Search

We are in a climate and ecological emergency, where climate change and direct anthropogenic interference with the biosphere are risking abrupt and/or irreversible changes that threaten our life-support systems. Efforts are underway to increase the resilience of some ecosystems that are under threat, yet collective awareness and action are modest at best. Here, we highlight the potential for a biosphere resilience sensing system to make it easier to see where things are going wrong, and to see whether deliberate efforts to make things better are working. We focus on global resilience sensing of the terrestrial biosphere at high spatial and temporal resolution through satellite remote sensing, utilizing the generic mathematical behaviour of complex systems—loss of resilience corresponds to slower recovery from perturbations, gain of resilience equates to faster recovery. We consider what subset of biosphere resilience remote sensing can monitor, critically reviewing existing studies. Then we present illustrative, global results for vegetation resilience and trends in resilience over the last 20 years, from both satellite data and model simulations. We close by discussing how resilience sensing nested across global, biome-ecoregion, and local ecosystem scales could aid management and governance at these different scales, and identify priorities for further work. This article is part of the theme issue ‘Ecological complexity and the biosphere: the next 30 years’.

Drylands occupy 45% emerged lands on Earth, are home of more than 2 billion people and are extremely vulnerable to climate change. Aridity increases is expected to influence the structure and functioning of drylands in a non-linear fashion. Yet, the prevalence and drivers of these abrupt changes in ecosystem structure and function remain poorly studied. We especially lack investigations of the changes of dynamical properties of these systems and how these dynamical properties relate to aridity. Those are key to understand the real menace of experiencing abrupt shifts with aridity increases in the near future.Here we used remote sensing tools to acquire dynamical trajectories of normalized vegetation indices (NDVI, surrogates of plant fractional cover) for more than 50,000 dryland sites. With this information we conducted analysis using machine learning processes to examine the relationship of aridity with some key dynamical properties of dryland ecosystems, including several aspects of resilience (ability to withstand fluctuations without changing the functioning of ecosystems), dynamical drivers of productivity, complexity of dynamical trajectories and abruptness of productivity changes.By doing so we provide a comprehensive assessment of aridity thresholds on dynamical properties of dryland productivity that show clear vulnerability of certain zones of the Earth exhibiting critical aridity thresholds previously identified through space. In particular, we show accumulation of abrupt shifts on aridity values characteristic of transition areas from semiarid to arid ecosystems. Furthermore, these values exhibit also nonlinear shifts in resilience of ecosystems and on the identity of key dynamical drivers. Our work paves the way to expand the incidence of aridity threshold from spatial to temporal implications, and highlights the necessity of developing strategies to protect and monitor especially vulnerable areas affecting more than one fifth of emerged lands.

The persistence and functionality of forest ecosystems are highly dependent on their resilience to the ongoing rapid changes in climate conditions and in natural and anthropogenic pressures. Experimental evidences of a sudden increase in tree mortality across different biomes are rising concerns about the ongoing changes in forest resilience. However, how forest resilience, which is the capacity to withstand and recover from perturbations, is evolving in response to global changes is not yet explored. Here, we integrate satellite-based vegetation indices with machine learning to show how forest resilience, quantified in terms of critical slowing down indicators, has changed over the period 2000-2020. We show that tropical, arid and temperate forests are experiencing a significant decline in resilience, likely related to the increase in water limitations and climate variability. On the contrary, boreal forests show an increasing trend in resilience, likely for the benefits of climate warming and CO2 fertilization in cold biomes, which may outweigh the adverse effects of climate change. These patterns emerge consistently in both managed and intact forests corroborating the existence of common large-scale climate drivers. Reductions in resilience are statistically linked to abrupt declines in forest productivity, occurring in response to a slow drifting toward a critical resilience threshold. We estimate that about 22% of intact undisturbed forests, corresponding to 3.32 Pg C of GPP, have already reached such critical threshold and are experiencing a further degradation in resilience. Altogether, these signals reveal a widespread and increasing instability of global forests and should be accounted for in the design of land-based mitigation and adaption plans.

