Journal of Ecology

1. Climate change increases the frequency and severity of drought events with strong repercussions on grassland ecosystems. While the effects of single drought events on ecosystem structure and functioning are well understood, it is largely unknown whether and how drought frequency modifies ecosystem responses to drought. 2. Here, we assessed how the increase in frequency of severe, annual summer drought impacted grassland communities. We examined these effects in a species-rich sub-alpine mountain meadow with a drought frequency of one, three, and 13 years, as well as ambient conditions. 3. We found that high drought frequency increased seasonal plant community synchrony through a reduction in species richness, a shift of plant functional groups, a loss of early-seasonal plant species, and the constrained establishment of seedlings throughout the growing season. These changes were associated with a decreased fraction of biomass as drought frequency increased. 4. Furthermore, we show that negative drought effects were enhanced with an increasing drought frequency, and that negative drought effects on plant communities outweighed the weak adaptive effects of species. 5. Synthesis. We conclude that single and low-frequency drought studies may not adequately predict longer-term changes in our rapidly shifting climate. With the ongoing increase in drought frequency due to climate change, we predict that grassland plant communities will increase in seasonal synchrony. We suggest that this increase in synchrony will leave ecosystems highly vulnerable to future disturbances, because asynchrony is a critical component of stability. Moreover, given the weak adaptive effects of plant species to long-term recurrent drought, we conclude that plant communities are unlikely to be able to adapt to the rapid increase in recurrent drought events.

Climate change is known to negatively impact tropical forests; yet how climate change impacts tree community persistence at local scales remains less clear. Using data from a long-term tropical forest census plot over 25 years, we constructed plant- plant interaction networks based on tree growth. We then quantified community persistence as feasibility domain of constituent species using recently developed frameworks of structural stability. We found a decrease in structural stability under warming and precipitation changes over time as evidenced by both direct environmental effects and indirect effects via network structure; and indirect effects were stronger than direct effects. Among these structures, facilitation:competition degree ratio and competitive transitivity were most positively related with structural stability. Our study highlights how the lens of plant-plant interaction networks can identify novel details about risk to tropical forest diversity under climate change at local scales. Insights from this work will be helpful for aligning forest management activities with areas under the greatest risk of species loss.

Flowering synchrony within a community may be facilitated by climatic factors and by ecological interactions that promote shared pollination services. In contrast, flowering asynchrony is promoted when there is inter-species competition for pollinators. In this multiyear study, we analyse the flowering phenology of a seasonal, herbaceous community (Kaas plateau) in Western Ghats, India to identify environmental constraints that may influence flowering synchrony. We addressed the questions: (a) Is flowering seasonality correlated with climatic factors? (b) Is there evidence for flowering synchrony within the community? (c) Do plant-pollinator interactions shift with flowering phenology? In Kaas, we recorded flowering phenology of 76 herbaceous species and found that climatic factors influenced their flowering phenology. We also identified the community to be composed of a few mass flowering (MF) species ([~]30%) and several non-mass flowering (nMF) species ([~]70%). Using two novel synchrony indices, temporal overlap (SItemp) and synchrony in abundance (SIabd), we also identified higher asynchronous flowering within the community than expected. Notably, species sharing the same floral colour, showed marked absence of synchrony, thus suggesting that competition and not pollinator-mediated facilitation drives flowering asynchrony within Kaas. Finally, pollination networks were observed to shift with flowering abundances within the community. Our findings reveal that even seasonal landscapes like the laterite plateaus, despite their short flowering season that last only 4-5 months, exhibit an overall asynchronous flowering phenology. And, the synchronous flowering for which the Kaas plateau is famous, was noted to be mostly due to a few mass flowering species that alter across years.

