Life history of the sand shrimp, Crangon septemspinosa Say, in a southern Gulf of St. Lawrence estuary.

ABSTRACT Body size, sex ratio and timing of reproduction of the sand shrimp, Crangon septemspinosa Say (1818), and the distribution of larvae and adults with respect to temperature and salinity, were studied in the estuary of the Kouchibouguac River, New Brunswick. Contrary to predictions, body size was not related to latitude or biogeographic zone. Reproduction in May and June by a cohort of large females, followed in summer and early fall by a cohort of smaller reproductive females, was consistent with previous reports of estuarine populations. Reproduction was continuous throughout the spring and summer, in contrast to the two distinct spawning seasons reported for other northerly populations. Larvae were most abundant in July. At least part of the adult population remained in shallow waters of the estuary throughout the summer, at temperatures up to 28[degrees]C, unlike the offshore migrations reported for most other populations during periods of warm water conditions.

KEY WORDS: shrimp, Crangon septemspinosa, estuary, Gulf of St. Lawrence, Kouchibouguac, larva, maturity, size relationships

INTRODUCTION

The sand shrimp or sevenspine bay shrimp, Crangon septemspinosa Say (1818), is an ecologically important species of coastal and estuarine waters of the western Atlantic. It is generally considered an arctic-boreal species (Haefner 1979), but its range extends from the Strait of Belle Isle (northern Gulf of St. Lawrence) to Florida (Squires 1996). The species' range extends well south of the boundary delineating warm-temperate from arctic-boreal distributions, regardless of whether one accepts that boundary as Cape Cod (Briggs 1974, Ayvazian et al. 1992), or Cape Hatteras (Buzas & Culver 1980). Given this extensive distribution, it is not surprising that life history and reproductive strategies vary among C. septemspinosa populations. Haefner (1979) described a pattern of 2 distinct broods per year (winter and summer) in the northern Gulf of St. Lawrence and Gulf of Maine, 2 "somewhat less distinct" broods (spring and fall) in Delaware Bay, and a single extended season (October to June) in Chesapeake Bay. Corey (1987) interpreted the same data to delineate northern (2 spawning periods/year) and southern (1 spawning season) reproductive zones, separated by a boundary located near 42[degrees] to 44[degrees] N latitude (Massachusetts to Maine). Temperature, often considered to be the most important ecologic factor in the distribution of marine animals (Hall 1964), was implied to be the defining environmental characteristic of these zones (Corey 1987).

Growth rates are also reported to vary throughout the range of C. septemspinosa apparently as a function of temperature and/or latitude. Growth rates were four times higher in Rhode Island waters where summer temperature averaged 20[degrees]C than in Delaware Bay where summer temperature reached 27[degrees]C (Price 1962, Wilcox & Jeffries 1973). This was explained as a consequence of increased metabolic costs at higher temperatures (Wilcox & Jeffries 1973). Alternatively, such a difference in growth rate may reflect counter gradient selection for rapid growth in higher latitudes (Conover 1990, Parsons 1997, Brown et al. 1998). Of the "rules" that have been formulated to explain the effects of temperature or latitude on growth or size, one of the best known is Bergmann's rule, describing increased size of homeotherms as an adaptive response of body surface: volume ratio to low temperatures (Vermeij 1978). Among poikilotherms, temperature-dependent growth rates can lead to larger cell and body size at lower temperatures (Vermeij 1978), so the effects of low temperature or of high latitude on body size may be similar for both homeotherms and poikilotherms although the mechanisms differ. Casanova (1981) therefore proposed an amended "Bergmann's rule" to explain the size relationships of marine invertebrates: "a species reaches its maximal size in high, intermediate or low latitudes of its area if it has a boreal, temperate or tropical distribution (respectively)". As a presumed subarctic-boreal species, Crangon septemspinosa is thus predicted to be larger in the northern portion of its range.

The southern Gulf of St. Lawrence is located within several 100 km of the northern limit of the range of the sand shrimp, however this area has biogeographic and climatic affinities with the waters south of Cape Cod (Bousfield & Thomas 1975). Shared characteristics of the two areas include relatively high (mean >18[degrees]C) summer temperatures, low tidal amplitude, and high proportion of sandy shoreline. The southern Gulf of St. Lawrence is, however, separated from Cape Cod by hundreds of kilometers of mainly rocky coastline with muddy estuaries that are exposed to low (mean <15[degrees]C) summer temperatures and high tidal amplitudes (Bousfield & Thomas 1975). The distribution of warm-water fauna is discontinuous on the Atlantic seaboard, with the southern Gulf described as a "warm-relict pocket" (Bousfield & Laubitz 1972). Dunbar et al. (1980) referred to the extreme southern portion of the Gulf of St. Lawrence, where our study area is located, as the "subtropics" of the Gulf. Even so, the estuaries and coastal waters of the southern Gulf of St. Lawrence, unlike those south of Cape Cod, are ice-covered from November or early December until April.

