Thursday, August 6, 2009 - 9:20 AM

COS 86-5: Large-scale damage assessment from Hurricane Dean in the dry tropical forests of the Yucatan Peninsula

Jennifer A. Holm1, Deborah Lawrence1, Laura Schneider2, John Rogan3, and Birgit Schmook4. (1) University of Virginia, (2) Rutgers University, (3) Clark University, (4) El Colegio de la Frontera Sur

Background/Question/Methods

Hurricanes are natural disturbances that have been affecting tropical forests in the Caribbean and in Central America for millennia.  This area encompasses a wide range of forest types, landscapes, and land uses all of which will respond differently to natural disturbances.  Thus, it can be expected that forest damage will differ across these varied forested landscapes. This study examines landscape scale changes in ecosystem processes, land cover change, and biogeochemical dynamics the dry tropical forests of the southern Yucatan Peninsula, which has a mosaic of many different landscapes. Hurricane Dean passed over the Southern Yucatan during August of 2007.  Vegetation and soil samples have since been collected in 93 plots within and outside protected areas, and across a wide range of land types during the summer of 2008. Natural observations are used to answer how the effects of hurricane disturbance differ across varying landscape categories.

Results/Conclusions

I measured the response of two dependent variables to four fixed independent variables. The response variables are: (1) the level of damage based upon on ground vegetation measurements of damage and (2) remotely sensed classifications of damage. These fixed categories are: (1) soil fertility, (2) forest type, (3) forest age, and (4) land use histories. In the past, the majority of hurricane disturbance studies have looked at homogeneous forest types and small scale plot levels. Damage analysis at the landscape level of the Southern Yucatan will allow for scaling up locally sampled disturbance damage. We have found that distribution of damage does vary over the region as a function of multiple factors. Likewise, there is a relationship between levels damage incurred to trees over this geographic distribution. This large scale analysis gives us a clearer projection of the fate of similar landscapes in the future and consequently how to prepare for long lasting damages.