Effects of land cover and climate on species richness of butterflies in boreal agricultural landscapes

This study examined how species richness of butterflies with differing population trends is affected by land cover and climate. Generalized linear models (GLM) with land cover, climate and geographical location as explanatory variables were constructed to explain variation in species richness of declining, stable and increasing butterflies in 10-km squares in Finland. The largest part of explained variation in species richness of all the three butterfly trend categories was related to spatially structured climate variables and pure geographical location. However, the relative importance of climate and land cover differed between the butterfly categories. Pure and spatially structured climate variables explained two-fold more variation in species richness in increasing butterflies than in stable and declining butterflies. On the other hand, pure and spatially structured land cover variables explained 2.5–6 times more variation in species richness in declining butterflies than in stable and increasing butterflies. Land cover variables differed in their effects on the species richness of the three trend categories: the amount of grassland had the greatest positive effect on the number of declining species, whereas the amount of forest was the most important land cover variable for species richness of stable butterflies and arable land and field-forest mosaic for species richness of increasing butterflies. The results suggest that habitat availability has the greatest effect on the trends of declining butterfly species, and climate in concert with habitat availability on the trends of increasing species.