Changes in herbaceous plant density and canopy cover of creosotebush (Larrea tridentata Sesse & Moc. ex DC) and associated shrubs following brush control treatments were measured in Sonoran and Chihuahuan Desert communities. Treatments were applied in 2 successive years st the Santa Rita Experimental range, Arizona, and 3 locations in Chihuahua, Mexico. Across all locations and years 1.5 kg/ha tebuthiuron (N-[5-(1,1-dimethylethyl)-1,3,4-thiadiazol-2-yl]-N,N'-dimethylurea) > 1.0 kg/ha tebuthiuron= disking = disking with furrowing >2-way railing >0.5 kg/ha tebuthiuron > land imprinting in reducing canopy cover of creosotebush and associated shrubs. At the Santa Rita Experimental Range annual precipitation was above long-term mean in 1982, 1983, 1984, and 1985; and grass density increased on all treated and untreated plots. Annual precipitation was below long-term mean during 1986 and 1987 and grass density decreased on both treated and untreated plots but did not decrease to pre-treatment densities. Forb densities were less than 3 plants/m2 throughout the study, except in 1987 when Russian thistle (Salsola iberica Sennen & Pau) increased on all plots. At the Chihuahuan locations, grass densities usually increased during the first year of the study, but very low precipitation throughout the study caused subsequent reductions in grass and forb densities. In dry years brush control treatments did not increase herbaceous plant density.

Knowledge of how resident wildlife populations respond to brush management strategies is especially limited for rangelands in the cross timbers vegetation type of Oklahoma. We examined how cottontail rabbit (Sylvilagus floridanus) density and habitat use were influenced by applications of tebuthiuron or triclopyr, with and without annual burning, on cross timbers rangeland. Line transect flush-counts, mark-recapture livetrapping, and fecal pellet counts were used to evaluate seasonal differences in population density among 5 brush control treatments. Cottontail rabbits (n = 225) were flushed along 362 km of line transects during 5 census periods. Density in winter was consistently lower than summer for all treatments, except for the untreated control in winter 1987. Line transect density estimates varied from 0 to 1.975 rabbits/ha and suggested that herbicide and annual burning treatments had a positive influence on cottontail rabbit populations compared to untreated controls. Mark-recapture density estimates did not differ among treatments. Fecal pellet counts were greater on herbicide-treated pastures than an untreated control in both spring and fall. Prairie-eastern redcedar (Juniperus virginiana L.) and forest-prairie ecotone habitats were utilized greater than expected by cottontail rabbits. Mature hardwood overstory and mixed-brush habitats were avoided. Tebuthiuron and triclopyr effectively deceased hardwood overstory and increased preferred habitats for cottontail rabbits.

A key to management of short duration grazing systems is maintaining proper rest periods for individual pastures, but information on the necessary length of rest periods for tallgrass prairie is limited. Research hypotheses for this study were that tallgrass prairie plant communities would respond differently to grazing schedules incorporating rest periods of varying lengths and that this response would be dependent on stocking rate. Treatments consisted of 3 grazing schedules (2, 3, or 4 rotation cycles per 152 day grazing season) and 2 stocking rates (1.6 and 2.2 times the moderate continuous rate). Plant frequency, standing crop, species composition, and forage utilization were sampled from 1985 to 1989. Precipitation was above average in 4 of the 5 study years. Grazing schedule did not affect any vegetation parameter over time. Stocking rate did not affect plant frequency or species composition. Standing crop was reduced and forage utilization increased at the higher stocking rate but these effects were consistent over time. Frequency of western ragweed [Ambrosia psilostachya DC.] and the relative species composition of the forb component increased in all grazed pastures compared to ungrazed pastures. The overall lack of major treatment effects was attributed to favorable precipitation, spring burning, and the initial high-seral successional stage of the experimental pastures.

Mineral salt supplements are used in attempts to reduce cattle losses to tall larkspur (Delphinium spp.). We determined the effects of a mineral salt mix on larkspur consumption, ruminal fluid kinetics, and water intake during 4 periods in June, July, and August, 1988 (Trial 1), and during an 18-day grazing period in August, 1989 (Trial 2). In 1988, 12 ruminally cannulated heifers were divided into 3 treatment groups: control with no access to mineral (CONT), 0.5 g mineral (LOW), and 1.0 g mineral kg body weight-1 day-1 (HIGH) dosed intraruminally. In 1989, 10 cows were allocated to either a control group or 0.75 g mineral kg body weight-1 day-1. During Trials 1 and 2, consumption of larkspur peaked at 5 and 7% of cattle diets, respectively; these low levels were attributed to drought. There were no differences (P>0.1) in consumption of total larkspur or larkspur plant parts. Mineral supplement increased water consumption (P<0.05) during Trial 1, but not during Trial 2. The HIGH group averaged 0.1 liters kg body weight-1 day-1 compared to 0.07 liters for the CONT and LOW groups. Ruminal fluid passage rate, turnover time, volume and fluid outflow rate (FOR) did not differ (P>0.05) among treatments during Trial 1, but FOR was increased by mineral treatment in Trial 2. Alkaloid concentration in larkspur deciined with maturity, and was apparently elevated by drought in Trial 2. This study found little indication that mineral salt supplement altered the amount of larkspur consumed by grazing cattle. Increased water intake one summer did not alter rumlnal fluid kinetics. If dietary minerals alter toxicity of larkspur to cattle, other mechanisms than those tested are responsible.

