Initial recruitment and growth of surfclams (Spisula solidissima Dillwyn) on the inner continental shelf of New Jersey.

ABSTRACT Surfclam (Spisula solidissima Dillwyn) larval settlement and the initial growth of recruits were studied on the inner shelf of New Jersey. Initial recruitment was measured by taking weekly benthic core samples during the summer settling season, and larval supply was characterized using meroplankton samples taken every four hours in July. The temporal variation in recruitment at two inshore stations (12-m depth) was linked to larval supply from the water column, and spatial differences (inshore vs. offshore) in recruitment also appeared to be related to larval supply. Spatial and temporal variation in larval concentrations was associated with wind-driven cross-shelf circulation. Contrasting recruitment patterns between the two inshore stations could not be explained by larval supply alone and were likely affected by near-bottom flows. Growth rates of initial surfclam recruits (with initial shell lengths <360 [micro]m) were estimated to be 10-20 [micro]m [d.sup.-1], and the growth rates of individuals >360 [micro]m shell length were 25-50 [micro]m [d.sup.-1]. This study provides realistic field estimates of early growth rates of surfclams and further evidence of the relationship between upwelling/ downwelling events and surfclam larval supply and initial recruitment on the inner continental shelf.

KEY WORDS: surfclam larvae, settlement, post-larval growth, recruitment

INTRODUCTION

The surfclam (Spisula solidissima Dillwyn) lives in a zone from the shallow subtidal out to depths of about 60 m, and ranges from the Gulf of St. Lawrence, Canada, south to North Carolina (Merrill & Ropes 1969). It is an important commercial species in the Mid-Atlantic Bight (NOAA/NMFS 2003). Surfclam populations are characterized by great year-to-year variation in recruitment success, especially in inshore areas (Murawski & Serchuk 1989, Weinberg 1993, 1999, Chintala & Grassle 2001). Pre- and postsettlement processes, both physical and biological, are likely responsible for spatial and temporal variation in recruitment success.

To understand factors affecting inshore surfclam larval settlement and initial recruitment a study was initiated at a Long-term Ecosystem Observatory at 15-m depth (LEO- 15) off southern New Jersey in 1993. The study area lies within one of the recurrent upwelling centers on the New Jersey coast (Glenn et al. 2004) and is characterized by episodic summer upwelling/downwelling events, which appear to be related to the arrival of high concentrations of surfclam larvae. Studies on larval surfclam supply with respect to these events (Ma & Grassle 2004, Ma 2005, Ma et al. 2006), larval settlement (Snelgrove et al. 1999, 2001, Weissberger & Grassle 2003, Ma 2005) and recruitment during the first year of life (Weissberger & Grassle 2003) form the background for the present study. All of these observations and experiments were focused on larvae spawned by inshore surfclam populations that experience rapidly warming temperatures in the late spring (Starypan 1976). At LEO-15, benthic studies (1993-2004) show that larval settlement from this inshore spawning population occurs chiefly in late June and July, with some markedly reduced settlement in August in some years (Weissberger & Grassle 2003, Snelgrove et al., 1999, 2001, Grassle unpubl, data). Offshore surfclams living under the influence of the Middle Atlantic cold pool do not spawn until the thermocline breaks down in the late summer or early fall (or in some instances when there is storm mixing). It is likely that these larvae are widely dispersed on the continental shelf in the fall and winter months, and we have evidence of surfclam larvae being present in the plankton at LEO-15 as late as February of the following year (Gregg, Tucker & Grassle, unpubl. data).

This study focused on benthic sampling at three LEO-15 stations during the expected peak in surfclam settlement in July 1998, together with frequent near-bottom sampling of surfclam larvae at the same stations. The year 1998 was chosen in part because initial surfclam recruitment was relatively high for the first time since 1993 (Weissberger & Grassle 2003). The primary goal was to identify the arrival of high concentrations of competent surfclam larvae, and to follow the growth of the consequent pulses of juvenile surfclams in the benthic samples at frequent enough intervals to provide estimates of growth rate in the earliest postsettlement stages (i.e., from a maximum shell length at settlement of approximately 290 [micro]m). One goal of our study was to describe surfclam growth during the earliest postsettlement stages, because it is likely that these growth rates will affect whether the juveniles find a size refuge from predation.

MATERIALS AND METHODS

Study Site

The study site was located in the vicinity of Beach Haven Ridge (Fig. 1). Samples were taken from three stations, two inshore (Sta. 9 and C at 12 m depth) and one offshore (Sta. C2 at 20 m depth). Sta 9 (39[degrees]27.69'N, 74[degrees]15.81'W) is on the inshore flank of the ridge with a coarse sand bottom, and Sta. C (39[degrees]27.85'N, 74[degrees]15.11'W) is on the offshore flank of the ridge with a finer sand bottom (Craghan 1995). Sta. C is about 1 km seaward of Sta. 9. Sta C2 (39[degrees]23.24'N, 74[degrees]12.75'W) is a sandy site, 8 km offshore from Sta. 9. At Sta. 9 and C bottom temperatures during the summer vary greatly during alternating periods of upwelling and downwelling, whereas the bottom temperature at Sta. C2 rarely exceeds 12[degrees]C during the summer months, except for occasional storm mixing (Glenn et al. 2004, Ma & Grassle 2004).

