partici- pation during the initial phase of post-metamorphic drifting in the gasteropod Lacuna spp. Mar. Ecol. Prog. Ser. 72:247-254. McGrath. D.. P. A. King & M. Reidy. 1994. Conditioning of artificial substrata and settlement of the marine mussel Mytilus edulis L.: a field experiment. Biol. Environ. Proc. R. Irish Acad. 94B:53-56. Newell. C. R.. H. Hidu. B. J. McAlice. G. Podniesinski. F. Short & L. Kindblom. 1991. Recruitment and commercial seed procurement of the blue mussel Mytilus edulis in Maine. ./. World Aquacult. Soc. 22:134- 152. Petersen, J. H. 1984. Larval settlement behaviour in competing species: Mytilus califonuanus Conrad and M. edulis L. J. Exp. Mar. Biol. Ecol. 82:147-159. Rees, C. B. 1954. Continuous plankton records: the distribution of lamel- libranch larvae in the North Sea. 1950-51. Bull. Mar. Ecol. 4:21-46. Sigurdsson. J. B.. C. W. Titman & P. A. Davies. 1976. The dispersal of young postlarval bivalve mollusc by byssus threads. Nature 262:386- 387. Sokal. R. S. & F.J. Rohlf 1981. In: Blume (ed.). Biometria. Madrid. 832 p. Widdows. J. 1991. Physiological ecology of mussel larvae. Aquaculture 94:147-163. Zar, J. H. 1984. Biostatistical Analysis. 2nd ed. Prentice-Hall. Englewood Cliffs, NJ. Journal of Shellfish Research, Vol. 16. No. 1, 87-89. 1997. FLUORESCENCE IN SITU HYBRIDIZATION OF VERTEBRATE TELOMERE SEQUENCE TO CHROMOSOME ENDS OF THE PACIFIC OYSTER, CRASSOSTREA GIGAS THUNBERG XIMING GUO AND STANDISH K. ALLEN, JR. Haskin Shellfish Research Laboratory Rutgers University 6959 Miller Avenue Port Norris, New Jersey 08349 ABSTRACT Fluorescence in situ hybridization (FISH) is useful in genomic research. We tested FISH in the Pacific oyster. Crassostrea gigas Thunberg, using metaphase chromosomes prepared from early embryos and all-human telomere and centromere probes. FISH with the all-human telomere probe produced strong hybridization signals at ends of all oyster chromosomes, suggesting that: (1) chromosomes from embryo preparation are suitable for FISH analysis; and (2) the vertebrate telomere sequence, (T,AG 3 )„. may be present in telomeres of the Pacific oyster. No interstitial sites were detected for the telomere sequence. FISH with the all-human centromere probe failed to detect any complementary sequences in oyster chromosomes. KEY WORDS: FISH, chromosome, telomere sequence, gene mapping, mollusc, Crassostrea gigas INTRODUCTION Fluorescence in situ hybridization (FISH) is a powerful tool in genomic analysis. By visualizing hybridization sites of a specific DNA probe. FISH permits the direct mapping of genes or DNA fragments to specific chromosomes and subchromosomal regions. Today. FISH is used in a variety of applications including the characterization and identification of chromosomes, the detection of aneuploidy, the physical mapping of genes and DNA fragments, the determination of linkage order, the detection of chromosomal deletions and arrangements, and comparative genome hybridiza- tion (Kallioniemi et al. 1992, Chowdhary et al. 1995. Matsuda and Chapman 1995, Wang et al. 1995, Pedersen et al. 1996). Despite the active use of FISH in other taxa, there have been few studies on FISH in molluscs. Only one study of FISH has been reported in oysters, where a 166-base-pair (bp) of satellite repeats was localized to two chromosomes (Clabby et al. 1996). FISH in mollusks is generally limited by a lack of reliable protocols and probes, not by a lack of interest. In fact, several important areas of genomic manipulation and analysis in molluscs have been difficult in the absence of FISH technology. One example is aneuploid research in oysters. Many types of aneuploids are viable and can be reliably produced in oysters (Guo et al. 1992a, Guo et al. 1992b, Guo and Allen 1994a). Some aneuploids, such as monosomies and trisomies, are especially useful for the identification and chromo- somal assignment of quantitative trait loci. The major obstacle in the development and use of aneuploid lines has been the inability to identify specific aneuploids, because chromosome identification by traditional banding is time consuming and less reliable in oys- ters than in other taxa. The development of FISH protocols and probes may provide effective methods of chromosome identifica- tion and pave ways for aneuploid research in oysters. One of the challenges for FISH and other chromosomal analy- ses in oysters is the difficulty to consistently obtain metaphase chromosomes. Because cell or tissue culture is not yet possible in oysters, chromosomes have to be prepared from adult tissues or embryos. Adult tissues usually have a low mitotic index and pro- duce highly contracted chromosomes. Although metaphase chro- mosomes can be more reliably obtained from early embryos, one major concern is whether the yolk materials from early embryos, which were inhibitory to trypsin G-banding. would also inhibit FISH. In this study, we tested FISH on Pacific oyster chromo- somes obtained from early embryos, using all-human telomere and centromere probes. MATERIALS AND METHODS Oyster metaphase chromosomes were obtained from 4-h-old embryos cultured at 25°C (Guo et al. 1992a). Eggs and sperm were obtained from mature oyster by stripping gonads. Eggs were passed through a 60-u.m nytex screen to remove large tissue debris and rinsed on a 20-u.m screen. Eggs were resuspended in seawater and fertilized by adding sperm suspension. Excessive sperm were removed at 15 min postfertilization (PF) by rinsing fertilized eggs on a 20-u.m screen. Embryos were resuspended and cultured for 4 h at 25°C. At about 4 h PF. embryos were harvested and treated with 0.005% colchicine for 15 min. After the removal of colchi- cine, nine parts of 0.075 M KC1 were added to every part of embryo suspension in a hypotonic treatment lasting for 8-10 min. Embryos were fixed with 1:3 acetic acid and methanol and stored at 4°C. For slide preparation, three drops of embryo suspension were loaded on each slide and air-dried at 45° angle. When more spreading is desired, slides were flooded with three drops of 1:1 methanol and acetic acid before drying. Slides were aged for 7 days before FISH analysis. FISH was conducted according to a protocol recommended by Oncor. Inc. (Gaithersburg, MD). Slides were pretreated with 2x SSC (pH 7.0) for 30 min at 37°C. dehydrated successively in 70. 80, and 95% ethanol for 2 min each, and air-dried. Denaturation was done by immersing slides in 70°C denaturation solution for 2 min. The denaturation solution consisted of one part of 20x SSC. two parts of distilled water, and seven parts of formamide. Slides were dehydrated in cold ethanol immediately after denaturation. Two digoxigenin-labeled probes were tested on oyster chromo- somes in this study; both were supplied by Oncor, Inc. One was an all-human telomere probe. (T 2 AG,) n , (Cat.# P5097), and the other was an all-human centromere probe (Cat.# P5095). Both probes were labeled with digoxigenin and came in Hybrisol VI (50% formamide. 2x SSC). Probes were denatured by incubation at 70°C for 5 min. Denatured probes were placed on ice until use. For hybridization, 30 mL of probes was placed on each slide, covered with a 22 x 50 mm glass coverslip. and sealed with rubber cement. Slides were incubated at 37°C for 1-2 h in a humidified chamber. After hybridization, slides were washed in 72°C 2x SSC for 5 min 87 88 Guo and Allen and stored in Ix PBD (phosphate-buttered detergent; OncorCat.# SI 370-7). Hybridization was detected with a digoxigenin- fluorescein isothiocyanate detection kit (Oncor Cat.# S1331-DF). Sixty microliters of detection reagent was applied to each slide, covered with a plastic coverslip, and incubated at 37°C for 5 min. Detection reagent was washed three times with 1 x PBD. Slides were counterstained with 18 mL of propidium iodide/antifade. covered with a coverglass, and readied for viewing. Ektachrome color slide film (400 ASA) was used for documentation. RESULTS Hybridization with the all-human telomere probe produced strong signals on termini of all oyster chromosomes (Fig. 1A). Washes at higher stringency (0.5X SSC) did not affect the hybrid- ization of the telomere probe to oyster chromosomes. Hybridiza- tion signals located exclusively at chromosome ends, and no in- terstitial sites were detected. Signals were weak in one or two chromosomal ends, probably as a result of the random variations in hybridization conditions. For some chromosomes, it was noticed that the hybridization signal on one of the sister chromosomes was stronger than the other. FISH with the all-human centromere probe failed to yield any hybridization sites on oyster chromosomes. To assure that the failure was not due to poor probe quality, we tested the all-human centromere probe on human metaphase chromosomes and ob- tained strong hybridization signals at the centromeres of human chromosomes (Fig. IB). DISCUSSION For oysters and many other marine molluscs, early embryos represent the best source for metaphase chromosomes (Guo et al. 1992a, Guo and Allen 1994b). Results with the telomere probe in this study clearly demonstrate that metaphases prepared from early embryos are suitable for FISH analysis. The yolk material, which is problematic for trypsin G-banding, did not inhibit DNA hybrid- ization and detection. The weak signals on a few chromosomes may be caused by random variation in hybridization conditions or by variation in the amount of telomere DNA among telomeres. Telomeres are the terminal protein-DNA structure located at the ends of all eukaryotic chromosomes. They protect linear chro- mosomes from DNA degradation, end-to-end fusion, rearrange- ments, and chromosome loss (Lewin 1994, Zakian 1995). The DNA component of telomeres usually consists of repeats of a simple sequence about 5-10 bp in length. Telomere sequences are highly conserved through evolution. All vertebrates studied so far, as well as the protozoa Trypanosoma and several slime molds and fungi, share the same telomere sequence, (T 2 AG,) n (Zakian 1995). Although telomere sequences in invertebrates are more variable, they share some similarities with each other and the vertebrate sequence (Zakian 1995). The successful hybridization of the all-human telomere probe to termini of oyster chromosomes strongly suggests that the ver- tebrate telomere sequence, (T 2 AG 3 ) n , may be present in the Pacific- oyster. It is possible that hybridization signals seen in this study were due to cross-hybridization of the vertebrate telomere se- quence to a similar, but different oyster telomere sequence. Cross- hybridization is usually eliminated by intensive washing, but washes at higher stringency in this study seemed to have no effect on signal intensity. Therefore, it is likely the vertebrate telomere sequence does exist in the Pacific oysters. Although telomere se- quences are unknown in most marine invertebrates, two marine worms (Polychaeta) have been shown by FISH to contain the vertebrate sequence (An et al. 1995). An insect telomere sequence, (T-.AG-,),,, has been identified in the silkworm (Bombyx won) and many other insects (Okazaki et al. 1993). Bulldog ants (Myrame- cia: Formicidae). however, have both the insect and the vertebrate sequences in their telomere region (Meyne et al. 1995). Cloning and sequencing studies are needed to define the oyster telomere sequence(s). and this study suggests that the human telomere se- quence can be used as a probe to screen genomic libraries for oyster telomere DNA. The failure for the all-human centromere probe to hybridize with oyster chromosomes is understandable. Oncor's all-human centromere probe consists of a selection of chromosome-specific a-satellite sequences — tandem repeats of 17 1 -bp units (Oncor Catalog). Most of these chromosome-specific sequences are de- signed to prevent cross-hybridization with other human chromo- somes or chromosomes from another taxa. Figure 1. FISH of the all-human telomere probe to ends of oyster chromosomes (A) and the all-human centromere probe to human chromosomes (B). FISH in C. gigas 89 ACKNOWLEDGMENTS We thank Dr. Jan Blancato and Mary Williams from Oncor, Inc. (Gaithersburg, MD). for technical assistance. Oncor. Inc., pro- vided probes and laboratory space for this study. This study is supported in part by a grant from the USDA/NRICGP. Publication NJAES D32100-03-97. LITERATURE CITED An. J. H. A.. I. Dominquez, A. S. Balajee, T. H. Hutchinson, D. R. Dixon & A. T. Natarajan. 1995. Localization of a vertebrate telomerric se- quence in the chromosomes of two marine worms (Phylum Annelida. Class Polychaeta). Chromosome Res. 3:507-508. Chowdhary. B. P.. C. Sena. I. Harbitz. L. Eriksson & I. Gustavsson. 1995. FISH on metaphase and interphase chromosomes demonstrates the physical order of the genes for GPI, CRC. and LIPE in pigs. Cytogenet. Cell Genet. 71:175-178. Clabby, C. U. Goswami. F. Flavin. N. P. Wilkins. J. A. Houghton & R. Powell. 1996. Cloning, characterization and chromosomal location of a satellite DNA from the Pacific oyster. Crassostrea gigas. Gene 168: 205-209. Guo. X. & S. K. Allen. Jr. 1994a. Viable tetraploids in the Pacific oyster {Crassostrea gigas Thunberg) produced by inhibiting polar body I in eggs from triploids. Mol Mar. Biol. Biotechnol. 