Fishing spiders, green sunfish, and a stream-dwelling water strider: male-female conflict and prey responses to single versus multiple predator environments

Many studies have experimentally addressed the effects of a particular predator species on prey behavior. In nature, however, prey frequently face multiple species of predators that often vary in their predatory mode and in their level of predation risk. Relatively few studies have considered prey responses under these complex conditions. In Kentucky, the stream-dwelling water strider (Aquariusremigis) coexists with many potentially dangerous predators, two of which are the green sunfish (Lepomiscyanellus) and the fishing spider (Dolomedesvittatus). Green sunfish occupy stream pools and attack water striders from below. In contrast, fishing spiders hunt along stream shorelines where they perch on overhanging vegetation or rocks and attack water striders near shore. We compared how A. remigis individuals respond to these two very different predators in pools with one or both predators. The presence of sunfish in pools had strong effects on male water strider behavior, including increased use of three types of refuge from sunfish (riffles, climbing out of the water, sitting on the water but at the edges of pools), decreased activity and a decreased number of aggressive males on the water. Spiders also influenced water strider behavior; male water striders avoided spiders by shifting away from the edges of pools. Comparisons of the effects of the two predator species showed that in general, antipredator responses by male water striders were stronger in pools with fish alone than in those with spiders alone. In the presence of both predators, male water strider behavior (microhabitat use and activity) was generally similar to behavior in the presence of fish alone. In contrast, female water striders showed no significant response to the presence of sunfish, and little response to the presence of spiders. This lack of response could be because females spent much of their time in refuges even in the absence of predators (apparently hiding from harassment by males). Both spiders and fish caused decreases in water strider mating activity. The presence of fish reduced both the number of matings per pool (mating frequency), and mean mating durations. Spiders induced a decrease in mean mating duration, but not in mating frequency. The largest reductions in mating activity occurred in pools with both predators present. Pools with either spiders or fish alone suffered 15–20% water strider mortality during our experiment (versus no mortality in predator-free pools). Extant theory suggests that when prey face conflicting microhabitat responses to two predators (as in this study), the predators should have facilitative effects on predation rates (i.e., prey that avoid one predator are often killed by the other and vice versa). Mortality rates in pools with both predators present, however, were not significantly different from that predicted by a null model of multiple predator effects. The lack of predator facilitation can be explained by the compensatory reductions in water strider activity and mating activity in the presence of both predators. Received: 26 August 1996 / Accepted: 12 June 1998     

Predation risk and social interference as factors influencing habitat selection in two species of stream-dwelling waterstriders

We investigated how an aggressive species of waterstrider, Aquariusremigis, and potential predators, green sunfish (Lepomis cyanellus),affected the habitat use and mating behaviors of a less aggressivecongeneric species, A. conformis. Although these species sometimesco-occur, A. remigis typically lives in small streams with fewor no fish, whereas A. conformis are typically in medium- orlarge-sized streams with large populations of potentially predatoryfish. We tested in separate experiments in seminatural streams:1) the effect of fish on behaviors of A. conformis; 2) the effectof A. remigis on A. conformis; and 3) the habitat use of A.conformis when given a choice between pools with A. remigisor fish. The first experiment showed no effect of fish on eithermating behaviors or microhabitat use of A. conformis. This isin surprising contrast to the strong effects of fish alreadydocumented in A. remigis. The second experiment showed thatthe mating activity of A. conformis was reduced when A. remigiswere present; hence, A. conformis should avoid A. remigis. Finally,when A. conformis were presented with a choice between two pools,one containing A. remigis and the other containing fish, bothsingle males and pairs of A. conformis chose the pools withfish. In contrast, the habitat use of single female A. conformiswas not affected by either fish or A. remigis. Results fromthese experiments demonstrate that closely related species exhibitcontrasting social and antipredator behaviors and that aggressivesocial behavior is an important determinant of habitat partitioning     

Interacting effects of predation risk and male and female density on male/female conflicts and mating dynamics of stream water striders

