Trichocorixa verticalis is a North American water bug (Heteroptera: Corixidae) that occurs in brackish and saline aquatic systems. Recently, it has been found invading three continents including Europe. Its invasive success has been attributed to the capacity of tolerating hypersalinity. We compared both the realized and standardized salinity niche of invasive T. verticalis and native Corixidae to verify if T. verticalis may fill in an unoccupied niche. We first established the field distribution of T. verticalis and native Corixidae along a salinity gradient. Second, we experimentally tested the salinity tolerance of T. verticalis and three common native Corixidae species. Of the seven Corixidae species found in the field study, three were positively related to the salinity gradient: S. selecta, S. stagnalis, and T. verticalis. T. verticalis showed the highest salinity optimum, however, after correcting for environmental background variation, salinity optima differed little among the three halophilic species. In the salinity tolerance experiment, S. selecta outperformed T. verticalis, which performed as well as S. stagnalis. Based on our experimental results, we cannot support the hypothesis that T. verticalis' invasion is mediated by a high salinity tolerance that allows this species to colonize habitats unoccupied by native Corixidae. Although these findings contrast with the field patterns in which T. verticalis showed the highest niche optimum and tolerance, these patterns may have been partly due to other environmental factors, particularly anthropogenic disturbance. Our comparative results are for adults only, and it remains possible that relative salinity tolerance patterns for juveniles differ from that for adults, which may add to the observed field pattern.

Summary1. The role of spatial processes in ecotoxicology is largely ignored. Yet, together with speciesspecific sensitivities to the pollutant, spatial processes may determine community recovery and the resulting community composition after a pollution event. 2. We investigated the isolation-specific potential for internal recovery by reproduction and external recovery by dispersal after a pesticide pulse in experimental aquatic insect communities along an isolation gradient. 3. External recovery was important in the univoltine species and in a multivoltine species that went locally extinct but only when a source population was nearby. Internal recovery occurred in all multivoltine species and was unexpectedly stronger in more isolated communities, probably because of release from a dispersal-limited key predator. 4. As a result, community recovery and resulting changes in composition strongly depended on isolation and species differences in the potential for external recovery through dispersal and internal recovery through reproduction. 5. Synthesis and applications. Our results indicate that while the immediate impact of a toxicant on natural communities is shaped by species sensitivities, their recovery is primarily dependent on the degree of isolation. Risk assessment and the protection of communities under toxicant threat may improve greatly from considering the spatial context: isolated communities and communities with poor dispersers should receive extra protection to safeguard their member species. Ideally, land use planning should strive to remediate isolation of natural communities under threat in agricultural landscapes. Where this is not possible, spatial regulation of pesticide application may considerably improve protection of extant diversity.

Recent work on habitat selection has shown that the perceived quality of habitat patches may depend on the quality of adjacent patches. However, it is still unclear how local habitat selection cues can alter distribution patterns in metacommunities at a larger (regional) scale. We studied mosquito oviposition in pond landscapes that differed in the proportion of bad patches with fish predation risk. Our experiment provided conclusive evidence for two local and two regional types of habitat selection. Good patches near bad patches were avoided (local risk contagion) while more distant good patches experienced increased oviposition (regional compression). Oviposition in bad patches increased when located next to good patches (reward contagion) or when there were no good patches regionally present (regional compromise). This complex colonisation behaviour involving compromises at different spatial scales forces experimenters to reconsider the independence of spatial replicates and challenges available theories to predict species distribution patterns.

There is increasing concern that synergistic interactions between stressors may result in accelerated biodiversity loss. Yet, the prevalence and magnitude of these interactions remain one of the largest uncertainties in projections of future ecological change. Synergistic interactions between pesticide stress and predation risk are receiving increasing attention because they indicate that standard pesticide tests in the absence of predation risk may underestimate effects that occur under natural conditions. We questioned whether synergisms (or interactions in general), by differently shaping individual species’ sensitivities, can modulate species sensitivity rankings (SSRs) in survival and growth rate. Using laboratory and outdoor mesocosm experiments with five species of water boatmen (Hemiptera: Corixidae) we studied mortality and growth rate over seven days of exposure. We evaluated the presence and nature of interactions between the pesticide endosulfan and predation risk and tested to what extent this affected the SSRs to endosulfan. The combined exposure to the pesticide and predation risk resulted in synergistic effects for survival (Sigara lateralis) and growth rate (Sigara iactans) and in an antagonistic effect for growth rate (Hesperocorixa linnaei and female Sigara striata). These results suggest that standard tests may underestimate the pesticide effect in a natural predator environment for some species. While the effects of these interactions on SSRs were not strong and SSRs remained largely similar in the absence and presence of predation risk, some obvious species rank shifts occurred for growth rate which may potentially affect community structure through changed competitive strength. The study of SSRs and their dependency on biotic stressors may provide a simple conceptual and predictive framework to increase our understanding of how stressors like pesticides may differentially affect community structure in the absence and presence of another stressor.

No abstract