Predicting reproductive success of insect- versus bird-pollinated scattered trees in agricultural landscapes

Scattered trees are set to be lost from agricultural landscapes within the next century without sustained effort to increase recruitment. Thus, understanding the reproductive dynamics of scattered tree populations will be critical in determining how they can contribute to population restoration. The distance between conspecifics should be a key predictor of reproductive success, as more isolated trees are expected to receive fewer pollinator visits and experience increased transfer of self-pollen during longer pollinator foraging bouts. Further, isolation effects should be greater in species with less mobile pollinators. Here we contrast the effects of plant isolation on reproductive success of two species of eucalypt "paddock trees", Eucalyptus camaldulensis and Eucalyptus leucoxylon, with insect- and bird-pollination, respectively. Seed production was not affected by tree isolation in either the insect- or bird-pollinated species and once outliers were removed, neither was there an effect on germination rate. As somatic mutations may bias outcrossing rate estimates, we initially screened our microsatellite markers for mutations but found no variation in microsatellite profiles throughout the canopy of trees. individual outcrossing rates did not decline with increasing tree isolation in either the insect- or bird-pollinated species, though there was considerable variation in these rates at large distances, suggesting that pollination becomes unreliable with increasing tree isolation. We found that pollination distances have likely increased in tree species in agricultural landscapes, and that this may be facilitated by introduced honeybees in the case of E. camaldulensis. We therefore suggest that even isolated trees of these species produce seed of sufficient quantity and quality to contribute to population restoration. (C) 2008 Elsevier Ltd. All rights reserved.