Ecology Letters

Predicting the stability of ecological communities in changing environments is challenging. Classical theory posits that community stability cannot be understood without considering interspecific interactions. A contrasting view is that species environmental responses and their variation (response diversity) influence stability to the extent that effects of interspecific interactions can be ignored. Surprisingly, few studies have evaluated the relative importance of interactions versus species responses. Moreover, trait-based measures of response diversity often show limited predictability. Here, we introduce community performance curves, the aggregate of species performance curves, as a powerful mechanistic link between community composition and stability. This approach reveals that species responses predict most of the variation in community stability in simulated communities, even when the strength of interspecific interactions varies. An experiment with ciliate communities corroborates these findings, while a literature review reveals how rarely both mechanisms are assessed jointly. By moving from summary traits to community performance curves, we reconcile the two perspectives: while species interactions undeniably shape community dynamics, community performance curves are sufficient to predict stability. This provides the opportunity to predict community stability, even when information about the multitude and diversity of interspecific interactions is unavailable.

Zooplankton communities are influenced by multiple environmental factors, including temperature, nutrient and resource availability, which fluctuate seasonally and across years. While long-term average effects can identify overall drivers, they may overlook dynamic, context-dependent effects that govern short-term changes in diversity and abundance. Understanding and disentangling both perspectives is crucial for identifying and estimating the drivers that shape community structure under varying environmental states. Here, we applied Empirical Dynamic Modeling (CCM, SMap) to a 12-year weekly zooplankton time series to identify causal environmental drivers of taxonomic and morphological diversity and quantify how the influence of each driver shifts over time. We contrast these results with static long-term average effects inferred from Generalized Linear Models which included predictor sets identified using covariate adjustment and accounting for temporal autocorrelation. Drivers linked to long-term average associations differed from those regulating short-term zooplankton dynamics, revealing a decoupling between mean environmental effects and the drivers of temporal variability. Temperature emerged as a persistent regulator of zooplankton dynamics across multiple diversity dimensions, while variables commonly associated with background trophic conditions (e.g. particulate organic matter) were primarily associated with long-term patterns and showed limited dynamical relevance. Importantly, we find evidence for morphological homogenisation in response to short-term fluctuations in chlorophyll a, which was not detectable in long-term average relationships. This contrast highlights that mean environmental associations do not necessarily reflect the mechanisms governing community dynamics. Impacts might be underestimated if average effects appear weak, or misinterpreted if arising mainly from shared trends or seasonality rather than direct mechanisms Integrating both perspectives clarifies the identity and role of environmental drivers, improving inference and prediction of zooplankton community change through time.

O_LIStage-dependent interactions, in which different life cycle stages (e.g., juveniles, adults) exert different per-capita competitive effects, are widespread across ecological communities. However, whether explicitly accounting for such ontogenetic variation improves forecasts of stochastic community dynamics remains unclear. We tested how the strength of stage dependence and species life-history strategy influence the predictive accuracy of community models that either include or ignore stage-specific interactions. C_LIO_LIWe constructed stochastic two-species competition models using stage-structured matrix population models spanning five virtual life histories along the fast-slow continuum. Density dependence was imposed separately on juvenile survival, adult survival, progression, retrogression, or fertility, and the strength of stage dependence varied from adult-driven to juvenile-driven competition. We then fitted deterministic projection models with and without stage-dependent interaction terms to simulated time series and quantified predictive performance over 100 time-step forecasts using mean absolute percentage error (MAPE). C_LIO_LIIncreasing stage dependence consistently reduced the predictive accuracy of models that ignored stage structure. However, absolute prediction errors remained small across all scenarios (MAPE < 0.7%), even under strong stage dependence. The influence of life-history strategy depended on which vital rate was density dependent: when juvenile survival was density dependent, faster life histories showed larger errors; when progression, retrogression, or fertility were density dependent, slower life histories exhibited greater errors; and when adult survival was density dependent, no consistent life-history effect emerged. Across simulations, temporal variation in population structure was low (coefficient of variation < 0.036), and prediction error was strongly associated with the magnitude of structural fluctuations rather than life-history pace per se. C_LIO_LISynthesis. Stage-dependent interactions can, in principle, alter stochastic competitive dynamics, but their practical importance for ecological forecasting depends on the extent to which population stage structure fluctuates through time. When environmental stochasticity dominates and stage structure remains near equilibrium, simpler models that ignore stage dependence provide robust approximations of community dynamics. Our results identify conditions under which demographic detail is necessary for forecasting and highlight the central role of structural variability in linking life-history strategy to community-level dynamics. C_LI

