Pollen and ovule production in wind-pollinated species with special reference to Juncus

Abstract

The reproductive biology of wind-pollinated species in terms of pollen and ovule production is rarely studied compared with zoophilous species, despite available hypotheses on the effect of growth form and life-history traits on reproductive investment. Here, we use published data and new data for species of Juncus and Luzula (Juncaceae) to test the hypotheses that, in wind-pollinated species, woody perennials should exhibit larger pollen–ovule (P/O) ratios than herbaceous species and that species with separate sexes have larger P/O ratios than homoecious species. In total, we report pollen and ovule production for 291 wind-pollinated species, including 19 Juncus and 5 Luzula species. Compared with other wind-pollinated species, Juncus exhibits unusually low P/O ratios (log P/O = 2.06 ± 0.46) because of high ovule production. We argue that the high ovule and seed production in Juncus, associated with frequent self-fertilization, may be beneficial in habitats preferred by the genus. In general, we found higher P/O ratios in woody perennials (log P/O = 4.37 ± 1.18) or in species with separate sexes (log P/O = 4.28 ± 1.12) than in herbaceous (log P/O = 3.51 ± 0.77) or homoecious (log P/O = 3.52 ± 0.80) species, respectively. However, when we analyzed woody perennials separately, we found no significant difference in P/O ratios between homoecious and nonhomoecious species. We argue that woody perennials, independent of dicliny, may be preferentially outcrossed and therefore exhibit decreased variation in mating systems compared with herbs. Because the degree of outcrossing correlates with P/O ratios, differences between homoecious and nonhomoecious woody perennials could be less pronounced.