Domestic livestock have the potential to function as ecosystem engineers in semiarid rangelands, but prevailing management practices largely emphasize livestock production and uniform use of vegetation. As a result, variation in vegetation structure might not occur at appropriate spatial and temporal scales to achieve some contemporary conservation objectives. Here, we introduce the utility of livestock as ecosystem engineers and address potential benefits and consequences associated with heterogeneity-based management practices for conservation grazing in the semiarid rangelands of the western North American Great Plains. To illustrate the potential value of this approach, we provide specific examples where engineering effects of livestock could alter vegetation heterogeneity at within-pasture (< 100 ha) and among-pasture (∼100 ha to thousands of hectares) scales to improve habitat for declining native grassland birds. Experimental evaluations of the efficacy of livestock to achieve desired modifications to vegetation structure are needed, along with the economic aspects associated with implementing heterogeneity-based management practices. Using livestock as ecosystem engineers to alter vegetation structure for grassland bird habitat is feasible in terms of application by land managers within the context of current livestock operations, and provides land managers important tools to achieve desired contemporary objectives and outcomes in semiarid rangelands of the western North American Great Plains.

Heathlands in the northwest of Spain have been traditionally used by domestic herbivores as a food resource. However, their abandonment in the past decades has promoted a high incidence of wildfires, threatening biodiversity. Sheep and goats exhibit different grazing behavior, affecting rangelands dynamics in a different way, but the botanical and structural composition may also affect such dynamics. The aim of this article was to compare the grazing effects of sheep and goats on three different heathland types: previously burned grass- or gorse (Ulex gallii Planchon)-dominated and unburned heather (Erica spp.)-dominated shrublands. Two grazing treatments (sheep or goats) were applied in each vegetation type in a factorial design with two replicates (12 experimental plots). A small fenced area was excluded from grazing in each plot (control treatment). The experiment was carried out from 2003 to 2006, and the grazing season extended from May to October–November. Plant cover, canopy height, and phytomass amount and composition were assessed in each plot. Results showed that goats controlled shrub encroachment, phytomass accumulation, and canopy height more than sheep in either burned grass– and gorse– and unburned heather–dominated shrublands. It was accompanied by a higher increase of herbaceous species under goat grazing. Nevertheless, plant dynamics showed different trends between the three vegetation types studied. Grazing effects were more important in previously burned grass-dominated heathlands than in unburned heather-dominated shrublands. At the end of the experiment (May 2006), shrub cover, height, and woody phytomass were significantly higher in the ungrazed enclosures than in the grazed plots. Small ruminant grazing, especially with goats, is proposed as an efficient tool to reduce shrub encroachment and woody phytomass accumulation in heathlands, thus reducing fire hazard, although these grazing effects depend on heathland composition.

The aim of this study, performed on 62 adult dry cashmere goats grazing upland perennial ryegrass–white clover pastures and naturally infected with gastrointestinal nematodes, was to investigate the effects of stocking rate (SR: 24 vs. 38 goats · ha⁻¹) and tannin-containing heather supplementation (H: Calluna vulgaris [L.] Hull, Erica spp.) vs. nonsupplementation on parasite burden, fecal egg counts (FEC), and live weight (LW) changes. Goats were randomly assigned to four treatments in a 2 × 2 factorial arrangement and grazed continuously from May to October. Six goats per treatment were slaughtered at the end of the grazing period, and adult worms in the abomasum and small and large intestines of each animal were recovered, counted, and identified. FEC was affected by SR (P < 0.01) but not by H. However, the SR × H interaction was significant (P < 0.05). FEC increased (P < 0.001) along the grazing season in all treatments, and the SR × time interaction was significant (P < 0.001). In general, mean total worm counts in abomasum and small intestine tended to be higher under high SR, although the differences were only significant (P < 0.01) in Trichostrongylus spp. counts. In goats managed under the high SR, the mean of total Teladorsagia circumcincta counts was lower (P < 0.01) in supplemented animals, but no differences were recorded for Trichostrongylus spp., Chabertia ovina, Oesophagostomum columbianum, and Trichuris ovis. The goats gained more LW (P < 0.001) under low SR and when they were heather-supplemented. No significant SR × H interaction was found for LW change. In conclusion, high stocking rate increases the infectivity risk of pasture and the supplementation of grazing goats with heather contributing to improve animals' performance. Notwithstanding, the effect of heather availability on nematode FEC reduction could be highly dependent on the climatic conditions.