Knowledge of species interactions unlocks our understanding of how ecological communities respond to climate change or habitat loss, explaining their resilience and robustness. Such knowledge requires inferring the presence, sign, and per capita strength of species interactions, as well as species intrinsic growth rates. While various studies have attempted to infer these parameters in isolation, none have successfully inferred them simultaneously. Here, we solve this grand challenge using an integrative approach combining ecological mechanistic models and statistical inference to simultaneously infer these parameters across time, capturing environmental variation and seasonality. We validate our approach on synthetic data in constant and changing environments, highlighting its ability to detect high-probability weak interactions - the key contribution of our method, and proving our ability to detect environmental changes. Applied to empirical data, it recovers the expectations from biological knowledge and unveils network rewiring. Our approach takes one step further to bridge the gap between mechanistic models and empirical ecology. It advances the understanding of ecological networks and their dynamics, thereby helping to validate existing hypotheses, spark new theories, and help guide ecological management and conservation.

Interaction networks, in which nodes represent species and edges represent direct interactions between species, have a long and impactful history in community ecology. However, co-occurrence networks, where edges represent statistical relationships among species presences or abundances, are often easier to construct from lab and field data. It is clear that co-occurrence edges often do not represent direct interactions, but frameworks for the interpretation of co-occurrence networks have not kept pace with their generation. It is therefore unclear when and how these networks can be used to gain insight into community dynamics. Here, we use a Generalized Lotka-Volterra-based model to explore the contexts in which emergent properties of species interaction networks are identifiable in their resulting co-occurrence networks. We find that, in spite of many differences in direct edges, key features of the true interaction network, such as unipartite modularity, high-degree nodes (hubs), and bipartite modularity and nestedness, can be preserved in co-occurrence networks. In contrast, node degree distributions are not preserved even in the most idealized scenarios. We propose that networks derived from large co-occurrence datasets could therefore be used in future empirical work to test existing hypotheses of how emergent network structures drive ecological community dynamics.

Additive habitat-amount models are widely used to infer independent configuration effects from observational landscape datasets, yet that inference depends on whether habitat amount and configuration are actually separable in the realised predictor space. Using a global multi-taxa forest dataset assembled from paired continuous and fragmented landscapes, this analysis evaluates that condition directly and shows that it is not met. Habitat amount and configuration remain embedded in a shared habitat-loss gradient with asymmetric nonlinear coupling that standard linear diagnostics do not capture, so near-zero additive fragmentation coefficients do not, by themselves, identify the intended ecological contrast. Under this geometry, the additive specification yields the classic cross-over suppressor signature: fragmentation aligns strongly with the fitted biodiversity gradient yet contributes almost no unique variance once habitat amount is included. When residual coupling is reduced to near zero, fragmentation coefficients shift uniformly negative for both local and landscape-scale diversity, and the same raw additive specification yields negative coefficients in high-cover landscapes, showing that the full-dataset null is geometry-conditional rather than stably ecological. The suppressor structure is absent in beta diversity, indicating that the attenuation is response-specific rather than a universal artefact of the dataset or modelling framework. Because these models are widely used to adjudicate fragmentation-per-se claims from observational data, this issue is a direct challenge to how null configuration coefficients have been interpreted across the fragmentation debate. These results show that a stable ecological-null interpretation is not supported in this dataset -- whenever the geometric constraint is reduced, the recoverable direction is uniformly and non-trivially negative. Habitat loss generates configuration change rather than the reverse, embedding asymmetric nonlinear coupling in the attainable predictor space before any landscape is sampled. In empirical landscape datasets, additive control by habitat amount becomes informative about configuration only when the realised predictor geometry has first been shown to support the ecological interpretation being drawn.

