This research addressed the hypothesis that spring precipitation data can be used to detect agricultural drought early in the growing season. The Rangetek range model was used to simulate yearly forage data based on historical precipitation and temperature records from the USDA-ARS Fort Keogh Livestock and Range Research Laboratory (Miles City, MT) and the Agriculture and Agri-Food Canada Manyberries Substation (Lethbridge, AB, Canada). Monthly total precipitation and monthly average maximum and minimum temperatures were used to develop regression equations predicting growing season forage production at the Fort Keogh Laboratory and Manyberries Substation. At Fort Keogh Laboratory, a combination of fall (October and November) and spring (April and May) precipitation were predictors of simulated forage yield index (P < 0.01, R2 = 0.84). At Manyberries Substation, April and May precipitation were predictors of simulated forage yield index (P < 0.01, R2 = 0.44). Using the actual forage data from Manyberries Substation yielded similar results, in that April, May, and June were predictors of forage production (P < 0.01, R2 = 0.50). Although the regression equation for actual forage production data from Manyberries Substation did indicate that July precipitation was a significant predictor, adding July precipitation did not increase the ability of the equation to detect reduced forage production. These results imply that annual forage production can be estimated with considerable confidence by July 1 and that forage produced by early July is a good indicator of total growing season forage production. Early season detection of drought effects on forage production provides much-needed flexibility in devising management alternatives to minimize the negative impacts of drought on rangelands and beef enterprises.

The objective of this study was to evaluate alternative drought management strategies for their effects on beef cowcalf enterprise profitability based on early detection of drought. A bioeconomic model was parameterized to represent a range-based cow-calf production system in the Northern Great Plains. The base management system was characterized by inputs required to maintain herd size of approximately 511 cows during an average climatic year with a fixed forage base. Treatments were factorially arranged where management (early vs. normal), intensity of drought (moderate, 20% reduction in available forage vs. severe, 40% reduction in available forage), and forage quality [average CP (%), ME (Mcal/kg), and NDF (%) vs. drought-affected values] were evaluated for effects on system performance. Early management (EM) included detecting drought by July 15 and weaning calves at 90 d. Normal management (NM) responded to drought by providing nutritional supplements as needed to maintain animal performance. A second bioeconomic computer model was used to simulate drylot performance for early-weaned calves. Treatments were evaluated based on their effects on ranch gross margins (RGM; gross revenue – variable costs). For EM, RGM was calculated with and without the drylot component. During drought, RGM was reduced compared with the base system: EM at 26 and 57% and NM at 33 and 72% for moderate and severe drought, respectively. For all levels of drought and forage quality, EM had equal or higher RGM than NM. Directly feeding EM calves was generally more efficient than feeding NM cows to produce milk to maintain calf performance.

Drought is an inherent trait of most rangelands and sound management necessitates managers address two fundamental questions when facing a drought situation. The first question is, ‘‘what is the probability that a useful amount of precipitation will be received over the period of concern?’’ and the second question is, ‘‘if it does rain, what will the impact be in terms of quantity and quality of herbage produced?’’ The objective of this study was to address the second question. Our hypothesis was that herbage growth response to above normal summer precipitation (i.e., 23 in July and August) would be limited in the northern Great Plains because of a general absence of productive warm-season species. Study plots were twelve 5 X 10-m non-weighing lysimeters. Treatments were: 1) simulated (i.e., rainout shelter imposed), severe spring drought (i.e., 1 May - 1 July) followed by ambient precipitation thereafter; 2) simulated, severe spring drought followed by ambient precipitation thereafter plus summer irrigation (i.e., July and August); 3) ambient precipitation only; and 4) ambient precipitation plus summer irrigation. Results indicated substantial herbage production can be expected in this region during summer when precipitation is well above average because of the positive growth response of blue grama (Bouteloua gracilis [H.B.K.] Lag. ex Griffiths), the dominant warm-season grass growing in this region. However, results also showed that level of production in the study situation (i.e., spring drought, wet summer) was only about 50% of that attained in a normal (i.e., wet spring/dry summer) year. Moreover, long-term weather data shows the probability of receiving 23 normal precipitation in both July and August (i.e., our irrigation treatments) is < 1%. Thus, although these rangelands possess the capacity to respond favorably to summer precipitation, the low probability of receiving substantial levels of summer precipitation ensures levels of ecological and economic risk remain high.

