UVM Theses and Dissertations
Format:
Print
Author:
DeSantis, S'ra
Dept./Program:
Plant and Soil Science
Year:
2010
Degree:
MS
Abstract:
This study was designed to characterize the extent to which farm management practices, biophysical features, and landscape variables affected plant species richness, abundance, and diversity of treed habitats on vegetable farms in the Champlain Valley of Vermont. Plants were inventoried in the overstories and understories of 12 treed habitats adjacent to organic fields and 4 treed habitats adjacent to conventional fields in Vermont, USA. A total of 18 independent variables were tested -nine landscape, three management, and six biophysical. In linear regressions and t tests, the overstory revealed more significant results than the understory. Landscape variables were the most influential factors determining plant diversity, followed by biophysical features. Management practices did not influence plant diversity metrics except for species diversity in the overstory, which surprisingly was higher in treed habitats on conventional farms.
In the overstory, all three dependent variables were higher in rural location and significantly correlated with decreasing levels of landscape heterogeneity, and increasing percent treed habitat, elevation, and percent organic matter. For the overstory, a multiple regression test revealed that elevation, percent organic matter, landscape heterogeneity, and percent treed habitat explained 74% of the variance in species richness, 80% of the variance in species abundance, and 66% of the variance in species diversity. Other studies have observed that increasing levels of landscape heterogeneity and treed habitats adjacent to organic farms corresponds to higher levels of plant diversity. However this study found the opposite, most likely because the studied region is situated in a forested matrix, as opposed to an agricultural matrix. Overall, 82 % of the species inventoried in the understory and overstory were native, revealing that treed habitats in agricultural landscapes have the capacity to conserve native plant diversity.
In the overstory, all three dependent variables were higher in rural location and significantly correlated with decreasing levels of landscape heterogeneity, and increasing percent treed habitat, elevation, and percent organic matter. For the overstory, a multiple regression test revealed that elevation, percent organic matter, landscape heterogeneity, and percent treed habitat explained 74% of the variance in species richness, 80% of the variance in species abundance, and 66% of the variance in species diversity. Other studies have observed that increasing levels of landscape heterogeneity and treed habitats adjacent to organic farms corresponds to higher levels of plant diversity. However this study found the opposite, most likely because the studied region is situated in a forested matrix, as opposed to an agricultural matrix. Overall, 82 % of the species inventoried in the understory and overstory were native, revealing that treed habitats in agricultural landscapes have the capacity to conserve native plant diversity.