We used a factorial experiment to examine interacting effectsof male density, female density, and sunfish (predation risk)on mating dynamics of the stream water strider (Aquarius remigis).Many of our results corroborated earlier studies on the isolatedeffects of each factor on mating behavior. The effect of eachfactor, however, depended on the other factors. For example,in low density pools, predation risk decreased male generalactivity, male/female harassment rates, mating activity, andmating duration and increased the large male mating advantage.At higher densities, however, water striders apparently enjoyed"safety in numbers" and did not alter their mating dynamicsin response to the presence of predators. Female activity showeda particularly complex response to male density and fish. Whenmales were scarce, fish caused females to reduce their activity.However, when males were abundant, fish increased female activity,probably because fish decreased male activity thus releasingfemales from harassment by males. The three treatment factorsalso had interacting effects on male mating success. In theabsence of fish, when females were scarce, increased male densityresulted in a decrease in mean male mating success; however,when females were abundant, increased male density enhancedmean male mating success. In contrast, in the presence of fish,male density had little effect on male mating success. Manyof the observed mating patterns can be explained by the effectsof ecological and social factors on male/female conflicts; thatis, on male harassment of females and female reluctance to mate.     

Mate disruption in the water strider Aquaris remigis Say, 1832 (Hemiptera: Gerridae)

This report describes the behaviour of a previously uncharacterised mating tactic amongst male water striders (Aquarius remigis Say, 1832), involving the physical break up of existing mating pairs. Using data from four separate laboratory observations on the water strider mating system, we show that this behaviour represents roughly 12.6% of all mating attempts by males. Furthermore, we demonstrate that males are successful in breaking up the existing mating pair in 15.6% of the cases, resulting in secure matings with the disrupted female 6.2% of time. We suspect that mate disruption may serve as an alternative means for acquiring mates by males with low mating success using conventional behaviours. Further research should be performed to determine the prevalence of this behaviour in various natural populations and the specific contexts in which this behaviour occurs.     

Asymmetric competition between distant taxa: poecilid fishes and water striders

Summary Competition for water surface prey between fish (Priapichtus annectens: Poeciliidae) and water striders (Potamobates unidentatus: Gerridae) was studied in the laboratory and in pools in a small tropical stream. Laboratory experiments showed that fish depressed activity and foraging success of water striders. Large fish (4–5 cm) had a greater effect than small fish (2–3 cm). The field experiment showed that competition was highly asymmetric. Presence of fish decreased water strider foraging success while the reverse interaction was insignificant. It is suggested that the higher individual foraging success of the fish, harassment of water striders by fish and the use of an exclusive resource, benthic invertebrates, by the fish, contribute to this pattern. Habitat use differed between the two species. Fish used the deeper parts of stream pools and water striders used the shallower parts of the pools. Asymmetric interference and exploitation competition may force water striders to use shallow edge habitats.     

Behavioural correlations across situations and the evolution of antipredator behaviour in a sunfish-salamander system

Behavioural correlations are at the heart of understanding how conflicting demands shape the evolution of ecologically important behaviours. Many studies have focused on the effects of negative behavioural correlations generated by time budget conflicts within situations. We examined an alternative possibility that involves positive behavioural correlations due to behavioural carryovers across situations. Specifically, we examined the role of behavioural carryovers in governing antipredator responses of streamside salamander larvae, Ambystoma barbouri, to predatory green sunfish, Lepomis cyanellus. Earlier work showed that these larvae suffer heavy sunfish predation due to high larval exposure to fish (high proportion of time spent out of refuge). Earlier work also showed that paradoxically, despite selection pressure from fish, these larvae show higher exposure in the presence of fish (poorer antipredator behaviour) than a sister species that inhabits fishless, ephemeral ponds. The standard time budget trade-off between feeding and antipredator behaviour does not appear to explain the observed antipredator behaviours. Instead, the present study shows that the relatively large proportion of time that larvae spend out of refuge (exposed) in fish pools in the daytime can be explained in part by behavioural correlations across situations. Specifically, larvae showed positive correlations among individuals in their daytime exposure in fish pools, nighttime exposure in fish pools, and exposure in fishless pools. The benefits of high exposure in fishless conditions (associated with high feeding and developmental rates) and high exposure in fish pools at night (necessary to drift out of fish pools) have apparently overridden the predation cost of being exposed in fish pools in the day. Behavioural correlations across situations might often result in ecologically important behaviours that appear maladaptive in an isolated context. Copyright 2003 The Association for the Study of Animal Behaviour. Published by Elsevier Science Ltd. All rights reserved.      

