Predictive Modeling of Floral Biodiversity and Invadedness of Michigan Prairie Fen Communities
Location
Hager-Lubbers Exhibition Hall
Description
Prairie fen communities are classified as rare and vulnerable by the Michigan Natural Features Inventory. They contain high levels of biodiversity, including 19 plant and 18 animal threatened or endangered species. The goal of this project is to develop a predictive multiple regression model that can be utilized by land managers and conservation organizations to efficiently and effectively allocate resources towards fen protection and restoration. Floral biodiversity was surveyed in 22 fens located in eight southern Michigan counties during the 2012 and 2013 field seasons, ranging from 1.5-148 acres. Fens were sampled based on regression equations developed using Coleman Rarefaction curves from eight fens surveyed during the 2012 field season to maintain equal sampling effort. Biodiversity indices, including Shannon's and Simpson's Diversity Indices and the Floristic Quality Index, were calculated for each fen used to compare plant dominance and evenness between fens. An NMDS was used to compare overall community composition between surveyed fens. Prairie fen connectivity, size, shape, underlying geology, and water table depth were measured and the surrounding environmental matrix of each fen quantified using Patch Analyst. A multiple regression predictive model was developed based on these environmental and physical fen attributes, the results of which will be presented.
Predictive Modeling of Floral Biodiversity and Invadedness of Michigan Prairie Fen Communities
Hager-Lubbers Exhibition Hall
Prairie fen communities are classified as rare and vulnerable by the Michigan Natural Features Inventory. They contain high levels of biodiversity, including 19 plant and 18 animal threatened or endangered species. The goal of this project is to develop a predictive multiple regression model that can be utilized by land managers and conservation organizations to efficiently and effectively allocate resources towards fen protection and restoration. Floral biodiversity was surveyed in 22 fens located in eight southern Michigan counties during the 2012 and 2013 field seasons, ranging from 1.5-148 acres. Fens were sampled based on regression equations developed using Coleman Rarefaction curves from eight fens surveyed during the 2012 field season to maintain equal sampling effort. Biodiversity indices, including Shannon's and Simpson's Diversity Indices and the Floristic Quality Index, were calculated for each fen used to compare plant dominance and evenness between fens. An NMDS was used to compare overall community composition between surveyed fens. Prairie fen connectivity, size, shape, underlying geology, and water table depth were measured and the surrounding environmental matrix of each fen quantified using Patch Analyst. A multiple regression predictive model was developed based on these environmental and physical fen attributes, the results of which will be presented.