The purpose of this study is to describe the population structure and some aspects of the life history of Crangon septemspinosa in a small estuary of the southern Gulf of St. Lawrence. Understanding the life history of southern Gulf of St. Lawrence populations is important for several reasons. Crangon has been demonstrated to be important in the food web. It forms a significant proportion of the diet of a number of estuarine fishes, and may also be a major food source for seabirds (Grabe 1980, Gilmurray & Daborn 1981, Lazarri et al. 1989, Blackwell et al. 1995, Robichaud-LeBlanc et al. 1997). In turn, Crangon may be a major predator of small fish (e.g., metamorphosing flatfish) (Keefe & Able 1994). Despite the ecological importance of the species, there have been suggestions that a commercial fishery should be developed in the southern Gulf. In 1971, sampling was conducted to evaluate the possibility of harvesting sand shrimp off Prince Edward Island (Lewis & Shanahan 1971), and in 1997 the species was the target of a small experimental fishery in Chaleur Bay (Hanson & Lanteigne 1999). There are no immediate plans for a commercial fishery of Crangon, but given the poor state of many "traditional" fisheries at present, it is likely that such a fishery will again be proposed. Very little is known of the ecology of this important species in the southern Gulf of St. Lawrence, other than the role of Crangon as a food source. In this article, we discuss body size, inferred age structure, sex ratio and timing of reproduction and the distribution of larvae and adults with respect to temperature and salinity in the Kouchibouguac estuary. Size, age and reproductive timing are compared with published data from other locations, to investigate the effects of gradients in latitude and temperature.

MATERIALS AND METHODS

Study Area

Crangon were sampled in the Kouchibouguac River estuary, located in Kouchibouguac National Park, New Brunswick, Canada (46.83[degrees]N, 65.05[degrees]W). The Kouchibouguac River flows into the Kouchibouguac Lagoon, which is separated from the Northumberland Strait (southern Gulf of St. Lawrence) by a chain of sandy barrier islands (Fig. 1). The estuary formed by the lagoon and the tidal waters (~15 km) of the Kouchibouguac River and several smaller rivers draining into the lagoon covers about 1451 ha (Beach 1988). The bottom sediments of the lagoon and the lower part of the river range from slightly gravelly sands to sandy muds (Beach 1988). Most areas of the lagoon and river are <1.5 m deep at low water, but a navigational channel 2-5 m deep has been dredged through the lagoon and the lower 5 km of river. The tidal range is normally <1 m. Outside of the channel, much of the lagoon and lower river is vegetated with eelgrass, Zostera marina. Shallow areas of the upper tidal river contain widgeon grass, Ruppia maritima.

[FIGURE 1 OMITTED]

Adult and Juvenile Shrimp

Samples of juvenile and adult shrimp were collected with a beach seine (in 1997, 1998 and 1999), Devismes trawl (1999) and epibenthic sled (1999). Beach seining was conducted approximately fortnightly from May 23 to October 24 1997, July 9 to September 29, 1998 and May 15 to October 20, 1999. In 1997 and 1998, samples were collected at three sites near the mouth of the river (Fig. 1). In 1999, the original three sites and seven additional sites in the river and lagoon were sampled. The beach seine (30 m long, 1.8-m bag, 13-mm mesh; in 1997 the bag contained a 1-mm mesh insert) was towed out from shore to a distance of approximately 15 m, parallel to the shore for 15 m, then dragged inshore. This method enclosed ~500 [m.sup.2] of water (Hanson & Courtenay 1995). Maximum depth of capture was ~1 m. An epibenthic sled (30 cm wide by 36 cm high, 1 mm mesh, mounted on runners to sample 12 cm off the bottom) was used fortnightly from June 22 to October 10, 1999 at four sites in the river and eight in the lagoon. Sites ranged in depth (low tide) from ~1 m to 3 m. The channel in the lagoon and river (depth, 3-6 m) was sampled approximately monthly from June 8 to October 26, 1999 using a Devismes trawl (4.9 m headline, 20 mm mesh; Conan et al., 1994). Temperature and salinity of near-surface (<0.5 m) depth and near-bottom (sampled with a Niskin bottle <1 m above bottom) waters were measured at all sites (except seined sites, where only near-surface conditions were measured), using a thermometer and refractometer.

Detailed biological data (length, weight and sex) were recorded in the laboratory for all Crangon sampled in 1997 and 1998, and those sampled by trawl or sled in 1999. Samples containing few Crangon were retained in their entirety, and a subsample of usually 40-50 individuals was collected from larger samples for laboratory analysis. In 1999, all Crangon collected by beach seine were counted and released.

Specimens retained for detailed analysis were placed on ice in a cooler and frozen as soon as possible. Total length (tip of antennal scale to end of telson), carapace length (tip of antennal scale to end of carapace), and wet weight (blotted on paper towels) were measured for thawed shrimp. Dry weight was measured after shrimp were held for 24 h at 70[degrees]C. Regressions of [log.sub.10] (total length) on [log.sub.10] (carapace length) [log.sub.10] (dry weight) on [log.sub.10] (total length) and [log.sub.10] (wet weight) on [log.sub.10] (total length) were calculated.

Adult shrimp were sexed externally, according to characteristics of the first and second pleopods (Squires 1990), and presence of an egg mass was recorded. Individuals of total length (TL) <20 mm could rarely be sexed and were considered to be juveniles, following the practice of Price (1962), Modlin (1980) and Corey (1981). Few shrimp [greater than or equal to] 25 mm TL could not be sexed based on the pleopods.

Because no internal sexual characteristics (e.g., oocyte development) were examined, the breeding season was defined as the period of time when ovigerous females were observed. Strictly speaking, this was the period of time between spawning and hatching during which eggs were incubated. Size at maturity was defined as the minimum size at which females produced eggs.