Experiments on the effects of food and density on voltinism in a stream-dwelling water strider (Aquarius remigis)

1. The stream water strider Aquarius remigis shows a latitudinal pattern of variation in voltinism. In general, populations with shorter growing seasons (e.g. in eastern Canada) tend to be univoltine (animals that reach adulthood in the summer overwinter before reproducing in the following spring), whereas populations with somewhat longer growing seasons (e.g. in the north-eastern United States) tend to be bivoltine. 2. This pattern was broken at our study site in the south-eastern United States (Kentucky) where A. remigis had a long growing season, but was almost always univoltine. In summer 1993, however, adult A. remigis in central Kentucky displayed a bivoltine reproductive cycle; i.e. individuals in some pools began breeding shortly after maturing to the adult stage. 3. A field survey documented a negative relationship between local water strider density and reproductive activity in prediapause adults. A laboratory experiment manipulating food availability and density, revealed that animals held at low density with high food levels displayed greater mating activity and egg production than did their counterparts at higher density or lower food levels. 4. A laboratory experiment also showed that high water strider density resulted in a greater frequency of very short pair durations (< 10 min). 5. Although the observed effects of density and food availability on mating activity of prediapause adults seem intuitively reasonable, they differ from the patterns observed in overwintered adults. The difference in reproduction patterns might be due to differences in selective pressures on prediapause vs. post-diapause adults.     

The effects of trout on water striders in stream pools

Summary Observations made in Rattlesnake Creek, Santa Barbara County, California, U.S.A. indicate that rainbow trout (Salmo gairdneri) affect the microdistribution of adult water striders (Gerris remigis). Gerrids occupied all areas of stream pools lacking trout but stayed in margins of pools containing trout. To test the hypothesis that the presence of trout controlled the distribution of water striders within pools, trout were removed from some pools and transferred to pools lacking trout. Undistrubed pools with and without trout served as controls for the trout transferrals. The results supported the hypothesis; within one week, gerrids moved to the edge of pools where trout were added and to the middle of pools where trout were removed. Time budgets and gut analyses suggest that local patterns of gerrid distribution were maintained by harrassment of gerrids by trout. At times trout had significant effects on components of gerrid fitness.     

Phylogeny and reclassification of species groups in Aquarius Schellenberg, Limnoporus Stål and Gerris Fabricius (Insecta: Hemiptera-Heteroptera, Gerridae)

Abstract. We investigated phylogenetic relationships of water striders (Hemiptera‐Heteroptera: Gerridae) from the three principal Holarctic genera, Aquarius Schellenberg, Limnoporus Stål and Gerris Fabricius with parsimony analyses of sixty‐six morphological characters and DNA sequences from mitochondrial (cytochrome c oxidase subunit I + II; large mitochondrial ribosomal subunit) and nuclear (elongation factor 1‐alpha) genes. The taxon sampling included all species of Aquarius and Limnoporus, and a dense, near complete, sample of Gerris species with representatives from all subgenera and species groups, and Gigantometra gigas (China) was selected as an outgroup species. A simultaneous analysis of all data sets gave eight equally parsimonious trees, and a strict consensus tree left only a few relationships within Gerris unresolved. While Limnoporus and Gerris each were resolved as monophyletic entities, Aquarius was found to be polyphyletic, because the Nearctic Aquarius remigis‐group, comprising A. remigis (Say), A. amplus (Drake and Harris), A. nyctalis (Drake and Hottes) and A. remigoides Gallant and Fairbairn, was placed as sister group to Gerris, while the Andean Aquarius chilensis (Berg) was sister group to all three genera. Remaining species of Aquarius comprised a sister group to the Gerris + the A. remigis‐group clade. Based on our phylogenetic reconstruction we discuss relationships within and among the three genera, reassess and diagnose species groups, and discuss zoogeographical relationships among all taxa.     

