Potential Project Description for 2010-11
Title: Prairie restoration: combining laboratory, field and statistical methods to enhance success
Domain Experts: Diane Angell, Biology; Kathy Shea, Biology
Prairie restoration: combining laboratory, field and statistical methods to enhance success Prairies are one of the most endangered biomes. In Minnesota, as in many other states the deep rich soils of the prairie were plowed up to cultivate corn, soy and other crops. In order to restore landscapes, time and energy is spent reconstructing prairies on lands that were once cultivated. Understanding how soils affect the germination and early seedling growth of prairie plants is therefore important. Soils that have been cultivated for many years have lost soil organic matter leading to a depletion of terrestrial carbon and nitrogen pools (Knops and Tilman 2000; Camill et al. 2004).
In ecology courses, students are collecting experimental data in growth chambers on the relative success of two species of prairie plant, in soils taken from a remnant prairie, a cultivated field, and a restored prairie. We are interested in whether cultivated soils are as conducive to prairie plant germination and growth as soils that support a rich set of prairie species. In some of the class projects half of the soil from each soil type has been sterilized to determine if rhizosphere differences are responsible for any observed variation. An immediate increase in soil carbon with the post-restoration of prairie plant species is thought to be influenced by the formation of microbial byproducts (McLauchlan, Hobbie and Post 2006). The experiment is set up as a 3x2x2 factorial design at the pot level, with data collected longitudinally. Basic data analysis is based on fixed effects analysis of variance by species with soil type and sterilization as fixed effects, but CIR students will have a chance to contribute to the research problem by conducting statistics workshops for ecology students, offering advice on the experimental design, and bringing more powerful statistical modeling tools to the data analysis, including mixed effects models, longitudinal data models, and survival analysis. Statistics students will analyze data collected in multiple years and look for any patterns in germination and growth of two prairie species with distinctive life history characteristics in different soil types.
Camill, P., M. J. McKone, S. T. Sturges, W. J. Severud, E. Ellis, J. Limmer, C. B. Martin, R. T.Navratil, Knops, J. M. H. and D. Tilman. 2000. Dynamics of soil nitrogen and carbon accumulation for 61 years after agricultural abandonment. Ecology 8:88-98.
McLauchlan, K.K., S. E. Hobbie, and W. M. Post. 2006. Conversion from agriculture to grassland builds soil organic matter on decadal timescales. Ecological Applications 16:143-153.Contact: Diane Angell (Biology), Kathy Shea (Biology)