Northeastern Naturalist Vol. 20, No. 4 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 609 2013 NORTHEASTERN NATURALIST 20(4):609–623 Occurrence and Distribution of the Freshwater Amphipods Gammarus pseudolimnaeus and Gammarus fasciatus in Southeastern Virginia Lindsey L. Postaski1,3, Gregory M. Capelli1, and Randolph M. Chambers2,* Abstract - The freshwater amphipod, Gammarus fasciatus, and a population that keys to Gammarus pseudolimnaeus are broadly sympatric in southeastern Virginia. By documenting the successful formation of pre-copulatory mate-guarding pairs between individuals collected from Virginia and New York, we confirmed the occurrence of a G. pseudolimnaeus population strongly disjunct from the previously described range in the Mississippi and St. Lawrence River drainage basins. Gammarus pseudolimnaeus appears restricted to high-quality, spring-fed streams that occur at low density across an increasingly fragmented natural landscape in Virginia. Gammarus fasciatus, however, occurs in lakes and streams of developed landscapes that typically are more degraded, and this species does not co-occur with G. pseudolimnaeus in high-quality habitat. Gammarus pseudolimnaeus reproduces year-round, with adults and juveniles of all size classes continuously present. Gammarus fasciatus reproduces primarily from February through June, at which point the adults die, and by late summer the population consists solely of immature individuals. If the two species overlapped in distribution, the smaller Gammarus fasciatus amphipods would have to compete with adult G. pseudolimnaeus for resources. We hypothesize that this competitively disadvantageous life cycle could account, in part, for the absence of G. fasciatus in high-quality streams occupied by G. pseudolimnaeus. Introduction The family Gammaridae, found in both marine and freshwater habitats, is the largest and most diverse of the eighty families that make up the order Amphipoda (Bousfield 1977). Gammarids are the only group of amphipods that occupy a broad range of continental freshwaters (Holsinger 1976), with the majority of species occurring in streams, rivers, and lakes of relatively high water quality (Holsinger 1972, MacNeil et al. 2001, Rinderhagen et al. 2000). As mostly shredders or collector-gatherers (Cummins 1973) with population densities of up to 10,000 m-2 (Smith 2001), gammarid amphipods are important in aquatic ecosystems for nutrient cycling (Hanson and Waters 1974) and energy flow (Marchant and Hynes 1981, Newman and Waters 1984). In addition, their sensitivity to a wide variety of pollutants makes them valuable bioindicators (Rinderhagen et al. 2000). Gammarus pseudolimnaeus Bousfield and G. fasciatus Say are two of the four most commonly collected amphipods in the eastern United States (Smith 2001). Gammarus pseudolimnaeus is widely distributed in the Mississippi and St. Lawrence River drainage basins, from Texas and Arkansas north to Wisconsin, 1Department of Biology, College of William and Mary, Williamsburg, VA 23187. 2Keck Environmental Lab, College of William and Mary, Williamsburg, VA 23187. 3Current address - Cardno ENTRIX, Metairie, LA 70002. *Corresponding author - rmcham@wm.edu. 610 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 Northeastern Naturalist Vol. 20, No.4 Ontario, and western Quebec (Bousfield 1958, Holsinger 1976). Throughout its range, G. pseudolimnaeus occupies large rivers, lakes, and ponds (Bousfield 1958). Gammarus fasciatus occurs in sympatry with G. pseudolimnaeus in the Great Lakes region of the United States (Holsinger 1972), and also occupies the St. Lawrence, Hudson, Delaware, and Susquehanna river drainage-basins (Bousfield 1958). Throughout its range, G. fasciatus primarily inhabits lakes and slow-moving, often turbid rivers (Bousfield 1958), but the species may also occur in springs and small streams, especially in the southern part of its range (Holsinge r 1976). We identified amphipods recently collected from numerous isolated populations in southeastern Virginia (VA) as G. pseudolimnaeus (Holsinger 1972, Zehmer et al. 2002). No previously published information is available on G. pseudolimnaeus and G. fasciatus distribution in the region. Preliminary observations suggest that these disjunct, southeastern populations of G. pseudolimnaeus occur in sympatry with G. fasciatus, but with a non-overlapping local distribution. Abiotic factors such as temperature (Smith 1973, Sprague 1963), particle size, current velocity, dissolved oxygen (Rees 1972), biotic factors including food availability (Dobson and Hildrew 1992), and predation (González and Burkart 2004)) can influence local gammarid distributions. Details reported on the reproductive cycles of G. pseudolimnaeus and G. fasciatus