Larval Shrimp

Larval shrimp were collected throughout the Kouchibouguac estuary in 1997 and 1998. In 1997, six stations were sampled fortnightly from May 20 to August 14, then monthly to November 12. In 1998, sampling at one or two sites was conducted on April 7 and April 30, and five stations were sampled monthly from May 28 to October 14. Seven stations in the lagoon were sampled every 1 to 3 days from July 5 to July 28, 1998. Larval shrimp were sampled using a conical plankton net, 0.5 m in diameter, and with a mesh size of 500 [micro]m. The net was equipped with a flowmeter (model 2030R, General Oceanics, Miami, Florida) to measure the volume of water filtered, and was towed obliquely for 5 min at approximately 0.5-1 m/s. Water from near-bottom (<1 m above bottom) and near-surface (< 0.5-m depth) was collected using a Niskin bottle, for determination of temperature and salinity with a thermometer (model C1067-855, VWR/Canlab, Halifax, Nova Scotia) and a refractometer (model A366ATC, Ben Meadows Co, Atlanta, Georgia), respectively. These measurements were usually supplemented with vertical profiles of temperature and salinity from a CTD (Conductivity-Temperature-Depth) apparatus (SBE19 Seabird Profiler, Seabird Electronics Inc., Belle, Washington).

RESULTS

Adult and Juvenile Shrimp

Adult and juvenile Crangon septemspinosa of total lengths (TL) from 11-84 mm were collected in the three years of study. Females (TL from 24-84 mm) were larger than males (TL from 19-61 mm). Ovigerous females were 32-79 mm TL. Specimens from all dates and sampling methods were pooled to determine relationships between carapace and total length, and between length and weight (Table 1).

Crangon were collected at temperatures of 4[degrees]C to 22[degrees]C and salinities of 2 [per thousand] to 26 [per thousand] in 1997 (Fig. 2). Females dominated the adult population from the start of sampling (day 143, May 23) until day 188 (July 7), after which males predominated. By the end of sampling on day 297 (October 24), only 16% of the adults were female. Ovigerous females were found throughout the season, but the proportion declined throughout the sampling season. Ovigerous females accounted for about half of the adult female population until mid-August (day 227) but decreased to 22% on day 297 (October 24). Juveniles were also found throughout the season. Juveniles accounted for 9% to 18% of the total Crangon population in May and June, but this proportion increased to 51% to 63% in July (days 188 and 201). In August and October, juveniles made up only 10% to 15% of the population.

[FIGURE 2 OMITTED]

There was a clear separation of two cohorts of females (small, nonreproducing females and larger ovigerous females) but only one obvious cohort of males in 1997 (Fig. 3). The two female cohorts were apparent only in the first two samples, on May 23 (day 143) and June 6 (day 157). All females larger than 46 mm TL were ovigerous on May 23, but by June 6 the proportion of ovigerous females had started to decline. There was no clear size separation of female cohorts in samples from June 20 (day 171) to October 24 (day 297). Length frequencies of male Crangon did not clearly indicate more than one cohort on any of the sampling dates.

[FIGURE 3 OMITTED]

In 1998, few Crangon were sampled by seining, probably because anoxic conditions had developed at the sites. Anoxia was attributed to the effects of a floating dock, and became more pronounced at the three original seining sites during each successive year of study from 1997 to 1999 (Klassen, unpublished data). Temperatures and salinities ranged from 14% to 23.5[degrees]C and from 7 [per thousand] to 22 [per thousand] in 1998. Juvenile Crangon were scarce and fewer than 15% of the females were ovigerous. Specimens from 1998 were included in the length-weight regressions but the data will not be further presented here because of the poor environmental conditions and small sample size.

In 1999, Crangon was present in the channel of both the river and lagoon on all dates sampled by Devismes trawl. Temperatures in both areas ranged from 5.5[degrees]C to 23[degrees]C (Fig. 4). Salinities in the lagoon were vertically homogenous at 27 [per thousand] to 30 [per thousand]. Bottom salinities at the river site approximated those of the lagoon but surface salinities were lower, decreasing from 19 [per thousand] to 25 [per thousand] in summer to 5 [per thousand] to 8 [per thousand] in September to October. The proportion of females in the population, and the proportion of the females that was ovigerous, were higher in the lagoon than in the river. As in the beach seined samples of 1997 (Fig. 2), the proportion of females in the population decreased throughout the season. Reproduction was continuous from June 28 (day 179) to August 31 (day 244), but as in 1997 the proportion of reproductive females declined throughout the season and few were collected in September or October. Unlike the situation described for 1997, only one size-based cohort of females was observed in the samples in 1999 (Fig. 5). However, in 1999, sampling did not begin until June 28, and by this date in 1997 it was no longer possible to discern two female cohorts in the beach seining samples. There were more ovigerous individuals among the larger females sampled on June 28, 1999, than among the smaller females, which was consistent with the 1997 samples. As in 1997, only one male cohort was discernable throughout the season.

[FIGURES 4-5 OMITTED]

Crangon were present at all 10 nearshore sites that were seined along the estuarine gradient from May to September 1999, and on all dates (Fig. 6). Abundance varied from 1-417 shrimp in a standardized haul enclosing approximately 500 [m.sup.2]. Maximum abundance occurred around July 22 (day 203). Crangon were collected in waters with surface temperatures of 12[degrees]C to 28[degrees]C and salinities of 4 [per thousand] to 32 [per thousand].

