Grazing herbivores reduce herbaceous biomass and fire activity across African savannas.

Fire and herbivory interact to alter ecosystems and carbon cycling. In savannas, herbivores can reduce fire activity by removing grass biomass, but the size of these effects and what regulates them remain uncertain. To examine grazing effects on fuels and fire regimes across African savannas, we combined data from herbivore exclosure experiments with remotely sensed data on fire activity and herbivore density. We show that, broadly across African savannas, grazing herbivores substantially reduce both herbaceous biomass and fire activity. The size of these effects was strongly associated with grazing herbivore densities, and surprisingly, was mostly consistent across different environments. A one-zebra increase in herbivore biomass density (~100 kg/km2 of metabolic biomass) resulted in a ~53 kg/ha reduction in standing herbaceous biomass and a ~0.43 percentage point reduction in burned area. Our results indicate that fire models can be improved by incorporating grazing effects on grass biomass.

Disease and weather induce rapid shifts in a rangeland ecosystem mediated by a keystone species (Cynomys ludovicianus).

Habitat loss and changing climate have direct impacts on native species but can also interact with disease pathogens to influence wildlife communities. In the North American Great Plains, black-tailed prairie dogs (Cynomys ludovicianus) are a keystone species that create important grassland habitat for numerous species and serve as prey for predators, but lethal control driven by agricultural conflict has severely reduced their abundance. Novel disease dynamics caused by epizootic plague (Yersinia pestis) within prairie dog colonies have further reduced prairie dog abundances, in turn destabilizing associated wildlife communities. We capitalized on a natural experiment, collecting data on prairie dog distributions, vegetation structure, avian abundance, and mesocarnivore and ungulate occupancy before (2015-2017) and after (2018-2019) a plague event in northeastern Wyoming, USA. Plague decimated black-tailed prairie dog populations in what was then the largest extant colony complex, reducing colony cover in the focal area from more than 10,000 ha to less than 50 ha. We documented dramatic declines in mesocarnivore occupancy and raptor abundance post-plague, with probability of occupancy or abundance approaching zero in species that rely on prairie dogs for a high proportion of their diet (e.g., ferruginous hawk [Buteo regalis], American badger [Taxidea taxus], and swift fox [Vulpes velox]). Following the plague outbreak, abnormally high precipitation in 2018 hastened vegetation recovery from prairie dog disturbance on colonies in which constant herbivory had formerly maintained shortgrass structure necessary for certain colony-associates. As a result, we observed large shifts in avian communities on former prairie dog colonies, including near-disappearance of mountain plovers (Charadrius montanus) and increases in mid-grass associated songbirds (e.g., lark bunting [Calamospiza melanocorys]). Our research highlights how precipitation can interact with disease-induced loss of a keystone species to induce drastic and rapid shifts in wildlife communities. Although grassland taxa have co-evolved with high spatiotemporal variation, fragmentation of the remaining North American rangelands paired with higher-than-historical variability in climate and disease dynamics are likely to destabilize these systems in the future.

A big data-model integration approach for predicting epizootics and population recovery in a keystone species.

Infectious diseases pose a significant threat to global health and biodiversity. Yet, predicting the spatiotemporal dynamics of wildlife epizootics remains challenging. Disease outbreaks result from complex nonlinear interactions among a large collection of variables that rarely adhere to the assumptions of parametric regression modeling. We adopted a nonparametric machine learning approach to model wildlife epizootics and population recovery, using the disease system of colonial black-tailed prairie dogs (BTPD, Cynomys ludovicianus) and sylvatic plague as an example. We synthesized colony data between 2001 and 2020 from eight USDA Forest Service National Grasslands across the range of BTPDs in central North America. We then modeled extinctions due to plague and colony recovery of BTPDs in relation to complex interactions among climate, topoedaphic variables, colony characteristics, and disease history. Extinctions due to plague occurred more frequently when BTPD colonies were spatially clustered, in closer proximity to colonies decimated by plague during the previous year, following cooler than average temperatures the previous summer, and when wetter winter/springs were preceded by drier summers/falls. Rigorous cross-validations and spatial predictions indicated that our final models predicted plague outbreaks and colony recovery in BTPD with high accuracy (e.g., AUC generally >0.80). Thus, these spatially explicit models can reliably predict the spatial and temporal dynamics of wildlife epizootics and subsequent population recovery in a highly complex host-pathogen system. Our models can be used to support strategic management planning (e.g., plague mitigation) to optimize benefits of this keystone species to associated wildlife communities and ecosystem functioning. This optimization can reduce conflicts among different landowners and resource managers, as well as economic losses to the ranching industry. More broadly, our big data-model integration approach provides a general framework for spatially explicit forecasting of disease-induced population fluctuations for use in natural resource management decision-making.