Predator Generalization Decreases the Effect of Introduced Predators in the San Marcos Salamander,Eurycea nana

The introduction of novel predators into an environment can have detrimental consequences on prey species, especially if these species lack the ability to recognize these predators. One such species that may be negatively affected by introduced predators is the federally threatened San Marcos salamander (Eurycea nana). Previous research found that predator‐naïve (captive‐hatched) salamanders showed decreased activity in response to the chemical cues of both a native fish predator (Micropterus salmoides) and an introduced fish predator (Lepomis auritus), but not to a non‐predatory fish (Gambusia geiseri). We tested the hypothesis that E. nana recognized the introduced Lepomis (and other non‐native Lepomis) because they share chemical cues with other native congeneric Lepomis predators in the San Marcos River. We examined the antipredator response of predator‐naïve E. nana to chemical cues from (1) a sympatric native sunfish (Lepomis cyanellus; Perciformes: Centrarchidae); (2) a sympatric introduced sunfish (L. auritus); (3) an allopatric sunfish (Lepomis gibbosus); (4) a sympatric non‐native, non‐centrarchid cichlid (Herichthys cyanoguttatum; Perciformes: Cichlidae); and (5) a blank water control to determine whether individuals make generalizations about novel predators within a genus and across a family. Exposure to chemical cues from all fish predator treatments caused a reduction in salamander activity (antipredator response). Additionally, there were no differences in the antipredator responses to each predatory fish treatment. The similar responses to all sunfish treatments indicate that E. nana shows predator generalization in response to novel predators that are similar to recognized predators. Additionally, the antipredator response to H. cyanoguttatum indicates that predator generalization can occur among perciform families.     

Biogeography and diversification of Holarctic water striders: Cenozoic temperature variation,habitat shifting and multiple intercontinental dispersals

It is now rare to find a semi‐aquatic organism group with which to vigorously test whether their diversification model and distribution pattern are closely related to the Cenozoic temperature variation. This hypothesis is explored for water striders of the genera Aquarius Schellenberg, Gerris Fabricius and Limnoporus Stål, which comprise a monophyletic clade with primarily Holarctic distribution. We sample almost 90% of the currently recognized Aquarius , Gerris and Limnoporus species. Five DNA fragments from 62 species are used to reconstruct a phylogram. Divergence time is estimated using Bayesian relaxed‐clock method and three fossil calibrations. We investigate diversification dynamics, biogeography and ancestral state reconstruction by using maximum‐likelihood, Bayesian and parsimony approaches. Our results showed that the crown of the three genera originated and underwent an initial diversification in Asia at 72 Ma (HPD: 59–86 Ma) in the Late Cretaceous, subsequently expanding into other regions via dispersal. The Bering Land Bridge was the major migration route between Eurasia and North America but was interrupted before the early Oligocene (34 Ma). Ancestors most likely used lentic habitats, and a minimum of two independent shifts to lotic habitats occurred in the initial diversification. Cenozoic temperature variation regulated the evolutionary history of Holarctic water striders of the genera Aquarius , Gerris and Limnoporus . Temperature warming during Stage I (52–66 Ma) was associated with the disappearance of shallow lentic habitats; this phenomenon forced certain lentic lineages to colonize new lotic habitats and promoted the diversification of lineages. Temperature cooling during Stage II (after 34 Ma) was associated with the fragmentation of water habitats of the ‘mixed‐mesophytic’ belt, resulting in the extinction of historical taxa and influencing close lineages that shaped the present disjunct Eurasian–North American distribution.     

