Evaluating northern quahog (= hard clam, Mercenaria mercenaria L.) restoration: are transplanted clams spawning and reconditioning?

ABSTRACT Spawner sanctuaries, harvest-free areas planted with high densities of adult clams, are currently being used to restore self-sustaining populations of Mercenaria mercenaria (L.) to Great South Bay, New York. To evaluate and guide this restoration, we monitored the condition and spawning of clams transplanted from two source locations in Long Island Sound since April 2004. Transplanted clams were in relatively high condition and gonad ripeness at time of transplant, spawned the first spring and/or summer after transplant, reconditioned and spawned in subsequent years, but rarely reconditioned to as high of levels as when they were first transplanted. All populations exhibited similar annual patterns of condition and gonad ripeness: both peaked in mid to late spring, dropped steeply through summer with spawning, and they were lowest in fall. In some years condition increased during fall, and the higher the condition attained by the end of fall (mid-December) the greater the peak in condition the following spring. Across years and populations, condition at the end of fall explained ~89% of the variance in spring peak condition. Consistent differences in condition through time among some populations suggested that location within the bay as well as clam size impact condition. Because of interannual and locational variability, long-term monitoring of this long-lived species is essential for determining factors affecting condition and reproduction, and the ultimate restoration of sustainable hard clam populations.

KEY WORDS: northern quahog, hard clam, Mercenaria mercenaria, bivalve reproduction, restoration, condition index, spawning, spawner sanctuary

INTRODUCTION

Bivalves play a vital role in estuarine ecosystems and are thus an essential component of ecosystem based management. They are responsible for benthic-pelagic coupling by filtering phytoplankton and seston and transporting this organic matter to the benthos, supplementing benthic food webs and accelerating nutrient cycling within the system (reviewed in Dame 1993, Smaal & Prins 1993, Prins et al. 1998). When abundant, the filtering capacity of bivalves can control phytoplankton populations in shallow low-energy embayments, leading to reductions in turbidity (Cloern 1982, Newell 1988, Dame 1993, Gerritsen et al. 1994). Increased water clarity results in greater abundances of sea grasses and other submerged aquatic vegetation, which in turn provides the foundation for diverse communities of invertebrates, fishes, and birds (Phelps 1994, Newell & Koch 2004). In addition, some bivalves act directly as foundation species, providing habitat and foraging grounds for communities of invertebrates and fishes (Wells 1961, Dame 1979, Coen et al. 1999, Peterson et al. 2003).

Many bivalves are also important fishery species, and overexploitation has led to severe declines in bivalve populations in many estuarine systems (Rothschild et al. 1994, Jackson et al. 2001, Peterson 2002, Kirby 2004, Kraeuter et al. 2005). Other factors, such as loss and destruction of habitat (Rothschild et al. 1994, Lenihan & Peterson 1998, MacKenzie 2007), hypoxia (Lenihan & Peterson 1998), harmful algal blooms (Bricelj & Lonsdale 1997), and disease (Ford & Tripp 1996, Ford 2001) also have negatively impacted bivalve populations. Declines in bivalve populations have had several ecological consequences, including reduced cropping of phytoplankton, increased turbidity, increased occurrence and duration of anoxic episodes in bottom waters, and reductions in benthic vegetation and invertebrate, fish, and bird species abundances (Newell 1988, Phelps 1994, Jackson et al. 2001).

Many different strategies have been used to supplement bivalve populations to enhance fisheries including the release of hatchery seed, transplant of bivalves from closed areas to cleaner waters for depuration, modification of bottom types to improve settlement and post settlement survival, and the use of spawner transplants to increase larval production (Kassner & Malouf 1982, Peterson et al. 1995, McHugh 2001). In recent years, an increasing recognition of the ecological importance of bivalves has shifted the focus of bivalve restoration efforts from fishery enhancement to ecosystem-based management (Jackson et al. 2001, Peterson et al. 2003, Coen et al. 2007). There are a growing number of efforts to restore self-sustaining bivalve populations in depleted areas as part of ecosystem-based strategies for restoring and maintaining ecosystem health and function. One such example is an effort currently undertaken by The Nature Conservancy (TNC) to restore the northern quahog, or hard clam, Mercenaria mercenaria, to the Great South Bay (GSB), a barrier island estuary on the south shore of Long Island, NY (Fig. 1).

The hard clam is an important ecosystem and fishery species from the Gulf of St. Lawrence to Florida. During the mid-1970s, the hard clam fishery of GSB provided over half of the nation's hard clam harvest (McHugh 1991, 2001). Because of over-harvest (Buckner 1984, Kraeuter et al. 2005) followed by blooms of the harmful alga, Aureococcus anophagefferens, or "brown tide," which first appeared in 1985 and has sporadically reoccurred ever since (Bricelj & Lonsdale 1997, Gobler et al. 2005), populations collapsed, contributing to a virtual collapse of the fishery. The filtering capacity of hard clam populations in GSB was estimated to have decreased from the equivalent of 40% of the bay per day in the 1970s, to only 4% per day in the 1990s (Kassner 1993). This loss of filtering capacity is suggested as one factor leading to the onset of brown tide blooms in the mid-1980s (Cerrato et al. 2004). A recent analysis of long-term data sets has demonstrated that M. mercenaria has been over-fished to recruitment-limiting levels in parts of GSB, suggesting that populations may not recover without intervention, even if fishing pressure is eliminated (Kraeuter et al. 2005).