[FIGURE 1 OMITTED]

Physical Data

Alongshore wind (SW/NE) and water temperature 1 m above the bottom were used to track upwelling and downwelling events. Wind data were from the Rutgers University Marine Field Station meteorological tower, where wind intensities were recorded 10 m above the ground. Water temperatures were recorded by thermistors on the zooplankton pump housing and on the pump frame.

Surfclam Larval Sampling

Three Moored, Automated, Serial, Zooplankton Pumps (MASZPs, Doherty & Butman 1990) were deployed at Sta. 9 and C from July 6-27 and at Sta. C2 from July 11-14, 1998. Samples (250 1) were taken automatically every 4 h, 1 m above the bottom (see Ma & Grassle 2004 for further details). The samples were filtered through 100-[micro]m mesh and organisms trapped on the filter were wound onto a take-up spool in a fixative compartment with 10% buffered formaldehyde. The take-up spool was retrieved after a 3-wk deployment. The meshes were cut into individual pieces corresponding to each sample and transferred to 70% ethanol with Rose Bengal. Each sample (no splitting) was examined under a dissecting microscope. Bivalve larvae with shell length >150 [micro]m were identified (Ma & Grassle 2004). In samples with high concentrations of surfclam larvae, shell lengths for up to 100 individuals were measured using an ocular micrometer to judge the readiness of the larvae to settle.

Benthic Sampling

Divers took 8 benthic core samples (7-cm diameter corer, area 38.5 [cm.sup.2]) at Sta. 9, C, and C2 at approximately weekly intervals in July, then at increasing intervals to monthly from September to December 1998. At Sta. 9 and C2 (rippled bottom) four pairs of samples (ripple trough and adjacent crest) were taken, with about 1 m between pairs. At Sta. C (no ripples) paired samples were taken at approximately the same intervals. Core samples were kept on ice for several hours until they were preserved in 95% ethanol and subsequently stained with Rose Bengal. The samples were sieved over a 106-[micro]m sieve and sorted under a dissecting microscope. When there were high concentrations of surfclams at Sta. C, a total of 200 individuals (100 haphazardly taken from each of two replicate cores) were measured. At Sta. 9, individuals from 6-8 replicate cores were combined to obtain 200 individuals for each sampling date in July. Initial surfclam recruitment at C2 was too low to achieve comparable sample sizes for measurement. Surfclam images were captured using a color camera linking a Zeiss dissecting microscope to a Macintosh computer and the IMAGE program (calibrated with a micrometer) was used to measure shell lengths.

Surfclam larvae settle at shell lengths from 230-290 [micro]m (Snelgrove et al., 1998), and surfclam juveniles were observed to increase in shell length -10 [micro]m per day in the laboratory (J. Bell, unpublished data). In the present study surfclams with shell lengths <360 [micro]m were considered to have settled recently, probably within the previous week (a time interval equivalent to the benthic sampling intervals in July). This cutoff assumed an upper limit for settling larvae of 290 [micro]m and 7 days of growth at 10 [micro]m [d.sup.-1]. The accumulation of surfclam individuals <360 [micro]m was used to estimate initial recruitment with the recognition that this estimation would incorporate early postsettlement mortality.

Data Analysis

A Computer-Assisted Analysis of Mixtures (C.A.MAN, Bohning et al. 1992) was used to find the number of cohorts and the mean shell lengths for each cohort. On each date/station combination, shell distributions were fitted to different numbers of cohorts. In general, the greater the number of cohorts included for fitting, the larger the likelihood of the fitting. Decisions on choosing the number of cohorts were based on likelihood adjusted by the number of estimated parameters (i.e., Akaike's Information Criterion, Weissberger & Grassle 2003) and visual fitting with the raw data in histograms. The initial values (mean and weight of each cohort) were estimated from raw data, and C.A.MAN calculated the maximum likelihood estimates for these parameters. Growth was estimated by following the cohorts over time. Urban (2002) pointed out that bivalve larvae and postlarvae grow linearly or exponentially whereas adult growth is asymptotic. Growth rates for the surfclams were calculated assuming both linear ([L.sub.t]+[DELTA]t = [L.sub.t] + r x [DELTA]t) and exponential growth ([L.sub.t+[DELTA]t = [L.sub.t] x [e.sup.r[DELTA]t).

