Invasive annual weeds negatively impact ecosystem services and pose a major conservation threat on semiarid rangelands throughout the western United States. Rehabilitation of these rangelands is challenging due to interannual climate and subseasonal weather variability that impacts seed germination, seedling survival and establishment, annual weed dynamics, wildfire frequency, and soil stability. Rehabilitation and restoration outcomes could be improved by adopting a weather-centric approach that uses the full spectrum of available site-specific weather information from historical observations, seasonal climate forecasts, and climate-change projections. Climate data can be used retrospectively to interpret success or failure of past seedings by describing seasonal and longer-term patterns of environmental variability subsequent to planting. A more detailed evaluation of weather impacts on site conditions may yield more flexible adaptive-management strategies for rangeland restoration and rehabilitation, as well as provide estimates of transition probabilities between desirable and undesirable vegetation states. Skillful seasonal climate forecasts could greatly improve the cost efficiency of management treatments by limiting revegetation activities to time periods where forecasts suggest higher probabilities of successful seedling establishment. Climate-change projections are key to the application of current environmental models for development of mitigation and adaptation strategies and for management practices that require amultidecadal planning horizon. Adoption of new weather technology will require collaboration between land managers and revegetation specialists and modifications to the way we currently plan and conduct rangeland rehabilitation and restoration in the Intermountain West.

Low seed germination and seedling establishment are the greatest challenges for revegetation success. Topographic microsites are known to enhance seed germination and seedling establishment due to their unique soil properties and provision of shelter from elements and herbivores; soil amendments can supply organic matter and nutrients for plant establishment and growth when limited. We investigated the effect of three topographic microsites and six soil amendments and their additive effects on three disturbed grasslands in central and southern Alberta, Canada. Treatments were topographic microsites of mounds, pits, and flats, with and without amendments (erosion control blanket, hay, straw, manure, hydrogel, control) and were seeded with four native grasses and three native forb species. Seedling emergence and survival and soil temperature and water content were assessed over two seasons and plant cover over three seasons. The effect of microsites and amendments was not additive. The addition of erosion control blanket, hay, and straw to flat sites was just as productive as on topographic microsites. These amendments increased grass and forb emergence and buffered soil temperature. Mounds increased first year forb emergence and reduced over winter survival rates for grasses and forbs. Pits were not beneficial for revegetation. The effect of topographic microsites and amendments was influenced by site conditions.

Big sagebrush (Artemisia tridentata Nutt.) restoration is needed across vast areas, especially after large wildfires, to restore important ecosystemservices. Sagebrush restoration success is inconsistent, with a high rate of seeding failures, particularly at lower elevations. Seed enhancement technologies may overcome limitations to restoration success. Seed pillows are one such technology designed to improve seed-soil contact in broadcast seedings by providing a favorable medium for seedling establishment and growth. Seed pillows have shown promising results in greenhouse studies; however, they have not been evaluated in the field. We compared broadcast-seeding seed pillows with broadcast-seeding bare seed in 2 yr across a large, burned elevation gradient. Compared with bare seed, we found no evidence that seed pillows improved sagebrush establishment and growth across the elevation gradient. Though our results suggest that seed pillows do not increase the likelihood of successful sagebrush restoration, they were successful at times when bare seeds were not, and the same was true for bare seeds. At least one of the two treatments was successful at 50% of the elevations over the 2 seeding yr. This suggests that a bet hedging approach, seeding both bare seed and seed pillows, may increase the probability of success. Further supporting the use of bet hedging, if both methods were used and seeding occurred in both years, success would have been 86%. Sagebrush density and cover varied by elevation. In the first-yr seeding, sagebrush density and cover generally increased with increasing elevation. In the second-yr seeding, sagebrush density and cover were greatest at the lowest and highest elevations. We speculate that at the lower elevations an unusually wet spring combined with limited herbaceous vegetation provided an ideal environment for sagebrush establishment and growth. Our results also demonstrate, counter to common assumptions, that lower elevations sagebrush seedings can be successful.