Justification for controlling honey mesquite (Prosopis glandulosa Torr. var. glandulosa) on rangelands has been traditionally related to enhanced livestock production from increased herbage production. More recently, however, it has been hypothesized that control would also increase off-site water yield. The objective of this 3-year study was to quantify the effects of control of individual honey mesquite trees inside nonweighable lysimeters on herbage standing crop, leaf area, and aboveground production. Utilizing frequent harvest techniques, estimated aboveground net primary production (ANPP) in intact tree lysimeters averaged 235 g/m2. Estimated ANPP in the treated lysimeters averaged 349 g/m2. The increased ANPP, following removal of the trees, resulted in significantly greater amounts of herbaceous leaf area and standing crop. The increase in ANPP was relatively uniform regardless of distance from the trunk of removed trees and was the result of increased production by those herbage species present at time of control rather than a shift in species composition. The dominant species in both treatments was Texas wintergrass (Stipa leucothrica Trin. & Rupr.). Sideoats grams [Bouteloua curtipendula (Michx.) Torr.] was a subdominant. The results, in combination with concurrent water yield studies, suggest control of honey mesquite will not enhance water yields dramatically in this region in the absence of livestock grazing.

Seed of 8 Triticeae species was planted in petri dishes (depth = 0) and 2, 4, 6, and 8 cm deep in soil in the greenhouse. Total emergence at 6 weeks after planting declined significantly with increased depth and the species by depth interaction was significant (P < 0.001). Dahurian wildrye (Elymus dahuricus Turcz. ex Griseb.) and tetraploid Russian wildrye (Psathyrostachys juncea [Fisch.] Nevski) exhibited similar emergence from deep seedlings and were superior to all other species except Altai wild ryegrass (Leymus angustus [Trin.] Pilger). The better emergence of the tetraploid Russian wildrye entry compared to diploid cultivar suggests that the establishment of the tetraploid cultivar will be less affected by poor seed depth control. Newly released cultivars of Dahurian wildrye will be less affected by variable seed depth than several of the species currently recommended for seedling rangelands.

Cattle trampling has been recommended to bury seeds and encourage seedling establishment but has not been compared with traditional seedbed preparation techniques. We compared seedling emergence of broadcast-seeded 'Vaughn' sideoats grams [Bouteloua curtipendula (Michx.) Torr], 'A-130' blue panic (Panicum antidotale Retz.), 'A-68' Lehmann lovegrass (Eragrostis lehmanniana Nees), and 'Cochise' atherstone lovegrass (Eragrostis lehmanniana Nees X E. tricophera Coss and Dur.) on lightly and heavily trampled seedbeds with that on undisturbed, land imprinted, and root-plowed or ripped seedbeds. We seeded and applied the treatments prior to summer rains on a sandy loam soil in southern Arizona for 3 years. In a wet year (1987) when surface soil water was estimated to be available for at least 24 consecutive days, heavy trampling and land imprinting increased emergence of blue panic and land imprinting increased emergence of Cochise lovegrass. In that year, lovegrass emergence was high even on undisturbed plots. In a moderately wet year (1988), surface soil water was available for periods of 6-9 days during seedling emergence and greater disturbance, either by heavy trampling, land imprinting and/or root plowing or ripping produced higher emergence than light trampling and nondisturbance. In a dry year (1989), surface soil water was available for periods of 2-3 days and seedling emergence was low and generally similar for all treatments. Sideoats grama emergence was low all 3 years, but was highest in 1988 when initial thunderstorms were followed closely by subsequent storms. Seedbed disturbance by cattle and mechanical methods may enhance revegetation in the Southwest in years of moderate precipitation but may be unnecessary in wet years or futile in dry years, depending on species and soils.

Harvesting trees for firewood in the oak hardwood rangelands of the western Sierra Nevada foothills creates slash that may be burned to improve livestock forage production or piled into brushpiles for wildlife. The economics of these actions are undocumented. We observed a firewood harvest that created 378 brushpiles averaging 13.6 m2 and 1.3 m high, and resulted in a forage loss of 1,807 kg dry weight, equivalent to 4.4 AUM. We projected the present net value of 5 management options concerning the removal or retention of brushpiles during a 15-year period. Inputs included revegetation of burned-brushpile sites, annual forage production on areas with the oak canopy removed, burning and reseeding costs, and income derived from cattle grazing and quail hunting. The options were: (A) burning all brushpiles and reseeding the burned sites; (B) option A without reseeding; (C) burning 235 brushpiles and reseeding, leaving 23 brushpiles/ha for quail; (D) option C without reseeding; (E) leaving all brushpiles. All but option B were economically feasible at a 4% interest rate; at an 8% interest rate, only options C-E were profitable. After 15 years, the accumulated returns per hectare at 4% for options A-E were 11.67, -3.97, 32.43, 22.29, and 23.35, respectively, and at 8%, -17.35, -25.74, 8.58, 3.02, and 17.98, respectively.