Main content area

Predicting spatial extent of invasive earthworms on an oceanic island

Paudel, Shishir, Wilson, Gail W. T., MacDonald, Beau, Longcore, Travis, Loss, Scott R.
Diversity & distributions 2016 v.22 no.10 pp. 1013-1023
boreal forests, islands, roads, vegetation cover, altitude, soil microorganisms, ecosystems, introduced plants, grasses, microbial communities, models, soil, earthworms, biodiversity, climatic factors, prediction, landscapes, vertebrates, microbial biomass, regression analysis, ground vegetation, streams, California
AIM: Invasions of non‐native earthworms into previously earthworm‐free regions are a major conservation concern because they alter ecosystems and threaten biological diversity. Little information is available, however, about effects of earthworm invasions outside of temperate and boreal forests, particularly about invasions of islands. For San Clemente Island (SCI), California (USA) – an oceanic island with numerous endemic and endangered plant and vertebrate species – we assessed the spatial extent and drivers of earthworm invasion and examined relationships between earthworms and plant and soil microbial communities. LOCATION: San Clemente Island, southern California, USA. METHODS: Using a stratified random sampling approach, we sampled earthworms, vegetation, soils and microbial communities across SCI. We examined the relationship between the presence of invasive earthworms and soil and landscape variables using logistic regression models and implemented a spatial representation of the best model to represent potential site suitability for earthworms. We evaluated the relationship between invasive earthworms and vegetation and microbial variables using ANOVA. RESULTS: We found that the likelihood of encountering earthworms increased close to roads and streams and in high moisture conditions, which correspond to higher elevation and a north‐eastern aspect on SCI. The presence of earthworms was positively associated with total ground vegetation cover, grass cover and non‐native plant cover; however, there was no significant relationship between earthworms and microbial biomass. These results suggest that the earthworm invasion on SCI is at an early stage and closely tied to roads and high moisture conditions. MAIN CONCLUSIONS: Climatic variables and potential sources of earthworm introduction and dispersal (e.g. roads and streams) should be broadly useful for predicting current and future sites of earthworm invasions on both islands and continents. Furthermore, the significant positive relationship between non‐native plant cover and invasive earthworm presence raises the possibility of an emerging invasional ‘meltdown’ on SCI. Additional study of earthworm invasions on human‐inhabited oceanic islands is necessary to identify additional invasions and their potential for negative impacts on unique insular biota.