Our ability to restore rangelands is limited, and it is unknown if seedling herbivory on its own, or in interaction with other stressors, is a major contributor to restoration failure. To address this, we conducted two experiments: a No Defoliation (ND) experiment (n=48), in which seedlings from three perennial grasses (crested wheatgrass [Agropyron cristatum {(L.} Gaertn.], bluebunch wheatgrass [Psuedoroegnaria spicata {Pursh} Á. Love], Sandberg bluegrass [Poa secunda J Presl]) were subjected to wet and dry water regimes for 4 mo, and a concurrent Defoliation (D) experiment (n = 95), in which seedlings were factorially assigned to two defoliation treatments—frequency (LOW, HIGH) and intensity (30% vegetation removal, 70% vegetation removal). Indicators of seedling performance were aboveground and belowground biomass (AGB and BGB), root:shoot ratio, tillering, and mortality. The effect size statistic, Hedge's g, allowed for comparisons between performance measures. Water stress induced reductions inmost performance measures: BGB (g=ND: -1.3; D: -1.6), root:shoot ratio (g=ND: n.s.; D: -0.2), and tillering (g= ND: -1.7; D: -1.2), though not significantly for all species. For ABG, water stress interacted with defoliation, reducing performance less at an intensity of 70% (g = -2.0) as opposed to 30% (g = -3.0), but not always significantly in the former. Water stress also caused less reduction in AGB when no defoliation occurred (ND: -0.8; g=D: -2.5). Intensity and frequency of defoliation interacted; seedlings were generally resistant to reductions in performance except at high frequency, 70% defoliation. Agropyron cristatum and P. spicata displayed similar sensitivity to treatments, mostly in terms of changes in AGB and BGB, while P. secunda also experienced increased mortality and reduced tillering. If these differences in sensitivity result in differential survival, herbivory could impact species postrestoration population demographics.

Wildlife-associated recreation and biodiversity are important management considerations on public and private rangelands, making it imperative that rangeland professionals explicitly take wildlife conservation into account in vegetation management planning and implementation. Here, we synthesize the literature reporting effects of brush management on wildlife and make recommendations for applying brush management to accomplish wildlife conservation objectives. Key observations arising from our synthesis are that habitat-related terminology is often misused in brush management literature. Recommending brush management as a “wildlife habitat improvement” tool is a non sequitur because habitat is species specific and brush management has different consequences for different species of wildlife and plants. Communication between resource managers and stakeholders can be improved by making it clear that habitat is species specific and then identifying what constitutes a benefit of brush management. Changes in resources resulting from brush management may not benefit targeted wildlife species unless these changes overcome some limiting factor or factors. Wildlife responses to brush management treatments are too complex to make broad generalizations because they are mediated by environmental factors and depend on the plant community, size and configuration of the area manipulated, type of treatment applied, and time since application. Prescriptions aimed at improving habitat for wildlife generalists may have relatively modest positive effects on that group but have potentially detrimental effects on specialists. Given this potential trade-off, an idea to consider is that it may be best to err on the side of using brush management as a tool to manage habitat for specialists. Brush management plans and recommendations should take into account trade-offs such as benefiting grassland wildlife at the expense of woodland species. Taking a broader “systems” perspective that balances needs of wildlife in conjunction with other ecosystem services affected by woody plant encroachment and brush management should be a goal of natural resource managers.

The fire-grazing interaction is well studied in mesic grasslands worldwide, but research is limited in semiarid systems. We examined the principal drivers and feedbacks of the fire-grazing interaction on the strength of cattle grazing selection, herbaceous biomass, crude protein, and vegetation structure and composition in two pastures in the Northern Great Plains. Cattle showed significant preference, use, and grazing utilization in recently burned patches that declined as time since fire increased. Cattle selection was driven by significantly increased crude protein in recent burns. Grazing utilization of 70% in patches with < 1 yr after fire established low herbaceous biomass, but the extent to which it was maintained varied with precipitation. Herbaceous biomass increased to nonburned levels 2 yr after fire, and crude protein decreased to nonburned levels 120 d after fire. Species composition was influenced primarily by site and year, though bare ground and litter were influenced by the fire-grazing interaction. Our data indicate that mixed-grass prairies of the Northern Great Plains are resilient to the fire-grazing interaction and that rest from grazing following fire is likely ecologically unnecessary. The use of the fire-grazing interaction is an alternative management strategy suitable for the Northern Great Plains, effectively increasing heterogeneity of grassland habitat.

