Fig. 4: The effects of protected area status and body mass on rates of change.
Results from protected and unprotected areas are shown in orange and grey, respectively. a, In all four road transect studies, median annual decline rates in UPAs exceeded those within the PAs assessed, significantly so in West Africa and Kenya. Boxplots show the median, first and third quartiles of the change in abundance within protected and unprotected areas in each of the four regions. Whiskers extend to ±1.5Ă— the interquartile range. Each point represents one species; n = 28 (West Africa), 15 (N. Cameroon), 22 (Kenya) and 25 (N. Botswana). b, The effects of site protection were more pronounced among large (≥1,300 g) than among small–medium raptors. Median decline rates in PAs vs UPAs differed by 30 percentage points among large raptors and by 18 percentage points among small–medium species. Boxplots show the median, first and third quartiles of the rate of change in abundance of large vs small–medium raptor species, inside vs outside of protected areas. Whiskers extend to ±1.5Ă— the interquartile range. Each point represents one species; n = 15 large, 27 small–medium species. c, Modelled relationship between the rate of change in abundance, body mass and protected area status (PAs vs UPAs). Notably, declines over three generation lengths exceeded the IUCN Vulnerable threshold (−30%, blue line) for the bulk of species in UPAs and for most large raptors within the PA types assessed. Fitted lines and shading indicate modelled change rates ±1 s.e.m. (model 13 in Extended Data Table 1).