Context: Changes in ecosystem resilience have been recently studied on various scales using remote sensing data, revealing various regions exhibiting decreasing resilience. However, the drivers of these changes have not been identified yet. Our study aims at filling this gap by exploring the factors that have caused the resilience of ecosystems to change during the last two decades. Methods: We investigate changes in vegetation resilience at the planetary scale, by quantifying two complementary aspects of resilience, namely sensitivity and autocorrelation, which are respectively associated with resistance and recovery abilities of ecosystems. We use a machine learning approach to identify the main environmental, climatic, and anthropogenic drivers of changes in resilience between two periods (the period 2000-2010 vs that of 2010-2020). Results: We find that in 26% of ecosystems worldwide, vegetation exhibits signs of resilience loss, and that the changes in climate conditions as well as the ecosystem’s intrinsic properties (aridity, elevation, anthropization) affect the way vegetation resilience has changed over time. Different biomes (forest, grasslands, and savannas) exhibit similar responses to their changing environment. Regions experiencing intense warming (>0.2ºC/decade) have shown a major loss in vegetation resilience. Decreasing productivity is associated with reduced resilience, and interacts with warming, exacerbating resilience loss of degraded lands. This shows that global warming and human activities are major drivers of losses in vegetation resilience across vegetation types. Conclusions: We reveal a decline in the capacity of a number of ecosystems to withstand perturbations, which should be accounted for in the management of vulnerable areas. Our results raise concerns about the persistence of ecosystems due to projected warming and expected intensification of human activities.

Forest ecosystems depend on their capacity to withstand and recover from natural and anthropogenic perturbations (that is, their resilience)1. Experimental evidence of sudden increases in tree mortality is raising concerns about variation in forest resilience2, yet little is known about how it is evolving in response to climate change. Here we integrate satellite-based vegetation indices with machine learning to show how forest resilience, quantified in terms of critical slowing down indicators3–5, has changed during the period 2000–2020. We show that tropical, arid and temperate forests are experiencing a significant decline in resilience, probably related to increased water limitations and climate variability. By contrast, boreal forests show divergent local patterns with an average increasing trend in resilience, probably benefiting from warming and CO2 fertilization, which may outweigh the adverse effects of climate change. These patterns emerge consistently in both managed and intact forests, corroborating the existence of common large-scale climate drivers. Reductions in resilience are statistically linked to abrupt declines in forest primary productivity, occurring in response to slow drifting towards a critical resilience threshold. Approximately 23% of intact undisturbed forests, corresponding to 3.32 Pg C of gross primary productivity, have already reached a critical threshold and are experiencing a further degradation in resilience. Together, these signals reveal a widespread decline in the capacity of forests to withstand perturbation that should be accounted for in the design of land-based mitigation and adaptation plans.

Species formation is a central topic in biology, and a large body of theoretical work has explored the conditions under which speciation occurs, including whether speciation dynamics are gradual or abrupt. In some cases of abrupt speciation, differentiation slowly builds up until it reaches a threshold, at which point linkage disequilibrium (LD) and divergent selection enter a positive feedback loop that triggers accelerated change. Notably, such abrupt transitions powered by a positive feedback have also been observed in a range of other systems. Efforts to anticipate abrupt transitions have led to the development of “early warning signals” (EWS), that is, specific statistical patterns preceding abrupt transitions. Examples of EWS are rising autocorrelation and variance in time-series data due to the reduction of the ability of the system to recover from disturbances. Here, we investigate whether speciation dynamics in theoretical models also exhibit EWS. Using a model of genetic divergence between two populations, we search for EWS before gradual and abrupt speciation events. We do so using six different metrics of differentiation: the effective migration rate, the number of selected loci, the mean fitness of our studied population, LD, FST, and Dabs, a metric analogous to DXY. We find evidence for EWS, with a heterogeneity in their strength among differentiation metrics. We specifically identify FST and the effective migration rate as the most reliable EWS of upcoming abrupt speciation events. Our results provide initial insights into potential EWS of impending speciation and contribute to efforts to generalize the mechanisms underlying EWS.

Global change encompasses many co-occurring anthropogenic drivers, which can act synergistically or antagonistically on ecological systems. Predicting how different global change drivers simultaneously contribute to observed biodiversity change is a key challenge for ecology and conservation. However, we lack the mechanistic understanding of how multiple global change drivers influence the vital rates of multiple interacting species. We propose that reaction norms, the relationships between a driver and vital rates like growth, mortality, and consumption, provide insights to the underlying mechanisms of community responses to multiple drivers. Understanding how multiple drivers interact to affect demographic rates using a reaction-norm perspective can improve our ability to make predictions of interactions at higher levels of organization-that is, community and food web. Building on the framework of consumer-resource interactions and widely studied thermal performance curves, we illustrate how joint driver impacts can be scaled up from the population to the community level. A simple proof-of-concept model demonstrates how reaction norms of vital rates predict the prevalence of driver interactions at the community level. A literature search suggests that our proposed approach is not yet used in multiple driver research. We outline how realistic response surfaces (i.e., multidimensional reaction norms) can be inferred by parametric and nonparametric approaches. Response surfaces have the potential to strengthen our understanding of how multiple drivers affect communities as well as improve our ability to predict when interactive effects emerge, two of the major challenges of ecology today.