Analysis of variance was carried out for densities of initial surfclam recruits (<360-[micro]m shell length) in July using a general linear model (GLM) in SAS. Dates and stations were fixed factors. The GLM procedure was used instead of ANOVA because of the variable number of replicate samples (6-8) completed for each station/date combination (Cody & Smith 1997). The data for number of surfclams were logarithmically transformed to remove heterogeneity in variance. Because of the unbalanced design, standard multiple comparison tests (such as Tukey, Duncan, or SNK tests) were not used (Cody & Smith 1997). Instead, multiple t-tests with an overall significance level [alpha] (= k x [alpha]') = 0.05 were used, where k is the number of comparisons and [alpha]' is the significant level for individual t-tests (Bonferroni method).

RESULTS

Physical Conditions and Surfclam Larval Concentrations in the Water Column, July 1998

Downwelling winds were interspersed with episodic upwelling winds before July 12, and bottom temperatures at the two inshore stations, Sta. 9 and C, varied between 13[degrees]C and 17[degrees]C until July 10 when temperatures increased to around 20[degrees]C from July 11-13 (Fig. 2). Upwelling winds began July 12 and persisted until late on July 24 when downwelling winds blew for 2 d. During the sustained upwelling wind, bottom temperatures were <13[degrees]C for >10 days. Temperatures increased near the end of the larval sampling period during the 2 d of downwelling winds. At offshore Sta. C2, bottom temperatures remained <12[degrees]C throughout July (Fig. 2).

[FIGURE 2 OMITTED]

Surfclam larval concentrations were relatively high and variable before July 13 at the two inshore stations (Fig. 2). At Sta. C there was a sharp peak in surfclam larval concentration on July 12-13, coinciding with the presence of relatively low salinity water (Ma and Grassle 2004). Surfclam larval concentrations were low (0-46 ind. 250 1-1) inshore after July 13 during the sustained upwelling. At offshore Sta. C2, surfclam larval concentrations were low ([less than or equal to] 20 larvae 250 [1.sup.-1]) from July 11-14 at the time when high concentrations were found inshore.

Except for the magnitude of the peak in larval surfclam concentration at Sta. C on July 12-13, the integrated larval supplies to Sta. 9 and C during July were very similar. Larvae present in the water column at that time were mostly of a size suggesting that they were competent to settle based on laboratory culture studies (Table 1). In Table 1, shell length distributions for surfclams in plankton samples where they were abundant are shown on 5 d between July 8 and 13. There was a small percentage of probable precompetent larvae (<230 [micro]m), 30-80% competent larvae (230-290 [micro]m) and a small fraction of juvenile surfclams >320-[micro]m shell length that had been resuspended from the bottom. This leaves a sizeable proportion (5-40% in the different samples) ranging in shell length from 290-320 [micro]m, which probably included both larvae and recently metamorphosed individuals.

Abundances of Initial Surfclam Recruits

At inshore Sta. 9, total surfclam densities were >20 ind. [core.sup.-1] in July and <5 ind. [core.sup.-1] in August and September (Fig. 3). Relatively high surfclam densities appeared again in November with the recruitment of fall-spawned larvae in October and November. This October-November pulse was observed at all three stations. At Sta. C, surfclam densities were very high (>200 ind. [core.sup.-1]) in July. At offshore Sta. C2, surfclam densities were <10 ind [core.sup.-1] throughout the summer and fall and about triple that in November with the fall recruitment (Fig. 3).

[FIGURE 3 OMITTED]

Analysis of variance for initial surfclam recruits (referring to individuals with shell lengths <360 [micro]m) in July showed that sampling date, stations, and the interaction between date and station had significant effects on densities (Table 2, Fig. 4). The pooled initial recruits from the three stations were significantly more abundant on July 6 and July 14. Initial recruitment was highest at Sta. C and lowest at Sta. C2. At Sta. C and C2, the number of initial recruits was significantly higher on July 6 and 14 than on July 23 and July 29 (Table 2). At Sta. 9, initial recruitment was significantly higher on July 6 and 14 and lowest on July 29.

[FIGURE 4 OMITTED]

The initial surfclam recruits present in the benthos on July 14 at all three stations were a likely consequence of the competent larvae present in the water column in the previous sampling interval (Fig. 4). Initial surfclam recruitment at each of the stations was roughly proportional to larval supply, with the highest larval supply and recruitment at Sta. C and the lowest at Sta. C2. However, the difference in initial recruitment between Sta. 9 and C was comparatively great (~10 times) compared with larval supplies to the two inshore stations. The average larval surfclam concentrations during July 6-14 were 79 ind. 250 [1.sup.-1] at Sta. 9 and 119 ind. 250 [1.sup.-1] at Sta. C. Low abundance of initial surfclam recruits in the benthos at all three stations at the end of July was correlated with low larval supply in the last two July sampling intervals.