3:42-50. Guo. X. & S. K. Allen. Jr. 1994b. The reproductive potential and genetics of tnploid Pacific oyster, Crassostrea gigas (Thunberg). Biol. Bull. 187:309-318. Guo, X.. K. Cooper, W. K. Hershberger & K. K. Chew. 1992a. Genetic consequences of blocking polar body I with cytochalasm B in fertilized eggs of the Pacific oyster. Crassostrea gigas: I. Ploidy of resultant embryos. Biol. Bull. 183:381-386. Guo. X.. W. K. Hershberger. K. Cooper & K. K. Chew. 1992b. Genetic consequences of blocking polar body I with cytochalasin B in fertilized eggs of the Pacific oyster, Crassostrea gigas: 11. Segregation of chro- mosomes. Biol. Bull. 183:387-393. Kallioniemi, A.. O.-P. Kalhoniemi. D. Sudar. D. Rutovitz. J. W. Gray, F. Waldman & D. Pinkel. 1992. Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 258:818-821. Lewin. B. 1994. Gene V. Oxford University Press. Oxford. England. Matsuda. Y. & V. M. Chapman. 1995. Application of fluorescence in situ hybridization in genome analysis of the mouse. Electrophoresis 16: 261-272. Meyne. J.. H. Hirai & H. T. Imai. 1995. FISH analysis of the telomere sequences of bulldog ants (Myrmecia: Formicidae). Chromosoma 104: 14-18. Okazaki, S., K. Tsuchida, H. Maekawa & H. Fujiwara. 1993. Identification of a pentanucleotide telomere sequence, (TTAGG)„. in the silkworm Bombyx mori and in other insects. Mol. Cell Biol. 13:1424-1432. Pedersen. C. S. K. Rasmussen & I. Linde-Laursen. 1996. Genome and chromosome identification in cultivated barley and related species of the Triticeae (Poaceae) by in situ hybridization with GAA-satellite sequence. Genome 39:93-104. Wang. Y. M.. S. Minoshima & N. Shimizu. 1995. COTI baning of human chromosomes using fluorescence in situ hybridization with CY3 label- ing. Jpn. J. Hum. Genet. 40:243-252. Zakian. V. A. 1995. Telomeres: beginning to understand the end. Science 270:1601-1607. Journal of Shellfish Research, Vol. 16, No. 1. 91-45, 1997. ANNUAL PATTERN OF SETTLEMENT IN POPULATIONS OF CHILEAN OYSTERS TIOSTREA CHILENSIS (PHILIPPE 1845) FROM NORTHERN NEW ZEALAND A. G. JEFFS, 1,3 S. H. HOOKER, 2 AND R. G. CREESE 3 Cawthron Institute Private Bag 2 Nelson, New Zealand 'School of Environmental and Marine Sciences University of Auckland Private Bag 92019 Auckland, New Zealand Leigh Marine Laboratory University of Auckland Private Bag 92019 Auckland, New Zealand ABSTRACT Patterns of larval settlement were examined in two populations of Chilean oysters. Tiostrea chilensis, in northern New Zealand. Artificial settlement surfaces were used to measure larval settlement rates over 36 mo at one site and 20 mo at the other. Larvae settled at both sites throughout the year, but with distinct peaks beginning in late winter and early spring and often continuing into early summer. This pattern was unlike those found previously in T. chilensis populations at higher latitudes. The annual pattern of larval settlement was closely related to the pattern of oyster brooding over the same period. Overall, the results suggest that colder water temperatures during winter (i.e., below 12°C) are important in synchronizing the annual cycle of larval production and settlement in this oyster species. KEY WORDS: Chilean oyster, Tiostrea chilensis, settlement, larvae. New Zealand, flat oyster, Ostreidae INTRODUCTION Reproduction in the Chilean oyster Tiostrea chilensis (Philippi, 1845) is characterized by an extended incubation period culminat- ing in the release of benthopelagic larvae that are capable of set- tling immediately (Hollis 1962. 1963, Millar and Hollis 1963. Cranfield 1968b, Bull 1971, Stead 1971. Westerskov 1980, Jeffs and Creese 1996). Most larvae are thought to settle within minutes of release, although in some populations of oysters, a small pro- portion of the larvae may become planktonic for up to 11 days (Cranfield 1968a. Cranfield 1968b, Stead 1971, Walne 1974. Westerskov 1980. DiSalvo et al. 1983, Cranfield and Michael 1989). Consequently, it could be expected that the annual pattern of larval settlement should closely follow the annual brooding cycle. This has been demonstrated for the short annual period of brooding and settlement in populations of T. chilensis in southern New Zealand (Cranfield and Allen 1977. Westerskov 1980). Previously we described marked differences in the annual pat- tern of brooding for this oyster from two sites in northern New Zealand, leading us to conclude that water temperature was im- portant in regulating the annual pattern of reproduction in this species (Jeffs et al. 1996). In this article, we report on the annual pattern of larval settlement in these two populations in relation to water temperatures. MATERIALS AND METHODS The location of the two study populations of Chilean oysters in the Manukau Harbour and Hauraki Gulf have been described pre- viously (Jeffs et al. 1996. Jeffs et al. 1997). Artificial settlement surfaces were deployed at each site to measure the intensity of oyster spatfall. This technique has been used in previous studies of T. chilensis and for other oyster species (Cole and Knight-Jones 1939. Korringa 1941. Shaw 1967, Cranfield 1968a, Cranfield 1968b, Cranfield 1970, Bull 1971, Hickman 1987). The artificial settlement surfaces consisted of cement-fiber board cut into plates that each measured 310 x 295 x 5mm. Frames were used to hold replicate plates horizontal, with the first plate positioned 40 mm from the seafloor and with subsequent plates spaced at 20 mm intervals above this. Each frame was permanently anchored to the seafloor with buried weight. Replicate frames were 10 m apart and placed within the main oyster bed. Three frames each holding three plates were deployed in the Manukau Harbour, and two frames each with four plates were deployed in the Hauraki Gulf. At monthly intervals, plates were removed from the frames and re- placed with clean plates. All oyster spat, both dead and alive, on each plate were identified and counted with the aid of a stereomi- croscope. Spat that died soon after settlement on the cement-fiber plates persisted because the lower valves of the prodissoconchs were always well cemented to the settlement plate. Therefore, the counts of settled spat could be expected to provide an accurate measure of the total number of spat arriving on the settlement plates for the period they were exposed. The mean number of spat for the sides of all settlement plates was calculated for each monthly sample, and for comparative purposes, the means were standardized as the number of spat settling per day of plate expo- sure for a square meter of settlement surface (i.e., #spat day" 1 m~ 2 ). Sampling began in December 1992 in the Manukau Harbour and in April 1994 in the Hauraki Gulf. For both populations, sampling continued until December 1995. For the duration of the study, water temperatures at each site were measured with handheld thermometers to the nearest 0.1 °C when researchers visited the study sites. The results of some of these measurements were reported previously (Jeffs et al. 1996). In addition, remote temperature-recording equipment (Data- sonde™ and Dataflow™) was deployed at both sites for the final 91 40 35 LU ^ 30 '>. 25 03 T3 CO | 20 B a. c 15 14000 (-> e 03 C/5 12000 0) GO >. O 10000 n 8000 0J 03 £ 03 <*— O 6000 0J -O E 3 Z 4000 03 0J 2000 T~ — i i Jan-93 May-93 Sep-93 Jan-94 May-94 Sep-94 Jan-95 May-95 Sep-95 Date Figure 1. The mean rate of larval settlement for monthly samples in the Manukau Harhour. with corresponding monthly estimates of larval production (from Jeffs et al. 1996a). / • 10000 9000 ■o oj n 8000 E 03 w 7000 0J 0) >. O 6000 h_ Q. OJ 03 5000 03 _l »4— o 1— 4000 -Q E 3000 z c m 0J 2000 i> 1000 ~i i i i i i 1 1 1 r May-94 Jul-94 Sep-94 Nov-94 Jan-95 Mar-95 May-95 Jul-95 Sep-95 Nov-95 Date Figure 2. The mean rate of larval settlement for monthly samples in the Hauraki Gulf, with corresponding monthly estimates of larval production (from Jeffs et al. 1996a). Annual Pattern of Settlement in T. chilensis 93 year of the study. These instruments recorded water temperature to the nearest 0.01 "C at 30 min intervals. Mean monthly temperatures were calculated for handheld thermometer recordings, and mean weekly temperatures were calculated for remote temperature re- cordings. RESULTS Annual Pattern of Larval Settlement At both sites. T. chilensis larvae were settling in every month of sampling. There were distinct peaks of settlement at both sites, although the peaks were much more pronounced in the Manukau Harbour (Fig. 1). Larval settlement peaked in the Manukau Har- bour from September to November of each year (Fig. 1 ). Peaks of larval settlement were found earlier in the Hauraki Gulf, from July to November of each year, with a further smaller peak in March to April 1995 (Fig. 2). The annual pattern of larval settlement at our two study sites corresponded closely with the annual pattern of larval production observed in these populations over the same period (Figs. 1 and 2; Jeffs et al. 1996). This relationship was particularly close for the Manukau Harbour population, albeit with a delay of 1 mo. For example, both the amplitude and the timing of the three annual peaks of settlement corresponded with those observed for larval production. At the Hauraki Gulf population, the relationship between the annual pattern of larval production and larval settlement was not as conspicuous. The peaks of larval settlement were less pronounced than in the Manukau Harbour, as were the peaks of larval produc- tion. Also, the peaks of larval settlement at this site appeared to follow peaks of brooding activity by 2 mo rather than 1 mo (Fig. 2). Annual Pattern of Water Temperatures At both sites, the mean weekly water temperatures for the year of 1994-1995 fluctuated seasonally, with the highest mean tem- peratures for both sites in the summer month of February (Figs. 3 and 4). The lowest mean temperatures were in August for the Hauraki Gulf and during July in the Manukau Harbour. The high- est and lowest temperatures recorded in this year were also during these months (Figs. 3 and 4). Temperatures over the year of 1994— 1995 were much more variable in the Manukau Harbour than in the Hauraki Gulf, as reflected in the greatest recorded temperature changes in 24 h (Figs. 3 and 4). Daily temperature fluctuations in the Manukau Harbour were associated with periods of spring and neap tides and were probably caused by the influx of cooler oce- anic waters and insolation of the harbor's shallow waters. Also, the oyster bed in the Manukau Harbour was exposed for short periods to ambient air temperatures during spring low tides. These periods Mean monthly temp, (hand readings) Mean weekly temp (remote sensor readings) Highest Temp. = 25.18 °C 2/12/95 1500 N.Z.S.T I 24 22 20 O O Q> 1 18 Q. E CD CD a 16 5 CO CD W I 14 12 - 10 I I \ I \ 7 \ \ \ X \ -J I l. \ \ \ I I 4 X X I o 1- r- •st- co ep N N CO i z z ll o o o o CL r^- cd t *"" in CD in o) h- 9? ^ c o CO CO CO ^ O) "~ O CO o ° o to .c CN CO O « t--: i_ ■sf *- £ CN ■<*■ CM I/) 3

o 2 18 Q. E TO I co c 03 0) TS 16 14 12 Mean monthly temp, (hand readings) Mean weekly temp, (remote sensor readings) ^^Highest Temp. 22.91 °C 2/22/95 1930 N.Z.S.T. X I r \~- X M- CO CO V-' N N Z z o o o CO ro E 0) r*. o T_ CM h- U) io (» CD i ■* ■* 01 CM CM en CM CM TO O O C o o O T O O r CM ^— ^f CM CM CM (/) 12°C). there is less synchronicity in spatfall. There is now a need to experimentally verify the precise nature of the relationship between water temperature and reproduction in T. chilensis by attempting to synchronize larval production and spatfall over a range of controlled water temperatures. ACKNOWLEDGMENTS We thank the many people who helped in the field and labo- ratory work for this research, especially Jo Evans. Barbara Hickey from the Auckland Regional Council assisted by providing access to Auckland Regional Council seawater temperature records. This work was funded by Contract 402 with the New Zealand Founda- tion for Science, Research & Technology. LITERATURE CITED Bayne. B. L. 1975. Reproduction of bivalve molluscs under environmen- tal stress, pp. 259-277. In: F. J. Vernberg. (ed.). Physiological Eco- logy of Estuarine Organisms. University of South Carolina Press. Co- lumbia. Bull. M. F. 1971. A preliminary study of the feasibility of oyster farming from rafts in Kenepuru Sound: growth rates, spat settlement and the spawning season. B.Sc.tHons.) Thesis (unpubl.). Victoria University, Wellington. New Zealand. 58 pp. Cole. H. A. & E. W. Knight-Jones. 1939. Some observations and experi- ments on the setting behaviour of larvae of Ostrea edulis. J. Cons. Perm. Int. Explor. Mer. 14:86-105. Cranfield, H. J. 1968a. An unexploited population of oysters. Oslrea lu- laria Hutton, from Foveaux Strait. Part I. Adult stocks and spatfall distribution. N.Z. J. Mar. Freshwater Res. 2:3-22. Cranfield. H. J. 1968b. An unexploited population of oysters. Ostrea lu- taria Hutton, from Foveaux Strait. Part II. Larval settlement patterns and spat mortality. N.Z. J. Mar. Freshwater Res. 2:183-203. Cranfield. H.J. 1970. Some effects of experimental procedure on settle- ment of Ostrea httaria Hutton. N.Z. J. Mar. Freshwater Res. 4:63-69. Cranfield. H. J. 1979. The biology of the oyster. Ostrea lutaria, and the oyster fishery of Foveaux Strait. Rapp. P. -v. Re'un. Cons. Int. Explor. Mer. 175:44-49. Cranfield. H. J. & R. L. Allen. 1977. Fertility and larval production of oysters in an unexploited population of oysters Ostrea httaria Hutton. from Foveaux Strait. N.Z. J. Mar. Freshwater Res. 11:239-253. Cranfield. H. J. & K. P. Michael. 