Species interactions determine the importance of response diversity for community stability to pulse disturbances

Species interactions determine the importance of response diversity for community stability to pulse disturbances | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL Ecology Letters This is a preprint and has not been peer reviewed. Data may be preliminary. 18 April 2025 V1 Latest version Share on Species interactions determine the importance of response diversity for community stability to pulse disturbances Authors : Charlotte Kunze 0000-0002-1130-7417 [email protected] , Owen Petchey 0000-0002-7724-1633 , Shyamolina Ghosh 0000-0002-5137-9933 , and Helmut Hillebrand 0000-0001-7449-1613 Authors Info & Affiliations https://doi.org/10.22541/au.174498373.31851247/v1 Published Ecology Letters Version of record Peer review timeline 574 views 391 downloads Contents Abstract Word counts Acknowledgements ABSTRACT 1. INTRODUCTION 2. METHODS 2.4 Meta Analysis of empirical communities 2.5 Community (in-)stability to disturbances 2.6 Species responses to pulse disturbances 2.7 Response diversity measures 2.8 Statistical Analysis 3. RESULTS 3.2 Meta-Analysis 4. DISCUSSION 5. CONCLUSION Tables Figure legends References Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Communities can buffer environmental change through the diverse responses of their species, often leading to greater emergent stability than expected from individual species. Metrics such as response dissimilarity and divergence capture this response diversity in fluctuating environments. Here, we test whether diverse species responses also stabilise community properties under pulse disturbance. Combining model simulations of multi-species communities with empirical data from a meta-analysis, we find that community stability was consistently determined by the mean species response, regardless of interaction strength. Contrastingly, response dissimilarity and divergence were only related to stability in the absence of interspecific interactions. While response diversity increases stability under fluctuating conditions, pulse disturbances often cause negative responses in most species, and stability is highest when species uniformly exhibit strong resistance or fast recovery. These results highlight that the role of response diversity in promoting community stability depends on the disturbance regime and is shaped by species interactions. Journal: Ecology Letters Article Type: Letter Title: Species interactions determine the importance of response diversity for community stability to pulse disturbances Running title: Response diversity under pulse disturbance Charlotte Kunze 1 , 0000-0002-1130-7417, [email protected] * Owen L. Petchey 2,3 0000-0002-7724-1633, [email protected] Shyamolina Ghosh 2,4 , 0000-0002-5137-9933, [email protected] Helmut Hillebrand 1,5,6 , 0000-0001-7449-1613, [email protected] *corresponding author Affiliations: 1 Institute for Chemistry and Biology of the Marine Environment (ICBM), School of Mathematics and Science, Carl von Ossietzky Universität Oldenburg, Ammerländer Heerstraße 114-118, 26129 Oldenburg, Germany 2 Department of Evolutionary Biology and Environmental Studies, University of Zurich, Winterthurerstrasse 190, 8057 Zürich, Switzerland. 3 Theoretical Sciences Visiting Program, Okinawa Institute of Science and Technology Graduate University, Onna, 904-0495, Japan 4 ECORISK Research Training Group, Osnabrück University, Seminarstr. 33, 49074 Osnabrück, Germany 5 Helmholtz-Institute for Functional Marine Biodiversity at the University of Oldenburg (HIFMB), Oldenburg, Germany. 6 Alfred-Wegener-Institute, Helmholtz Centre for Polar and Marine Research, Bremerhaven, Germany. Statement of authorship C.K. and H.H. designed the study. C.K. collected the data and performed the meta-analysis. S.G. wrote the model, C.K. and O.L., performed the model analyses. C.K. wrote the first draft of the manuscript. All authors contributed to manuscript drafts. Data accessibility statement: Code for the simulations is now available at the public repository https://github.com/charlyknz/response-diversity-pulse-pert/tree/main\#. Code for the meta-analysis is available at https://github.com/charlyknz/MetaMultistab. Data are published on Figshare: https://doi.org/10.6084/m9.figshare.28803887.v1 Key words: biodiversity, compensatory dynamics, meta-analysis, response diversity, population, perturbation, variability, resilience Word counts Abstract: 150 Main Text: 4996 Figures in main text: 4 Tables in main text: 2 References: 50 + 30 Meta-Analysis studies Acknowledgements C.K. and H.H. were supported by DFG funding (HI848/29-1). H.H. was supported additionally by HIFMB, a collaboration between the Alfred-Wegener-Institute, Helmholtz-Centre for Polar and Marine Research, and the Carl-von-Ossietzky University Oldenburg and the Ocean Floor Excellence Cluster (EXC2077). OLP was supported by SNSF Project funding 10002183 - Improving predictions of community and ecosystem stability by exploring and validating response diversity measures. SG would like to thank University of Zurich, and Deutsche Forschungsgemeinschaft (grant number RTG 3004-1) for funding. The work described in this paper in part results from the activities and support of the Response Diversity Network (). We thank all authors who shared raw data on species-specific responses and Francesco Polazzo for helpful comments on an earlier version of this manuscript. ABSTRACT Communities can buffer environmental change through the diverse responses of their species, often leading to greater emergent stability than expected from individual species. Metrics such as response dissimilarity and divergence capture this response diversity in fluctuating environments. Here, we test whether diverse species responses also stabilise community properties under pulse disturbance. Combining model simulations of multi-species communities with empirical data from a meta-analysis, we find that community stability was consistently determined by the mean species response, regardless of interaction strength. Contrastingly, response dissimilarity and divergence were only related to stability in the absence of interspecific interactions. While response diversity increases stability under fluctuating conditions, pulse disturbances often cause negative responses in most species, and stability is highest when species uniformly exhibit strong resistance or fast recovery. These results highlight that the role of response diversity in promoting community stability depends on the disturbance regime and is shaped by species interactions. 1. INTRODUCTION As Earth’s ecosystems face unprecedented and often human-induced changes, predicting how ecological communities will respond to changing environmental conditions has become a major scientific challenge. Ecological stability captures the ability of an ecosystem to withstand and recover from disturbances (Pimm, 1984) and is a multidimensional concept. It encompasses a variety of different metrics (Grimm and Wissel, 1997; Ives and Carpenter, 2007; Pimm, 1984) that have been sorted (Donohue et al., 2016, 2013), decomposed (Hillebrand et al., 2018), and recently reassembled into an integrative metric of Overall Ecological Vulnerability (OEV) (Urrutia‐Cordero et al., 2021). Stability can be measured at the level of individual species (population stability) and of emergent properties of the whole community, such as total biomass. Community properties are often found to be more stable than the properties of their component species (Tilman, 1999, 1996). Hereby, diverse communities fluctuate less than species-poor ones (“insurance hypothesis”, Yachi and Loreau, 1999) because they have a greater chance of harbouring species that are resistant to disturbances (Yachi and Loreau, 1999) or because higher richness allows higher asynchrony of species abundances (Hautier et al., 2014; Loreau and de Mazancourt, 2013). Hence, the buffering of community fluctuations can be based on mechanistic replacements (compensatory dynamics) but also increases the chance for different species responses by chance (portfolio