In the Nebraska Sandhills, one of the largest contiguous grassland ecoregions remaining in North America, sandy textured soils are stabilized by fine root biomass from predominantly warm-season grasses. Concerns over destabilization have led to management that aims to avoid an undesirable state change toward mobile sand dunes. In 2012, the Sandhills experienced extreme drought conditions that coincided with the worst wildfire year on state record. According to state-and-transition models and ecosystem managers, the combination of wildfire and drought conditions should cause a state transition due to a lack of recovery of grassland vegetation and a loss of sand dune stability. To test this hypothesis, we implemented a time-since-fire study to track biomass recovery of Sandhills grassland vegetation following a wildfire on The Nature Conservancy's Niobrara Valley Preserve in burned and unburned areas. Two yr following the wildfire, aboveground herbaceous biomass in burned areas had recovered to levels that did not differ from unburned areas, maintaining the stability of the sand dunes. This provides evidence that counters current land management frameworks that portray Sandhills grassland as highly vulnerable to destabilization when wildfires occur during severe drought conditions.

Defoliation intensity and timing are two important factors determining plants response to grazing. These factors can be managed by adjusting stocking rate and applying a grazing strategy. In a 6-yr clipping experiment conducted in northwestern Argentina, we assessed the effect of different defoliation intensities (∼30%, ∼50%, and ∼70% removal of the annually produced aboveground biomass) and simulated grazing strategies (continuous grazing, two-paddock rest-rotation, three-paddock rest-rotation, dormant season grazing) on plots of three C4 native bunchgrasses (Pappophorum vaginatum, Trichloris crinita, and Digitaria californica). Response variables were mean and trend of clipped-off biomass during the 6 yr of treatments, number of inflorescences, and aboveground biomass produced on the year following treatments end (to evaluate residual effect of treatments). Results were species dependent. Mean clipped-off biomass increased with defoliation intensity in T. crinita and D. californica. However, defoliation intensity negatively affected clipped-off biomass trend in T. crinita and the production of P. vaginatum and T. crinita during “residual effect” evaluation. The three species responded positively at least in one response variable to the amount of rest periods in the grazing strategy. Our results are not fully consistent with the concept that forage production is more influenced by defoliation intensity than by grazing strategy: In two of the three species, grazing strategy presented greater impact on response variables than defoliation intensity. When significant “defoliation intensity × grazing strategy” was detected, intensity tended to be more detrimental as grazing strategy allows fewer rest periods. We observed a residual effect of treatments in the three species (generally, negative effect of defoliation intensity and positive effect of grazing strategies with more rest periods). Our results show that dormant season utilization and rest periods are beneficial for maximizing mean clipped-off biomass and ensuring clipped-off biomass trend. High defoliation intensities can maximize short-term clipped-off biomass, but it may produce negative residual effects and trends.

Fire kills Wyoming big sagebrush (Artemisia tridentata Nutt. ssp. wyomingensis Beetle & Young) and promotes cheatgrass (Bromus tectorum L.), a highly flammable and invasive annual in sagebrush communities with compromised resistance. To focus management on resistance and resilience of Wyoming big sagebrush communities with varying species composition, we studied 51 paired sites with burned and unburned areas. We quantified soil surface and foliar cover in 12 cover groups. Comparisons identified vegetation or soil surface factors that significantly (p ≤ 0.05) correlated (Spearman's rank correlation coefficient = ρ) to burned area community composition. Cheatgrass cover in burned areas was greater where unburned areas had more cheatgrass cover (ρ = 0.75), litter cover (ρ = 0.31), and sagebrush plant canopy volume (ρ = 0.40), and less bare soil (ρ=-0.39) and cryptogamcover (ρ=-0.32). Cheatgrass cover in burned areas was not significantly correlated with unburned area perennial grass or forb cover. Burned area perennial grass cover appeared to be related to more perennial grass (ρ = 0.77) and native forb cover (ρ = 0.30), but less cheatgrass cover (ρ = -0.39) in unburned areas. Burned area native herbaceous dominance (native minus exotic herbaceous foliar cover) correlated with less cheatgrass cover (ρ = -0.65) and sagebrush canopy volume (ρ = -0.34) in unburned areas and with more perennial grass (ρ = 0.30) and sagebrush relative cover (ρ = 0.39) in adjacent unburned areas. Postfire site dominance could be of either native or exotic plants where cheatgrass cover on adjacent unburned sites was < about 15%. Native species however, never dominated or increased in dominance where cheatgrass was above 15%. Results suggest that cheatgrass cover before a fire played a strong role in determining postfire plant communities; this suggests management should focus on prefire and postfire management of cheatgrass and litter. Perhaps prescriptions and priorities should bemore nuanced on the basis of driving variables of postfire response hypothesized to be cheatgrass, perennial grass, and shrub abundance.