tend to vary by study and geographic location. For example, according to Hynes and Harper (1972), G. pseudolimnaeus reproduces from February through July in Ontario, but Miller (1982) described a longer reproductive period lasting from mid-January through September in Wisconsin. Waters (1981) observed an annual life cycle of G. pseudolimnaeus in Minnesota similar to the period reported by Hynes and Harper (1972), suggesting that reproductive activity virtually ceased in winter months. For G. fasciatus, ovigerous females have been observed from May to September (Bousfield 1958), but from February to April in more southern parts of its range (Holsinger 1976). In fact, most life history studies of G. pseudolimnaeus and G. fasciatus were conducted decades ago in the northern portion of their distributional ranges; to date no studies have documented the reproductive cycle of G. fasciatus in southeastern VA, nor of G. pseudolimnaeus recently discovered there. To fill these information gaps, we sought to document the occurrence of G. pseudolimnaeus in southeastern VA by confirming our species identifications using observations of reproductive behavior with conspecifics from the St. Lawrence River drainage basin. We also documented the distribution of G. pseudolimnaeus and G. fasciatus in a small VA watershed, and described the physicochemical conditions in which these species occur. Finally, we documented the timing of reproduction by G. pseudolimnaeus and G. fasciatus, and examined factors influencing the distribution of sympatric gammarid species in local aquatic environments. Field Site Description All studies were completed in a freshwater lake and streams in southeastern VA (Fig. 1). Lake Matoaka is a 16-ha, hyper-eutrophic lake located on the College of William and Mary campus in Williamsburg, VA. The lake is fed by five small Northeastern Naturalist Vol. 20, No. 4 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 611 streams; of these, three are perennial, first-order streams (Strawberry Creek, Pogonia Creek, Berkeley Creek) that enter from the western portion of the watershed which is dominated by forest (72–96%; Wach and Chambers 2007) . These streams arise as springs in which dissolved oxygen is typically near saturation (Zehmer et al. 2002) and temperature does not exceed 25 °C (L. Postaski, unpubl. data). The primary source of organic matter in these streams is allochthonous material from the surrounding deciduous forest preserve (Mahon 1997). We identified the most abundant invertebrate within these streams as the amphipod G. pseudolimnaeus (Zehmer et al. 2002). Three other streams were included for study, all of which are dominated by G. fasciatus (Fig. 1). Crim Dell Creek and College Creek discharge into Lake Matoaka, and both streams are significantly degraded, as evidenced by lower forest cover in their watersheds (25 and 69% forest, respectively; T.M. Russell, Kekc Lab, Figure 1. Site map of study location in southeastern Virginia, showing Lake Matoaka and the six streams sampled for amphipods. 612 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 Northeastern Naturalist Vol. 20, No.4 College of William and Mary, Williamsburg, VA, pers. comm.), and by reduced dissolved oxygen levels, high turbidity from channel erosion, and low species diversity (L. Postaski, unpubl. data). Downstream from Lake Matoaka, Paper Mill Creek is a tributary of College Creek, and it receives runoff from two golf courses and from local roads, with only 40% forest cover in the watershed (T.M. Russell, pers. comm.). The annual water temperature fluctuation in Paper Mill Creek, Crim Dell Creek, and College Creek is greater than the temperature fluctuation observed in the streams from the western portion of the Lake Matoaka watershed (L. Postaski, unpubl. data). Methods Documentation of G. pseudolimnaeus As with many other crustaceans, male gammarids guard females as the females approach molting. In G. pseudolimnaeus, males use their gnathopods to attach themselves to the dorsal side of a female approximately four days before the female molts. The pair remains intact until the female begins to molt; after molting the male fertilizes eggs released into the marsupium. Because this behavior is considered species-specific, formation of pre-copulatory mate guarding (PCMG) pairs between suspected and known G. pseudolimnaeus individuals was used to verify the species’ occurrence in southeastern VA. In October 2008, PCMG pairs of amphipods from southeastern VA were collected from Strawberry Creek using a hand-held aquarium net. PCMG pairs of G. pseudolimnaeus from the St. Lawrence River drainage basin were collected similarly from Spring Creek, a perennial, first-order stream in the town of Caledonia, NY (see Sutton 