The impact of varied calving and weaning times on post-weaning production of steer calves was evaluated in a 3yr study. Steers (approximately 12 steers per calving-weaning treatment per year) born in late winter or early spring were weaned at 190 or 240 d of age, and steers born in late spring were weaned at 140 or 190 d of age after grazing with their dams on native range. Steers were pen-fed a growing diet until approximately 375 kg BW. They were then moved to an individual feeding facility and fed a higher-energy diet. Steers were allotted to harvest dates based upon visual estimates of degree of fat cover. Mixed model procedures were used to evaluate fixed effects of treatment and random effects of year and year × treat1This research was conducted under a cooperative agreement between USDA-ARS and the Montana Agric. Exp. Sta. USDA-ARS is an equal opportunity and affirmative action employer and all agency services are available without discrimination. Mention of a proprietary product does not constitute a guarantee or warranty of the product by USDA, Montana Agric. Exp. Sta., or the authors and does not imply its approval to the exclusion of other products that may also be suitable. 2To whom correspondence should be addressed: elaine@larrl.ars.usda.gov 3Retired. ment. Predefined contrasts were used to delineate treatment effects. Initial steer BW was affected by calving system and age at weaning. No treatment differences in ADG during the growing phase or ADG, feed intake, or feed efficiency during the finishing phase occurred. Steers averaged 537 ± 4 kg at harvest and age at harvest did not differ among treatments. Late winter steers had greater hot carcass weights, marbling scores, and quality grades than early spring steers. Calving later in the calendar year without altering weaning times produced younger calves at weaning with lighter BW at the start of the feeding period, which required greater number of days from weaning to harvest and produced carcasses with less marbling and lesser quality grades.

In environments where cool-season forages are the major forage resource, shifting calving season from winter to spring may match up forage nutrient supply with animal nutrient demands and thereby reduce purchased feed cost. However, calf BW at weaning in the fall may be less. The objective of this experiment was to determine the effect of calving season and age at weaning on stocker and finishing phase performance. Cows were assigned to late winter (LW), early spring (ES), or late spring (LS) calving seasons. Steers in the LW and ES groups were weaned at 190 and 240 d of age. Steers in the LS group were weaned at 140 and 190 d of age. Delaying the calving season reduced (P < 0.01) the age and BW of the calf at entry into the stocker phase and finishing phases. For the steers grown and developed in Oklahoma, overall stocker gain was similar (P > 0.10) among the 6 treatments groups. Steers developed in Montana had greater stocker ADG than steers grown in Oklahoma, and stocker ADG decreased (P < 0.01) as weaning age decreased. Steers born in LS produced carcasses that had less ( P < 0.05) marbling, fat thickness, quality grade, and yield grade than steers born in LW or ES, but had greater (P < 0.10) ADG. Finishing steers on pasture with ad libitum access to high-energy feed produced leaner carcasses, but pasture-finished steers were not as efficient in converting feed DM to gain as steers finished in dry lot. In a vertically integrated enterprise, delaying the calving season resulted in more of the BW gain being achieved in the finishing phase.