Ecological systems comprise a complex array of interacting processes that manifest across multiple scales. Effective management of natural ecosystems has to be underpinned by an understanding of how the scaling of these processes influences system integrity and stability. This is particularly true in semiarid rangelands, which display strong relationships between pattern and process that are fundamental to maintaining ecosystem function. Grazing can disrupt the scaling of these relationships and the mechanistic coupling between pattern and process, undermining the health of grazed semiarid rangelands. This is due to possible hysteretic responses in key system components to increases and decreases in grazing disturbance. We used data from a semiarid rangeland in northern Australia to test for hysteretic responses in system components after the removal of cattle grazing. We found an uncoupling of spatial linkages between vegetation and soil moisture in a severely degraded plot that was not evident in less intensively grazed or recovering plots. Recovering plots protected from grazing for 20 yr showed a scale of spatial linkage between vegetation and soil moisture, and soil organic matter and mineralization flush, of a scale much coarser than that of degrading plots. These findings provide evidence for hysteretic recovery from grazing and demonstrate that comparison of the spatial patterns of vegetation and soil properties is essential for capturing the true state of ecological functionality in this system. This has important implications for assessing ecological function in systems typified by strong natural environmental variation or in which data for pristine conditions are lacking.

Lactation can have significant costs to individual and population-level productivity because of the high energetic demands it places on dams. Because the difference in condition between lactating and dry Rocky Mountain elk (Cervus elaphus nelsoni) cows tends to disappear as nutritional quality rises, the magnitude of that difference could be used to relate condition to habitat quality or the capability of habitats to support elk. We therefore compared nutritional condition of ≥ 2.5-yr-old lactating and dry cows from six free-ranging Rocky Mountain elk populations throughout the United States. Our goal was to quantify differential accrual of body fat (BF) reserves to determine whether the condition of dry and lactating cows could be used to define relevant management thresholds of habitat quality (i.e., relative carrying capacity) and consequently potential performance of elk populations. Levels of BF that lactating cows were able to accrue in autumn and the proportional difference in BF between dry and lactating cows in autumn were related (F_1–2,10 ≥ 16.2, P < 0.001). Models indicated that elk experienced no negative effects of reproduction on condition when lactating cows were able to accrue ≥ 13.7% BF in autumn. When lactating cows are accruing ≤ 7.9% BF, elk are in a nutritionally stressed condition, which may be limiting population performance. Using the logistic model to predict relative proximity to ecological carrying capacity (ECC), our population-years ranged from 3–97% of ECC and proportion of the population lactating (an index of calf survival) was negatively related to proportion of ECC. Results indicate that the proportional difference in accrual of BF between lactating and dry cows can provide a sensitive index to where elk populations reside relative to the quality of their range.

The dehesa is defined as an agroforestry system that is characteristic of the southwestern Iberian Peninsula, where grassland is combined with evergreen species of the genus Quercus. Those systems have been gradually transformed from the Mediterranean forest into a unique kind of pastoral woodland by means of an agricultural use. Dehesas occupy more than 6 million ha, and the livestock systems that are based in them are of vital importance for their sustainability. The present work classifies, describes, and evaluates the sustainability of these systems in the Spanish region of Extremadura (southwestern Spain). To this end, we apply a methodological adaptation of the Framework for the Evaluation of Management Systems incorporating Sustainability Index (MESMIS). MESMIS is based on the evaluation of basic attributes of sustainability from indicators that allow one to make a simultaneous and comparative analysis of different types of farms. For the study, 69 farms were selected at random, and were classified using multivariate techniques into four types according to their level of intensification and productive orientation. The results were used to obtain an overall value of sustainability from a technical economic perspective for each farm type present in the dehesa. The mixed systems (beef cattle–sheep–Iberian pigs) have been found to be the most sustainable in general terms. The high–stocking rate sheep dehesas are the least sustainable, although at present, they are the most profitable. The other two groups analyzed, “low–stocking rate sheep farms” and “beef cattle farms,” had intermediate and similar scores. Mixed livestock dehesa farms are the closest to the traditional systems with a highly diverse production, an optimal use of the system's resources, and little dependence on external subsidies. In the present context, with uncertainties about European Union subsidies, this type of farm should be a goal for dehesa farmers.