1989. Larvae of the incubatory oyster Tiostrea chilensis (Bivalvia: Ostreidae) in the plankton of central and southern New Zealand. N.Z. J. Mar. Freshwater Res. 23:51-60. DiSalvo. L. H., E. Alarcon & E. Martinez. 1983. Induced spat production from Ostrea chilensis Philippi 1845 in mid-winter. Aquacultnre 30: 357-362. Gleisner. A. 1981. Ciclo reproductivo y desarrollo larval de Ostrea chil- ensis Philippi (Bivalvia. Ostreidae) en el Estuario Quempillen, Chiloe. Tesis, Facultad de Medicina Veterinaria. Universidad Austral de Chile, Valdivia. Chile. Hickman. R. W. 1987. Growth, settlement, and mortality in experimental fanning of dredge oysters in New Zealand waters. N. Z. Fish. Tech. Rep. 1:1-18. Hollis, P. J. 1962. Studies on the New Zealand mud-oyster Ostrea lutaria Hutton, 1873. M.Sc. Thesis (unpubl.). Victoria University. Wellington. New Zealand. 167 pp. Hollis, P.J. 1963. Some studies on the New Zealand oysters. Zoo. Pub. Victoria University. Wellington 31:1-28. Jeffs, A. G. & R. G. Creese. 1996. Overview and bibliography of research on the Chilean oyster Tiostrea chilensis (Philippi. 1845) from New Zealand waters. J. Shellfish Res. 15:305-311. Jeffs, A. G., R. G. Creese & S. H. Hooker. 1996. Annual pattern of brood- ing in populations of Chilean oysters. Tiostrea chilensis, (Philippi. 1845) from northern New Zealand. J. Shellfish Res. 15:617-62. Jeffs, A. G„ R. G. Creese & S. H. Hooker. 1997. The potential for Chilean oysters. Tiostrea chilensis (Philippi. 1845), from two populations in northern New Zealand as a source of larvae for aquaculture. Aquacult. Res. 28:(ln press). Korringa, P. 1941. Experiments and observations on swarming, pelagic life and setting in the European flat oyster, Ostrea edulis L. Arch. Neerl. Zool. 5:1-249. Lepez, M. I. 1983. El cultivo de Ostrea chilensis en la zona central y sur de Chile. Mems. Asoc. Latinoam. Acuiatlt. 5:117-127. Millar. R. H. & P. J. Hollis. 1963. Abbreviated pelagic life of Chilean and New Zealand oysters. Nature 197:512-513. Osorio, R. C. 1979. Moluscos marinos de importancia economica en Chile. Biol. Pesq. Chile 11:3-47. Padilla. M.. M. Mendez & F. Casanova. 1969. Observaciones sobre el comportamiento de la Ostrea chilensis en Apiao. Bol. Inst. Fom. Pesq. Santiago 10:1-28. Price, K. S. & D. Maurer. 1971. Holding and spawning Delaware Bay oysters (Crassostrea virginica) out of season. 11. Temperature require- ments for maturation of gonads. Proc. Natl. Shellfish. Assoc. 61:29-34. Shaw, W. N. 1967. Seasonal fouling and oyster setting on asbestos plates in Broad Creek. Talbot County, Maryland, 1963-65. Chesapeake Sci. 8:228-236. Soli's, I. F. 1967. Observaciones bioldgicas en ostras (Ostrea chilensis Phil- ippi) en Pullinque. Biol. Pesq. Chile 2:51-82. Soli's, I. F. 1973. Valoracion de colectores de larvas de ostras (Ostrea chilensis Philippi) en Pullinque. Biol. Pesq. Chile 6:5-23. Stead. D. H. 1971. Observations on the biology and ecology of the Foveaux Strait dredge oyster (Ostrea lutaria Hutton). N.Z Fish. Tech. Rep. 68:1^19. Walne. P. R. 1974. Culture of Bivalve Molluscs. 50 Years' Experience at Conwy. Fishing News Books Ltd. Farnham, Surrey. England. 189 pp. Westerskov. K. 1980. Aspects of the biology of the dredge oyster Ostrea lutaria Hutton. 1873. Ph.D. Thesis (unpubl.). University of Otago. Dunedin, NZ. 192 pp. Winter, J. E., C. S. Gallardo. J. Araya. J. E. Toro & A. Gleisner. 1983. Estudios en la ostricultura Quempillen, un estuario del sur de Chile. Parte II. La influencia de los factores ambientales sobre el crecimiento y los periodos de reproduction en Ostrea chilensis. Mems. Asoc. Lati- noam. Acuicult. 5:145-159. Winter. J. E., J. E. Toro, J. M. Navarro. G. S. Valenzuela & O. R. Chaparro. 1984. Recent developments, status, and prospects of molluscan aquacul- ture on the Pacific coast of South America. Aquaculture 39:95-134. Journal of Shellfish Research, Vol. 16. No. 1.97-101. 1997. BYSSUS PRODUCTION IN SIX AGE CLASSES OF THE SILVER-LIP PEARL OYSTER, PINCTADA MAXIMA (JAMESON) JOSEPH J. TAYLOR, 1 2 * ROBERT A. ROSE, 1 AND PAUL C. SOUTHGATE 2 1 Pearl Oyster Propagators Pty. Ltd. 4 Daniels Street Ludmilla, N. T. 0820, Australia 2 Aquaculture, School of Biological Sciences James Cook University of North Queensland Townsville, QUI. 4811, Australia ABSTRACT Two experiments were conducted to study byssus production of silver-lip (or gold-lip) pearl oysters. Pinctada maxima, from six different age classes. In the first experiment. 75- or 120-day-old P. maxima were removed from their point of attachment by severing of the byssus and were placed in clear plastic Petri dishes. The production of byssal threads and the behavior of the pearl oysters were monitored over a 120-h period. Emerging byssal threads were pinkish before changing to green. Juveniles at 75 days of age began reattaching faster than 120-day-old juveniles. However, after the first 12 h. older individuals had produced significantly more (p < 0.001) byssal threads than the younger individuals and produced significantly more (p < 0.001) byssal threads over the 120-h period. Additionally, byssal thread production for the younger juveniles did not increase significantly (p > 0.05) after 48 h. whereas byssal thread production from older animals continued to increase significantly (p < 0.001 ) after this period. The maximum number of threads produced by a single individual in the older age class was 30. compared with 16 in the younger age class. Juvenile P. maxima were observed to voluntarily eject the byssal apparatus, move, and reattach within 24 h. Reattachment after voluntary ejection of the byssus was faster than that after mechanical severing. In the second experiment, older P. maxima aged 7. 9, 1 1. or 13 mo were removed from their nets after severing of the byssus with a scalpel. These oysters were placed in nets in an area of either strong (2.5-3.5 knots h" 1 ) or mild (<1 knot h -1 ) current. Pearl oysters placed in a mild current reattached faster than those in a strong current. However, after 4 days, pearl oysters aged 13 mo in strong current had produced significantly more threads (p < 0.05) than those in the mild current, and the same was true for 1 1-mo-old pearl oysters by Day 5. From Day 5 onward, there were generally more threads produced by pearl oysters in strong current compared with mild current; however, these differences were not significant (p > 0.05) for pearl oysters aged 9 and 7 mo. By the end of the 1 1-day experiment, 9-mo-old oysters had produced significantly more byssal threads than any other age class, and there were significant differences between all age classes in the number of threads produced. The results of these simple experiments provide useful information on the time for reattachment of different age classes of P. maxima in a variety of culture conditions after mechanical severing of the byssus. KEY WORDS: Aquaculture, pearl oysters. Pinctada. byssus. attachment INTRODUCTION The foot and byssal gland of the silver-lip (or gold-lip) pearl oyster, Pinctada maxima (Jameson), provide mobility and anchor- age, respectively. The foot, as in all pearl oysters, is a tongue- shaped organ, the bulk of which is a system of multidirectional fibers (Farn 1986). Retractor and levator muscles control foot movement, and extensive blood-filled spaces within the foot pro- vide hydrostatic strength and flexibility (Velayadin and Gandhi 1987). At the proximal end of the foot is the byssal gland, which secretes byssus fibers that pass down a tubular pedal groove (Farn 1986). Muscular contractions of the foot cause the formation of the discoid attachment and stem of each byssal thread. Attachment takes place as the tip of the foot touches the substrate. Byssal secretions harden quickly in seawater. securing the pearl oyster to the substrate (Herdman 1903. Dharmaraj and Alagarswami 1987). Pinctada fucata (Kafuku and Ikenoue 1983), Pinctada marga- ritifera (Nichols 1931), and P. maxima (Saville-Kent 1890. Saville-Kent 1893) juveniles are able to sever their byssal attach- ment, change position, and reattach. P. maxima ceases to use the byssus as a point of anchorage at about 3 y of age. when it is "Correspondence to: Joseph J. Taylor. Aquaculture. School of Biological Sciences. James Cook University of North Queensland. Townsville. Qld. 481 1, Australia. sufficiently heavy to avoid being moved by ocean currents. How- ever, large (3- to 5-kg) wild P. maxima have been found with byssal threads attached to rubble (R.A. Rose, unpubl. data, 1984- 1988). In contrast P. fucata and P. margaritifera maintain byssal attachment as an anchorage system for life (Gervis and Sims 1992). In aquaculture facilities, regular grading increases growout ef- ficiency by separating faster growers from slower growers and removing individuals that are not growing at a profitable rate. This is particularly important in pearl oyster culture because the timing of the implantation of the pearl nucleus depends on the size of the oyster. As with other byssally attached bivalves, grading requires breaking the byssus to remove animals from their point of attach- ment (Bourne et al. 1989. Heasman et al. 1994). Generally, the byssus of P. maxima is severed with a scalpel or razor blade before grading. For commercial rearing of P. maxima, the period required for reestablishment of the byssus is important because of the com- mon practice of using pressurized water for routine cleaning of pearl oysters. If pearl oysters have not reestablished a firm anchor- age, this method of cleaning may prove harmful or even fatal. P. maxima growers recognize weak byssal attachment or de- tachment by juvenile pearl oysters as a sign of ill health (J. Jor- gensen, M Pieper, and S. Arrow, pers comm., 1993-1997). De- tachment, accompanied by other symptoms such as mantle retrac- tion, has been observed before and during mass mortality incidents 97 98 Taylor et al. (losses of up to 75% of the population) in juvenile P. maxima (J.J. Taylor and R.A. Rose, unpubl. data, 1993-1997). Knowing how long it takes for the byssus to regenerate and if the time required varies as pearl oysters grow would therefore aid growout manage- ment and provide valuable data for general health monitoring in commercial operations. To this end, this study determined the time required for reattachment for six age classes of P. maxima juve- niles, after mechanical severing of the byssus. MATERIALS AND METHODS Experiment I Plastic Petri dishes were used as experimental settlement sub- strata. The Petri dishes provided a clear substrate through which attachment was observed and individual byssal threads could be counted over time. Byssus were observed by inverting the Petri dishes under a dissecting microscope. Juveniles of two age classes were used: 75-day-old juveniles, with mean (±SE) dorsoventral shell height and anteroposterior length of 6.7 ± 0.5 and 10.9 ± 1.0 mm. respectively, and 120-day-old juveniles, with mean (±SE) shell height and length of 14.9 ± 1 and 20.7 ± 1.4 mm. respec- tively. Fourteen juveniles of each age class were used for the experiment. At Time 0, juveniles were removed from their point of attachment by severing of the byssus with a scalpel blade. Two juveniles of the same age class were placed in each of 14 Petri dishes. Juveniles were monitored every 30 min for the first 3 h. Juveniles were inspected for the following 6 days, and attached byssal threads were counted. Experiment 2 This experiment was conducted with pearl oysters of 7. 