Genetic diversity shapes the evolutionary potential of plant populations. For outcrossing plants, genetic diversity is influenced by effective population size and by dispersal, first of paternal gametes through pollen, and then of paternal and maternal gametes through seeds. Forest loss often reduces genetic diversity, but the degree to which it differentially impacts the paternal and maternal contributions to genetic diversity and the spatial scale at which these impacts are most pronounced are poorly understood. To address these questions, we genotyped 504 seedlings of the animal-dispersed palm Oenocarpus bataua collected from 29 widely distributed sites across Ecuador and decomposed the contribution of paternal and maternal gametes to overall genetic diversity. The amount of forest cover at a landscape scale (> 10 km radius) had an equally significant positive association with both male and female gametic diversity. In addition, there was a significant positive association between forest cover and effective population size. Stronger fine-scale spatial genetic structure for female versus male gametes was observed at sites with low forest cover, but this did not scale up to differences in male vs female gametic diversity. These findings show that reductions in forest cover at spatial scales much larger than those typically evaluated in ecological studies lead to significant, and equivalent, decreases of diversity in both male and female gametes, and that this association between landscape level forest loss and genetic diversity may be driven directly by reductions in effective population size of O. bataua, rather than by indirect disruptions to local dispersal processes. 

We sampled O. bataua seedlings across Ecuador from January through May 2017 in forested areas on both sides on the Andes mountain range. We used Qiagen DNeasy 96 kits to extract DNA from leaf and the outer seed coat. We then amplified 11 microsatellite loci using polymerase chain reaction (PCR) and scored microsatellite genotypes using the program GeneMarker v. 1.85 (SoftGenetics®) using a protocol previously used for O. bataua in Ecuador. We extracted DNA from leaf and seed tissue from 641 seedlings from 29 sites across Ecuador. 138 seed samples did not amplify well and were omitted from analyses, leaving 504 samples with both leaf and corresponding seed tissue.

The file with multilocus leaf genotypes was used for population genetic analyses and to estimate the lower 95% CI for effective population size. The genotypes for leaf and seed tissue were used to estimate alpha diversity of male and female gametes. We have included the paper with the code for the modified gametic extraction in R. Missing alleles/genotypes are coded with NA. Below is the metadata for both sheets: