Constraints on flowering phenology in a phryganic (East Mediterranean shrub) community

The phenological pattern of flowering at the community level was studied in a Greek phryganic ecosystem near Athens for 4 consecutive years. Flowering is strongly seasonal: 80% of the insect-pollinated flora, which consists of 133 species, blooms between February and June. There is a variably expressed secondary flowering period in autumn. The pollinating fauna follows a strongly correlated pattern of abundance. Two types of plants were distinguished pauciflorous species bearing < 10 flowers that are large compared to the plant body, and multiflorous species with many small flowers. Pauciflorous species flower in the winter half of the year, while multiflorous species flower mainly in the summer half. The mean flower life spans are 9 and 3 days, respectively. The duration of flowering (DF) for each species is 55 days on average, which is long compared to other communities. The DF shows year-to-year variations, concomitant with the vicissitudes of the climate. The start of flowering of a species is statistically correlated with the temperature in the previous month, not with rainfall; its end date of flowering only partly compensates for the time gained or lost. DF is maximal in winter. The average flower life span of species flowering at any given date varies strongly and independently of the average DF. We tested the hypothesis that flowering phenology is set by phylogenetic and life form constraints. This could not be corroborated for phylogeny, evidently because of the overriding influence of the mediterranean climate, and probably also for biogeographical reasons. In contrast, life forms and multiflorous and pauciflorous species show strong differences. Many (51) of the species are therophytes; we tested the hypothesis that because of their annual habit they would be more dependent on pollination than perennials. Thus we anticipated that therophytic species would be differentiated from perennials in their flowering phenologies. This is not corroborated. We therefore conclude that the seed bank plays a role that is analogous to that of a perennial plant body.