An elemental and stable isotope assessment of water strider feeding ecology and lipid dynamics: synthesis of laboratory and field studies

1. Despite the ubiquity and abundance of water striders (Hemiptera: Gerridae) in temperate streams and rivers and their potential usefulness as sentinels in contaminant studies, little is known about their feeding ecology and lipid dynamics. 2. In this study we used stable isotopes of carbon (δ¹³C) and nitrogen (δ¹⁵N) and elemental carbon to nitrogen ratios (C/N) to assess dietary habits and lipid content, respectively, for water striders. 3. To determine diet‐tissue fractionation factors, nymphs of the most common species in New Brunswick, Canada, Aquarius remigis were reared in the laboratory for 73 days and exhibited rapid isotopic turnover in response to a switch in diet (C half‐life = 1.5 days, N half‐life = 7.8 days). Their lipid content increased towards the end of the growing season and resulted in lower δ¹³C values. Diet‐tissue fractionation factors were established after correction of δ¹³C data for the confounding effect of de novo lipid synthesis (strider δ¹³C_adj– diet δ¹³C_adj = 0.1‰, strider δ¹⁵N – diet δ¹⁵N = 2.7‰). 4. Water striders from the majority of 45 stream sites (83%) in New Brunswick had less than 50% contribution of aquatic carbon to their diets but showed a gradual increase in the contribution of this carbon source to their diet with increasing stream size. 5. These data indicate that striders exhibit a strong connection to terrestrial carbon sources, making them important users of energy subsidies to streams from the surrounding catchment. However, this dependence on terrestrial organic matter may limit their utility as indicators of contamination of aquatic systems by heavy metals and other pollutants.     

Mosquito female response to the presence of larvivorous fish does not match threat to larvae

1. Several species with complex life‐history traits such as amphibians and insects with aquatic immature stages and terrestrial adults avoid ovipositing in pools containing larvivorous fish. This avoidance response was assumed to be a general one for most fish species. 2. The generality of ovipositing Culex to the presence of three, widespread larvivorous fish species was tested in a set of field experiments with artificial oviposition pools using caged fish. 3. Larval performance was further examined under actual predation by these three fish species. 4. Results show that ovipositing females responded strongly to the presence of caged mosquitofish, Gambusia affinis, while showing no significant response to the presence of caged green sunfish, Lepomis cyanellus, or the pirate perch, Aphredoderus sayanus. All three fish species consumed similar amounts of larvae. 5. This is the first example of species‐specific response differences to predators during mosquito oviposition habitat selection. These results point to the existence of predator‐released kairomones affecting mosquito behaviour. These kairomones may either be species‐specific or vary in concentration among fish, and probably have an important role in the understanding of mosquito spatial distribution.     

Signalling Asymmetry in the Communication of the Water Strider Aquarius remigis in the Context of Dominance and Spacing in the Non-mating Season

The role of high-frequency ripple signals (HF signals) made by males of the water strider Aquarius remigis was studied in the contexts of competition for food and general spacing behaviour during the non-mating season. HF signals were played back through the ripple-producing legs of males during dyadic agonisdc encounters, using a signal-driven wire coil to oscillate a magnet glued to a leg. These signals were in addition to normal signals. The additional signals significantly increased the number of retreats by non-magneted males, showing that the signals increased the dominance of a non-territorial male. Hence, our results increase the number of contexts in which HF signals of A. remigis function. Males, but not females, avoided a site occupied by a magneted dead male through which HF signals were played. Thus, the communication system used by A. remigis males during competition for food seemed to be ignored by females, suggesting sex-specific signals even in a non-mating context. Evolutionary models of signalling often assume that contestants have evolved the same repertoire of signals in order to resolve conflicts peacefully. This water strider system thus poses an interesting challenge for future theoretical and empirical research on communication asymmetry.     