Short-term Growth of Surfclams

Three modes were identified by the C.A.MAN procedure for the surfclams sampled on July 6 at Sta. 9 and C and two modes for the other three dates in July (Fig. 5). Growth rates were calculated when advancement of modes with time could be established (lines shown on Fig. 5). Only modes with weights (proportions) >0.2 (20%) were included for growth estimation (the weights of all modes at one date/station combination add up to 1). Both linear ([micro]m [d.sup.-1]) and exponential ([d.sup.-1]) growth rates were calculated. Two estimates were made at Sta. 9 and six at Sta. C. (Fig. 5 and Fig. 6). Only the results using the linear growth model are shown. Linear growth rates for initial surfclam recruits (<360-[micro]m shell length) ranged from 9.7-18.1 [micro]m [d.sup.-1]. Exponential growth rates for the same individuals ranged from 0.029-0.052 [d.sup.-1]. Linear growth rates of individuals >360-[micro]m shell length ranged from 25.0-51.6 [micro]m [d.sup.-1] (0.021-0.081 [d.sup.-1] for exponential growth rates). There were too few surfclam recruits at Sta.C2 to permit an analysis of growth, and they ranged mainly from 260-960 [micro]m shell length in July (more similar to Sta. 9).

[FIGURES 5-6 OMITTED]

The surfclams recruiting at Sta. C in the July 6-14 interval did not appear as a separate mode on July 14 (Fig. 5). The first mode on July 14 at Sta. C was likely composed of survivors from the first mode on July 6 and larvae settling during July 6-14. This could have resulted in an underestimate of the growth rates of surfclams in the first mode on July 6. Visually selecting the shell length with the highest frequency (highest bar in Fig. 5) on July 14 as the "mode" for survivors from July 6 would increase the linear growth-rate estimate by only 4 [micro]m [d.sup.-1].

The intersampling growth periods at Sta. 9 and C in July provided a contrast in temperature regimes: interval 1 (July 6-14) was characterized by temperatures varying from 13[degrees]C to 22[degrees]C, including some very rapid increases and decreases; interval 2 (July 14-23) was characterized by uniformly low temperatures (12[degrees]C to 13[degrees]C) during upwelling; interval 3 (July 23-29) was more like interval 1, with temperatures varying from 12[degrees]C to 23[degrees]C at Sta. C and 13[degrees]C to 20[degrees]C at Sta. 9 (Fig. 1). There appeared to be no relationship between growth rates and the prevailing temperature regimen (Fig. 6).

DISCUSSION

There was good correspondence between the timing of the highest initial surfclam recruitment and surfclam larval supply at the two inshore stations in July 1998. There was also a rough correlation at all three stations between the magnitude of larval surfclam supply during the period July 6-14 and the abundance of initial surfclam recruits, with Sta. C having the highest larval supply and the highest abundance of recruits and Sta. C2 the least. This relationship between surfclam larval supply and initial recruitment had previously been demonstrated experimentally at the same three stations in 1997 (Ma 2005), although in that year the correlation was significant for Sta. C and C2 but not for Sta. 9. At Sta. 9 the abundance of initial surfclam recruits was less than would have been expected given the similarity in larval supply between Sta. 9 and C, the two inshore stations, which are only one kilometer apart. Ma (2005) has hypothesized that the generally lower initial surfclam recruitment at Sta. 9 compared with Sta. C over several years (Weissberger & Grassle 2003, Quijon & Grassle, in prep.) could be a consequence of stronger near-bottom flows at Sta. 9, especially during storms. These flows could inhibit surfclam larval settlement and/or be responsible for resuspending sediment and winnowing initial recruits out of the sediment and depositing them on the crest and seaward side of Beach Haven Ridge. This hypothesis is partly supported by the difference in sediment grain size between Sta. 9 and C (Craghan 1995), namely coarser sediment characterizes Sta. 9, and it may also explain the great difference (~ one order of magnitude) in initial surfclam recruitment between the two inshore stations during downwelling in July 1998 in the face of relatively similar integrated larval supplies.

Based on shell lengths, many of the surfclams present in core samples inshore on July 6 settled in June, probably in association with several upwelling/downwelling periods during that month, namely relaxation or downwelling on June 9-12 and June 17-30. After the initial larval settlement pulses in June and early July, a sharp decline in surfclam densities was observed, and there were few or no individuals remaining in the sediments in August and September. Assuming that there was no new settlement and the decline in density was caused by mortality only, the estimated mortalities (z) using an exponential survival equation [N.sub.t+[DELTA]t] = [N.sub.t] x [e.sup.-z[DELTA]t] (e.g., Hilborn & Walters 1992) were 0.21-0.22 [d.sup.-1] from July 23-29 and from July 29 to August 13. The absence of larval settlement was reflected in the very low numbers of surfclams with shell lengths <360 [micro]m on July 23 (immediately after upwelling), on July 29, and August 13, especially at Sta. C. However, the estimated mortalities could also be affected by resuspension and bedload transport of surfclam recruits to or from other locations.

