Source-linked AI summary
Eco-evolutionary feedbacks - theoretical models and perspectives
Lynn Govaert, Emanuel A. Fronhofer, Sébastien Lion, Christophe Eizaguirre, Dries Bonte, Martijn Egas, Andrew P. Hendry, Ayana De Brito Martins, Carlos J. Melián, Joost A. M. Raeymaekers, Irja I. Ratikainen, Bernt-Erik Saether, Jennifer A. Schweitzer, Blake Matthews
TL;DR
Feedbacks connect ecology and evolution across biological systems, but progress is constrained by subtle distinctions among thematic subdisciplines. This overview synthesizes theoretical models and their assumptions, finding that eco-evolutionary feedbacks significantly change models and that density- and frequency-dependent selection are key ingredients.
Problem
Feedbacks are central to ecology and evolutionary biology, but subtle distinctions among thematic subdisciplines hinder concerted progress.
Method
The paper provides an overview of theoretical models of eco-evolutionary feedbacks and emphasizes their underlying assumptions and modelling formalisms.
Results
Including eco-evolutionary feedbacks in theoretical models significantly changes their settings and trait correlations, while density- and frequency-dependent selection are key ingredients.
Takeaways & Limitations
Eco-evolutionary models are characterized by the combination of demographic and evolutionary formalisms, with density- and frequency-dependent selection central to feedbacks.
Takeaways & Limitations
Individual-based models can reduce generality.
Abstract
from arXiv · showhide
1. Theoretical models pertaining to feedbacks between ecological and evolutionary processes are prevalent in multiple biological fields. An integrative overview is currently lacking, due to little crosstalk between the fields and the use of different methodological approaches. 2. Here we review a wide range of models of eco-evolutionary feedbacks and highlight their underlying assumptions. We discuss models where feedbacks occur both within and between hierarchical levels of ecosystems, including populations, communities, and abiotic environments, and consider feedbacks across spatial scales. 3. Identifying the commonalities among feedback models, and the underlying assumptions, helps us better understand the mechanistic basis of eco-evolutionary feedbacks. Eco-evolutionary feedbacks can be readily modelled by coupling demographic and evolutionary formalisms. We provide an overview of these approaches and suggest future integrative modelling avenues. 4. Our overview highlights that eco-evolutionary feedbacks have been incorporated in theoretical work for nearly a century. Yet, this work does not always include the notion of rapid evolution or concurrent ecological and evolutionary time scales. We discuss the importance of density- and frequency-dependent selection for feedbacks, as well as the importance of dispersal as a central linking trait between ecology and evolution in a spatial context.
1 Introduction
Eco-evolutionary feedbacks connect ecological properties such as demography with evolutionary change and back again. This overview addresses fragmented theoretical approaches by organizing models across biological and spatial complexity.
- Ecology and evolution have often been studied in isolation, despite feedbacks being central to both fields.
- Rapid evolution on ecological timescales renewed interest in eco-evolutionary dynamics and feedbacks.
- Feedbacks can generate spatial variation in biotic interactions, affect population regulation and community dynamics, and promote species coexistence.
- The review surveys theoretical EEF work across methodological approaches and thematic subdisciplines whose distinctions are often subtle or semantic.
- It organizes the overview by community complexity and spatial complexity to summarize formalisms, assumptions, and existing theoretical work.
2 Formalisms used for modelling EEFs
Eco-evolutionary feedbacks can be modeled by coupling demographic formalisms with evolutionary approaches such as adaptive dynamics and quantitative genetics. These formalisms differ in time-scale assumptions, tractability, generality, and suitability for rapid evolution or spatial complexity.
- Adaptive dynamics: Adaptive-dynamics models assume rare mutations, ecological equilibration, and evolution through a sequence of allele substitutions measured by invasion fitness.
- Adaptive dynamics: In adaptive dynamics, feedback occurs because mutant invasion fitness depends on ecological conditions created by the resident community.
- Quantitative genetics: Quantitative-genetics models track moments of trait distributions and can couple selection gradients to ecological dynamics without requiring ecological equilibrium.
- Quantitative genetics: Quantitative-genetics models can focus on short-term dynamics, making them potentially applicable when rapid evolution matters in experiments or field studies.
- Demographic formalisms: ODEs, matrix or integral projection models, and individual-based models provide demographic frameworks that can be coupled with evolutionary formalisms.
- Demographic formalisms: Individual-based models incorporate stochasticity, spatial structure, and kin competition, but their complexity reduces generality and analytical tractability.
3 EEFs within populations
Within populations, eco-evolutionary feedbacks link density, traits, genotype or morph frequencies, and environmental conditions across temporal and spatial scales. Dispersal is a central spatial link because it connects demography with gene flow and evolution.
- Single populations: Single-species feedbacks commonly connect population size with heritable traits, while spatial models also connect local and regional demography with trait values.
- Spatially structured populations: Dispersal is central to spatial EEFs because it affects demography and mediates evolution through gene flow.
- Single populations: Density- and frequency-dependent selection allow population density or morph frequency to alter selection, while evolved competitive abilities feed back on density.
- Single populations: In constant environments, quantitative-genetics models found that evolution maximizes mean fitness, whereas fluctuating population sizes can change this outcome.
- Single populations: Evolutionary rescue models examine persistence through adaptation, whereas evolutionary suicide occurs when trait change degrades viability and causes extinction.
- Spatially structured populations: Spatial feedbacks link patch colonization and extinction with disperser frequency or mean dispersal rate, which can then alter patch dynamics.
