COMPETATIVE EVQUIVALENCY OF CULTIVAR AND NON-CULTIVAR DOMINANT GRASSES IN AN EXPERIMENTAL RESTORATION

Author

Lewis Reed, Southern Illinois University CarbondaleFollow

Date of Award

5-1-2010

Degree Name

Master of Science

Department

Plant Biology

First Advisor

Baer, Sara

Second Advisor

Gibson, David

Abstract

AN ABSTRACT OF THE THESIS OF LEWIS KENNEDY REED, for the Master of Science degree in PLANT BIOLOGY presented on October 30, 2009, at Southern Illinois University Carbondale TITLE: COMPETATIVE EVQUIVALENCY OF CULTIVAR AND NON-CULTIVAR DOMINANT GRASSES IN AN EXPERIMENTAL RETORATION MAJOR PROFESSORS: Dr. Sara G. Baer and Dr. David J. Gibson Multiple population sources of species for use in prairie restoration exist, including cultivars and non-cultivars of dominant native grasses. However, little is known about the competitive equivalency of different population sources of dominant C4 grasses and whether intraspecific variation in their competitive effect on the community scales to affect ecosystem assembly. In 2006, an experimental restoration was established in a former agricultural field using cultivars and non-cultivars of the dominant grasses (Andropogon gerardii, Sorghastrum nutans, and Schizachyrium scoparium) and two different species pools of non-dominant species containing equal richness and distribution of species among functional groups. I evaluated inter - and intra-specific variation in the competitive equivalency of A. gerardii, S. nutans, and S. scoparium in terms of community response to their individual and complete (all three species) removal. Removals were maintained for two growing seasons, though little maintenance was needed after the initial treatment. The competitive effect of each grass species and population source was determined by calculating a response ratio for percent cover and above ground primary productivity between removal and control plots for each dominant grass individually, forbs, non-dominant grasses, and legumes which were each analyzed using a mixed model procedure for a split-split-plot randomized block design. Effects of removals on overall community composition were assessed using non-metric multi-dimensional scaling (NMDS) and analysis of similarity (ANOSIM). In addition I monitored changes in photosynthetically active radiation (PAR) at the soil surface and inorganic soil nitrogen. Based on the 2008 data (after two years of maintaining removals), several important differences were detected between the focal species and in some cases the two sources of the same species in terms of their effects on neighbors, inorganic soil nitrogen, and PAR. Many of these differences depended on which species pool the comparison took place in. A 3-way interaction was detected between species pool, source, and removal treatment in the volunteer forb group (F3, 60 =3.28, p = 0.0268). Volunteer forbs showed a positive response to removal of cultivars of the dominant grass functional group in one species pool but not the other. A three-way interaction was detected between species pool, source, and removal treatment in terms of Bray-Curtis similarity (F3, 60 = 2.91, p = 0.0417). Within one of the species pools, similarity values of communities were higher between removals and controls in cultivar plots than in non-cultivar plots where A. gerardii was removed. While NMDS ordination showed separation of some plots by dominant grass source, within group variation was higher than among group variation and ANOSIM deemed this separation insignificant. Planted forb ANPP exhibited a significantly negative response to removal of S. nutans that was not observed in the other removal treatments (F3,41.1 = 3.09, p = 0.038) suggesting facilitation by the dominant grass on planted forbs. The aboveground net primary productivity (ANPP) of the subordinate community (i.e. all species except the dominant grasses) exhibited a 3-way interaction among species pool, dominant grass source, and species removed (F3,61 = 2.76, p=0.0499). This interaction resulted from a stronger negative response of subordinate community ANPP to cultivars of the dominant grass functional group removal than non-cultivars of this group that only occurred in one of the species pools. There was a significant main effect of species removed on %PAR at ground level (F3, 60 = 4.84, p = 0.0044). All removal treatments allowed higher penetration of PAR to ground level, with S. nutans and all dominant grass removal having the strongest effect. Inorganic N availability was lower in response to removal of A. gerardii cultivars compared to the removal of non-cultivars of this species in one species pool but not the other. Removal of cultivar S. scoparium lead to a positive response in total inorganic nitrogen while removal of non-cultivars of this species lead to a negative response in total inorganic nitrogen in this same species pool (F3, 51 = 3.61, p = 0.018). Results demonstrate that inter- and intra-specific variation among these dominant species affect some aspects of community structure and ecosystem properties, but these effects are not consistent among dominant species and among subordinate species pools. These complex interactions may have important implications for restoration and land management.