[FIGURE 6 OMITTED]

Larval Shrimp

Larval Crangon were collected throughout the estuary. No larvae were found on the first sampling date, May 20, 1997 (day 140) (Fig. 7), when ovigerous females were at their highest levels (Fig. 2). The first larvae were collected on June 3 (day 154). A peak of larval abundance in mid to late July (days 196 and 210) corresponded to the peak of juveniles in the beach seining samples. The last larvae were collected on August 12 (day 224), although sampling continued until November (day 316).

[FIGURE 7 OMITTED]

Larvae were not present in samples collected through the ice on April 7, 1998 (day 97), or in open-water samples from April 30 (day 120) (Fig. 7). The date of first collection of larvae in 1998 was May 28 (day 148). Larval abundance peaked in mid-July (about day 200). In 1998, larvae were collected two months later into the fall than in 1997 (until the last sampling date, October 14 [day 287]), although autumn water temperatures were similar in the 2 years.

Larvae were collected at temperatures from 9[degrees]C to 27[degrees]C (Fig. 8). Peak abundances occurred in July at 19[degrees]C to 22[degrees]C. Larvae were found in waters with surface salinity ranging from 0 [per thousand to 31 [per thousand], and bottom salinity ranging from 16 [per thousand] to 31 [per thousand]. Mean abundance tended to be higher at salinity >15 [per thousand].

[FIGURE 8 OMITTED]

DISCUSSION

Size of Crangon was not consistent with Bergmann's rule, which predicted size should be greatest in the arctic-boreal part of the species range or decrease with latitude south from the northern Gulf of St. Lawrence. There was no clear north-south gradient in size of either sex (Table 2). Female Crangon at Kouchibouguac were the largest reported in the literature, with smaller females documented to the north and south. Males were within the size ranges documented elsewhere. Overall, the pattern of maximum size and size at maturity we recorded more closely resembled those of the population of Chesapeake Bay/York River than those of populations closer to Kouchibouguac. Mean or median sizes might have provided a more meaningful basis of comparison. They were not used here, because they were likely more biased by sampling method than the measures compared here, minimum and maximum sizes.

Body size of Crangon was also inconsistent with a temperature-based gradient. The annual temperature regimen at Kouchibouguac is intermediate between the colder waters of the northern Gulf of St. Lawrence (Port au Port Bay and Conception Bay) and the Gulf of Maine (Passamaquoddy Bay, Lamoine, Penobscot estuary, and Georges Bank), and the warmer waters of Delaware Bay and Chesapeake Bay. If temperature or climate rather than latitude determined Crangon size then one would expect shrimp from Kouchibouguac to be intermediate in size between those from the warmer and colder waters; this was not observed. It is possible, however, that the migratory behavior of Crangon from Kouchibouguac in response to water temperature was different from that of other populations, as will be discussed later. If so, Kouchibouguac Crangon might be experiencing wanner mean summer temperatures than many of the more southerly populations, and this might account for the large body size.

Seasonal changes in size of ovigerous females were consistent with previous reports of Crangon populations in estuaries, but these seem to differ from observations in coastal or offshore areas. The pattern of reproduction early in the season by a cohort of large females and later reproduction by a cohort of smaller females has been reported in Delaware Bay, Mystic, Passamaquoddy Bay and Conception Bay (Corey 1981, Hadjistephanou 1978, Modlin 1980, Price 1962). By contrast, there appear to be no seasonal trends in size of spawning females at Lamoine or on Georges Bank. The Lamoine population appeared to lack the smaller sized reproductive cohort; ovigerous females were relatively scarce and the smallest ovigerous female was 44 mm TL (Haefner 1972). The opposite situation, lacking a cohort of large females, was reported on Georges Bank. Whiteley (1948) found that ovigerous females of 20-25 mm TL were common in early spring catches in plankton nets from Georges Bank. This differed from estuarine studies, in which small ovigerous females did not become common until much later in the spawning season. The maximum size in the planktonic Georges Bank population was 35 mm TL, only about half the size of the largest ovigerous females reported in most other studies.

Based on the cohorts observed in spring/early summer of 1997, the maximum age of females in the Kouchibouguac population seemed to be 2 y, and of males seemed to be 1 y. There was no evidence of a third cohort of females, although age 3 females have been reported from Mystic (Modlin 1980) and Delaware Bay (Price 1962). The male life cycle in the Kouchibouguac estuary also seemed to be a year shorter than that recorded at Delaware Bay. (Price 1962). As in Kouchibouguac, only two cohorts of females and one of males were observed in Passamaquoddy Bay (Corey 1981). This suggests that the Kouchibouguac and Passamaquoddy shrimp may have taken about a year less to attain sizes similar to or slightly smaller than those recorded in southern portions of the range such as Mystic and Delaware Bay, indicating faster growth rates at higher latitudes, consistent with countergradient growth. Again, the literature indicates a different strategy in growth and maturity for offshore populations. Whiteley (1948) judged that the planktonic population collected from Georges Bank had a life span of about 1 y, and, as mentioned earlier, the maximum size recorded in this spawning population was ~35 mm TL. In many of the inshore populations, females were just starting to reproduce at this size.