Information about forage productivity and its interactions with cultural practices or climatic variation is necessary to plan livestock management and to increase production without damaging the environment. Remote sensing provides a valuable data source to achieve these goals. Here we characterize forage production over a large region (92 million hectares) by analyzing spatial, seasonal, and interannual variability with Normalized Difference Vegetation Index (NDVI) data. We identified 23 homogeneous zones that enclose multiple counties with similar characteristics of land use and productivity. A long-term series (1981–2000) of Advanced Very High Resolution Radiometer images were used to calculate monthly NDVI and the annual integral of NDVI (I-NDVI), which is an estimate of primary productivity, for each county. County agricultural land use data were used to resolve pure forage and crop NDVI patterns over time using a spectral unmixing model. The annual integral of NDVI was significantly associated with geographic longitude and average precipitation but not with latitude. Improved relationships between forage production and I-NDVI can be obtained by collecting more accurate forage estimates in the field and calculating radiation use efficiencies. Images of high temporal resolution allow the inference of seasonal changes, and images of high spatial resolution allow a more precise description of the forage resources.

In arid ecosystems, a few large summer rains frequently differentiate wet years from dry ones. However, use of this additional water by plants has limited experimental evidence. We applied a 16-mm summer water pulse (12% of mean annual precipitation) to two plant communities of the Patagonian steppe, and compared responses of three dominant species, which can be ordered by decreasing xerophytism and increasing rooting depth and summer activity: 1) colapiche, evergreen dwarf shrub (Nassauvia glomerulosa [Lag.] Don); 2) coirón amargo, evergreen grass (Stipa speciosa Trin. et Rupr.); and 3) neneo, drought deciduous shrub (Mulinum spinosum [Cav.] Pers.). Shallow-rooted species (S. speciosa and N. glomerulosa), which use water from dry soil layers, showed a greater leaf water potential response to watering than deep-rooted species (M. spinosum). Leaf water potential response was greater and quicker in xerophytic species than in mesophytic ones (N. glomerulosa > S. speciosa > M. spinosum). However, this response only translated into leaf conductance and transpiration responses for coirón amargo, probably because the species with winter phenological cycle (N. glomerulosa) is less able to utilize summer water inputs than species with a summer phenological cycle (S. speciosa). The lack of response of deep-rooted M. spinosum in leaf conductance, transpiration, and photosynthesis may have been due to the high leaf water potential of control plants. Instead, in S. speciosa and N. glomerulosa net photosynthesis decreased below zero following watering, suggesting the start of growth pulses. The complex chain of plant processes triggered by rainfall, and the constraints imposed to different species by rooting depth, phenology, and xerophytism, could explain the frequent low response to both large rainfall events and above-average rainfall years in this arid community. Our results suggest that, paradoxically, water may be suboptimally used at local scales in arid rangeland ecosystems.

Efficacy of fire in reducing shrub density is low in plant communities where most woody plants resprout from stem bases and crowns following fire. Our objective was to determine the relationship of shrub mortality and recovery from summer fire to prefire shrub structural characteristics. A randomized, complete block design with two treatments (burned and control) and three blocks was used in the experiment. Within each block and treatment combination, we randomly selected 40 individuals each of brasil (Condalia hookeri M. C. Johnst.), huisache (Acacia farnesiana [L.] Willd.), and spiny hackberry (Celtis ehrenbergiana [Klotzsch] Liebm.). We estimated height, canopy diameter, number of stems, stem diameter, and distance to the nearest shrubs before ignition of fires. Fires were ignited during July and August 2001. Survival, sprout number, height, and total plant height were estimated 47–52 wk postburn. Mortality of brasil was 26 times greater on burned sites than on control sites, but mortality of huisache and spiny hackberry was negligible. Mortality of brasil varied from 0% to 68% among blocks. Postburn height and number of sprouts increased with preburn shrub height and number of stems, indicating that longer intervals of time between fires that allow shrub growth facilitate more rapid postfire recovery. Factors other than the preburn shrub structural characteristics we measured appear to influence postfire shrub survival most strongly, although these characteristics are useful in predicting postfire sprout production and shrub height.