1995). From each collection, we separated males from local PCMG pairs and transported them in continuously oxygenated containers (water temperature ≤25 °C) to the other state, where we collected females from local PCMG pairs. Then, we placed 5 transported males and 1 local female in a 12- cm x 12-cm plastic container with equal amounts of water from the respective original streams to a depth of 5 cm. We observed amphipods until mate guarding occurred, and then the PCMG pair was transferred to another container of the same type and observed for 1 h. We conducted 25 trials with VA males in NY. Due to higher mortality of NY males in transport, only 10 trials of NY males in VA were conducted. We also conducted 25 trials using suspected VA G. pseudolimnaeus paired with VA G. fasciatus (males with females, watched for one hour for evidence of PCMG pair formation). Gammarid amphipod distribution We compared the environments in five streams and in Lake Matoaka where gammarid amphipods occurred (College Creek was not included). We sampled each stream at upper, middle, and lower reaches; in Lake Matoaka, we sampled along the shoreline. Along each reach, we used a random numbers table (1–10) to determine the distance (meters) traveled downstream or along shore to a sampling location. We took samples from the area with the largest amount of leaf litter within one meter of Northeastern Naturalist Vol. 20, No. 4 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 613 the randomly chosen sampling location. In June and November 2008, and February 2009, we collected amphipods at each site using a 20-cm-diameter metal cylinder. We drove the cylinder into the sediment, then visually gridded it, and estimated the percentage of leaf litter cover within the cylinder to the nearest 10%. Amphipods were then collected by disturbing the substrate within the confined area using a handheld aquarium net and sweeping the area to collect the dislodged individuals. Sweep sampling in the cylinder continued until three consecutive sweeps yielded no amphipods. We preserved amphipods in 70% ethanol for subsequent identification. We also recorded dissolved oxygen, temperature, conductivity, and calcium from the middle of each stream site and from the lake surface during the three sampling periods. Amphipod abundance by species was plotted as a function of percentage of leaf-litter cover (arcsine square-root transformed), and analyzed using Pearson correlation. Timing of reproduction Conclusions regarding timing of reproduction and other life-history information were determined based on the results of year-long sampling for gammarid amphipods in local streams. We collected samples weekly from 18 March 2008 to 16 March 2009 from the same sampling reach along the Lake Matoaka shoreline and from Strawberry Creek. Organisms were collected by scooping leaf-litter patches with a D-frame net (0.3m-wide at base, 750-μm mesh). We finished sampling when a maximum of three net samples was taken, or when collections yielded at least 100 individuals. We used existing taxonomic keys to confirm identification of amphipod species (Holsinger 1972, Smith 2001). Body length, from the tip of the telson to the base of the antennae, was measured using a micrometer. We recorded the number of amphipods engaged in pre-copulatory mate guarding for each species in each sample. Based on preliminary data on relative size, amphipods were categorized as adults (body length ≥8.0 mm) or juveniles (body length <8.0 mm). Prior studies from more northern locations grouped G. pseudolimnaeus and G. fasciatus into adult and juvenile categories ≥6 mm and <6 mm, respectively (Hynes 1955, Hynes and Harper 1972, Miller 1982). In our VA samples, however, mean body length of G. fasciatus PCMG individuals (i.e., sexually mature) was 9.9 mm (n = 100); while the mean body length of G. pseudolimnaeus PCMG individuals was 9.3 mm (n = 100), suggesting that a body length ≥8 mm was an appropriate designation for adult G. pseudolimnaeus and G. fasciatus. Monthly variation in the percentage of adults and juveniles for each species was determined with a chisquare goodness-of-fit test. In June 2009, when sampling in Lake Matoaka yielded very few adult G. fasciatus, sediment samples were collected from the lake bottom to locate smaller juvenile amphipods (≈2 mm body length). Using a plastic scoop, a 500-cm3 sediment sample was collected from the top 1–4 cm of lake sediment. Samples were sorted, and amphipods were preserved in 70% ethanol for measurement separate from the weekly tallies of adult and juvenile amphipods. 