Drought is common in rangeland environments and an understanding of its impacts on the structure and function of rangeland ecosystems is paramount for developing effective management strategies. This research was the second of a series of studies investigating the impacts of varying seasonal droughts on northern Great Plains rangelands. Research was conducted on native rangeland during the 1998 through 2001 growing seasons. Study plots were twelve 5 3 10 m nonweighing lysimeters. An automated rainout shelter was used to establish drought conditions on 6 lysimeters during April, May, and June of 1998 and 1999. Single-day, flash grazing events were imposed at the beginning of May, June, and July. Grazing treatments were 1) graze during the 2 years of drought and the year after; 2) graze during the 2 years of drought and rest the year after; and 3) rest all years. Results showed that the intense spring drought reduced soil water content in the upper 30 cm of the soil profile and subsequently reduced total herbage production 20% to 40%; cool-season perennial grasses were the primary contributor to the reduction and cool-season annual grasses were secondary. Periodic grazing during drought had minimal impact on herbage production, whereas impacts on nondrought plots ranged from moderate enhancement to moderate suppression, with effects varying depending on functional group. Results also showed that substantial recovery occurred during the 1st postdrought year, with near full recovery realized within 2 years.

A 3-yr study evaluated late winter (Feb), early spring (Apr), and late spring (Jun) calving systems in conjunction with varied weaning strategies on beef cow and calf performance from Northern Great Plains rangelands. Crossbred cows were randomly assigned to one of three calving systems (on average n= 168.calving system(-1).yr(-1)) and one of two weaning times (Wean 1, 2) within each calving system. The Feb and Apr calves were weaned at 190 and 240 d of age, whereas Jun calves were weaned at 140 and 190 d of age. Breeding by natural service occurred in a 32-d period that included estrous synchronization. Cows were managed throughout the year as appropriate for their calving season. Quantity and quality of hay and supplements were provided based on forage and weather conditions, physiological state of the cows, and available harvested feed resources within a year. After weaning, two-thirds of the early weaned steers were fed in confinement in Montana, and one-third were shipped to Oklahoma and were grazed or fed forage. One-half of the early weaned heifers grazed seeded pastures, and the other half was fed in confinement. Early weaned calves were weighed on approximately the same day as late-weaned calves. Birth weight and overall rate of gain from birth to weaning did not differ for calves from the three calving systems. Calf weaning weight differed by weaning age within calving system (P = 0.001), and calves from the Jun calving system that were weaned at 190 d of age tended (P = 0.06) to be lighter than calves of the same age from the Feb or Apr calving systems. Cow BW change and BCS dynamics were affected by calving system, but the proportion of cows pregnant in the fall was not. Cows suckled until later dates gained less or lost more BW during the 50 d between the first and second weaning than dry cows during this period. The previous year's weaning assignment did not affect production in the following year. Estimated harvested feed inputs were less for the Jun cows than for the Feb and Apr cows. We conclude that season of calving and weaning age affect outputs from rangeland-based beef cattle operations.

Data from a 3-yr study in Montana were utilized to evaluate impacts of season of calving, weaning strategy, and retained ownership of steer calves on enterprise profitability. Calving seasons were late winter (LW), early spring (ES), or late spring (LS). Each season had 2 weaning times: 190 (LW190, ES190) or 240 (LW240, ES240) d for LW and ES, and 140 (LS140) or 190 (LS190) d for LS. Backgrounding options included shipping steers to Oklahoma (OK1), or backgrounding in Montana to a constant age (MT2) or weight (MT3). Steers from OK1 and MT2 were finished in Oklahoma in confinement or via self-feeders on pasture and harvested in Texas. Steers in MT3 were finished in Montana in confinement and harvested in Colorado. Performance of each system was modeled based on actual animal performance, market prices, and variable input costs. When calves were sold at weaning, gross margins per cow were greatest for LS190 (P < 0.05) and lowest for LW240. During backgrounding, costs of gain were similar among cow-calf systems, and gross margins per steer were greatest for LS140 (P < 0.05), but not different among backgrounding systems. During finishing, costs of gain were greatest for steers from MT2 due to transportation costs to Oklahoma (P < 0.05), and gross margin per steer favored MT3 (P < 0.05). Gross margin for a ranch with a fixed land base did not differ among systems if calves were sold at weaning, but was greatest for LS systems after backgrounding or finishing (P < 0.05).