Natural populations exhibit complex class structures that profoundly shape evolutionary trajectories. While evolutionary demography provides a formal framework to predict adaptation using invasion fitness, the high mathematical dimensionality of these models often precludes analytical solutions, obscuring biological interpretation and hindering the analysis of long-term evolutionary outcomes. Because current reduction techniques remain fragmented, a unifying theoretical foundation is critically needed. Here, we introduce "structural evolutionary invasion analysis," a systematic framework that integrates two complementary tools to simplify complex life cycles. First, we formulate the "invasion determinant," an algebraic method that yields a direct scalar condition for mutant invasion. Second, we develop the Projected Next-Generation Matrix (PNGM), which structurally compresses life-cycle graphs by eliminating secondary classes. We demonstrate that this reduction is mathematically equivalent to separating dynamical timescales, explicitly preserving Fishers reproductive values for the retained focal classes. Crucially, under the standard assumption of weak selection, our synthesized framework guarantees that all properties of evolutionary singularities--including their location, convergence stability, and evolutionary stability--are strictly identical to those derived from the full, unreduced model. Illustrated with diverse ecological examples, this framework provides modellers with a rigorous and tractable toolkit for decoding state-dependent selection in high-dimensional populations.

O_LIUnderstanding functional community structure and the niche-based mechanisms that enable coexistence among sympatric species is essential for explaining how biodiversity is maintained in natural systems, and for anticipating how ecological communities will respond to ongoing environmental change. Stable isotope analysis provides a process-oriented perspective on resource use by integrating information across time and space, thereby allowing reconstruction of realised isotopic niches that reflect multiple dimensions of ecological differentiation. C_LIO_LIWe applied this framework to a community of ungulates in the Central-Eastern Italian Alps, including red deer (Cervus elaphus), roe deer (Capreolus capreolus), and Alpine chamois (Rupicapra rupicapra). Using stable isotope ratios in summer-grown hair segments ({delta}13C, {delta}15N, {delta}34S, {delta}18O, {delta}2H), we quantified species-specific n-dimensional niche hypervolumes within a Bayesian framework and estimated niche regions, overlap probabilities, univariate differentiation and multivariate structure. C_LIO_LIDespite broad dietary overlap typically observed among these ungulates, we found clear isotopic niche segregation, with mean pairwise overlap consistently remaining below 40%. Three dimensions emerged as primary drivers of differentiation: water sourcing ({delta}18O), diet quality ({delta}15N), and habitat openness ({delta}13C). Specifically, chamois appeared to derive more water from plants in their diet rather than from drinking, and to consume a higher-quality diet compared to Cervids. Red deer relied more heavily on forested habitats for resource use compared to roe deer and chamois, and additional isotopic differences between red deer and roe deer may stem from fine-scale abiotic conditions like microclimate and topography. We found no isotopic evidence for differential niche breadth among the three ungulate species. C_LIO_LITogether, these patterns highlight functional differentiation across multiple ecological axes, offering mechanistic insight into how these ungulates segregate realised niche space despite substantial potential for resource overlap. This multi-element isotope perspective underscores the value of integrative, process-based approaches for understanding current coexistence as well as improving predictions of how mammal communities may reorganise under accelerating environmental change. C_LI