Predicting spatial-temporal patterns of diet quality and large herbivore performance using satellite time series.

Adaptive management of large herbivores requires an understanding of how spatial-temporal fluctuations in forage biomass and quality influence animal performance. Advances in remote sensing have yielded information about the spatial-temporal dynamics of forage biomass, which in turn have informed rangeland management decisions such as stocking rate and paddock selection for free-ranging cattle. However, less is known about the spatial-temporal patterns of diet quality and their influence on large herbivore performance. This is due to infrequent concurrent ground observations of forage conditions with performance (e.g., mass gain), and previously limited satellite data at fine spatial and temporal scales. We combined multi-temporal field observations of diet quality (weekly) and mass gain (monthly) with satellite-derived phenological metrics (pseudo-daily, using data fusion and interpolation) to model daily mass gains of free-ranging yearling cattle in shortgrass steppe. We used this model to predict grazing season (mid-May to October) mass gains, a key management indicator, across 40 different paddocks grazed over a 10-year period (n = 138). We found strong relationships between diet quality and the satellite-derived phenological metrics, especially metrics related to the timing and rate of green-up and senescence. Satellite-derived diet quality estimates were strong predictors of monthly mass gains (R2  = 0.68) across a wide range of aboveground net herbaceous production. Season-long predictions of average daily gain and cattle off-mass had mean absolute errors of 8.9% and 2.9%, respectively. The model performed better temporally (across repeated observations in the same paddock) than spatially (across all paddocks within a given year), highlighting the need for accurate vegetation maps and robust field data collection across both space and time. This study demonstrates that free-ranging cattle performance in rangelands is strongly affected by diet quality, which is related to the timing of vegetation green-up and senescence. Senescing vegetation suppressed mass gains, even if adequate forage was available. The satellite-based pseudo-daily approach presented here offers new opportunities for adaptive management of large herbivores, such as identifying within-season triggers to move livestock among paddocks, predicting wildlife herd health, or timing the grazing season to better match earlier spring green-up caused by climate change and plant species invasion.

At high stocking rates, cattle do not functionally replace wild herbivores in shaping understory community composition.

Over a quarter of the world's land surface is grazed by cattle and other livestock, which are replacing wild herbivores, potentially impairing ecosystem structure, and functions. Previous research suggests that cattle at moderate stocking rates can functionally replace wild herbivores in shaping understory communities. However, it is uncertain whether this is also true under high stocking rates and the effects of wild herbivore on plant communities are moderate, enhanced, or simply additive to the effects of cattle at high stocking rates. To evaluate the influence of cattle stocking rates on the ability of cattle to functionally replace wild herbivores and test for interactive effects between cattle and wild herbivores in shaping understory vegetation, we assessed herbaceous vegetation in a long-term exclosure experiment in a semi-arid savanna in central Kenya that selectively excludes wild mesoherbivores (50-1000 kg) and megaherbivores (elephant and giraffe). We tested the effects of cattle stocking rate (zero/moderate/high) on herbaceous vegetation (diversity, composition, leafiness). We also tested how those effects depend on the presence of wild mesoherbivores and megaherbivores. We found that herbaceous community composition (primary ordination axis) was better explained by the presence/absence of herbivore types than by total herbivory, suggesting that herbivore identity is a more important determinant of community composition than total herbivory at high cattle stocking rates. The combination of wild mesoherbivores and cattle stocked at high rates led to increased bare ground and annual grass cover, reduced perennial grass cover and understory leafiness, and enhanced understory diversity. These shifts were weaker or absent when cattle were stocked at high stocking rates in the absence of wild mesoherbivores. Megaherbivores tempered the effects of cattle stocked at high rates on herbaceous community composition but amplified the effects of high cattle stocking rate on bare ground and understory diversity. Our results show that cattle at high stocking rates do not functionally replace wild herbivores in shaping savanna herbaceous communities contrary to previous findings at moderate stocking rates. In mixed-use rangelands, interactions between cattle stocking rate and wild herbivore presence can lead to non-additive vegetation responses with important implications for both wildlife conservation and livestock production.