Megafauna associations with deepwater corals (Primnoa spp.) in the Gulf of Alaska

Our study showed that one species of water strider (Aquarius najas) dominated the insect fauna (>90% of the biomass) on and near the surface of a small stream in southern Sweden, but the diet of Daubenton's bats (Myotis daubentonii), regularly feeding over the same stream, contained <1% of these insects. To explain why the bats did not eat water striders in proportion to the apparent abundance of these insects, we tested three hypotheses, with the following results: (1) The water striders did not respond to artificial bat calls and therefore their defence against bats is not based on ultrasonic hearing. (2) Water striders hand-fed to the bats were eaten with similar frequency as palatable controls (caddis flies), so there is no evidence that water striders are unpalatable to bats. (3) Attempted prey-captures by the bats occurred less often than expected within 1 m of the bank of the stream (the bank consisted of a stone wall), indicating that the bats' flight or, alternatively, their detection or capture of insects, was difficult within that area, possibly constrained by clutter. At dusk and just before the bats emerged to feed, the water striders moved away from the open water. They remained motionless close to the bank for as long as the bats were active.     

SEXUAL SELECTION AND THE EVOLUTION OF SEXUAL SIZE DIMORPHISM IN THE WATER STRIDER,AQUARIUS REMIGIS

Sexual size dimorphism (SSD) is often attributed to sexual selection, particularly when males are the larger sex. However, sexual selection favoring large males is common even in taxa where females are the larger sex, and is therefore not a sufficient explanation of patterns of SSD. As part of a more extensive study of the evolution of SSD in water striders (Heteroptera, Gerridae), we examine patterns of sexual selection and SSD in 12 populations of Aquarius remigis. We calculate univariate and multivariate selection gradients from samples of mating and single males, for two sexually dimorphic traits (total length and profemoral width) and two sexually monomorphic traits (mesofemoral length and wing form). The multivariate analyses reveal strong selection favoring larger males, in spite of the female-biased SSD for this trait, and weaker selection favoring aptery and reduced mesofemoral length. Selection is weakest on the most dimorphic trait, profemoral width, and is stabilizing rather than directional. The pattern of sexual selection on morphological traits is therefore not concordant with the pattern of SSD. The univariate selection gradients reveal little net selection (direct + indirect) on any of the traits, and suggest that evolution away from the plesiomorphic pattern of SSD is constrained by antagonistic patterns of selection acting on this suite of positively correlated morphological traits. We hypothesize that SSD in A. remigis is not in equilibrium, a hypothesis that is consistent with both theoretical models of the evolution of SSD and our previous studies of allometry for SSD. A negative interpopulation correlation between the intensity of sexual selection and the operational sex ratio supports the hypothesis that, as in several other water strider species, sexual selection in A. remigis occurs through generalized female reluctance rather than active female choice. The implications of this for patterns of sexual selection are discussed.     

Role of body size in dominance interactions between male water striders, Aquarius paludum

Water striders are a model system for the study of sexual size dimorphism, but the effect of body size on the dominance relationship between individuals has not been experimentally tested. In 34 staged contests between males of the water strider Aquarius paludum, we determined the effect of body size difference between contestants on the outcome of the aggressive interactions. In contests between a large and a small male, the larger individuals won the interactions significantly more often than expected by chance. This is the first experimental evidence for the importance of body size in pair-wise contests among water striders.     

Sources of character conflict in a clade of water striders (Heteroptera: Gerridae)

