Seedling recruitment of Dalmatian toadflax, (Linaria genistifolia ssp. dalmatica (L.) Maire and Petitmengin (Scrophulariaceae)), was examined in a 2-year field study in Montana using overseeding and plant/insect exclusion methods, to determine whether it was more limited by seed availability or interspecific plant competition. Overseeding test plots with toadflax seed had no effect on seedling recruitment. Exclusion of plant competition (via herbicide application and pruning) significantly increased total, and cumulative seedling recruitment of Dalmatian toadflax on the last sampling date in 3 of 4, and 2 of 4 cases examined, respectively. Insect exclusion (via insecticide application) significantly increased total seedling recruitment of Dalmatian toadflax on the last sampling date in only 1 of 4 cases examined, and had no effect on cumulative seedling recruitment of Dalmatian toadflax on the last sampling date. We conclude that seedling recruitment in Dalmatian toadflax was more strongly influenced by plant competition than herbivory in our study. Hence, microsite limitation (i.e., competition for "safe sites for germination") rather than seed limitation appears to play a more important role in toadflax seedling recruitment. In light of this, current biological control agents that impact seed production will likely have minimal capabilities of influencing toadflax density. Thus, a premium should be placed on establishing biological control agents that are able to cause significant damage to the stem and root system of Dalmatian toadflax, and in maintaining a healthy plant community that, through interspecific competition, will negatively affect toadflax seedling recruitment.

High concentrations of condensed tannins in browse impair brush clearing by goats. We studied the effect of polyethylene glycol (PEG, MW 4000), a polymer that binds condensed tannins, on the feeding behavior of Damascus goats (Capra hircas) on a range dominated by tannin-rich lentisk (Pistacia lentiscus L.). This was done with or without a nutritious alternative to browse (alfalfa hay) available at pasture. In phase 1, no hay was provided to goats; in phase 2, hay was distributed daily in the field. In both phases, 6 goats had free access to PEG while at pasture, while 6 goats that grazed separately on another paddock did not. All goats received each night an allowance of concentrate (400 g day-1 of 40% ground corn grain, 40% ground barley, 17% soybean meal and 3% of a mineral-vitamin premix, and contained as fed 16% crude protein (CP) and 2.66 Mcal kg-1 of Metabolizable Energy). The 2 groups of goats alternated daily between paddocks. Goats supplemented with PEG spent more time browsing lentisk than goats in the control group (73 and 41%, respectively, P < 0.0001). Goats in the control group spent more time foraging on dry grasses than their PEG-fed counterparts (28 and 12%, respectively, P < 0.0001). Goats from the PEG group gained body weight at a higher rate than controls. The daily intake of PEG was 450 g, with an intake rate of 1.2 g sec-1. Supplemental alfalfa hay substituted partly for dry grasses in goats' diets, but did not modify the percent of time goats in either treatment spent browsing lentisk. Our data suggest that PEG has the potential to increase intake of tannin-rich species, even where alternative fodder of better nutritional quality is present. However, self-feeding of PEG may not be the best way to provide PEG because goats may ingest more PEG than needed to annul the aversive effects of tannins on food intake.

Effective rangeland weed programs require the ability to predict plant community responses to management. Our objective was to develop regression equations to predict the plant community after control with picloram using the pre-treatment plant community. Five transects were established from dense spotted knapweed (Centaurea maculosa Lam.) in the center of each patch to an area of low or no spotted knapweed occurrence on the outside of the patch. Transects ended in areas dominated by Idaho fescue (Festuca idahoensis Elmer). Twenty permanent plots (20 x 50 cm, spacing along the transect ranged from 1/2 to 2 m) were placed along this gradient. Pre-treatment density and cover of all species were sampled in each plot. Biomass of all species was harvested in plots adjacent to the transect. Picloram (4-amino-3,5,6-trichloropicolinic acid) was applied along each transect at a rate of 0.28 kg a.iha-1 in October 1996 to each plot. Density, cover, and biomass of all species were re-sampled in August 1998. Regression models were fit using perennial grasses, Idaho fescue, forbs, species richness, and species diversity after treatment as predicted variables. All predicted variables were indigenous species. Regressor variables used were site, transect, and spotted knapweed, a spotted knapweed quadratic component, indigenous perennial grasses, Idaho fescue, indigenous forbs, species richness, and species diversity sampled in the first year (1996) prior to treatment. The best predictive models for assessing post-management indigenous perennial grass, Idaho fescue, and species richness were based on density. The best models predicting post-management forbs and species diversity were based on cover and biomass, respectively. In 4 out of the 5 models, for a given post-management parameter, an important predictor in the model was its pre-management regressor variable. Additionally, pre-management spotted knapweed was a relatively unimportant preditar in most models. The model predicting species diversity based on density (pre-treatment) predicted an increase in species diveristy 2 years after management. This study indicated that it may be feasible to use pre-management spotted knapweed was a relatively unimportant predictor in most models. The model predicting species diversity based on density (pre-treatment) predicted an increase in species diversity 2 years after management. This study indicated that it may be feasible to use pre-management plant community data to predict post-management plant community response for spotted knapweed-infested rangeland using picloram.