The mortality of juvenile surfclams was likely caused by multiple impacts of numerous predator species. They include several species of brachyuran crabs (Stehlik 1993), hermit crabs (Newby 2005), crangonid shrimp (Viscido 1994, Viscido et al., 1997), naticid snails (Weissberger & Grassle 2003, Quijon & Grassle, in prep.) and starfish (Weissberger & Grassle 2003). It is likely that the relative importance of the predators varies from year to year depending on the timing of settlement and initial abundance of the juvenile stages that prey on the smallest surfclams. For example, the 1993 summer year class of surfclams was preyed on by the naticids Neverita duplicata and Euspira heros and the sea star Asterias forbesi. By virtue of the high density of recently settled sea stars beginning in July of that year at Sta. 9 and C, and their high feeding rates, their impacts on juvenile surfclam density were estimated to be much higher (4,500 surfclams [m.sup.-2] [d.sup.-1]) than the impacts of naticid predation (90 surfclams [m.sup.-2] [d.sup.-1]) (Weissberger & Grassle 2003). In 1998 no sea stars were observed in the benthic samples so it is likely that the declining densities in August and September in the present study were a combination of crustacean and naticid predation (Quijon & Grassle, in prep).

Total surfclam densities increased in October and November at all three stations. More than 80% of individuals in November at the two inshore stations were <430-[micro]m shell length, and >50% were <430-[micro]m shell length at the offshore station. It is highly likely that these increases in densities were caused by new larval settlement and recruitment in the fall, the result of spawning by surfclams living below the summer thermocline (Tarnowski 1982, Wagner 1984, Weissberger & Grassle 2003, Ma & Grassle 2004). When the thermocline breaks down, warmer water reaches the bottom, inducing the offshore clams to spawn. Pulses of high larval surfclam concentrations were previously observed in the water column in the fall (September to October) at a southern New Jersey site within 80 km of the LEO-15 research area (Haskin et al. 1981, Tarnowski 1982). We also have unpublished data showing that fall-spawned surfclam larvae may persist in the water column at LEO-15 into December and even the following February (Gregg, Tucker, & Grassle unpubl, data), though whether the larvae found in the plankton in February settle and metamorphose successfully is unknown.

Urban (2002) evaluated the applicability of several standard models to the growth of different developmental stages of pearl oysters, including larvae, plantigrades, postlarvae, and juveniles. Like previous authors he concluded that the larvae and plantigrades grow linearly or exponentially (e.g., Bayne 1965, Beaumont & Budd 1982, Rose & Baker 1994). The linear growth rates for initial surfclam recruits (<360 [micro]m) in our study were 9.7-18.1 [micro]m [d.sup.-1]. Subsequent growth rates of juveniles >360-[micro]m shell length were somewhat higher (25.0-51.6 [micro]m [d.sup.-1]). These growth rates were comparable to those estimated by Weissberger and Grassle (2003) for clams settling in similar densities in 1993 (14.6-63.5 [micro]m [d.sup.-1]). No clear trend of increased linear growth rate with increased initial size (an indication of exponential growth) was observed in our study. The lower growth rates for the initial surfclam recruits (<360 [micro]m) may be a result of the period of time between loss of the larval velum and formation of labial palps at metamorphosis and development of functional suspension-feeding structures (fused siphons and gills). This has not been examined directly in Spisula solidissima, but in other bivalves these developmental processes may take one to several days depending on temperature (e.g., Bayne 1971, Reid et al. 1992).

Plotting 1993 growth rates directly against bottom temperature, Weissberger and Grassle (2003) found some correlation up until September. In the present study there was no evidence of any growth rate/temperature correlation for the smallest clams in July 1998. Of course many factors may affect the growth of bivalve recruits, not just temperature. They include population density and the availability of appropriate food. At different intertidal sites in Maine, surfclam juveniles grew faster at sites with warmer water temperatures and higher levels of chlorophyll a (Davis et al. 1997). Because upwelling at the LEO-15 sites is associated with high POC and chlorophyll a values (Glenn et al. 2004), it is possible that growth of recently settled surfclams during the July upwelling (July 14-23) was enhanced by an increased flux of food particles to the bottom, thus counterbalancing the negative effects of low prevailing temperatures on growth. The LEO-15 research area also represents one of the recurrent hypoxic nodes associated with summer upwelling along the New Jersey coast, but the potential effects of high variance in bottom water [O.sub.2] concentrations during July (Boehme et al. 1998) on both juvenile surfclam growth and mortality have not been studied.