Plant traits related to precipitation sensitivity of species and communities in semiarid shortgrass prairie.

Understanding how plant communities respond to temporal patterns of precipitation in water-limited ecosystems is necessary to predict interannual variation and trends in ecosystem properties, including forage production, biogeochemical cycling, and biodiversity. In North American shortgrass prairie, we measured plant abundance, functional traits related to growth rate and drought tolerance, and aboveground net primary productivity to identify: species-level responsiveness to precipitation (precipitation sensitivity Sspp ) across functional groups; Sspp relationships to continuous plant traits; and whether continuous trait-Sspp relationships scaled to the community level. Across 32 plant species, we found strong bivariate relationships of both leaf dry matter content (LDMC) and leaf osmotic potential Ψosm with Sspp . Yet, LDMC and specific leaf area were retained in the lowest Akaike information criterion multiple regression model, explaining 59% of Sspp . Most relationships between continuous traits and Sspp scaled to the community level but were often contingent on the presence/absence of particular species and/or land management at a site. Thus, plant communities in shortgrass prairie may shift towards slower growing, more stress-resistant species in drought years and/or chronically drier climate. These findings highlight the importance of both leaf economic and drought tolerance traits in determining species and community responses to altered precipitation.

Managing risks related to climate variability in rangeland-based livestock production: What producer driven strategies are shared and prevalent across diverse dryland geographies?

Rangeland-based livestock production (RBLP) primarily occurs in drylands where interannual variation in rainfall directly and indirectly affects economies, plant primary productivity (forage production), and livestock reproduction and mortality. Tight ecological and economic links to climate variation constrain production in dryland systems, but producers have a breadth of strategies to reduce climate-related risks and maintain RBLP. Research on these strategies has focused on context-specific tactics linked to specific systems and/or geographies. Inspired by studies that look for broader patterns to offer frameworks for discourse and to advance collective knowledge, we review global literature to identify risk management strategies related to climate variability that are in widespread use across dryland rangeland systems and geographies. We organize strategies within three key decision areas for producers engaged in RBLP: profit and return options, land use, and herd management. Across the decision areas, four strategies emerge as playing a strong role in risk management across the globe, with refinements based on local conditions. These shared and prevalent producer driven strategies are dynamic management of forage supply (in the decision area of land use), dynamic management of animal demand (in the area of herd management), and diversification and use of social networks (both of which apply across all three decision areas). Within each of the decision areas, we found diversification reduces climate related risks but has circumstances under which it is less effective; for example, large landholders already buffered to risk via landscape diversity benefit less from livelihood diversification. In practice, implementation of the four strategies often results in livestock producers who do not maximize short-term profits but instead prioritize land resilience, large herd sizes, lifestyle goals, and longer-term economic sustainability. In this synthesis, we considered existing producer strategies for reducing risk related to climate related variability -- an intrinsic and defining characteristic of dryland rangelands -- in order to highlight valuable areas in which research can support problem solving across diverse RBLP geographies and economies, especially in a changing climate.

Thresholds are in the eye of the beholder: plants and wildlife respond differently to short-term cattle corrals.