Declines in European bird populations are reported for decades but the direct effect of major anthropogenic pressures on such declines remains unquantified. Causal relationships between pressures and bird population responses are difficult to identify as pressures interact at different spatial scales and responses vary among species. Here, we uncover direct relationships between population time-series of 170 common bird species, monitored at more than 20,000 sites in 28 European countries, over 37 y, and four widespread anthropogenic pressures: agricultural intensification, change in forest cover, urbanisation and temperature change over the last decades. We quantify the influence of each pressure on population time-series and its importance relative to other pressures, and we identify traits of most affected species. We find that agricultural intensification, in particular pesticides and fertiliser use, is the main pressure for most bird population declines, especially for invertebrate feeders. Responses to changes in forest cover, urbanisation and temperature are more species-specific. Specifically, forest cover is associated with a positive effect and growing urbanisation with a negative effect on population dynamics, while temperature change has an effect on the dynamics of a large number of bird populations, the magnitude and direction of which depend on species' thermal preferences. Our results not only confirm the pervasive and strong effects of anthropogenic pressures on common breeding birds, but quantify the relative strength of these effects stressing the urgent need for transformative changes in the way of inhabiting the world in European countries, if bird populations shall have a chance of recovering.

Managing ecological communities requires fast detection of species that are sensitive to perturbations. Yet, the focus on recovery to equilibrium has prevented us from assessing species responses to perturbations when abundances fluctuate over time. Here, we introduce two data-driven approaches (expected sensitivity and eigenvector rankings) based on the time-varying Jacobian matrix to rank species over time according to their sensitivity to perturbations on abundances. Using several population dynamics models, we demonstrate that we can infer these rankings from time-series data to predict the order of species sensitivities. We find that the most sensitive species are not always the ones with the most rapidly changing or lowest abundance, which are typical criteria used to monitor populations. Finally, using two empirical time series, we show that sensitive species tend to be harder to forecast. Our results suggest that incorporating information on species interactions can improve how we manage communities out of equilibrium.

Lakes are often described as sentinels of global change. Phenomena like lake eutrophication, algal blooms, or reorganization in community composition belong to the most studied ecosystem regime shifts. However, although regime shifts have been well documented in several lakes, a global assessment of the prevalence of regime shifts is still missing, and, more in general, of the factors altering stability in lake status, is missing. Here, we provide a first global assessment of regime shifts and stability in the productivity of 1,015 lakes worldwide using trophic state index (TSI) time series derived from satellite imagery. We find that 12.8% of the lakes studied show regime shifts whose signatures are compatible with tipping points, while the number of detected regime shifts from low to high TSI has increased over time. Although our results suggest an overall stable picture for global lake dynamics, the limited instability signatures do not mean that lakes are insensitive to global change. Modeling the interaction between lake climatic, geophysical, and socioeconomic features and their stability properties, we find that the probability of a lake experiencing a tipping point increases with human population density in its catchment, while it decreases as the gross domestic product of that population increases. Our results show how quantifying lake productivity dynamics at a global scale highlights socioeconomic inequalities in conserving natural environments., Proceedings of the National Academy of Sciences, 119 (35), ISSN:0027-8424, ISSN:1091-6490

International audience; Regular vegetation patterns have been predicted to indicate a system slowing down and possibly desertification of drylands. However, these predictions have not yet been observed in dryland vegetation due to the inherent logistic difficulty to gather longerterm in situ data. Here, we evaluate the theoretical prediction that regular vegetation patterns are associated with empirically derived temporal indicators (autocorrelation, variance, responsiveness) of critical slowing down in a dryland ecosystem in Sudan using different remote sensing products. We use recently developed methods using remote-sensing EVI time-series in combination with classified regular vegetation patterns along a rainfall gradient in Sudan to test the predicted slowing down. We tested our empirical findings against theoretical predictions from a stochastic version of a spatial explicit model that has been used to describe vegetation dynamics in drylands under aridity stress. Overall, three temporal indicators (responsiveness, temporal autocorrelation, variance) show slowing down as vegetation patterns change from gaps to labyrinths to spots towards more arid conditions, confirming predictions. However, this transition exhibits non-linearities, specifically when patterns change configuration. Model simulations reveal that the transition between patterns temporarily slows down the system affecting the temporal indicators. These transient states when vegetation patterns reorganize thus affect the systems resilience indicators in a non-linear way. Our findings suggest that spatial self-organization of dryland vegetation is associated with critical slowing down, but this transition towards reduced resilience happens in a non-linear way. Future work should aim to better understand transient dynamics in regular vegetation patterns in dryland ecosystems, because long transients make regular vegetation patterns of limited use for management in anticipating critical transitions.