effect, Doak et al., 1998). Elmqvist et al. (2003) described this asynchrony and difference in species responses to disturbances as response diversity. Unlike functional diversity, which focuses on the variation in traits that constrain species’ roles in an ecosystem, response diversity specifically captures the variation in species’ responses to change (Elmqvist et al., 2003; Mori et al., 2013). Suding et al. (2008) introduced the terms “effect traits” aligned with functional diversity and “response traits” aligned with response diversity. Under changing environmental conditions, the stability of functional properties can rely on both, functional redundancy (multiple occurrence of the same effect trait) and response diversity, i.e. multiple response traits of functionally similar species (Lawton and Brown, 1994; Suding et al., 2008; Yachi and Loreau, 1999). High response diversity enables compensatory dynamics, where declines in one species’ biomass are offset by other species increases because of their differing environmental responses (Gonzalez and Loreau, 2009; Micheli et al., 1999; Mori et al., 2013). Empirical evidence on how response diversity affects stability is scarce as the advancement of the theoretical concept has outpaced the empirical test of it (Ross and Sasaki, 2023). Most empirical studies have measured response diversity as a response variable for stability of ecosystems but not as predictor of stability (Ross et al., 2023). Of the studies empirically linking response diversity to stability of aggregate community properties (biomass or abundance), those conducted in fluctuating environments have highlighted a positive effect of response diversity on temporal stability at the community level (Leary and Petchey, 2009; Ross et al., 2023). Studies exploring the consequences of pulse disturbances find contrasting relationships between response diversity and resistance as well as resilience (Baskett et al., 2014; Bhaskar et al., 2017). For example, Bhaskar et al. (2017) found that secondary forest recovery rates after a hurricane were slower with increasing response diversity, measured as variation in species-specific responses. However, response diversity was positively related to resistance in this study, emphasising the often-observed trade-off between resistance and recovery rate (Donohue et al., 2013; Hillebrand et al., 2018; Hillebrand and Kunze, 2020). These studies suggest the need to consider different disturbance regimes (fluctuations vs. pulses) to operationalise the context of response diversity for empirical ecology. There is also the question of what information to use for measuring response diversity. One option is to derive response diversity from fundamental responses of species in isolation (Brennan and Collins, 2015; Leary and Petchey, 2009; Ross et al., 2023). Alternatively, it could be derived from realised responses in the presence of other species, thus including the effects of species interactions (Baskett et al., 2014; Bhaskar et al., 2017). Fundamental species responses are often the basis for modelling approaches, but their empirical application is constrained by the need to isolate and study each species individually - an approach that is generally only practical in low-diversity or artificial communities. Moreover, the applicability of fundamental species responses could be low in situations where species interactions are important. In contrast, realised species responses, i.e., responses estimated in the presence of effects of other species, may offer a more precise depiction of real-world conditions. However, these responses are also likely to be highly context-specific, i.e. most suitable in the community context in which they are measured and less so in other community contexts. Theoretical (Ives et al., 1999; Ives and Cardinale, 2004) and empirical studies (Ruiz‐Moreno et al., 2024) collectively suggest that response diversity plays a more important role in driving community dynamics than interspecific interactions. Whereas evidence from a meta-analysis indicates that altered species interactions may have a greater impact on population distribution and abundance than the direct effects of climate change (Ockendon et al., 2014). This leaves a knowledge gap about the influence of species interactions on the role of response diversity for community stability and about the influence of using fundamental or realised species responses. Here, we ask (i) whether response diversity is related to community stability under pulse disturbances, (ii) whether the relationship between response diversity and stability differs when response diversity is derived from species’ fundamental responses versus realised responses, and (iii) whether this relationship depends on the strength of interspecific interactions. We use two complementary approaches: mathematical simulations of competitive Lotka-Volterra dynamics and empirical results obtained from a meta-analysis of pulse-disturbance experiments to test hypothesis H1 that i ncreasing response diversity increases the stability of properties of the community after pulse disturbance. The empirical data only allow deriving response diversity from realised responses, i.e., in the presence of other species and under an unknown average interaction strength that cannot be quantified. The simulations provide insight into response diversity, both through fundamental differences between species and through the realised responses in the simulations, with varying average interaction strengths. This allowed us to test H2, that the effect of response diversity differs between fundamental and realised responses and H3 that the effect of response diversity on community stability depends on the strength of interspecific interactions. 2. METHODS 2.2 Simulated communities We simulated a multi-species community using a discrete-time version of the classical Lotka-Volterra model (De Mazancourt et al., 2013) with temperature-dependent vital rates (Vasseur, 2020): \begin{equation} r_{i}\left(t\right)=\ln{N_{i}\left(t+1\right)-\ln{N_{i}\left(t\right)\ }}\nonumber \\ \end{equation} \(=r_{\text{mi}}[1-\frac{N_{i}\left(t\right)+\sum_{j\ \neq i}{\alpha_{\text{ij}}N_{j}(t)}}{K_{i}}]\)(Eq. 1) where \(N_{i}\left(t\right)\) corresponds to the biomass of species i at time t. The \(r_{\text{mi}}\) is the intrinsic rate of natural increase, which is given by the difference between the birth rate \(\left(b_{0,i}\right)\ \)and death rate \((d_{0,i})\).\(K_{i}\) is carrying capacity, and \(\alpha_{\text{ij}}\), the interspecific competition coefficient describing the effect of species i on j . While r and K are temperature-dependent variables, \(\alpha\) is independent of the environment. The carrying capacity K of species i is given as \(K_{i}=\frac{r_{\text{mi}}}{\beta+\delta}\ \)(Vasseur, 2020; Eq 14.6) where \(\beta\) and \(\delta\) are the density dependent constants. Temperature dependence was incorporated into the birth and death rate: \begin{equation} b_{0,i}\left(T\right)=abe^{\frac{{-\left(T-b_{opt,i}\right)}^{2}}{s_{i}}}\nonumber \\ \end{equation} \(d_{0,i}\left(T\right)=ade^{\text{zT}}\) (Vasseur, 2020; Eq 14.5) where, \(a_{\ }\)is an intercept, \(b_{opt,i}\ \)is the temperature at which intrinsic growth rate is highest, \(s\) is the width of intrinsic growth rate-birth rate function, \(z\) is the slope of death rate-temperature function and scales the effect of temperature on death rate (in °C) to mimic the Arrhenius relationship. Combining the birth rate temperature function and death rate temperature function gives the left-skewed shape common for temperature performance curves (Vasseur, 2020). Creating communities We created communities with S = 10 species. The species in a community varied in their optimum temperature for birth (\(b_{opt,i}\)) such that some species had maximum birth rates at lower temperatures, some had maximum birth rate at higher temperatures, and some had maximum birth rate at intermediate temperatures. The values of \(b_{opt,i}\) of the species in a community were drawn from a uniform distribution with a defined mean and range. When the mean of the distribution