Contrasting heterozygosity-fitness correlations between populations of a self-compatible shrub in a fragmented landscape.

The mechanisms underlying heterozygosity-fitness correlations (HFCs) are subject of intense debates, especially about how important population features such as size or degree of isolation influence HFCs. Here, we report variation in HFCs between Large and Small populations of a self-compatible shrub (Myrtus communis) occurring within an extremely fragmented landscape. In each of the five study populations, we obtained data on both heterozygosity and fitness for 9-12 maternal families (i.e. offspring from the same mother plant). Whereas heterozygosity explained most of the variance (60-86 %) in growth rate of seedling families within Large populations, this relationship was absent within Small populations. Our results suggest that inbreeding may explain the observed HFCs within Large populations, and that different genetic processes (such as genetic drift and/or selection) could have overridden HFCs within Small populations. While it is difficult to draw general conclusions from five populations, we think our results open new research perspectives on how different genetic processes underlie variation in HFCs under different population contexts. Our study also points to a need for further attention on the complex relationships between heterozygosity in self-compatible plants and their progeny in relation to mating system variation. Finally, our results provide interesting new insights into how population genetic diversity is maintained or lost in a highly fragmented landscape.