The seasonal timing of spawning in the southern Gulf of St. Lawrence differed from populations of adjacent water bodies to the north and south. In Kouchibouguac, the proportion of ovigerous females was highest on the earliest sampling date, May 23; ovigerous females are frequently captured in smelt nets set under the ice of neighboring estuaries during the commercial smelt fishery (J.M. Hanson, pers. comm.). This fishery closes in mid to late March, so ovigerous females must be present at least 2 mo before our planktonic collections indicated the hatching of larvae in late May or early June. In the only other southern Gulf of St. Lawrence Crangon study, Needler (1941) observed that eggs hatched in Malpeque Bay, Prince Edward Island, from mid-June through July. This egg hatching period would have started a few weeks after the timing we observed based on larval distributions. Needler did not indicate when the females became ovigerous. In the northern Gulf of St. Lawrence, Squires (1996) found no ovigerous females in Port au Port Bay in May, but 41% were ovigerous in June, and large numbers of larvae appeared for the first time in June. Unlike our population, there were no ovigerous females present in winter. Hadjistephanou (1978) inferred the presence of ovigerous females in winter in Conception Bay from the appearance of juveniles in spring, but the earliest month he collected ovigerous females was April. In the nearest population studied to the south of Kouchibouguac, ovigerous females collected in Passamaquoddy Bay in March all carried newly extruded eggs, whereas eggs collected from May to August were at various stages of development. This suggested that the first spawning period of the season occurred in or after March (Corey 1981). Ovigerous female Crangon have also been collected from the offshore neuston of the Bay of Fundy in March and October (but not in July-August), but the egg stage was not determined (Locke & Corey 1988). Larvae were present in the offshore Bay of Fundy neuston in July, August and October but not in March (Locke & Corey 1988), which is consistent with Corey's inferences for the Passamaquoddy Bay population. The presence of juveniles and small adults along the Maine coast in April and May, however, led Haefner (1972) to conclude that winter spawning took place in the Gulf of Maine.

There was no evidence in the Kouchibouguac population of the two distinct spawning seasons reported for northerly populations, including those at Port au Port Bay, Passamaquoddy Bay, Lamoine and Mystic River (Corey 1987). At Kouchibouguac, there appeared to be continuous spawning throughout the summer with a single peak in abundances of larvae and juveniles in July and declining proportions of ovigerous females from spring to autumn. In this, the Kouchibouguac population more closely resembled the southern than the northern populations of Crangon. By contrast, Squires (1996) found two peaks in occurrence of ovigerous females (June and August) and Corey (1981) found two peaks in the occurrence of postlarval young in Passamaquoddy Bay, from mid-July to early August and from late August to early October.

In many other populations, offshore migration of adults and juveniles is believed to reduce the exposure of the shrimp to high summer temperatures, but Crangon were collected in the inshore waters of Kouchibouguac at temperatures up to 28[degrees]C. Likewise, Squires (1996) found Crangon to be abundant in Port au Port Bay, which reached 26[degrees]C in summer. By contrast, Haefner (1969) determined that the upper temperature limit for maximum survival of Crangon from Maine waters was 18[degrees]C. In Mystic River, shrimp moved offshore when temperatures exceed 20[degrees]C (Modlin 1980). The temperature during the spawning season in Delaware Bay was >21[degrees]C for most of the season, and apparently did not induce migration to deeper waters (Price 1962). Because we did not simultaneously sample the nearshore and offshore populations, we cannot rule out the possibility that some offshore movement occurred, but evidently at least a portion of the Kouchibouguac population tolerated high summer temperatures. Offshore movement of the larger females as the waters began to warm might account for their disappearance from the nearshore population by June, and is consistent with observations of the Penobscot and Mystic populations (Haefner 1979, Modlin 1980).

Larval Crangon were present at temperatures up to 27[degrees]C, and annual variation in the timing of occurrence of larval Crangon in the plankton was not obviously related to temperature regimes. The earliest observations in 1997 and 1998 were at similar times despite earlier warming of the water column in 1998, and the duration of larval occurrence in the plankton differed by two months in the autumn despite similar seasonal temperatures in the 2 y. Nevertheless, some aspects of the distribution of larvae with respect to temperature in this study were very clear. Larvae first appeared at 9[degrees]C to 10[degrees]C and peak abundances occurred at 19[degrees]C to 22[degrees]C. Temperature relationships observed by Wehrtmann (1994) off Chesapeake Bay were quite different. Wehrtmann found highest larval abundance at 10[degrees]C to 14[degrees]C, a smaller peak in abundance at 2.7[degrees]C to 4.5[degrees]C, and very few larvae at 14[degrees]C to 26[degrees]C.

Adults were found nearshore throughout the entire sampled salinity range, from 0 [per thousand] to 29 [per thousand], although the experiments of Haefner (1969) predicted that the tolerated salinity range should be narrow at high temperatures. This does not seem to have been the case at Kouchibouguac, where the salinity range of Crangon collected at temperatures of 25[degrees]C to 28[degrees]C was potentially as great as 2 [per thousand] to 32 [per thousand]. These were, however, surface salinities measured during beach seining, and the halocline of the Kouchibouguac River typically occurs at a depth of 1 m, which approximates the maximum depth of sampling by beach seine. Thus, Crangon located just below the halocline could have experienced salinities up to 10 [per thousand] higher than the minimum measured salinity (Locke, unpublished data). The presence of Crangon in inshore waters of 0 [per thousand] surface salinity would not have been predicted from the literature; the lowest reported record was at a salinity of 4 [per thousand] (Price 1962, Haefner 1979). Haefner (1976) recorded Crangon from salinities as low as 10 [per thousand] to 12 [per thousand] but noted that they were more commonly caught at salinities >16 [per thousand]. The limit of salinity tolerance for ovigerous females may be in the range of 15 [per thousand] to 17 [per thousand] (Price 1962, Haefner 1976). The intolerance of ovigerous females to lower salinities might account for the higher proportion of ovigerous females in trawl samples from the Kouchibouguac lagoon compared with the estuarine river.