was low, then species in the community tended to grow better at lower temperatures. When the range of the uniform distribution was small, the species tended to grow best at similar temperatures. Intra- and interspecific competition was as follows. Intraspecific interactions (\(\alpha_{\text{ii}}\)) were always given a value of 1 (\(\alpha_{\text{ii}}=1\)). Interspecific interactions (\(\alpha_{\text{ij}}\)) were the absolute value of draws from a normal distribution with a mean of 0 and a specified standard deviation. When the standard deviation was small, interspecific interactions were relatively weak. When the standard deviation was larger, interspecific interactions were stronger. Pulse disturbance Each community experienced two independent simulations, one control and one pulse disturbance. Each simulation ran for 750-time steps in total. The first 500 times steps of both control and pulse simulations were at control temperature (22 °C). At time step 500 the temperature of the pulse simulation was decreased by 7 °C (from 22°C to 15°C) in one time step, was kept constant at 15°C for 50-time steps. After that, temperature returned to 22°C in one time step. Communities in the control simulations experienced constant temperatures of 22 °C. Durations and temperatures were selected so that 500-time steps allowed transients to pass and ensured that all communities reached an equilibrium before the pulse. Only data from timepoint \(\geq\)500 contribute towards the analysis. Effects of interactions and response diversity Effects of the strength of interspecific interaction were examined by systematically varying the average interaction strength of communities (by varying the value of\(\mathbf{\alpha}_{\mathbf{\text{ij}}}\mathbf{\_sd}\)). There were 21 levels of interaction strength, with\(\mathbf{\alpha}_{\mathbf{\text{ij}}}\mathbf{\_sd}\) taking values from 0 to 0.5 in steps of 0.025. A value of 0 gave a community with no interspecific interaction, whereas a value of 0.5 gave a community with strong interspecific interactions. Effects of response diversity were examined by systematically varying the distribution of the species temperature optima in a community. The parameters \(\mathbf{b}_{\mathbf{opt,mean}}\) and\(\mathbf{b}_{\mathbf{opt,range}}\) were used to do this. A community with a high value of \(\mathbf{b}_{\mathbf{opt,mean}}\) contained species with high temperature optimum and were therefore mostly negatively affected by a pulse perturbation to lower temperature. A community with high value of \(\mathbf{b}_{\mathbf{opt,range}}\)contained species with very different temperature optima, so that some might respond positively to the pulse disturbance, and some respond negatively. Varying these two community-level parameters independently of each other created a set of communities that varied in their response diversity. There were 15 values of \(\mathbf{b}_{\mathbf{opt,range}}\)from 15 to 22 in 0.5°C steps and 9 values of\(\mathbf{b}_{\mathbf{opt,range}}\) from 3 to 7 in 0.5°C steps. To account for the stochasticity introduced by random selection of interaction strengths and temperature optima, we replicated each combination of\({\mathbf{\alpha}_{\mathbf{\text{ij}}}}_{\mathbf{\text{sd}}}\mathbf{,\ }\)\(\mathbf{b}_{\mathbf{opt,mean}}\), and\(\mathbf{b}_{\mathbf{opt,range}}\) five times. This gave a total of 14’175 independent communities. 2.4 Meta Analysis of empirical communities We extracted species-specific information from a published meta-analysis on the effect of pulse disturbances on community stability (Hillebrand and Kunze, 2020). The original meta-analysis was performed in 2018 based on a search at the Web of Science (.com/WOS, assessed April 3 rd , 2018) using the search terms ‘(experiment* or manipulat* or mesocosm* or microcosm*) AND recover* AND (disturb* or perturb* or pulse) AND (communit* or composit* or diversit* or assembl*)’. We used the published meta-analysis because it identified key factors influencing community stability in the face of pulse disturbances and allowed us to focus on disentangling species-specific responses. From the 110 publications of the original analysis, we retrieved those experiments which fulfilled the following selection criteria: 1) The study contained data on genus or species level. 2) The study reported data for all species in the community. 3) At least 3 time points were sampled in the experiments. 4) The study contained an undisturbed control. These criteria led to a database comprising 98 experiments from 30 publications (see Appendix S2). For each time point we obtained means for the available univariate response variable (abundance, biomass) for control and treatment. Most studies reported abundance and only in the groups of organisms and the duration of the experiments, we converted the duration of the experiments to a scale of 0 to 1, with 1 being the last sampling date (Hillebrand and Kunze, 2020). 2.5 Community (in-)stability to disturbances To assess community stability, we used the Overall Ecological Vulnerability (OEV), a metric of instability (Urrutia‐Cordero et al., 2021). The OEV is calculated as the area under the curve (AUC) of community biomass relative to an undisturbed control. In this framework, the total response of the disturbed community (Treat.Tot) is normalised to the community response in control conditions (Con.Tot) as response ratio (RR) of this difference. \begin{equation} tot.RR=\ \frac{(Treat.Tot-Con.Tot)}{\left(Treat.Tot+Con.Tot\right)}\ \nonumber \\ \end{equation}\begin{equation} OEV=AUC\left(\text{tot.RR}\right)\nonumber \\ \end{equation} The OEV encompasses several dimensions of post-disturbance-stability and directly correlates with the overall effect size of the disturbance: A larger OEV indicates greater variability, weaker resistance to disturbance, slower recovery (resilience) and/or incomplete recovery (for overviews of these stability dimensions see Donohue et al., 2013; Hillebrand et al., 2018; Urrutia‐Cordero et al., 2021). In contrast to the original OEV, we consider both negative and positive deviations, allowing us to explore the relationship between community stability and realised species responses, which can differ by showing negative or positive deviations from the control. 2.6 Species responses to pulse disturbances To assess the effect of the disturbance on each species, we standardised species-specific responses to the disturbance with their response in control conditions. We distinguished between species fundamental responses and species realised responses to pulse disturbances (Fig. 1a). Species fundamental responses reflect their fundamental niche-based responses to disturbance obtained from their performance along an environmental gradient in isolation. Specifically, species fundamental responses are determined as the difference between the intrinsic growth rate (IGR) under undisturbed, control conditions and the disturbed environment, i.e. the IGR effect (Fig. 1). IGR effect = IGR disturbed – IGR control The IGR effect becomes negative when the IGR in the disturbed environment is smaller than in the control and positive when the IGR is larger in the disturbed environment than in the control. The IGR effect is similar to calculating the first derivative (slope) of the temperature response curve. Species realised responses reflect the difference in the net responses of species to a disturbance within the community relative to an undisturbed control, incorporating both the sensitivity of individual species to disturbances and the outcomes of interactions (Fig. 1). Specifically, species realised responses are assessed from species biomass/abundance in the disturbed community relative to their biomass/ abundance in the undisturbed control community. We base our calculation on the OEV instability metric, which is also the basis for our calculation of community stability, and when transferred to population level responses, reflects species absolute contribution to stability (Kunze et al., 2024). Specifically, realised responses are determined as the AUC of the biomass response ratio, i.e. the difference in species-specific biomass in treatment (Treat.spp) and in control (Con.spp) divided by the summed biomass of the same species in treatment and control for each single time point: \begin{equation} RR=\ \frac{(Treat.spp-Con.spp)}{\left(Treat.spp+Con.spp\right)}\ \nonumber \\ \end{equation} If RR equals zero, the species is unaffected by disturbance at this time point. Negative RR reflects biomass declines, positive RR reflect biomass increases for a species. The AUC of this deviation (AUC.RR) directly correlates with the overall effect size of the disturbance on the species in the community, i.e. their realised responses . Negative values indicate a disadvantageous effect of the disturbance on the species over time by decreasing biomass production, whereas positive values indicate a beneficial effect of the disturbance over time by facilitating biomass production compared with the control. 