Impact of surrounding landscape on nectarivorous bat abundance and pollen movement along riparian corridors in a Mexican tropical dry forest

Background/Question/Methods

Recent studies in landscape ecology have found significant effects of the landscape matrix (the heterogeneous mix of ecosystems and land use types surrounding a focal site) on patterns of organism abundance and movement. The goal of this study was to evaluate the impact of the landscape matrix on pollinator abundance and contemporary pollen-mediated gene flow in a bat-pollinated tropical tree species, Crescentia alata.  This tree species is often found in ephemerally dry riverbeds, which may act as corridors for pollinator movement. The condition of the surrounding matrix (dense continuous forest vs. more permeable fragmented forest), may impact how bat pollinators, and consequent pollen-mediated gene flow, move through these corridors. We used 5 sites within continuous forest matrix and 4 sites within fragmented forest matrix in and around the Chamela-Cuixmala Biosphere Reserve in Jalisco, Mexico, to compare flower abundance and bat pollinator abundance between these sites. We also examined preliminary data on the genetic structure of the pollen pool, to evaluate differences in the number of effective pollinators (N_ep) at continuous and fragmented sites.

Results/Conclusions

A total of 366 nectarivorous bats were mist-netted near flowering trees over 25 nights in Summer 2011, at 2-3 nights per site. There was a higher correlation between the number of nectarivorous bats caught and number of open flowers next to nets, in fragmented matrix sites versus continuous matrix sites. This pattern was driven by the bat species Glossophaga soricina, which was captured more frequently than the other two nectarivorous species captured (Leptonycteris yerbabuenae and Choeroniscus godmani). Trees in the fragmented matrix sites also produced more flowers. This higher bat-flower correlation in fragmented matrix sites corresponds to preliminary genetic data from 2010, which showed slightly higher genetic structure of the pollen pool from continuous forest maternal trees (Φ_{ST cont} = 0.213 vs. Φ_{ST frag} =0.188) indicating more restricted movement in the continuous forest landscape. These structure estimates translate into a higher number of effective pollinators per tree in fragmented sites (N_{ep cont} = 2.35 vs. N_{ep frag} = 2.66), suggesting that the landscape matrix may affect bat pollinator movement, and that removal of vegetation may potentially increase pollinator visitation to this tree species.