Seed dispersal and FST
Our results did not support the hypothesis that gravity-mediated seed dispersal increases population differentiation compared to wind or animal dispersal (Givnish, 2010) (Fig. 1d). This is in line with previous findings suggesting that the genetic structure of nuclear markers is largely driven by pollen flow (Petit et al., 2005; Sork, Nason, Campbell, & Fernandez, 1999; Skogen et al., 2019), and that the effect of seed dispersal is only detectable in the population genetic structure of chloroplast genes (Duminil et al., 2007). However, we note that gravity dispersal resulted in highly variable FSTvalues, potentially due to unrecorded secondary seed vectors. FST values for animal dispersal were also highly variable, which suggests that different animals could have different effects on population differentiation. Thus overall, as with vertebrate pollination, we suspect that more fine-scaled classifications of dispersers may improve our understanding of their effects on plant population genetic structure. Testing this idea, however, requires more detailed data on animal dispersal modes, which can be difficult to characterize. For example, in our study many species have a mix of seed dispersers, including small to large mammals and birds (like most Arecaceae, Fabaceae, Fagaceae, Myrtaceae, Sapotaceae, among others), making it difficult to assign plants to a disperser-specific taxonomic affiliation or foraging behavioral trait.