Table 2. Model statistics of PGLS-models a)-e). To test our main hypotheses (see Fig. 1) we fitted five separate models. We tested the effect of direction of shifts in a) phenology and b) NRB (Table 1a) on population trend (continuous). We tested the effect of c) categorical shift in phenology (Table 1a) on shift in NRB (continuous). To test the effect of different combinations of the responses on population trends, and which hypothesis provides a more viable strategy for species (as opposed to which strategy is the more common), we tested the effect of the d) combined response groups (Fig. 1b; RGs in Table 1b) and e) hypothesis-wise responses (Fig. 1b; Hypotheses in Table 1b) on population trend (continuous).