Incongruence among trees reconstructed with different data may stem from historical (gene tree‐species tree conflict) or process (character change biases) phenomena. Regardless of the source, incongruent data, as determined with “global” measures of homoplasy, have often been excluded from parsimony analysis of the combined data. Recent studies suggest that these homoplasy measures do not predict the contribution of each character to overall tree structure. Branch support measures identify, on a character to node basis, sources of support and conflict resulting from a simultaneous analysis of the data. We implement these branch support measures to identify sources of character conflict in a clade of water striders consisting of Gerris Fabricius, Aquarius Schellenberg, and Limnoporus Stål species. Separate analyses of morphology, mitochondrial cytochrome oxidase I (COI), large mitochondrial ribosomal subunit (16SrRNA), and elongation factor‐1α (EF‐1α) data resulted in cladograms that varied in resolution and topological concordance. Simultaneous analysis of the data resulted in two trees that were unresolved for one node in a strict consensus. The topology agreed with current classification except for the placements of Aquarius chilensis and the Aquarius remigis species group closer to Gerris than to congeneric species. Branch support measures indicated that support derived from each data set varied among nodes, but COI had an overall negative effect on branch support. However, Spearman rank correlation of partitioned branch support values indicated no negative associations of branch support between any data sets and a positive association between EF‐1α and 16SrRNA. Thus incongruence among data sets was not drastic and the gene‐tree versus species tree phenomenon was not implicated. Biases in character change were a more likely reason for incongruence, although saturation curves and incongruence length difference for COI indicated little potential for homoplasy. However, a posteriori inspection of COI nucleotide change with reference to the simultaneous analysis tree revealed AT and codon biases. These biases were not associated with branch support measures. Therefore, it is difficult to predict incongruence or identify its cause. Exclusion of data is ill advised because every character is potentially parsimony informative.     

Effect of predatory fish presence on habitat use and diel movement of the stream amphipod, Gammarus minus

SUMMARY.

• 1 We examined how the presence of predatory fish affected macro—and microhabitat use and movement patterns of the amphipod, Gammarus minus Say, in a second order stream.
• 2 Among macrohabitats, amphipods were consistently more abundant in runs than in pools. Densities in pools were low regardless of fish presence, whereas low densities in runs were correlated with fish presence. Siltation and the lack of coarse substrate particles probably accounted for the scarcity of G. minus in pools. Among microhabitats, G. minus was more abundant in leaf litter and gravel than in silt/sand substrates. Fish presence did not affect microhabitat use in the field. Laboratory manipulations showed, however, that predation rates by green sunfish (Lepomis cyanellus Rafinesque) were significantly lower on amphipods in leaf litter than in other microhabitats.
• 3 Most movement by G. minus in runs occurred at night and was directed downstream. In the laboratory, amphipods significantly reduced swimming activity at night in response to water conditioned with green sunfish. In the field, however, amphipods were only slightly less active in runs at night when exposed to chemical cues of confined green sunfish. The distribution of amphipods in streams is influenced by the presence of fish, but chemical cues alone may not be important in triggering predator avoidance behaviours in nature.

     

GENETIC POPULATION STRUCTURE AND LEVELS OF GENE FLOW IN THE STREAM DWELLING WATERSTRIDER,AQUARIUS (=GERRIS) REMIGIS (HEMIPTERA: GERRIDAE)

Gene flow, in combination with selection and drift, determines levels of differentiation among local populations. In this study we estimate gene flow in a stream dwelling, flightless waterstrider, Aquarius remigis. Twenty-eight Aquarius remigis populations from Quebec, Ontario, New Brunswick, Iowa, North Carolina, and California were genetically characterized at 15 loci using starch gel electrophoresis. Sampling over two years was designed for a hierarchical analysis of population structure incorporating variation among sites within streams, streams within watersheds, watersheds within regions, and regions within North America. Hierarchical F statistics indicated that only sites within streams maintained enough gene flow to prevent differentiation through drift (Nm = 27.5). Above the level of sites within streams gene flow is highly restricted (Nm ≤ 0.5) and no correlation is found between genetic and geographic distances. This agrees well with direct estimates of gene flow based on mark and recapture data, yielding an N_e of approximately 170 individuals. Previous assignment of subspecific status to Californian A. remigis is not supported by genetic distances between those populations and other populations in North America. Previous suggestion of specific status for south-eastern A. remigis is supported by genetic distances between North Carolina populations and other populations in North America, and a high proportion of region specific alleles in the North Carolina populations. However, because of the high degree of morphological and genetic variability throughout the range of this species, the assignment of specific or subspecific status to parts of the range may be premature.