Data compilation on magnitude of size changes from literature
We examined publications that report seasonal variation in skull size and/or brain mass in wild populations of the common shrew (S. araneus ). From those publications, we additionally used values on total body mass when reported. For each study, we extracted the percentage of change from the first summer size peak to the winter minimum (decrease), and from winter to the second summer peak (regrowth) of each metric. Whenever estimations of the decrease and regrowth were not directly reported, we calculated them based on the published data. We determined the first size peak as the month with the highest mean value for juveniles; the winter minimum as the month with the lowest mean value for winter subadults; and the second size peak as the month with highest mean value for adults. Summer juveniles are immature young individuals born in late spring or summer; winter subadults are immature individuals, which are ca. six months old; adults are individuals in spring and summer which were born the previous year. As S. araneus has a maximum life span of 13-18 months, there is no overlap of sexually mature individuals from two generations. When sample size in a given month was low, we joined data from two or more consecutive months to determine the corresponding size extreme. The amount of change was calculated as the difference between mean values in the size extremes. We added to this data compilation from the literature our own data on the changes in braincase height, brain mass and/or body mass from four new populations (see below).
We determined coordinates and altitude for all locations as geographic variables. We extracted 19 bioclimatic variables from WorldClim Global Climate Data version 1.4 for all locations (Hijmans et al. , 2005) and used averaged values from data collected along 1960 - 1990 (see results section for details).
Following the criteria listed above, we also compiled the same information on seasonal morphological variation in other wild mammal species. However, the low number of publications prevented any statistical analysis.