2.7 Response diversity measures Response diversity can be measured in several ways, including with the two measures response dissimilarity and response divergence (Ross et al., 2023), alongside the mean species response (see Fig.1 & Table 2 for an overview of used metrices). However, we are aware that the mean response is not always independent of response diversity in our simulations (Appendix Figs. S2,3). Response diversity was assessed using two complementary components: the dissimilarity metric and the divergence metric (Ross et al., 2023). Both measures of response diversity were assessed from fundamental species responses (IGR effect), i.e. fundamental response diversity and from species realised responses (AUC.RR), i.e. realised response diversity. The dissimilarity-based metric captures the overall magnitude of response diversity among species responses (Leinster and Cobbold, 2012). Specifically, dissimilarity is calculated based on pairwise Euclidean distances of species responses to disturbance between all pairs of species in the community, respectively for realised and fundamental responses. Dissimilarity is low when species responses are identical, and high when species responses are dissimilar, but independent of the actual direction of responses. For example, a dissimilarity of one indicates that all species respond identically. The divergence metric aims to quantify the extent to which species declines are offset by gains (Yachi and Loreau, 1999), and considers whether responses differ in direction. Instead of calculating the first derivatives of species performance-environment relationships and assessing their interspecific variation (Ross et al., 2023), we evaluate the divergence between species’ responses to disturbance relative to the control (both realised and fundamental responses, respectively). Response divergence is bound between zero and one, where zero indicates that all response values are either negative or positive, and one indicates that the most negative value is equal in magnitude to the most positive value (Table 2). The species mean response was separately calculated from species realised responses and from fundamental responses. A mean of zero results from balance in the negative and positive responses, i.e., the sum of the positive values equals the sum of the negative values. Positive and negative deviations from zero hence indicate a positive or negative trend in the mean response. 2.8 Statistical Analysis All calculations and simulations were performed using R version 4.4.0 (2024-04-24) (R Core Team, 2024). The meta-analysis was performed using the ‘metafor’ package (Viechtbauer, 2010), visualisations and data transformations were performed using the packages cowplot (Wilke, 2020), tidyverse (Wickham et al., 2019), ggpubr (Kassambara, 2020). The AUC was estimated using the auc() function of the MESS package (Ekstrøm, 2022) with linear splines. For analysis of the simulation results, we removed very rare species from the analyses (fewer than 10 entries with abundances higher than 10 -3 ). We did this because small changes in species-specific abundance in the treatment relative to the control have minimal effect on overall abundance and its stability. However, these small changes can result in disproportionately large species-specific AUCs because they are measured on a relative scale. For example, an increase in abundance from 10 -10 to 10 -8 , results in very high AUC values (>150) but has virtually no effect on community stability. The threshold thus prevented overestimating the impact of very rare species for the calculation of realised responses. In contrast to the simulations, empirical data from the meta-analysis allowed only calculating realised species responses, as studies did not contain monospecific information, and thus fundamental responses could not be obtained. Community stability was assessed from the summed biomass/ abundance (depending on study) of all species within a community analogous to the model simulations. To determine the relationship between response diversity and mean species responses on community stability, we performed an unweighted meta-analysis with community instability as response variable, mean species responses, response dissimilarity and divergence as moderators, and experiment and study ID as random factors. 3. RESULTS 3.1 Simulations Species’ fundamental responses In the absence of interspecific interactions (Fig. 2a-c), the simulations yielded strong relationships between community instability (OEV) and the mean response as well as both metrics of fundamental response diversity. Instability was lowest when the mean response was zero, whereas positive or negative mean responses led to higher instability (Fig. 2a) with the sign conserved, as treatments could deviate positively or negatively from control. Highest response diversity, measured as dissimilarity (Fig. 2b) or divergence (Fig 2c) was associated to lowest instability (OEV ≈ 0). Reductions in fundamental response diversity resulted in positive or negative OEV, the magnitude of instability was correlated to the magnitude of response diversity (Fig. 2b, c). Adding interspecific interactions of intermediate (Fig. 2d-f) and high (Fig. 2g-i) strength to the simulations resulted in similarly shaped relationships between OEV and mean response as well as response diversity but shifted towards more negative deviations in OEV. This shift towards more negative deviations in OEV is the result of interactions in our model simulations that are purely competitive and reduce the net growth rates of species. The mean response showed the same monotonic relationship with community instability as interaction strength increased (Fig. 2a,d,g), but with increasing variability. Moreover, a more positive mean fundamental response was required for lowest instability (OEV ≈ 0). Similarly, increasing interaction strength maintained the V-shaped relationship between fundamental response dissimilarity and community OEV (Fig. 2b,e,h) and between fundamental response divergence and community OEV, respectively (Fig. 2c,f,i). However, highest response diversity was no longer associated with lowest instability (OEV ≈ 0), but negative OEV. Reductions in response diversity then led to even more negative OEV or less negative OEV turning positive. Lowest instability was consequently associated with intermediate response diversity. Species’ realised responses The simulations allowed testing whether the response diversity – stability relationships differ when based on realised responses instead of fundamental responses. In the absence of species interactions, the monotonic relationship between community instability and the mean realised response (Fig. 3a) as well as the V-shaped relationship to both metrics of realised response diversity (Fig. 3b-c) mirrored the patterns observed for fundamental species responses (cf Fig. 2a-c). However, the relationships were shifted quantitatively, such that lowest instability (OEV ≈ 0) was observed at slightly positive mean realised responses (Fig. 3a) and below the maximum response diversity (Fig. 3b-c). Increasing interaction strength led to greater heterogeneity in the relationship between response diversity measures and community instability (Fig. 3 d-i). The relationship between realised mean response and the OEV showed greater variation with increasing interaction strength and a shift towards higher mean response for maximum stability (cf. Fig. 3d,g with Fig. 3a). While these results mirrored closely the analyses of the fundamental mean response (see above), the same comparison for response diversity metrics showed qualitative changes for metrics based on realised metrics. For realised response dissimilarity, the relationship turned monotonic with increasing interaction strengths (Fig. 3b,e,h). Low response dissimilarity resulted in large negative OEV, increasing dissimilarity led to high variation in OEV with negative, positive, or OEV close to zero, while high dissimilarity resulted in large positive OEV (Fig. 3e,h). For response divergence, the V-shaped relationship found in simulations without interactions (Fig. 3c) gradually disappeared with increasing interaction strength (Fig. 3f,i). 