9. II, and 13 mo of age. Before the start of the experiment, the number of byssal threads of 25 randomly selected individuals from each age class was counted before the oysters were removed from their point of attachment. The same animals were also measured and weighed. Oysters of 12 mo of age were held in 8-pocket panel nets (Gervis and Sims 1992), while all other age classes were held in 28-pocket panel nets. Only 10 pearl oysters were placed in each net, and 16 nets were used for each age class. Half of the pearl oysters from each age class were placed in an area of either strong current (2.5-3.5 knots h _1 ) or mild current (<1 knot h -1 ) near the island of Bacan. Maluku Utara, Indonesia (lat. 1°S, long. 127°E). The number of byssal threads produced by each pearl oyster was counted on Days 1 to 7 after the start of the experiment and again on Days 9 and 1 1 . Statistical Analysis Data were compared using analysis of variance (Sokal and Rohlf 1981 ); means were compared using Fisher's Protected Least Significant Difference test from the StatView 1 statistical program, version 4.5, for Macintosh computers (Abacus Concepts, StatView 1992). Homogeneity of variances was confirmed using Cochran's test (Snedcore and Cochran 1967). RESULTS Experiment I Figure 1 shows the number of byssal threads produced by pearl oysters, in each of the two age classes, at intervals during the 120-h experiment. After only 2 h. 6 of the 14 younger individuals (75 days old) were able to hold position when inverted and washed gently with seawater. However, no byssal threads were evident at this time and position appeared to be maintained by the foot alone. Younger individuals showed significantly greater byssal thread production than older individuals ( 120 days old) during the first 3 h of the experiment (p < 0.001; Fig. 1). Within 12 h, all pearl oysters, in both age classes, had formed a byssal attachment. After 12 h. older individuals had produced significantly more byssal threads than younger individuals (p < 0.001: Fig. 1). The total number of threads produced over the 120-h period also differed significantly (p < 0.001); the older pearl oysters produced 21.9 ± 1.1 threads (mean ± SE), and the younger pearl oysters produced 11.3 ± 0.8 threads. Byssal thread production for the younger ju- veniles did not increase significantly (p > 0.05) after 48 h, whereas older individuals produced significantly more threads each day from 24 h onward (p < 0.001 ). New and emerging byssal threads appeared pinkish. Within a few hours, they began to change color, initially becoming trans- lucent before gaining a greenish hue. The color darkened and the threads thickened over time. The point of attachment was splayed (Figs. 2 and 3), and the fibers of the byssal threads were obvious at the point of attachment. On flat surfaces, byssal threads were arranged in a radial pattern (Fig. 2). Where juveniles had moved to the edge of the Petri dish and attached to the dish wall, the threads were attached predominantly in a single direction (Fig. 3). In many instances, byssal threads were ejected from the byssal gland and were observed with one end still attached to the Petri dish and the other end floating free (Fig. 4). In some cases, the entire byssus was jettisoned and the oysters had moved some dis- tance before reattaching. This loss and replacement of byssal threads occurred within 24 h. Experiment 2 The mean (±SE; n = 25) shell length, shell height, and wet weight (WW) and the number of byssal threads (BT) for each age class at the start of the experiment are shown in Figure 5. The 13-mo-old P. maxima had significantly fewer (p < 0.01) byssal threads (8.9 + 0.7) than did any other age class. The number of byssal threads counted from 1 1 -mo-old P. maxima did not differ significantly (p > 0.05) from those from the 9- or 7- mo-old indi- viduals, but the 7-mo-old individuals had significantly fewer bys- sal threads (p < 0.05) than did the 9-mo-old oysters (Fig. 5). Significant differences resulted when the ratios of WW to BT were compared (Fig. 6). The WW/BT ratio for 13-mo-old individuals Time in hours Eigure 1. Byssus thread production over time (mean ± SE; n = 14) in 75- and 120-day-old P. maxima juveniles. Byssal Attachment of Pearl Oysters 99 Figure 2. Byssal threads of a juvenile P. maxima attached to the flat surface of a Petri dish and arranged in a radial pattern. This juvenile attached to the flat surface in the center of a Petri dish. Note: H. the hinge of the oyster; BN, the byssal notch: S, the splayed end of the byssal threads at the point of attachment. Figure 4. A juvenile /'. maxima (far right) that has detached, changed position, and reattached, leaving ejected byssal threads behind. Note: H, the hinge of the oyster; N, new byssal threads; E. ejected byssal threads that are still attached to the surface of a Petri dish. was significantly greater (P < 0.01) than that for any other age class. The WW/BT ratio became significantly less {P < 0.01 ) with each age class, with the exception of the 9- and 7-mo-old pearl oysters, where the WW/BT ratio did not differ significantly (p > 0.05). After mechanical severing of the byssal threads, differences in byssus production were noted due to both age and current strength. Regardless of age. significantly more byssal threads were pro- duced by oysters in the mild current area during the first day of the experiment (Table 1 ). This was also true after Day 2 for all but the 13-mo-old oysters. From Day 5 onward, there were generally more threads produced by pearl oysters in strong current compared with mild current; however, these differences were not significant (P > 0.05) for pearl oysters aged 9 and 7 mo. By the end of the experi- ment, pearl oysters aged 9 mo had produced significantly more threads (/> < 0.01) than any other age class and differences in the number of threads produced were significant (p < 0.0 1 ) between all age classes. A number of individuals from each age class ejected the origi- nal byssal plug from the shell cavity. After 1 1 days, no oyster in any of the age classes had produced the number of threads that Figure 3. Byssal threads of a juvenile P. maxima attached near to the wall of a Petri dish with the threads predominantly in a single direc- tion. Note: H, the hinge of the oyster; BN. the byssal notch; S, the splayed end of the byssal thread at the point of attachment; W, the wall of the Petri dish. were counted at the start of the experiment and some of the older individuals did not reattach at all. DISCUSSION Juveniles of 75 days of age began reattaching faster than 1 20- day-old juveniles. However, after the first 12 h, older P. maxima produced significantly more threads than the younger individuals and significantly more threads over the 120-h experiment. More- over, byssal thread production for the younger juveniles did not increase significantly after 48 h, whereas production from older animals continued to increase significantly. This suggests that younger pearl oysters regain maximal anchorage after a shorter period than older pearl oysters. The maximum number of threads produced by a single individual in the older age class was 30 compared with 16 in the younger age class. A stronger anchorage may have been required by the older individuals to compensate for greater resistance to water currents due to larger surface area. Saville-Kent (1890, 1893) reported that juvenile P. maxima of a size range between 8 and 65 mm had 30—40 byssal threads. Rose and Baker ( 1994) reported the average number of byssal threads in 10- to 15-mm P. maxima juveniles to be approximately 20: neither study reported differences in byssal production with age or be- tween size classes. Juvenile P. maxima have the ability to sever the byssus, move, and reattach (Saville-Kent 1890, 1893). This behavior was ob- served in this study with juveniles moving and reattaching with the same or a greater number of threads within a 24-h period. One 100 80 ■J 60 | 401 20 D i.: □ G3 Q 1,1 SH WW Figure 5. Mean (±SEl shell height (SHl, shell length (SL), wet weight (WW), and number of byssal threads (BT) of four age classes of P. maxima. Gl, 13 mo old; G2, 11 mo old; G3, 9 mo old; G4. 7 mo old. Values with the same superscript for each variable are not signifi- cantly different (p > 0.05). 100 Taylor et al. CO 14 12 10 Gl G2 G3 G4 Age Class Figure 6. Mean (±SE) ratio of wet weight (WW) to number of byssal threads (BT) of four age classes of P. maxima. Gl, 13 mo old; G2, 11 mo old; G3, 9 mo old; G4. 7 mo old. Values with the same superscript for each variable are not significantly different (p > 0.05). younger individual moved twice within a 24-h period and pro- duced a total of 20 new byssal threads. This suggests that when a pearl oyster voluntarily ejects the byssus. it can reattach more rapidly than after mechanical severing. P. maxima juveniles can also eject individual threads, which will allow minor positional changes, perhaps to adjust to water flow without losing the security of the entire byssus. The number of byssal threads counted at the start of the second experiment shows the reduction in byssal thread production as P. maxima ages. Presumably, a point is reached where the increased water resistance, due to greater surface area, is offset by the greater stability resulting from increased mass. Eventually, the weight of the pearl oyster is such that byssal attachment loses importance as the main means of maintaining position. The large differences in the ratio of wet weight to the number of byssal threads supports this notion. Even in very large specimens of P. maxima (>1 kg WW), the byssus may still be observed, even though there is no attachment to substrata and therefore no byssal anchorage (R.A. Rose, unpublished data 1984-1988). In the second experiment. P. maxima initially reattached with a greater number of threads when placed in relatively calm water with little current. It appears that stronger current made initial reattachment more difficult. This may be because the net holding the oysters was less stable under these conditions. After initial attachment, greater numbers of threads were produced by pearl oysters in the strong current, indicating that additional threads were required to secure the pearl oysters in the nets. This was particularly the case for the oldest and largest individuals, which in some cases, failed to reattach during the 1 1-day experiment. The results of this study indicate that byssus production in P. maxima TABLE 1. Mean (±SE) byssal thread production of four age classes of silver-lip pearl oyster, P. maxima, placed in an area of either strong current (SO or mild current (MC). Day Current Gl G2 G3 G4 1 SC 0.2 ±0.1"' 0.5±0.1 bl ii ±0.1" 1.3 ±0.1" MC o.5±o.i a - 0.9±.01 ba 1.8±0.1 c - 2.1 ±o.r : 2 SC 1.(1 ±0.2''' 1.1 ±0.2 hl 2.6±0.2 C| 2.9 ±0.2" MC 1.1 ±0.2" 1.9 ± 0.2 b2 3.3 ± 0.2 c2 3.3 ± 0.2 c2 3 SC 1.8 ± 0.2 J| 2.2 ±0.2 bl 4.3 ± 0.3 C| 4.5 ±0.2" MC I.5 + 0.2" 1 2.4±0.2 bl 4.1 ±0.3" 4.5 ±0.2" 4 SC 2.7 ±0.3''' 3.1 ±0.3 M 5.6 ±0.3" 5.7 ±0.2" MC 1 .9 ± 0.3" 2 3.1 ±0.3 bl 5.4 ±0.3" 6.0 ±0.3" 5 SC 3.8 ± 0.4" 4.9±0.3 hl 7.1 ±0.4" 7.1 ±0.3" MC 2.5 ± 0.4 s2 3.6 ± 0.3 b - 7.2 ±0.4" 6.8 ±0.3" 6 SC 4.8 ±0.4-" 5.6 ±0.4"' 8.7 ±0.4" 8.6 ±0.4" MC 3.0 ± 0.4" : 4.5 ± 0.3 b: 8.2 ±0.4" 8.1 ±0.3" 7 SC 5.0 ±0.4" 6.4±0.4 hl 9.7 ±0.4" 8.7 ± 0.3 d ' MC 3.5 ± 0.4"- 5.9±0.3 bl 9.6 + 0.4" 8.6±0.4 dl 9 SC 7.0 ±0.4"' 8.6 ± 0.4 bl 11.9 ±0.5" 9.8±0.4 dl MC 5.0 ± 0.4 a2 7.4 ± 0.4 b: 11.2 ±0.5" 10 1 ±0.4 dl 11 SC 7.8 ±0.4 al 9.8±0.5 M 13.0 ±0.6" ll.()±0.4 dl MC 6.2 ± 0.5 a2 9.0 ± 0.4 b2 12.0 ±0.5" 10.1 ±0.4 dl Gl, pearl oysters aged 13 mo; G2, pearl oysters aged 11 mo; G3. pearl oysters aged 9 mo and G4, pearl oysters aged 7 mo. Means with the same superscript (alphabetical across rows, numerical down columns) are not significantly different (p > 0.05). may be the result of a subtle relationship between the stability of the substratum, the resistance of a pearl oyster to a given current, and the size and weight of the individual. In this experiment, the pearl oysters were maintained under typical farm conditions — vertically orientated in panel nets. The results may have differed had the pearl oysters been placed flat on the seabed, where resis- tance to the current would have been reduced. The results of these simple experiments provide useful infor- mation on the time required for reattachment after mechanical severing of the byssus of P. maxima. Where possible, it is advised that newly graded pearl oysters, or oysters that have been trans- ferred to new nets, should be placed in areas with calm water for a minimum of 24 h to allow a reasonable degree of reattachment before moving them into areas of higher current or wave action. ACKNOWLEDGMENTS This study was conducted at the Darwin Hatchery Project pearl oyster hatchery and at the KRI pearl project in Indonesia, both operated by Pearl Oyster Propagators Pty. Ltd. Thanks are due to Nurhayati and Hasbuana for their technical assistance. LITERATURE CITED Bourne. N. F.. C. A. Hodgson & J. N. C. Whyte. 1989. A Manual for Scallop Culture in British Columbia. Canadian Technical Report of Fisheries and Aquatic Sciences No. 1694. 215 pp. Dharmaraj, S. K. & K. Alagarswami. 1987. Some aspects of physiology of Indian pearl oysters, pp. 21-28. In: K. Alagarswami (ed.). Pearl Cul- ture. Bulletin of the Central Marine Fisheries Research Institute, No. 39. Central Marine Fisheries Research Institute. Cochin. India. Farn, A. E. 1986. Pearls Natural. Cultured and Imitation. Butterworth Gem Books, London. 1 50 pp. Gervis. M. H. & N. A. Sims. 1992. The Biology and Culture of Pearl Oysters (Bivalvia: Pteriidae). ICLARM Stud. Rev. 21. ODA, London. 49 pp. Heasman. M. P., W. A. O'Connor & A. W. J. Frazer. 1994. Detachment of the commercial scallop. Pecten fumatas, spat from settlement sub- strates. Aquaculture 123:401— 107. Herdman, W. A. 1903. Report to the Government of Ceylon on the Pearl Oyster Fisheries of the Gulf of Manaar. Part 1 . The Royal Society of London. London. 146 pp. Kafuku, T. & H. Ikenoue. 1983. Pearl oyster (Pinctadafucata). pp. 161- 171. In: Modem Methods of Aquaculture in Japan. Development in Aquaculture and Fisheries Science. Elsevier Scientific Publishing Co.. Amsterdam. Byssal Attachment of Pearl Oysters 101 Nichols. A. G. 1931. On the breeding and growth rate of the black-lip pearl oyster 'Pinctada margaritifera). Rept. Gt. Barrier ReefComm. 3:26- 31. Rose. R. A. & S. B. Baker. 1994. Larval and juveniles culture of the Western Australian silver- or goldlip pearl oyster. Pinctada maxima Jameson (Mollusca: Pteriidae). Aquaculture 126:35-50. Saville-Kent. W. 1890. On the experimental cultivation of the mother-of- pearl shell Meleagrina margaritifera in Queensland. Rep. An.sr. Assoc. Adv. Sci. 2:541-548. Saville-Kent. W. 1893. Pear! and pearl-shell fisheries, pp. 1075-1078. In: The Great Barrier Reef of Australia: Its Products and Potentialities. W. H. Allen and Co.. London. Snedcore. G. W. & W. G. Cochran. 1967. Statistical Methods. 6th ed. University of Iowa Press. Ames. IA. 593 pp. Sokal, R. R. & F. J. Rohlf. 1981. Biometry. Freeman. New York. 859 pp. Velayadin. T. S. & A. D. Gandhi. 