614 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 Northeastern Naturalist Vol. 20, No.4 Results and Discussion Documentation of G. pseudolimnaeus In all VA/NY trials, a PCMG pair developed within 3 minutes and was sustained for the 1-h observation period. This rapid occurrence of sustained PCMG behavior between NY and VA populations is strong evidence that the disjunct populations are conspecific. In contrast, no PCMG pairs formed in the 25 trials between VA G. pseudolimnaeus and VA G. fasciatus. The occurrence of interspecific PCMG in amphipods has never been documented. The European species Gammarus pulex L. and Gammarus duebeni Lilljeborg will not form PCMG pairs (Dick and Elwood 1992), and in previous field trials from our study site, G. pseudolimnaeus and G. fasciatus did not form PCMG pairs (G.M. Capelli, pers. observ.). As with mating systems in general, interspecific pairing should be strongly selected against to prevent inappropriate wasted reproductive effort. Species-specific pheromones are probably involved in amphipod signaling (Dunham 1978), and interspecific morphological differences also may be sufficient to prevent PCMG (J. Holsinger, Old Dominion University, Norfolk, VA, pers. comm.). Although documentation of viable offspring from interstate pairs would be more fully conclusive, mate guarding provides strong evidence that amphipods from VA and NY are conspecific. That these disjunct populations are conspecific raises questions about the spatial separation of habitat. Gammarus pseudolimnaeus is not known to occur within any other area of the Susquehanna River drainage except for a small tributary of the Patuxent River in Maryland, that might support a population (G. Capelli, pers. observ.). Perhaps G. pseudolimnaeus was at one time much more widely distributed both locally and probably regionally as well. In southeastern VA, most headwater streams in which G. pseudolimnaeus now occurs flow into either impoundments or other streams with somewhat degraded environmental conditions, e.g., with silty substrates, higher summer temperatures, reduced oxygen, and generally poorer water quality. In addition, stream habitats where G. pseudolimnaeus may have occurred historically would have been more broadly connected (Larsen 1998), because the entire Susquehanna River drainage area from NY to VA was freshwater until the end of the last ice age about 15,000 years ago. Since then, with warming global temperatures, sea level rise, and formation of the Chesapeake Bay estuary, freshwater tributaries to the Susquehanna River have been effectively segregated from one another. Thus, the current distribution of G. pseudolimnaeus in VA may reflect, at least in part, the fragmentation of habitat driven by warming temperatures, estuary formation, and most recently, general alteration of aquatic habitats by humans. As an alternate to invoking habitat fragmentation to account for the disjunct populations in VA, the occurrence of G. pseudolimnaeus in the Chesapeake Bay basin could be the result of recent introduction. Little is known about the passive transport of amphipods from one drainage basin to another, and amphipods are not generally adapted to withstand drought and other adverse environmental conditions (Smith 2001). Some proposed dispersal methods include transport on the feet of waterfowl (Figuerola and Green 2002), transport on the fur of aquatic mammals (Peck Northeastern Naturalist Vol. 20, No. 4 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 615 1975), and arrival on aquatic plants and stocked fish (Mills et al. 1993). Evidence for each mechanism, however, is anecdotal, with little quantitative information available. Although other species have been introduced via human activities including fishing, we have no evidence linking the occurrence of G. pseudolimnaeus to any of them. Gammarid amphipod distribution The abundance of G. pseudolimnaeus and G. fasciatus fluctuated over the three sampling dates and was greatest for both species in November (Table 1). Our results suggest that variable seasonal abundance may be connected to availability of deciduous leaves in the streams either as a food source or as cover from predators; the number of amphipods present and leaf-litter coverage were positively correlated (Fig. 2; r = 0.76, P < 0.001). Although we did not measure it, the number of amphipods could vary seasonally or spatially not only with relative leaf abundance, but also with leaf quality. For example, Carrick et al. (2011) found the amount of algae growing on leaves covaried with the abundance of gammarid amphipods. Finally, amphipod numbers are lower and individual size is greater in sections of Lake Matoaka streams occupied by fish (Wach and Chambers 2007), suggesting possible top-down regulation of amphipod abundance. Table 1. Number of amphipods and associated water quality parameters (mean ± s.d.) at six sites sampled in June and November 2008, and February 2009. G.p. = total number of G. pseudolimnaeus in three samples comprising 942 cm2; G.f.