Coevolution describes evolutionary change in two or more interacting populations or species in which each population imposes selective pressures on the others and evolutionary responses in one participant influence subsequent selection in another. The concept is particularly relevant to interactions involving predators and prey, parasites and hosts, plants and pollinators, plants and herbivores, competitors, and mutualistic partners. Coevolution does not simply mean that two species evolve at the same time. The defining feature is reciprocal evolutionary influence through interaction.
Genetic changes in one population can alter the fitness consequences of traits in another, creating feedback between interacting evolutionary trajectories. Coevolution can produce specialisation, reciprocal adaptation, escalation, stabilising interactions, trade offs, or diversification, depending on ecological circumstances and the genetic architecture of the interacting traits.
Predator prey relationships provide a classic context for coevolution. Predators can impose selection favouring prey traits that reduce detection, capture, or consumption, while prey can impose selection favouring predator traits that improve hunting efficiency. Such interactions can produce changes in camouflage, speed, defensive structures, sensory abilities, toxins, behaviours, or hunting strategies. However, evolutionary change does not necessarily produce a simple continuous increase in performance because energetic costs and ecological trade offs constrain adaptation. A defensive structure may reduce predation but require additional resources, while a predator adaptation may improve capture of one prey type but reduce efficiency against another. Coevolution therefore often produces dynamic compromises rather than unlimited escalation.
Host parasite interactions are another major example because parasites depend on hosts while hosts evolve mechanisms that resist or tolerate infection. Host immune responses can select for parasite strategies that evade recognition, suppress immunity, change surface antigens, or alter transmission timing. Parasites can in turn impose selection favouring host genetic variants associated with resistance or tolerance. The result can produce geographically variable evolutionary dynamics because parasite populations, host populations, environments, and migration patterns differ among locations. Genetic diversity can be maintained when resistance provides advantages in some environments but imposes costs in others. Coevolutionary interactions can therefore contribute to balancing selection, local adaptation, and substantial genetic variation within populations.
Plant animal interactions can also produce reciprocal evolutionary effects. Plants may evolve nectar characteristics, floral shapes, odours, colours, flowering times, chemical defences, or structures that influence interactions with pollinators and herbivores. Animals may evolve feeding structures, sensory preferences, detoxification mechanisms, or behaviours that affect their interactions with plants. Some specialised pollination systems show close matching between floral morphology and pollinator anatomy, although apparent matching does not by itself prove strict pairwise coevolution. Many plants interact with multiple pollinators and herbivores, while animals interact with many plant species, producing diffuse rather than one to one selection. Coevolution can therefore occur within complex ecological networks rather than isolated species pairs.
Mutualisms can also generate reciprocal evolutionary change when both partners gain fitness benefits from an interaction. Pollination, seed dispersal, nutrient exchange, and symbiotic associations can create selection favouring efficient cooperation. However, mutualisms are not necessarily permanently beneficial under every condition. Environmental changes, resource availability, partner abundance, and cheating strategies can alter the costs and benefits. Selection may favour traits that increase a partner's benefit when doing so indirectly increases the focal organism's fitness, but it may also favour exploitation if benefits can be obtained without providing reciprocal services. Coevolutionary stability therefore depends on ecological and genetic conditions that influence partner choice, sanctions, rewards, dependence, and transmission.
Coevolution can occur at different spatial and temporal scales. Reciprocal selection may operate within populations, between geographically separated populations, or across broad species distributions. Geographic mosaics can arise when an interaction is strong in some locations and weak in others, producing different evolutionary trajectories across a species range. Gene flow can transfer adaptations between populations, potentially reducing or redistributing local differentiation. Temporal changes can occur because environmental conditions alter the strength of selection, while seasonal interactions can generate repeated cycles. Mathematical models and experimental evolution can investigate these dynamics by tracking allele frequencies, phenotypes, fitness, and interaction outcomes over successive generations.
Evidence for Coevolution requires more than observing that two interacting organisms possess apparently complementary traits. Researchers may seek evidence that interaction strength affects selection, that traits influence fitness, and that evolutionary change in one species alters selection in the other. Reciprocal transplant experiments, controlled feeding or infection experiments, population genetic analyses, comparative studies, and longitudinal measurements can provide stronger evidence. Correlation between traits can also result from common environmental pressures, shared ancestry, or nonreciprocal adaptation. Phylogenetic comparative methods can help determine whether trait associations remain after accounting for shared ancestry. The concept therefore requires careful testing rather than assuming that every ecological interaction is coevolutionary.
A common misconception is that Coevolution necessarily produces perfectly matched species or obligate partnerships. Many interacting organisms remain generalists, and selection can favour flexibility rather than extreme specialisation. Another misconception is that coevolution always benefits both participants. Antagonistic coevolution can involve reciprocal harm, while mutualistic interactions involve benefits that may change with ecological context. Coevolution is also not equivalent to evolutionary change caused by any environmental interaction because reciprocal selection is central to the concept. Its broader significance lies in demonstrating that evolution occurs within ecological networks, meaning that the evolutionary trajectory of a population can depend partly on the traits and responses of other organisms.