International audience

Datasets and documentation of all analytical procedures from a scientific paper, Anticipating the occurrence and type of critical transitions, that will be formally published inScience Advances. Authors: Florian Grziwotz+, Chun-Wei Chang+, Vasilis Dakos, Egbert van Nes, Markus Schwarzländer, Oliver Kamps, Martin Heßler, Isao Tokuda, Arndt Telschow*, Chih-hao Hsieh* +Equal contribution; *Correspondence to: Arndt Telschow (telschow@uni-muenster.de) and Chih-hao Hsieh (chsieh@ntu.edu.tw) Terms of Use for the datasets (I) You must acknowledge the use of content. (II) Monitoring data is made available for use in activities of a non-profit nature only. (III) Users must contact the database administrator listed in Table S2 before using the dataset for any publications, including conference presentations as well as handouts and presentation materials for meetings such as committees and councils.Co-authorship may be required for some publications depending on the way the time series data is to be used. Note for analyzing empirical dataset: This R file demonstrates the DEV analysis applied in the following five datasets (Fig. 2A-O) (1) Microcosm experiment data provided by Dr. Egbert H. van Nes (Fig. 2A-C; microcosm.txt).The complete dataset analyzed in the original paper (Veraart et al. 2012 Nature) has been published on online DANS repository, doi:10.17026/dans-ztg-93aw (2) Voice experiment data are provided by Dr. Isao Tokuda (Fig. 2D-F; voice.txt) (3) Cellular ATP experiment data are offered by Dr. Markus Schwarzlander (Fig. 2G-I; mitochondria.txt) (4) Greenhouse earth climate data is publicly available and were downloaded from the World Data Center for Paleoclimatology, National Geophysical Data Center, Boulder, Colorado (http:// www.ncdc.noaa.gov/paleo/data.html).This dataset has also been used for computing the other EWS in a previous study (Dakos et al. PNAS 2008)(Fig. 2J-L; climate.txt) (5) Frequency data provided by Bonneville Power Administration (Fig. 2M-O; blackout.txt). Note for analyzing synthetic dataset: To significantly reduce the computation time running the analysis including 100 replicates, the analyses presented in the R code for producing Fig. 1, 3, and S14 are based on one time series replicate per mathematical model.Due to process noise (as well as observation error), the DEV results produced from analyzing one replicate look more variable and change less smoothly.However, the findings based on analyzing one replicate still reveal the same dynamical behaviours qualitatively (e.g., increasing trend of |DEV|)and quantitatively (e.g., |DEV|->1 at the bifurcation point). To obtain the results presented in the main text, the R-file can be easily modified in four steps: 1. Calculate all results of this R-file 100 times. 2. Save each set of the results. 3. Average across all results associated with the same time step. 4. Plot the averaged results instead of the results based on one replicate as shown in this file.

Ecosystems under stress may respond abruptly and irreversibly through tipping points. Although much is explored on the mechanisms that affect tipping points and alternative stable states, little is known on how ecosystems with alternative stable states could have emerged in the first place. Here, we explore whether evolution by natural selection can lead to the emergence of ecosystems with tipping points using a toy model inspired by shallow lakes. Shallow lakes are the best-known example where a shift from a submersed macrophyte dominated state to a floating macrophyte dominated state in response to excess nutrient loading corresponds to a tipping point between alternative stable states. We model the evolution of the macrophyte's living depth in the lake, we identify the conditions under which the ancestor macrophyte population diversifies, and investigate whether alternate stable states dominated by different macrophyte phenotypes occur. Given the trade-off between access to light and nutrient along the water column, we show that asymmetry in competition for light is required for diversification, while alternative stable states require an additional competitive asymmetry, a priority effect to nutrient access. We find that eco-evolutionary dynamics may produce alternative stable states, but only under a restrictive range of conditions. Our analysis suggests that competitive asymmetries along opposing resource gradients may allow alternative stable states to emerge by natural selection.

Contains fulltext : 286284.pdf (Publisher’s version ) (Closed access)

International audience; Despite many sectors of society striving for sustainability in environmental management, humans often fail to identify and act on the connections and processes responsible for social-ecological tipping points. Part of the problem is the fracturing of environmental management and social-ecological research into ecosystem domains (land, freshwater, and sea), each with different scales and resolution of data acquisition and distinct management approaches. We present a perspective on the social-ecological connections across ecosystem domains that emphasize the need for management reprioritization to effectively connect these domains. We identify critical nexus points related to the drivers of tipping points, scales of governance, and the spatial and temporal dimensions of social-ecological processes. We combine real-world examples and a simple dynamic model to illustrate the implications of slow management responses to environmental impacts that traverse ecosystem domains. We end with guidance on management and research opportunities that arise from this cross-domain lens to foster greater opportunity to achieve environmental and sustainability goals.