Shrub recruitment in arid and semiarid regions often occurs in pulses controlled by specific weather events. Previous research suggested that Wyoming sagebrush in Wyoming is no exception. We examined four species/ subspecies of sagebrush in Nevada, in 2009 and 2010, to discover if evidence of recruitment pulses was contained in the annual growth-ring records. Sagebrush species and subspecies occur on a wide variety of ecological sites that require different management strategies. Species included black sagebrush (Artemisia nova A. Nelson), Wyoming big sagebrush (Artemisia tridentata subsp. wyomingensis Beetle & Young), Lahontan sagebrush (Artemisia arbuscula subsp. longicaulis Winward & McArthur), and low sagebrush (Artemisia arbuscula Nutt. ssp. arbuscula). Eighty stem sections were collected from each of 24 stands (6 stands per species or subspecies) at different geographic locations along east-west or north-south gradients where each species or subspecies naturally occurred. Annual growth-ring analysis was used to determine the year of establishment and the relationship between recruitment and weather events. Results indicated stand ages and locations were different (P > 0.001) among species and subspecies, and years of recruitment were strongly correlated with local and hemispheric weather patterns. Linear and multiple regressions modeled recruitment pulses for all four species. Weather-based predictor variables indicated complex interactions between recruitment and climatic controls. Pacific Decadal Oscillation (PDO) index variables were prominent predictors for all four species at their associated sites. Other important local weather variables included total annual precipitation the year before recruitment, the year of recruitment, and the year following recruitment. In Nevada and the Great Basin, it is imperative that successful sagebrush seeding technologies are discovered and implemented. Ecological restoration and postfire rehabilitation methods should be timed correctly with respect to precipitation patterns (positive phase PDO) and/ or designed to mimic conditions responsible for natural sagebrush recruitment.

Ecological restoration is predicated on our abilities to discern plant taxa. Taxonomic identification is a first step in ensuring that plants are appropriately adapted to the site. An example of the need to identify taxonomic differences comes from big sagebrush (Artemisia tridentata). This species is composed of three predominant subspecies occupying distinct environmental niches, but overlap and hybridization are common in ecotones. Restoration of A. tridentata largely occurs using wildland collected seed, but there is uncertainty in the identification of subspecies or mix of subspecies fromseed collections. Laboratory techniques that can determine subspecies composition would be desirable to ensure that subspeciesmatch the restoration site environment. In this study, we use spectrophotometry to quantify chemical differences in the water-soluble compound, coumarin. Ultraviolet (UV) absorbance of A. tridentata subsp. vaseyana showed distinct differences among A.t. tridentata and wyomingensis. No UV absorbance differences were detected between A.t. tridentata and wyomingensis. Analyses of samples from > 600 plants growing in two common gardens showed that UV absorbance was unaffected by environment. Moreover, plant tissues (leaves and seed chaff) explained only a small amount of the variance. UV fluorescence of water-eluted plant tissue has been used for many years to indicate A.t. vaseyana; however, interpretation has been subjective. Use of spectrophotometry to acquire UV absorbance provides empirical results that can be used in seed testing laboratories using the seed chaff present with the seed to certify A. tridentata subspecies composition. On the basis of ourmethods, UV absorbance values < 2.7 would indicate A.t. vaseyana and values > 3.1 would indicate either A.t. tridentata or wyomingensis. UV absorbance values between 2.7 and 3.1 would indicate a mixture of A.t. vaseyana and the other two subspecies.

Future climate projections indicate temperatures in the shortgrass steppe region are likely to increase up to 3°C by midcentury, with a corresponding reduction in soil moisture even without precipitation deficit. Although periodic drought is a natural disturbance in shortgrass rangeland, negative effects on characteristic shortgrass species are possible as the frequency and severity of drought events increase in comparison with recent historic norms. As part of a study intended to detect vegetation changes at a shortgrass steppe site on Colorado's eastern plains, frequency and canopy cover percentage were measured in 37 permanently marked plots over a period of 17 yr. The study period included the two lowest total annual precipitation yr (2002 and 2012) in the period of record for regional weather stations, exceeding even the driest years of the extended 1930s drought. Growing season mean temperatures during those drought years were 1°C and 1.6°C above the 1971–2000 average, respectively. Three of the six perennial grass species monitored showed a decline over the period of the study. Blue grama (Bouteloua gracilis), a dominant and important forage species in the shortgrass steppe, declined in both cover and frequency, while alkali sacaton (Sporobolus airoides), not an important forage species, slightly increased. In addition to changes in graminoid dominance, we observed an increase in cholla (Cylindropuntia imbricata) and a decrease in sandsage (Artemisia filifolia) densities between 1999 and 2015. Even if total productivity of the shortgrass steppe ismaintained under warming and drying conditions, changes in species composition have implications for rangeland quality with regard to its use for livestock grazing and use bywildlife such as small mammals and songbirds.

