Understanding the mechanisms that shape ecosystem resistance to increasing livestock grazing pressure, a major driver of land degradation, is essential for predicting its impacts and informing sustainable land management strategies. This issue is particularly relevant in drylands, which host half of the world's livestock production and are highly vulnerable to desertification caused by overgrazing. Here we conduct a standardized field survey across 73 dryland sites in 25 countries to simultaneously evaluate how climatic, edaphic, vegetation and grazing-related factors influence ecosystem resistance-defined here as the capacity to maintain vegetation cover under increasing grazing pressure. We found that increasing grazing pressure reduced vegetation cover in 80% of sites, with an average decline of 35%. Plant species richness emerged as the strongest predictor of ecosystem resistance, with higher richness associated with lower vegetation cover loss. Functional trait data indicated that this positive effect was mainly explained by complementarity in trait values among plants, rather than by functional redundancy. Our results indicate that conserving plant diversity is key to strengthening ecosystem resistance and sustaining dryland functioning under intensifying grazing pressure.

ABSTRACT Drylands worldwide are predominantly used for animal husbandry and provide critical ecosystem services such as forage provision and carbon sequestration. Sustainable livestock grazing management in these ecosystems relies on accurate forage dynamics assessments. Forage dynamics largely result from the characteristics of the plant populations composing the community, the traits of domestic herbivores and the management strategies that influence consumption and selectivity. However, traditional forage evaluations often focus on total productivity, plant cover or biomass measurements, overlooking plant‐size population structures that control forage dynamics as perceived by grazers. We developed species‐specific allometric equations for dominant perennial grasses of the Patagonian steppes to estimate aboveground biomass categories related to the provision of forage and other ecosystem services—including total, green, senescent, standing‐dead and crown biomass—based on plant size. Then, we applied the equations to assess the effects of sheep grazing intensity on population plant‐biomass structures and forage biomass in commercial‐sized paddocks. The non‐destructive allometric method was highly accurate ( R 2 ranged 0.77–0.90; p < 0.001 for green and total biomass models), allowing efficient biomass estimations across different plant morphologies under varying grazing pressures. Notably, grazing intensification did not affect total grass green biomass, but significantly reduced forage biomass by 70% ( p < 0.001) due to the substantial population decline in sheep‐preferred species, highlighting the differential impacts of selective grazing on specific biomass components. By enabling a finer understanding of plant–animal interactions and species‐specific biomass dynamics, our method supports informed management strategies to sustain critical ecosystem services in Patagonian steppes and other drylands grazed by selective herbivores.

1. The functional consequences of life form coexistence in mixed woody-herbaceous communities remain a key question within the biodiversity-ecosystem functioning (BEF) framework. Our aim was to test three models proposed for the relationship between vegetation physiognomy and ecosystem functioning in drylands. Furthermore, these natural systems provide an opportunity to disentangle the ecological mechanisms driving biomass production, which have often been oversimplified by using species richness as a proxy for diversity in randomly assembled artificial communities. 2. We conducted a field-based removal experiment in a Patagonian shrub-grass steppe, creating experimental communities dominated by a single life form (i.e. grasses or shrubs) and comparing them to mixed communities. We estimated green biomass at the peak of the growing season as a proxy for above-ground net primary production (ANPP; gm(-2)year(-1)) in three censuses conducted during the first 5 years after removals. Furthermore, we applied an ecological adaptation of the Price equation to decompose ANPP differences into components that reveal the mechanisms driving the response. 3. We found that (i) ANPP of shrubs decreased when coexisting with grasses; (ii) ANPP of grasses increased when coexisting with shrubs; (iii) total ANPP was maximized at intermediate densities of life forms. 4. Our analysis revealed that life form coexistence maximizes ANPP through a positive effect exerted by shrubs on grasses, rather than through a more exhaustive use of the limiting resource (i.e. water). 5. Synthesis. The relationship between life form dominance and ANPP remains an unresolved question in woody-herbaceous communities. More than 20 years ago, three alternative models were proposed, yet they remain unchallenged. Here, we provide evidence supporting the optimum model, which predicts higher ANPP when herbaceous and woody species co-dominate the community. Although BEF studies have identified niche complementarity as the main mechanism explaining this pattern, our results suggest that facilitation plays a more important role in the functioning of the Patagonian shrub-grass steppe. By incorporating community assembly processes into ecosystem functioning research, we provide empirical evidence that positive interactions are crucial for ecosystem functioning in mixed woody-herbaceous communities.

