Department of Computer Science, Michigan Technological University, Houghton, MI 49931-1225, USA
Comparison of root distributions of species in North American grasslands using GIS
Article first published online: 24 FEB 2009
1997 IAVS - the International Association of Vegetation Science
Journal of Vegetation Science
Volume 8, Issue 4, pages 587–596, August 1997
How to Cite
Sun, G., Coffin, D. P. and Lauenroth, W. K. (1997), Comparison of root distributions of species in North American grasslands using GIS. Journal of Vegetation Science, 8: 587–596. doi: 10.2307/3237211
- Issue published online: 24 FEB 2009
- Article first published online: 24 FEB 2009
- Received 28 May 1996; Revision received 2 February 1997; Accepted 12 March 1997.
- Growth form;
- Quantitative root distribution;
- Anon. (1986)
Abstract. We quantified the spatial distribution of roots of individual plants using detailed drawings from the literature of species of grasses, forbs, and shrubs in the Central Great Plains grasslands of North America. We scanned each two-dimensional drawing electronically and used ARC/INFO, a Geographic Information System, to calculate root length (cm) and density (cm root length/cm soil) with depth in the soil profile. We then selected one of three mathematical models that best fit the data, and classified each species as either shallow-, medium- or deep-rooted. 66 root drawings from 55 species were evaluated. Most plants were shallow-rooted with largest root densities occurring at depths < 20 cm; most maximum rooting depths were > 1m. Grasses had the shallowest maximum depth and shrubs the deepest. Deep-rooted forbs had the smallest root densities by depth. Most plants had two sections to their distribution of root density: an initial increase from the soil surface followed by a decrease in density with increasing depth. Our results suggest that the abundance and importance of different species and growth forms in North American grasslands is related to similarities and differences in the spatial distributions of their root systems. Our approach provides quantitative information on root distributions to be used for comparisons among species, and in simulation modeling analyses that could not be done with conventional methods that are qualitative in nature.