COS 6-2
Changes in local to landscape tree species diversity from large catastrophic wind disturbance events in the Northwestern Amazon

Monday, August 10, 2015: 1:50 PM
320, Baltimore Convention Center
Sami Walid Rifai, School of Forest Resources and Conservation, University of Florida, Gainesville, FL
Stephanie A. Bohlman, School of Forest Resources and Conservation, University of Florida, Gainesville, FL
Jose David Urquiza Muñoz, Facultad de Ciencias Forestales,, Universidad Nacional Amazonía Peruana, Iquitos, Peru
Fredy F. Ramírez Arévalo, Facultad de Ciencias Forestales, Universidad de la Amazonía Peruana, Iquitos, Peru
Robinson Negron-Juarez, Earth Sciences Division, Lawrence Berkeley National Lab
Jeffrey Q. Chambers, Earth Science Division, Climate Sciences, Lawrence Berkeley National Laboratory, University of California Berkeley, Berkeley, CA
Background/Question/Methods

The Northwestern Amazon hosts hyperdiverse forests that experience high precipitation and convective activity, which promote a highly active wind disturbance regime that kills trees over varying spatial scales - and at varying degrees of severity. Blowdowns emanating from downbursts can kill a few clustered trees to hundreds of thousands of trees in concentrated areas with spatial extents that can be in excess of 30 square kilometers. We identified a chronosequence of accessible blowdowns in the department of Loreto, Perú using Landsat 5 TM imagery. Linear spectral unmixing was used to estimate the pre- and post- disturbance change in the pixel fraction of non-photosynthetic vegetation, the ΔNPV. Previous work has shown ΔNPV to be an effective linear predictor of within pixel adult tree mortality. We installed forest inventory plots in and outside of five blowdowns that occurred in Loreto between 1988 and 2009, where trees > 10 cm DBH were measured and identified to species. Forest type, and pre-existing level of landscape species richness varied across sites - hence complicating a direct comparison using space for time substitution. Through scaling species richness estimators, we examined how alpha and beta diversity were affected as a consequence of disturbance intensity, and time since disturbance.

Results/Conclusions

The estimated species richness was lower within the blowdowns than in the site-paired intact forests for years one through 16, yet had recovered by 22 years following the disturbance event. Estimated species richness generally declined with increasing disturbance severity, yet this effect was weakest at the youngest site. Using the Chao-Jaccard abundance based estimator, compositional dissimilarity between intact and disturbed forests generally increased with time since disturbance. The successional trajectories of post disturbance species assemblages on terra firme forests were somewhat similar, with fast-growing pioneer species colonizing forest patches that had experienced high mortality from the wind disturbance. Yet the successional trajectory of the post-disturbance forest across all sites was not easily predictable. At one flood-plain forest site, very few pioneers were found in the species assemblage of the blowdown even after six years had passed. Ultimately, catastrophic wind disturbance promoted lower local scale species richness in these high diversity tropical forest ecosystems, but wind disturbance did seem to marginally promote greater landscape scale diversity because of the distinct species assemblages resulting from the disturbance.