3.2 Meta-Analysis In the meta-analysis, we only had access to realised responses and lacked information on the mean interspecific interaction strengths, as the original studies typically did not include monocultures or pairwise interaction data. Hence, we only assessed the relationship between realised response diversity and the mean response with community instability. Similar to the simulations, community instability was significantly related to the mean response, showing the same monotonic relationship (Fig. 4a; Appendix Table S1; slope = 0.90 \(\pm\) SE = 0.40, p 0.05). Largest (positive or negative) OEV occurred at low response divergence, otherwise response diversity was not related to community OEV. 4. DISCUSSION Our results highlight the limited role of response diversity for community stability under pulse disturbances. This does not contradict its successful application in understanding stability under environmental fluctuations (Leary and Petchey, 2009; Ross et al., 2023) but emphasises that these mechanisms do not translate easily to other types of environmental change. The stability of both simulated and empirical communities was consistently determined by mean response, largely independent of interspecific interaction strength, rather than response diversity (rejecting H1). Community stability was related to the mean response regardless of whether it was derived from realised or fundamental responses (partly accepting H2). High response diversity resulted in high community stability in the absence of species interactions. However, the relationship between response diversity and community stability weakened with increasing interaction strength, with a stronger effect on realised responses than on fundamental responses (accepting H3). Combining a meta-analysis and model simulations, our study provides mechanistic insights into the stabilising role of response diversity for community stability in the context of pulse disturbances. 4.1 Response diversity in a pulse disturbance context Our combined model simulations and meta-analysis of experimentally disturbed communities with unknown interaction strengths revealed that, contrary to expectations, response diversity did not determine community stability under pulse disturbance (Figs. 2, 3). This contrasts with findings from fluctuating environments, where high response diversity typically enhances stability (Leary and Petchey, 2009; Ross et al., 2023). Fluctuating environments are characterised by recurrent positive and negative deviations from a long-term mean (Harris et al., 2018; Ives and Carpenter, 2007). Here, high response diversity promotes stability as diverse species responses increase the likelihood of including species with environmental optima both above and below the mean, facilitating compensatory dynamics (Mori et al., 2013). As such, in fluctuating environments diverse species responses enable a community to buffer environmental changes and maintain functioning across varying conditions (Downing et al., 2008; Yachi and Loreau, 1999). Pulse disturbances, however, differ fundamentally in nature. Defined by their abrupt onset and short duration, they shift environmental conditions in a single direction (e.g., a temporary drop in temperature) before conditions recover (Donohue et al., 2016). Under such disturbances, species with optima above the mean increase response diversity but are disproportionately negatively affected, ultimately reducing community stability. In a pulse disturbance context, stability thus relies less on diverse responses and more on the presence of a few species with high resistance or rapid recovery, allowing the community to maintain functionality if there is functional redundancy (Hillebrand and Kunze, 2020). Rather than by response diversity, community stability was consistently determined by the mean species response to pulse disturbance for both simulated and empirical communities. Stability required a balance in species responses—not just variation, but symmetry around zero, i.e. equal variation above and below zero. This balance can emerge when all species show no response to the disturbance (low dissimilarity/divergence), all species respond differently (high dissimilarity/divergence) (Appendix Figs. S2,3), or a mixture of these occurs, so long as responses are centred on zero. Similarly, one other study found that species coexistence did not depend on the variation in species responses but on the mean response across species, when analysing the size of the feasibility domain in response to pollution and warming for simulations and empirical macroinvertebrate communities (De Laender et al., 2023). Moreover, a recent study turned this ‘imbalance’ in species responses into a metric and showed that in fluctuating environments higher balance in species responses (lower imbalance) is related to greater temporal stability (Polazzo et al., 2025). Consequently, response diversity alone is not sufficient for stability; rather, the distribution and symmetry of species responses relative to zero play a crucial role. The results of the meta-analysis largely reflected the results of the simulations for realised responses, suggesting that the observed patterns are independent of the introduced disturbances but might occur because of the nature of the disturbance type (pulse). While our model simulations introduced a single pulse disturbance, comprising of a drop in temperature, the meta-analysis covers a wide range of disturbances such as pollution, species removal, drought, or flood and therefore represents an ideal test case. Moreover, other measures of stability obtained from empirical data, that is resistance, temporal variability, and recovery (Hillebrand et al., 2018), mirrored the patterns found for OEV (Appendix Fig. S4), reinforcing that the OEV reflects the multidimensional nature of stability (Urrutia‐Cordero et al., 2021). 4.2 The influence of interspecific interaction strength Increasing interaction strength weakened the relationship between both response diversity and the mean species response with community stability, with a stronger influence on realised than on fundamental responses (Figs. 3,4; Appendix Figs. S5,6). In the absence of species interactions, increasing response diversity increases the likelihood of including species with optimal traits that allow species persistence for the given disturbance (Yachi and Loreau, 1999). Once interactions are introduced, species no longer respond solely based on their intrinsic traits, as interspecific interactions shape their realised responses (Lajaaiti et al., 2024). As a result, even species with optimal traits may experience reduced biomass or be outcompeted, weakening the previously strong link between response diversity and stability, because of lower asynchrony in species responses under pulse disturbance. Indeed, increasing interaction strength led to more negative realised species responses, as species with higher T opt and higher competitiveness showed greater biomass loss (Appendix Fig. S7). On the other hand, species with lower T opt and higher competitiveness gained even more biomass because of their competitive advantage under