1987. Morphology and anatomy of In- dian pearl oyster, pp. 4-12. In: K. Alagarswami (ed.). Pearl Culture. Bulletin of the Central Marine Fisheries Research Institute. No. 39. Central Marine Fisheries Research Institute. Cochin. India. Journal of Shellfish Research. Vol. 16. No. ]. 103-110. 1997. BREEDING CYCLE OF PEARL OYSTERS Pinctada mazatlanica AND Pteria sterna (BIVALVIA:PTERIIDAE) AT BAHIA DE LA PAZ, BAJA CALIFORNIA SUR, MEXICO PEDRO SAUCEDO AND MARIO MONTEFORTE Centra de Investigaciones Biologicas del Noroeste, S.C. P.O. Box. 12S La Paz. B.B.S.. Mexico ABSTRACT The breeding cycles of pearl oysters Pinctada mazatlanica and Pteria sterna were studied from June 1992 to May 1993 as part of a Pearl Culture Program in Bahfa de La Paz. Gonad samples of 20 oysters of each species were collected monthly (480 over the annual cycle) and processed for histological examination. We studied the annual breeding cycle of both species, the sex ratio as a function of time, and the size of the oysters. The results obtained by histological analysis were confirmed by similar changes in a gonadosomatic index. Gametogenesis was continuous throughout the year in both species. P. mazatlanica spawned once a year (September), when water temperature reached 29.5°C. It is a protandrous hermaphrodite in which sex reversal was observed in oysters larger than 100-mm shell height. The female:male sex ratio was 0.12:1. Gonad maturity was found in oysters larger than 39 mm. P. sterna spawned twice a year (February and May), when water temperature was 22.2 and 23.4°C. There was not enough evidence to conclude that P. sterna was a protandrous hermaphrodite. If that were the case, sex reversal would have occurred in oysters larger than 50-mm shell height. The female:male sex ratio was 0.38:1. Gonad maturity was seen in oysters larger than 40 mm. KEY WORDS: Pearl oysters, breeding cycle, reproduction, repopulation, Bahfa de La Paz INTRODUCTION In Mexico, natural populations of the native species Pinctada mazatlanica (Hanley, 1856) and Pteria sterna (Gould. 1851) are now in a critical situation because of overexploitation. The uncon- trolled pearl fishery carried out in Bahfa de La Paz for more than 400 years depleted the natural stocks along the coast almost en- tirely by 1940 (Sevilla 1969. Shirai and Sano 1979. Carino 1987. Carino and Caceres-Martinez 1990. Monteforte 1990. Monteforte 1991. Monteforte and Carino 1992, Carino and Monteforte 1995). At present, both species are under a legal ban decreed on the pearl oyster fishery (Diario Oficial de la Federacion 1939) and are con- sidered "species under special protection" (Diario Oficial de la Federacion 1994). However, illegal extractions have continued, impeding the natural recovery of broodstock. The presence of pearl oysters in the Baja California Peninsula has played an important role in the social and economic develop- ment of the region, mainly in Bahfa de La Paz. Therefore, the urgent need to apply strategies for conservation, extensive culture, repopulation, and recovery of the nacre resource has been empha- sized on several occasions (Sevilla 1969, Dtaz-Garces 1972. Mar- tinez 1983, Monteforte 1990. Monteforte 1991, Monteforte and Carino 1992, Saucedo and Monteforte 1994, Saucedo et al. 1994). The success of aquaculture of pearl oysters requires a proper knowledge of the biology and ecology of the species. To under- stand the population dynamics of the wild stock and, more re- cently, the development of pearl-culture strategies, it is essential to understand the reproductive biology of pearl oysters (Tranter 1958a. Sevilla 1969. Rose et al. 1990, Garcfa-Dominguez et al. 1996). There have been a number of studies of the reproductive biol- ogy of the genus Pinctada. These studies reveal that most aspects of the breeding cycle of pearl oysters are common to all species. They seem to be functional protandrous hermaphrodites (maturing as males and changing to females at a certain size, regulated by Correspondence to: Pedro Saucedo. Centra de Investigaciones Biologicas del Noroeste. S.C. Division de Biologfa Marina. P.O. Box. 128. La Paz. Baja California Sur, Mexico. internal and external processes). The ratio of males to females tends toward 1:1 with increasing age (Gervis and Sims 1992). However, little is known about the reproductive biology of P. mazatlanica and P. sterna. Before this study, the only data avail- able were those of Sevilla ( 1969), who made a precise microscopic description of the gonad anatomy of P. mazatlanica, pointing out each phase of the breeding cycle and its seasonal occurrence. No study has been done on P. sterna. In 1987, we started an applied research program on pearl oyster culture and pearl induction at the Centro de Investigaciones Bio- logicas del Noroeste, in Bahfa de La Paz. Parallel to technological development aiming at production through extensive culture and induction to pearl formation, we are also searching for recovery and conservation strategies. The objective of this study is to de- scribe the annual breeding cycle of the pearl oysters P. mazat- lanica and P. sterna, obtained from extensive culture and kept under bottom-culture conditions in Bahfa de La Paz. MATERIALS AND METHODS Oysters used in this study were collected in 1991 at Isla Gavi- ota and reared by extensive culture at Caleta El Merito, located on the southwest coast of Bahia de La Paz, between 24°46' and 24°07'N, and 1 10°38' and 110°18'W (Fig. 1). The selection of Caleta El Merito as the study area was made because of its climatic, geomorphologic. and oceanographic con- ditions, which were adequate for the development of the study. A more detailed description of the area is provided in Alvarez- Borrego and Schwartzlose (1979), Osuna-Valdez (1986). Murillo ( 1987), and Monteforte and Carino (1992). In April 1992. 480 oysters (240 of each species) were trans- ferred to bottom-culture conditions, using plastic pearl cages (70- cm length, 40-cm width, and 20-cm height) placed on a submerged shelf at 10-m depth. Sixteen cages were placed on the bottom of the study area (eight per species), each one containing 30 oysters. The initial size range varied from 39.5 to 136.5-mm shell height for P. mazatlanica (mean, 80.55; SD, 20.25) and from 41 . 1 to 89.2 mm for P. sterna (mean. 66.12: SD. 12.30). Twenty oysters of each species were collected monthly using 103 104 Saucedo and Monteforte Figure 1. Location of the study area (Caleta El Merito) inside Bahia de La Paz, indicated bv closed diamond. SCUBA gear, and they were preserved in 10<7<- formalin for 48 h. Before dissection, the following shell measurements were taken with plastic calipers (±0.01 mm) according to Hynd's expressions ( 1955): height or dorsoventral measurement, length or anteropos- terior measurement, thickness, wet weight of the oyster with shell, wet weight without shell, and wet weight of the visceral mass in which the gonadal tissue is intermingled. This latter sample, al- ways excised between the labial palps, near the foot, and the in- testine tube, was processed for histological examination. Samples were embedded in paraffin, sectioned at 7 or 8 p.m. and stained by the hemotoxylin-eosin technique. They were analyzed with a com- pound microscope at low magnifications (lOx and 40x) and were photographed through the microscope. To analyze the breeding cycle of both species, and especially to understand the seasonal changes occurring in the gonads, we used five broad gametogenic stages, using the schemes developed by Sevilla (1969) for P. mazatlanica, and Rose et al. (1990) for P. maxima. The stages are: (1) indeterminate or inactive, (2) devel- oping or near-ripe, (3) maturity or ripe. (4) spawning, and (5) spent. We also calculated the total female:male sex ratio of both spe- cies, the sex ratio as a function of time, and the sex ratio related to the size of the oysters. Shell height was used as the most adequate indicator of growth. It is considered the largest dimension of the oyster measured at right angles to the hinge line, excluding the growth processes (Hynd 1955). At the same time as histological analysis was done, a gonado- somatic index (GI) was calculated with the oyster's measurements originally taken, using the equation proposed by Sastry (1970): GI = GW/ WWS x 100 This is obtained by dividing the gonad weight of the animal (GW) by its wet weight without shell (WWS). multiplied by 100. Finally, the relationship between the GI and the monthly changes in (he water temperature during the annual cycle was also studied. RESULTS Gonad Developmental Stages Gametogenesis was found to be a continuous process through- out the annual cycle in both species. However, many of the stages of the breeding cycle overlapped in time within the same gonad, so their classification into any gametogenic stage was sometimes dif- ficult to determine. The most important microscopic characteristics of the gonad anatomy are described as follows: Indeterminate or Inactive There is no evidence of gonad development. Instead, the gonad consists mainly of connective tissue. Follicles are completely empty and may contain some phagocytes. Gonads are not able to be classified as to sex (Fig. 2A). Developing or Near-Ripe The production of gametes begins. At first, follicles are small and poorly developed. Oogonia in the ovary and spermatogonia in the testis are mainly connected to the follicular wall (Figs. 2B and 3A). As gametogenesis proceeds, different stages of gametes can be observed. In the testis, primary and secondary spermatocytes proliferate rapidly. In the ovary, connected and some free oocytes with little or no yolk expand into the lumen (Figs. 2C and 3A). At the final stages of gametogenesis. spermatids and some spermato- zoa, or free oocytes with yolk and nucleolus, are common in the follicles. During this stage, the amount of connective tissue rapidly decreases and almost disappears. Maturity or Ripe The gonad has grown and enlarged as a compact and uniform mass, in which the individual follicles are distended and hard to distinguish. Connective tissue has been reduced to a small and thin layer in the distal regions of the gonad. The follicular lumen is filled mainly with polygonal-shaped free oocytes with yolk and nucleolus (in the ovary) or with spermatozoa clearly defined by their eosinophilic tails (in the testis). Some isolated pockets of developing oogonia or spermatids can be observed (Figs. 2D and 3C). Spawning This phase is easy to detect because of the expulsion of ga- metes. Follicles are broken, distended, and partially empty. The lumen is filled with residual free oocytes or thin spermatozoa, showing signs of regression (Figs. 2E and 3D). Spent Follicles have become extremely thin, and the lumen is prac- tically empty, with some isolated pockets of residual oocytes or spermatozoa. This phase is characterized by the rapid proliferation Breeding Cycle of Pearl Oysters 105 0>' ' Figure 2. Sexual phases of male gametogenesis in P. mazatlanica and P. sterna. (A) Indeterminate phase in /'. sterna, showing empty follicles with some phagocytes (ph); (B) early gametogenesis in P. sterna, in which small follicles (fo( contain spermatogonia (sg) connected to the follicular wall, primary and secondary spermatocytes (sc) expanding toward the lumen, and some spermatozoa (sp) filling the center; (C) adyanced gametogenesis in P. mazatlanica with distended follicles containing large amounts of spermatozoa; (D) maturity stage in P. mazatlanica, characterized by the presence of follicles packed ytith spermatozoa almost exclusively; (E) spawning in P. mazatlanica with broken and partially empty follicles containing residual spermatozoa Irs), and the presence of different kinds of phagocytes: (K) spent stage in P. sterna, in which emptj and collapsed follicles contain high phagocytic activity; connective tissue is deyeloping again. Scale bar. 25 pm. 106 Saucedo and Monteforte Figure 3. Sexual phases of female gametogenesis in P. mazatianica and P. sterna. (A) Early gametogenesis in P. mazatianica showing poorly developed follicles (fo) with small oogonia (og) connected to the follic- ular wall, lacking yolk and nucleolus, and auxiliary cells (ac); (B) advanced gametogenesis in P. mazatianica, in which immature pe- duncle-shaped oocytes (po) are still present in the follicle, together with polygonal-shaped free oocytes (ocl with yolk and nucleolus; (C) ma- turity in P. sterna, containing mainly free-shaped oocytes filling the lumen: ID) spawning in /'. sterna, in which follicles are thin and col- lapsed, with residual oocytes (rol; (E) spent stage in P. mazatianica, with residual oocytes showing signs of regression. Scale har, 25 uni. of different kinds of phagocytes surrounding the gametes. The oysters were spent in October and November. Gonad development connective tissue has started to develop again (Figs. 2F and 3E). continued in November and lasted until May. Mature oysters were seen from February to May. P. mazatianica Breeding Cycle The breeding cycle of P. mazatianica is shown in Figure 4. In June and July, oysters at different developmental stages are com- mon; a large number of them have started gonad development, others have reached maturity, and another group was found to be spent. Spawning took place in September and October, and most Sex Ratio This was completely skewed to the male sex. From the total sample analyzed, 77% were male, 9% female, and the last 13% were indeterminate. The female:male