Pierre-Olivier Montiglio

Contaminants as a neglected source of variation in behaviour (from Raphaël Royauté and me)


Raphael Royauté and I recently published a small commentary on how contaminants can alter phenotypic variation in animal populations. We think this a fascinating idea from an evolutionary standpoint and a nice opportunity to collaborate with ecotoxicologists.


The main idea is pretty simple: exposure and accumulation to a lot of contaminants (toxic byproducts of human activities such as pesticides or heavy metals) are determined by how animals acquire resources. For example, foraging patterns and diet, habitat use, or dominance and aggressiveness can all affect which contaminants an animal will encounter and how much exposure it will experience. Contaminants, in turn, affect all these resource acquisition behaviours. Food intake, activity patterns, and predation responses are all affected by organosphophate pesticides, and anxiolytics amongst other. Raphael has some very interesting work on this (read it here).



Animal behaviour is also closely associated with life history, or resource allocation strategy. For example, individuals with lower survival rates or with earlier should express risk taking behaviour to a greater extent. Contaminants surely affect many aspects of an animal’s life history, like its growth rate, survival and fecundity, and so contaminants could have additional, long lasting effects on behaviour through life history and resource allocation patterns.



These two sets of effects between behaviour and contaminant exposure or accumulation can feedback into each other. As a result, individuals in a population will experience differing rates of exposure and accumulation. Contaminants will not only affect the average behaviour expressed by a group of exposed individuals, but also affect the amount of behavioural variation that is expressed in the whole population. For example, individuals with higher foraging activity may accumulate a larger dose of contaminant over their lifespan. Whether a contaminant amplifies or masks phenotypic variation depends on how it affects behaviour, and how behaviour affects contaminant exposure (Figure 1).


Honey bee workers feed on nectar and pollen contaminated with neurotoxic insecticides. Returning foragers also expose other bees and larvae of the colony to contaminated food. Depending on their social role within the hive, individuals may be exposed to different doses of insecticides. Insecticides impair foraging activity, navigation skills, olfactory memory and learning. A feedback loop could exist between the social role or behaviour of individuals and their contamination (picture credit: en.wikipedia.org).


Contaminant-driven changes in behavioural variation may be transient. For example, at low, nonlethal doses, exposure to pesticides will affect animals for a few minutes or hours only. In some other cases, the changes in behavioural variation may be long lasting. Endocrine disrupters can have developmental effects, affecting behaviour in a permanent way. In all cases, contaminants effects on behaviour variation could have important implication for the ability of populations to respond to their environment.



Feedbacks between contaminants and behaviour are fascinating from an evolutionary ecological point of view. Think about how much phenotypic variation is actually driven by contaminants in natural populations of fishes, bees, or birds? How consistent is the effect contaminants on phenotypic variation and how does it affect the population’s response to selective pressures? Behavioural, or life history variation also have huge impacts on the interaction between the population and its community, or affect its dynamic over time.


Pharmaceuticals released into streams can affect the boldness of individuals. Bolder individuals end up accumulating more contaminants than shy co-specific, leading to strong differences in boldness among exposed individuals. Read the cool results from Brodin et al. at doi: 10.1126/science.1226850. (picture credit: en.wikipedia.org).


Contaminant – behaviour feedbacks also have implication for ecological risk assessments and ecotoxicology. It prompts us to take into account the variation in phenotype and how it is affected by contaminants when conducting toxicological assays. Toxicological assays available for the major classes of contaminants (heavy metals, pesticides, pharmaceuticals, etc) typically shows important variability in the individuals’ response to contaminants. Understanding, or even simply modelling this variability would greatly improve predictive power.



Investigating these feedbacks is surely a complex thing to do in most systems. It requires to have a mechanistic approach, assessing the effect of the contaminant on behaviour, and then the effect of behaviour on contaminant exposure and accumulation, potentially by monitoring changes in life history. Of course all of this involves following temporal variation in behaviour and contaminant levels in marked individuals. This surely is a challenge. Yet, these types of approaches are already implemented in research programs following hormones in wild populations instead of contaminants. We see these challenges as a great opportunity for collaboration between evolutionary ecologists and ecotoxicologists. Part of Raphael’s PhD thesis was focused on this topic, but we are also interested in getting in touch with researchers having an ecotoxicological background!