Background and aimsIntraspecific variation in functional traits is essential for the evolutionary success of organisms. The co-variation between trait variation and environment, as well as between different traits, can help us to understand which ecological factors drive habitat adaptation, and to what extent adaptation may be constrained by trait correlations and trade-offs. In managed grasslands, plants experience a combination of competition, recurrent biomass removal and nutrient pulses. Each of these ecological challenges requires specific plant tolerances, and populations should locally adapt if intraspecific variation exists in these traits. MethodsWe studied variation in land use-related traits in the common grassland plant Plantago lanceolata. In a common environment, we quantified the competitive ability (R*), clipping tolerance and responses to a nitrogen pulse of plants from 54 populations with different land use intensities across Germany. Key resultsWe found significant population differentiation in competitive ability but there was little evidence that trait variation was related to land use intensity. There was a positive relationship between competitive ability and clipping tolerance at the population level, indicating a genetic, and possibly functional, link between these two traits. In contrast, clipping tolerance and nitrogen responses were negatively correlated at the levels of plant individuals, indicating a physiological trade-off between plant responses to these two land-use processes. ConclusionsOur results show that there is substantial intraspecific variation in some of the key functional traits for plant success in managed grasslands, and that rapid evolution and adaptation is therefore possible in these traits.

O_LIGrasslands are facing a rapid decline worldwide. Among the primary threats to these ecosystems are changes in land use, such as land abandonment and forestation, which promote forest vegetation in detriment of grassland plant diversity. To support the conservation and restoration of grasslands, it is key to understand what ecological processes limit the recovery of their biodiversity and functions after perturbations. However, we still know little about the legacy effects of forestation on the ecological mechanisms involved in the recovery of grasslands, especially concerning long-lasting impacts on plant-pollinator interaction networks and plant reproduction. C_LIO_LIHere, we aim to fill this gap of knowledge by comparing the plant and pollinator diversity, the degree of network generalization, the pollination success and pollen limitation of native plant species in 30 plant-pollinator networks of old-growth and restored grasslands of different ages since recovery (from 2 to 84 years). We hypothesized that past forestation would exert long-lasting legacy negative effects on plant richness and plant-pollinator networks, increasing pollen limitation for native plants, thereby delaying community recovery in restored grasslands. C_LIO_LIResults showed that restored grasslands exhibited significantly lower plant richness, less specialized (more generalized) interaction networks, lower pollination success and pollen-limited seed reproduction of native plants compared to old-growth grasslands. Meanwhile, the degree of network specialization and pollination success gradually increased with time after grassland restoration initiated. Overall, network generalization, which was caused by low plant richness, reduced pollination and reproduction success in native grassland plants, and degraded pollination networks and functions could recover in restored grasslands with continuous management. These findings imply that plant diversity restoration was slow because of the negative feedbacks associated with low plant richness and consequently, generalized plant-pollinator interaction networks, which diminished native plant reproduction in restored grasslands. C_LIO_LISynthesis and applications Our findings suggest that the recovery of specialized plant-pollinator networks by enhancing plant diversity is essential for restoring pollination function. For quicker grassland restoration, it may be effective to facilitate the establishment of highly specialized pollination networks by seeding or planting diverse native plants collected from neighbouring areas while avoiding genetic contamination. C_LI

Grasslands face an uncertain future due to climate change. Although there is increased interest in the interdependencies of different biodiversity components, the effects of climate change on these relationships remain understudied. One of these is the richness-evenness relationship (RER), which is sensitive to altered species abundances in relation to richness. This relationship may be important as evenness and richness jointly shape diverse ecosystem functions, such as stability and productivity. As evenness affects productivity differently in low and high richness communities, the richness-evenness relationship is important to investigate, especially under climate change. Here, we assess the effects of increased CO2 concentrations, temperature, and drought on the RER in a subalpine long-term (2010 - ongoing) grassland climate change experiment, and test whether these effects can be buffered by reseeding. We provide evidence that climate change alters the RER in our experiment, and that these changes occur independently of changes in richness and evenness separately. Reseeding erases the differences in RER between treatments and controls but fails to restore the negative RER initially found in controls. Further, we show that the dominant grass species in our system (Arrhenatherum elatius) responds differently to each climate change factor, with opposite effects in high vs. low richness plots, thereby largely determining the direction of the RER. These results suggest that the RER can reveal additional insights on community responses to climate change and represents a different signal than evenness or richness alone. A more nuanced approach integrating evenness and maximizing richness in seed mixtures could be an important step forward to better match restoration treatments to particular community types and global change drivers.