Larvae were found throughout the entire salinity range in the Kouchibouguac estuary, in waters with surface salinities from 0 [per thousand] to 31 [per thousand], but not in waters with bottom salinities <16 [per thousand]. According to the literature, larvae prefer higher salinities. In Chesapeake Bay the newly hatched larvae occurred at 22 [per thousand] to 33.6 [per thousand], and the older larvae at salinities >30 [per thousand] (Wehrtmann 1994). In the Mystic River, larvae were collected from a wide range of salinities, from 9 [per thousand] to 25 [per thousand], but with the largest concentration in surface water that averaged 20 [per thousand] (Modlin 1980). Regnault (1971) determined that larval development was possible from 20 [per thousand] to 30 [per thousand]. In Kouchibouguac, larvae could maintain themselves in salinities of 16 [per thousand] or higher at any depth in the lagoon, or by remaining below the halocline in the river channel virtually anywhere we sampled. We have no data on depth distribution of larval Crangon in the Kouchibouguac estuary. Sandifer (1973, 1975) found all larval stages at all depths in Chesapeake Bay, and indeed the near-surface zone contained 59% of the larvae. He indicated that the vertical distribution of larval Crangon in Chesapeake Bay would result in advection out of the bay by currents and hypothesized that the major recruitment mechanism for bay populations was immigration of juveniles and adults. If, in contrast, larvae in Kouchibouguac needed to stay in bottom waters of the river to satisfy a salinity requirement, this could limit the extent to which they were advected out of the estuary.

The reduction in numbers of Crangon at the beach seined sites in 1998 and 1999 is consistent with reports of the intolerance of the species to anoxia. Ovigerous females that were notably lacking from the site, are reported to be the most sensitive to low oxygen levels (Haefner 1979).

The population dynamics of Crangon septemspinosa in the Kouchibouguac estuary evidently differ in many respects from those of populations elsewhere in the species range. A number of questions about the biology of Crangon in the southern Gulf of St. Lawrence remain unanswered by this study. Perhaps the biggest gap is the lack of simultaneous nearshore and offshore sampling, particularly as it addresses the question of Crangon movements between the estuary and the offshore environment of the Northumberland Strait. Are there seasonal onshore-offshore migrations? To what extent does Crangon use the estuary as habitat as opposed to the coastal area, and is there population exchange between the two? Do Crangon in the southern Gulf of St. Lawrence belong to a metapopulation, or are there distinct populations associated with the offshore areas and with each of the estuaries? Data on the winter ecology of the species, when the estuaries and the gulf are ice-covered, are entirely absent. We have no knowledge of the recruitment of Crangon in the coastal area. We are also lacking data on young juvenile Crangon between the postlarval benthic settlement stage, which is ~4 mm TL (Needler 1941), and the smallest benthic stage we collected, which was 11 mm TL. There have been no studies of feeding of sand shrimp on the prey assemblage of the southern Gulf of St. Lawrence. Finally, we have no knowledge of abundance of the species, and are just starting to understand its life history in this unique Canadian environment.

ACKNOWLEDGMENTS

The authors thank D. Argument, I. Aube, O. Bernard, P. Charbonneau, J. Connop, J. Dumont, M. Knockwood, F. Leblanc, A. Poulin, J. Richard, S. Richardson, J. Rogers, J. Roy and C. Vardy for technical assistance; E. Tremblay and I. Kaczmarska for logistical support; J. M. Hanson and S. Courtenay for comments on the manuscript; and Fisheries and Oceans Canada and Kouchibouguac National Park for financial support.

LITERATURE CITED

Ayvazian, S. G., L. A. Deegan & J. T. Finn. 1992. Comparison of habitat use by estuarine fish assemblages in the Acadian and Virginian zoogeographic provinces. Estuaries 15:368-383.

Beach, H., 1988. The resources of Kouchibouguac National Park: resource description and analysis. Environment Canada Parks. Kouchibouguac National Park, Kouchibouguac, NB. March 1988.

Blackwell, B. F., W. B. Krohn & R. B. Allen. 1995. Foods of nestling Double-crested Cormorants in Penobscot Bay, Maine, USA: temporal and spatial comparisons. Colonial Waterbirds 18:199-208.

Bousfield, E. L. & D. R. Laubitz. 1972. Station lists and new distributional records of littoral marine invertebrates of the Canadian Atlantic and New England regions. National Museum of Natural Sciences, Ottawa. Ontario: Publications in Biological Oceanography 5. 50 pp.

Bousfield, E. L. & M. L. H. Thomas. 1975. Postglacial changes in distribution of littoral marine invertebrates in the Canadian Atlantic region. Proc. Nova Scotian Inst. Sci 27(Suppl. 3):47-60.

Briggs, J. C. 1974. Marine zoogeography. New York: McGraw-Hill Book Co.

Brown, J. J., A. Ehtisham & D. O. Conover. 1998. Variation in larval growth rate among striped bass stocks from different latitudes. Trans. Am. Fish. Soc. 127:598-610.

Buzas, M. A. & S. J. Culver. 1980. Foraminifera: distribution of provinces in the western North Atlantic. Science 209:687-689.

