Flowering phenology, floral traits and pollinator composition in a herbaceous Mediterranean plant community.

The relationships between flowering plants and their insect visitors were studied in a Mediterranean grassland in north-east Spain. Floral traits (size, shape, symmetry, and colour), floral rewards (pollen and nectar), flowering period, and floral visitors were recorded for the 17 most abundant plants in the community. Flowering was year-round, but most species flowered in spring. The three species that flowered after spring had small flowers, but the distribution of floral features (including rewards offered) did not show a strong seasonality. Ants contributed 58.5% to the flower visits recorded. Other frequent visitors were beetles (12%), flies (9.5%), honey bees (6.4%), wild bees (6.4%), and wasps (5.2%). Honey bees were most abundant in April, wild bees from April to July, beetles from May to July, and ants from May to September. The lack of tight plant-insect associations was the rule, with most plant species visited by a rather diverse array of insects representing two or more orders. The plant species having narrower spectra of visitors either had flower rewards exposed or attracted mostly illegitimate visitors. By means of correspondence analysis four categories of plants were defined according to their main groups of visitors: (1) honey bees and large wild bees; (2) large wild bees; (3) ants and beetles; and (4) beetles and small-sized bees. The Mantel test was used to calculate correlations among four matrices representing similarities in visitors attracted, floral morphological traits, pollen-nectar rewards, and blooming time, respectively. In spite of seasonality shown by the different insect groups, results indicate that the observed patterns of visitor distribution among plants were most affected by pollen-nectar rewards.