The constant provision of plant productivity is integral to supporting the liability of ecosystems and human wellbeing in global drylands. Drylands are paradigmatic examples of systems prone to experiencing abrupt changes in their functioning. Indeed, space-for-time substitution approaches suggest that abrupt changes in plant productivity are widespread, but this evidence is less clear using observational time series or experimental data at a large scale. Studying the prevalence and, most importantly, the unknown drivers of abrupt (rather than gradual) dynamical patterns in drylands may help to unveil hotspots of current and future dynamical instabilities in drylands. Using a 20-y global satellite-derived temporal assessment of dryland Normalized Difference Vegetation Index (NDVI), we show that 50% of all dryland ecosystems exhibiting gains or losses of NDVI are characterized by abrupt positive/negative temporal dynamics. We further show that abrupt changes are more common among negative than positive NDVI trends and can be found in global regions suffering recent droughts, particularly around critical aridity thresholds. Positive abrupt dynamics are found most in ecosystems with low seasonal variability or high aridity. Our work unveils the high importance of climate variability on triggering abrupt shifts in vegetation and it provides missing evidence of increasing abruptness in systems intensively managed by humans, with low soil organic carbon contents, or around specific aridity thresholds. These results highlight that abrupt changes in dryland dynamics are very common, especially for productivity losses, pinpoint global hotspots of dryland vulnerability, and identify drivers that could be targeted for effective dryland management., Proceedings of the National Academy of Sciences of the United States of America, 119 (43), ISSN:0027-8424, ISSN:1091-6490

International audience; Pairs of plants and pollinators species sometimes consistently interact throughout time and across space. Such consistency can be interpreted as a sign of interaction fidelity, that is a consistent interaction between two species when they co-occur in the same place. But how common interaction fidelity is and what determines interaction fidelity in plant–pollinator communities remain open questions.We aim to assess how frequent is interaction fidelity between plants and their pollinators and what drives interaction fidelity across plant–pollinator communities.Using a dataset of 141 networks around the world, we quantify whether the interaction between pairs of plant and pollinator species happens more (‘interaction fidelity’) or less (‘interaction avoidance’) often than expected by chance given the structure of the networks in which they co-occur. We also explore the relationship between interaction fidelity and species' degree (i.e. number of interactions), and the taxonomy of the species involved in the interaction.Our findings reveal that most plant–pollinator interactions do not differ from random expectations, in other words show neither fidelity nor avoidance. Out of the total 44,814 co-occurring species pairs we found 7,877 unique pair interactions (18%). Only 551 (7%) of the 7,877 plant–pollinator interactions did show significant interaction fidelity, meaning that these pairs interact in a consistent and non-random way across networks. We also find that 39 (0.09%) out of 44,814 plant–pollinator pairs showed significant interaction avoidance. Our results suggest that interactions involving specialist species have a high probability to show interaction fidelity and a low probability of interaction avoidance. In addition, we find that particular associations between plant and insect orders, as for example interactions between Hymenoptera and Fabales, showed high fidelity and low avoidance.Although niche and neutral processes simultaneously influence patterns of interaction in ecological communities, our findings suggest that it is rather neutral processes that are shaping the patterns of interactions in plant–pollinator networks.

Indicators to predict ecosystem state change are urgently needed to cope with the degradation of ecosystem services caused by global change. With the development of new technologies for measuring ecosystem function with fine spatiotemporal resolution over broad areas, we are in the era of 'big data'. However, it is unclear how large, emerging datasets can be used to anticipate ecosystem state change. We propose the construction of indicators based on functional variables (flows) and state variables (pools) to predict future ecosystem state changes. The indicators identified here may be useful signals for doing so. In addition, functional indicators have explicit ecological meanings that can identify the ecological mechanism that is causing state changes, and can thus be used to improve ecosystem models.

There is growing concern about the dire socioecological consequences of abrupt transitions between alternative ecosystem states in response to environmental changes. At the same time, environmental change can trigger evolutionary responses that could stabilize or destabilize ecosystem dynamics. However, we know little about how coupled ecological and evolutionary processes affect the risk of transition between alternative ecosystem states. Using shallow lakes as a model ecosystem, we investigate how trait evolution of a key species affects ecosystem resilience under environmental stress. We find that adaptive evolution of macrophytes can increase ecosystem resilience by shifting the critical threshold, which marks the transition from a clear-water state to a turbid-water state to a higher level of environmental stress. However, following the transition, adaptation to the turbid-water state can delay the ecosystem recovery back to the clear-water state. This implies that restoration could be more effective when implemented early enough after a transition occurs and before organisms adapt to the alternative state. Our findings provide new insights into how to prevent and mitigate the occurrence of regime shifts in ecosystems and highlight the need to understand ecosystem responses to environmental change in the context of coupled ecological and evolutionary processes.