Plant communities within a metacommunity can vary widely in their degree of invasion by introduced species. Disturbance, propagule pressure, and biotic resistance are common explanations for this variation, but empirical evidence for these hypotheses is mixed. Alternatively, the community assembly framework predicts that local assembly filters determine both native and exotic composition, but lower trait variation in the introduced species pool may exclude them from certain sites. We examined evidence for this framework using observational data from forests and woodlands of Long Island, NY, USA. These forests vary in vegetation composition and invasion along a soil gradient. They are also highly disturbed and fragmented, yet some stands have almost no introduced plants. Using data collected in 1998 and 2021-22, we quantified relationships between community composition, soil characteristics, and functional traits for native and exotic assemblages, as indicators of environmental filtering. We found similar trait-environment relationships in native and introduced species, suggesting that both groups follow the same local assembly rules. Introduced species were predominantly found in sites with more nutrient-rich soils and were absent from sites with nutrient-poor soils. At the regional scale, the exotic species pool was biased toward trait values favored in more nutrient-rich environments, particularly high growth rates and low leaf C:N ratios, which explains their absence from nutrient-poor environments. These patterns were consistent over time, and stands that were uninvaded in 1998 remained so in 2021-22, supporting the robustness and reliability of short-term studies. This study shows that invasion patterns in plant communities can be explained by the assembly rules that govern native species. By linking local environmental filtering with regional species pool characteristics, this work advances our understanding of how some communities remain uninvaded despite high disturbance and propagule pressure. Overall, these results highlight the utility of the community assembly framework, and emphasize the importance of regional processes in constraining the local distribution of introduced species.

Theory suggests that different components of environmental fluctuations, from daily and seasonal cycles to multidecadal trends, can have distinct and even opposing effects on species abundances and community dynamics, depending on their specific adaptations. But empirical research that deconstructs the influence of these different cycles on communities is lacking. Here, we used long-term biological monitoring data together with flow records of rivers across New Zealand to (i) investigate the role of fast, slow, and seasonal river-flow fluctuations in structuring macroinvertebrate communities; and (ii) to assess whether life-history and mobility traits mediate the response. Using joint species distribution models, we found striking differences in taxon and community responses to the different components of river flow variation. Responses to slow fluctuations were generally stronger and better predicted by traits, while responses to seasonal fluctuations were highly heterogeneous. Fast increases in flow, typical of flooding events, had pervasive negative effects on species abundances, but the severity of impact partly depended on mobility traits. Our results suggest that different ecological mechanisms underpin the response to distinct environmental fluctuations, highlighting the value of jointly considering multiple temporal scales of variation and species functional traits to understand and predict how communities reorganise under fluctuating environmental regimes.

The Biodiversity-Ecosystem Functioning (BEF) literature has shown species diversity to be essential for ecosystem functioning and services. Yet although acquiring information through interspecific networks can impact ecosystem functioning, it is unclear how it is modulated by species diversity. Eliciting vocal responses using predator models across a latitudinal gradient, we first show that the species diversity of birds increases public information about predation both in the low-cost system of mobbing and in the higher-cost system of alarm calls. A similar result was also found across a fragment area gradient for mobbing; this system was then used to test how species diversity affects interspecific information flow in mobbing communities. We set up two BEF playback experiments, manipulating the species richness level of the playback sound files by varying the number of species producing mobbing calls (one, two, four, eight species). In an experiment in which the call rate across treatments was held constant, and only heterospecific responses were counted, increasing species richness of the sound files increased the number of species and individuals responding, the number of calls produced and their frequency range, and decreased latency to call. An experiment in which call rate increased with the addition of species in each treatment showed a similar, but stronger pattern. There was little evidence that the signals of one particular species changed responses. This supports the hypothesis that the species diversity of a community is a key component influencing the quantity and quality of information flow inside it.

Cross-scale integration remains a persistent challenge in ecology. Mechanistic network models have advanced this integration by linking individual behavior to community dynamics. Their complexity, however, often limits exploration to numerical simulations, which tend to be insufficient for fully unveiling the fundamental rules governing system behavior. Extracting these rules requires moving beyond numerical observation to establish exact, analytical constraints. Here, a complete mathematical analysis of a mechanistically detailed plant-pollinator model is presented. This cross-scale analysis decouples transient and equilibrium dynamics, proving that pollination strictly gates plant persistence while recruitment competition caps equilibrium abundance. The precise behavioral mechanisms scaling up to determine network stability are determined: nestedness stabilizes communities by generating floral reward gradients that guide adaptive foraging, whereas connectance destabilizes by eroding these rescue pathways. Additionally, native community persistence and biological invasions are conceptually unified; a single, multi-scale reward threshold (R*) is shown to govern both native survival and alien establishment. These analytical derivations are distilled into conceptual frameworks and visual summaries accessible for empiricists interested in theory and conceptual unification. By translating numerical observations into rigorous, trait-grounded proofs, this analysis demonstrates that complex, cross-scale networks are tractable, revealing the precise conditions under which communities assemble, persist, and collapse.

