The number of species effect on the dispersal - range size relationship
Including a large number of species is advantageous for various reasons. Besides increasing the statistical power of the study, including more species may also capture a larger variation in both dispersal-related traits and range sizes essential to capture the relationship. The number of species used to examine the dispersal-range size relationship across studies ranged from 5 (for reef fishes; Zapata & HerrĂ³n 2003 and Lester & Ruttenberg 2005) to 10,338 (for birds; Sheard et al. 2020). A relatively small sample size (i.e., few species investigated), can result from a) study objective, such as the focus on a particular genus that comprises few species, or from b) data limitations, e.g. because trait data are only available for a subset of the species of interest. While data limitations are understandable because it is often difficult to gather information for all species in a large, wide-spread, species-rich clade, it may also come with the risk that the subsample of species and traits are biased and thus not representative for our understanding of the relationship between dispersal and range size. This may lead to biased outcomes (Alzate et al. 2019a), even after data imputation to resolve data gaps (Johnson et al 2021). It is noteworthy that a small sample size might be sufficient if the species pool is also small, so the level of completeness might be the critical issue here. However, we were not able to evaluate this because most studies did not report on sampling completeness.