Rangelands are governed by threshold dynamics, and factors such as drought, wildfire, and herbivory can drive change across thresholds and between ecological states. Most work on this topic has focused on shifts in a single response variable, vegetation, and little research has considered how to reconcile responses of more than one variable to determine whether a system has undergone a genuine state change. In sub-Saharan Africa, mobile overnight livestock corrals (bomas) can be used by managers to precipitate ecological transitions from areas dominated by bare ground to productive ecosystem hotspots (glades) that are attractive to wild herbivores. We asked how long bomas must be occupied by cattle before undergoing a state change, considering both plant and animal response variables, to glade ecosystem hotspots. We tested five durations of boma occupation: 0, 4, 7, 14, and 28 days. Each treatment was replicated five times, and we assessed vegetation as well as herbivore dung (as a proxy of use) at multiple time points over 3 yr following boma abandonment. Vegetation in 7-, 14-, and 28-d boma duration treatments appeared to undergo a complete transition to glade-like plant communities, whereas the shortest 4-d treatment had not converted to a glade plant community by year 3. Wildlife responses appeared to lag behind vegetation responses, with transitions to glade-like herbivore use occurring only in the longest duration (14- and 28-d) treatments. Our results show that different response variables, when considered individually, may provide incomplete or misleading information about state changes. Although shorter-occupied bomas might be effective for reducing bare ground, they may not attract enough wild herbivores to constitute crossing into an alternative state. Understanding threshold dynamics associated not only with vegetation responses but with a broader suite of response variables is challenging, but will provide a more complete representation of ecosystem function and greater opportunity for more successful ecosystem management.

Herbivory and drought generate short-term stochasticity and long-term stability in a savanna understory community.

Rainfall and herbivory are fundamental drivers of grassland plant dynamics, yet few studies have examined long-term interactions between these factors in an experimental setting. Understanding such interactions is important, as rainfall is becoming increasingly erratic and native wild herbivores are being replaced by livestock. Livestock grazing and episodic low rainfall are thought to interact, leading to greater community change than either factor alone. We examined patterns of change and stability in herbaceous community composition through four dry periods, or droughts, over 15 years of the Kenya Long-term Exclosure Experiment (KLEE), which consists of six different combinations of cattle, native wild herbivores (e.g., zebras, gazelles), and mega-herbivores (giraffes, elephants). We used principal response curves to analyze the trajectory of change in each herbivore treatment relative to a common initial community and asked how droughts contributed to community change in these treatments. We examined three measures of stability (resistance, variability, and turnover) that correspond to different temporal scales and found that each had a different response to grazing. Treatments that included both cattle and wild herbivores had higher resistance (less net change over 15 years) but were more variable on shorter time scales; in contrast, the more lightly grazed treatments (no herbivores or wild herbivores only) showed lower resistance due to the accumulation of consistent, linear, short-term change. Community change was greatest during and immediately after droughts in all herbivore treatments. But, while drought contributed to directional change in the less grazed treatments, it contributed to both higher variability and resistance in the more heavily grazed treatments. Much of the community change in lightly grazed treatments (especially after droughts) was due to substantial increases in cover of the palatable grass Brachiaria lachnantha. These results illustrate how herbivory and drought can act together to cause change in grassland communities at the moderate to low end of a grazing intensity continuum. Livestock grazing at a moderate intensity in a system with a long evolutionary history of grazing contributed to long-term stability. This runs counter to often-held assumptions that livestock grazing leads to directional, destabilizing shifts in grassland systems.

Herbivore effects on productivity vary by guild: cattle increase mean productivity while wildlife reduce variability.

Wild herbivores and livestock share the majority of rangelands worldwide, yet few controlled experiments have addressed their individual, additive, and interactive impacts on ecosystem function. While ungulate herbivores generally reduce standing biomass, their effects on aboveground net primary production (ANPP) can vary by spatial and temporal context, intensity of herbivory, and herbivore identity and species richness. Some evidence indicates that moderate levels of herbivory can stimulate aboveground productivity, but few studies have explicitly tested the relationships among herbivore identity, grazing intensity, and ANPP. We used a long-term exclosure experiment to examine the effects of three groups of wild and domestic ungulate herbivores (megaherbivores, mesoherbivore wildlife, and cattle) on herbaceous productivity in an African savanna. Using both field measurements (productivity cages) and satellite imagery, we measured the effects of different herbivore guilds, separately and in different combinations, on herbaceous productivity across both space and time. Results from both productivity cage measurements and satellite normalized difference vegetation index (NDVI) demonstrated a positive relationship between mean productivity and total ungulate herbivore pressure, driven in particular by the presence of cattle. In contrast, we found that variation in herbaceous productivity across space and time was driven by the presence of wild herbivores (primarily mesoherbivore wildlife), which significantly reduced heterogeneity in ANPP and NDVI across both space and time. Our results indicate that replacing wildlife with cattle (at moderate densities) could lead to similarly productive but more heterogeneous herbaceous plant communities in rangelands.