Theory predicts that demographic performance should peak at the core of species ranges and decrease toward their limits. Yet, empirical correlations between population growth rate and species distribution remain weak for most tree species. Part of the problem may arise from the difficulty of integrating multiple demographic processes across the complex life cycle of a forest, and from the significant variability among individuals and locations. It remains unclear if the mismatch between performance and distribution arises from modelling limitations or if climate is simply a poor predictor of species performance across distributions. Here, rather than asking whether demographic performance correlates with species distributions, we ask how climate and competition jointly shape population growth rate for 31 tree species across eastern North America. By combining flexible nonlinear hierarchical models for growth, survival, and recruitment with explicit uncertainty propagation, we use Integral Projection Models to address key gaps in previous studies. Perturbation analyses revealed that population growth rate was consistently more sensitive to mean annual temperature than to conspecific or heterospecific competition across all species. We further examined how sensitivities to climate and competition varied across species thermal ranges. The dominance of climate over competition increased toward both cold and hot range limits, while sensitivity to competition generally declined from cold to hot limits. Notably, these patterns emerged along the continental thermal gradient shared across species rather than within each species individual range, suggesting that range-edge demographic responses may arise as a community-level phenomenon. Across species, the largest source of variability remained the local plot conditions captured by random effects, likely reflecting differences in soil conditions, drainage, and disturbance history. Together, these results may provide a mechanistic pathway underlying the performance declines predicted by range-limit theories, and offer a basis for understanding how forest populations and communities may reorganize in response to ongoing climate change and shifting disturbance regimes.

O_LIEcological communities are complex and exhibit considerable spatial variability, presenting challenges in accurately understanding these systems. A primary obstacle in ecological research is the existence of missing links between species: inevitable unobserved interactions that limit our comprehension of ecological networks and their response to change. While link prediction methods have been developed to address this challenge, most approaches overlook the intrinsic spatial variability of ecological systems. C_LIO_LIWe introduce a flexible, spatially explicit framework based on matrix decomposition that leverages latent structural patterns to predict missing interactions and their strength, without requiring species traits or environmental data. The framework integrates information from paired auxiliary and target networks (locations) using thresholded SVD for link prediction. We applied it to plant-pollinator networks across the Canary Islands, performing pairwise predictions between locations, comparing them to within-location predictions (as a control), and quantifying how spatial variability influences predictive performance. C_LIO_LIPredictions revealed that latent network structure contains substantial predictive information, with F0.5 scores consistently exceeding a random baseline (mean F0.5 = 0.67 {+/-} 0.02 SD), while being less sensitive to interaction strength. The method enabled identifying plausible gaps in the data and producing ecologically coherent predictions. Incorporating information from auxiliary locations enhanced predictive accuracy in certain cases, but success depended on spatial context: predictions were most reliable when derived from nearby, ecologically similar locations, and declined with increasing geographic and ecological distance, consistent with a distance-decay effect. C_LIO_LIWe conclude that the predictability of missing links is spatially variable, reflecting both network and species-level heterogeneity. These patterns provide insights into network structure and the ecological processes shaping it, complementing trait-based approaches. While network structure offers rich predictive information, spatial context is essential for applying it effectively: ignoring spatial variability can obscure ecological signals and inflate predictive error. Our framework is computationally efficient, transferable, and readily applicable to any system with spatial or temporal replication. It can be used for a variety of ecological contexts, including island systems, fragmented landscapes, and environmental gradients, making it a practical and scalable tool for advancing link prediction in ecology. C_LI

