The argument rests on a well-known mathematical pitfall. Ecological processes are overwhelmingly non-linear, and Jensen’s inequality, described in the literature as the fallacy of the average, shows that for non-linear functions the response of a system to average conditions differs from the average response to variable conditions. When ecologists aggregate individuals into species or population-level summaries and then assume equilibrium conditions, they risk drawing erroneous conclusions about how real systems respond to real, fluctuating environments. This matters more than ever because global change drivers — climate change, habitat loss, overexploitation, pollution and invasive species — often act synergistically, pushing ecosystems out of equilibrium and into transient dynamics that can last decades or even centuries.

The consequences of ignoring individual variation are not abstract. Take dispersal, the process that drives invasions, range shifts and the success of species establishment. Individual differences in dispersal ability, often linked to behavioural traits such as boldness or to genetic differences, can determine whether an invasive or colonising species establishes at all, with only specific phenotypes or genotypes managing the journey. Similarly, biodiversity responses to land-use change, landscape fragmentation and habitat loss have recently been tied to individual-level variation in behavioural and other traits. Individual responses play a pivotal role in priority effects, food-web dynamics, predator-prey and host-pathogen interactions, animal migration, community assembly and stability, nutrient supply, the formation of novel ecosystems and even the provisioning of ecosystem services such as fisheries yields.

Despite this growing evidence, few studies have causally linked individual variation to complex biodiversity dynamics and ecosystem functioning. Individual-based concepts have been developed in theoretical ecology, most notably through agent-based modelling, but these approaches have largely been confined to single species or simplified modules of pairwise interactions, with forest gap models among the few exceptions. The authors identify two main obstacles that have held back a truly integrated individual-based ecology: first, the difficulty of collecting and handling individual-specific data at scale, and second, insufficient computational power. Even disciplines where inter-individual differences are extensively documented, such as behavioural ecology, plant ecology and evolutionary biology, have rarely explored the consequences of that variation for higher ecological levels, often remaining focused on single species or pairwise interactions.

What has changed, the researchers contend, is technology. Recent massive advances in sensors and analytical tools now enable high-resolution data collection on a much larger scale, across multiple species, taxonomic groups and ecosystems. High-throughput animal tracking systems can monitor the physiology and behaviour of interacting individuals in real time. Multi-omics and meta-barcoding approaches can identify intraspecific and cryptic diversity and assess the distribution of community traits. In situ animal personality tests, advanced remote observations capable of detecting unique patterns of individual organisms, and mesocosm and microcosm experiments of experimental evolution are providing an empirical foundation that simply did not exist before. Machine learning and other advanced analytical approaches now provide the ability to process and interpret big datasets, linking data from large numbers of interacting individuals with high-resolution environmental data and system-level information.

The conceptual framework proposed in the article is called Individual-Based Global Change Ecology, or IBGCE. It aims to capture, understand and predict the emergent response of ecological systems to key drivers of global change by taking full account of the role of individual variation. The approach applies to all forms of life, from bacteria to plants, fungi, invertebrates and vertebrates, and places particular emphasis on resilience, non-linear transitions and emergent properties of ecosystems, functions and services. Individuals, in this view, are the elementary particles of ecological systems: variation and interactions between individuals scale up to emergent properties at the population, community and ecosystem levels, and these levels are interconnected through both bottom-up and top-down feedback loops that must be elucidated through an individual-based lens.

The road ahead, according to the authors, requires a systematic combination of empirical, experimental and modelling studies across systems, while accounting for multiple drivers of global change and their interactions, whether gradual or abrupt, predictable or stochastic. Field and experimental work on trait variation — in movement and decision making in animals, or in functional traits in plants and microbial communities — coupled with data science approaches, has already allowed the development and testing of new hypotheses on emergent phenomena. The next step is to systematically investigate how individual-level variation translates into community and ecosystem-level processes across organisational, taxonomic and ecological scales, from microbes to top predators, in both terrestrial and aquatic systems, and across metapopulations, metacommunities and biodiversity-ecosystem functioning relationships.

Theory must evolve alongside the data. The article highlights recent examples of integrating individual-based concepts from movement ecology into broader biodiversity theory, incorporating eco-evolutionary dynamics into community assembly, refining foraging theory to include adaptive behaviour, and extending modern coexistence theory with the complementary concept of coviability. Benchmark models and a common toolset are needed to advance individual-based theory and cross-fertilise with whole-system experiments. Among the promising technical solutions are adiabatic approximations, a method well established in physics for studying self-organisation, which allows quantitative assessment of how individual-level parameters affect aggregated population-level parameters. Alternatively, genetic algorithms can evolve simplified, robust rules for individual decision making that reproduce observed spatial and temporal patterns, while machine learning techniques, including neural networks and deep learning, can uncover those patterns in the first place. The authors pose a fundamental question that remains open: should model parsimony be the central consideration, or does the accuracy of predictions in capturing key patterns override parsimony?

The implications extend beyond theory into conservation practice. Individual variation is rarely explicitly considered in conservation and management decisions, even though it is highly relevant to stakeholders, biodiversity conservation and ecosystem services. Large-scale harvest experiments coupled with high-resolution monitoring have shown how populations and ecosystem functions respond to trait-selective fishing, and how changes in behavioural traits can feed back to limit catchability, harvest and monitoring itself. Model-based results also highlight the importance of accounting for differences in animal behaviour in biodiversity conservation, particularly in reintroduction programmes where only limited numbers of individuals can be released. Understanding the feedbacks between individual variation and ecosystem management, the authors argue, will help refine existing approaches that still largely focus on populations, communities or aggregated landscape-level measures.

In the long term, the researchers envision the individual-based perspective reaching beyond the laboratory and the field into museums and schools, where IBGCE data, models and simulations could power interactive educational multimedia for inquiry-based learning and data-driven dialogue formats that involve the wider public in the process of acquiring knowledge. The stakes could hardly be higher. Future scenario analyses and predictions that ignore individual responses and rapid adaptations of organisms, along with the cascading eco-evolutionary processes that shape communities, run the risk of being short-sighted and potentially biased. By synthesising knowledge across systems and scales — linking ecology and evolution, aquatic and terrestrial systems, micro- and macroscales, and urban and rural landscapes — Individual-Based Global Change Ecology aims to provide a causal understanding of the mechanisms underlying eco-evolutionary dynamics, and a new foundation for adaptive and sustainable management under the non-equilibrium, non-stationary conditions that now define the Anthropocene.

Subject of Research: A proposed paradigm shift toward Individual-Based Global Change Ecology that treats individual organisms as fundamental agents of ecological change

Article Title: The need for an individual-based global change ecology

Article References: Jeltsch, F., Roeleke, M., Abdelfattah, A., Arlinghaus, R., Berg, G., Blaum, N., De Meester, L., Dittmann, E., Eccard, J. A., Fournier, B., Gaedke, U., Gallagher, C., Govaert, L., Hauber, M., Jeschke, J. M., Kramer-Schadt, S., Linstädter, A., Lucke, U., Mazza, V., … Zurell, D. (2025). The need for an individual-based global change ecology. Individual-based Ecology, 1, 1-18. https://doi.org/10.3897/ibe.1.148200

Image Credits: AI Generated

DOI: 10.3897/ibe.1.148200

Keywords: individual-based ecology, global change ecology, biodiversity loss, ecosystem degradation, individual variation, adaptive behaviour, agent-based modelling, machine learning, eco-evolutionary dynamics, conservation, Anthropocene, predictive ecology