Regular vegetation patterns have been predicted to indicate a system slowing down and possibly desertification of drylands. However, these predictions have not yet been observed in dryland vegetation due to the inherent logistic difficulty to gather longer-term in situ data. Here, we use recently developed methods using remote-sensing EVI time-series in combination with classified regular vegetation patterns along a rainfall gradient in Sudan to test these predictions. Overall, three temporal indicators (responsiveness, temporal autocorrelation, variance) show slowing down as vegetation patterns change from gaps to labyrinths to spots towards more arid conditions, confirming predictions. However, this transition exhibits non-linearities, specifically when patterns change configuration. Model simulations reveal that the transition between patterns temporarily slows down the system affecting the temporal indicators. These transient states when vegetation patterns reorganize thus affect the systems resilience indicators in a non-linear way. Our findings suggest that spatial self-organization of dryland vegetation is associated with critical slowing down, but this transition towards reduced resilience happens in a non-linear way. Future work should aim to better understand transient dynamics in regular vegetation patterns in dryland ecosystems, because long transients make regular vegetation patterns of limited use for management in anticipating critical transitions.

Ecological communities and other complex systems can undergo abrupt and long‐lasting reorganization, a regime shift, when deterministic or stochastic factors bring them to the vicinity of a tipping point between alternative states. Such changes can be large and often arise unexpectedly. However, theoretical and experimental analyses have shown that changes in correlation structure, variance, and other standard indicators of biomass, abundance, or other descriptive variables are often observed prior to a state shift, providing early warnings of an anticipated transition. Natural systems manifest unknown mixtures of ecological and environmental processes, hampered by noise and limited observations. As data quality often cannot be improved, it is important to choose the best modeling tools available for the analysis.We investigate three autoregressive models and analyze their theoretical differences and practical performance. We formulate a novel probabilistic method for early warning signal detection and demonstrate performance improvements compared to nonprobabilistic alternatives based on simulation and publicly available experimental time series.The probabilistic formulation provides a novel approach to early warning signal detection and analysis, with enhanced robustness and treatment of uncertainties. In real experimental time series, the new probabilistic method produces results that are consistent with previously reported findings.Robustness to uncertainties is instrumental in the common scenario where mechanistic understanding of the complex system dynamics is not available. The probabilistic approach provides a new family of robust methods for early warning signal detection that can be naturally extended to incorporate variable modeling assumptions and prior knowledge., The article is concerned with early warning signals in one‐dimensional dynamical systems. We present a new probabilistic method for detecting rising lag‐1 autocorrelation and benchmark it against previously published approaches.

1. Climate warming is causing changes in the physics of deep lakes, such as longer summer stratification, increased water column stability, reduced ice cover, and a shallower depth of winter overturns. An ultimate consequence of warming would be a transition to a different mixing regime. Here we investigate the role of physical, chemical, and biological feedback mechanisms that unfold during a shift in mixing regime, and whether these feedbacks could prompt and stabilise the new regime. Although climate, interannual temperature variation, and lake morphometry are the main determinants of a mixing regime, when climate change causes shifts in mixing regime, internal feedback mechanisms may gain in importance and modify lake ecosystem functioning. 2. We review the role of these feedbacks in three mixing regime shifts: from polymictic to seasonally stratified, from dimictic to monomictic, and from holomictic to oligomictic or meromictic. 3. Polymictic lakes of intermediate depth (c. 3–10 m mean depth) could experience seasonal stratification if a stratification event triggers phytoplankton blooms or dissolved organic matter release, reducing transparency and therefore further heating the surface layer. However, this feedback is only likely to have influence in small and clear lakes, it would be easily disturbed by weather conditions, and the resulting stratified state does not remain stable in the long term, as stratification is lost in winter. 4. The ice‐albedo feedback might cause an accelerated shift from ice‐covered (dimictic) to ice‐free (monomictic) winters in sufficiently deep (mean depth 50 m or more) lakes, where temperature memory is carried over from one winter to the next. Nevertheless, there is an ongoing debate into whether this process can persist during natural weather variations and overcome self‐stabilising mechanisms such as thermal insulation by snow. The majority of studies suggest that a gradual transition from dimictic to monomictic is more likely than an abrupt transition. 5. A shift from a holomictic to a meromictic regime can occur if anoxia is triggered by incomplete mixing and an increase in deep‐water density—through the accumulation of solutes—exceeds a density decrease by hypolimnetic warming. A shift to meromixis would strongly alter the biology of a lake and might be difficult to reverse. If solutes accumulate only minimally in the hypolimnion, an oligomictic regime is formed, in which years with complete and incomplete mixing alternate. 6. Understanding the importance of feedback mechanisms and the role of biogeochemistry when lakes shift in mixing regime could lead to a better understanding of how climate change affects lake ecosystems.