= total number of G. fasciatus in three samples comprising 942 cm2; temperature in ºC; dissolved oxygen in mg L-1; conductivity in μS cm-1; calcium in mg L-1. Month/Site G.p. G.f. Temp. Oxygen Conductivity Calcium JUN 2008 Strawberry Creek 331 0 19.4 ± 1.5 8.9 ± 0.5 221 ± 31 126 ± 10 Berkeley Creek 207 0 19.3 ± 1.6 8.9 ± 0.7 70 ± 32 26 ± 17 Pogonia Creek 209 0 19.1 ± 0.8 8.9 ± 0.6 62 ± 34 38 ± 13 Lake Matoaka 0 108 26.5 ± 0.3 3.0 ± 4.5 277 ± 21 111 ± 12 Crim Dell Creek 0 219 25.3 ± 0.5 5.2 ± 0.9 757 ± 136 255 ± 17 Paper Mill Creek 0 588 25.8 ± 0.2 8.1 ± 0.6 1050 ± 40 137 ± 49 NOV 2008 Strawberry Creek 576 0 11.3 ± 2.3 9.6 ± 1.5 250 ±32 193 ± 32 Berkeley Creek 532 0 12.1 ± 1.0 8.8 ± 0.8 146 ± 58 40 ± 18 Pogonia Creek 338 0 9.1 ± 0.5 9.5 ± 0.7 136 ± 29 44 ± 14 Lake Matoaka 0 257 19.3 ± 2.9 7.9 ± 0.7 274 ± 19 117 ± 16 Crim Dell Creek 0 367 12.2 ± 0.5 5.4 ± 1.0 759 ± 99 207 ± 60 Paper Mill Creek 0 232 9.6 ± 2.9 9.1 ± 0.7 1181 ± 189 152 ± 27 FEB 2009 Strawberry Creek 212 0 10.1 ± 1.8 8.9 ± 0.3 240 ± 29 142 ± 18 Berkeley Creek 192 0 8.1 ± 1.2 9.0 ± 1.1 130 ± 27 34 ± 15 Pogonia Creek 232 0 7.9 ± 1.2 8.8 ± 0.4 60 ± 26 48 ± 14 Lake Matoaka 0 159 6.9 ± 1.2 11.6 ± 4.4 263 ± 42 105 ± 12 Crim Dell Creek 0 207 8.9 ± 1.5 5.6 ± 1.1 598 ± 90 192 ± 57 Paper Mill Creek 0 415 6.9 ± 1.9 7.0 ± 1.8 959 ± 281 154 ± 35 616 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 Northeastern Naturalist Vol. 20, No.4 In comparing the distribution of the two gammarid amphipods in southeastern VA, we found the species were completely isolated with no syntopic occurrences at the local level (Table 1). Gammarus pseudolimnaeus was exclusive to Berkeley Creek, Pogonia Creek, and Strawberry Creek; G. fasciatus was exclusive to Lake Matoaka, Crim Dell Creek, and Paper Mill Creek. Temperature and dissolved oxygen content tended to be more variable at sites containing G. fasciatus, and we considered these to be lower quality sites (Table 1), although we have no data comparing, for example, nutrient or trophic status of the sites (e.g., Dodds et al. 1998). Measurements of conductivity and calcium appeared to be within tolerable ranges of both species (Zehmer et al. 2002). Gammarus fasciatus often is associated with habitats with lower water quality relative to those occupied by G. pseudolimnaeus, but we suspect G. fasciatus should be able to colonize areas of higher water quality. Smith (1973) describes G. fasciatus as surviving at temperatures that range from 10–30 °C, whereas the much narrower range at which G. pseudolimnaeus survives is 15–18 °C. In contrast, the literature suggests that G. pseudolimnaeus would not be able to tolerate the low summer dissolved oxygen levels present in areas where G. fasciatus predominates (Lake Matoaka: 3.02 ± 4.5 mg/L; Crim Dell Creek: 5.2 ± 0.9 mg/L). For example, Hoback and Barnhart (1996) tested the effects of low dissolved oxygen on G. pseudolimnaeus by examining survival under hypoxic conditions. Negative physiological and behavioral Figure 2. Relationship between leaf-litter coverage and amphipod abundance, from monthly sample collections of Gammarus fasciatus and G. pseudolimnaeus (sample n = 162). Percent cover data were arcsine square-root transformed prior to analysis. Northeastern Naturalist Vol. 20, No. 4 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 617 effects, some of which directly affected reproduction, were observed to occur below 6 mg/L (Hoback and Barnhart 1996). The literature also suggests that G. pseudolimnaeus would not be able to tolerate the high summer water temperatures in habitats occupied by G. fasciatus (Lake Matoaka: 26.5 ± 0.3 °C, Crim Dell Creek: 25.3 ± 0.5 °C; Paper Mill Creek: 25.8 ± 0.2 °C). For example, Smith (1973) found that G. pseudolimnaeus could only tolerate temperatures in the range of 21–26 °C for brief periods. Peak reproduction of G. pseudolimnaeus occurs at 18 °C, with higher temperatures resulting in the production of fewer eggs (Smith 1973). Lake Matoaka, Crim Dell Creek, and Paper Mill Creek would not be able to sustain ecologically significant populations of G. pseudolimnaeus because of this physiological constraint. Hynes and Harper (1972) reported a similar temperature restriction on the distribution of Gammarus lacustris limnaeus Smith along the entire length of a spring-fed stream in Ontario. The species was absent from the connected lake and its outflow, where summer temperatures rose into a range generally avoided by G. lacustris limnaeus (Hynes and Harper 1972). Although the absence of G. pseudolimnaeus from some of the habitats we studied may be ascribed to reduced water quality, the restriction of G. fasciatus to the same lower-quality