A threshold represents a point in space and time at which primary ecological processes degrade beyond the ability to self-repair. In ecosystems with juniper (Juniperus L. spp.) encroachment, ecological processes (i.e., infiltration) are impaired as intercanopy plant structure degrades during woodland expansion. The purpose of this research is to characterize influences of increasing juniper on vegetation structure and hydrologic processes in mountain big sagebrush–western juniper (Artemisia tridentata Nutt. subsp. vaseyana [Rydb.] Beetle–Juniperus occidentalis Hook.) communities and to identify and predict states and thresholds. Intercanopy plant cover and infiltration rates were sampled in relation to juniper canopy cover. Study plots, arranged in a randomized complete-block design, represented low shrub–high juniper, moderate shrub–moderate juniper, and high shrub–low juniper percentage of canopy cover levels at four primary aspects. In field plots, percentage of plant cover, bare ground, and steady-state infiltration rates were measured. In the laboratory, juniper canopy cover and topographic position were calculated for the same area using high-resolution aerial imagery and digital elevation data. Parametric and multivariate analyses differentiated vegetation states and associated abiotic processes. Hierarchical agglomerative cluster analysis identified significant changes in infiltration rate and plant structure from which threshold occurrence was predicted. Infiltration rates and percentage of bare ground were strongly correlated (r²  =  0.94). Bare ground was highest in low shrub–high juniper cover plots compared to both moderate and high shrub–low juniper cover levels on south-, east-, and west-facing sites. Multivariate tests indicated a distinct shift in plant structure and infiltration rates from moderate to low shrub–high juniper cover, suggesting a transition across an abiotic threshold. On north-facing slopes, bare ground remained low, irrespective of juniper cover. Land managers can use this approach to anticipate and identify thresholds at various landscape positions.

Selective foraging among free-ranging herbivores can make measuring botanical composition of diets challenging. Using near-infrared reflectance spectroscopy (NIRS) on feces for predicting botanical components of individual animal diets is a novel method for studying diet selection. This study was conducted to determine the ability of fecal NIRS to predict the percentage of consumption of Leymus chinensis (Trin.) Tzvel., a dominant species in north China, by sheep (Ovis aries L.). The calibration data set consisted of 47 diets of known L. chinensis composition, paired with corresponding fecal spectra. These pairs were generated in a trial using restricted feeding. Validation pairs (n  =  9) were collected in a similar trial that used ad libitum feeding. Derived coefficients of determination (R²) and standard error of calibration were 0.99% and 2.2% for partial least squares (PLS) regression and 0.89% and 7.3% for stepwise regression, respectively. Derived coefficients of determination (r²) and standard error of prediction were 0.78% and 4.8% for PLS regression and 0.90% and 3.2% for stepwise regression, respectively. PLS regression resulted in better calibration than stepwise regression, but when the calibration data set was small, stepwise regression improved the precision and accuracy of predictions compared with the PLS regression. Results of the present study show that a fecal NIRS equation developed from a restricted feeding trial could be used to predict the percentage of L. chinensis in fecal materials collected from voluntary feeding trials.

Belowground plant biomass plays a critical role in the maintenance of riparian ecosystems and generally constitutes the majority of the total biomass on a site. Despite this importance, belowground dynamics of riparian plant species are not commonly investigated, in part because of difficulties of sampling in a belowground riparian environment. We investigated the field utility of a root-ingrowth sampling technique for measuring root production. We established four streamside sampling sites in southeastern Oregon, and randomly located four plots within each site. In each plot we established two 7.6-cm–diameter sand-filled ingrowth cores in September of 2004. In September of 2005 we harvested the cores with the use of a vacuum sampling technique in which a 5.1-cm–diameter camphored polyvinyl chloride casing was driven into the center of the root core and sand and root materials were suctioned out. Root-length density was determined by computer image analysis, and roots were dried and weighed to determine production by weight. Results indicate that root-length density averaged 7.2 (± 0.7) cm · cm⁻³ across sites and root-production index was 356.7 (± 20.6) g · m⁻². Our index to root production by weight was consistent with previous estimates of annual root production reported in the literature. Our sampling technique proved to be a practical solution for root sampling in riparian environments, and helps overcome some of the difficulties in sequential coring of saturated soils. Use of any ingrowth core technique to index root production can potentially bias production estimates because of the artificial, root-free environment of the core. However, these biases should be consistent across sites, making ingrowth cores useful for determining differences between manipulative treatments.