The objective of this paper is to examine the sustainability of rangeland agriculture (i.e., managed grazing) on a world-wide basis, with a focus on North America. Sustainability is addressed on three fronts: 1) ecological, 2) economic, and 3) social acceptance. Based on previous and on-going research, we suggest that employment of science-based rangeland grazing management strategies and tactics can ensure ecological sustainability. The formidable challenge in employing such technology centers around the need to balance efficiency of solar energy capture and subsequent harvest efficiencies across an array of highly spatially and temporally variable vegetation growing conditions using animals that graze selectively. Failure to meet this fundamental challenge often accelerates rangeland desertification processes, and in some instances, enhances rate and extent of the invasion of noxious weeds. We also suggest that the fundamental reason that ecologically sound grazing management technologies are often not employed in the management of grazed ecological systems is because social values drive management decisions more so than ecological science issues. This is true in both well-developed societies with substantial economic resources and in less-developed societies with few economic resources. However, the social issues driving management are often entirely different, ranging from multiple-use issues in developed countries to human day-to-day survival issues in poorly developed countries. We conclude that the long-term sustainability of rangeland agriculture in 1) developed societies depends on the ability of rangeland agriculturalists to continually respond in a dynamic, positive, proactive manner to ever-changing social values and 2) less-developed societies on their ability to address the ecological and social consequences arising from unsustainable human populations before the adoption of science-based sustainable rangeland management technologies.

The impact of varied calving and weaning times on post-weaning production of steer calves from the Northern Great Plains was evaluated in a 3-yr study. Steers (n = 215) born in one of three calving seasons (late winter (LW), early spring (ES), or late spring (LS)) were weaned at 4 (LS1), 6 (LW1, ES1, LS2), or 8 (LW2, ES2) mo of age after grazing with their dams on native range. Later weaned cow- calf pairs continued to graze native range until weaning. Steers were pen-fed a corn silage and alfalfa hay-based diet until the weaning group averaged 375 kg. They were then moved to an individual feeding facility and fed a higher energy diet. Steers were individually allotted to harvest dates based upon visual estimates of fat thickness. Data were analyzed as a completely random design with fat thickness as a covariate using mixed model procedures. Year and year by treatment were random effects. Non- orthogonal estimates were used to delineate treatment effects. Initial steer weights averaged 216 ± 12 kg but were affected by calving season and age at weaning, with LS2 steers weighing 24 ± 10 kg less (P < 0.05) than the average of the LW1 and ES1, LW and ES steers weaned at 6 mo averaging 26 ± 8 kg less (P < 0.01) than those weaned at 8 mo of age, and LS1 weighing 26 ± 11 kg less (P < 0.05) than LS2. There were no treatment differences in ADG during the growing or finishing phases. Total days to harvest averaged 311 ± 15 d and differed between LW and ES steers weaned at 6 versus 8 mo of age due to a 37 ± 12 d difference (P < 0.01) in time to reach harvest. Total days to harvest did not differ between LS steers weaned at 4 versus 6 mo of age. Steers averaged 527 ± 12 kg at harvest and weights were 23 ± 10 kg less (P < 0.01) for ES than LW. Differences in production of steers among calving and weaning strategies may be related to differences in harvest weights and time on feed as affected by weaning weights.