sex ratio was 0.12:1. The sex ratio related to oyster size revealed that P. mazatianica matured as male and tended to be a protandrous hermaphrodite. Breeding Cycle of Pearl Oysters 107 £ BO- > ZO z 111 o UJ u. Ill 40- > 1 Ul a 20 JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY TIME (months) £77j INDETERMINATE gg] DEVELOPING ~J MATURITY ^B SPAWNING y/y SPENT Figure 4. Sexual gametogenic stages in P. mazatlanica during an an- nual cycle. / 100-^ 90-^ (/ ^7 (tit '*' 1 —. ao- g 7 J? >- 70- o / W 60 3 o H j DC u. UJ 40- 11 < 30- Ul a 20 _ fit rFYV I 1 vT v 10- \ ) L P.ii ' FEMALE . /(J/ INDETERM JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY TIME (months) Figure 6. Temporal sex ratio in P. mazatlanica. This is because males were seen from 40 to 150-mm shell height. Females, on the other hand, never appeared under 100 mm. Inde- terminate oysters were observed from 40 to 80 mm (Fig. 5). Figure 6 shows the percentage of females and males as a func- tion of time. Once again, males were present during the entire annual cycle, but with a higher incidence in January, February, and April, months in which gonads were in active development or ripe. Females appeared only from June to August and from February to May, again, the months with greatest reproductive activity. Gl This index was found to be a good indicator of the reproductive activity of the animals because it revealed a close relationship with the reproductive activity of the oysters (Fig. 7). The highest values of the index denoted an increase in the reproductive activity, and gonads were found in active development or mature (July 1992 and from January to May 1993). The lowest value, in September, coincided with the spawning. Figure 7 also shows the relationship between the Gl and the water temperature. Once again, in Sep- tember, when the water temperature was 29.5°C. spawning oc- 39-49 59-69 79-89 99-109 119-129 139-149 49-59 69-79 89-99 109-119 129-139 SHELL HEIGHT (mm) curred. After December, increasing water temperature resulted in gradual increases in the values of the Gl. P. sterna Breeding Cycle The breeding cycle of P. sterna is shown in Figure 8. Some gonad development can be detected in June. In July, oysters were spent, although the largest part of the sample was found to be indeterminate. This stage was present almost all year. Gametogen- esis was seen continuously from August 1992 to May 1993. Spawning took place in February. We detected a new, short breed- ing cycle, including gametogenesis. maturity, and a second spawn- ing in May. Sex Ratio Once again, the sex ratio was skewed to the male sex. From the total sample. 48% were male, 19% were female. 0.6% were her- maphrodite, and 32% were indeterminate. The female:male sex ratio was 0.38:1. 7 20 JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY TIME (months) r -» GONADOSOMATIC INDEX TEMPERATURE Figure 5. Size-related sex ratio in P. mazatlanica. Figure 7. Relationship between the Gl and the water temperature for P. mazatlanica. 108 Saucedo and Monteforte 70- 60- 50 40- 30- 20- 10- JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR HAY TIME (months) J INDETERMINATE £2 DEVELOPING J MATURITY Hi SPAWNING '//, SPENT Figure 8. Sexual gametogenic stages in P. sterna during an annual cycle. Figure 9 shows the percentage of females and males. The size- range analysis suggests that P. sterna can be a protandrous her- maphrodite. Males were present from 40 to 85-mm shell height. with a higher incidence between 40 and 55 mm. Females appeared after 50-55 mm but were not represented in all of the size ranges. Their maximum number was observed between 60 and 65 mm and 85 and 90 mm. Indeterminate oysters were present in all of the size ranges (Fig. 9). Figure 10 shows the sex ratio as a function of time. Once again, the male sex was present over the entire annual cycle, but with a higher incidence in January, February, and April during the spawn- ings. Females appeared from August on and were present the rest of the annual cycle. This behavior, unlike that of P. mazatlanica, could indicate that this species is a multispawner. GI The index seems to describe adequately the reproductive ac- tivity of the oysters. The relationship between the index and the reproductive activity of the oysters was relatively close (Fig. 1 1 ). The highest peaks in the values of the index, recorded in December and April, indicated the gonads to be in active development or sexually ripe. The lowest values of the index, detected in March MALE FEMALE NDETERM 40-45 5D-55 60-65 70-75 80-85 45-50 55-60 65-70 75-60 85-90 SHELL HEIGHT (mm) JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY TIME (months) Figure 10. Temporal sex ratio in P. sterna. and May, coincided with the two spawings observed in the year. The GI and the water temperature had an inverse relationship. DISCUSSION As suggested by Giese and Pearse (1974). pearl oysters from temperate regions generally exhibit discrete and regular breeding seasons. Evidence found in this study indicated that P. mazatlanica and P. sterna followed a clear annual breeding cycle. Gametogen- esis was found to be a continuous process throughout the annual cycle in both species. Changes observed in the reproductive activ- ity of the oysters during the annual cycle were regulated mostly by seasonal changes in the water temperature. Earlier work done on pearl oyster reproduction from different parts of the world con- firms Orton's rule: "if temperature conditions are constant or nearly so and the biological conditions do not vary much, animals will breed continuously" (Orton 1929 in Chellam 1987). Tranter (1958b-d) observed a definite annual reproductive cycle for Pinctada albina and Pinctada margaritifera from the Torres Strait, Queensland. Sevilla (1969) also found a breeding cycle with continuous gametogenesis in P. mazatlanica from Ba- hia de La Paz. Mexico. Rose et al. ( 1990) noted the same pattern of gametogeneis for Pinctada maxima from Eighty-Mile Beach, Western Australia, and Garcfa-Domi'nguez et al. (1996) made JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY TIME (months) GONADOSOMATIC INDEX TEMPERATURE Figure 9. Size-related sex ratio in P. sterna. Figure 11. Relationship between the (il and the water temperature for P. sterna. Breeding Cycle of Pearl Oysters 109 similar observations for wild P. mazatlanica from Isla Espfritu Santo. Baja California Sur. Mexico. However, some differences in the overall pattern of gametogen- esis of pearl oysters can be detected. P. albina breeds annually and spawns once between April and May (Tranter 1958c), whereas P. maxima breeds annually, but spawns twice, from October to De- cember and from February to April (Rose et al. 19901. P. marga- ritifera also has a bimodal spawning pattern, from March to Au- gust and from September to February (Tranter 1958d). A smaller species. Pinctada fucata, has two spawning peaks, during June to September and December to February, at Tuticorin Harbour (Chel- lam 1987). The tendency for a bimodal spawning pattern has been described for several marine invertebrates (Giese 1959, Giese and Pearse 1974). During our study, P. mazatlanica bred annually and spawned once, from September to October, when the water temperature rose to 29-30°C. Histological evidence indicates a second short spawn- ing could have occurred in June or July, triggered again by changes in water temperature. Unfortunately, the monthly sampling used for collecting the gonads did not allow us to detect this. P. sterna also bred annually, but spawned twice during the annual cycle. The maximum peak was from February to March, when the water temperature decreased to 24°C. The lesser peak was from April to May. when the water temperature dropped to 22.8°C. Once again, the histological evidence suggests the possibility of a third spawn- ing, during June or July. We believe that P. sterna is potentially capable of spawning throughout the year, because mature gonads of both sexes were present almost all year. Rose et al. ( 1990) found mature oysters outside the main breeding period, suggesting that P. maxima is also capable of spawning all year. The histological gonad analysis revealed that spawning in P. mazatlanica and P. sterna was not complete during the breeding season, and a large number of residual gametes were present after the spawning at the spent stage. We found phagocytic activity in oysters of both species that had recently spawned, indicating the presence of gonad regression. Tranter (1958c) and Rose et al. (1990) detected incomplete spawning and gonad regression in P. albina and P. maxima. However, regression in P. sterna was in- complete after the first massive spawning in February. We suggest the possibility of a second short breeding cycle in which animals avoid the spent stage and pass directly to gametogenesis after the first spawning. Similar observations were made by Sevilla ( 1969) for P. mazatlanica. Pearl oysters seem to be functional protandrous hermaphro- dites, with sexes separated by time. However, bisexual phases may occur in the same gonad, although they appear to be transitional and nonfunctional, as suggested by Rose et al. (1990) for P. maxima. Two cases of hermaphroditism were detected in P. sterna. both after 60-mm shell height. Similarly, Garcfa-Domfnguez et al. (1990) found two hermaphrodite specimens in wild P mazat- lanica. In this study, we managed young oysters instead of adults, especially in P. mazatlanica. a larger species. The total sex ratio was completely skewed to the male sex in both species, and there- fore, they behaved as protondrous hermaphrodites. Particularly in P. mazatlanica. the female:male sex ratio of 0.12:1 and the mean sex-reversal size detected at 100 mm confirmed this. No females were observed under 100 mm. Garcfa-Domfnguez et al. (1996) found a different female:male sex ratio of 1.33:1 in P. mazatlanica from Isla Espfritu Santo. However, although females outnumbered males, the study was carried out with larger individuals (adults). ranging from 72 to 176-mm shell height. A sex ratio of 1:1 was observed for P. maxima from Eighty-Mile Beach. Western Aus- tralia, at 200-mm shell height (Rose et al. 1990). For P. sterna, there was not enough evidence to conclude that the species were protandrous hermaphrodites, although the female: male sex ratio was 0.38:1. Females were present >50-mm shell height, but males kept appearing with high frequency up to 85 mm. Apparently, all members of the genus Pinctada exhibit this capacity to change sex at a certain size, after male maturity has been reached. Previous descriptions of this phenomenon have been made by Sevilla ( 1969) for P. mazatlanica; Wada (1953a), Tranter (1958a), and Rose et al. (1990) for P. maxima; Wada ( 1953) and Ojima and Maeki (1955) for Pinctada martensii; Tranter (1958d) for P. margaritifera; and Tranter (1959) for P. fucata. However, change in sex can be reversible and may be brought about by stress (Cahn 1949. Tranter 1958a-d. Chellam 1987, Rose et al. 1990). The ability of sex reversal has been observed in other bivalves like the Ostreidae, Teredinidae, and Pectinidae. and as hypothesized by Tranter (1958b) for P. albina. it may be explained by a "weak hereditary sex-determining mechanism." Histological data demonstrated that male maturity was reached at 39 to 49-mm shell height for P. mazatlanica. This size range, corresponding to 8 months olds, was reached using organisms reared by extensive culture. P. sterna maturity was detected at 40- to 45- mm shell height, corresponding to 1 1 months olds. How- ever, because P. sterna is a relatively small species, we believe that male maturity can be attained at a lower size, and the density inside the pearl cages could have somehow inhibited gonad maturity. Chellam ( 1978) found for P. fucata, a similar species in size, male maturity within 8 months and spawning at 9 months in Tuticorin Harbour. Tranter (1958a) also noticed P. albina. another small species, to be mature and spawn at 4 months. Male maturity occurs for P. maxima at 110-120 mm during the first year (Rose et al. 1990). Full maturity is not attained by P. margaritifera until the second year (Crossland 1957). The GI was a useful quantitative method for estimating the reproductive activity of pearl oysters. Resting on the assumption that the ratio of body parts varies little with change in size of the animal (Giese and Pearse 1974). we were able to measure the relative reproductive condition of the oysters of different sizes and to compare changes in their gonads at different times. In species possessing little nutritive tissue in the gonads, like pearl oysters, an increase in GI was interpreted as a buildup of gametogenesis. with a decrease interpreted as spawning (Giese 1959). Because weight and volume values increase by approximately the cube of linear dimensions (Galtsoff 1931 ), care should be taken to equate dimen- sions when volumetric and linear measurements are used, as oc- curred in this study. However, a limitation of the GI is that, unless accompanied by microscopic examination of the gonads, it indicates little as to what is occurring within the gonads. Therefore, if used as a single method, it could not be considered a reliable tool for studying the breeding cycle of pearl oysters and for understanding the seasonal changes occurring in their gonads. For pearl-culture programs, histology, coupled with gonad index measurements, is recom- mended for understanding the overall pattern of reproduction in pearl oysters. Other aspects of the reproductive biology of pearl oysters P. mazatlanica and P. sterna are yet to be studied. To improve the techniques and strategies for spat collection, cultivation, growth, and especially, the production of high-quality cultured pearls, an- 110 Saucedo and Monteforte nual and biannual trials on the breeding cycle of pearl oysters are recommended. ACKNOWLEDGMENTS We dedicate this work to the memory of Don Gaston Vives, pioneer of pearl culture in the world. The study was conducted as part of an institutional program of the Centro de Investigaciones Biologicas del Noroeste (CIBNOR), Mexico. It has also been funded by the International Foundation for Science (IFS of Swe- den) since 1990, the Consejo Nacional de Ciencia y Tecnologfa (CONACYT-Mexico) since 1990, and the Sistema de Investiga- dores del Mar de Cortes (SIMAC-Mexico) since 1994. We appre- ciate the invaluable help of the Pearl Oyster Research Group (Grupo Ostras Perleras) of CIBNOR. for all of the SCUBA diving support during the in situ study. Special thanks to Victor Perez, Horacio Bervera, Humberto Wright, and Sandra Morales. We are also indebted to M.C. Federico Garcia Domtnguez. CICIMAR, who provided important assistance during the histological analysis. Finally, we thank Dr. Ellis Glazier, CIBNOR, for the editorial help on the English language manuscript. LITERATURE CITED Alvarez-Borrego, S. & R. A. Schwartzlose. 