Disentangling patterns and processes that define biodiversity distribution around the world has excited ecologists and conservation biologists for centuries. We now know that two of the main drivers of biodiversity changes are climate and land use. Investigating how species respond to such changes across long periods of time is therefore of major importance. In this thesis I investigate how species distribution modelling techniques (SDMs) can aid in analyzing the impacts of past and present changes in climatic and land use conditions on biodiversity (pollinators: bees, butterflies and hoverflies), if these drivers of change have always had the same importance for driving the species distributions, how species respond to these drivers depending on their functional traits and if the recent biodiversity patterns are constrained by past landscape conditions. The results suggest that SDMs are tools that can effectively aid when investigating how environmental changes impact the distribution of biodiversity. They show that the importance of drivers of species distributions may change across time, highlighting a possible limitation on model forecasting to future environmental conditions. They also show that present pollinator distributions have been constrained by the historical landscape conditions and that different pollinator groups present a varied set of responses to environmental changes, which greatly depend on the species functional traits. This shows the low value of one-fits-all type of conservation measures and underlines the need for multi species analysis across long periods of times to disentangle the long lasting effects of changes in climate and land use conditions on biodiversity.