Drivers of inter-annual variation and long-term change in High-Arctic spider species abundances

Understanding how species abundances vary in space and time is a central theme in ecology, yet there are few long-term field studies of terrestrial invertebrate abundances and the determinants of their dynamics. This is particularly relevant in the context of rapid climate change occurring in the Arctic. Arthropods can serve as strong indicators of ecosystem change due to their sensitivity to increasing temperatures and other environmental variables. We used spider samples collected by pitfall trapping from three different habitats (fen, mesic and arid heath) in High-Arctic Greenland to assess changes in individual species abundances over an 18-year period (1996-2014). We calculated annual abundances of each species using a conventional method and compared this to a technique that corrected for the influence of short-term weather variation on arthropod activity. The latter method used the area under the curve of a fitted generalized additive model to measure annual change in abundance of each species. Abundances calculated using each of the two methods did not differ greatly over time nor in direction of climate effects, suggesting that short-term weather-driven activity does not influence interpretation of long-term trends. We used model selection to determine which climatic variables and/or previous years' abundance best explained annual variation in species abundances over this period. We identified and used 28 566 adult spiders that comprised eight species. Most notably, the abundances of some species (Collinsia thulensis and Erigone psychrophila) have declined during this 18-year period, in response to rising temperatures and snow depth dynamics, which affected snowmelt timing and moisture availability. No species increased in abundance through the study period. Since some species showed no trend in abundance through time and climatic effects were habitat-specific, continued climate change may also affect local species interactions. Long-term monitoring programmes are an extremely valuable means through which to generate empirical data about changes in populations and communities; this is especially pertinent in the Arctic, where climate change is occurring at a rapid rate.