Pollination in small islands by occasional visitors: the case of Daucus carota subsp. commutatus (Apiaceae) in the Columbretes archipelago, Spain

Abstract

This study investigates the pollination ecology and related floral traits of the species Daucus carota subsp. commutatus in the isolated archipelago of Columbretes, E. Spain, where bees are absent. Two populations were studied: a small population found on a relatively large island (Grossa) inhabited nowadays by three people; and a larger population on a smaller non-inhabited island (Foradada). The plant, found also in other rocky coastal enclaves of the west Mediterranean, is totally self-compatible, andromonoecious, exhibiting strong protandrous dichogamy and sequential flowering of umbel orders, with limited probability of self pollination at the plant level. Pollination on both study islands is carried out solely by flies, mainly by Eristalis tenax (Syrphidae), a long distance visitor and efficient pollinator, but irregular migrant to the archipelago; and Lucilia sericata (Calliphoridae), an abundant flower visitor but less efficient pollinator dependent on human and avian wastes for its larval growth. The small population on Grossa appears not to suffer pollen limitation. Seed set is reduced, however, in the large population on Foradada, which we attribute to the low pollinator capital of this island which is too small to attract and sustain migrant hoverflies and where no human and avian wastes are available to sustain blowfly growth. We conclude that in isolated environments with reduced or unpredictable pollinator capital, long distance migrant pollinators, such as E. tenax, play an invaluable conservation and perhaps evolutionary role for the native plants. Because these islands have been always practically uninhabited we argue that the absence of adequate local pollinator capital is perhaps compensated by the selection of particular floral traits, which differentiate the study subspecies from other conspecifics. Such traits are: the exposure of nectar during male vs. female stage to enhance quick pollen flow; the extended stigma receptivity period (up to 10 days) resulting in a total flower life span of up to 22 days, enabling an adequate pollen deposition on stigmas even under low pollen transfer rates. In this respect, the low flower sex ratio (hermaphrodite : staminate) compared to other conspecifics studied elsewhere may be also relevant.