Invasion by alien plants is frequently attributed to increased resource availabilities. Still, our understanding is mainly based on effects of single resources despite the fact that plants rely on multiple resources. How multiple resources affect success of alien plants remains largely unexplored. Here, with two common garden experiments, one in China and one in Germany, we tested whether nutrient and light availabilities affected the competitive outcomes between alien and native plants. We found that under low resource availabilities or with addition of only one type of resource aliens were not more competitive than natives. However, with a joint increase of nutrients and light intensity, aliens outcompeted natives. Our finding indicates that addition of multiple resources could greatly reduce the number of limiting factors (i.e. niche dimensionality), and that this favors the dominance of alien species. It also indicates that habitats experiencing multiple global changes might be more vulnerable to plant invasion.

Plant phenology controls resource acquisition, survival and reproductive success, yet it is still commonly reduced to a limited set of calendar-based metrics, such as discrete dates of budburst or senescence. By contrast, chronobiology quantifies biological rhythms from full activity profiles using metrics that capture timing, regularity and amplitude of rest-activity cycles. Here, we bridge these two perspectives by developing a pheno-clock framework that translates plant phenology into chrono-ecological properties. We applied actigraphy-inspired metrics to multi-year satellite time series of European beech (Fagus sylvatica) forests across their European range. From daily photothermal activity profiles, we derived indices describing the strength, fragmentation and amplitude of annual rest-activity rhythms and related them to classical phenological metrics and regional climate. Our results reveal a marked asymmetry between spring and autumn phenology, indicating that canopy decline is governed by the cumulative organization of annual energy input, whereas canopy activation is dominated by short-term forcing. Across biogeographical regions, beech forests segregate into distinct pheno-chronotypes that differ in the timing and consolidation of rest and activity phases rather than in growing-season length alone. These chrono-ecological patterns suggest that climate filters not only when forests grow, but also how they structure their annual rhythmicity. By importing chronobiology into plant phenology, the pheno-clock framework provides a transferable approach to describe and compare seasonal strategies in plants, opening new avenues to link phenological diversity, functional traits and ecosystem responses to environmental change.

Flowering time underpins plant fitness, species coexistence, and ecosystem functioning. While global warming consistently advances flowering, the influence of water availability remains unclear. We hypothesized that this inconsistency reflects the overlooked timing of drought. In 200 experimental Mediterranean annual-plant communities, we imposed early-, mid-, and late-season dry periods and grew plants in monocultures and mixtures to disentangle physiological and competition-mediated responses. Early and late droughts shortened flowering duration: early drought delayed onset, late drought advanced termination. Some shifts were direct, others emerged through competition. A new community-level index revealed greater phenological segregation in mixtures, showing that plasticity alone can generate niche separation under competition. Both early and late droughts further enhanced this segregation. Together, our results demonstrate that the seasonal timing of drought governs flowering responses through both direct physiological pathways and indirect biotic interactions, emphasizing rainfall seasonality as a key driver of ecological responses to climate change. Impact statementSince flowering time is crucial to ecosystem functioning, shifts in the timing of drought could have far-reaching effects on the performance and resilience of ecological communities. Our study shows that droughts effect on flowering is dependent on their timing and influenced by interactions with neighboring species. This means that understanding how plants respond to changing water conditions requires looking beyond single species, considering both seasonality and community interactions.