pulse disturbance (temperature drop). This competitive release allowed biomass losses in some species to be offset by biomass gains in others, leading to compensatory dynamics (Gonzalez and Loreau, 2009; Micheli et al., 1999). As increasing interaction strength led to greater biomass losses of sensitive species and pronounced reductions in emergent community functions, a higher mean species response, that is higher biomass production, is required to compensate for these losses and maintain community stability. In other words, compensatory dynamics do not result from high diversity in species responses alone but require the ability of some species to maintain or increase biomass production under increasing interspecific interactions (Ernest and Brown, 2001). Other studies in fluctuating environments have shown that response diversity is more important than species interactions for community stability (Ives et al., 1999; Polazzo et al., 2025; Ruiz‐Moreno et al., 2024). Indeed, the importance of response diversity for community stability may be determined by the intensity of environmental change (Mori et al., 2013). When environmental changes are extreme, and particularly when coupled with species extinctions, response diversity may become more important for community stability than species interactions. Combined with our results, this highlights that the stabilising mechanisms of response diversity are context-dependent and vary between different disturbance types. 4.3 Limitations and recommendations for future studies Here, we show the limitations of response diversity to explain community stability in response to pulse disturbances. In such pulse disturbance contexts, there is often a single favourable response that allows species persistence (Arnoldi et al., 2019; Hillebrand and Kunze, 2020), whereas in fluctuating environments, maintaining stability over time requires high response diversity (Ross et al., 2023; Ross and Sasaki, 2023). Given the increasing importance of extreme events with proceeding climate change (Bathiany et al., 2018; Harris et al., 2018), we see a pressing need to further explore the role of response diversity for disturbances besides fluctuating environments. Future studies should systematically compare different disturbance types (e.g., pulse, press, and fluctuations) to better understand the generality of response diversity as a stability mechanism. Given the apparent differences between realised and fundamental response diversity measures with increasing interaction strength, we show here the benefit of assessing response diversity and the mean response from both, species fundamental responses in isolation and their realised responses in community. For communities with unknown interspecific interactions, it is difficult to interpret patterns that emerge as communities change when exposed to disturbances. To better understand community stability to disturbance, a holistic approach including both species realised responses and fundamental responses of species in isolation is advisable. This requires bottom-up experiments that build up the level of complexity and allow explicit exploration of the role of interspecific interactions in shaping community stability. 5. CONCLUSION Here we show that the importance of response diversity for community stability is context dependent and differs depending on the prevailing strength of interspecific interactions. Although the mean species response, rather than existing metrics of response diversity, determined community stability in our study, this does not mean that a diverse set of species responses is not important for maintaining community stability under external forcing. Rather, it is not enough to have variation in responses, but there needs to be equal variation above and below zero for communities with no interactions, and slightly higher mean responses with increasing interaction strengths to buffer the negative effects on community function in the face of pulse disturbance. Indeed, there is often only one good response that allows species to persist through pulse disturbances (Arnoldi et al., 2019; Hillebrand and Kunze, 2020). To ensure ecosystem resilience to global change it is therefore essential to maintain as much biodiversity as possible (Dee et al., 2019, 2017; Ives and Carpenter, 2007). Tables Table 1: Overview of model parameters and variables. Species level variables and parameters \(\mathbf{r}_{\mathbf{i}}\left(\mathbf{t}\right)\) Realised growth rate of species i. \(=r_{\text{mi}}[1-\frac{N_{i}\left(t\right)+\sum_{j\ \neq i}{\alpha_{\text{ij}}N_{j}(t)}}{K_{i}}]\) (De Mazancourt et al., 2013) \(\mathbf{N}_{\mathbf{i}}\left(\mathbf{t}\right)\) The biomass of species i at time t. \(\mathbf{r}_{\mathbf{m,i}}\) Intrinsic rate of natural increase. defined by the birth and death rate Birth rate minus death rate. \(\mathbf{b}_{\mathbf{0,i}}\left(\mathbf{T}\right)\) Temperature-dependent birth rate with Gaussian function. \(b_{0,i}\left(T\right)=a_{b}e^{\frac{{-\left(T-b_{opt,i}\right)}^{2}}{s_{i}}}\) (Vasseur, 2020; Eq 14.5) \(\mathbf{b}_{\mathbf{opt,i}}\) Temperature optimum of birth rate of species i Drawn from a uniform distribution with mean \(\mathbf{b}_{\mathbf{opt,mean}}\mathbf{\ }\)and range \(\mathbf{b}_{\mathbf{opt,range}}\) \(\mathbf{a}_{\mathbf{b,i\ }}\) Birth rate at optimum temperature 0.3; same for all species in all simulated communities. \(\mathbf{s}_{\mathbf{\text{i\ }}}\) Width of temperature response curve of birth rate. 10; same for all species in all simulated communities. \(\mathbf{d}_{\mathbf{0,i}}\left(\mathbf{T}\right)\) Temperature-dependent death rate with Arrhenius function. \(d_{0,i}\left(T\right)=a_{d}e^{\text{zT}}\) (Vasseur, 2020; Eq 14.5) \(\mathbf{a}_{\mathbf{d,i\ }}\) Constant of temperature-death rate function 0.1; same for all species in all simulated communities. (Vasseur, 2020) \(\mathbf{z}\) Slope of temperature-death rate function. 0.05; same for all species in all simulated communities. \(\mathbf{K}_{\mathbf{i}}\) Carrying capacity of species i . \(K_{i}=\frac{r_{\text{mi}}}{\beta+\delta}\) (Vasseur, 2020; Eq 14.6) \(\mathbf{\beta,\ \delta}\) Density dependent constants for species carrying capacity. 0.001; same for all species in all simulated communities. Community level parameters \(\mathbf{\alpha}_{\mathbf{\text{ij}}}\) The strength of interspecific interaction between species i and j . Absolute value of draws from a normal distribution with mean 0 and standard deviation \(\mathbf{\alpha}_{\mathbf{\text{ij}}}\mathbf{\_sd}\) \(\mathbf{\alpha}_{\mathbf{\text{ij}}}\mathbf{\_sd}\) Strength of interspecific competition in a community. Ranging from 0-0.5. Low values create a community with weak competitive interactions, high values create a community with strong interactions. \[\mathbf{b}_{\mathbf{opt,mean}}\] Mean interspecific value of optimum temperature. Controls response diversity of a community. Ranging from 15-22 °C. Low values make a community of species with low temperature optima. \[\mathbf{b}_{\mathbf{opt,range}}\] Range of interspecific optimum temperatures. Controls response diversity of a community. Ranging from 3-7 °C. Low values make a community of species all the similar temperature optima. S Species richness 10; same for all communities T Temperature Control = 22°C Pulse = 15°C Table 2 : Overview and interpretation of response diversity metrics, i.e. dissimilarity and divergence, and the mean species response. Response Dissimilarity The overall magnitude of response diversity among species’ responses \[\ {}^{q}D^{Z}\left(p\right)=\ \sum{(p_{i}{{(Zp)}_{i}^{q-1}}^{\ })}^{\frac{1}{1-q}}\] where q - sensitivity parameter that was set to zero (Ross, Petchey, et al. 2023), \(p_{i}\ \)– species responses, i.e. fundamental responses (IGR_effect) and realised responses (AUC of effect size RR) Z - dissimilarity matrix and \({(Zp)}_{i}\ \)- the ordinariness of the ith species within the community where \({(Zp)}_{i}=\ \sum_{j=1}^{s}{Z_{\text{ij}}p_{i}}\) 1<y<2 y = 1, all species respond identically y = 2, species responses are as dissimilar as possible Equation (1) in Leinster and Cobbold (2012), with modifications