As the tempo of biological invasions increases, explanations and predictions of their impacts become more crucial. Particularly with regard to biodiversity, we require elucidation of interspecific behavioural interactions among invaders and natives. In freshwaters in The Netherlands, we show that the invasive Ponto-Caspian crustacean amphipod Dikerogammarus villosus is rapidly eliminating Gammarus duebeni, a native European amphipod, and Gammarus tigrinus, until now a spectacularly successful invader from North America. In the laboratory, survival of single (unguarded) female G. duebeni was significantly lower when male D. villosus were free to roam as compared with isolated within microcosms. In addition, survival of paired (guarded) female G. duebeni was significantly lower when male D. villosus as compared with male G. duebeni were present. D. villosus killed and consumed both recently moulted and, unusually, intermoult victims. Survival of G. tigrinus was significantly lower when D. villosus were free to roam as compared with isolated within microcosms and, again, both moulted and intermoult victims were preyed upon. Male D. villosus were significantly more predatory than were females, while female G. tigrinus were significantly more often preyed upon than were males. Predation by D. villosus on both species occurred over a range of water conductivities, an environmental feature previously shown to promote amphipod coexistence. This predatory invader is predicted to reduce further the amphipod diversity in a range of freshwater habitats in Europe and North America.

Abstract:To assess the increasing threats to aquatic ecosystems from invasive species, we need to elucidate the mechanisms of impacts of current and predicted future invaders. Dikerogammarus villosus, a Ponto-Caspian amphipod crustacean, is invading throughout Europe and predicted to invade the North American Great Lakes. European field studies show that populations of macroinvertebrates decline after D. villosus invasion. The mechanism of such impacts has not been addressed empirically; however, D. villosus is known to prey upon and replace other amphipods. Therefore, in this study, we used microcosm and mesocosm laboratory experiments, with both single and mixed prey species scenarios, to assess any predatory impact of D. villosus on a range of macroinvertebrate taxa, trophic groups, and body sizes. Dikerogammarus villosus predatory behaviour included shredding of prey and infliction of "bite" injuries on multiple victims. Dikerogammarus villosus killed significantly greater numbers of macroinvertebrates than did the native Gammarus duebeni, which is currently being replaced by D. villosus. This invader thus appears to impact on freshwater ecosystems through its exceptional predatory capabilities. We predict that future invasions by D. villosus will have serious direct and indirect effects on freshwaters, with its invasion facilitated in a larger "invasional meltdown" in regions like the North American Great Lakes. Résumé

As biological invasions continue, interactions occur not only between invaders and natives, but increasingly new invaders come into contact with previous invaders. Whilst this can lead to species replacements, co-existence may occur, but we lack knowledge of processes driving such patterns. Since environmental heterogeneity can determine species richness and co-existence, the present study examines habitat use and its mediation of the predatory interaction between invasive aquatic amphipods, the Ponto-Caspian Dikerogammarus villosus and the N. American Gammarus tigrinus. In the Dutch Lake IJsselmeer, we found broad segregation of D. villosus and G. tigrinus by habitat type, the former predominating in the boulder zone and the latter in the soft sediment. However, the two species co-exist in the boulder zone, both on the short and longer terms. We used an experimental simulation of habitat heterogeneity and show that both species utilize crevices, different sized holes in a plastic grid, non-randomly. These amphipods appear to optimise the use of holes with respect to their 'C-shape' body size. When placed together, D. villosus adults preyed on G. tigrinus adults and juveniles, while G. tigrinus adults preyed on D. villosus juveniles. Juveniles were also predators and both species were cannibalistic. However, the impact on G. tigrinus of the superior intraguild predator, D. villosus, was significantly reduced where experimental grids were present as compared to absent. This mitigation of intraguild predation between the two species in complex habitats may explain the co-existence of these two invasive species.

1. Data from field surveys, laboratory experiments and computer simulations of community dynamics revealed that a novel interaction among intraguild predation, physiological adaptation and environment may explain the complex distributions of two putatively competing aquatic amphipods. 2. Gammarus pulex and G. tigrinus both thrive in fresh and oligohaline waters in western Europe. However, the native European G. pulex excludes the invading North American G. tigrinus from freshwaters of relatively low conductivity, whereas the reverse occurs at higher conductivities. Additionally, there is much spatio‐temporal fluctuation in the patterns of coexistence of these species. 3. Laboratory experiments in The Netherlands and Ireland revealed that mutual predation of moulting individuals occurred frequently between these species. However, predation frequencies were differentially in favour of G. pulex under the ionic conditions to which this species is physiologically adapted (freshwater). On the other hand, predation was not differential under the ionic conditions to which G. tigrinus is physiologically adapted (oligohaline water). 4. A mathematical model, which extends the logistic equation to include mutual intraguild predation, simulated interactions over a range of values of relevant population parameters. This indicated that G. pulex would be excluded when balanced instantaneous rates of mutual predation were combined with the known greater reproductive output of G. tigrinus. However, this reproductive advantage is overcome by any relatively small bias in the instantaneous rate of predation favouring G. pulex, leading to the exclusion of G. tigrinus. This occurs even when the reproductive advantage to G. tigrinus is relatively large. Moreover, the model generated ‘switches’ in species dominance that are determined by the relative values of reproductive rate and mutual predation. The time taken to ‘switch’ may explain the transient periods of apparent coexistence of these species observed in the field. 5. The complex community dynamics of such species may thus be understood in terms of variation in the intensity of species interactions mediated by behavioural, physiological and environmental factors.