Three studies were conducted to evaluate late summer protein supplementation for growing steers on Northern Great Plains rangeland. In Experiment 1, crossbred yearling steers (N = 80 per year, mean initial live-weight = 275 kg) were allotted to 1 of 2 treatments replicated in 3 pastures in each of 3 years. Treatments were summer-long grazing with or without protein supplementation in late summer. Protein supplement (26% crude protein) was fed at a rate of 1.68 kg (dry matter basis) every third day. In 1995, a third treatment was added to additional pastures consisting of 1.62 kg (dry matter basis) of a 40% crude protein supplement fed every third day. There was no weight gain response to protein supplementation. In Experiment 2, yearling steers grazing rangeland from May to September were fed either no supplement, 1.5 kg of a 22% crude protein safflower meal-based supplement, 1.2 kg of 26% soybean meal-based supplement or 1.2 kg of a 26% safflower and soybean meal-based supplement every third day in late summer. Live-weight gain, forage intake, and digestibility were not affected by supplementation. A third experiment using ruminally cannulated steers fed grass hay and the 3 protein supplements based on safflower and soybean meals showed an increase in ruminal ammonia concentrations but no other appreciable change in ruminal fermentation with protein supplementation. Supplementation with as much as 648 grams of protein every third day was not a viable means to increase gains of steers grazing Northern Great Plains rangelands during late summer under the conditions of this experiment.

A 3-year study was conducted to evaluate grazing strategies for production of growing cattle during summer on Northern Great Plains rangeland. Crossbred yearling steers (N = 123 per year, avg initial weight = 275 kg) were allotted to 1 of 2 treatments replicated in 3 pastures. Treatments were season-long grazing of pastures at recommended stocking rates assuming a 4-month grazing period or intensive-early grazing of pastures stocked at the same rate assuming only a 2-month grazing season. Precipitation in 1993 was 169% of normal resulting in greater forage quality than in other years and no differences were observed in weight gains between treatments during 1993. In 1994 and 1995, steers in the intensive-early stocked pastures gained less weight during the 2 months of grazing than did those in the season-long stocked pastures; however, gain per hectare was greater in the intensive-early stocked pastures. Intensive-early stocking with growing steers may be a viable means to overcome limited forage quality during late summer in the Northern Great Plains and to maximize forage utilization in years of abundant forage.

Understanding the interacting effects of drought and grazing on runoff, erosion, and nutrient transport is essential for improved rangeland management. Research was conducted at the Fort Keogh Livestock and Range Research Laboratory located near Miles City, Mont. using 12, non-weighing lysimeters for 3 years. During years 1 and 3, no drought treatment was imposed. For year 2, one half of the lysimeters were covered to implement a drought treatment. The 3 grazing treatments were ungrazed, grazed during but not after drought, and grazed during and after drought. Runoff, sediment yield, and an array of nutrients in the runoff water were measured from the lysimeters. First year base line data with no grazing or drought treatments applied indicated no significant differences among lysimeters. Below normal precipitation occurred during year 2, resulting in no runoff from the drought treatment and negated the "non-drought" control. This prevented a direct assessment of the interaction among the drought and grazing treatments for this year. The drought treatment did produce significant reductions in water, sediment, and nutrient yield. No grazing impact was observed during year 2. The third year with more normal precipitation, there was a trend toward increased runoff, sediment, and nutrient yield from the second year drought treatment lysimeters. In the third year, both grazing treatments showed significantly greater runoff, sediment, and nutrient yield than the ungrazed treatment. Runoff and sediment yield tended to increase from the combination of drought and grazing treatments. The observed increases in runoff and sediment and reduced water quality from the drought and grazing treatments were measured against controls and when compared to the natural variability and water quality standards, they were concluded to be minimal.

