Brain size affects the behavioural response to predators in female guppies (Poecilia reticulata)

Abstract

Large brains are thought to result from selection for cognitive benefits, but how enhanced cognition leads to increased fitness remains poorly understood. One explanation is that increased cognitive ability results in improved monitoring and assessment of predator threats. Here, we use male and female guppies (Poecilia reticulata), artificially selected for large and small brain size, to provide an experimental evaluation of this hypothesis. We examined their behavioural response as singletons, pairs or shoals of four towards a model predator. Large-brained females, but not males, spent less time performing predator inspections, an inherently risky behaviour. Video analysis revealed that large-brained females were further away from the model predator when in pairs but that they habituated quickly towards the model when in shoals of four. Males stayed further away from the predator model than females but again we found no brain size effect in males. We conclude that differences in brain size affect the female predator response. Large-brained females might be able to assess risk better or need less sensory information to reach an accurate conclusion. Our results provide experimental support for the general idea that predation pressure is likely to be important for the evolution of brain size in prey species.

Keywords: Poecilia reticulata; artificial selection; brain size; guppy; predator inspection; predator response.

Figure 1.

Parameters of inspection behaviour for the different treatments presented as boxplots, indicating the median and quartiles with whiskers reaching up to 1.5 times the interquartile range. The violin plot outlines illustrate kernel probability density, i.e. the width of the shaded area represents the proportion of the data located there. Significance is based on LMMs, see text and table 1 (*p < 0.05, **p < 0.01 and ***p < 0.001). (a) Total time spent inspecting the predator model per fish in the shoal. (b) Total number of inspections per shoal. (c) Average duration of those inspections. Note the use of log₁₀ scales.

Figure 2.

Overview of positional data during trials with a model predator, presented as statistical heat maps. First row (a) shows the overall result of each sex, second row (b) splits this up per brain size, and the third row (c) divides the brain size results for females up into the different shoal sizes. Each cell (pixel) in the heat maps shows the p-value of a non-parametric test. In the orange and blue maps, we tested against the median density to visualize the areas of the tanks that were visited more or less than expected. The pink and green maps show the statistical comparison between groups; i.e. males and females or small- and large-brained individuals. Black rectangles indicate predator position. Numbers between parentheses denote sample sizes. Note the nonlinear axis in the legend.

Figure 3.

Density distributions of the distance to the predator for (a) small- and (b) large-brained females, split by shoal size (colours). As each trial has an individual density distribution, lines represent the median of those distributions for each group with the shaded areas being bootstrapped 95% CIs. As the expected distance distribution is determined by the shape of the tank, relative density distributions are shown, where the expected distribution (when tank locations would be drawn randomly from a two-dimensional uniform distribution) has been subtracted.