Plasticity in climate change responses

Abstract

Recent research has shown that climate change can both induce and modulate the expression of plastic traits but our understanding of the role of phenotypic plasticity as an adaptive response to climate change is limited. In this review, we dissect the mechanisms and impact of phenotypic plasticity as a response to accumulating climatic pressures on the individual, species and community levels. (i) We discuss how plasticity can affect individuals, populations and community dynamics and how climate change can alter the role of plasticity. We hypothesise that some pathways to phenotypic plasticity such as irreversible and anticipatory organismal responses will be reduced under increasing climate change. (ii) We then propose an integrated conceptual framework for studying phenotypic plasticity to advance our understanding of the feedbacks between the different levels of biological organisation. (iii) By formulating as yet unaddressed research questions within and across levels of biological organisation, we aim to instigate new research on phenotypic plasticity and its role in climate change responses.

Keywords: climate change; community; ecological feedbacks; microbiome; molecular pathways; phenotypic plasticity; species interactions; warming.

Fig. 1

Relationships between the different aspects of phenotypic plasticity and corresponding traits (PP&T) and their possible changes caused by climate warming. Each rectangle represents the diversity of plastic traits and their quantitative expression before (blue unfilled areas) and after (orange‐filled areas) warming. The arrows between the rectangles indicate possible combinations of the different aspects of plasticity. In this hypothetical example, the overall amount of phenotypic plasticity (i.e. the number of phenotypically plastic traits and the overall magnitude of their expression, represented by the rectangle area) declines after warming (orange area is smaller) because more traits become non‐plastic and/or diminish in plasticity (lost PP&T) relative to “novel” plastic traits or increased levels of plasticity (novel PP&T) after warming. These changes and the proportion of PP&T that is not affected by warming (preserved PP&T) will likely not be uniform across the different aspects of plasticity. For example, anticipatory PP&T and irreversible PP&T can be lost significantly more than gained and the opposite may hold for responsive PP&T after warming. The silhouettes show examples of plasticity and traits that could be directly or indirectly affected by climate warming (double‐headed arrows: reversible plasticity; unidirectional arrows: irreversible plasticity). Left (from the top): reversible plasticity: huddling behaviour in penguins; active plasticity: defence formation in Daphnia; anticipatory plasticity: hibernation in toads. Right (from the top): irreversible plasticity: seasonal pigmentation in butterflies; passive plasticity: degrowth of planarians after resource depletion; responsive plasticity: insect larvae attack on plant followed by release of toxins (modified images from PhyloPic; planarian picture by Noah Schlottman https://creativecommons.org/licenses/by‐sa/3.0/).

Fig. 2

Conceptual framework integrating bottom‐up and top‐down approaches for understanding phenotypic plasticity at different levels of biological organisation. Phenotypic plasticity (curved black arrow on image on bottom row) occurs at the individual level and can be induced by abiotic environmental responses (e.g. warming illustrated by orange colours) as well as by biotic changes in population (e.g. density) and community (e.g. predation, competition, resources, parasites) state. The response to the environmental cue can be modified by the microbial community (microbiome) composition associated with the individual. Here, we highlight an example of phenotypic plasticity in response to warming, the type of plasticity response (active versus passive, reversible versus irreversible and responsive versus anticipatory) and the type of trait (morphology, behaviour, life history, learning, physiology or novel “hidden” traits) that can show the plastic response (orange arrows and rectangles at the individual level). Phenotypic plasticity at the individual level can have bottom‐up effects at the population and community level. For example, higher temperatures can result in a smaller body size in Daphnia, which can modify interactions with predatory aquatic invertebrates that prefer smaller prey (Pastorok, 1981) and consequently, for example alter prey density, with further effects on population and community level (orange arrows and rectangles connecting the individual, population and community level). Altered ecological interactions at the community or population level could further induce different phenotypic plasticity responses, studied from a top‐down approach. This illustrates the many potential ways phenotypic plasticity can result in changed interactions affecting individual‐, population‐ (e.g. intraspecific interactions, indicated by thin black double‐headed arrows on the left‐hand panel) and community‐level processes (e.g. interspecific processes, indicated by thin black double‐headed arrows on the left‐hand panel) resulting in many potential feedbacks between different levels of ecological organisation in all directions (large double‐headed arrows on left‐hand side).