1979. Masas de agua del Golfo de California. Ciencias Marinas 6:43-63. Cahn. A. R. 1949. Pearl Culture in Japan. Fishery Leaflet 357. United States Department of the Interior. Fish and Wildlife Service, Washing- ton. DC. 91 pp. Carino, M. M. 1987. Le mythe perlier dans l'histoire coloniale de la Sud- californie. These de Maitrise en Histoire Universite de Paris VII, Jus- sieu. 164 pp. Carino M. M. & C. Caceres-Martinez. 1990. La perlicultura en la Peninsula de Baja California a principios de siglo. Serie Cientifica (Special Num- ber AMAC l):l-6. Carino, M. M. & M. Monteforte. 1995. History of pearling in the Bay of La Paz, South Baja California. Mexico (1533-1914). Gems Gemol. 31: 88-I0S. Chellam, A. 1987. Biology of pearl oyster, pp. 13-21. In: K. Alagarswami (ed.). Pearl Culture. Bulletin 39. Central Marine Fisheries Research Institute. Cochin. India. Crossland, C. 1957. The cultivation of the mother-of-pearl oyster in the Red Sea. Aust. J. Freshwater Res. 8:1 1 1-135. Diario Oficial de la Federacion. 1939. Related to the Legal Permanent Ban Decreed on Pearl Oyster Fisheries by the Mexican Government. Norma oficial del Globierno de los Estados Unidos Mexicanos, 2pp. Diario Oficial de la Federacion. 1994. Related to the Determination of Terrestrial and Aquatic Species Under the Categories Endangered. Threatened. Rare, and Under Special Protection. Norma oficial del Globierno de los Estados Unidos Mexicanos, 2pp. Diaz-Garces, J. 1972. Cultivo experimental de madreperla Pinctada mazatlanica Hanley, 1 856, en la Bahia de La Paz, Mexico. Memoriae IV Congreso Na- cional de Oceanografia. Mexico, D.F.. November 17-19. pp. 429—442. Galtsoff, P. S. 1931. The weight-length relationship of the shells of the hawaiian pearl oyster Pinctada sp. Am. Nat. 65:423—433. Garcfa-Domfnguez. F., B. P. Ceballos- Vazquez & A. Tripp. 1996. Spawn- ing cycle of the pearl oyster, Pinctada mazatlanica (Hanley, 1856), (Pteriidae) at Isla Espintu Santo, Baja California Sur. Mexico. /. Shell- fish Res. 15(2):297-303. Gervis, M. N. & N. S. Sims. 1992. Biology and Culture of Pearl Oysters (Bivalvia: Pteriidae). Overseas Development Administration of the United Kingdom. International Center for Living Aquatic Resources Management, Manila, Philippines. 49 pp. Giese, A. C. 1959. Reproductive cycles of some west coast invertebrates, pp. 625-638. In: R. Withrow (ed.). Photoperiodism and related phe- nomena in plants and animals. American Association of Science, Pub- lication No. 55. Washington. D.C. 238 pp. Giese. A. C. & J. S. Pearse. 1974. Introduction and general principles, pp. 3-21. In: A. C. Giese and J. S. Pearse (eds). Reproduction of Marine Invertebrates. Vol. I. Acoelomated and Metazoans. Academic Press Inc.. New York. 284 pp. Hynd. J. S. 1955. A revision of the Australian pearl-shells genus Pinctada (Lamellibrancha). Aust. J. Mar. Freshwater Res. 6:98-137. Martinez. A. 1983. Prospeccion de los bancos de madreperla en el Golfo de California de 1962 a 1965. M.Sc. Thesis. CICIMAR-I.P.N.. La Paz, Mexico. 77 pp. Monteforte. M. 1990. Ostras perleras y perlicultura: situacion actual en los principales pafses productores y perspectivas para Mexico. Serie Ci- entifica (Special Number AMAC 11:13-18. Monteforte, M. 1991. Las perlas. leyenda y realidad: un proyecto actual de investigacion cientifica. Panorama 38:28-35. Monteforte. M. & M. Carino. 1992. Exploration and evaluation of natural stocks of Pearl Oysters Pinctada mazatlanica and Pteria sterna (Bi- valvia: Pteriidae) in La Paz Bay, South Baja California. Mexico. Ambio 21:314-320. Murillo. J. 1987. Algunas caracterfsticas paleoceanograficas y cuerpos de agua inferidos a partir de registros micropaleontologicos (Radiolaria) en la Bahia de La Paz, Baja California Sur, Mexico. B.Sc. Thesis. Universidad Autiinoma de Baja California Sur, La Paz. 68 pp. Ojima. Y. & K. Maeki. 1955. Some observations on the spermatogenesis and oogenesis in the pearl oyster Pinctada martensii (Dunker). Kwansei Gakuin Univ. Ann. Stud. 3:12-14. Osuna-Valdez, I. 1986. Evolocion Holocenica de la Laguna de La Paz, B.C.S.. Mexico. B.Sc. Thesis. Universidad Autonoma de Baja Califor- nia Sur. La Paz. 112 pp. Rose. R. A.. R. E. Dybdahl & S. Harders. 1990. Reproductive cycle of the Western Australian silver-lip pearl oyster Pinctada maxima (Jameson) (Mollusca: Pteriidae). J. Shellfish Res. 9:261-272. Sastry. A. N. 1970. Reproductive physiology variation in latitudinally separated populations of the bay scallop Aequipecten irradians. Lamark. Biol. Bull. 138:56-65. Saucedo. P. & M. Monteforte. 1994. Breeding cycle of pearl oysters Pinctada mazatlanica and Pteria sterna in Bahia de La Paz. South Baja California, Mexico. (Abstract Pearls "94). J. Shellfish Res. 13:348-349. Saucedo, P.. M. Monteforte, H. Bervera, V. Perez & H. Wright. 1994. Repopulation of natural beds of pearl oysters Pinctada mazatlanica and Pteria sterna in Bahi'a de La Paz. South Baja California. Mexico. (Abstract Pearls '94). J. Shellfish Res. 13:349-351. Sevilla. M. L. 1969. Contribucibn al conocimiento de la madreperla Pinctada mazatlanica (Hanley. 1845). Rev. Soc. Me.x. Hist. Nat. 30:223-262. Shirai, S. & Y. Sano Y. 1979. Reporte preliminar sobre los recursos de ma- dreperla y su cultivo en aguas protegidas en Baja California Sur. Institute for development of Pacific Natural Resources. Mei. Japan. Secretaria de Pesca. Baja California Sur, Mexico. Internal Report. 55 pp. Tranter, D.J. 1958a. Reproduction in Australian pearl oysters (Lamelli- branchia). I. Pinctada albina (Lamark): Primary gonad development. Aust. J. Mar. Freshwater Res. 9:135-143. Tranter. D.J. 1958b. Reproduction in Australian pearl oysters (Lamelli- branchia). II. Pinctada albina (Lamark): Gametogenesis. Aust. J. Mar. Freshwater Res. 9:144-158. Tranter. D.J. 1958c. Reproduction in Australian pearl oysters (Lamelli- branchia). III. Pinctada albina (Lamark): Breeding season and sexual- ity. Aust. J. Mar. Freshwater Res. 9:191-216. Tranter. D. J. 1958d. Reproduction in Australian pearl oysters (Lamelli- branchia). IV. Pinctada margaritifera (L.). Aust. J. Mar. Freshwater Res. 9:509-523. Tranter. D. J. 1959. Reproduction in Australian pearl oysters (Lamellibranchia). V. Pinctada fucata (Gould). Aust. J. Mar. Freshwater Res. 10:45-66. Wada. S. 1953. Biology of the silver-lip pearl oyster Pinctada maxima (Jameson). 2. Breeding season. Margarita 1:15-28. Journal of Shellfish Research, Vol. 16. No. 1. 111-114. 1997. THE EFFECT OF PENTACHLOROPHENOL ON PYRIDINE NUCLEOTIDE PRODUCTION IN OYSTER HEMOCYTES: NADPH AND IMMUNOMODULATION CAL BAIER-ANDERSON AND ROBERT S. ANDERSON University of Maryland Program in Toxicology Chesapeake Biological Laboratory P.O. Box 38 Solomons. Maryland 20688 ABSTRACT Increased NADPH production coincides with the generation of reactive oxygen species (ROS) by immunostimulated hemocytes of the oyster, Crassostrea virginica. The effects of a putative environmental immunotoxicant on NADPH production and the subsequent effects on ROS generation are reported here. Oyster hemocytes were exposed in vitro to a range of sublethal concentrations of the biocide pentachlorophenol (PCP) for 20 h. The cells were then assayed for both NADPH and superoxide generation following immunostimulation. The results indicate that PCP partially inhibits the production of both NADPH and super- oxide in a dose-dependent manner. Significant decreases in NADPH production were observed at 500 ppb. whereas significant decreases in superoxide generation were evident at 1.000 ppb. The decrease in NADPH production could represent a mechanism underlying the observed decrease in ROS production following PCP incubation. KEY WORDS: notoxicity Oyster hemocytes, pentachlorophenol. superoxide production, NADPH production, reactive oxygen species, immu- INTRODUCTION In the oyster Crassostrea virginica (Gmelin 1791). hemocytes are presumed to be important in the defense against pathogens (Anderson 1994). The production of reactive oxygen species (ROS) by hemocytes is one of the most prominent and intensely studied cell-mediated putative defense mechanisms in molluscs (Wishkovsky 1988, Adema et al. 1991, Anderson et al. 1995). There is a growing body of evidence that exposure to xenobiotics can modulate immune, responses in aquatic organisms. The sup- pression of ROS production after in vitro exposure to toxicants has been used as a sensitive biomarker for immunomodulation (Tarn and Hinsdill 1990). Although several chemical compounds have been shown to suppress ROS production in fish (Roszell and Anderson 1993, Anderson and Brubacher 1992, Warinner et al. 1988, Elasseret al. 1986) and oysters (Larson et al. 1989, Fisher et al. 1990, Roszell and Anderson 1992, Anderson et al. 1994). it is not clear if this is a manifestation of a generalized stress response or if toxicant-specific mechanisms are involved. The elucidation of specific mechanisms of toxicity would enhance the utility of im- munomodulatory responses as biomonitoring tools. ROS production in vertebrate phagocytes is well characterized (Robinson and Badwey 1992) and serves as the presumptive model for ROS production in bivalves (Anderson 1994). After immuno- stimulation, the membrane-associated enzyme NADPH oxidase catalyzes the transfer of a single electron from NADPH to mo- lecular oxygen, producing the superoxide anion (0 2 ~). Superoxide, while cytotoxic in itself, can be further metabolized to more highly toxic species. It can undergo dismutation to hydrogen peroxide, which, in turn, may be converted by Fenton chemistry into hy- droxyl radicals (Halliwell and Gutteridge 1989). The enzyme my- eloperoxidase, found in mammalian neutrophils as well as oyster hemocytes. catalyzes the production of hypohalous acids, such as HOC1. from H 2 2 (Rosen and Klebanoff 1985). Several other interactions among the ROS are possible, generating a variety of oxygen species. Each of these reactive oxygen compounds has cytotoxic properties, inducing lipid peroxidation and enzyme in- activation. Increased production of NADPH, chiefly by an up-regulation of the pentose phosphate pathway, is essential to the production of O-T by NADPH oxidase. The purpose of this study was to measure the effects of a putative immunotoxicant on NADPH production in oyster hemocytes and to examine the relationship between NADPH and 2 ~ production. The pesticide selected for this study, pentachlorophenol (PCP). is a potent general biocide and common aquatic pollutant (Ahlborg and Thunberg 1980). PCP is an uncou- pler of oxidative phosphorylation; the resultant decrease in ATP production and subsequent effect on NADPH production were postulated to contribute to the observed immunotoxicity. MATERIALS AND METHODS Hemocyte Collection Oysters, collected from the Wicomico River in St. Mary's County. MD, were maintained in a flow-through tank under am- bient temperature (4—7°C) and salinity (10 ppt) conditions. The hemocytes were extracted as previously described (Anderson et al. 1995). Briefly, the oysters were notched and hemolymph was col- lected with a syringe from the adductor muscle sinus. The pooled hemolymph (nine oysters/pool) was plated onto glass Petri dishes and incubated at room temperature (22-23°C) for 15 min. Non- adherent cells were gently removed by rinsing with filtered ambi- ent sea water (FA), and the adherent cells were incubated in FA at room temperature for an additional 2 h. After the incubation, the cells were collected by gentle aspiration and centrifuged at 300 g for 15 min. The FA was decanted, and the cell pellet was resus- pended in Hanks Balanced Salt Solution (HBSS), made iso- osmotic with NaCl, and augmented with 1 mg/mL glucose and 3% antibiotic/antimycotic solution (Sigma). Pentachlorophenol Exposure Hemocytes (10 6 cells/mL) were incubated with PCP ranging from 100 to 1,000 ppb (0 2 _ studies) or 10 to 1.000 ppb (NADPH studies) for 20 h at room temperature. Stock solutions, made from water-soluble sodium PCP (Aldrich), ranged from 1.0 to 100 ppm such that an equal volume of stock was added to each treatment vial. Viability after incubation was assessed by use of the trypan HI 112 Baier-Anderson and Anderson blue exclusion assay (equal volume of 0.4% trypan blue in HBSS and cells suspended in HBSS: incubation time = 5-10 min) and was based on four pools of cells. Measurement of NADPH Production NADPH production by chemically stimulated oyster hemocytes was estimated with the CellTiter 96 AQ kit (Promega). which is a colorimetric assay based on the production of reduced pyridine nucleotides. After a 20-h incubation with PCP. each pool of cells was centrifuged, decanted, resuspended in fresh HBSS (iso-osmotic, augmented with 1 mg/mL glucose), and divided equally into 96-well plates (200,000 cells/well, six wells/treatment per pool.) Superoxide dismutase (300 U/mL, final concentration) was added to each well to prevent any spurious interaction be- tween O-T and the assay reagents. The NADPH oxidase stimulator phorbol 12-myristate 13-acetate (PMA, Sigma) was added to half of the wells of each treatment group to give a final concentration of 0.001 mM. Cells were incubated for 20 min at room tempera- ture, and then the reagent mixture was added. Incubation continued for an additional 60 min. at which time the color change was read on a Bio-Rad model 2550 EIA Reader at 492 nm. Three separate pools of hemolymph were analyzed to permit statistical evaluation. The use of this assay to evaluate NADPH production with immunostimulation represents a novel application because the Promega CellTiter 96 AQ kit is traditionally used to characterize viability or cell proliferation. The kit consists of two reagents: the sulfated, water-soluble tetrazolium. 