- Spatially structured populations: Spatial evolution can produce both rescue and suicide, with dispersal changes affecting recolonization probabilities, extinction, and population-density distributions.
4 EEFs involving two species
Two-species eco-evolutionary feedbacks arise through competition, predator-prey interactions, and host-parasite dynamics. Models link interspecific or epidemiological densities to trait evolution and back to ecological dynamics, with outcomes including divergence, altered cycles, stability changes, and diversification.
- Interspecific competition: Competition can drive trait evolution that changes selection pressures on both species, producing ecological character displacement and potentially speciation.
- Interspecific competition: Assortative mating can generate reproductive isolation and increased diversity, while non-random mating is required for evolutionary branching.
- Interspecific competition: Landscape structure strongly influences diversity outcomes generated by competition and assortative mating in individual-based models.
- Predator-prey interactions: Predator-prey feedbacks couple predator or prey densities with defence, offense, growth, mortality, and fecundity trade-offs.
- Predator-prey interactions: Rapid evolution in prey defence can produce antiphase predator-prey cycles rather than the 1/4-lag cycles predicted by non-evolutionary models.
- Predator-prey interactions: Feedbacks can stabilize or destabilize predator-prey dynamics depending on genetic variation and trade-off shapes.
- Host-parasite interactions: Host-parasite models link susceptible or infected-host densities with virulence and resistance evolution, and high mutation rates can produce overlapping ecological and evolutionary timescales.
- Host-parasite interactions: High-dimensional host-parasite feedbacks can support pathogen diversification and evolutionary branching in host resistance.
5 EEFs in a community and ecosystem context
Eco-evolutionary feedbacks in communities and ecosystems link trait evolution with changes in soils, species interactions, community diversity, and food-web structure. Models capture these feedbacks across abiotic environments, spatial communities, and trophic networks, while differing in how evolution is represented.
- 5.1 Feedbacks between organisms and abiotic environments: Increasing the direct benefit of soil nutrient conditioning is predicted to select for higher values of soil-conditioning traits.The model assumes a genetic link between plants and soils.
- 5.1 Feedbacks between organisms and abiotic environments: Plant-soil interactions exemplify niche construction because evolving plant traits alter soil conditions, which change selection on plant traits.These feedbacks can be modeled using individual-based models or extended resource-competition models.
- 5.2 Feedbacks in communities: Evolution can maintain, increase, or decrease phenotypic, species, and functional diversity, depending on the modeled eco-evolutionary processes.Martín et al. showed that eco-evolutionary feedbacks can maintain phenotypic diversity.
- 5.2 Feedbacks in communities: In spatial communities, frequency-dependent competition generates local trait shifts that alter global trait distributions, species differentiation, and community diversity.Competition occurs between neighboring individuals, so local traits and spatial locations change selection pressures.
- 5.2 Feedbacks in communities: Climate-change feedbacks can continue generating species extinctions after climate stabilization, producing additional diversity loss.A spatially explicit individual-based model jointly represented genetic variation, dispersal, competition, species sorting, and adaptation.
- 5.3 Feedbacks in food webs: Food-web models represent feedbacks in which trait divergence changes species interactions, producing further evolution, branching, specialization, and altered network stability.Examples include trophic species clusters, evolutionary branching, increased mean abundances, and reduced temporal abundance variation.
- 5.3 Feedbacks in food webs: Eco-evolutionary food-web models remain rare, and many existing models add species from a pool rather than modeling speciation from intrinsic food-web dynamics.Modeling speciation as an ecological consequence is identified as a major challenge.
6 Discussion
Theoretical eco-evolutionary feedback models span biological fields and spatial and organizational scales, but remain fragmented by disciplinary boundaries and differing formalisms. The review identifies coupled demographic–evolutionary modelling, density- and frequency-dependent selection, dispersal, and more mechanistic integration as central themes for understanding these feedbacks.
- Including eco-evolutionary feedbacks significantly changes theoretical models’ dynamics and outcomes.
- Density- and frequency-dependent selection are key ingredients because they can emerge from ecological settings and generate feedbacks between ecology and evolution.
- Eco-evolutionary feedbacks are deeply rooted across evolutionary ecology, including predator–prey, host–parasite, speciation, branching, character-displacement, metapopulation, and niche-construction models.
- Dispersal links ecology and evolution spatially by affecting densities, mediating gene flow, and evolving itself.
- Eco-evolutionary feedbacks do not require rapid or contemporary evolution; they can also occur over longer timescales.
- The core modelling strategy couples demographic and evolutionary formalisms, although these approaches differ in timescale assumptions and sources of genetic variation.
- Future models should mechanistically represent phenotypic and genotypic variation over tens to hundreds of generations to assess whether feedbacks are time-dependent and prevalent.
- Greater realism creates a trade-off with generalism: individual-based models capture complex situations but can make mechanisms difficult to identify.
Author’s contributions
The study was conceived and led by Lynn Govaert and collaborators, with all authors contributing to the literature analysis, revisions, and manuscript development.
- Lynn Govaert, Emanuel A. Fronhofer, and Blake Matthews conceived the study.
- Lynn Govaert and Emanuel A. Fronhofer led the study.
- All authors performed the literature search and analysis.
- Lynn Govaert, Emanuel A. Fronhofer, and Blake Matthews drafted the manuscript.
- All authors contributed to revisions.
Data accessibility
All papers used for this review are cited in the text.
- All papers used for this review are cited in the text.