Casanova, J.-P. 1981. Nouvelles formulations des regles ecologiques connues sous le nom de regle de Bergmann et loi de Jordan. J. Plankton Res. 3:509-529.

Conan, G. Y., M. Comeau, C. Gosset, et al. 1994. The Bigouden Nephrops trawl, and the Devismes trawl, two otter trawls efficiently catching benthic stages of snow crab (Chionoecetes opilio), and American lobster (Homarus americanus). Can. Tech. Rep. Fish. Aquat. Sci. 1992. 27 pp.

Conover, D. O. 1990. The relation between capacity for growth and length of growing season: evidence for and implications of countergradient variation. Trans. Am. Fish. Soc. 19:416-430.

Corey, S. 1981. The life history of Crangon septemspinosa Say (Decapoda, Caridea) in the shallow sublittoral area of Passamaquoddy Bay, New Brunswick, Canada. Crustaceana 41:21-28.

Corey, S. 1987. Reproductive strategies and comparative fecundity of Crangon septemspinosa Say (Decapoda, Caridea). Crustaceana 52:25-28.

Dunbar, M. J., D. C. MacLellan & A. Filion. A. & D. Moore, 1980. The biogeographic structure of the Gulf of St. Lawrence. Montreal: McGill University Marine Sciences Centre Manuscript 32.

Gilmurray, M. C. & G. R. Daborn. 1981. Feeding relations of the Atlantic silverside Menidia menidia in the Minas Basin, Bay of Fundy. Mar. Ecol. Prog. Ser. 6:231-235.

Grabe, S. A. 1980. Food of age 1 and 2 Atlantic tomcod, Microgadus tomcod, from Haverstraw Bay, Hudson River, New York. Fish. Bull. 77:1003-1006.

Hadjistephanou, N. A. 1978. Feeding experiments and some aspects of the biology of the sand shrimp, Crangon septemspinosa, in Long Pond, Newfoundland. M.Sc. Thesis, Memorial University, Saint John's, NF, 165 pp.

Haefner, P. A. 1969. Temperature and salinity tolerance of the sand shrimp, Crangon septemspinosa Say. Physiol. Zool. 42:388-397.

Haefner, P. A. 1972. The biology of the sand shrimp, Crangon septemspinosa, at Lamoine, Maine. J. Elisha Mitchell Sci. Soc. 88:36-42.

Haefner, P. A. 1976. Seasonal distribution and abundance of sand shrimp Crangon septemspinosa in the York River-Chesapeake Bay estuary. Chesapeake Sci. 17:131-134.

Haefner, P. A. 1979. Comparative review of the biology of North Atlantic caridean shrimps (Crangon), with emphasis on C. septemspinosa. Bull Biol. Soc. Wash. 3:1-40.

Hall, C. A., Jr. 1964. Shallow-water marine climates and molluscan provinces. Ecology 45:226-234.

Hanson, J. M. & S. C. Courtenay. 1995. Seasonal abundance and distribution of fishes in the Miramichi estuary. Can. Spec. Publ. Fish. Aquat. Sci. 123:141-160.

Hanson, J. M. & M. Lanteigne. 1999. Results of an experimental shrimp fishery in Chaleur Bay, southern Gulf of St. Lawrence. N. Am. J. Fish. Manage. 19:713-723.

Keefe, M. L. & K. W. Able. 1994. Contributions of abiotic and biotic factors to settlement in summer flounder, Paralichthys dentatus. Copeia 1994(2):458-465.

Lazzari, M. A., K. W. Able & M. P. Fahay. 1989. Life history and food habits of the grubby, Myoxocephalus aeneus (Cottidae), in a Cape Cod estuary. Copeia 1989(1):7-12.

Lewis, B. & F. Shanahan. 1971. Sand shrimp assessment, final report, 1971. Unpublished report, Prince Edward Island Department of Fisheries and Fisheries Service, Canada Department of the Environment. 17 pp.

Locke, A. & S. Corey. 1988. Taxonomic composition and distribution of Euphausiacea and Decapoda (Crustacea) in the neuston of the Bay of Fundy, Canada, J. Plankton Res. 10:185-198.

Modlin, R. F. 1980. The life cycle and recruitment of the sand shrimp, Crangon septemspinosa, in the Mystic River estuary, Connecticut. Estuaries 3:1-10.

Needler, A. B. 1941. Larval stages of Crago septemspinosus Say. Trans. Roy. Can. Inst 23:193-199.

Parsons, K. E. 1997. Contrasting patterns of heritable geographic variation in shell morphology and growth potential in the marine gastropod Bembicium vittatum: Evidence from field experiments. Evolution 51: 784-796.

Price, K. S. 1962. Biology of the sand shrimp, Crangon septemspinosa, in the shore zone of the Delaware Bay region. Chesapeake Sci 3:244-255.

Regnault, M. 1971. Croissance au laboratoire de Crangon septemspinosa Say (Crustacea Decapoda, Natantia), de la metamorphose a la maturite sexuelle. Bull. Mus. Nat. Hist. Nat. 42:1108-1126.

Robichaud-LeBlanc, K. A., S. C. Courtenay & J. M. Hanson. 1997. Ontogenetic diet shifts in age-0 striped bass, Morone saxatilis, from the Miramichi River estuary, Gulf of St. Lawrence. Can. J. Zool. 75:1300-1309.

Sandifer, P. A. 1973. Distribution and abundance of decapod crustacean larvae in the York River estuary and adjacent lower Chesapeake Bay, Virginia, 1968-1969. Chesapeake Sci. 14:235-257.