Growing evidence of global declines in pollinator abundance and diversity has raised concerns about the resilience of pollination systems. When subjected to stressors, each nested component of the pollination system (communities, populations, and colonies) can respond in either a smooth linear fashion, or in an abrupt nonlinear manner. Threshold and tipping point responses to stress are of particular concern because they result in sudden changes with little warning; such changes may lead to persistent non-functional states that are difficult to reverse. Here, we review evidence for threshold and tipping point responses at the colony, population and community levels of the pollination system. We find that while there are strong theoretical reasons to expect tipping point and threshold responses at all three levels of the pollination system, evidence in the field is lacking for all levels except the colony level. While this is encouraging, caution is still warranted as tipping point and threshold responses—by their very nature—may not be apparent until they are underway. Moreover, we propose that the interaction of nonlinear stress responses across different levels of the pollination system can increase the risk of cascading failures. We therefore suggest a cautious approach toward the management of pollination systems. Since environmental change will almost certainly continue to accelerate, understanding the potential for thresholds, tipping points and cascading failures is key to safeguarding global pollination systems.

AimThe impact of global change on biodiversity is commonly assessed in terms of changes in species distributions, community richness and community composition. Whether and how much associations between species, i.e. the degree of correlation in their spatial co-occurrence, are also changing is much less documented and mostly limited to local studies of ecological networks. In this study, we quantify changes in large-scale patterns of species associations in bird communities in relation to changes in species composition.LocationFrance.Time period2001-2017.Major taxa studiedCommon breeding birds.MethodsWe use network approaches to build three community-aggregated indices reflecting complementary aspects of species association networks. We characterise the spatio-temporal dynamics of these indices using a large-scale and high-resolution dataset of bird co-abundances of 109 species monitored for 17 years (2001-2017) from 1,969 sites across France. We finally test whether spatial and temporal changes in species association networks are related to species homogenisation estimated as the spatio-temporal dynamics of β-diversity and the proportion of habitat generalists. The consistency of these relationships is tested across three main habitats, namely woodland, grassland and human settlements.ResultsWe document a directional change in association-based indices in response to modifications in β-diversity and in the proportion of generalists in space and time. Weaker associations and sparser networks were related to lower β-diversity and a higher proportion of generalists, suggesting an overlooked aspect of biotic homogenisation affecting species associations. We report that this overall pattern is not constant across habitats, with opposite relationships between biotic homogenisation and change in species association networks in urban versus forest communities suggesting distinct homogenisation processes.Main ConclusionsAlthough species association contain only partial signatures of species interactions, our study highlights that biotic homogenisation translates to finer changes in community structure by affecting the number, strength and type of species associations.

Temporal dynamics in ecological variables are usually assessed using linear trends or smoothing methods. Those trends qualitatively summarise the increase or decrease in the variable of interest over a given time period. Yet, linear trends do not capture changes in the direction or in the rate of change of indices such as population trajectories, that may typically occur when conditions improve or worsen following conservation actions or environmental disturbances. In a similar way, non-linear methods while aiming to fully characterise population trajectories, fail to end up with a standard classification. Here, we propose and test a simple method to classify the trajectory of a given ecological variable according to its trend and acceleration. Our method is based on fitting a second order polynomial that is used to describe a trajectory according to its direction, velocity, and curvature (accelerated or decelerated). We apply this method to the temporal dynamics of bird populations monitored by the French Breeding Bird Survey as a case study. Using data for more than 100 species monitored from 1989 to 2017 in more than 2000 sites, we show that one quarter of the studied species have dynamics that can be better described by our polynomial approach than typically-used linear analysis. We also show how it can be used to analyse indicators constructed with multi-species indices. Our method can be applied to any type of ecological variable either to classify trajectories of ecological variables in time or trajectories of ecological variables across pressure gradients. Overall, our results suggest a more systematic investigation of non-linear trajectories when analysing the dynamics of ecological variables.