Woody encroachment threatens the biodiversity, functioning, and ecosystem services of savannas worldwide. Few systematic—experimental—manipulative studies have evaluated the mechanisms that determine demographic transitions of encroaching species across broad environmental gradients. We evaluated cattle and herbaceous vegetation impacts on two widespread tree-encroaching species in Neotropical savannas, Vachellia caven and Vachellia aroma. At two savanna sites with contrasting aridity and productivity, a humid, high-productivity site and a semiarid, low-productivity site, we manipulated cattle grazing (grazed vs. ungrazed) and herbaceous vegetation (present vs. removed) to evaluate their effects on Vachellia seed loss, seedling emergence, survival, growth, and establishment. In the low-productivity site, seed loss (e.g., by granivory) was higher in ungrazed plots with herbaceous vegetation than in all other treatments. Vachellia seedling emergence rate did not vary significantly among treatments or sites. At both sites, cattle grazing decreased final survival (0.33 in ungrazed vs 0.01 grazed plots) and seedling growth in height. Herbaceous vegetation also decreased Vachellia seedling survival (0.004 vs. 0.197). Finally, regardless of the site, Vachellia seedling establishment was higher in ungrazed plots without herbaceous vegetation. Our findings suggest that cattle effects on Vachellia seedlings are consistent among sites with contrasting environmental conditions and could therefore be effectively managed to address woody encroachment in savannas. We developed a conceptual model suggesting that grazing management should consider that increasing stocking rates after seedling emergence may enhance the likelihood of encounters between cattle and tree seedlings. But grazing intensification should not compromise grass regrowth capacity, to maximize tree-grass competitive interactions.

Vegetation patchiness is a key characteristic of drylands and closely linked with their functioning. Deviations from pure power-law distributions of patch sizes have been suggested as indicators of degradation and early warning signals of approaching desertification thresholds. The general objective of our study is to evaluate the usefulness of patch-size distributions as indicators of vegetation degradation caused by grazing and drought. To achieve this, we utilized field data and data generated by individual-based simulation models to examine how vegetation patchiness changes in Patagonian drylands under different rainfall scenarios and grazing conditions. First, we compared patch-size distributions predicted by two models for semi-arid grass steppes and for arid shrub steppes under current rainfall conditions with field data from two grazing scenarios (ungrazed and overgrazed). Second, we used the models to simulate the long-term effects of grazing on vegetation patchiness under both drier and wetter conditions. Our field data revealed that the impact of grazing depends on aridity, with stronger effects observed in the semi-arid steppes than in the arid steppes. Our simulation results indicate that changes in rainfall regimes have a stronger effect on patch-size distributions than grazing. Wetter conditions increased the number of medium-to-large-sized patches in both steppes, regardless of grazing. Deviations from power-law distributions were not directly related to overgrazing and degradation; however, grazing and changes in rainfall caused systematic changes in the parameters of the power-law patch-size distributions, supporting their usefulness as one indicator in the assessment of degradation.