Sandifer, P. A. 1975. The role of pelagic larvae in recruitment to populations of adult decapod crustaceans in the York River estuary and adjacent lower Chesapeake Bay, Virginia. Est. Coast. Mar. Sci 3:269-279.

Squires, H. J. 1990, Decapod Crustacea of the Atlantic coast of Canada. Can, Bull Fish. Aquat: Sci. 221. 532 pp.

Squires, H. J. 1996, Decapod crustaceans of Newfoundland, Labrador and the Canadian eastern Arctic. Can. Tech. Rep. Fish. Aquat. Sci 2359. 234 pp.

Vermeij, G. J. 1978. Biogeography and adaptation: patterns of marine life. Cambridge, MA: Harvard University Press.

Wehrtmann, I. S. 1994. Larval production of the caridean shrimp, Crangon septemspinosa, in waters adjacent to Chesapeake Bay in relation to oceanographic conditions. Estuaries 17:509-518,

Whiteley, G. C., Jr. 1948. The distribution of larger planktonic Crustacea on Georges Bank; Ecol. Monogr. 18:233-264.

Wilcox, J. R. & H. P. Jeffries, 1973. Growth of the sand Shrimp, Crangon septemspinosa, in Rhode Island. Chesapeake Sci. 14:201-205.

Williams, A. B. 1965. Marine decapod crustaceans of the Carolinas. Fish. Bull. 65:1-298.

A. LOCKE, (1) * G. J. KLASSEN, (2) R. BERNIER (1) AND V. JOSEPH (1)

(1) Fisheries and Oceans Canada, Gulf Fisheries Centre, P.O. Box 5030, Moncton, New Brunswick, E1C 9B6, Canada; (2) Tau BioDiversity, 49 Parkindale Rd., Pollett River, New Brunswick, E4Z 3A7, Canada

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

Table 1.
Regression equations ([+ or -] standard error of estimate, SE) of
relationships among total length (TL) in mm, carapace length (CL)
in mm, wet weight (WW) and dry weight (DW) in mg for Crangon
septemspinosa from Kouchibouguac estuary.

                                Independent          Coefficient
     Dependent Variable           Variable          ([+ or -] SE)

[log.sub.10] CL               [log.sub.10] TL   0.765 ([+ or -] 0.058)
[log.sub.10] WW (juveniles,
  males and non-ovigerous
  females)                    [log.sub.10] TL   3.064 ([+ or -] 0.035)
[log.sub.10] WW (ovigerous
  females)                    [log.sub.10] TL   2.783 ([+ or -] 0.074)
[log.sub.10] DW (juveniles,
  males and non-ovigerous
  females)                    [log.sub.10] TL   3.357 ([+ or -] 0.062)
[log.sub.10] DW (ovigerous
  females)                    [log.sub.10] TL   2.911 ([+ or -] 0.102)

                                     Intercept                  [r.
     Dependent Variable            ([+ or -] SE)          N    sup.2]

[log.sub.10] CL                0.884 ([+ or -] 0.013)   1535    0.75
[log.sub.10] WW (juveniles,
  males and non-ovigerous
  females)                    -2.353 ([+ or -] 0.102)   1297    0.86
[log.sub.10] WW (ovigerous
  females)                    -1.813 ([+ or -] 0.087)    357    0.80
[log.sub.10] DW (juveniles,
  males and non-ovigerous
  females)                    -3.565 ([+ or -] 0.179)   1297    0.70
[log.sub.10] DW (ovigerous
  females)                    -2.712 ([+ or -] 0.118)    357    0.70

TABLE 2.
Total (TL) and carapace (CL) lengths of Crangon septemspinosa from
populations listed in order from northern to southern latitudes.

                                             Largest Female

                                Latitude
       Location              ([degrees]N)   TL (mm)   CL (mm)

Port au Port Bay, NF               48          --       12
Conception Bay, NF                 47          77        --
Kouchibouguac, NB                  46          84       14.7
Passamaquoddy Bay, NB              45          --       14.1
Lamoine, ME                        44          69        --
Penobscot estuary, ME              44          --        --
Georges Bank                       41          35        --
Delaware Bay, DE                   38          70        --
Chesapeake Bay/York R., VA         37          78        --
North Carolina                     34          60        --

                                  Smallest
                              Ovigerous Female        Largest Male

       Location              TL (m(mm)   CL (mm)   TL (mm)   CL (mm)

Port au Port Bay, NF             --         4         --       12
Conception Bay, NF               45        --         56       --
Kouchibouguac, NB                32         5         61       12
Passamaquoddy Bay, NB            --         6         --        8
Lamoine, ME                      40        --         69       --
Penobscot estuary, ME            35        --         --       --
Georges Bank                   20-25       --         --       --
Delaware Bay, DE                 --        --         46       --
Chesapeake Bay/York R., VA       31        --         61       --
North Carolina                   25        --         --       --

       Location               Reference

Port au Port Bay, NF         Squires 1996
Conception Bay, NF           Hadjistephanou 1978
Kouchibouguac, NB            this paper
Passamaquoddy Bay, NB        Corey 1981
Lamoine, ME                  Haefner 1972
Penobscot estuary, ME        Haefner 1972
Georges Bank                 Whiteley 1948
Delaware Bay, DE             Price 1962
Chesapeake Bay/York R., VA   Haefner 1976
North Carolina               Williams 1965