[FIGURE 1 OMITTED]

To restore hard clam populations and the ecosystem functions they provide, TNC has been transplanting adult hard clams since April 2004 onto 54 [km.sup.2] of privately owned bottomlands in central GSB where fishing had removed most clams (Fig. 1). These bottomlands constitute about one-fifth the total area of GSB, and the hope is that restoration efforts will have positive impacts on the entire bay. Throughout this property, clams are planted on small plots, typically less than 0.5 ha but as large as 3.5 ha, at densities from 4.3-33.9 clams [m.sup.-2], creating a network of separate transplant populations. As of June 2008, TNC has transplanted ~2.9 million clams into GSB to 50 separate sites. Because the bottomlands are privately owned and harvest is restricted, the entire area is a sanctuary for hard clam spawning and recruitment. The goal is to boost the spawning stock past recruitment limiting levels, thereby allowing hard clams to naturally repopulate and become self-sustaining once again. Clams are transplanted at high densities on numerous small plots, rather than spread out evenly across the property, to help increase fertilization efficiency. Peterson (2002) suggested that the creation of spawner sanctuaries may be a particularly effective restoration strategy for M. mercenaria because of this species' relatively long lifespan, and lack of reproductive senility (Peterson 1983). The longer transplant populations persist, the greater their potential contributions to recruitment, provided energy requirements necessary for spawning are met.

We evaluated the potential effectiveness of hard clam spawner sanctuaries in GSB by monitoring condition index and gonad ripeness of transplanted clams from the time of initial transplants in spring 2004 through the fall of 2007. Our goals were 2-fold: (1) to determine if transplanted clams were spawning and reconditioning each year post transplant and (2) to compare condition and spawning among populations and among years to guide the adaptive management of this restoration effort.

METHODS

Source and Transplant Locations

All transplanted clams were purchased from commercial fishermen and would have otherwise been sold to seafood distributors. Transplant populations originated from one of two source locations in western Long Island Sound: Greenwich Core (GC), CT (73[degrees]34'W, 41[degrees]01'N) or Oyster Bay (OB), NY (73[degrees]30'W, 40[degrees]55'N) (Fig. 1). The initial transplant populations in 2004 were from GC. TNC first transplanted clams from OB in 2005, and continued transplanting clams from both sources each subsequent year. All transplantations from GC were made in April or December because of seasonal shellfish closures in GC from May through November. Clams from OB were transplanted in spring, summer, and fall. Clams were not transplanted January to March from either source.

We monitored the condition index and gonad ripeness of clams transplanted to 7 of the 50 sites on TNC's bottomlands in GSB (73[degrees]06'W, 40[degrees]41'N) (Fig. 1). These transplant populations varied in source population, season and year of transplant, plot size, clam density, and average clam size (Table 1). The transplant populations are designated by the year (04, 05, 06, 07) and season (F = fall, S = spring, Su = summer) of transplant and the source location (e.g., 04FGC, 04SGC, 05SuOB).

All populations were monitored from the time of transplant, except for population 05SuOB that was planted in summer 2005, but not sampled until April 2006. Monitoring continued through mid-December 2007 for all populations except for sites 05FGC and 05SuOB, in which monitoring was terminated at the end of August 2007 because of poor survivorship.

To compare the performance of transplanted clams from the two source locations, pairs of transplant populations were established side-by-side in 2005 (05SuOB and 05FGC), and again in the spring of 2007 (07SOB and 07SGC). The paired populations were planted within 50 m of one another to control for local environmental differences. The 2005 populations were planted in September and December, and the 2007 populations were planted in April and May.

Sample Collection, Handling, and Measurement

At each site a sample of 20 clams was collected with a clam rake from a boat on each sampling date. Sampling frequency varied over the course of the study, both within and among years. In general, sampling was most frequent from April to September, which includes the spawning season for hard clams in Great South Bay (May to August, Kassner & Malouf 1982, Eversole 2001). We sampled approximately once per week during this period in 2004 and 2005, and once every 2 wk in 2006 and 2007 (except for 05FGC and 05SuOB, which were sampled once per week in 2006).

Sampling was less frequent, once per month, during fall (October through December) in 2004 and 2005. We sampled approximately once every 2 wk in the falls of 2006 and 2007, to better describe patterns of fall reconditioning after summer spawning. Winter (January to March) samples were collected in 2007, with 2-3 sampling dates per site.

Sampled clams were kept on ice packs in coolers for 1-2 h until arrival at the laboratory. They were scrubbed clean of sediment and placed in 12[degrees]C to 15[degrees]C temperature-controlled recirculating aquaria until processed, usually within 24 h Aquaria were maintained at this temperature to prevent spawning because of sudden temperature increases, but to allow clams to actively filter. Clams were not removed from aquaria for measurement until after they had gaped and extended siphons to filter, which usually occurred within 2 h; this ensured that the mantle cavity was filled with water prior to assessing whole live weight (see below).

For each clam, we: (1) quantified clam size, (2) computed a condition index, (3) determined sex, and (4) ranked gonad ripeness. The height, length, and width of each clam (Buckner 1984) were measured with calipers to describe size. To compute condition index (see below), we measured (1) whole live wet weight (shell and tissue), (2) wet shell weight, and (3) dry tissue weight. Ranks of gonad ripeness (see below) were based on visual observations of dissected gonad tissue, as well as microscopic examination of wet smears of gonad tissue, which were used to verify sex.

Condition Index

Condition index (CI) is a ratio of tissue weight to internal shell cavity capacity and provides a measure of the nutritive status of a bivalve (Crosby & Gale 1990). Seasonal changes in condition index are associated with reproductive cycles in several bivalve species including M. mercenaria (Ansell et al. 1964, Ansell & Lander 1967). Condition index increases as gonad tissue proliferates and ripens, and drops during spawning caused by loss of tissue mass as gametes are released. These changes in condition with changes in gonad mass reflect the large fraction of total annual organic growth, between 30% and 50%, allocated to reproduction by M. mercenaria (Eversole 2001). We considered peak spring condition to be an important gauge of spawning potential and fecundity in transplant populations.

