Search

Refuges that result from environmental heterogeneity within ecosystems have an important yet under-appreciated role in maintaining native community diversity in face of exotic invasion. The objective of our study was to determine if different refuge types constrain invasion impacts on native biodiversity at the whole ecosystem-scale of the Upper St. Lawrence River. We focused on the voracious round goby fish as a sentinel exotic species whose spatial distribution within this ecosystem is also representative of the species distributions of several other Ponto-Caspian invaders. We first explored if wetlands were acting as unknown refuges in reducing the local abundance of the round goby fish. We then tested the relative influence of a known broad-scale conductivity gradient compared with local wetlands on structuring the composition, diversity, and abundance of native macroinvertebrate and fish communities inside and outside each of these refuge types. We found that the two types of refuges, broad-scale conductivity gradient and local wetlands, limited round goby abundances at the whole ecosystem scale, and structured macroinvertebrate and fish community diversity. The broad-scale conductivity refuge was twice stronger than wetlands in limiting round goby abundance. Although wetlands were effective in constraining round goby abundance, the direct effect of wetlands rather than their indirect effects through limiting round goby abundance, were more powerful in explaining maintenance of macroinvertebrate and fish community diversity in invaded high conductivity waters. Our findings underscore the important role of environmental heterogeneity in producing different types of refuges that buffer invasion effects in freshwater ecosystems, and we advocate the preservation of wetlands as a part of this process.

Evolutionary approaches are gaining popularity in conservation science, with diverse strategies applied in efforts to support adaptive population outcomes. Yet conservation strategies differ in the type of adaptive outcomes they promote as conservation goals. For instance, strategies based on genetic or demographic rescue implicitly target adaptive population states whereas strategies utilizing transgenerational plasticity or evolutionary rescue implicitly target adaptive processes. These two goals are somewhat polar: adaptive state strategies optimize current population fitness, which should reduce phenotypic and/or genetic variance, reducing adaptability in changing or uncertain environments; adaptive process strategies increase genetic variance, causing maladaptation in the short term, but increase adaptability over the long term. Maladaptation refers to suboptimal population fitness, adaptation refers to optimal population fitness, and (mal)adaptation refers to the continuum of fitness variation from maladaptation to adaptation. Here, we present a conceptual classification for conservation that implicitly considers (mal)adaptation in the short‐term and long‐term outcomes of conservation strategies. We describe cases of how (mal)adaptation is implicated in traditional conservation strategies, as well as strategies that have potential as a conservation tool but are relatively underutilized. We use a meta‐analysis of a small number of available studies to evaluate whether the different conservation strategies employed are better suited toward increasing population fitness across multiple generations. We found weakly increasing adaptation over time for transgenerational plasticity, genetic rescue, and evolutionary rescue. Demographic rescue was generally maladaptive, both immediately after conservation intervention and after several generations. Interspecific hybridization was adaptive only in the F1 generation, but then rapidly leads to maladaptation. Management decisions that are made to support the process of adaptation must adequately account for (mal)adaptation as a potential outcome and even as a tool to bolster adaptive capacity to changing conditions.

The evolutionary consequences of temporal variation in selection remain hotly debated. We explored these consequences by studying threespine stickleback in a set of bar-built estuaries along the central California coast. In most years, heavy rains induce water flow strong enough to break through isolating sand bars, connecting streams to the ocean. New sand bars typically re-form within a few weeks or months, thereby re-isolating populations within the estuaries. These breaching events cause severe and often extremely rapid changes in abiotic and biotic conditions, including shifts in predator abundance. We investigated whether this strong temporal environmental variation can maintain within-population variation while eroding adaptive divergence among populations that would be caused by spatial variation in selection. We used neutral genetic markers to explore population structure and then analysed how stickleback armor traits, the associated genes Eda and Pitx1 and elemental composition (%P) varies within and among populations. Despite strong gene flow, we detected evidence for divergence in stickleback defensive traits and Eda genotypes associated with predation regime. However, this among-population variation was lower than that observed among other stickleback populations exposed to divergent predator regimes. In addition, within-population variation was very high as compared to populations from environmentally stable locations. Elemental composition was strongly associated with armor traits, Eda genotype and the presence of predators, thus suggesting that spatiotemporal variation in armor traits generates corresponding variation in elemental phenotypes. We conclude that gene flow, and especially temporal environmental variation, can maintain high levels of within-population variation while reducing, but not eliminating, among-population variation driven by spatial environmental variation.