3-(4.5-dimethylthiazol-2-yl)- 5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt (MTS), and phenazine methosulfate (PMS), which acts as an electron transfer intermediary. Reduced pyridine nucleotides complex with PMS. allowing for a two-electron transfer to occur, forming A'-methyl dihydrophenazine. This compound is capable of a single-electron transfer to the tetrazolium. resulting in the for- mation of formazyl and phenazyl radicals. The transfer of the second electron results in the production of a blue formazan and the regeneration of PMS (Dunigan et al. 1995). This reaction can be monitored colorimetrically at 490 nm. Phagocytic cells pro- duced NADPH on appropriate immunostimulation; this was de- tected by absorbance readings that were elevated over the con- comitant baseline production of reduced pyridine nucleotides (NADPH and NADH) by the cells. Measurement of Superoxide Production The effect of PCP on 2 ~ production was tested by the use of lucigenin-augmented chemiluminescence (CL). After incubation with PCP. lucigenin (frw-A'-methylacridinium nitrate. Sigma) was added to each incubation vial (25 u,M. final concentration), and the vials were loaded into a Packard Tri-Carb 1900 CA liquid scintil- lation analyzer, adapted for single-photon counting. After baseline measurements were obtained, zymosan (Sigma), at a concentration of 4 mg/mL. was added, and CL was monitored for approximately 2 h. Cell-free experiments with xanthine/xanthine oxidase as the O^-generating system demonstrated that the presence of PCP does not interfere with the detection of O," (p = 0.4069. data not shown). The assay was repeated four times with four pools of hemolymph. Statistics All data were analyzed with the Prism™ (GraphPad) statistics package. Viability data were analyzed by the use of analysis of variance (ANOVA). For both NADPH and O-T production data, repeated-measures ANOVA was used, because variation between pools was expected to be significant. Both baseline NADPH pro- duction (unstimulated cells) and net NADPH production (esti- mated by subtracting the absorbance of unstimulated cells from the absorbance of stimulated cells) were evaluated. CL data were ana- lyzed in terms of both area under the curve and peak CL minus unstimulated baseline. Two post-hoc tests were used when the ANOVA showed significant (p < 0.05) variation between study groups: Dunnet's multiple comparison test to identify differences between control and treatment means, and trend analysis to test for a linear relationship among the means. RESULTS Exposure to =£1.000 ppb PCP did not significantly affect hemocyte viability (p = 0.3886, data not shown). The experimen- tal design incorporated the use of repeated-measures ANOVA to analyze the baseline NADPH, net NADPH. and Or data; the pairing was statistically significant at p < 0.001. p < 0.005. and p < 0.0001. respectively, indicating that the matching was effective and that the use of this method was appropriate. Exposure to PCP resulted in no significant differences in the baseline production of NADPH (p = 0.8451, data not shown): however, significant in- hibition of stimulated net NADPH production (p = 0.0137) was evident at both 500 and 1.000 ppb (Fig. 1 ). The evaluation of the integrated area under the curve for 2 ~ generation indicated sig- nificant decreases at 1,000 ppb (p = 0.0178; Fig. 2). The analysis of peak minus baseline data gave similar results (p = 0.0166. data not shown). In all instances, the relationships between the control and treatment means were linear (p < 0.005). indicating significant dose-response relationships. DISCUSSION In mammals, the enzyme NADPH oxidase facilitates the trans- fer of an electron from NADPH to molecular oxygen to produce O,". Although direct evidence of the presence of this enzyme in oyster hemocytes is lacking, there is circumstantial evidence of its existence. Hemocytes share several important characteristics with mammalian blood cells, including phagocytic capacity, the pro- duction of ROS, and the generation of effector molecules such as lysozyme that may function in concert with ROS. Like their raam- 150 125 100 o c i=? X o" 075 a. o 3 i 050- 025 0000 Control 10 ppb 100 ppb 500 ppb 1000 ppb Figure 1. NADPH production by oyster hemocytes after 20-h in vitro exposure to PCP (200,000 cells/well, n = 3). Stars indicate means that differ significantly from control (Dunnet's post-hoc test, p < 0.05). Stim, stimulated; Unstim. unstimulated. Effect of PCP on NADPH Production in Hemocytes 113 1500000 Im O ■D C 3 n 0> 0) 1000000 < £ "So *■» nj 500000 O) 0) control 100 ppb 500 ppb 1000 ppb Figure 2. I.ucigenin-augmented CL after 20-h in vitro exposure to PCP (10" cells/ml., integrated area under the eur\e, n = 4). Star indi- cates mean that differs significantly from control ( Dunne! 's post-hoc test, p < 0.05). malian counterparts, hemocytes respond to immunostimulation by zymosan or PMA by producing ROS. although their response is lower in magnitude. Inhibitors of NADPH oxidase activity have been shown to decrease CL in the hemocytes of molluscs such as Lymnaea stagnalis ( Adema et al. 1 993 ) and Mytilus edulis ( Noel et al. 1993). Therefore, it is not unreasonable to assume that an enzyme homologous to NADPH oxidase may be active in oyster hemocytes. Recent evidence indicates that ROS production by oyster hemocytes may not be elicited, or may even be suppressed, by exposure to certain viable, potentially pathogenic bacteria (Bramble and Anderson, in press) or the protozoan parasite Per- kinsus marinus (La Peyre et al. 1995). suggesting that other factors may also contribute to cell-mediated defense. However, by anal- ogy to the better characterized mammalian systems. ROS produc- tion may be considered a measure of the defensive capacity of oyster hemocytes and the inhibition of this response could impose a limitation on disease resistance. In phagocytic cells, the NADPH used to fuel the respiratory burst is generated by the pentose phosphate pathway, which uses glucose-6-phosphate (G6P) supplied by glycolysis or glycogen metabolism. ATP is required in several steps in the production of G6P by glycolysis, as well as in the activation of the enzymes phosphorylase kinase and glycogen phosphorylase — both of which are necessary for the cleavage of GIP from glycogen. ATP may also be necessary for the assemblage of NADPH oxidase. PCP, an uncoupler of oxidative phosphorylation, is similar to dinitrophenol (Cantelmo et al. 1978) in that it acts as a proton shuttle across mitochondrial membranes, depleting the proton gradient required for ATP production. Because ATP is required for NADPH pro- duction. PCP has the potential to limit the availability of NADPH for ROS production. The data presented here indicate that in vitro exposure to PCP inhibits NADPH production by immunostimulated hemocytes. Al- though both O-T production and NADPH production by oyster hemocytes are decreased after exposure to PCP, significant de- creases in NADPH appear at lower concentrations of PCP than does decreased 2 " production. These results suggest that the de- creased NADPH production is the proximate cause of the de- creased ROS production. The fact that NADPH appears to be slightly more sensitive to PCP implies that the coupling between NADPH production and superoxide generation is not tightly linked. This is not unexpected, because NADPH is an important cofactor in numerous other cellular functions unrelated to the pro- duction of ROS. In summary, oyster hemocytes exposed in vitro to PCP exhib- ited decreased CL in response to phagocytic stimulation. They also demonstrated significantly decreased NADPH production in response to chemical stimulation with PMA. Because NADPH is a required cofactor in the production of 2 ~. it appears that the immunosuppressive action of PCP results from the reduced NADPH production. Other possible explanations for decreased 2 ~ production include direct interference with NADPH oxi- dase assembly or altered cellular redox status (decreased NADPH: NADP + ). leading to lipid peroxidation and enzyme inactivation. Because 2 ~ production may be an element of microbicidal defense in oyster hemocytes. the inhibition of this pathway by exposure to environmental contaminants could have dire con- sequences in terms of resistance to infectious diseases (Anderson et al. 1996. Chu and Hale 1994). However, a concrete relation- ship between decreased 2 ~ production and increased suscep- tibility to disease cannot be established until the role of 2 ~ production in disease resistance in oysters is completely charac- terized. LITERATURE CITED Adema. CM.. W. P. W. van der Knaap & E. Sminia. 1991. Molluscan hemocyte-mediated cytotoxicity: the role of reactive oxygen interme- diates. Rev. Aquat. Sci. 4:201-223. Adema. C. M.. E. C. van Deutekom-Mulder. W. P. W. van der Knapp & T. Sminia. 1993. NADPH-oxidase activity: the probable source of reac- tive oxygen intermediate generation in hemocytes of the gastropod Lymnaea stagnalis. J. Leak. Biol. 54:379-383. Ahlborg. V. G. & T. M. Thunberg. 1980. Chlorinated phenols: occurrence, toxicity, metabolism, and environmental impact. Crii. Rev. To.x. 7:1-35. Anderson. R. S. 1994. Blood cell-mediated oxyradical production in aquatic species: implications and applications, pp. 241-265. In: D. C. Malins and G. K. Ostrander (eds.). Aquatic Toxicology: Molecular. Biochemical and Cellular Perspectives. CRC Press, Boca Raton, FL. Anderson. R. S. & L. L. Brubacher. 1992. In vitro inhibition of medaka phagocyte chemiluminescence by pentachlorophenol. Fish Shellfish Immunol. 2:299-310. Anderson. R. S.. L. M. Mora & L. L. Brubacher. 1995. Luminol-dependent chemiluminescence in molluscs, pp. 101-108. In: J. S. Stolen. T. C. Fletcher, S. A. Smith. J. T. Zelikoff. S. L. Kaatari. R. S. Anderson. K. Soderhall. and B. A. Weeks-Perkins (eds.). Techniques in Fish Immu- nology-4. Immunological and Pathological Techniques of Aquatic In- vertebrates. SOS Publications. Fairhaven. NJ. Anderson. R. S.. L. M. Mora & S. A. Thompson. 1994. Modulation of oyster (Crassostrea virginica) hemocyte immune function by copper, as measured by luminol-enhanced chemiluminescence. Comp. Bio- chem. Physiol. 108C:2 15-220. Anderson. R. S.. M. A. Unger & E. M. Burreson. 1996. Enhancement of Perkinsus marinus disease progression in TBT-exposed oysters {Cras- sostrea virginica). Mar. Environ. Res. 42:177-180. Bramble. L. A. & R. S. Anderson, in press. Modulation of Crassostria virginica hemocyte reactive oxygen species production by Listonella anguillarum. Dev. Comp. Immunol. Cantelmo. A. C, P. J. Conklin, F. R. Fox & K. R. Rao. 1978. Effects of sodium pentachlorophenate and 2.4-dinitrophenol on respiration in crustaceans, pp. 251-264. In: K. R. Rao (ed.). Pentachlorophenol: Chemistry, Pharmacology, and Environmental Toxicology. Plenum Press. New York. Chu. F.-L. E. & R. C. Hale. 1994. Relationship between pollution and 114 Baier-Anderson and Anderson susceptibility to infectious disease in the eastern oyster, Crassostrea virginica. Mar. Environ. Res. 38:243-256. Dunigan. D. D.. S. B. Waters & T. C. Owen. 1995. Aqueous soluble tet- razolium/formazan MTS as an indicator of NADH- and NADPH- dependent dehydrogenase activity. Biotechniques 19:640-649. Elasser. M. S., B. S. Roberson & F. M. Hetrick. 1986. Effects of metals on the chemiluminescent response of rainbow trout iSalmo gairdneri) phago- cytes. Vet. Immunol. Immunopathol 12:243-250. Fisher, W. S., A. Wishkovsky & F.-L. E. Chu. 1990. Effects of tributyltin on defense-related acti