A variety of methods have been used to measure condition index, making comparisons between studies extremely difficult (Crosby & Gale 1990, Ranier & Mann 1992). Crosby and Gale (1990), evaluated the three primary methods used for the computation of CI, and recommended using gravimetric measures of internal shell cavity capacity. Following this recommendation, we measured condition index using the formula of Lawrence and Scott (1982):

Condition Index (CI) = dry soft tissue wt (g) x 100/ internal shell cavity capacity (g)(1)

The internal shell cavity capacity was determined by subtracting the shell weight from the total whole live weight. Clams were made to close underwater prior to weighing so that the mantle cavity was completely filled with seawater.

Soft tissue was collected in preweighed aluminum trays and dried in convection drying ovens at 75[degrees]C for one week before weighing. The trapped seawater inside clams was discarded, but body fluids exuded during tissue removal and dissection were collected and dried with the soft tissue. Wet shells were blotted dry and weighed immediately after opening and removal of all tissue. A preliminary analysis of 502 clams found a strong linear correlation between wet and oven-dried shell weights ([r.sup.2] = 0.9997); with dry shells about 2.8% lighter than wet shells. Patterns of CI in M. mercenaria are therefore identical for measurements made with wet or dry shell weights, with only a small difference in the magnitude of CI. Abbe and Albright (2003), found a similar tight relationship between wet and dry shell weights in the oyster Crassostrea virginica, and recommended that wet shell weights be used because of improved accuracy and reduced sample processing rime.

Gonad Ranking

The gonad tissue of hard clams is intertwined with digestive and other tissues, and therefore, it is not possible to cleanly remove and weigh separately. To quickly and reliably estimate reproductive status of clams, we ranked gonad ripeness ranging from 0 (least ripe) to 4 (most ripe). Ranks were based on 4 attributes: (1) the amount of gonad tissue relative to other tissues within the viscera, (2) the general texture and appearance of the gonad tissue, (3) the relative number of gametes observed in a tissue smear, and (4) for females, the appearance of eggs (Table 2). In general, as gonads ripened, gonad tissue proliferated, became of the dominant tissue type, and eventually bulged the body wall. When ripe, hard clam gonad tissue was typically white, but sometimes appeared very light brown. Ripe tissue was soft, easily removed with forceps, and discharged a milky liquid packed with gametes. In unripe states, little or no gonad tissue was apparent, tissue was harder to remove with forceps, and gametes were either not present or were found in low concentrations. In fall and winter, undischarged eggs were often empty or had reduced contents, most likely caused by resorption.

To assess gonad rank, the viscera was removed from the shell and cut parallel to the gape. The amount, appearance, and texture of gonad tissue were visually inspected and forceps were used to extract a piece of gonad tissue to smear on a slide. The wet smear was examined under a compound microscope at x200 to determine sex by the presence of eggs or sperm. If no gametes were found, the sex was recorded as unknown, and a gonad rank of 0 assigned. Otherwise, gamete concentration was categorized as low (gametes cover ~ <25 % of field of view), moderate (~25-75%), or high (~ >75%), and the presence of abnormal eggs (those with reduced contents) was noted.

Statistical Analyses

The software package Statistica (verson 6.1, StatSoft Inc.) was used for statistical tests. Single factor ANOVAs followed by Tukey HSD tests to examine pair wise differences were used to test for differences in CI of clams among populations and among dates. This analysis was used to compare clam condition among transplant dates, to compare condition between newly transplanted clams and clams that had spent at least one spawning season in GSB, and to test for differences in peak spring conditions among populations within years, as well as among years within populations.

To test for correlations between gonad rank and CI for each transplant population over time we used a nonparametric Spearman ranked correlation test. Pearson's product-moment correlation coefficient was computed to test for the correlation between clam condition at the end of fall (mid-December) and peak condition the next spring.

Repeated-measures ANOVAs were used to compare the condition of clams between transplant populations through time. Separate tests were used to compare populations 05SuOB and 05FGC, and populations 07SOB and 07SGC, which were planted side-by-side to test for the effects of different source populations whereas controlling for local environmental differences. A repeated-measures ANOVA also was used to compare condition through time between 2 populations from the same source (GC) transplanted to opposite sides of GSB in the spring and fall of 2004 (04SGC and 04FGC; Figure 1) to test for local environmental differences. These populations were compared from April 2006 through December 2007, after both populations had spent at least one spawning season in GSB and the effects of prior conditioning at the source location had dissipated.

RESULTS

Condition and Spawning in First Year Post Transplant

For 6 of the 7 transplant populations, 4 from Greenwich Cove and 2 from Oyster Bay, CI was measured on the date of transplant, providing an indication of the condition of clams at the source locations in Long Island Sound. The initial condition of clams significantly differed among the 6 transplant populations, and ranged from 7.9-10.6 (F = 13.05, P < 0.0001; Fig. 2). Condition did not differ between source populations (GC and OB) in the same season and year, or between clams from the same source transplanted in December or the following spring prior to the spawning season (Tukey HSD, P > 0.76; Fig. 2). However, condition significantly differed among years of transplant.

[FIGURE 2 OMITTED]

At the time of each transplant, newly transplanted clams were in significantly higher condition than those that had spent at least one spawning season in the bay (Figs. 3 and 4). This was true for new transplants from GC in fall 2004 (F = 58.03, P < 0.0001), fall 2005 (F = 18.48, P < 0.0001), and spring 2007 (F = 26.87, P < 0.0001), and for new transplants from O B in fall 2006 (F = 36.31, P < 0.0001) and spring 2007 (F = 28.74, P < 0.0001). Clams transplanted in fall maintained high condition and gonad ripeness through winter into spring (Figs. 4 and 5). In all cases, clams spawned during the first spring and summer after transplant, as evidenced by steep drops in condition and gonad ripeness (Fig. 4 and 5).

