Water potential ('P) values were determined for aqueous coiloids of four molecular sizes of polyethylene glycol (PEG) using freezing-point depression and vapor-pressure deficit methods. A significant third-order interaction exists between the method used to determine 4,., PEG molecular PEG [an inert, nonionic, long-chain polymer: HOCH2-(CH2-0-CH2), CH20H (3)] has been widely used to maintain experimental media at predetermined 4.2 values (2, 7, 11, 14-16).Several researchers have reported that PEG has toxic effects on plants (5, 10). Lagerwerff et al. (8) attributed PEG 6,000 toxicity to associated heavy metals and recommended dialysis or passage through ion exchange columns to remove these impurities. Lawlor (9) suggested that PEG 1,000, 4,000, and 20,000 caused plant desiccation by blocking pathways of H20 movement.In addition to possible toxic effects, debate exists concerning the appropriate method to use for determining 'a of PEG media (8). Applegate (1) noted that the freezing-point depression of PEG 4000-Hoagland solutions of different concentrations deviated from the expected value predicted by van MATERIALS AND METHODS A laboratory advanced osmometer, model 3L (Advanced Instruments Inc., Needham Heights, MA) was used to determine the freezing-point depression of media. Vapor-pressure deficit measurements were obtained for the same media with a dew-point microvoltmeter, model HR-33, and thermocouple chamber, model C-5 1 (Wescor Inc., Logan, UT). All dew-point measurements were done at 23C. Both instruments were calibrated using the same standards. Readings in mosmol and ,uv were converted to corresponding '1. values for statistical analysis. Polythylene glycol was purchased from J. T. Baker Chemical, Phillipsburg NJ.

The effects of fire season on forb diversity patterns, density, and composition were determined for a northern Mixed Prairie site, USA. Repeated spring burns (dormant season), summer burns, fall burns (dormant season), and unburned treatments were compared over a 3-yr period characterized by wet and dry moisture conditions. Alpha and beta diversity were highest on unburned and summer burn treatments, while landscape mosaic diversity was highest on fall burns. Forb density was highest on fall and spring burn sites. Nine forb species comprised 82~o of total densities and were significantly affected by fire season and year to year variations in moisture. Forb composition for unburned and spring burn treatments was similar, but both treatments were different from the summer burn and fall burn treatments which were similar to each other. Fire alone did not appear to be an intense enough disturbance to initiate drastic changes in the forb component of vegetation patches. Specific fire seasons did appear to either mask or enhance forb structure arising from other disturbance(s). Fire season also affected the scales of forb organization in the landscape. Contrasting spatial characteristics of the forb component of prairie plant communities may provide a diagnostic technique for exposing the interaction of disturbances at different temporal and spatial scales.Nomenclature:

No abstract

Native herbivores are generally preferred over introduced species for managing prairie preserves. However, contemporary prairies are fragments of the original ecosystem, and their size and management may not match the foraging and/or social behavior of native herbivores. We conducted field studies from May 1988 through April 1991 within a 3000‐ha sandhills prairie to describe bison (Bos bison) and pocket gopher (Geomys bursarius) use of native Great Plains grassland at a size relevant to current land ownership and management units. The distribution and diet of bison and pocket gophers were compared to vegetation patterns at several scales. The study area is characterized by a matrix of warm‐season grasses occurring on four range sites (topo‐edaphic units). Patches characterized by forbs, shrubs, and trees occur within the matrix communities associated with the four range sites. Bison distribution appears to be influenced by a complex interaction between seasonal forage availability, herd size, and habitat openness. Bison diets during the growing season are strongly linked to the warm‐season grasses of the prairie matrix. Pocket gopher activity is ubiquitous throughout the study area, but is particularly high in forb patches and appears to be responsive to bison impacts. Pocket gopher diets are strongly linked to forbs, both in the forb patches and in the prairie matrix. Field data indicate that bison and pocket gophers use herbage resources that have distinctly different distributions within the sandhills prairie. It appears that bison remain well adapted to prairie remnants of the size we studied, and through interactions with other herbivores and vegetation patterns, will support a diverse landscape.

This study was designed to: 1) compare the landscape distribution patterns of bison on fire-managed prairie remnants in the tallgrass (Oklahoma), and mixed prairie (Nebraska); and 2) identify the extent to which fire and range site [topoedaphic classification of the landscape] affect bison distribution patterns. This research was conducted at 2 sites: the Niobrara Valley Preserve (1990-1996), and the Tallgrass Prairie Preserve (1993-1996). At both preserves, bison selected burned areas during the growing season for 1-3 years and mostly avoided old burns and unburned areas. There was an interaction between fire and range site in selection patterns. In the absence of fire, bison mostly avoided both the Choppy Sand range sites at the Niobrara Valley Preserve and Loamy Prairie range sites at the Tallgrass Prairie Preserve. When they were burned, however, these sites were highly selected. The main difference in bison selection patterns between the Tallgrass Prairie Preserve and the Niobrara Valley Preserve was observed during the dormant season. In the Tallgrass Prairie Preserve, burned sites continue to be preferred during the dormant period for an average of 2 years while in the Niobrara Valley Preserve selections were random. These differences can be explained by 2 mechanisms: the fall and winter re-growth of forage at the more southern latitude and the significant physiognomic changes that fire can cause in tall grass prairies. Our study documents a continuing interaction between the ecological processes of the fire regime and bison distribution and abundance within 2 of the major prairie landscapes of the Great Plains, and provides critical details for understanding this relationship.