Integrating use of seedings of perennial cool-season grasses with native range is used to increase available forage and for maintaining a high plane of nutrition for grazing livestock. Our goal was to evaluate performance of yearling cattle and stand persistence of 3 released wheatgrass cultivars. Twice replicated 3-ha pastures were seeded to 'Rosana' western wheatgrass [Pascopyron smithii (Rydb.) A. Love], 'Luna' pubescent wheatgrass [Elytrigia intermedia (Host) Nevski], and 'Hycrest' crested wheatgrass {[Agropyron cristatum (L.) Gaertn.] ssp. desertorum (Fisch. ex Link) A. Love} in autumn 1994. Yearling steers (n = 8) grazed from 9 May to 12 June 1997 and 24 April to 15 June 1998. Yearling heifers grazed from 27 April to 18 June 1999. Hycrest produced the largest herbage standing crop in spring 1997 (912 kg ha(-1)) and 1998 (1,223 kg ha(-1)) (P < 0.05), but by spring 1999 standing crops averaged 656 kg ha(-1), and cultivars were not significantly different. Digestible organic matter standing crop did not differ among pastures of seeded species, but declined (P < 0.05) from May to June each year. Crude protein standing crop varied among cultivars (P < 0.05) in April and May 1998 and May 1999; however, no clear trends emerged. Crude protein standing crops consistently declined from April-May to June. Average daily gains were similar among cultivars in 1997, but greater (P < 0.05) on Hycrest (1.28 kg day(-1)) than Rosana (1.03 kg day(-1)) in 1998. Gains on Hycrest (0.74 kg day(-1)) and Rosana (0.78 kg day(-1)) were greater (P < 0.05) than on Luna (0.52 kg day(-1)) in 1999. These findings show in some years, Hycrest provided more forage in spring than Rosana, and will allow an increase in livestock numbers. Based on the encroachment of invading species, persistence of Luna is marginal of the 335-mm precipitation zone in the Northern Great Plains.

An understanding of the impacts of grazing during and following drought on rangeland ecosystems is critical for developing effective drought management strategies. This study was designed to examine the effects of drought and grazing on blue grama [Bouteloua gracilis (H.B.K) Lag. ex Griffiths] and western wheatgrass [Pascopyrum smithii Rydb. (Love)] tiller growth dynamics. Research was conducted from 1993 to 1996 at the Fort Keogh Livestock and Range Research Laboratory located near Miles City, Mont. An automated rainout shelter was used during 1994 to impose a severe late spring to early fall (May to October) drought on 6 of twelve, 5- x 10-m non-weighing lysimeters. Twice replicated grazing treatments were: 1) grazed both the year of (1994) and the year after (1995) drought; 2) grazed the year of and rested the year after drought; and 3) no grazing either year. Drought had minimal impact on tiller relative growth rates of plants grazed twice, although it reduced (P less than or equal to 0.01) rates of axillary tiller emergence for blue grama (79%) and western wheatgrass (91%). Defoliation periodically increased relative growth rates (P less than or equal to 0.05) and tiller emergence (P less than or equal to 0.01) of both species. Neither drought nor grazing affected tiller densities or tiller replacement rates of either species nor did they affect productivity of blue grama. Drought, however, reduced (P less than or equal to 0.01) productivity of western wheatgrass 50% in 1994 whereas grazing reduced productivity (P less than or equal to 0.01) by 46% in 1994 and 69% in 1995. Moderate stocking levels (40-50% utilization) during and after drought did not adversely affect the sustainability of these dominant native grasses.

Presence of annual bromes (Bromus spp.), introduced annual weedy grasses, can alter seasonal patterns of forage production and quality and require management changes for efficient use of infested rangelands. We determined biological impacts of the presence of brome by comparing livestock performance on brome infested rangeland to similar sites on which brome had been suppressed by autumn application of atrazine [6-chloro-N-ethyl-N'-(1-methylethyl)-1,3,5-triazine-2,4-diamine] at 0.56 kg ha(-1) in 1992 and 1993. Each treatment was randomly assigned to three, 12-ha pastures. Vegetation was measured for 5 months (May to September) each year from 1993 to 1995. Each pasture was stocked with 8 crossbred steers of British breed origin (Bos taurus) from mid-May to mid-September 1993 and 1995 and to mid-August 1994. Initial body weights averaged 329 kg SD = 31 in 1993, 273 kg SD = 14 in 1994, and 272 kg SD = 21 in 1995. Brome suppression and environment influenced plant species in diets, diet quality, and livestock performance. Brome suppression reduced percentage of annual grasses in diets from 14% to 10%. Annual grasses were replaced in the diet by a variety of forb and grass species {western wheatgrass [Pascopyrum smithii Rydb. (Love)], and blue grama [Bouteloua gracilis [H.B.K.] Lag. ex Griffiths]}, with specific replacement depending on year and month. Steer gains were increased from 0.92 to 1.04 +/- 0.02 kg head(-1) day(-1) (P < 0.02) and from 69 to 81 +/- 2.8 kg ha(-1) (P < 0.05) with brome suppression. This experiment demonstrated that improvement in livestock performance can be expected with the suppression of annual bromes on semiarid rangelands.

