Inbreeding exposes deleterious recessive alleles in homozygotes, lowering fitness and generating inbreeding depression (ID). Both purging (via selection) and fixation (via drift) should reduce segregating deleterious mutations and ID in more inbred populations. These theoretical predictions are not well-tested in wild populations, which is concerning given purging/fixation have opposite fitness outcomes. We examined how individual- and population-level inbreeding and genomic heterozygosity affected maternal and progeny fitness within and among 12 wild populations of Impatiens capensis. We quantified maternal fitness in home sites, maternal multilocus heterozygosity (using 12,560 SNPs), and lifetime fitness of selfed and predominantly outcrossed progeny in a common garden. These populations spanned a broad range of individual- (f_i = -0.17–0.98) and population-level inbreeding (F_IS = 0.25–0.87). More inbred populations contained fewer polymorphic loci, less fecund mothers, and smaller progeny, suggesting higher fixed loads. However, despite appreciable ID (mean: 8.8 lethal equivalents per gamete), ID did not systematically decline in more inbred population. More heterozygous mothers were more fecund and produced fitter progeny in outcrossed populations, but this pattern unexpectedly reversed in highly inbred populations. These observations suggest that persistent overdominance or some other force acts to forestall purging and fixation in these populations.

We measured the reproductive output (a component of fitness) of Impatiens capensis plants in each of 12 home field sites (N = 808 individuals). These field sites occurred in fragmented floodplain forests and marches surrounded by an urban/agricultural matrix in Wisconsin, USA. We define reproductive output as the number of cleistogamous and chasmogamous structures on each plant during peak flower (binned into buds, open flowers, and seed capsules and bare pedicles). We also measured the height of each plant. We collected cleistogamous and chasmogamous seeds (N = 6,859) from as many of these mother plants as possible (N = 718 maternal plants with seeds collected). We germinated the seeds in a greenhouse, transplanted the seedlings into a common garden, and measured lifetime fitness (total number of ripe seed capsules produced) of each progeny (N = 2,260) in the common garden. A subset of the maternal plants (N = 296) were genotyped for a previous study (Toczydlowski and Waller, 2019, Molecular Ecology), at thousands of SNP loci using genotype-by-sequencing (1 x 100 base pair reads). We used this SNP dataset to calculate individual maternal multilocus heterozygosity (sMLH), individual inbreeding (f), and population-level inbreeding (F_IS ). We then compared the fitness of CH and CL progeny at each major life stage, estimated cumulative inbreeding depression and inbreeding load across the lifetime of the progeny, and tested how maternal sMLH related to both maternal and progeny fitness. Our main goal was to test if inbreeding depression declines in more inbred populations.

This data package includes the files needed to reproduce the analyses for the paper: Toczydlowski, R. H., and Waller, D. M. 2023. Failure to purge: Population and individual inbreeding effects on fitness across generations of wild Impatiens capensis. Evolution, under review.

More detailed methods are available in the README for this data package and in the associated scientific paper.

A dynamic version of this data package is also housed in a git repository on BitBucket at:
https://bitbucket.org/toczydlowski/inbreeding_effects_in_impatiens