Fig. 1
Effects of phenology advances on abundance and distribution trends depend on voltinism, illustrated with example species. The flowchart describes the main conclusions of this study: climate-driven phenology advances have a positive effect on abundance trends in multivoltine species, but a neutral or negative effect on abundance trends in univoltine species (depending on habitat specialisation). In turn, abundance trends have a positive effect on trends in distribution size and northern range margin, regardless of voltinism. Trends in emergence date, abundance, distribution size and northern range margin are depicted for two butterflies: Silver-studded Blue Plebejus argus, and Small Blue Cupido minimus. a, b P. argus has retracted in distribution size (−0.2 %/yr) and range margin (−1.8 km/yr), whereas C. minimus has expanded in distribution size (1.3 %/yr) and range margin (7.4 km/yr). Distribution size depicted as number occupied hectads in 1995–2014 (black circles), 1995–2004 only (orange circles) and 2005–2014 only (blue circles). Range margins are depicted for 1995 (orange lines) and 2014 (dashed black lines). c, d Voltinism of univoltine P. argus and multivoltine C. minimus shown by observed abundance on transect counts across all sites and years (grey circles; counts > 100 omitted) and GAM-fitted curves (blue lines). e–h Both P. argus and C. minimus have advanced their phenology (0.25 days/yr and 0.46 days/yr, respectively); P. argus has declined in abundance (−5.5 %/yr), but C. minimus has increased in abundance (3.8 %/yr). Observed peak day of first-generation emergence (e, f) and mean abundance per recording event (g, h) in each year is shown (grey circles), and points at the same site are connected (grey lines); overall trend across the duration of the study period (Supplementary Data 1) is shown (black dashed line) with 95% confidence intervals (grey shading)