waters may be ascribed to biotic interaction. MacNeil et al. (2000) observed similar, mutually exclusive distributions of Gammarus spp. and Crangonyx pseudogracilis Bousfield within two river systems in Ireland, and concluded Gammarus spp. precluded Crangonyx pseudogracilis from higher quality habitats. We observed strong seasonal variation in stream leaf litter that serves as both food and shelter for amphipods, but we saw no evidence suggesting that differences in the quantity or quality of this food resource (Gee 1988) could account for the differential distribution of the species. However, amphipods may compete for resources (Van Dolah 1978), and many gammarids are predatory or cannibalistic, with larger individuals typically consuming small individuals (McGrath et al. 2007, Polis 1981). Dick et al. (1990) and Dick and Elwood (1992) used mutual, but differential predation to describe the elimination of G. duebeni celticus (Stock & Pinkster) from certain habitats by Gammarus pulex. Although interspecific predation between G. pseudolimnaeus and G. fasciatus has not been documented in the field, size-asymmetric, intraspecific predation has been observed in laboratory settings (L. Postaski, pers. observ.). We hypothesize that habitat partitioning between species is achieved by interspecific interactions that displace G. fasciatus from high-quality habitats occupied by G. pseudolimnaeus and restrict G. fasciatus distribution to habitats with lower water quality. Timing of reproduction The relative percentage of adult and juvenile G. pseudolimnaeus in stream samples did not vary significantly throughout the year (chi-square goodness of fit: P > 0.05), with PCMG pairs found in each sample every month (Fig. 3). In contrast, only juvenile G. fasciatus occurred in all samples throughout the year (Fig. 4); adult G. fasciatus were found every month except September and October. The percentage 618 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 Northeastern Naturalist Vol. 20, No.4 Figure 3. Structure of Gammarus pseudolimnaeus populations in southeastern Virginia streams. Top panel: Monthly relative percentage of adults and juveniles. Middle panel: Monthly percentage of adults forming pre-copulatory mate-guarding pairs. Bottom panel : Monthly size distribution of adults and juveniles. Northeastern Naturalist Vol. 20, No. 4 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 619 Figure 4. Structure of Gammarus fasciatus populations in southeastern Virginia streams. Top panel: Monthly relative percentage of adults and juveniles. Middle panel: Monthly percentage of adults forming pre-copulatory mate-guarding pairs. Bottom panel: Monthly size distribution of adults and juveniles. 620 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 Northeastern Naturalist Vol. 20, No.4 of mating adult G. fasciatus increased from March through August (Fig. 4). In July and August, all adults collected were found in mating pairs. In September and October, however, no adults were found in mating pairs. The relative percentage of adult and juvenile G. fasciatus present in the population varied significantly throughout the year (Fig. 4; chi-square goodness of fit: P < 0.05). Populations of G. fasciatus were dominated by very small individuals (mean body length = 2.2 mm) that were burying into the sediment collected in June from Lake Matoaka. A high percentage of PCMG pairs suggested a peak in reproductive activity of G. pseudolimnaeus during the coldest months of the year (December and January; Fig. 3). Kostalos (1979) similarly reported an increased occurrence of PCMG pairs of Gammarus minus Say during the coldest months (December, January, and February) in a small Pennsylvania stream. In addition to the occurrence of PCMG pairs all year, we also observed fairly equal numbers of adults and juveniles throughout the year (Fig. 3), suggesting that the population of G. pseudolimnaeus is continuously replenished. The continuous reproduction we observed is evidence that life history of G. pseudolimnaeus in southeastern VA differs from what has been observed farther north, where reproductive periods occur between mid-January and early fall (Hynes and Harper 1972, Miller 1982). Previous studies have documented G. fasciatus reproducing from March and April through September and November (Bousfield 1973, Hynes 1955). Based on the presence of PCMG pairs, we found reproductive activity of G. fasciatus occurring from November to August with an abrupt cessation in reproductive activity from September through October, concomitant with the decline and subsequent disappearance of adults from the population during these months (Fig. 4). Interestingly, as the percentage of adult G. fasciatus decreased, the percentage of PCMG pairing