Are cattle surrogate wildlife? Savanna plant community composition explained by total herbivory more than herbivore type.

The widespread replacement of wild ungulate herbivores by domestic livestock in African savannas is composed of two interrelated phenomena: (1) loss or reduction in numbers of individual wildlife species or guilds and (2) addition of livestock to the system. Each can have important implications for plant community dynamics. Yet very few studies have experimentally addressed the individual, combined, and potentially interactive effects of wild vs. domestic herbivore species on herbaceous plant communities within a single system. Additionally, there is little information about whether, and in which contexts, livestock might functionally replace native herbivore wildlife or, alternatively, have fundamentally different effects on plant species composition. The Kenya Long-term Exclosure Experiment, which has been running since 1995, is composed of six treatment combinations of mega-herbivores, meso-herbivore ungulate wildlife, and cattle. We sampled herbaceous vegetation 25 times between 1999 and 2013. We used partial redundancy analysis and linear mixed models to assess effects of herbivore treatments on overall plant community composition and key plant species. Plant communities in the six different herbivore treatments shifted directionally over time and diverged from each other substantially by 2013. Plant community composition was strongly related (R2  = 0.92) to residual plant biomass, a measure of herbivore utilization. Addition of any single herbivore type (cattle, wildlife, or mega-herbivores) caused a shift in plant community composition that was proportional to its removal of plant biomass. These results suggest that overall herbivory pressure, rather than herbivore type or complex interactions among different herbivore types, was the main driver of changes in plant community composition. Individual plant species, however, did respond most strongly to either wild ungulates or cattle. Although these results suggest considerable functional similarity between a suite of native wild herbivores (which included grazers, browsers, and mixed feeders) and cattle (mostly grazers) with respect to understory plant community composition, responses of individual plant species demonstrate that at the plant-population-level impacts of a single livestock species are not functionally identical to those of a diverse group of native herbivores.

Grazing intensity differentially regulates ANPP response to precipitation in North American semiarid grasslands.

Grazing intensity elicits changes in the composition of plant functional groups in both shortgrass steppe (SGS) and northern mixed-grass prairie (NMP) in North America. How these grazing intensity-induced changes control aboveground net primary production (ANPP) responses to precipitation remains a central open question, especially in light of predicted climate changes. Here, we evaluated effects of four levels (none, light, moderate, and heavy) of long-term (>30 yr) grazing intensity in SGS and NMP on: (1) ANPP; (2) precipitation-use efficiency (PUE, ANPP : precipitation); and (3) precipitation marginal response (PMR; slope of a linear regression model between ANPP and precipitation). We advance prior work by examining: (1) the consequences of a range of grazing intensities (more grazed vs. ungrazed); and (2) how grazing-induced changes in ANPP and PUE are related both to shifts in functional group composition and physiological responses within each functional group. Spring (April-June) precipitation, the primary determinant of ANPP, was only 12% higher in NMP than in SGS, yet ANPP and PUE were 25% higher. Doubling grazing intensity in SGS and nearly doubling it in NMP reduced ANPP and PUE by only 24% and 33%, respectively. Increased grazing intensity reduced C3 graminoid biomass and increased C4 grass biomass in both grasslands. Functional group shifts affected PUE through biomass reductions, as PUE was positively associated with the relative abundance of C3 species and negatively with C4 species across both grasslands. At the community level, PMR was similar between grasslands and unaffected by grazing intensity. However, PMR of C3 graminoids in SGS was eightfold higher in the ungrazed treatment than under any grazed level. In NMP, PMR of C3 graminoids was only reduced under heavy grazing intensity. Knowing the ecological consequences of grazing intensity provides valuable information for mitigation and adaptation strategies in response to predicted climate change. For example, moderate grazing (the recommended rate) in SGS would sequester the same amount of aboveground carbon as light grazing because ANPP was nearly the same. In contrast, reductions in grazing intensity in NMP from moderate to light intensity would increase the amount of aboveground carbon sequestrated by 25% because of increased ANPP.