Patch dynamics in ecology have been useful for understanding the functioning of diverse biomes, focusing primarily on how processes maintain the vegetation mosaic. However, less is known about how the mosaic structure influences processes within each patch, completing the pattern–process feedback. We present a conceptual framework exploring the importance of neighborhood structure and composition on patch dynamics. We described the neighborhood for each focal patch based on the neighbor's distance (Voronoi area), the size of the neighbors, and the neighborhood species composition. We tested this model in grazed and ungrazed Patagonian steppes, where mosaic structure and patch dynamics are associated to shrub life stages. Using field data spanning 26 years, we evaluated how mosaic attributes influence the establishment, growth and mortality of long‐lived dominant shrub species, while considering climate variability and the effect of grazing by domestic livestock. Our results show that larger neighborhood areas and lower intraspecific ratios reduced shrub establishment and growth, while increasing mortality. Therefore, the neighborhood structure and composition alter patch dynamics. Grazing had species‐specific effects on establishment and growth but did not alter the relationship between neighborhood variables and focal shrubs. Our results demonstrate that shrub population dynamics – and thus patch dynamics – are influenced by the structure and composition of the vegetation mosaic. This long‐term study enhances our understanding of the mosaic maintenance feedback involving patterns and processes in ecosystems where woody species play a significant role in its dynamics, and in a context of increasing environmental pressures.

Understanding how environmental changes impact arthropod communities requires multi-scale studies. Despite their importance, research exploring arthropod diversity across regional gradients while accounting for local factors, such as vegetation structure, remains scarce. This dual approach provides deeper insights into the influence of regional processes on local community dynamics. We employed pitfall traps to investigate diversity and composition changes in ground-dwelling beetles across a latitudinal gradient and local vegetation mosaic within the Occidental District of Patagonia, Argentina. Five sites were selected along the 850 km geographical gradient to capture regional diversity, while within each site, 10 traps were set in shrub-covered areas and 10 in bare-ground areas to assess local variability. The northernmost sites exhibited the highest diversity and abundance, and distinctive species compositions. Species composition differed significantly between sites north and south of latitude 42º, with Carabidae and Tenebrionidae as key families representing this variation. Local microsite characteristics, particularly shrub presence, significantly shape beetle communities, sometimes surpassing the influence of latitudinal changes. Both, regional and local (vegetation mosaic) factors control community composition. The differences in species composition around latitude 42º mirrors patterns seen in other taxa, indicating that historical processes, such as biogeographical processes, substantially affect beetle communities at a regional scale. Our findings underscore the critical need to conserve structurally diverse habitats, such as shrub-grassland mosaics and native tussock grass areas, to support key beetle assemblages and ecosystem functions. Protecting these microhabitats and recognizing endemic beetles as indicators can strengthen conservation planning and promote ecological resilience under increasing climate and land-use pressures.

Sexual reproduction, growth, and survival are crucial demographic strategies for plant population viability. Here, we propose a conceptual model predicting demographic responses of species based on their ecological strategy and the heterogeneity of environmental conditions within a biogeographical unit and then applied it to a case study from a 5˚ latitudinal gradient in the Patagonian steppes. We also aim to disentangle genetic from environmental effects on demographic responses. We performed in situ and common garden experiments with two species from six local populations of the Occidental Phytogeographical District of the Patagonian steppes. Species differ in key ecological traits, and thus fit into Grime's model for evolutionary strategies in plants: one as competitive species and the other as stress‐tolerant species. We calculated population growth rate (λ) and performed elasticity analyses to compare the contribution of each demographic strategy to population fitness between species and among local populations distributed along 600 km latitudinal gradient with differences in mean annual precipitation (MAP). We highlight four results. First, the competitive species change from sexual reproduction to growth as MAP increases. Second, the stress‐tolerant species relied on growth and survival along the MAP gradient. Third, interannual variation in resource availability modulated demographic responses for both strategies. Fourth, based on the comparison of the in situ and common garden experiments, we submit that demographic responses were genetically driven. Our study shows that demographic responses can be roughly predicted by the ecological strategy across environmental gradients. We show that differences arise not only between species, but also were genetically driven differences within species among local populations. Scaling up plant‐level responses to population‐level dynamics allows for a process‐based understanding of current and future biogeographical species organization. Furthermore, conservation and restoration efforts should be guided by demographic strategies underlying population viability.