[FIGURE 3 OMITTED]

Condition and Spawning in Subsequent Years

After the first spawning season, transplant populations showed similar and recurring annual patterns of condition index and gonad rank (Figs. 4 and 5). Condition index and gonad rank were significantly correlated in each transplant population, with condition increasing as gonads ripened, and dropping with spawning (P < 0.05; Fig. 5). In general, condition and gonad rank peaked in May to mid-June, then dropped steeply through the summer, and was lowest in the early fall. In 2 of the 4 y (2005, 2006), condition increased appreciably from mid-October through December. However, fall reconditioning was virtually absent in 2004 and 2007. Condition changed little during the cold winter months, with small decreases observed from February to April in 2007. Condition and gonad rank began to rise again in mid-spring, with the steepest rise typically from mid-April into May, leading up to the annual peak in condition. With the exception of population 04SGC in 2006, the transplant populations never reconditioned to their initial transplant level (Fig. 4).

[FIGURE 4 OMITTED]

Whereas seasonal patterns of condition were similar from year to year, levels of condition and gonad rank varied among years (Figs. 4 and 5). Spring peaks in condition varied significantly in population 04SGC over 4 y (F = 18.803, P < 0.0001), in population 04FGC over 3 y (F = 12.30, P < 0.0001), and in populations 05FGC (F = 9.43, P < 0.01) and 05SuOB (F = 23.73, P < 0.0001) over 2 y. In population 04SGC, our longest data set, spring peak condition was significantly higher in 2006, the population's third spawning season in GSB, than in all other years (Fig. 6). In the other 3 populations, spring peak condition was significantly higher in the first spawning season after transplant than in following years (Fig. 6).

Differences in spring peak condition were largely explained by differences in condition at the end of the preceding fall (Fig. 7). The higher the condition attained by the end of fall, the greater the peak in condition the next spring; fall condition explained ~89% of the variance in spring peak condition (r = 0.942, P < 0.0001).

Variation Among Populations

Condition varied within a year among transplant populations. Significant differences in spring peak condition were found between 2 transplant populations in 2005 (F = 33.91, P < 0.0001), 4 transplant populations in 2006 (F = 3.83, P < 0.05), and 7 transplant populations in 2007 (F = 13.35, P < 0.001). These differences were solely driven by differences between newly transplanted clams and those that had been in GSB for at least one spawning season (Fig. 8). Newly transplanted clams almost always had significantly higher peaks in condition. Within each year there was no significant difference in spring peak condition among newly transplanted populations, or among populations in GSB for at least one spawning season.

There were significant differences in condition through time among transplanted populations. We compared clam condition in populations 04SGC and 04FGC from April 2006 through December 2007 with repeated measures ANOVA, after both populations had been in GSB for at least one spawning season and the influences of prior conditioning at source locations had dissipated. These two populations were from the same source location, but transplanted to opposite sides of GSB, with 04SGC on the south side and 04FGC on the north side (Fig. 1). Planting densities and plot sizes also differed substantially between the 2 populations (Table 1). Condition significantly differed between the two populations, with 04SGC maintaining a higher condition than 04FGC through most of 2006 and 2007 (F = 87.46, P < 0.0001; Fig. 9). There was no significant interaction between time and population, indicating that the pattern of change in condition was similar between the two populations.

Clams from the two source populations differed in size. Clams from OB were significantly smaller than GC clams (t = 93.64, P < 0.001). OB transplants averaged 64.34 [+ or -] 8.82 mm (mean [+ or -] SD; n = 1353) in shell length, whereas GC transplants averaged 88.68 [+ or -] 8.00 mm (n = 3854). We used repeated measures ANOVAs to test for the effect of source population on condition by comparing clams transplanted side by side in the same year from the two sources. Separate tests were used to compare populations 05SuOB and 05FGC and 07SOB and 07SGC. Whereas the initial conditions were similar in each pair, condition through time significantly differed between the pair of transplant populations planted in 2005 (F = 59.99, P < 0.0001), and the pair planted in 2007 (F = 102.02, P < 0.0001). In both cases, condition was greater on most dates for clams from GC, the larger clam, than those from OB (Fig. 10).

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

DISCUSSION

Condition and Spawning of Transplanted Clams

We found that adult clams transplanted into a harvest-free sanctuary in Great South Bay spawned, reconditioned, and spawned again for several years after transplant. Spawning was evidenced by steep drops in condition and gonad ripeness that occurred between May and September of each year. Condition typically started to decline in May or June, but was delayed until July in 2004. A complimentary study in 2004 (Perino 2006, Perino et al. 2008) identified hard clam larvae in the plankton of Great South Bay in July and August, coinciding with spawning that year. Histological studies have indicated that spawning can begin as early as late May and continue into August in GSB, with peak spawning in June or July (Kassner & Malouf 1982, Eversole 2001).

We observed large variation in peak CI during spring, just prior to spawning, among years within populations. Highest spring condition was strongly positively correlated with condition at the end of the previous fall (Fig. 7), suggesting that environmental conditions during fall greatly affect subsequent spawning. The extent of reconditioning during fall in GSB was variable, and in some years (2004, 2007) did not occur. Clams at the source locations were in high condition with ripe gonads in December of each year, indicating strong reconditioning after spawning and suggesting a more favorable environment during fall for clam growth at these Long Island Sound locations than in central GSB.

Within each year, spring condition was typically highest in newly transplanted populations, reflecting their prior conditioning at the source locations, and lower in populations that had spent at least one spawning season in GSB. This condition suggests that the greatest single-season spawning contribution of transplanted clams is the first year posttransplant.