increased, suggesting male G. fasciatus were spending more time mateguarding to ensure reproductive success—a strategy used by Crustacea as the rate at which females are encountered decreases (Jormalainen 1998). Conclusions We confirmed the disjunct occurrence of G. pseudolimnaeus in southeastern VA. Gammarus fasciatus and G. pseudolimnaeus exhibit non-overlapping distributions in six streams from this region. Gammarus pseudolimnaeus occupies cool, spring-fed streams with high water quality, whereas G. fasciatus occupies relatively warm, less-oxygenated waters. Although typically given a higher tolerance value for stream conditions (e.g., Bode et al. 1996, Maxted et al. 2000), Gammarus fasciatus should be able to survive the physicochemical conditions where G. pseudolimnaeus—with a lower tolerance value (Hilsenhoff 1987)—occurs, but differences in life history may convey a competitive advantage to G. pseudolimnaeus. Gammarus fasciatus populations are dominated by juveniles with few to no adults in September and October, whereas adult G. pseudolimnaeus are present and reproduce year-round. Whether via more efficient resource exploitation, interference competition, or interspecific predation (Dick et al. 1993), the outcome of biotic interaction is that the species partition habitats, with G. pseudolimnaeus occupying higher-quality streams. Northeastern Naturalist Vol. 20, No. 4 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 621 Acknowledgments We thank the William and Mary Biology Department for financial support and Timothy Russell for logistical support. Literature Cited Bode, R.W., M.A. Novak, and L.E. Abele. 1996. Quality assurance work plan for biological stream monitoring in New York State. New York State Department of Environmental Conservation, Albany, NY. Bousfield, E.L. 1958. Fresh-water amphipod crustaceans of glaciated North America. Canadian Field Naturalist 72:55–113. Bousfield, E.L. 1973. Shallow-water Gammaridean Amphipoda of New England. Cornell University Press, Ithaca, NY. 312 pp. Bousfield E.L. 1977. A new look at the systematics of gammaridean amphipods of the world. Crustaceana Supplement 4:282–316. Carrick, H.J., K.L. Dananay, R.A. Eckert, and K.J. Price. 2011. Decomposition during autumn foliage leaf-fall in wetlands situated along a biogeochemical gradient in Pennsylvania, USA. Freshwater Ecology 21:1–17. Cummins, K.W. 1973. Trophic relations of aquatic insects. Annual Review of Entomology 18:183–206. Dick, J.T.A., and R.W. Elwood. 1992. Coexistence and exclusion among Gammarus species: Behavioural avoidance of interspecific precopulation by male G. pulex (Amphipoda). Oikos 64:541–547. Dick, J.T.A., D.E. Irvine, and E.W. Elwood. 1990. Differential predation by males on moulted females may explain the competitive displacement of Gammarus duebeni by G. pulex (Crustacea: Amphipoda). Behavioral Ecology and Sociobiology 26: 41–45. Dick, J.T.A., I. Montgomery, and R. Elwood. 1993. Replacement of the indigenous amphipod Gammarus duebeni celticus by the introduced G. pulex: Differential cannibalism and mutual predation. Journal of Animal Ecology 62:79–88. Dobson, M., and A.G. Hildrew. 1992. A test of resource limitation among shredding detritivores in low order streams in southern England. The Journal of Animal Ecology 61:69–77. Dodds, W.K., J.R. Jones, and E.B. Welch. 1998. Suggested classification of stream trophic state: Distributions of temperate stream types by chlorophyll, total nitrogen, and phosphorus. Water Research 32:1455–1462. Dunham, P.J. 1978. Sex pheromones in Crustacea. Biological Review 53:555 –583. Figuerola, J., and A.J. Green. 2002. Dispersal of aquatic organisms by waterbirds: A review of past research and priorities for future studies. Freshwater Biology 47:483–494. Gee, J.H.R. 1988. Population dynamics and morphometics of Gammarus pulex L.: Evidence of seasonal food limitation in a freshwater detritivores. Freshwater Biology 19:333–343. González, M.J., and G.A. Burkart. 2004. Effects of food type, habitat, and fish predation on the relative abundance of two amphipod species, Gammarus fasciatus and Echinogammarus ischnus. Journal of Great Lakes Research 30:100–113. Hanson, D.L., and T.F. Waters. 1974. Recovery of standing crop and production rate of a Brook Trout population in a flood-damaged stream. Transactions of the American Fisheries Society 103:431–439. 622 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 Northeastern Naturalist Vol. 20, No.4 Hilsenhoff, W.L. 1987. An improved biotic index of organic stream pollution. Great Lakes Entomolology 20:31–39. Hoback, W.W., and M.C. Barnhart. 