O_LIMegaherbivores are known to strongly influence multiple ecological processes, but their bottom-up impacts on vegetation via nutrient redistribution remain poorly understood, particularly in mesic ecosystems. Here, we investigated how woody plant communities are influenced by large nutrient inputs in the form of dung deposited by Asias largest herbivore, the Asian elephant (Elephas maximus), in the tropical forests of southern India. C_LIO_LIWe conducted field and mesocosm experiments on woody saplings to examine three mechanisms through which dung deposition by elephants can alter plant community assembly. Specifically, we tested if elephant dung input 1) creates hotspots of plant growth, and shapes plant communities by altering 2) the negative density-dependent effects of neighborhood competition, and 3) interference competition between plant functional types differing in nutrient limitation, namely nitrogen-fixers and non-nitrogen-fixers. C_LIO_LIOur findings show that dung deposition by elephants can generate fine-scale spatial differences in woody sapling communities by creating local growth hotspots and altering competitive interactions. We analyzed relative growth rate and final sapling size in the field experiment, and found that average-sized and large saplings receiving dung inputs were buffered against the negative density-dependent effects of neighborhood competition. In the mesocosm experiment, non-nitrogen-fixing species (which are nitrogen-limited) outcompeted nitrogen-fixers in accumulating biomass when supplied with elephant dung. These outcomes were associated with changes in their relative competitive strength which was stronger for non-nitrogen fixers under dung treatment and for nitrogen-fixers under control. Such bottom-up effects on plant growth and competition can be of substantially large magnitude, as we estimated that elephants in these forests create a total of 11000 such nutrient-rich sites /km2/year, with each elephant redistributing [~]130 kg Nitrogen/year through this pathway and each site receiving [~]20 g Nitrogen. C_LI SynthesisOur findings on the outcomes of sapling competition highlight the role of nutrient redistribution by megaherbivores as an underappreciated driver of species interactions that can alter plant communities at fine scales, effects that are widespread across megaherbivore habitats. Such bottom-up effects of megaherbivores, along with their top-down effects, have important conservation implications and can help in restoring species interactions and spatial heterogeneity in plant communities in defaunated habitats.

Interannual variability in grassland primary production is strongly driven by precipitation, nutrient availability and herbivory, but there is no general consensus on the mechanisms linking these variables. If grassland biomass is limited by the single most limiting resource at a given time, then we expect that nutrient addition will not affect biomass production at arid sites. We conducted a distributed experiment manipulating nutrients and herbivores at 44 grassland sites in 8 regions around the world, spanning a broad range in aridity. We estimated the effects of 5-11 years of nutrient addition and herbivore exclusion treatments on precipitation sensitivity of biomass (proportional change in biomass relative to proportional change in rainfall among years), and the biomass in the driest year (to measure treatment effects when water was most limiting) at each site. Grazer exclusion did not interact with nutrients to influence driest year biomass or sensitivity. Nutrient addition increased driest year biomass by 74% and sensitivity by 0.12 (proportional units), and that effect did not change across the range of aridity spanned by our sites. Grazer exclusion did not interact with nutrients to influence sensitivity or driest year biomass. At almost half of our sites, the previous year's rainfall explained as much variation in biomass as current year precipitation. Overall, our distributed fertilization experiment detected co-limitation between nutrients and water governing grasslands, with biomass sensitivity to precipitation being limited by nutrient availability irrespective of site aridity and herbivory. Our findings refute the classical ideas that grassland plant performance is limited by the single most limiting resource at a site. This suggests that nutrient eutrophication will destabilize grassland ecosystems through increased sensitivity to precipitation variation.

Competition among plants can lead to allocation strategies that favor individual competitiveness at the expense of group-level productivity. However, the role of root and shoot responses in driving these outcomes is intensely debated. Experimental approaches often have difficulty disentangling above- and belowground interactions due to the confounding effects of pot size and nutrient distribution. Here, we used physical dividers and varied inter-plant distances (3-24 cm) in soybean to isolate competitive interactions while controlling for these confounding effects. Simultaneous above- and belowground competition increased relative root allocation, but reduced total biomass. Aboveground competition alone had stronger effects, triggering shade avoidance and reducing both shoot and root biomass. This likely overrode belowground cues. These results underscore the significance of neighbor-induced responses, particularly under full competition, as pivotal drivers of allocation patterns and promising targets for breeding strategies that enhance collective crop performance.