Sub-optimal environmental conditions during fall may result in reduced fecundity the following spring, and produce a high variance in interannual recruitment. Many environmental factors are known to affect growth, condition, and reproduction in hard clams, including water temperature and salinity, sediment characteristics, food quantity and quality, water flow, noxious algae, infectious diseases, and biotic interactions (reviewed in Grizzle et al. 2001, Eversole 2001). Greenfield et al. (2005) found substantial differences in phytoplankton communities between OB and GSB and suggested that higher growth rates of juvenile claros in OB were caused by a more nutritional diet. Additional studies, including field and laboratory experiments, are required to determine the specific suite of factors impacting condition and spawning of clams in GSB.

Alternatively, the reduced condition of transplanted clams may reflect local adaptation of the source population. Perhaps the GSB environment is not less favorable, just different, and the transplanted clams are less adapted to these conditions. The condition patterns of GSB native clams compared with transplanted clams are needed to assess the role of local adaptation and how it may affect current restoration strategies.

Lessons for Adaptive Management

The current hard clam restoration effort by TNC in Great South Bay uses information obtained from monitoring and experimental studies to modify strategies and guide implementation to improve chances of success, and thus is an example of adaptive management. Success will ultimately be measured by the size and sustainability of future hard clam populations. Our findings provide useful information that can be used to guide and improve the success of the current restoration strategy.

[FIGURE 8 OMITTED]

Identifying areas where adult clams thrive versus where they do less well will help 2-fold. First, the best areas for clams can be targeted as priority areas for transplant, which can then serve as sources of larvae for other areas in the bay. The potential for larval transport and retention in the Bay is also important. However, at present little is known about dispersal patterns or the potential for larval retention in GSB. Second, differences among sites can be used to determine the dominant factors that affect clam growth and reproduction and to assess the relative impacts of anthropogenic-controlled factors on clam performance (e.g., nutrient loading, habitat alteration). We found that clam condition varied with location within GSB; clams were in better condition in the southern part of the bay relative to a northern transplant site. However, this comparison was confounded by clam density, which differed 5-fold between the sites (4.3 versus 22.9 clams [m.sup.2]). The clams planted at the higher density (north) had consistently lower condition than those at the lower density (south). Other studies have found that high clam densities (80 clams [m.sup.-2], Peterson & Beal 1989, 67 clams [m.sup.-2], Malinowski 1993) inhibit growth and condition in M. mercenaria, possibly because of local food depletion, but that these density effects disappear at 27 clams [m.sup.-2] (Malinowski 1993). The transplant densities in our study were below the threshold where growth would be expected to be inhibited, and therefore unlikely to explain the differences we saw.

[FIGURE 9 OMITTED]

The selection of source populations is also an important consideration when establishing spawner sanctuaries. We found that clams from both sources in Long Island Sound (GC and OB) spawned after transplant and reconditioned for repeated spawning in the transplant sites. In both cases, source clams were in high condition at the time of transplant, which was important for spawning in the first spawning season post transplant.

The two source populations differed significantly in size, with GC clams >33% larger than OB clams (Table 1). For restoration, the optimal size for transplants needs to balance the higher fecundity of bigger clams (Bricelj & Malouf 1980), and the potential longer life of smaller (younger) clams (Buckner 1984). Recent observations in GSB indicate that the smaller OB transplants are likely to suffer greater predator mortality than the larger GC clams (LoBue et al. 2008). The potentially higher fecundity of the larger clams, combined with their lower susceptibility to predation, suggests that for this restoration effort large clams may be more beneficial. More longitudinal data are needed to determine the balance between fecundity, longevity, and size-specific mortality caused by predation for these sites before an optimal transplant size can be determined.

[FIGURE 10 OMITTED]

Clams should be transplanted when they are in high condition to reduce mortality from transplant stress and maximize the first spawning. We found that clams from both sources were in high condition in December and early spring, with no difference in clam condition between these two periods. Clams transplanted in December maintained high condition through winter and reached peak spring condition similar to spring transplants. Thus, late fall and early spring are suitable times to transplant clams from these sources. Summer may be a less favorable time to transplant as condition is reduced during spawning.

Long-term monitoring is essential to help guide the adaptive management of restoration strategies, especially for long-lived species such as M. mercenaria. We found that the condition and spawning of clams was highly variable among years; a single-year snapshot would not provide an accurate or complete picture of the effectiveness of restoration efforts. If we ended our study after the 2004 spawning season, we could have concluded that transplanted clams spawn. If we ended our study after 2 y, after the 2005 season, we may have concluded that transplanted clams only spawn their first year and then fail to recondition (Fig. 4). Four years of monitoring has shown that transplanted clams can indeed spawn and recondition for at least several years post transplant but that condition and spawning are highly variable among years.

Are Spawner Sanctuaries an Effective Restoration Tool?

The use of spawner transplants to supplement natural recruitment of hard clams in GSB dates back at least to the 1960s (Kassner & Malouf 1982) when hard clams were abundant and transplants were used to supplement the fishery. The goal was not only to supplement the larval pool but to extend the spawning season by transplanting clams from colder, northern locations in the hope that they would spawn later than native clams, thereby increasing the likelihood that larvae would encounter favorable conditions and increase potential recruitment. Similar to our study, Kassner and Malouf (1982) found that hard clams transplanted into GSB from western Long Island Sound spawned in their first spring-summer after transplant. Additionally, they found that the transplants spawned at the same time as native clams, and thus there was no extension of the spawning season. They concluded that ar that rime spawner transplants contributed little to recruitment as the number of transplants paled in comparison with native stocks.