Classical niche theory predicts that species with more similar traits experience stronger competition, and thus, trait dimensionality, here defined as the number of independent traits per species in a community, plays a critical role in coexistence. Despite this, current studies often assume that communities are structured by only a few key traits. Here, we leveraged a theoretical framework that integrates phylogenetic information with repeated instances of community assembly to estimate the number of ecologically relevant traits required to support observed species richness in experimental and natural plant communities. We found that the inferred trait dimensionality is surprisingly high, often exceeding species richness, suggesting that many traits contribute to coexistence. Furthermore, we explored drivers of grassland trait dimensionality, and it depends in complex ways on area, species pool size, and latitude. Our findings indicate that local coexistence may rely on a larger number of traits than previously assumed, challenging low-dimensional trait-based views of community structure.

Species tolerating the same environmental conditions can potentially colonize and thrive in the same habitats and eco-regions. Are any pair of those species equally probable to co-occur in the same community? Can we quantify the propensity of two species to co-occur together? Here, we focus on a simple but largely overlooked community-level pattern: the co-occurrence-occupancy curve, which relates the tendency of species to co-occur with others to their total occupancy across sites. We first define this empirical curve and then derive its expected shape under a random null model that assumes site equivalence and species independence. Building on these results, we introduce the Species Association Index (SAI), an occupancy-standardized measure that quantifies the tendency of a species to associate with others independently of its overall frequency of occurrence. The SAI enables meaningful comparisons among species with contrasting occupancies and provides a transparent benchmark against which departures from neutrality can be assessed. We illustrate the approach using two contrasting systems--tropical rain forest trees on Barro Colorado Island and organisms from Mediterranean rocky shores--highlighting both the generality of the co-occurrence-occupancy framework and its limitations.

Predicting the composition and dynamics of ecological communities is challenging because complexity increases rapidly with species richness. A common strategy is to adopt a reductionist framework in which community dynamics are inferred from simpler components, such as population-level parameters or organismal traits. However, it remains unclear at which level of biological organization ecological predictability emerges. Here we experimentally test this reductionist cascade in marine phytoplankton communities. We first ask whether multispecies dynamics can be quantitatively predicted from demographic parameters measured in monocultures and species pairs. We then test whether these predictive parameters can themselves be inferred from organismal traits, focusing on cell size. We find that community composition is highly reproducible and can be accurately predicted from population-level parameters measured in simpler experimental settings. In contrast, these parameters do not show systematic relationships with cell size and cannot be predicted from this commonly used trait. These results demonstrate that ecological predictability emerges at the population level, where demographic parameters capture the combined effects of underlying biological processes, but resist further reduction to simple trait-based descriptions, suggesting that ecological interactions reshape organismal performance across levels of organisation.

A major goal of community ecology lies in the deciphering of the processes underlying species distribution. A widespread approach to this question is to identify patterns in species community data and relate them to possible processes. Joint Species Distribution Models (JS-DMs) offer one way to do so through the infernece of association networks that describe patterns of statistical correlations and dependencies between species, but it is unclear what processes can explain the presence of such correlations. While it has now been established that there is no equivalence between JSDM-inferred associations and biotic interactions, the later remain one possible explanation, among others, for the former. However, to our knowledge, there is no specific study of the statistical patterns induced by different types of interactions or of the conditions under which they may or may not appear as statistical correlations / dependencies in species communities. To explore these questions, we propose a "virtual ecologist" approach that consists in simulating community data based on abiotic and biotic processes with the VirtualCom model that emulates the effects of environmental processes and of competition and facilitation interactions. Then, we study to what extent JSDMs retrieve correlations between species that match the simulated interactions. We show that these interactions are better identified when using JSDMs that model partial correlations between species rather than marginal ones. We further demonstrate how critical it is to correctly model abiotic effects in order to identify biotic ones and that the "correct modelling" of these effects depend on the type of interactions at stake.