The LTAR Grazing Land Common Experiment at the Central Plains Experimental Range: Collaborative adaptive rangeland management.

Semiarid rangelands throughout the western Great Plains support livestock production and many other ecosystem services. The degree to which adaptive multi-paddock (AMP) grazing management approaches can help achieve desired ecosystem services remains unclear. At the Central Plains Experimental Range in northeastern Colorado, a management-science partnership with a diverse stakeholder group is comparing collaborative adaptive rangeland management (CARM), designed to incorporate AMP principles, to traditional rangeland management (TRM), consisting of season-long grazing during the growing season. Each treatment was implemented on a set of 10, 130-ha pastures paired by soils, topography, and plant communities to evaluate how CARM affects vegetation (composition and production), livestock production (steer weight gain), and wildlife habitat (vegetation structure for grassland birds). For the first 5 years of the experiment, CARM cattle were managed as a single herd using AMP grazing with planned year-long rest in 20% of the pastures. Relative to TRM, CARM enhanced heterogeneity in vegetation structure across the landscape, benefiting two grassland bird species. However, this came at the cost of 12%-16% lower steer weight gains in CARM versus TRM and declining populations of a third bird species of conservation concern in both treatments. Here we discuss how increased understanding of ecological and social processes during the experiment's first 5 years led to changes in the CARM treatment and management objectives during the next 5 years. We also discuss how innovations in remote sensing, environmental sensors, ecosystem modeling, social learning, and economic analyses are being integrated into and supported by the CARM experiment.

Edge-effect interactions in fragmented and patchy landscapes.

Ecological edges are increasingly recognized as drivers of landscape patterns and ecosystem processes. In fragmented and patchy landscapes (e.g., a fragmented forest or a savanna with scattered termite mounds), edges can become so numerous that their effects pervade the entire landscape. Results of recent studies in such landscapes show that edge effects can be altered by the presence or proximity of other nearby edges. We considered the theoretical significance of edge-effect interactions, illustrated various landscape configurations that support them and reviewed existing research on this topic. Results of studies from a variety of locations and ecosystem types show that edge-effect interactions can have significant consequences for ecosystems and conservation, including higher tree mortality rates in tropical rainforest fragments, reduced bird densities in grassland fragments, and bush encroachment and reduced wildlife densities in a tropical savanna. To clarify this underappreciated concept and synthesize existing work, we devised a conceptual framework for edge-effect interactions. We first worked to reduce terminological confusion by clarifying differences among terms such as edge intersection and edge interaction. For cases in which nearby edge effects interact, we proposed three possible forms of interaction: strengthening (presence of a second edge causes stronger edge effects), weakening (presence of a second edge causes weaker edge effects), and emergent (edge effects change completely in the presence of a second edge). By clarifying terms and concepts, this framework enables more precise descriptions of edge-effect interactions and facilitates comparisons of results among disparate study systems and response variables. A better understanding of edge-effect interactions will pave the way for more appropriate modeling, conservation, and management in complex landscapes.

Graduate student's guide to necessary skills for nonacademic conservation careers.

Graduate education programs in conservation science generally focus on disciplinary training and discipline-specific research skills. However, nonacademic conservation professionals often require an additional suite of skills. This discrepancy between academic training and professional needs can make it difficult for graduate students to identify the skills and experiences that will best prepare them for the conservation job market. We analyzed job advertisements for conservation-science positions and interviewed conservation professionals with experience hiring early-career conservation scientists to determine what skills employers of conservation professionals seek; whether the relative importance of skills varies by job sector (government, nonprofit, and private); and how graduate students interested in careers in conservation science might signal competency in key skills to potential employers. In job advertisements, disciplinary, interpersonal, and project-management skills were in the top 5 skills mentioned across all job sectors. Employers' needs for additional skills, like program leadership, conflict resolution and negotiation, and technical and information technology skills, varied across sectors. Our interview results demonstrated that some skills are best signaled to employers via experiences obtained outside thesis or dissertation work. Our findings suggest that graduate students who wish to be competitive in the conservation job market can benefit by gaining skills identified as important to the job sector in which they hope to work and should not necessarily expect to be competent in these skills simply by completing their chosen degree path.