The situation in GSB is substantially different today. Presently the transplant effort is not to supplement a fishery, although this may be one benefit of restoration. Rather, present efforts are part of an ecosystem-based management plan to restore depleted populations and the ecosystem services they once provided. Hard clam populations are dramatically smaller today in GSB than they were in the 1970s, as reflected by an ~99% decline in annual harvest levels since the 1976 peak (LoBue et al. 2008). Fishery-independent estimates of clam abundances indicate that populations have declined substantially from 1978 to 2003 (Kraeuter et al. 2005). Extensive surveys of the TNC bottomlands in GSB found an average adult density of just 0.175 clams [m.sup.-2] in 2006, and a total population size of approximately 9.5 million clams (LoBue et al. 2008). As of June 2008, ~2.9 million clams had been transplanted, supplementing native spawning populations on TNC's bottomlands by ~30%. Diver surveys at 6 transplant locations found that posttransplant survival was highly variable across sites and averaged 52.6% after 30 mo (LoBue et al. 2008).

The current contribution of transplanted clams to the total larval pool may be even greater than that suggested just by their numbers. Transplant populations are likely to have higher fertilization efficiencies than native clams because of higher densities producing shorter nearest-neighbor distances. Kraeuter et al. (2005) modeled spawner/recruit relationships for hard clams in GSB, and found that recruitment fails when densities fall to 0.5-0.75 adults [m.sup.2] and suggested that reduced fertilization success at low density could be a cause. The optimal density that balances fertilization efficiency with competition for resources has not been determined for M. mercenaria and will depend on location-specific factors such as mixing and food availability. Experimental studies are needed to determine optimal planting densities for hard clams in GSB.

Previous efforts in the 1960s to 1980s to increase clam populations in GSB did not always protect transplanted clams from harvest. The current hard clam restoration strategy transplants clams into a privately owned harvest-free sanctuary. Thus, the contribution of transplanted clams to annual recruitment may extend for years, as the transplanted clams do indeed survive, recondition and spawn in subsequent years.

The first step toward reestablishing sustainable populations of hard clams is to increase the number of clams that are conditioning and spawning. Other bottlenecks to population recovery may occur at any of the hard clam life stages. Attention must also be paid to larval production, transport and retention, larval survival, growth and settlement, and all of the factors that affect growth and survival of juveniles. This will include the impacts of predation, as well as water and plankton quality. All potential bottlenecks must be identified and alleviated if restoration is to be successful. Monitoring in combination with experimental studies is an essential component of a restoration strategy, especially for long-lived species such as hard clams and is needed to detect success, especially in the light of interannual variability.

ACKNOWLEDGMENTS

We thank the undergraduates who assisted with laboratory processing of clam samples, especially Michael Martinsen, Annie Coccari, Liam Dolgin, Adriane McCoy, and Connie Catalfamo. Paul Bordeau helped with statistical analyses and Adam Starke and Laurie Perino assisted with sampling. We acknowledge the use of the Functional Ecology Research and Training Laboratory. This research was supported by The Nature Conservancy through grants from Suffolk County, the NOAA Community-Based Restoration Program, and the New York State Department of State, and through private donations. This work is contribution number 1174 in Ecology and Evolution from Stony Brook University.

LITERATURE CITED

Abbe, G. R. & B. W. Albright. 2003. An improvement to the determination of meat condition index for the eastern oyster Crassostrea virginica (Gmelin, 1791). J. Shellfish Res. 22:747-752.

Ansell, A. D. & K. F. Lander. 1967. Studies on hard-shell clam, Venus mercenaria, in British waters. III. Further observations on seasonal biochemical cycle and on spawning. J. Appl. Ecol. 4:425-435.

Ansell, A. D., F. A. Loosmore & K. F. Lander. 1964. Studies on the hard-shell clam, Venus mercenaria, in British waters. II. Seasonal cycle in condition and biochemical composition. J. Appl. Ecol. 1:83-95.

Bricelj, V. M. & D. J. Lonsdale. 1997. Aureococcus anophagefferens: Causes and ecological consequences of brown tides in U.S. mid-Atlantic coastal waters. Limnol. Oceanogr. 42:1023-1038.

Bricelj, V. M. & R. E. Malouf. 1980. Aspects of reproduction of hard clams (Mercenaria mercenaria) in Great South Bay, New York. Proc. Natl. Shellfish. Assoc. 70:216-229.

Buckner, S. C. 1984. Aspects of the population dynamics of the hard clam, Mercenaria mercenaria L., in Great South Bay, New York. Ph.D. Dissertation, State University of New York at Stony Brook. 217 pp.

Cerrato, R. M., D. A. Caron, D. J. Lonsdale, J. M. Rose & R. A. Schaffner. 2004. Effect of the northern quahog Mercenaria mercenaria on the development of blooms of the brown tide alga Aureococcus anophagefferens. Mar. Ecol. Prog. Ser. 281:93-108.

Coen, L. D., M. W. Luckenbach & D. L. Breitburg. 1999. The role of oyster reefs in essential fish habitat: A review of current knowledge and some new perspectives. In: L. R. Benaka, editor. Fish habitat: essential fish habitat and rehabilitation. American Fisheries Society, Symposium 22, Bethesda, Maryland, pp. 438-454.

Coen, L. D., R. D. Brumbaugh, D. Bushek, R. Grizzle, M. W. Luckenbach, M. H. Posey, S. P. Powers & S. G. Tolley. 2007. Ecosystem services related to oyster restoration. Mar. Ecol. Prog. Ser. 341:303-307.

Cloern, J. E. 1982. Does the benthos control phytoplankton biomass in San Francisco Bay? Mar. Ecol. Prog. Ser. 9:191-202.

Crosby, M. P. & L. D. Gale. 1990. A review and evaluation of bivalve condition index methodologies with a suggested standard method. J. Shellfish Res. 9:233-237.