as described Ross et al., 2023 Response Divergence The extent to which species declines are offset by gains \[\frac{\left(\max\left[x\right]-\min\left[x\right]\right)-||\max\left[x\right]\left|-\right|min[x]||}{(\max\left[x\right]-\min\left[x\right])}\] Where x - species responses, i.e. fundamental responses (IGR_effect) and realised responses (AUC of the effect size RR) at a given environmental state relative to an undisturbed control. This is similar to calculating the first derivative of a performance–environment relationship 𝑓(𝐸) (Ross et al., 2023) 0<y<1 y = 0 responses are identical, that is all response values are negative or all are positive y = 1 responses are dissimilar, and the most negative value is equal in magnitude to the most positive value (Ross et al., 2023) Mean response The mean response of species in a community to the disturbance \[\ \frac{\sum_{i=1}^{n}p_{i}}{n}\] \({\text{where\ }p}_{i}\ \)– species responses, i.e. fundamental responses (IGR_effect) and realised responses (AUC of the effect size RR) \(y\mathbb{\in R}\) y0 species are positively affected on average Figure legends Fig. 1: Calculation of response diversity metrics, dissimilarity and divergence, and the mean species response of fundamental species responses and realised responses. Based on species fundamental reaction norms along an environmental gradient, species responses to a disturbance can be assessed as the effect on their intrinsic growth rate (IGR) (first row). In comparison, species realised responses in the community can be assessed as the area under the curve (AUC) of the response ratio (RR), given by the difference in species responses to the disturbance (D biomass ) relative to the control (C biomass ). Response diversity can be assessed using the two consecutive measures: response dissimilarity and divergence of fundamental species responses in isolation (fundamental response diversity) and their realised responses in the community (realised response diversity). The mean of species fundamental and realised responses gives an idea on the average effect of the disturbance on species, respectively. Fig. 2: Community instability, calculated as the OEV, as function of the mean fundamental response, fundamental response dissimilarity, and fundamental response divergence for communities without interactions (a-c), intermediate interaction strength (d-f), and high interaction strength (g-i). Community instability decreases for balanced mean fundamental responses (a) and increasing fundamental response diversity (b, c), however this relationship weakened with increasing interaction strength (d-i). Each point represents one model community (subset of n= 2025 communities), average strength of species interactions increases from top to bottom row. Fig. 3: Community instability, measured as OEV, as a function of mean realised response, realised response dissimilarity, and realised response divergence for no interactions (a-c), intermediate interaction strength (d-f), and high interaction strength (g-i). Community instability was low for balanced mean realised response (a) and increasing realised response diversity (b, c), but only without species interactions. Each point represents one model community (subset of n= 2025 communities), average strength of species interactions increases from top to bottom row. Fig. 4: Community instability, measured as OEV, as a function of mean realised species responses (a), realised response dissimilarity (b), and realised response divergence (c). Community instability was low when realised responses were balanced (a), but was unrelated to realised response diversity metrics (b, c). Each point represents one community in the meta-analysis (n = 98). References 1. Arnoldi, J., Loreau, M., Haegeman, B., 2019. The inherent multidimensionality of temporal variability: how common and rare species shape stability patterns. Ecol Lett 22, 1557–1567. https://doi.org/10.1111/ele.13345Baskett, M.L., Fabina, N.S., Gross, K., 2014. Response Diversity Can Increase Ecological Resilience to Disturbance in Coral Reefs. The American Naturalist 184, E16–E31. https://doi.org/10.1086/676643Bathiany, S., Dakos, V., Scheffer, M., Lenton, T.M., 2018. Climate models predict increasing temperature variability in poor countries. Sci. Adv. 4, eaar5809. https://doi.org/10.1126/sciadv.aar5809Bhaskar, R., Arreola, F., Mora, F., Martinez-Yrizar, A., Martinez-Ramos, M., Balvanera, P., 2017. Response diversity and resilience to extreme events in tropical dry secondary forests. Forest Ecology and Management 426, 61–71. https://doi.org/10.1016/j.foreco.2017.09.028Brennan, G., Collins, S., 2015. Growth responses of a green alga to multiple environmental drivers. Nature Clim Change 5, 892–897. https://doi.org/10.1038/nclimate2682De Mazancourt, C., Isbell, F., Larocque, A., Berendse, F., De Luca, E., Grace, J.B., Haegeman, B., Wayne Polley, H., Roscher, C., Schmid, B., Tilman, D., Van Ruijven, J., Weigelt, A., Wilsey, B.J., Loreau, M., 2013. Predicting ecosystem stability from community composition and biodiversity. Ecology Letters 16, 617–625. https://doi.org/10.1111/ele.12088Dee, L.E., Cowles, J., Isbell, F., Pau, S., Gaines, S.D., Reich, P.B., 2019. When Do Ecosystem Services Depend on Rare Species? Trends in Ecology & Evolution 34, 746–758. https://doi.org/10.1016/j.tree.2019.03.010Dee, L.E., De Lara, M., Costello, C., Gaines, S.D., 2017. To what extent can ecosystem services motivate protecting biodiversity? Ecol Lett 20, 935–946. https://doi.org/10.1111/ele.12790De Laender, F., Carpentier, C., Carletti, T., Song, C., Rumschlag, S.L., Mahon, M.B., Simonin, M., Meszéna, G., Barabás, G., 2023. Mean species responses predict effects of environmental change on coexistence. Ecology Letters 26, 1535–1547. https://doi.org/10.1111/ele.14278Doak, D.F., Bigger, D., Harding, E.K., Marvier, M.A., O’Malley, R.E., Thomson, D., 1998. The Statistical Inevitability of Stability‐Diversity Relationships in Community Ecology. American Naturalist 151, 264–76. https://doi.org/10.1086/286117Donohue, I., Hillebrand, H., Montoya, J.M., Petchey, O.L., Pimm, S.L., Fowler, M.S., Healy, K., Jackson, A.L., Lurgi, M., McClean, D., O’Connor, N.E., O’Gorman, E.J., Yang, Q., 2016. Navigating the complexity of ecological stability. Ecol Lett 19, 1172–1185. https://doi.org/10.1111/ele.12648Donohue, I., Petchey, O.L., Montoya, J.M., Jackson, A.L., McNally, L., Viana, M., Healy, K., Lurgi, M., O’Connor, N.E., Emmerson, M.C., 2013. On the dimensionality of ecological stability. Ecol Lett 16, 421–429. https://doi.org/10.1111/ele.12086Downing, A.L., Brown, B.L., Perrin, E.M., Keitt, T.H., Leibold, M.A., 2008. Environmental fluctuations induce scale-dependent compensation and increase stability in plankton ecosystems. Ecology 89, 3204–3214. https://doi.org/10.1890/07-1652.1Ekstrøm, C.T., 2022. MESS: Miscellaneous Esoteric Statistical Scripts.Elmqvist, T., Folke, C., Nyström, M., Peterson, G., Bengtsson, J., Walker, B., Norberg, J., 2003. Response diversity, ecosystem change, and resilience. Frontiers in Ecology and the Environment 1, 488–494. https://doi.org/10.1890/1540-9295(2003)Ernest, S.K.M., Brown, J.H., 2001. HOMEOSTASIS AND COMPENSATION: THE ROLE OF SPECIES AND RESOURCES IN ECOSYSTEM STABILITY. Ecology 82, 2118–2132. https://doi.org/10.1890/0012-9658(2001)082[2118:HACTRO]2.0.CO;2Gonzalez, A., Loreau, M., 2009. The Causes and Consequences of Compensatory Dynamics in Ecological Communities. Annu. Rev. Ecol. Evol. Syst. 40, 393–414. https://doi.org/10.1146/annurev.ecolsys.39.110707.173349Grimm, V., Wissel, C., 1997. Babel, or the ecological stability discussions: an inventory and analysis of terminology and a guide for avoiding confusion. Oecologia 109, 323–334. https://doi.org/10.1007/s004420050090Harris, R.M.B., Beaumont, L.J., Vance, T.R., Tozer, C.R., Remenyi, T.A., Perkins-Kirkpatrick, S.E., Mitchell, P.J., Nicotra, A.B., McGregor, S., Andrew, N.R., Letnic, M., Kearney, M.R., Wernberg, T., Hutley, L.B., Chambers, L.E., Fletcher, M.