Timing of flowering is shifting with climate change. Although climate-driven shifts in phenology sometimes affect seed production, whether changing phenology will scale up to affect population dynamics of long-lived plants remains largely unknown, particularly under changing precipitation. Understanding how phenology affects persistence and extinction risk is a pressing need given contemporary biodiversity loss. We combined nearly a decade of demographic censuses and a four-year phenological survey in a rainfall manipulation experiment to examine the effects of experimental drought and irrigation on flowering phenology, vital rates (e.g., survival and individual growth), and population growth in the perennial herb Lomatium utriculatum. We found that drought advanced flowering by 3.3 days on average, and that earlier-flowering plants produced more seeds regardless of treatment. However, both rainfall treatments reduced seed production compared to controls. We quantified the phenology-mediated and direct, non-phenological effects of rainfall manipulation on population growth rates using integral projection models and a life table response experiment. Drought and irrigation increased {lambda} through increased individual growth, but these effects were partially negated by treatment-driven declines in seed output. In contrast, changes to seed production resulting from shifting flowering times had negligible effects on population growth. Our results suggest that climate-driven phenological shifts may only marginally impact population dynamics in perennial plants and highlight that assessing phenologys consequences for persistence under climate change must also account for direct demographic effects of the climate driver(s) themselves. SignificanceWill changing flowering times under climate change increase extinction risk in plant populations? Despite well-documented earlier flowering and its influence on the number of offspring produced, how changing flowering times will affect population growth or decline is still mostly unknown. We study this in a perennial wildflower subject to changes in rainfall. While we found that drought meant earlier flowering and that, all else equal, early flowering meant more seeds, these effects only marginally affected population growth. Instead, population growth was influenced mostly by rainfall-driven changes to individual plant growth. While shifting flowering times remain an important indicator of climate change, assessing extirpation in plants requires considering flowering times as only one of many life cycle processes changing with climate.

Plants introduced to degraded ecosystems in the framework of ecosystem restoration are commonly challenged by novel environmental conditions. Consequently, plant functional traits can differ between restored and reference sites, even within individual species. Studies on such intraspecific variation mainly focused on vegetative traits, while timing of life history events, phenology, received less attention so far. To address this gap, we focused on reproductive phenology of 16 flowering plant species and compared it between 47 restored meadows and 16 reference meadows in the same region. We found that plants in restored meadows flowered on average two days earlier compared to the reference, semi-natural meadows. This trend was particularly strong among early-flowering species. The potential reasons for these phenological differences are environmental differences, such as warmer microclimatic conditions and different soil properties in restored meadows; and differences in management practices, such as earlier mowing. Our findings contribute to a deeper understanding of restoration outcomes and underscore the importance of considering restoration-induced phenological shifts in conservation and management practices.

O_LIGrasslands comprise 40% of terrestrial ecosystems and are globally important for food production, carbon storage, and other ecosystem services. However, grasslands in many areas are becoming increasingly exposed to extreme wet and dry periods resulting from global temperature increases. C_LIO_LITherefore, understanding how grasslands will respond to climate change is a pressing issue for managing changes to biodiversity and ecosystem service provision. C_LIO_LIHere, we use experimental manipulations of precipitation (50% increase and 50% decrease of growing-season precipitation) to investigate the resistance of the diversity and productivity of a calcareous grassland community recovering from historical agricultural conversion. C_LIO_LIWe found that decreasing growing season precipitation led to reductions of mean productivity (25 % decrease in peak above-ground biomass) and its temporal stability (54 % increase in biomass variance across years). However, the grassland community composition was resistant to the precipitation manipulations, with no clear difference in community compositional turnover, dissimilarity, or biodiversity indices. Furthermore, the precipitation manipulations had no effect on the path of ongoing (30 year) recovery of grassland plant diversity from the period of previous agricultural conversion. C_LIO_LIWhile the diversity of this calcareous grassland was resistant to precipitation extremes (at least in the short term), sustained reductions in growing-season precipitation reduced productivity and its temporal stability demonstrating that different properties of grasslands can vary in their responses to changes in precipitation. C_LI