There are several generalizations or assumptions concerning rangeland hydrology and erosion relationships found in the literature and in the management arena. These generalizations have found their way into rangeland models, where modelers have assumed that diverse rangeland types can be lumped or averaged together in some way to develop one algorithm or equation to describe a process or relationship across the entire spectrum of rangeland types. These assumptions and modeling approaches based on the universal concept may not be appropriate for diverse rangeland types. This paper presents a comprehensive data set of vegetation, soils, hydrology, and erosion relationships of diverse western rangelands, and utilizes the data to assess the validity of the various assumptions/generalizations for rangelands. The data set emphasizes the difficulty in understanding hydrologic responses on semiarid rangelands, where the relationship between plant/soil characteristics and infiltration/erosion is not well established. When all sites were pooled together, infiltration and sediment production were not correlated with any measured vegetation or soil characteristic. A myriad group of factors determine infiltration and erosion, and is dependent on rangeland type and site conditions. The infiltration and erosion responses and correlation/regression analyses presented highlight the risk of using generalized assumptions about rangeland hydrologic response and emphasize the need to change the current modeling approach. Universal algorithms to represent the response of all rangeland types, such as the pooled multiple regression equations presented, will not provide sufficient accuracy for prediction or assessment of management. We need to develop a rationale to organize rangeland types/vegetation states according to similarities in relationships and responses. These functional rangeland units would assist in the development of more accurate predictive equations to enhance model performance and management of rangelands.

Stocking rate is generally considered to be the most important management factor in sustainable grazing of Great Plains grasslands over management periods of 10 to 20 years or longer. Most studies to determine optimum stocking rates have compared only 2 or 3 discrete stocking rates. Our objective was to determine cow, calf, and economic performance on sand sagebrush (Artemisia filifolia Torr.) rangeland as a continuous function of stocking rate. Replicated stocking rates of 0.11, 0.15, and 0.22 head ha-1 were studied over an 8-year period. Cow weight declined as stocking rate increased in drought years but was not affected by stocking rate in wetter years. Weaning percentage was not affected by stocking rate but variation within treatment groups was high. Calf birth weight and weaning weight both declined as stocking rate increased. Comparing 0.11 and 0.22 head ha-1, calf production cow-1 declined from 206 to 144 kg cow-1 as stocking rate increased but calf production ha-1 increased from 22.6 to 31.7 kg calf ha-1. Net returns were maximized at 7.87 ha-1 year-1 at a stocking rate of 0.172 head ha-1, well within the range of experimental treatments. Net returns were within 5% of maximum between stocking rates of 0.156 and 0.183 head ha-1. The variability of all responses increased as stocking rate increased. Simulation indicated that improved livestock prices and increased animal productivity shifted the economic optimum stocking rate to higher levels, which would put more pressure on the conservation ethic of land managers.

We used microhistological analyses of fresh fecal pellets to determine seasonal diets of desert bighorn sheep (Ovis canadensis mexicana Merriam 1901) in northwestern Sonora, Mexico from April 1997 to December 1998. We identified 41 plant species (22 browse, 10 forbs, 5 grasses, and 4 succulents) in diets of bighorn sheep. We found no differences between diets of males and females, and diet diversity between sexes was similar (P 0.05). Diet included: browse (45.7%), forbs (32.0%), succulents (17.8%), and grasses (4.5%). The consumption of succulents was higher during spring, decreased during summer, increased in autumn, and decreased in winter. Consumption of forbs was higher during winter and summer. Globemallow (Sphaeralceae spp.), desert agaves (Agave spp.), range ratany (Krameria parvifolia Benth.), buck-wheatbrush (Eriogonum spp.), foothill palo verde (Cercidium microphyllum [Torrey] Rose Johnst.), Engelmann prickly pear (Opuntia engelmanii Salm-Dyck), desert ironwood (Olneya tesota A. Gray), and elephant tree (Bursera microphylla A. Gray) were consumed throughout the study. As biologists identify potential release sites for restoration of bighorn sheep in Mexico, studies of diet composition will provide managers with information for successful translocations.