In a recent Tansley Review, Holdo and Nippert (2023) provided a comprehensive analysis of existing models that explain tree-grass coexistence along precipitation gradients in savannas. They proposed three key elements that a definitive savanna model should include and predicted the upper boundary pattern of tree cover based on the ‘Sankaran curve’ depending on soil texture. We tested this boundary pattern of tree cover in drylands using global field data (98 sites from 25 countries and six continents) accounting from a wide range of environmental conditions and grazing pressures. We found that the upper boundary on tree cover in global drylands increases with mean annual precipitation before reaching a threshold of 453 mm. Increasing grazing pressure significantly decreased this upper limit. However, the upper boundary on tree cover did not stabilize but instead decreased beyond this precipitation threshold. Our findings partially support the three coexistence mechanisms proposed by Holdo and Nippert and indicate some features that may be specific for water-limited rangelands: a lower precipitation threshold for the upper boundary on tree cover, a declining tree cover beyond the threshold, and an interplay between resource- and disturbance-based mechanisms. Overall, our findings advance our understanding of tree-grass coexistence in drylands and provide new elements to develop a definitive model of such coexistence across terrestrial ecosystems.

Increases in the abundance of woody species have been reported to affect the provisioning of ecosystem services in drylands worldwide. However, it is virtually unknown how multiple biotic and abiotic drivers, such as climate, grazing, and fire, interact to determine woody dominance across global drylands. We conducted a standardized field survey in 304 plots across 25 countries to assess how climatic features, soil properties, grazing, and fire affect woody dominance in dryland rangelands. Precipitation, temperature, and grazing were key determinants of tree and shrub dominance. The effects of grazing were determined not solely by grazing pressure but also by the dominant livestock species. Interactions between soil, climate, and grazing and differences in responses to these factors between trees and shrubs were key to understanding changes in woody dominance. Our findings suggest that projected changes in climate and grazing pressure may increase woody dominance in drylands, altering their structure and functioning.

Vegetation patchiness is a distinctive feature of arid ecosystems that strongly shapes plant-soil interactions. While fertility islands are recognized as a critical plant legacy influencing ecosystem functioning, it is not clear whether there is a legacy associated with fungi symbionts in soils. We compared the legacies of arbuscular mycorrhizal fungi (AMF) and dark septate endophyte (DSE) inoculum potential in soil, to determine if they respond to the plant functional type (PFT, grasses and shrubs) and if they overlap spatially with the nutrient legacies. We estimated the soil inoculum potential of AMF and DSE and the soil organic carbon (C), phosphorus (P), and nitrogen (N) concentrations in the soil microsites occupied by the dominant grass and shrub species, and in bare soil microsites distributed in a 6-hectare grazing exclosure in the Patagonian steppe. The AMF inoculum potential was higher in soil microsites conditioned by grasses than in microsites conditioned by shrubs. Instead, the DSE inoculum potential did not differ among microsites and was higher than that of AMF. Furthermore, C and P concentrations were higher in microsites conditioned by shrubs, contrary to AMF inoculum potential. We found that grasses and shrubs have decoupled soil legacies that may alleviate nutrient limitation. While shrubs increase nutrient concentrations, grasses increase AMF inoculum potential. Moreover, the homogeneous distribution of DSE inoculum potential highlights the need to further investigate the role of this group of fungi and its interactions with AMF and soil nutrients in the plant-soil interface.