The relationship between energy supply and biodiversity is a longstanding question in ecology. Although a monotonic increase in diversity with energy availability is often assumed, unimodal species-energy relationships have been widely documented across ecosystems, and their origin from first principles remains unclear. Here, we develop a geometric framework that recasts ecological feasibility in explicitly energetic terms. By treating total energy supply as a system-level constraint on an energy-based network model, we define nested feasibility domains in the space of energy capture rates and quantify feasibility probabilities as their volume ratios. We show that the probability of initializing a feasible network increases monotonically and saturates with energy supply, whereas the probability of sustaining steady-state biomass follows a unimodal relationship--revealing a bounded energetic window within which network maturation is most likely. Extending this analysis to all candidate subcommunities via feasibility partitions, we find that different community sizes are most feasible at different energy levels, and that average diversity itself peaks at intermediate supply. Together, these results suggest that energetic constraints determine the diversity of ecological networks not through energy scarcity alone, but through the geometric interplay between external energy supply and internal energy exchange. Author SummaryWhy do many ecosystems show the highest biodiversity not where energy is most abundant, but at intermediate levels? This unimodal species-energy relationship has been documented across grasslands, wetlands, and rainforests, yet its origin from first principles has remained unclear. We approached this question by developing a simplified model that treats ecological networks as energy-processing systems. In this model, each species captures energy from the environment and exchanges it with others, and the total energy available to the network is explicitly limited. By measuring how the likelihood of species coexistence changes with energy supply within this framework, we found that while a minimum energy threshold is needed for any community to persist, too much energy can paradoxically reduce the chance of long-term coexistence. This creates a bounded energy window most favorable for community persistence. When we extended the analysis to all possible subsets of species, we found that different-sized communities are most likely to persist at different energy levels, and that overall expected diversity peaks at intermediate supply. These results suggest a possible geometric origin for why more energy does not always support more species, providing a theoretical baseline for connecting the structure of energy flow within networks to observed biodiversity patterns.

Projecting species responses to changing temperatures remains a major challenge in ecology. Central to this effort is harnessing our understanding of species thermal physiological traits, which underlie ectotherm fitness. These traits are typically characterized by describing performance across temperatures (thermal performance curve, TPC), and/or tolerance limits, which capture endpoints of biological failure. Despite their importance, we still lack an understanding of the functional relationship between these traits, limiting our ability to integrate them into comprehensive vulnerability assessments. Using a synthesized dataset of >100 ectotherms, we tested how heat tolerance (CTmax) relates to key TPC traits: thermal optima, thermal maxima, and the supra-optimal range of temperatures where performance is positive. Across taxa, TPC traits were positively related to CTmax, demonstrating a link between heat tolerance and temperature-dependent performance at sub-critical temperatures. While acute locomotor performance scaled proportionally with CTmax, metabolic processes and sustained locomotion scaled sub-proportionally, suggesting decoupling of CTmax and performance among high-CTmax species. This suggests that using CTmax as a comparative metric may overestimate thermal safety margins for metabolic processes critical to growth. Our results indicate that while CTmax and TPCs reflect shared underlying constraints--particularly in acute neuro-muscular traits--their relationship is dependent on timescale and the TPC response trait. Our findings connect our understanding of the processes that maintain performance over thermal gradients with those that cause performance to fail, improving our ability to project species persistence in a warming world. SignificanceClimate warming is increasingly reshaping the thermal environments that govern species persistence worldwide. Predicting vulnerability requires integrating multiple aspects of thermal biology, yet relationships among widely used thermal traits remain poorly understood. By synthesizing data from more than 100 ectotherm species, we quantify links between acute heat tolerance and traits describing sustained biological function across temperatures. We show that performance at relatively benign temperatures and performance at thermal extremes are coupled, but this coupling is strongly process and timescale dependent, with close correspondence for short term locomotion but weaker coupling for metabolic processes. Our results link the processes that maintain performance across temperatures with those that cause failure, fundamentally advancing our projections of species performance in a warming world.