Other studies that have examined early growth of surfclams have generally been conducted over much longer time intervals and in some cases under relatively nonnatural field conditions (e.g., Belding 1910, Starypan 1976, Jones et al., 1978), and the initial lengths of the surfclams were at least several millimeters. In a study of surfclams using the growth bands laid down in the shell, Chintala and Grassle (2001) found that surfclams in the inner shelf environment off New Jersey generally reached shell lengths of 10.3-22.0 mm when the first band was laid down, with no significant differences in mean shell length among nine nearshore zones. Although these shell lengths are considerably less than in some other studies (e.g., Jones et al. 1983), they are consistent with previous estimates for the LEO-15 area, namely that surfclams settling in July 1993 reached a maximum shell length of 16.0 mm by December (Weissberger & Grassle 2003).

One potential complication for field studies of bivalve growth rates using modal length frequency analyses is that size-selective predation may effectively remove particular size classes. There was some suggestion that naticid predators could have selectively removed surfclams 2-5 mm in shell length during the late summer and fall in 1993, and that Asterias forbesi might have targeted clams 5-12 mm in length in October and November of the same year (Weissberger & Grassle 2003). The effects of size-selection by small crustacean predators of surfclams in the LEO-15 research area remain largely unstudied. In late July 1998, there were many fewer surfclams >1,000-[micro]m shell length at Sta. 9 than at Sta. C (Fig. 5). This may be an example of initial predator swamping at Sta. C where juvenile surfclams were an order of magnitude more abundant than at Sta. 9 in early July. This would allow some individuals to find a size refuge from predators preying selectively on the smallest individuals.

This study has provided further evidence of the close relationship between upwelling/downwelling events in June and July with surfclam larval supply and initial recruitment in June and July in the inner shelf environment. In addition, the study has provided realistic field estimates of early growth rates under both upwelling and downwelling conditions. Analyses of length frequency distributions and abundances of juvenile surfclams have provided further estimates of the limits on early growth in the nearshore environment and of the probable impacts of the many surfclam predators. Such knowledge is important to the understanding of surfclam recruitment processes and to management of the surfclam fishery in the Mid-Atlantic Bight. Potentially, this kind of information can have implications for shellfish species in other geographic regions.

ACKNOWLEDGMENTS

The authors thank Rose Petrecca and the IMCS diving team for collecting benthic core samples and for deploying the moored zooplankton pumps. Cheryl Ann Zimmer kindly loaned the zooplankton pumps. Amy Stawicki helped sort the benthic core samples and measure the surfclams, and Alex Figueroa helped sort zooplankton samples. Eric Weissberger provided advice on the use of statistical algorithms in C.A.MAN. Pedro Quijon shared the data for surfclam shell lengths in 1998 and provided helpful comments on an early draft. Comments from anonymous reviewers improved the manuscript. This study was supported by a NOAA/ NURP grant.

LITERATURE CITED

Bayne, B. L. 1965. Growth and the delay of metamorphosis of the larvae of Mytilus edulis (L.). Ophelia 2:147.

Bayne, B. L. 1971. Some morphological changes that occur at the metamorphosis of the larvae of Mytilus edulis. In: D. J. Crisp, editor. Fourth European Marine Biology Symposium. Cambridge, UK: Cambridge University Press. pp. 259-280.

Beaumont, A. R. & M. D. Budd. 1982. Delayed growth of mussel (Mytilus edulis) and scallop (Pecten maximus) veligers at low temperatures. Mar. Biol. 71:97-100.

Belding, D. L. 1910. The growth and habits of the sea clam (Mactra solidissima). Reports of the Commissioner of Fish and Game 1909. Commonwealth of Massachusetts. Public Document No. 25:26-41.

Boehme, S. E., C. L. Sabine & C. E. Reimers. 1998. C[O.sub.2] fluxes for a coastal transect: a time-series approach. Mar. Chem. 63:49-67.

Bohning, D., P. Schlattmann & B. Lindsay. 1992. Computer-assisted analysis of mixtures (C.A.MAN): statistical algorithms. Biometrics 48: 283-303.

Chintala, M. M. & J. P. Grassle. 2001. Comparison of recruitment frequency and growth of surfclams, Spisula solidissima (Dillwyn, 1817), in different inner-shelf habitats of New Jersey. J. Shellfish Res. 20: 1177-1186.

Cody, R. P. & J. K. Smith. 1997. Applied statistics and the SAS programming language. Upper Saddle River, New Jersey: Prentice Hall. 445 pp.

Craghan, M. 1995. Topographic changes and sediment characteristics at a shoreface sand ridge--Beach Haven Ridge, New Jersey. MSc thesis, Rutgers University. 123 pp.

Davis, C. V., K. C. Scully & S. E. Shumway. 1997. Juvenile and yearling growth of Atlantic surfclams Spisula solidissima (Dillwyn, 1817) in Maine. J. Shellfish Res. 16:161-168.

