Implications for biodiversity conservation in fragmented landscapes
Our results are consistent with the landscape planning strategy proposed by Arroyo-Rodríguez et al. (2020) for forest biodiversity protection in human-dominated naturally forested regions. They suggested that 40% of the landscape should be dedicated to forest, and that three quarters of that forest should be in small patches while the remaining quarter should be in a single large patch (or a few large patches). Retaining a large patch would support the persistence of herptile populations, while small patches would maximize the diversity and persistence of other taxa.
Most notably, patch-size thresholds used to prioritize habitat for biodiversity protection are orders of magnitude larger than the sizes of the small patches that we found to have disproportionately high biodiversity value on a per-area basis (Fig. 4, vertical lines; reviewed in Fahrig et al. (2022)). These thresholds may be imposed for practical reasons (Armsworth et al. 2018), but they are not supported by biological evidence. In addition, while some species requiring large amounts of habitat likely have already disappeared from human-dominated regions, there is no evidence that these species are generally negatively affected by fragmentation per se (independent of habitat loss, Fahrig (2017); Fig. 1, left panel). If anything, the opposite is confirmed in previous analyses suggesting that many small patches harbor both more specialist species and more species of conservation concern than few large patches (Fahrig 2020; Riva & Fahrig 2022). Therefore, protecting a very large number of small patches that sum to a high total habitat density across a landscape appears to be an effective solution for sustaining populations of most species. This seems to be the case for carnivores across Europe, a group expected to rely strongly on large patches, but in fact showing widespread population recoveries across patchy anthropogenic landscapes (Chapronet al. 2014).
We caution that our results apply in regions with patches similar in size to those analyzed here, which were generally centered around 10-ha (Fig. 4). It is possible that analyses using much smaller patches would not show the predominant SS > SL pattern. However, as the small patches in the datasets were generally much smaller than those normally considered for protection, our conclusion that the default priority for habitat protection should shift to several small patches seems robust. Additionally, we lack detailed information about the spatial extents of the studies, which could conceivably affect the SLOSS outcomes. A recent meta-analysis suggests that the extent of most studies in ecology is < 100,000 ha (Estes et al. 2018). The datasets we used here from Chase et al. (2019) appear to be consistent with this pattern based on inspection of maps in the papers describing each dataset. Further work is needed to determine whether the generality of our results is affected by the spatial extent of the study area or by the spatial arrangement of small and large patches.