PLANT COMMUNITY RESPONSE TO REGIONAL SOURCES OF DOMINANT GRASSES IN GRASSLANDS RESTORED ACROSS A LONGITUDINAL GRADIENT

Author

Laurel Wilson, Southern Illinois University CarbondaleFollow

Date of Award

5-1-2014

Degree Name

Master of Science

Department

Plant Biology

First Advisor

Gibson, David

Abstract

Questions: Do extended phenotypes of regional sources of Andropogon gerardii and Sorghastrum nutans exist? How does using a mix of dominant grass sources affect the assembling plant community? Do regional sources of Andropogon gerardii and Sorghastrum nutans differ in their use of nitrogen, water and light resources? Location: Kansas, Illinois, USA. Methods: Four reciprocal common gardens experiments were established across a longitudinal climate gradient characterized by a decrease in aridity in western Kansas (COLBY), central Kansas (HAYS), eastern Kansas (PMC) and southern Illinois (ARC). At each site, plots were seeded with dominant grasses (Andropogon gerardii and Sorghastrum nutans) sourced from central Kansas (CKS), eastern Kansas (EKS), southern Illinois (SIL), or a mix of all three regional sources (MIX). All plots were also seeded with the same suite of subordinate species. Species composition was measured during the 4th year of restoration. Resource variables (soil inorganic nitrogen, water and light captured) were measured throughout the season at ARC. Results: The greatest variation between communities occurred at HAYS and ARC, between plots seeded with CKS and SIL sources. At these sites, plots seeded with the local source had the lowest diversity and cover of non-dominant species. Compositional variation between plots seeded with different dominant grass sources was found exclusively at ARC between CKS and SIL plots. Differences between locally seeded plots and plots seeded with a MIX of dominant grass sources were contingent upon site. At ARC, MIX seeded plots had higher diversity than SIL sourced plots. There were no differences in inorganic nitrogen or water content between plots with different dominant grass sources at ARC. Leaf area index was greater in locally seeded SIL plots than in plots seeded with CKS or EKS sources. Conclusions: The results of this study did not provide clear evidence of an extended phenotype associated with regional sources of A. gerardii and S. nutans. Differential sequestration of above-ground (light) resources by the dominant grasses, however, represents a possible mechanism by which an extended phenotype could arise. Seed mixes that contain non-local sources could be used in restoration without risk to the assembling plant community although genetic diversity and potential out-breeding effects should first be considered.