Abstract
As primarily sessile organisms, plants often show a non-random spatial distribution of genotypes over distance. This process known as fine-scale spatial genetic structure (FSGS) has been suggested through systematic reviews to depend on life form, mating system, and pollen and seed dispersal vectors, while there is no consensus on its behaviour due to external factors, such as anthropogenic habitat changes. By conducting a systematic review and global meta-analysis of empirical FSGS studies, we aimed to evaluate how anthropogenic habitat fragmentation and degradation influence the strength of FSGS in plant populations by means of the Sp statistic. Moreover, we tested how pollination and seed dispersal vectors contribute to the variation of the Sp statistic. We retrieved 243 FSGS studies from 1960 to 2020 of which only 65 were informative for the systematic review. Most empirical studies comprised outcrossers (84%) and trees (67%), with few herbs (23%) and scarce annual species (2%). In weighted meta-analyses for 116 plant populations (31 studies), we did not detect significant effects in the magnitude of effect sizes for the Sp statistic among undisturbed, degraded and fragmented habitats. Results showed significant effects for seed dispersal vectors, but not for pollination. Overall, we observed high variation among the effect sizes (not related to the goodness-of-fit of mixed models) of habitat status, pollination and seed dispersal categories, which precludes identifying biological trends on the Sp statistic. More empirical studies are needed that contrast multiple plant populations in disturbed versus undisturbed habitats, and by increasing the taxonomic groups, such as herbs and annual plants.