Shrubs are important plant species in grassland ecosystems worldwide, and their density and cover have been gradually increasing globally. However, the interaction effect of grazing and aridity on population recruitment and population growth of shrub species in grasslands has not been examined explicitly. We examined sapling establishment, sexual recruitment, population mortality, and population growth of Caragana stenophylla along a climatic aridity gradient and a grazing intensity gradient in the Inner Mongolia Steppe, using manipulative field experiments. Sapling establishment, sapling height, and sexual recruitment of C. stenophylla decreasedas climatic aridity and grazing intensity increased. The negative effects of grazing on sapling establishment and sexual recruitment gradually increased as climatic aridity increased. The effect of climatic aridity and grazing on population mortality was influenced by sexual recruitment. In the combined treatments of climatic aridity and grazing, population mortality was relatively high when sexual recruitment was relatively high, while population mortality increased as climatic aridity and grazing increased when sexual recruitment was relatively low. C. stenophylla population increased under relatively low drought stress and mild grazing but declined under strong drought stress and/or severe grazing. Our results suggested that to maintain viable Caragana populations, appropriate grazing policies must be made according to climate aridity gradient.

A better understanding of site-specific factors such as soil salinity that regulate plant invasions is needed. We conducted a 3-mo greenhouse study to evaluate the salinity responses of three local maternal sources of Lepidium alyssoides, which is an indigenous species shown to aggressively colonize disturbed shrubland sites in the southwestern United States, including those affected by high salinity and sodicity. Results indicated that there were little or no population effects on plant evapotranspiration (ET), growth, and tissue Na and Cl concentrations. Significant reductions in seedling growth and ET were largely independent of various isosmotic saline irrigation solutions that included NaCl, Na₂SO₄, and CaCl₂, each at-0.1MPa and-0.2MPa, suggesting that ET and growth were controlled by solution osmotic potential. The combined Na and Cl concentrations in leaves were 9–10% of dry weight with no visible sign of injury. However, increasing leaf mortality and abscission as a proportion of total leaf production was observed in the high-salt treatments (-0.2 MPa), with a combined Na and Cl concentration reaching 16% with high NaCl. Under saline conditions, considerable foliage salt loads of this species could deposit high-salt litter to potentially alter a landscape to its own favor and to the detriment of other salt-sensitive species. Results of this study add to a limited quantitative database on site-specific salinity factors governing plant invasions by showing the potential for these populations to behave invasively under saline conditions and, thus, potential for soil salinity assessment to predict incipient populations. However, due to its halophytic traits and indigenous status, L. alyssoides may alternatively provide ecosystem services to salinized shrublands of the arid and semiarid southwestern United States.

Salt cedar (Tamarix spp.) readily invades and dominates riparian areas and lake basins throughout the western United States. Traditional control efforts (chemical and mechanical control) are expensive and provide limited long-term control. The salt cedar leaf beetle (Diorhabda elongata) provides a method of biological control through reduction in cover. However, population establishment of leaf beetles in some locations is often difficult because of environmental conditions. In previous research, goats readily consumed salt cedar, offering an alternative method of reducing salt cedar cover. For this study, we determined if sheep would consume salt cedar and consume a similar amount as goats. Twelve Rambouillet and 12 Suffolks lambs were fed salt cedar once daily (Trial 1) and three times daily (Trial 2). Intake of salt cedar by sheep was compared between breeds and with intake of salt cedar by goats (n = 10). Salt cedar was fed once a day in Trial 1 for 30 min over 15 d. Intake was recorded daily for individual animals. In Trial 2, salt cedar was offered three times daily for 13 d with intake recorded. There were no differences (P > 0.05) between breeds of sheep. In addition, sheep consumed more salt cedar than goats except on the last day of the study. When salt cedar was offered three times daily, both breeds of sheep increased intake and gained weight over the 13 d of feeding in Trial 2. By the end of the study, intake appeared to still be increasing. Collectively, these results illustrate that both Rambouillet and Suffolk sheep will consume a similar amount of salt cedar as goats and will provide another species of livestock that can be potentially used to reduce salt cedar cover.

