Does the spatial variation in selective pressures explain among-site differences in seed mass? A test with Buxus balearica

Positive effects of seed size on germination and survival can be offset by a greater probability of predation or a poorer dispersal of larger seeds. We hypothesized that spatial variation in local selective pressures acting on seed mass may lead to differences in both optimal and observed seed mass among discrete populations. We first examined the variation in seed mass across a total of 14 populations of Buxus balearica; in six of such populations, we further examined the temporal variation by measuring seed mass during three consecutive years. Second, we evaluated seed mass effects on different recruitment phases (seed rain to seedling establishment) in three populations. Lastly, for these three populations, we estimated the resulting overall phenotypic selection on seed mass during recruitment and compared the observed and the predicted optimal seed masses. Most variation (c. 70%) in seed mass occurred among populations, and although we found inter-annual variation in seed mass, the differences in seed mass among populations were consistent over time. Conflicting selective pressures on seed mass appeared during recruitment, and their direction and strength varied among populations, depending on the relative local importance of seed predation vs. germination and establishment. Observed seed mass matched predicted optimal seed mass in two of the three examined populations, suggesting local adaptative responses to the spatial mosaic of selective pressures.