Dame, R. F. 1979. The macrobenthic fauna of a South Carolina oyster reef. Proc. Natl. Shellfish. Assoc. 69:6-10.

Dame, R. F. 1993. The role of bivalve filter feeder material fluxes in estuarine ecosystems. In: R. F. Dame, editor. Bivalve Filter Feeders in Estuarine and Coastal Ecosystem Processes. Berlin: SpringerVerlag. pp. 245-269.

Eversole, A. G. 2001. Reproduction in Mercenaria mercenaria. In: J. N. Kraeuter & M. Castagna, editors. Biology of the Hard Clam. Amsterdam: Elsevier. pp. 221-256.

Ford, S. E. 2001. Pests, Parasites, Diseases, and Defense Mechanisms of the Hard Claro, Mercenaria mercenaria. In: J. N. Kraeuter & M. Castagna, editors. Biology of the Hard Clam. Amsterdam: Elsevier. pp. 591-621.

Ford, S. E. & M. R. Tripp. 1996. Diseases and defense mechanisms. In: V. S. Kennedy, R. I. E. Newell & A. F. Eble, editors. The Eastern Oyster Crassostrea virginica. Maryland Sea Grant College, College Park. pp. 581-660.

Gerritsen, J., A. F. Holland & D. E. Irvine. 1994. Suspension feeding bivalves and the late of primary production: An estuarine model applied to the Chesapeake. Estuaries 17:403-417.

Gobler, C. J., D. J. Lonsdale & G. L. Boyer. 2005. A review of the causes, effects, and potential management of harmful brown tide blooms caused by Aureococcus anophagefferens (Hargraves et Sieburth). Estuaries 28:726-749.

Greenfield, D. I., D. J. Lonsdale & R. M. Cerrato. 2005. Linking phytoplankton community composition with juvenile-phase growth in the northern quahog Mercenaria mercenaria (L.). Estuaries 28:241-251.

Grizzle, R. E., V. M. Bricelj & S. E. Shumway. 2001. Physiological Ecology of Mercenaria mercenaria. In: J. N. Kraeuter & M. Castagna, editors. Biology of the Hard Clam. Amsterdam: Elsevier. pp. 305-371.

Jackson, J. B. C., M. X. Kirby, W. H. Berger, K. A. Bjorndal, L. W. Botsford, B. J. Bourque, R. H. Bradbury, R. Cooke, J. Erlandson, J. A. Estes, T. P. Hughes, S. Kidwell, C. B. Lange, H. S. Lenihan, J. M. Pandolfi, C. H. Peterson, R. S. Steneck, M. J. Tegner & R. R. Warner. 2001. Historical overfishing and the recent collapse of coastal ecosystems. Science 293:629-638.

Kassner, J. 1993. Possible effects of reduced hard clam abundance in Great South Bay. On the Water July/Aug: 4-5. Cornell Cooperative Extension of Suffolk County, Riverhead, NY.

Kassner, J. & R. E. Malouf. 1982. An evaluation of "spawner transplants" as a management tool in Long Island's hard clam fishery. J. Shellfish Res. 2:165-172.

Kirby, M. X. 2004. Fishing down the coast: Historical expansion and collapse of oyster fisheries along continental margins. P. Natl. Acad. Sci. 101:13096-13099.

Kraeuter, J. N., S. Buckner & E. N. Powell. 2005. A note on a spawner-recruit relationship for a heavily exploited bivalve: The case of northern quahogs (hard clams), Mercenaria mercenaria in Great South Bay New York. J. Shellfish Res. 24:1043-1052.

Lawrence, D. R. & G. I. Scott. 1982. The determination and use of condition index of oysters. Estuaries 5:23-27.

Lenihan, H. S. & C. H. Peterson. 1998. How habitat degradation through fishery disturbance enhances impacts of hypoxia on oyster reefs. Ecol. Appl. 8:128-140.

LoBue, C. P., C. Clapp & M. H. Doall. 2008. Spawner sanctuaries: A tool for restoring hard clams (Mercenaria mercenaria) and improving water quality in the Great South Bay, Long lsland NY (Phase 2), and advancing the restoration of hard clams (Mercenaria mercenaria) in Long Island estuaries. Final Completion Report to NOAA Community-Based Habitat Restoration Program.

MacKenzie, C. L, Jr. 2007. Causes underlying the historical decline in eastern oyster (Crassostrea virginica Gmelin, 1791) landings. J. Shellfish Res. 26:927-938.

Malinowski, S. M. 1993. Effects of density on adult growth, survivorship, fecundity, and recruitment of quahogs. In: M. A. Rice & D. Grossman-Garber, editors. Proceedings of the 2nd Rhode Island Shellfish Industry Conference. Rhode Island Sea Grant. pp. 79-90.

McHugh, J. L. 1991. The hard clam fishery past and present. In: J. R. Schubel, T. M. Bell & H. H. Carter, editors. The Great South Bay. Stony Brook, NY: SUNY press, pp. 55-63.

McHugh, J. L. 2001. Management of hard clam stocks, Mercenaria mercenaria. In: J. N. Kraeuter & M. Castagna, editors. Biology of the hard clam. Amsterdam: Elsevier. pp. 631-649.

Newell, R. I. E. 1988. Ecological changes in Chesapeake Bay: Are they the result of overharvesting the American oyster, Crassostrea virginica? In: M. P. Lynch & E. C. Krome, editors. Understanding the estuary: advances in Chesapeake Bay Research. Solomons, MD: Chesapeake Research Consortium Publication 129 (CBP/TRS 24/ 88). pp. 536-546.

Newell, R. I. E. & E. W. Koch. 2004. Modeling seagrass density and distribution in response to changes in turbidity stemming from bivalve filtration and seagrass sediment stabilization. Estuaries 27:793-806.