The long-term sustainability of modern agriculture is examined in an ecological context. As an aid to defining agriculture, animal agriculture, and sustainable agriculture, a broad overview of the structural and functional aspects of ecosystems is presented. The longterm sustainability of two corn production systems and four beef cattle production systems is then examined relative to energy output/cultural energy input ratios. Results indicate that as corn yields increase, ecological efficiencies decrease dramatically. But analyses of the four beef cattle production systems show an even more startling effect in that cultural energy inputs (i.e. fossil fuels) far exceed energy outputs. This low level of efficiency was found to be largely the result of the interaction effects of the high levels of cultural energy required to maintain a productive cow herd and grow and finish calves in the rather harsh environment of the Northern Great Plains. Results pointedly reveal the high level of dependency of the U.S. beef cattle industry on fossil fuels. These findings in turn bring into question the ecological and economic risks associated with the current technology driving North American agriculture.

Drought and herbivory frequently influence North American rangelands. While these influences may temporarily reduce vegetative cover, their mutual influence on the available seedbanks which might occupy new safe sites is unclear. We examine effects of drought and grazing upon pre- and post-drought plant root distribution and germinable seed bank to determine 1) if the response of root distributions to drought depends upon grazing use and 2) if the presence of germinable seeds is altered significantly by drought and grazing. Using twelve, 5 X 10 m nonweighing lysimeters with an automated rainout shelter, we documented root intercepts in situ using a minirhizotron from 1993-1996, Seed bank samples were incubated in a greenhouse to determine seedling emergence. Roots were fewer in shallow soil layers in grazed plots than ungrazed plots by the end of the study, irrespective of drought, Roots in deeper (Bw horizon) soil layers were fewer during drought, but were not influenced by grazing. Seed bank composition results suggest that perennial grasses were a small portion of the seed bank. Cool-season annual grass seeds accumulated after drought. Without drought, forb seed banks increased with grazing, Thus while shallow roots may decrease during drought, in the year following drought grazing may decrease aboveground net primary production, and allow Large accumulations of cool-season annual grass seed in a northern mixed grass prairie.

The effects of animal age and sex on chemical and botanical composition of diets of cattle grazing native rangelands were evaluated in a 2-year study. Samples were collected monthly from June through October using esophageally cannulated suckling calves, yearling heifers, mature cows, and mature steers. Dietary crude protein and digestibility differed among animal classes, but these differences varied over time. These 2 diet quality indicators did not vary in the same manner over time for all animal classes, Dietary crude protein varied from a low of 7.2% for steers in August 1994 to a high of 14.3% for heifers in June 1993, In vitro digestibility varied from a low of 50.7% for cows in October 1993 to a high of 74.3% for calves in June 1993, Botanical composition of diets varied with animal class and sampling date with interactions among these. Cool-season grasses accounted for an average of 70% of the diet with a range of 33 to 90%, Shrubs varied from 1 to 61% of the diet. Differences in chemical composition among age and sex classes of cattle grazing native rangeland during the growing season may be partially related to differences in botanical composition of diets. Animals used to obtain diet samples should, therefore, be of similar physiological state and age as animals being monitored for performance.