We have previously demonstrated that recurrent application of glyphosate causes dramatic shift in the vegetation structure of the native grasslands of Flooding Pampa. As these structural changes might alter functional processes such as primary production, carbon, nitrogen and phosphorus cycling, this study aims to evaluate functional changes associated with the application of glyphosate in these temperate grasslands. We measured aboveground net primary production (ANPP) during two consecutive years, and the concentration of organic carbon, nitrogen and phosphorus in the soil during the following six years after primary production measurements ended in glyphosate treated and non- treated (control) paddocks of a commercial livestock far. We related the vegetation data, basal cover, species richness and diversity, obtained in a previous study conducted in the same paddocks of the livestock farm, with ANPP data obtained in this one. Late summer applications of glyphosate greatly reduced the biomass contribution of warm-season perennial grasses and legumes and increased the contribution of cool season annual grasses, altering the seasonal pattern of ANPP. As the reduction of the spring and summer productivity could not be compensated by the increase of cool-season productivity, the annual ANPP was lower in the glyphosate-treated paddocks than in control paddocks. Glyphosate applications also decreased soil organic carbon and phosphorus concentration, probably because of the reduction of ANPP, the changes of its seasonal distribution and the shift in the floristic composition of the community, which may modify the amount and quality of the litter. We found a linear positive relationship between basal cover, species richness and species diversity with ANPP, which suggest that the negative effects on ecosystem functioning would be directly related with the changes in vegetation structure caused by glyphosate application.

The highly centralized management of rangelands in northwest Iran has caused their degradation. Rangeland Management Cooperatives (RMCs) have been taken into account by the Iranian researchers and practitioners as the best mode of managing and tackling these resources. In this regard, stakeholders' participation (i.e., the rangers) in such institutions is a substantial issue because without their close collaboration, any management scheme is likely to fail or succeed partially. Therefore, this study investigates the rangers' participation in RMCs using the theory-triangulation method. We developed the main research question: how the explanatory variables, extracted from the social-psychological models, influence rangers' participation in RMCs. A sample of 200 rangers participated in the survey method, of which we received 179 completed self-reported questionnaires. The reliability of the questionnaire was calculated using the Confirmatory Factor Analysis and Kuder-Richardson 21, the metrics that measure the consistency of items in indicator variables with the interval and binary scales, respectively. The results of path analysis unveil that job satisfaction and progressivism have a direct effect on participation, and the improved economic conditions of industries developed by the RMCs, good intrarelation, fatalism, progressivism, optimism, and cost-benefit indirectly influence participation via job satisfaction. On the basis of these results, it is concluded that to increase rangers' participation in RMCs, which is a key factor in preventing the degradation of rangelands, RMCs' officials need to improve the local industries benefiting from the rangelands and upgrade intracommunication skills via training. It is also suggested that all rangers, even those with fatalistic beliefs, need to be included in RMCs' participatory activities. Finally, it is needed to assess progressivist rangers' needs, promote optimism, and visualize the economic, social, and conservation benefits of the participation in RMCs.

The Borana pastoral systemhas long been regarded as amodel for sustainable resource use in eastern Africa. Recent growth in human and livestock populations, however, has contributed to amarked decline in rangeland condition, as well as increasing poverty. Another trend is fewer pastoralists controlling more resources. Today, for example, only 10% of households own 60% of all livestock. This wealthyminority has become increasingly important but has received little research attention. We wanted to learn how such elites perceive system change and how they innovate when accumulating or managing their assets. Twelve wealthy men were interviewed. They noted that the pastoral system is in sharp decline,with the most serious livestock-production constraints including chronic shortages of forage and labor. The average value of the physical and financial assets held by these men was estimated as at least USD $164,000, about 62-times that held by poor households. The average investment portfolio was composed of livestock (two-thirds of total value), while savings accounts in local banks and urban real estate (largely housing) made up the remainder. Livestock in general—and cattle in particular—were the riskiest physical assets given recurrent effects of drought and forage scarcity on animal productivity andmortality. When asked to identify future investment priorities, the men said that investing in urban real estate and their children was now preferred to investing inmore livestock; their tradition of steady livestock reinvestment has thus changed. Recent urban growth in the rangelands has given the wealthy elite new investment options that offset heightened risks of animal losses. Urban investments are important because they could facilitate town development and provide incentives to improve rangemanagement via destocking. Outreach programs focused on the diversification of pastoral assets could includewealthy pastoralists as opinion leaders and accelerate positive change here.