The response of epiphytes to anthropogenic disturbance of pine-oak forests in the highlands of Chiapas, Mexico

Epiphytes contribute up to over 30% of the vascular plant diversity of old-growth tropical rain forests, but little is known about their occurrence in exploited forests that are structurally altered. Here, I estimate epiphyte species richness and biomass on 35 oak host trees, equally divided over several trunk diameter classes, in each of 16 sites in anthropogenic disturbed montane pine-oak forests at similar elevations. The sites vary in their amount of disturbance from scarcely to heavily disturbed, related to the structure of the forest, and in their underlying distance in the landscape with a maximum of ca. 25 km. The explainable variation in the epiphyte vegetation is partitioned between environmental and spatial variables, using multivariate data analysis. In total, 74 species of epiphytes were encountered. Bromeliads dominate at all sites in terms of their biomass. Forest disturbance, measured as the proportion of sprouted oaks, has an expected negative effect on the epiphyte biomass and alpha diversity per ground surface area, but also on the epiphytes on the remaining and re-grown trees. This pattern was sustained after corrections for differences in host-tree size were made. Nevertheless, the epiphytes show resilience to disturbance if forests are selectively logged to spare big trees. Such remnant trees are essential for epiphytes that require the presence of accumulated suspended soil and they may also serve as nearby epiphyte seed sources for the re-growing trees. Anthropogenic disturbances also force a shift from mesic to more drought tolerant species. Hence, disturbance exerts an influence on the regional distribution of epiphytes, but a similar amount of variation (20¿30%, depending on the used spatial descriptors) could entirely be attributed to the geographic position of the sites, independent from the influence of environmental variables. Nearby sites are more similar in epiphytes than distant sites (Mantel¿s r = 0.44, P = 0.002) and I postulate that this relates to differences in seed supply, viewing the development of the epiphyte community from a dispersal assembly perspective. For epiphyte conservation purposes, it is recommended to adopt a management of selective logging instead of cyclic clear-cutting and to spare large trees during logging. In addition, it is wise to try to attain a homogeneous distribution of epiphyte protective reserves within a uniform physiographic regio