Tropical forests provide ecosystem services that are inarguably vital to global health, from biodiversity conservation to climate regulation, yet these forests are threatened by anthropogenically-driven climate change and the highest rates of deforestation worldwide. More than 90% of tropical plant species are animal-pollinated and are wholly dependent on these mutualisms for reproduction, making the disruption of pollination services one of the greatest threats to tropical biodiversity. Climate and deforestation may synergistically impact fundamental density-dependent plant gene flow processes by altering a plant population’s spatial structure, the composition of its pollinator community, and the plants’ phenology, potentially leading to phenological mismatches. In this study, we quantified the phenology, conspecific density, and pollinator community of a common understory tree, Miconia affinis, across seven biogeographic regions in Central Panama over two years, one of which occurred during the 2016 El Niño Southern Oscillation. We then used spatially explicit reproductive and population genetic data to analyze the effects of phenological synchrony, tree size, conspecific neighborhood density, local kinship, and pollinator community composition on the seed set and pollen dispersal patterns of these 62 trees across both a typical non-El Niño (2013) an El Niño season (2016).
Results/Conclusions
We found that M. affinis phenological patterns, specifically the date of flowering onset, was nearly a month later for the majority of regions during the El Niño season compared to the non-El Niño season. Pollinator community composition was also significantly different between the two seasons across all regions. Our models revealed that the El Niño event and local kinship were the two most important factors negatively impacting seed set, with local kinship having a negative impact across both seasons. Interestingly, local kinship had a significantly larger negative effect on seed set during the El Niño climate event, indicating that regions with lower levels of kinship (greater local genetic diversity) exhibited greater reproductive success and appear to be ‘buffered’ from the negative reproductive effects typically associated with severe drought-like conditions. Additionally, we found that pollen dispersal distances were greater for trees with higher conspecific neighborhood density, indicating that positive density-dependence may play an important role in pollen-mediated gene flow. By incorporating genetic and trait-based data into the quantification of plant reproductive success, we highlight the critical role that landscape genetic diversity plays in the resilience of tropical plant populations to deforestation and extreme climate events.