Under siege: Isolated tributaries are threatened by regionally impaired metacommunities

Abstract

Pristine streams are often targeted as conservation priorities because of their ability to preserve regional biodiversity. However, isolation within heavily degraded regions likely alters important metapopulation and metacommunity processes (e.g., rescue and mass effects), affecting the ability of in-tact communities to boost regional conditions. To test this hypothesis, we sampled invertebrate communities and physicochemical conditions from 168 streams within the mountaintop removal-valley fill mining region of West Virginia. We used redundancy analysis to first test for significant effects of local (observed physicochemical conditions) and neighborhood (streams within a 5km buffer) degradation on assemblage structure across all taxa and stress tolerance (low, moderate, high) and dispersal (low, high) categories. We then used generalized linear and hurdle models to characterize changes in community metrics and individual taxa, respectively. Local condition consistently explained the majority of variation (partial R(2) up to 5× those of neighborhood condition) in community structure and was the only factor affecting sensitive taxa. Neighborhood condition explained significant variation in moderately tolerant taxa with low dispersal capacity and highly tolerant taxa, regardless of dispersal. Decreased occurrence (Baetis) and abundance (Maccaffertium) of key taxa and corresponding metrics (%E, %EPT) suggest decreased dispersal and associated mass and rescue effects within degraded neighborhoods. Decreased neighborhood conditions also resulted in the proliferation of tolerant taxa (Chironomidae, Chimarra, Hemerodromia). Our results suggest communities within even the most pristine streams are at risk when isolated within heavily impacted regions. Consequently, protection of regional species' pools in heavily impacted regions will require more than simply conserving un-impacted streams.

Keywords: Headwater conservation; Local vs. regional effects; Mountaintop mining; Multiple spatial scales; Stream metacommunities.