-S., Keatley, M.R., Woodward, C.A., Williamson, G., Duke, N.C., Bowman, D.M.J.S., 2018. Biological responses to the press and pulse of climate trends and extreme events. Nature Clim Change 8, 579–587. https://doi.org/10.1038/s41558-018-0187-9Hautier, Y., Seabloom, E.W., Borer, E.T., Adler, P.B., Harpole, W.S., Hillebrand, H., Lind, E.M., MacDougall, A.S., Stevens, C.J., Bakker, J.D., Buckley, Y.M., Chu, C., Collins, S.L., Daleo, P., Damschen, E.I., Davies, K.F., Fay, P.A., Firn, J., Gruner, D.S., Jin, V.L., Klein, J.A., Knops, J.M.H., La Pierre, K.J., Li, W., McCulley, R.L., Melbourne, B.A., Moore, J.L., O’Halloran, L.R., Prober, S.M., Risch, A.C., Sankaran, M., Schuetz, M., Hector, A., 2014. Eutrophication weakens stabilizing effects of diversity in natural grasslands. Nature 508, 521–525. https://doi.org/10.1038/nature13014Hillebrand, H., Kunze, C., 2020. Meta‐analysis on pulse disturbances reveals differences in functional and compositional recovery across ecosystems. Ecol Lett 23, 575–585. https://doi.org/10.1111/ele.13457Hillebrand, H., Langenheder, S., Lebret, K., Lindström, E., Östman, Ö., Striebel, M., 2018. Decomposing multiple dimensions of stability in global change experiments. Ecol Lett 21, 21–30. https://doi.org/10.1111/ele.12867Ives, A.R., Cardinale, B.J., 2004. Food-web interactions govern the resistance of communities after non-random extinctions. Nature 429, 174–177. https://doi.org/10.1038/nature02515Ives, A.R., Carpenter, S.R., 2007. Stability and Diversity of Ecosystems. Science 317, 58–62. https://doi.org/10.1126/science.1133258Ives, A.R., Gross, K., Klug, J.L., 1999. Stability and Variability in Competitive Communities. Science 286, 542–544. https://doi.org/10.1126/science.286.5439.542Kassambara, A., 2020. ggpubr: “ggplot2” Based Publication Ready Plots.Kunze, C., Bahlburg, D., Urrutia‐Cordero, P., Striebel, M., Kelpsiene, E., Langenheder, S., Donohue, I., Hillebrand, H., 2024. Partitioning species contributions to ecological stability in disturbed communities. Ecological Monographs e1636. https://doi.org/10.1002/ecm.1636Lajaaiti, I., Kéfi, S., Arnoldi, J.-F., 2024. How biotic interactions structure species responses to perturbations. Proc. R. Soc. B 291. https://doi.org/10.1098/rspb.2024.0930Lawton, J.H., Brown, V.K., 1994. Redundancy in Ecosystems, in: Schulze, E.-D., Mooney, H.A. (Eds.), Biodiversity and Ecosystem Function. Springer Berlin Heidelberg, Berlin, Heidelberg, pp. 255–270. https://doi.org/10.1007/978-3-642-58001-7_12Leary, D.J., Petchey, O.L., 2009. Testing a biological mechanism of the insurance hypothesis in experimental aquatic communities. Journal of Animal Ecology 78, 1143–1151. https://doi.org/10.1111/j.1365-2656.2009.01586.xLeinster, T., Cobbold, C.A., 2012. Measuring diversity: the importance of species similarity. Ecology 93, 477–489. https://doi.org/10.1890/10-2402.1Loreau, M., de Mazancourt, C., 2013. Biodiversity and ecosystem stability: a synthesis of underlying mechanisms. Ecol Lett 16, 106–115. https://doi.org/10.1111/ele.12073Micheli, F., Cottingham, K.L., Bascompte, J., Bjornstad, O.N., Eckert, G.L., Fischer, J.M., Keitt, T.H., Kendall, B.E., Klug, J.L., Rusak, J.A., 1999. The Dual Nature of Community Variability. Oikos 85, 161. https://doi.org/10.2307/3546802Mori, A.S., Furukawa, T., Sasaki, T., 2013. Response diversity determines the resilience of ecosystems to environmental change. Biological Reviews 88, 349–364. https://doi.org/10.1111/brv.12004Ockendon, N., Baker, D.J., Carr, J.A., White, E.C., Almond, R.E.A., Amano, T., Bertram, E., Bradbury, R.B., Bradley, C., Butchart, S.H.M., Doswald, N., Foden, W., Gill, D.J.C., Green, R.E., Sutherland, W.J., Tanner, E.V.J., Pearce‐Higgins, J.W., 2014. Mechanisms underpinning climatic impacts on natural populations: altered species interactions are more important than direct effects. Global Change Biology 20, 2221–2229. https://doi.org/10.1111/gcb.12559Pimm, S.L., 1984. The complexity and stability of ecosystems. Nature 307, 321–326.Polazzo, F., Hämmig, T., Petchey, O.L., Pennekamp, F., 2025. The imbalance of nature: The Role of Species Environmental Responses for Ecosystem Stability. https://doi.org/10.1101/2025.01.30.635685R Core Team, 2024. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria.Ross, S.R.P. ‐J., Petchey, O.L., Sasaki, T., Armitage, D.W., 2023. How to measure response diversity. Methods Ecol Evol 14, 1150–1167. https://doi.org/10.1111/2041-210X.14087Ross, S.R.P. ‐J., Sasaki, T., 2023. Limited theoretical and empirical evidence that response diversity determines the resilience of ecosystems to environmental change. Ecological Research 1440-1703.12434. https://doi.org/10.1111/1440-1703.12434Ruiz‐Moreno, A., Emslie, M.J., Connolly, S.R., 2024. High response diversity and conspecific density‐dependence, not species interactions, drive dynamics of coral reef fish communities. Ecology Letters 27, e14424. https://doi.org/10.1111/ele.14424Suding, K.N., Lavorel, S., Chapin, F.S., Cornelissen, J.H.C., Díaz, S., Garnier, E., Goldberg, D., Hooper, D.U., Jackson, S.T., Navas, M., 2008. Scaling environmental change through the community‐level: a trait‐based response‐and‐effect framework for plants. Global Change Biology 14, 1125–1140. https://doi.org/10.1111/j.1365-2486.2008.01557.xTilman, D., 1999. The ecological consequences of changes in biodiversity: A search for general principles. Ecology 80, 1455–1474.Tilman, D., 1996. Biodiversity: Population Versus Ecosystem Stability. Ecology 77, 350–363. https://doi.org/10.2307/2265614Urrutia‐Cordero, P., Langenheder, S., Striebel, M., Angeler, D.G., Bertilsson, S., Eklöv, P., Hansson, L., Kelpsiene, E., Laudon, H., Lundgren, M., Parkefelt, L., Donohue, I., Hillebrand, H., 2021. Integrating multiple dimensions of ecological stability into a vulnerability framework. Journal of Ecology 110, 374–386. https://doi.org/10.1111/1365-2745.13804Vasseur, D.A., 2020. The impact of temperature on population and community dynamics, in: Theoretical Ecology. Oxford University Press, pp. 243–262. https://doi.org/10.1093/oso/9780198824282.003.0014Viechtbauer, W., 2010. Conducting meta-analyses in R with the metafor package. Journal of Statistical Software 36, 1–48. https://doi.org/10.18637/jss.v036.i03Wickham, H., Averick, M., Bryan, J., Chang, W., McGowan, L.D., François, R., Grolemund, G., Hayes, A., Henry, L., Hester, J., Kuhn, M., Pedersen, T.L., Miller, E., Bache, S.M., Müller, K., Ooms, J., Robinson, D., Seidel, D.P., Spinu, V., Takahashi, K., Vaughan, D., Wilke, C., Woo, K., Yutani, H., 2019. Welcome to the tidyverse. Journal of Open Source Software 4, 1686. https://doi.org/10.21105/joss.01686Wilke, C.O., 2020. cowplot: Streamlined Plot Theme and Plot Annotations for “ggplot2.”Yachi, S., Loreau, M., 1999. Biodiversity and ecosystem productivity in a fluctuating environment: The insurance hypothesis. Proc. Natl. Acad. Sci. U.S.A. 96, 1463–1468. https://doi.org/10.1073/pnas.96.4.1463 Crossref Google Scholar Information & Authors Information Version history V1 Version 1 18 April 2025 Peer review timeline Published Ecology Letters Version of Record 31 Dec 2025 Published Copyright This work is licensed under a Non Exclusive No Reuse License. Collection Ecology Letters Keywords biodiversity compensatory dynamics meta-analysis perturbation population resilience response diversity variability Authors Affiliations Charlotte Kunze 0000-0002-1130-7417 [email protected] University of Oldenburg View all articles by this author Owen Petchey 0000-0002-7724-1633 University of Zurich View all articles by this author Shyamolina Ghosh 0000-0002-5137-9933 University of Zurich View all articles by this author Helmut Hillebrand 0000-0001-7449-1613 University of Oldenburg View all articles by this author Metrics & Citations Metrics Article Usage 574 views 391 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Charlotte Kunze, Owen Petchey, Shyamolina Ghosh, et al. Species interactions determine the importance of response diversity for community stability to pulse disturbances. Authorea . 18 April 2025. DOI: https://doi.org/10.22541/au.174498373.31851247/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.174498373.31851247/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'a00c9cbd5eb0f047',t:'MTc3OTYyODk3OA=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();