Doherty, K. W. & C. A. Butman. 1990. A time- or event-triggered automated, serial plankton pump sampler. In: D. Frye, E. Stone & A. Martin, editors. Advanced Engineering Laboratory Project Summaries-1989. WHOI Technical Report 20. pp. 15-23.

Glenn, S., R. Arnone, T. Bergmann, W. P. Bissett, M. Crowley, J Cullen, J. Gryzmski, D. Haidvogel, J. Kohut, M. Moline, M. Oliver, C. Orrico, R. Sherrell, T. Song, A. Weidemann, R. Chant & O. Schofield. 2004. Biogeochemical impact of summertime coastal upwelling on the New Jersey shelf. J. Geophys. Res. 109:C12S02.

Haskin, H. H., M. L. Tarnowski & R. R. Schneider. 1981. Report on the surf clam management project for New Jersey, July 1, 1978 to June 30, 1979, unpublished report to the Division of Fish, Game and Shellfisheries, Department of Environmental Protection, State of New Jersey.

Hilborn, R. & C. J. Walters. 1992. Quantitative fisheries stock assessment: choice, dynamics & uncertainty. New York: Routledge, Chapman & Hall. 570 pp.

Jones, D. S., I. Thompson & W. Ambrose. 1978. Age and growth rate determinations for the Atlantic surf clam Spisula solidissima (Bivalvia: Mactracea), based on internal growth lines in shell cross-sections. Mar. Biol. 47:63-70.

Jones, D. S., D. F. Williams & M. A. Arthur. 1983. Growth history and ecology of the Atlantic surfclam, Spisula solidissima (Dillwyn), as revealed by stable isotopes and annual shell increments. J. Exp. Mar. Biol. Ecol. 73:225-242.

Ma, H. 2005. Spatial and temporal variation in surfclam (Spisula solidissima) larval supply and settlement on the New Jersey inner shelf during summer upwelling and downwelling. Estuar. Coast. Shelf Sci. 62:41-53.

Ma, H. & J. P. Grassle. 2004. Invertebrate larval availability during summer upwelling and downwelling on the inner continental shelf off New Jersey. J. Mar. Res. 62:837-865.

Ma, H. J. P. Grassle & R. J. Chant. In press. Vertical distribution of bivalve larvae along a cross-shelf transect during summer upwelling and down-welling. Mar. Biol.

Merrill, A. A. & J. W. Ropes. 1969. The general distribution of the surf clam and ocean quahog. Proc. Nat. Shellfish. Assoc. 59:40-45.

Murawski, S. A. & F. M. Serchuk. 1989. Enhanced shellfish harvesting and its management: the offshore clam fishery of the eastern United States. In: J. F. Caddy, editor. Marine invertebrate fisheries: their assessment and management. New York: Wiley & Sons. pp. 479-506.

Newby, S. G. 2005. Predator related turbulence and the response of juvenile Atlantic surfclams, Spisula solidissima. 2005 Aquatic Sciences Meeting, Salt Lake City, Utah (Abstract).

NOAA/NMFS 2003. 37th Northeast Regional Stock Assessment Workshop (37th SAW). National Marine Fisheries Services, Woods Hole, Massachusetts. pp. 284-433.

Quijon, P. A. & J. P. Grassle (in prep). Naticid snail predation on early post-settlement juvenile surfclams (Spisula solidissima) on the New Jersey inner shelf, 1998-2001.

Reid, R. G. B., R. F. McMahon, D. O. Foighil & R. Finnigan. 1992. Anterior inhalant currents and pedal feeding in bivalves. Veliger 35: 93-104.

Rose, R. A. & S. B. Baker. 1994. Larval and spat culture of the Western Australian silver- or goldlip pearl oyster, Pinctada maxima Jameson (Mollusca: Pteriidae). Aquaculture 126:35-50.

Snelgrove, P. V. R., J. P. Grassle & C. A. Butman. 1998. Sediment choice by settling larvae of bivalve, Spisula solidissima (Dillwyn), in flow and still water. J. Exp. Mar. Biol. Ecol. 131:171-190.

Snelgrove, P. V. R., J. P. Grassle, J. F. Grassle, R. F. Petrecca & H. Ma. 1999. In situ habitat selection by settling larvae of marine soft-sediment invertebrates. Limnol. Oceanogr. 44:1341-1347.

Snelgrove, P. V. R., J. F. Grassle, J. P. Grassle, R. F. Petrecca & K. I. Stocks. 2001. The role of colonization in establishing patterns of community composition and diversity in shallow-water sedimentary communities. J. Mar. Res. 59:813-831.

Starypan, G. L. 1976. Abundance, size distribution, and early growth of Spisula solidissima in the inshore waters of the south Jersey coast. MS thesis, Rutgers University, New Brunswick, NJ. pp. 113.