Changing conditions may lead to sudden shifts in the state of ecosystems when critical thresholds are passed. Some well‐studied drivers of such transitions lead to predictable outcomes such as a turbid lake or a degraded landscape. Many ecosystems are, however, complex systems of many interacting species. While detecting upcoming transitions in such systems is challenging, predicting what comes after a critical transition is terra incognita altogether. The problem is that complex ecosystems may shift to many different, alternative states. Whether an impending transition has minor, positive or catastrophic effects is thus unclear. Some systems may, however, behave more predictably than others. The dynamics of mutualistic communities can be expected to be relatively simple, because delayed negative feedbacks leading to oscillatory or other complex dynamics are weak. Here, we address the question of whether this relative simplicity allows us to foresee a community's future state. As a case study, we use a model of a bipartite mutualistic network and show that a network's post‐transition state is indicated by the way in which a system recovers from minor disturbances. Similar results obtained with a unipartite model of facilitation suggest that our results are of relevance to a wide range of mutualistic systems., While detecting upcoming critical transitions in complex ecosystems is challenging, predicting what comes after a critical transition is terra incognita altogether. Here, we address the question of whether the relative simplicity of the dynamics of complex mutualistic communities allows us to foresee a community's future state.

Stochastic ecological dynamics result from both transient and asymptotic features of the underlying system, yet explanations for observed patterns often emphasize asymptotics. For example, an ecological state (e.g., a particular population size or community composition) that occurs frequently and/or persists for a meaningful duration might be assumed to be a stable equilibrium, even though transients can also persist for a long time and may recur. In this paper, we consider one particular pattern—a bimodal distribution of states as a system is observed through time—and consider alternative causes for this pattern. First, we consider the “asymptotic” explanation that each mode corresponds to a distinct stable state. Second, we consider that one mode might correspond to a long transient. We explore the dynamics that result from each of these causes in a classic bistable model, focusing particularly on the degree of environmental stochasticity needed to generate a bimodal distribution of states in each case. Our results highlight that observations of a system’s dynamics do not provide enough information to determine the number and location of stable states. We conclude that a more serious and systematic consideration for the possible role of transients in driving observed dynamics will lead to stronger insights and understanding.

International audience; Understanding the stability of ecological communities is a matter of increasing importance in the context of global environmental change. Yet, it has proved to be a challenging task. Different metrics are used to assess the stability of ecological systems, and the choice of one metric over another may result in conflicting conclusions. Although each of the multitude of metrics is useful for answering a specific question about stability, the relationship among metrics is poorly un-derstood. Such lack of understanding prevents scientists from devel-oping a unified concept of stability. Instead, by investigating these relationships we can unveil how many ‘dimensions’ of stability there are (i.e. in how many independent components stability metrics can be grouped), which should help building a more comprehensive con-cept of stability. Here, we simultaneously measured 27 stability met-rics frequently used in ecological studies. Our approach is based on dynamical simulations of multispecies trophic communities under different perturbation scenarios. Mapping the relationships between the metrics revealed that they can be lumped into three main groups of relatively independent stability components: ‘early response to pulse’, ‘sensitivities to press’ and ‘distance to threshold’. Selecting metrics from each of these groups allows a more accurate and com-prehensive quantification of the overall stability of ecological com-munities. These results contribute to improving our understanding and assessment of stability in ecological communities.

River catchments are highly complex systems characterized by several properties such as self-organization, multi-scale variability, hydraulic and topographic gradients, patchiness and heterogeneity, resilience and a hierarchical structure. These features, coupled with several geomorphological, anthropogenic and climatic drivers, are expected to influence the surface water composition over different temporal and spatial scales. The knowledge of these complex interlinks plays a key role in both river basin management and predictability to potential pollution events. Nevertheless, due to the considerable amount of factors involved in the analysis, the unique combination of attributes characterizing each catchment and the lack of data at an adequate scale, it still remains unclear which of the environmental parameters have a major influence on the water chemistry. In this work, the hierarchy of the variability in the chemical composition of 160 water samples collected in 2017 throughout the Tiber River Basin, the largest catchment in Central Italy (17,156 km2), was explored. The results obtained by using advanced statistical methods, including the Compositional Data Analysis, highlighted different sources of variability linked to the geological (low variability) and anthropogenic origin (high variability) of the main solutes. Furthermore, for each sampling site, the corresponding watershed was calculated from the Digital Terrain Model using a Geographical Information System-based elaboration. The aim was to evaluate the relationships between the landscape morphological properties of the watersheds, such as elevation, drainage area, slope or other morphometric indexes and the physical-chemical parameters of the river waters on the basis of different geological and topographical settings of the basin. The outcomes proved to be particularly useful to discriminate between water chemistry mainly influenced by surface run-off processes and that affected by ground water circulation.