Population structure effect on reproduction of a rare plant: beyond population size effect

Abstract

Several studies have evaluated the effect of population size on plant reproductive output, but there are few studies concerning the effect of other population structure variables on plant reproduction. In this study, we went beyond population size effect and we tested the effect of some population spatial variables such as plant density, plant aggregation, population structure, and population identity on plant reproductive success and fitness in large populations (>1000 individuals) of Centaurea hyssopifolia Vahl., an endemic species of central Spain. We analysed the reproductive output (capitula, flowers, and seeds) and another two components of plant fitness (seed mass and germinability) in 350 individuals from seven populations in 2002 and 2003. All populations were similar in size but differed in other population spatial characteristics (density, plant aggregation pattern, population plant size structure). We used Morisita’s aggregation index to characterize the level of plant aggregation within populations. Population plant size structure variable determined the proportion of different sized plants in a population. We used generalized linear mixed models to model the contribution of these population spatial variables to several reproductive parameters. Our results showed that whereas the aggregation index exerted a positive control on plant seed set, plant density had a negative effect, and plant population size structure did not significantly influence any response variable. Density only exerted a negative effect on seed set. On the contrary, plant aggregation had a positive effect. Seed mass was also larger in more aggregated populations, although no differences in seed germinability were observed. We detected the effect of population attributes on reproduction in terms of seed set, but we did not detect any effect on other overall reproductive variables measured at plant level. To sum up, our results showed that if population fragments were large enough, spatial population attributes became crucial factors for plant reproductive output and fitness. These easily measurable population variables may improve the conservation management of rare plant species.