While the relative importance of climate filtering is known to be higher for woody species assemblages than herbaceous assemblage, it remains largely unexplored whether this pattern is also reflected between the woody overstory and herbaceous understory of forests. While climatic variation will be more buffered by the tree layer, the understory might also respond more to small-scale soil variation, next to experiencing additional environmental filtering due to the overstorys effects on light and litter quality. For (sub)tropical forests, the understory often contains a high proportion of fern and lycophyte species, for which environmental filtering is even less well understood. We explored the proportional importance of climate proxies and soil variation on the species, functional trait and (functional) diversity patterns of both the forest overstory and fern and lycophyte understory along an elevational gradient from 850 to 2100 m a.s.l. in northern Taiwan. We selected nine functional traits expected to respond to soil nutrient or climatic stress for this study and furthermore verified whether they were positively related across vegetation layers, as expected when driven by similar environmental drivers. We found that climate was a proportionally more important predictor than soil for the species composition of both vegetation layers and trait composition of the understory. The stronger than expected proportional effect of climate for the understory was likely due to fern and lycophytes higher vulnerability to drought, while the high importance of soil for the overstory seemed driven by deciduous species. The environmental drivers affected different response traits in both vegetation layers, however, which together with additional overstory effects on understory traits, resulted in a strong disconnection of community-level trait values across layers. Interestingly, species and functional diversity patterns could be almost exclusively explained by climate effects for both vegetational layers, with the exception of understory species richness. This study illustrates that environmental filtering can differentially affect species, trait and diversity patterns and can be highly divergent for forest overstory and understory vegetation, and should consequently not be extrapolated across vegetation layers or between composition and diversity patterns.

O_LIThe quantum yield () of photosynthesis represents the maximum efficiency of light use as indicated by the initial slope of photosynthetic light response curves. Understanding is crucial for accurate modeling of photosynthesis and terrestrial carbon cycle. Despite its importance, the spatial and temporal variations in at large scales remain largely elusive. C_LIO_LIWe leveraged long-term eddy-covariance observations from 90 sites globally and examined the spatiotemporal variations in due to climatic drivers, using statistical and machine learning approaches. C_LIO_LIWe found significant spatial variability in across and within biomes, primarily driven by atmospheric vapor pressure deficit and soil moisture variations. Meanwhile, the temporal changes in are mainly driven by the negative effect of vapor pressure deficit, which weakens the positive effects of elevated CO2 and leaf area index. C_LIO_LIOur results highlight the dominant role of vapor pressure deficit in controlling the spatiotemporal variations of as well as the unneglectable impacts of soil water content, CO2, and leaf area on . Those new results provide insights for improving the representation of in ecosystem photosynthesis models. C_LI

