Fig. 3: Associations of variation in provincial colonisation rate of butterflies with land use and temperature change.
Scatterplots show provincial colonisation rate (number of colonising species per decade) as a function of (a) forest cover, (b) grassland cover, (c) temperature change, and (d) human settlements. Each dot represents data for one province per period with different colours of the dots for each of the four time periods. Dark blue lines and blue shading illustrate the predicted mean lines with 95% CI obtained from the GLMM (n = 201). The thin trend lines of different colours represent the associations based on linear regressions as estimated separately for each of the four time periods. Red asterisks indicate significant interactions with time period, °0.05 < P < 0.1, *P < 0.05, and **P < 0.01, see Table 1a and Table S1 for details. Bottom panel (e) shows how the proportion of each habitat type changed over time (one box every 10th year between 1900 and 2010, and one box for 2019). Each province contributes one value per time point. Boxes indicate interquartile range (IQR), vertical line denote median, upper, and lower whisker third quartile +1.5 × IQR and 1st quartile −1.5 × IQR range, and dots represent outliers. Note that the scale on the vertical axis is different for human settlements. Data for (e) were obtained from the HILDA GIS product102,103,104.