Diet selectivity by domestic herbivores controls plant community structure and dynamics and may induce rangeland degradation, particularly in drylands. However, management decisions frequently ignore herbivore selectivity. Here, we studied how grass morphology controls sheep selectivity for individual plants, and how this selectivity interacts with grazing intensity to determine population plant-size distributions and the forage supply.In Patagonian steppes, we manipulated the plant morphology (size and standing-dead proportion) of three dominant grass species differentially preferred by sheep for 4 years. Then, we evaluated how these morphological alterations affected intra- and inter-specific preference patterns. We also evaluated how grazing intensity (ungrazed, moderate grazing and intensive grazing) affected the plant-size distribution of the three species, the forage supply, and its accessibility.For the highly preferred species, herbivores selected plants that were either naturally or experimentally short, with low proportion of standing-dead biomass. In contrast, morphological changes did not alter the within-species selectivity of the least preferred species.Grazing intensity strongly changed the population plant-size distribution of preferred species in ways that resembled the experimental manipulations of morphology. Moderate grazing showed the greatest morphological heterogeneity among individuals. When integrating the green biomass of forage species' individuals at ecosystem level, we found that the forage supply was the highest in ungrazed sites and decreased as grazing intensity increased. However, considering the dissuading effect of the standing-dead proportion of plants, the accessible forage was the highest under moderate grazing.Synthesis and applications. Our findings (i) showed that, within preferred species, sheep selectivity at individual-plant level is controlled by morphological characteristics that determine accessibility to green high-quality biomass. This effect was as important as that of species identity; (ii) empirically proved plant-animal positive feedback at individual level; and (iii) revealed how the individual-plant selectivity scales-up at population level and controls the forage supply, but also its accessibility. Our complementary approach generates critical knowledge for developing management practices to control key forage species defoliation and to adjust the grazing pressure to the offer of accessible forage, avoiding the common carrying capacity overestimation. These aspects are essential for sustainable production in grazed drylands.

Global drylands are threatened by grazing pressure intensification and climate change, which act as major drivers of land degradation. Detecting this process at an early stage is essential for predicting losses of ecological functions and for restoration management. Vegetation patch-size distribution is an indicator of dryland multifunctionality and has been proposed as a warning signal for the onset of degradation processes. However, we proposed and tested a general model that stresses that patchiness may fail to detect degradation of the forage provision, depending on plant community species composition. This is a key aspect since forage provision is strongly associated with human well-being in drylands. We hypothesized that grazing-induced changes in patchiness and forage provision converge in drylands dominated by forage species but are decoupled in those dominated or co-dominated by non-forage species. We tested the conceptual model in a unique regional-scale gradient with strong ecological differences but a common biogeographical and human impact history to reduce local contingencies effects. We compared datasets of grazing intensification impacts on (i) plant cover and patch-size distribution and (ii) plant density and plant-size distribution of dominant forage grasses (a proxy of forage provisioning). We showed that there is a decoupling between grazing-induced changes in vegetation patchiness and forage provisioning, particularly in drylands where non-forage species are dominant. In these drylands, plant cover and patch-size distribution were slightly affected by grazing intensification, whereas plant density of forage species was decimated and their plant-size distributions were strongly skewed towards small sizes. Synthesis. Our dryland conceptual model suggests that global change impacts on forage species populations can be detected even before changes in patch-size distribution and plant cover. Our findings support the model and indicate that the population status (plant density and plant-size distribution) of forage species allows for predicting forage dynamics and is useful to the early detection of losses of ecosystem services linked to human well-being in drylands.

Livestock rearing is increasing in savannas, so the ability of trees to regrow after consumption represents one of the critical aspects of savanna structure and functioning. Here, we identified specific traits—which may explain the mechanisms behind defoliation tolerance—in saplings of two encroacher tree species (Vachellia caven and Vachellia aroma). We carried out common garden and field experiments where manual defoliation treatments simulated the cattle grazing regime employed in the study region. At the end of the experiments, we recorded growth variables and, in the common garden experiment, root reserves concentration. In the common garden, defoliation decreased height and basal diameter growth but did not affect the aerial relative growth rate. Also, defoliation increased the number of branches per plant and decreased root relative growth rate. Starch concentration was higher or similar in defoliated plants than controls, depending on the species. In the field, defoliation decreased both species’ height and basal diameter growth. We found that tree saplings tolerated defoliation, and the mechanisms behind tolerance would be linked mainly to the axillary bud activation and not to storage reserve mobilization. Over time, these plant architectural changes might complicate cattle movement and management in these systems. From a long-term perspective, livestock rearing might retard juvenile recruitment into reproductive-size classes (e.g., adults) by reducing overall tree growth rates.

