Home
Up
Biodiversity
Bobcats
Blues
Cormorants
Grassland birds
Bat acoustics

BIODIVERSITY

Predicting the effects of land-use change on Vermont's biodiversity.   

Principal Investigator:  Therese Donovan, Austin Troy, Alexey Voinov

Cooperators:

Graduate Students:  

Post Docs:

Funding Sources:

Abstract: Land use change involves the conversion of one land cover type to another, and often results in the loss and fragmentation of existing wildlife habitat. Although large areas of contiguous habitat still remain in the Northern Forest , increasing development pressures results in the conversion of forests land types to urban, suburban, or exurban land types. In Vermont, for example, the rate of land development is ~ 260% greater than the rate of population growth, where land growth consists primarily of “small urban build up,” i.e., small, unconnected blocks of land that develop within a matrix of less developed land – an effective facilitator of habitat fragmentation for many wildlife species.  These trends indicate that research focused on understanding the impacts of land use change on the region’s native flora and fauna is much needed.  Because different species are expected to respond to land use and climate change in different ways, these assessments should target both individual species and changes in species richness and biological community composition over time.  

In 2003-2004,  we surveyed plants and animals at over 200 sites across Vermont.  The dataset consists of 36,000 bird records, 7,000 amphibian/reptile records, 200 carnivore records, and several thousand invertebrate records and tree records that were collected  Sites were separated by ≥ 8 km, and consisted of 4 survey stations that were separated by 500 meters.  Survey methods were taxon-specific, with either site or station as the sampling unit of interest (e.g., Maneras et al. in press, Long et al. 2007 a, b).  Local vegetation was quantified at each station with standard sampling techniques, and  landscape pattern surrounding each station was characterized with FRAGSTATS® at taxon-specific buffers, based on 2001 National Landcover.  The data for vertebrates were analyzed as presence-absence on a species-by-species basis with a single-season occupancy-modeling framework with local, landscape, and climatic variables as covariates, which allowed us to generate probability of occurrence maps for multiple taxa at a 30m² spatial scale across the entire state of Vermont. These maps provide the baseline data upon which future land use change scenarios can be quantitatively assessed.   

Publications:

Selected Links:

Send mail to tdonovan@uvm.edu with questions or comments about this web site.
Last modified: March 12, 2008