Perino, L. L. 2006. Using PCR to determine the accuracy of morphological identification of Mercenaria mercenaria (L.) larvae. M. S. Thesis, State University of New York at Stony Brook. 30 pp.

Perino, L. L., D. K. Padilla & M. H. Doall. 2008. Testing the accuracy of morphological identification of northern quahog larvae. J. Shellfish Res. 27:1081-1085.

Peterson, C. H. 1983. A concept of quantitative reproductive senility: Application to the hard clam, Mercenaria mercenaria (L.)? Oecologia 58:164-168.

Peterson, C. H. 2002. Recruitment overfishing in a bivalve mollusk fishery: Hard clams (Mercenaria mercenaria) in North Carolina. Can. J. Fish. Aquat. Sci. 59:96-104.

Peterson, C. H. & B. F. Beal. 1989. Bivalve Growth and Higher Order Interactions: Importance of Density, Site, and Time. Ecology 70:1390-1404.

Peterson, C. H., J. H. Grabowski & S. P. Powers. 2003. Estimated enhancement of fish production resulting from restoring oyster reef habitat: Quantitative valuation. Mar. Ecol. Prog. Ser. 264:249-264.

Peterson, C. H., H. C. Summerson & J. Huber. 1995. Replenishment of hard clam stocks using hatchery seed: Combined importance of bottom type, seed size, planting season, and density. J. Shellfish Res. 14:293-300.

Phelps, H. L. 1994. The Asiatic clam (Corbicula fluminea) invasion and system-level ecological change in the Potomac River estuary near Washington, D.C. Estuaries 17:614-621.

Prins, T. C., A. C. Smaal & R. F. Dame. 1998. A review of the feedbacks between bivalve grazing and ecosystem processes. Aquat. Ecol. 31:349-359.

Ranier, J. S. & R. Mann. 1992. A comparison of methods for calculating condition index in eastern oysters, Crassostrea virginica. J. Shellfish Res. 11:55-58.

Rothschild, B. J., J. S. Ault, P. Goulletquer & M. Heral. 1994. Decline of the Chesapeake Bay oyster population: a century of habitat destruction and overfishing. Mar. Ecol. Prog. Ser. 111:29-39.

Smaal, A. C. & T. C. Prins. 1993. The uptake of organic matter and the release of inorganic nutrients by bivalve suspension feeder beds. In: R. F. Dame, editor. Bivalve filter feeders in estuarine and coastal ecosystem processes. Berlin: Springer-Verlag. pp. 271-298.

Wells, H. W. 1961. The fauna of oyster beds with special reference to the salinity factor. Ecol. Monogr. 31:239-266.

MICHAEL H. DOALL, (1) * DIANNA K. PADILLA, (1) CARL P. LOBUE, (2) CHRIS CLAPP, (2) ANNA R. WEBB (1) AND JESSE HORNSTEIN (1)

(1) Department of Ecology and Evolution, Stony Brook University, Stony Brook, New York 11794-5245.

(2) The Nature Conservancy, Long Island Chapter, Cold Spring Harbor, New York

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

TABLE 1.
Characteristics of seven transplant hard clam populations
monitored in Great South Bay. Identifications include the year
and season of transplant and source location, Greenwich Cove CT
(GC) or Oyster Bay NY (OB). Mean length is the average
([+ or -] SD) of all clams sampled from each transplant
population, with sample size in parenthesis.

                                                    Planting
                                                     Density
         Time of                     Plot Size       (clams
ID       Transplant       Source       (ha)        [m.sup.-2])

04SGC    April 2004         GC          3.5            4.3
04FGC    December 2004      GC          0.2           22.9
05SuOB   September 2005     OB          0.1           27.7
05FGC    December 2005      GC          0.1           18.8
06FOB    December 2006      OB          0.4           10.3
07SGC    April 2007         GC          0.2           15.9
07SOB    May 2007           OB          0.2           29.6

           Mean Length [+ or -] SD
ID              (mm) (n)                   Monitoring Period

04SGC    88.05 [+ or -] 7.85 (1573)     April 2004 to Dec. 2007
04FGC    89.98 [+ or -] 7.81 (1207)     Dec. 2004 to Dec. 2007
05SuOB   63.03 [+ or -] 9.17 (724)      April 2006 to Aug. 2007
05FGC    87.82 [+ or -] 8.04 (798)      Dec. 2005 to Aug. 2007
06FOB    68.45 [+ or -] 7.39 (352)      Dec. 2006 to Dec. 2007
07SGC    89.03 [+ or -] 8.85 (276)      April 2007 to Dec. 2007
07SOB    62.55 [+ or -] 7.90 (277)      May 2007 to Dec. 2007

TABLE 2.
Gonad Ripeness Ranking. Ranks from 0-4 were based on several
attributes including the amount, appearance and texture of
gonadal tissue, the amount of gametes in wet tissue smears,
and the appearance of eggs. If gametes covered <25% of the
field of view (under x200 magnification) the concentration
was considered low, moderate if 25-75%, and high when >75%
of the field of view. Abnormal eggs were empty or had
reduced contents.

Gonad
Rank      Amount of Gonadal Tissue        Texture and Appearance

0        Not visible                    Not visible
1        Little or none                 Light-brown to brown
2        Moderate; similar to other     Tough to soft; light-brown
           tissues                        to white
3        Moderate to high; more         Soft; white to light brown;
           prevalent than other           may have milky white
           tissues                        discharge
4        Large, often distending        Soft; white, with milky
           body wall; more prevalent      white discharge
           than other tissues

Gonad                                       Proportion Abnormal
Rank          Gamete Abundance                     Eggs

0        None                           None
1        Low                            Low to high
2        Low to moderate                Low to high
3        Moderate to high               None to moderate
4        High                           None