Abstract Woody encroachment in savannas is a complex and global phenomenon that has negative impacts on the forage availability and livestock yield. We examined the impact of cattle grazing and herbaceous vegetation on Vachellia caven and Vachellia aroma, the main encroacher tree species in Neotropical savannas, where livestock production is the principal activity. Our experiments were conducted across a gradient of aridity and productivity (low-, medium-, and high-productivity sites), covering the central distribution of savannas in Argentina. We carried out manipulative experiments with cattle grazing (grazed–ungrazed) and herbaceous vegetation (with–without) to examine Vachellia seed loss (e.g. granivory), seedling emergence, survival, growth, and establishment. In the medium-productivity site, seed loss was higher under ungrazed conditions with herbaceous vegetation. Across all sites, cattle grazing decreased the final seedling survival regardless of the presence of the herbaceous vegetation. Herbaceous vegetation increased tree seedling survival in the medium-productivity site but decreased it in the low- and the high-productivity sites. Overall, the effect of grazing on tree establishment was neutral in the medium-productivity site as a consequence of increasing seed availability and decreasing seedling survival. However, seedling establishment was higher under ungrazed conditions and without herbaceous vegetation in the high- and low-productivity sites, because of the negative effects of grazing and herbaceous vegetation on seedling survival. From this demographic approach, we identified an ecological window during which grazing could be effectively managed to control the early stages of woody encroachment if stocking is adjusted and synchronized with tree species life history and site productivity.

Questions: (1) Is the co-dominance of grasses and shrubs a general feature of Patagonian steppes? (2) At a local scale, does it change with domestic grazing intensity or decadal changes in precipitation?Locations: Question 1. Patagonia region in southern South America. Question 2. Shrub-grass steppes (45 & DEG;24 & PRIME;S, 70 & DEG;17 & PRIME;W) subjected to different grazing intensity for decades.Methods: We sampled vegetation to estimate perennial grass and shrub cover at two scales: regional and local. At the regional (phytogeographical) scale, we sampled grazed steppes under different climates and soils (n = 324 sites over ca. 487,000 km(2)). At the local scale, we sampled paddocks subjected to known grazing management, from commercial intensive densities of sheep to total (experimental) exclusion. Also, in the same site, we sampled permanent plots (25 years old) in a grazing exclosure to study the effects of precipitation variation.Results: Mixed communities are the predominant form of community organization in the Patagonia region. At the regional scale, shrub cover was negatively related to grass cover and communities where shrubs had higher cover than grasses represent 54% of the data set. At the local scale for an arid steppe, perennial grasses had higher cover than shrubs, regardless of grazing pressure. The shrub-to-grass ratio was increased only by intensive grazing. Permanent plots in a grazing-excluded paddock indicate that decadal xeric or humid climate periods decrease or increase total cover, respectively. Surprisingly, the relative shrub-to-grass ratio did not change significantly.Conclusions: Although perennial grasses and shrubs cover are negatively correlated, they coexist in the Patagonian steppes under a wide range of climate, soil and grazing conditions. At a local scale, selective herbivory promotes species replacement but did not substantially change the life-form covers. Water shortages, as well as periods of water surplus, similarly affect both life forms without compromising their coexistence.