Stehlik, L. L. 1993. Diets of the Brachyuran crabs Cancer irroratus, C. borealis, and Ovalipes ocellatus in the New York Bight. J. Crustacean Biol. 13:723-735.

Tarnowski, M. L. 1982. Temporal distribution of surf clam larvae off southern New Jersey. J. Shellfish Res. 2:108.

Urban, H. J. 2002. Modeling growth of different developmental stages in bivalves. Mar. Ecol. Prog. Ser. 238:109-114.

Viscido, S. V. 1994. Seasonal and spatial variations in the marine epibenthic decapod crustacean community at Beach Haven Ridge, NJ. MS thesis, Rutgers University, Camden, NJ.

Viscido, S. V., D. E. Stearns & K. W. Able. 1997. Seasonal and spatial patterns of an epibenthic decapod crustacean assemblage in north-west Atlantic continental shelf waters. Estuar. Coast. Shelf Sci. 45:377-392.

Wagner, E. S. 1984. Growth rate and annual shell structure patterns in a single year class of surfclams Spisula solidissima off Atlantic City, New Jersey. MS thesis, Rutgers University, New Brunswick, NJ. 161 pp.

Weinberg, J. R. 1993. Surf clam populations of the Middle Atlantic, Southern New England, and Georges Bank in 1992. NEFSC Lab. Ref. Doc. 93-01. 14 pp.

Weinberg, J. R. 1999. Age-structure, recruitment, and adult mortality in populations of the Atlantic surfclam, Spisula solidissima, from 1978 to 1997. Mar. Biol. 134:113-125.

Weissberger, E. J. & J. P. Grassle. 2003. Settlement, first-year growth, and mortality of surfclams, Spisula solidissima. Estuar. Coast. Shelf Sci. 56:669-684.

HONGGUANG MA, * JUDITH P. GRASSLE AND JEANINE M. ROSARIO

Institute of Marine and Coastal Sciences, Rutgers, The State University of New Jersey, 71 Dudley Road, New Brunswick, New Jersey 08901

* Corresponding author. E-mail: [e-mail]

TABLE 1.
Shell length distribution of surfclams in plankton samples when
high concentrations were present (bivalve larvae >150 pm were
identified). Laboratory studies (Snelgrove et al. 1998) indicate
that larvae are ready to settle at shell lengths (SL) 230-290 pm,
so larvae <230 [micro]m SL were probably not competent, and some
individuals >320 [micro]m SL were likely to be resuspended juveniles.

Station,              <230     230-290    290-320
Date (hours)        [micro]m   [micro]m   [micro]m

Sta. 9
  July 8 (7:00)        0          29         41
  July 9 (19:00)       1          71         24
  July 10 (3:00)       1          38         19
  July 11 (11:00)      0          70         24
  July 12 (11:00)      0          39         42
Sta. C
  July 8 (7:00)        15         39         41
  July 10 (15:00)      0          78         20
  July 11 (19:00)      0          74         13
  July 12 (3:00)       18         70         9
  July 12 (19:00)      3          79         17
  July 12 (23:00)      11         67         20
  July 13 (3:00)       13         76         11
  July 13 (7:00)       20         74         4

Station,              >320
Date (hours)        [micro]m    Total

Sta. 9
  July 8 (7:00)        30        100
  July 9 (19:00)       4         100
  July 10 (3:00)       2          60
  July 11 (11:00)      7         100
  July 12 (11:00)      1          82
Sta. C
  July 8 (7:00)        5         100
  July 10 (15:00)      2         100
  July 11 (19:00)      13        100
  July 12 (3:00)       3         100
  July 12 (19:00)      1         100
  July 12 (23:00)      2         100
  July 13 (3:00)       0         100
  July 13 (7:00)       2         100

TABLE 2.
Analysis of variance of benthic abundances of initial surfclam
recruits (<360 pm shell length) on four sampling dates in July 1998
at three LEO-15 stations. Means with common letters not
significantly different (with an overall significance level a = 0.05).

                        Sum of      Mean
Sources        DF       Squares     Square      F value     P

Date            3           13.37        4.46       65.36   0.0001
Station         2           21.59       10.80      158.39   0.0001
Date x
  station       6            2.18        0.36        5.33   0.0001
Error          69            4.70       0.068

Stations     July 6      July 14     July 23     July 29

Mean comparison within stations
  Sta. 9    1.215 (a)   1.211 (a)   0.750 (b)   0.226 (c)
  Sta. C    2.365 (a)   2.084 (a)   1.146 (b)   0.930 (b)
  Sta. C2   0.613 (a)   0.479 (a)   0.151 (b)   0.050 (b)

Pooled mean comparison among dates
                1.398       1.258   0.682 (b)   0.402 (c)

Pooled mean comparison among stations: (a) 1.631 (Sta. C);
(b) 0.850 (Sta. 9); (c) 0.323 (Sta. C2).