DISCUSSION
We investigated the habitat fragmentation-related effects (e.g., the consequences of reduced size and increased isolation) on the demographic and genetic processes shaping the metapopulation of two epiphyllous bryophyte species with contrasting mating systems. We showed that fragmentation induces marked effects on the functional connectivity of both plant populations in less than half a century of experimental size and connectivity reduction. The long-term reduced population density in small patches (1- and 10-ha) (Sierra et al., 2019b), has led to accelerated genotypic differentiation (genetic drift) compared to the populations in large patches (100-ha fragment and continuous forests). This result supports our hypothesis (H1) that spatial variation in patch size and isolation are important determinants of the metapopulation genetic process (stochastic loss of genotypes) in a fragmented forest, despite the notion of bryophytes as being highly vagile organisms (Vanderpoorten et al., 2019)
We observed that species’ mating system confers different sensitivity to the effects of habitat fragmentation, where a shift in the genetic structure (diversity and allele frequency: FST ) of the unisexual plant was less negatively impacted by forest fragmentation than of the bisexual counterpart, contrary to our expectations (H2). However, this should be taken with caution since is expected that genetic diversity meltdown will take a longer time to be apparent than changes in allele frequency in response to habitat fragmentation (Lowe et al., 2005). In the case of the unisexual species the observed differentiation in 1-ha might indicate initial signs of diversity loss in the following generations under the current landscape fragmented scenario.
The metapopulation migration network of the two species shows differences in functional connectivity, with the unisexual species more inter-connected among patches than the bisexual species, contrary to our predictions (H3). For R. flaccida (unisexual) we observed a patchy metapopulation with a highly interconnected migration network, with all patches exchanging migrants among them, irrespective of size and isolation. However, positive isolation by distances in small patches suggests that migrants are from nearby source populations. WhileC. surinamnesis (bisexual) present symmetrical migration only between small patches, and no sign of isolation by distance among populations. Stochastic extinctions are expected to increase in small patches, but genetic diversity might be maintained through colonization from nearby source populations if functional connectivity is maintained (Auffret et al., 2017). Despite the patch size effect on the metapopulation demography and genetics of both epiphyllous bryophyte species, small patches remain valuable for conservation assessment, as they can act as a source of migrants to other forest fragments and continuous forests, as observed for Macronesian oceanic islands (Laenen et al., 2011; Patiño et al., 2015).