Fig. 3

Proposed shift of phenotypic plasticity types relative to length of ontogeny and lifespan under climate change. Without climate change, the environmental fluctuations and associated cues are more predictable (illustrated by the blue, more regular waves) and all types of phenotypic plasticity (PP) may be used across the whole lifespan of long‐ and short‐lived organisms (arrows). Under climate change, the environmental fluctuations (in particular weather extremes) and associated cues (illustrated by orange waves with more irregular amplitude and frequency) become less predictable. The same environmental fluctuation can be perceived differently depending on the organism and the spatio‐temporal scale (i.e. resolution) at which it interacts with the environment. Larger organisms generally have longer lifespans, influencing the perceived magnitude of change. When faced with the same environmental fluctuation, smaller short‐lived animals with faster metabolism may perceive it as a major cue, whereas larger long‐lived animals with slower metabolic rates may perceive it as environmental noise or as an oscillating stressor. Therefore, a potential stressor only acquires meaning depending on the focal timescales at which the organism operates and its generation time (e.g. Einum & Burton, ; Dupont et al., 2024). Thus, in long‐lived organisms such as African elephants that live up to 70 years, have a 2‐year gestation period and mature after ca. 15 years, reliance on irreversible PP both during ontogeny and adult life could lead to maladaptation and irreversible PP traits may therefore decrease under climate change (grey text). In toads that live about 10–12 years, the tadpole development of a few weeks is more likely to fall into a predictable time window, where the anticipated environment at egg laying matches the environment during tadpole development, while the long‐lived adult will more likely face unpredictable environmental fluctuations. In this case, different types of PP will be relevant for different life stages (arrows). Anticipatory PP can remain adaptive in long‐lived animals, for example if the cue triggers a reversible response and the duration of the plastic phenotype falls into a time window with environmental conditions matching the cue (typically a short‐term, rapid behavioural or physiological response following the cue), or an induced irreversible trait is still adaptive or neutral in the novel condition. However, we predict an overall decline in anticipatory PP (grey text). Short‐lived organisms such as Drosophila flies that can complete their life cycle during a few weeks within predictable environmental conditions can usually rely on all forms of PP (arrows), including irreversible and anticipatory PP. Ultimately, the repertoire of different plastic responses used by a species will depend on the costs and benefits of each plasticity type for the organism and the context of local climate change drivers. The lifespan and stages of the different organisms are not to scale.

Fig. 4

Dynamics of a predator–prey model with an active inducible antipredator response (see Yamamichi et al., 2019) that we extended to include temperature dependence in the attack rates. The dashed vertical line indicates a distinction between the dynamics before (time < 150) and during (time > 150) environmental change; the colour intensity of the red bar above each panel indicates the severity of warming. (A) Attack rates of the predator on undefended (a _u) and defended (a _d) prey phenotypes. With gradual warming (between times 150 and 900), both attack rates gradually increase twofold. (B) Impact of warming on the dynamics without plasticity: increased overexploitation by the predator results in more severe cycles. (C, D) Impact of reversible plasticity on the dynamics, showing a stabilising effect that is counterintuitively enhanced by the effect of warming. In both C and D, prey estimate predation risk by the abundance of predators and their conspecifics (Tollrian et al., 2015) to determine whether they should express a defended or undefended phenotype. In C, the model also incorporates the increased attack rate (i.e. individuals consider the same predator density at higher temperature to constitute a higher predation risk), while in D, their choices are based only on the predator and prey abundances.

Fig. 5

Regime shift as a function of rates of trait change and of environmental stress increase. Although the risk of regime shifts increases with faster environmental change, this risk can be reduced by rapid phenotypic trait changes. Phenotypic trait changes can shift a critical threshold to higher levels of environmental stress, which increases stress tolerance. Therefore, the speed of trait change determines the rate at which the system (e.g. population, community, ecosystem) becomes more tolerant to stress, avoiding tipping into an alternative contrasting state (from Chaparro‐Pedraza, 2021).