Herbivory and drought interact to enhance spatial patterning and diversity in a savanna understory.

The combination of abiotic stress and consumer stress can have complex impacts on plant community structure. Effective conservation and management of semi-arid ecosystems requires an understanding of how different stresses interact to structure plant communities. We explored the separate and combined impacts of episodic drought, livestock grazing, and wild ungulate herbivory on species co-occurrence and diversity patterns in a relatively productive, semi-arid Acacia savanna. Specifically, we analyzed 9 years of biannual plant community data from the Kenya long-term exclosure experiment, a broad-scale manipulative experiment that has excluded different combinations of large mammalian herbivores from 18 4-ha plots since 1995. During droughts, we observed low species diversity and random species co-occurrence patterns. However, when rain followed a major drought, areas exposed to moderate cattle grazing displayed high species diversity and evidence of significant species aggregation. These patterns were not apparent in the absence of cattle, even if other large herbivores were present. To explore possible mechanisms, we examined patterns separately for common and rare species. We found that aggregation patterns were likely driven by rare species responding similarly to the availability of open micro-sites. Our results indicate that in a productive, fire-suppressed savanna, the combination of periodic drought and moderate cattle grazing can enhance plant biodiversity and fine-scale spatial heterogeneity by opening up space for species that are otherwise rare or cryptic. Our findings also emphasize that domestic herbivores can have significantly stronger impacts on plant community dynamics than wild herbivores, even in an ecosystem with a long history of grazing.

Can initial intraspecific spatial aggregation increase multi-year coexistence by creating temporal priority?

Both intraspecific spatial aggregation and temporal priority effects have the potential to increase long-term species coexistence. Theory and models suggest that intraspecific aggregation can facilitate coexistence via limited dispersal or asymmetric interaction distances. During community assembly, intraspecific aggregation may also delay interactions between more and less competitive species, thus creating opportunities for priority effects to facilitate longer-term coexistence. Few empirical studies have tested predictions about aggregation and coexistence, especially in the context of community assembly or ecological restoration. We investigated (1) impacts of intraspecific aggregation on the assembly of eight-species communities over three years, (2) the scale dependence of these impacts, and (3) implications for California prairie restoration. We planted eight native species in each of 19, 5 m wide, octagonal plots. Species were either interspersed throughout the plot or aggregated into eight, 2.2-m(2), wedge-shaped, monospecific sectors. Over three years, species diversity declined more quickly in interspersed plots than in aggregated plots. Two species had higher cover or increased more in interspersed than aggregated plots and were identified as "aggressives." Four species had higher cover or increased more in aggregated than interspersed plots and were identified as "subordinates." Within aggregated plots, aggressive species expanded beyond the sector in which they were originally seeded. Cover of aggressive species increased faster and reached higher values in sectors that were adjacent to the originally planted sector, compared to nonadjacent sectors. Cover of aggressive species also increased more and faster near plot centers, compared to plot edges. Areas near plot centers were representative of smaller aggregation patches since species were planted closer to heterospecific neighbors. Two subordinate species maintained higher cover near plot edges than near plot centers. Moreover, two subordinate species maintained higher cover when seeded in sectors farther away from aggressive species. These results suggest that initial intraspecific aggregation can facilitate species coexistence for at least three years, and larger aggregation patches may be more effective than smaller ones in the face of dispersing dominants. The creation of temporal priority effects may represent an underappreciated pathway by which intraspecific aggregation can increase coexistence. Restorationists may be able to maintain more diverse communities by planting in a mosaic of monospecific patches.

Grasses and browsers reinforce landscape heterogeneity by excluding trees from ecosystem hotspots.

