An Extended Flowering and Fruiting Season Has Few Demographic Effects in a Mediterranean Perennial Herb

The Mediterranean perennial herb Lobularia maritima shows an exceptionally extended flowering and fruiting that lasts for 10 months, from early September to late June. We hypothesized that such an extended phenology may be a flexible mechanism that enhances population persistence in variable Mediterranean environments, as fecundity in one part of the season could compensate for reproductive failure in another part of the season. We predicted that (1) fecundities throughout the year would significantly contribute to mean population growth rate l, (2) negative covariances among fecundities in different periods of the year would reduce variance in population growth rate, and (3) extended flowering and fruiting would enhance long-term stochastic population growth rates. We used a periodic matrix model to describe the demography of a L. maritima population over five years, in which each year was divided into six periods of two months each. Population growth rates varied from a low of 0.25 to a high of 4.81 over the five years of study. Periodic elasticity analyses revealed that fecundities in the first two periods of the year, i.e., the autumn periods, were the major determinants of population growth in most of the years, rather than fecundities over the whole flowering and fruiting season. Variation in the autumn fecundities also made the largest contribution to the observed variance in l. Hence, unlike earlier generalizations, highly variable parameters also had relatively high elasticities. As predicted, negative covariances appeared between the fecundities, but only between those of the two autumn periods. These negative covariances were interpreted as a flexible mechanism to tailor the onset of reproduction to the start of the autumn rains, thereby reducing the variance in l. We used stochastic matrix models to investigate the effect of different reproductive strategies on stochastic population growth rate. We created hypothetical populations with either an even distribution of fecundity throughout the year or with concentrated flowering and fruiting in autumn or in spring. Stochastic population growth rates were high if a major part of the reproduction remained in autumn. A hypothetically even distribution of fecundities over the entire flowering season increased both the mean and the variance in l, with little overall effect on stochastic population growth rate. All together, our results indicate that a reproductive failure of L. maritima in autumn cannot be compensated for by reproduction in winter and spring. The extended flowering and fruiting season thus has very little effect on the population dynamics of the species. Alternative explanations for the unusual phenology of L. maritima in the Mediterranean environment are discussed.