1996. Lethal limits and sublethal effects of hypoxia on the amphipod Gammarus pseudolimnaeus. Journal of the North American Benthological Society 15:117–126. Holsinger, J.R. 1972. Biota of Freshwater Ecosystems Identification Manual No. 5. The Freshwater Amphipod Crustaceans (Gammaridae) of North America. US Environmental Protection Agency, Washington, DC. Holsinger, J.R. 1976. The Freshwater Amphipod Crustaceans (Gammaridae) of North America. Water Pollution Control Research Series. US Environmental Protection Agency, Cincinnati, OH. Hynes, H.B.N. 1955. The reproductive cycle of some British freshwater Gammaridae. Journal of Animal Ecology 24:352–387. Hynes, H.B.N., and F. Harper. 1972. The life histories of Gammarus lacustris and G. pseudolimnaeus in southern Ontario. Crustaceana Supplement 3:329–341. Jormalainen, V. 1998. Precopulatory mate-guarding in crustaceans: Male competitive strategy and intersexual conflict. The Quarterly Review of Biology 73:275–304. Kostalos, M.S. 1979. Life history and ecology of G. minus Say (Amphipoda, Gammaridae). Crustaceana 37:113–122. Larsen, C.E. 1998. The Chesapeake Bay: Geologic product of rising sea level. Available online at http://pubs.usgs.gov/fs/fs102-98/. Accessed 10 January 2011. MacNeil, C., J.T.A. Dick, and R.W. Elwood. 2000. Differential physico-chemical tolerances of amphipod species revealed by field transplantations. Oe cologia 124:1–7. MacNeil, C., W.I. Montgomery, J.T.A. Dick, and R.W. Elwood. 2001. Factors influencing the distribution of native and introduced Gammarus spp. in Irish river systems. Archives of Hydrobiology 151:353–368. Mahon, S. 1997. Distribution and ecology of freshwater amphipods in the Lake Matoaka/ College Woods area. M.A. Thesis. The College of William and Mary, Williamsburg, VA. Marchant, R., and H.B.N. Hynes. 1981. The distribution and production of Gammarus pseudolimnaeus (Crustacea, Amphipoda) along a reach of the Credit River, Ontario. Freshwater Biology 11:169–182. Maxted J.R., M.T. Barbour, J. Gerritsen, V. Poretti, N. Primrose, A. Silvia, D. Penrose, and R. Renfrow. 2000. Assessment framework for mid-Atlantic coastal-plain streams using benthic macroinvertebrates. Journal of the North American Benthological Society 19:128–144. McGrath, K.E., E. Peters, J. Beijer, and M. Scheffer. 2007. Habitat-mediated cannibalism and microhabitat restriction in the stream invertebrate Gammarus pulex. Hydrobiologia 589:155–164. Miller, S.A. 1982. The life history of Gammarus pseudolimnaeus Bousfield in a central Wisconsin stream (Amphipoda, Gammaridea). Crustaceana 43:89–99. Mills, E.L., J.H. Leach, J.T. Carlton, and C.L. Secor. 1993. Exotic species in the Great Lakes: A history of biotic crises and anthropogenic introductions. Journal of Great Lakes Research 19:1–54. Newman, R.M., and T.H. Waters. 1984. Size-selective predation on Gammarus pseudolimnaeus by trout and sculpins. Ecology 65:1535–1545. Peck, S.B. 1975. Amphipod dispersal in the fur of aquatic mammals. Canadian Field Naturalist 89:181–182. Northeastern Naturalist Vol. 20, No. 4 L.L. Postaski, G.M. Capelli, and R.M. Chambers 2013 623 Polis, G.A. 1981. The evolution and dynamics of intraspecific predation. Annual Reviews of Ecology and Systematics 12:225–251. Rees, C.P. 1972. The distribution of the amphipod Gammarus pseudolimnaeus Bousfield as influenced by oxygen concentration, substratum, and current velocity. Transactions of the American Microscopy Society 91:514–529. Rinderhagen, M., J. Ritterhoff, and G.P. Zauke. 2000. Crustaceans as bioindicators. Pp 161–194, In A. Gerhardt (Ed.). Biomonitoring of Polluted Water, Vol. 9. Uetikon-Zuerich, Oldenburg, Germany. 320 pp. Smith, D.G. 2001. Pennak’s Freshwater Invertebrates of the United States: Porifera to Crustacea (4th Edition). John Wiley and Sons, New York, NY. 648 pp. Smith, W.E. 1973. Thermal tolerance of two species of Gammarus. Transactions of the American Fisheries Society 102:431–433. Sprague, J.B. 1963. Resistance of four freshwater crustaceans to lethal high temperature and low oxygen. Journal of the Fisheries Research Board of Cana da 20:387–415. Sutton, W.L. 1995. Biological monitoring of Oatka Creek in upstate New York. New York State Department of Environmental Conservation, Division of Fish, Wildlife and Marine Resources, Albany, NY. 29 pp. Van Dolah, R.F. 1978. Factors regulating the distribution and population dynamics of the amphipod Gammarus palustris in an intertidal salt marsh community. Ecological Monographs 48:191–217. Wach, E.A., and R.M. Chambers. 2007. Top-down effect of fish predation in Virginia headwater streams. Northeastern Naturalist 14:461–470. Waters, T.F. 1981. Seasonal patterns in production and drift of Gammarus pseudolimnaeus in Valley Creek, Minnesota. Ecology 62:1458–1466. Zehmer, J.K., S.A. Mahon, and G.M. Capelli. 2002. Calcium as a limiting factor in the distribution of the amphipod Gammarus pseudolimnaeus. American Midland Naturalist 148:350–362.