| Summary: | Questions Species distribution models have traditionally relied heavily on temperature and precipitation, often ignoring other potentially important variables. However, recent advances have shown other climatic variables, including snow cover and solar radiation, may strongly improve predictions of species occurrence. Wind has long been known to have mechanical and physiological impacts on plants, but has not yet received adequate attention as a driver of species distributions. Location Marion Island, sub-Antarctic. Methods Using data from 1,440 plots in a chronically windy system, we test if wind stress (a combination of wind exposure and wind speed) improves species distribution models of vascular plant species, examining predictions for both species occurrence and cover. Results Wind stress was a significant predictor of the occurrence of 12 of the 16 species, even after accounting for seven other ecophysiologically important abiotic variables. Species showed differential responses to wind, but wind stress was among the four most important drivers for the majority of species when modelling occurrence patterns (10 of 16) and variation in cover (12 of 16). Further, wind stress was more important than all temperature and precipitation variables in predicting the occurrence of six species (and three species' cover). Conclusions Wind conditions were most influential for species that are characteristic of open, wet environments and for pteridophyte species, likely due to high wind speeds and exposure increasing the potential for moisture loss. This research highlights the value of incorporating wind metrics into species distribution models, particularly under changing wind patterns. Peer reviewed
|
|---|