In many grasslands, species with specific traits occupy unique temporal positions within communities. Such intra-annual segregation is predicted to be greatest in systems with high intra-annual climate variability because fluctuating environmental conditions provide opportunities for temporal niche partitioning among species. However, because most studies on intra-annual community dynamics have been conducted at individual sites, relationships between intra-annual climate variability and seasonal community dynamics at global scales have not yet been identified. Furthermore, the same characteristics that promote species-specific responses to fluctuations in environmental conditions may also drive species-specific responses to global change drivers such as eutrophication. Research provides evidence that eutrophication alters inter-annual plant community dynamics yet understanding of how it alters intra-annual dynamics remains limited. We used early-season and late-season compositional data collected from 10 grassland sites around the world to ask how intra-annual variability in precipitation and temperature as well as nutrient enrichment shape intra-annual species segregation, or seasonal {beta}-diversity, in plant communities. We also assessed whether changes in the abundances of specific functional groups including annual forbs, perennial forbs, C3 and C4 graminoids, and legumes underpin compositional differences between early- and late-season communities and treatments. We found that intra-annual temperature variability and seasonal {beta}-diversity were positively related but observed no relationship between intra-annual precipitation variability and seasonal {beta}-diversity. This suggests that positive relationships between -diversity and intra-annual temperature variability identified in earlier studies may be underpinned by the positive influence of intra-annual temperature variability on temporal segregation of species within growing seasons. We found that nutrient enrichment increased seasonal {beta}-diversity via increased turnover of species between early- and late-season communities. This finding mirrors patterns observed at inter-annual scales and suggests fertilization can alter compositional dynamics via similar mechanisms at varied temporal scales. Finally, fertilization reduced the abundance of C4 graminoids and legumes and eliminated intra-annual differences in these groups. In contrast, fertilization resulted in intra-annual differences in C3 graminoids which were not observed in control conditions, and increased abundance of C3 graminoids and annual forbs overall. Our study provides new insight into how intra-annual climate variability and nutrient enrichment influence biodiversity and seasonal dynamics in global grasslands.

Seasonally Tropical Dry Forests (STDFs) are widespread throughout the world and store significant amounts of carbon; however, they are often overlooked in spatial assessments compared to humid tropical forests. The Caatinga, which is the largest seasonally dry tropical forest in South America, covers approximately 862,000 km{superscript 2} in northeastern Brazil and supports millions of people. Unfortunately, its carbon dynamics has not been thoroughly quantified, especially after centuries of land-use transformation and wood extraction that have significantly diminished its biomass stocks. In this study, we model the potential aboveground biomass (AGB) that Caatinga could sustain under current climatic, atmospheric, and topographic conditions. We integrated data from 301 geo-referenced plots along with high-resolution environmental predictors. The principal Component Analysis revealed two main gradients: a hydro-thermal axis dominated by precipitation, temperature, and severity of drought, and a topographic axis reflecting slope, ruggedness, and terrain position. Random Forest models, validated through both random and spatial cross-validation, explained a significant amount of variation in AGB (R{superscript 2} = 0.81, RMSE = 21 Mg ha). Our findings indicated that AGB is highly sensitive to water availability. In particular, biomass increased sharply when the maximum cumulative water deficit (MCWD) was less than -500 mm and annual rainfall exceeded approximately 1,100 mm. In contrast, elevated vapour pressure deficit (VPD) and potential evapotranspiration (PET) were associated with reduced carbon storage. Topographic heterogeneity further influenced AGB, with rugged and concave terrains supporting potential biomass levels more than twice as high as those found in flat, convex areas. Our predictive map reveals a mosaic of low-biomass cores and localized high-biomass refugia, highlighting the dual influence of hydroclimatic and topographic factors. These findings reposition Caatinga as a heterogeneous and dynamic potential carbon reservoir - historically degraded by changes in land-use but still capable of storing substantial carbon. This research offers critical insights for the restoration, conservation and mitigation efforts for climate change in tropical drylands worldwide.

The positive effects of large herbivores on plant diversity in grasslands have so far been mainly attributed to increased light availability or suppressed dominance, and thus to the consequences of aboveground biomass consumption (trophic effects). However, these insights are mainly derived from short-term experiments. Using a 46-year experiment in a salt marsh, comparing cattle grazing, mowing (as a proxy of aboveground consumption) and the ungrazed control, we found that the non-trophic effects (e.g. trampling, deposition of urine and dung) of large herbivores on plant diversity increased over time, exceeding that of the trophic effects after 23 years. This long-term accumulation of non-trophic effects through slow ecosystem-level feedback highlights the sustainability of using low to moderate densities of large herbivores to conserve plant diversity. Our results emphasize the need for the conservation and re-introduction of large herbivores, domestic or wild, to sustain long-term grassland plant diversity.