Woody encroachment in savannas represents an ecological process of current global interest given its negative impact on ecosystem functioning, particularly on forage production. Traditional savanna models propose competition and niche differentiation as the main mechanisms allowing tree-grass coexistence. Demographic models, instead, propose abiotic and biotic factors as bottlenecks controlling vital rates and transitions from seeds to adult trees. The role played by domestic grazing on woody encroachment is yet controversial. Here, using a multistage tree life approach, we combine both models and evaluate the role of grazing and herbaceous vegetation on woody recruitment in a Neotropical savanna dominated by Vachellia caven, a successful and widely spread encroacher tree species. We performed three experiments to evaluate seed predation, seedling emergence and survival of V. caven by manipulating cattle grazing (grazed and ungrazed areas) and herbaceous vegetation presence (vegetated and unvegetated). Finally, we combined the results of the three experiments to estimate the probability of plant recruitment across these experimental factors. Grazing decreased seed predation by half, did not modify seedling emergence and decreased seedling survival. Herbaceous vegetation did not affect seed predation nor seedling emergence rate, but increased seedling survival. Overall, the net effect of grazing on V. caven recruitment was neutral since the increase in seed availability due to the reduction in seed predation rate was compensated by the negative effect of grazing on seedling survival. Our analysis revealed that cattle grazing and herbaceous vegetation had contrasting effects on the seed and seedling life stages. We propose that in order to restrain the early stages of encroachment, cattle grazing pressure could be managed following the seasonality of demographic tree transitions. Through rotational grazing amongst paddocks, stocking rates could be relaxed during the primary dispersal stage to maximize granivory, and then increased to enhance the chance of seedling consumption and trampling.

Questions Arid communities are strongly limited by soil resources including water and nitrogen (N). Plants compete for N with other plants and microorganisms, which are also limited by carbon (C). We propose that above- and below-ground plant responses to soil resources are modulated by community structure (species relative abundances, "mass ratio hypothesis") and species traits (relative growth rates - RGRs). We evaluated the single and combined effects of soil N and C addition on the above- and below-ground biomass accumulation of perennial grass patches in an arid community, and the mechanisms involved in their responses. Location Patagonian steppe, Argentina. Methods We added N (2 g N m(-2); NH4NO3) and C (330 g C m(-2); sucrose) to 1-m(2) field plots in a factorial design. After two years, we harvested above-ground (n = 5 plots) and below-ground biomass (n = 10 soil cores) and sorted it by species. We measured potential soil respiration as a proxy of microbial activity. Results Total above-ground biomass increased by 55% as a result of N and decreased by 45% as a result of C addition, in relation to controls. C addition reduced total below-ground biomass by 42%. The above-ground differences were associated with changes in the biomass of dominant species according to their RGRs. Poa ligularis (dominant, high RGR) increased by 92% as a result of N addition while Pappostipa speciosa (dominant, low RGR) decreased by 55% as a result of C addition. Intermediate and subordinate grasses did not modify their biomass, independently of their RGR. Potential soil respiration was three times higher in plots with C addition than in control plots. Conclusions Community biomass was explained by a combination of mass ratio hypothesis and specific RGR, as dominant grasses controlled above-ground community responses to N (high-RGR species) and C addition (low-RGR species). Our findings highlight the independence between the above- and below-ground processes and the importance of considering community equitability and species characteristics to predict plant community responses to changes in soil resources.

Aims Soil processes in arid ecosystems are strongly controlled by resource scarcity. Grazing intensification can induce changes in ecosystem processes through multiple pathways, adding new constraints to those of local conditions. We focus on grazing-induced changes in litter traits and soil environment that may affect litter decomposition and N dynamics in temperate grass-shrub steppes. Methods We performed three litterbag decomposition experiments to evaluate: i) the effect of litter traits in a common garden (ex situ), ii) the effect of soil environment using a foreign common litter substrate (in situ), and iii) the interactive effects of litter traits and soil environment (grazed vs. exclosure communities, reciprocal transplants in situ). Field experiments were replicated in three blocks with paired plots under a long-term exclosure (> 25 years) and under year-round sheep grazing. Local litter included mixtures of species of grasses and shrubs, separately. Results Grazing exclusion did not alter litter decomposition rates, either through changes in litter traits or in soil environment. Nevertheless, N released during grass litter decomposition was 286% higher in exclosures than in grazing communities. The difference was associated to changes in litter C:N ratio. The effects were maintained when results were integrated to the entire litter community. Conclusion Our study suggests that litter decomposition rates in arid steppes are strongly controlled by local drivers. Ungulate grazing does not have an important influence on litter mass loss, but it can exert a strong control on N flux during decomposition, by changing grass litter traits.