Spatial heterogeneity in woody cover affects biodiversity and ecosystem function, and may be particularly influential in savanna ecosystems. Browsing and interactions with herbaceous plants can create and maintain heterogeneity in woody cover, but the relative importance of these drivers remains unclear, especially when considered across multiple edaphic contexts. In African savannas, abandoned temporary livestock corrals (bomas) develop into long-term, nutrient-rich ecosystem hotspots with unique vegetation. In central Kenya, abandoned corral sites persist for decades as treeless 'glades' in a wooded matrix. Though glades are treeless, areas between adjacent glades have higher tree densities than the background savanna or areas near isolated glades. The mechanisms maintaining these distinctive woody cover patterns remain unclear. We asked whether browsing or interactions with herbaceous plants help to maintain landscape heterogeneity by differentially impacting young trees in different locations. We planted the mono-dominant tree species (Acacia drepanolobium) in four locations: inside glades, far from glades, at edges of isolated glades and at edges between adjacent glades. Within each location, we assessed the separate and combined effects of herbivore exclusion (caging) and herbaceous plant removal (clearing) on tree survival and growth. Both caging and clearing improved tree survival and growth inside glades. When herbaceous plants were removed, trees inside glades grew more than trees in other locations, suggesting that glade soils were favorable for tree growth. Different types of glade edges (isolated vs. non-isolated) did not have significantly different impacts on tree performance. This represents one of the first field-based experiments testing the separate and interactive effects of browsing, grass competition and edaphic context on savanna tree performance. Our findings suggest that, by excluding trees from otherwise favorable sites, both herbaceous plants and herbivores help to maintain functionally important landscape heterogeneity in African savannas.

Direct and indirect effects of a keystone engineer on a shrubland-prairie food web.

Keystone engineers are critical drivers of biodiversity throughout ecosystems worldwide. Within the North American Great Plains, the black-tailed prairie dog is an imperiled ecosystem engineer and keystone species with well-documented impacts on the flora and fauna of rangeland systems. However, because this species affects ecosystem structure and function in myriad ways (i.e., as a consumer, a prey resource, and a disturbance vector), it is unclear which effects are most impactful for any given prairie dog associate. We applied structural equation models (SEM) to disentangle direct and indirect effects of prairie dogs on multiple trophic levels (vegetation, arthropods, and birds) in the Thunder Basin National Grassland. Arthropods did not show any direct response to prairie dog occupation, but multiple bird species and vegetation parameters were directly affected. Surprisingly, the direct impact of prairie dogs on colony-associated avifauna (Horned Lark [Eremophila alpestris] and Mountain Plover [Charadrius montanus]) had greater support than a mediated effect via vegetation structure, indicating that prairie dog disturbance may be greater than the sum of its parts in terms of impacts on localized vegetation structure. Overall, our models point to a combination of direct and indirect impacts of prairie dogs on associated vegetation, arthropods, and avifauna. The variation in these impacts highlights the importance of examining the various impacts of keystone engineers, as well as highlighting the diverse ways that black-tailed prairie dogs are critical for the conservation of associated species.

Local adaptation to precipitation in the perennial grass Elymus elymoides: Trade-offs between growth and drought resistance traits.

Understanding local adaptation to climate is critical for managing ecosystems in the face of climate change. While there have been many provenance studies in trees, less is known about local adaptation in herbaceous species, including the perennial grasses that dominate arid and semiarid rangeland ecosystems. We used a common garden study to quantify variation in growth and drought resistance traits in 99 populations of Elymus elymoides from a broad geographic and climatic range in the western United States. Ecotypes from drier sites produced less biomass and smaller seeds, and had traits associated with greater drought resistance: small leaves with low osmotic potential and high integrated water use efficiency (δ13C). Seasonality also influenced plant traits. Plants from regions with relatively warm, wet summers had large seeds, large leaves, and low δ13C. Irrespective of climate, we also observed trade-offs between biomass production and drought resistance traits. Together, these results suggest that much of the phenotypic variation among E. elymoides ecotypes represents local adaptation to differences in the amount and timing of water availability. In addition, ecotypes that grow rapidly may be less able to persist under dry conditions. Land managers may be able to use this variation to improve restoration success by seeding ecotypes with multiple drought resistance traits in areas with lower precipitation. The future success of this common rangeland species will likely depend on the use of tools such as seed transfer zones to match local variation in growth and drought resistance to predicted climatic conditions.