There is evidence that animal personality traits can have spill-over effects for sexual selection, with studies reporting that male behavioural types are associated with success during pre- and post-copulatory sexual selection. Given these links between personality and sexual traits, and the evidence that their expression can depend on an individual’s nutritional status (i.e. condition), a novel prediction is that changes in a male’s diet should alter both the average expression of personality and sexual traits, and their covariance. We tested these predictions using the guppy Poecilia reticulata, a species previously shown to exhibit strong condition dependence in ejaculate traits and a positive correlation between sperm production and individual variation in boldness. Contrary to expectation, we found that dietary restriction – when administered in mature adult males – did not affect the expression of either behavioural (boldness and activity) or ejaculate traits, although we did find that males subjected to dietary stress exhibited a positive association between sperm velocity and boldness that was not apparent in the unrestricted diet group. This latter finding points to possible context-dependent patterns of covariance between sexually selected and personality traits, which may have implications for patterns of selection and evolutionary processes under fluctuating environmental conditions.

This dataset contains measures for two behavioural traits (boldness and activity) and two ejaculate traits (sperm number and sperm velocity)  from male guppies that had been fed either high diet quantity (1.9mg per day) or low diet quantity (0.9 mg per day) for one month prior to trials. During feeding trials, opaque screens were placed between adjacent individual male tanks to avoid behavioural interactions between test subjects. To ensure that the males were stimulated to produce sperm throughout the course of the study, female guppies were placed within view of each tank, as male guppies remain sexually active in the presence of females 

Males were tested for behavioural traits at the end of the diet treatments, acorss two repeated measures assays (with one day recovery in between). Each fish was placed initially inside an uncovered circular refuge of opaque PVC (, which itself was placed within a circular arena  A video camera was secured above the arena to allow the fish’s movements to be observed without human disturbance. After a 5-minute period, during which the fish was allowed to acclimatize to its new surroundings, the door of the refuge was gently pulled away, allowing the fish to emerge. Using video footage, the time (in seconds) taken by an individual to emerge from the refuge was recorded. If the individual emerged within 10 minutes, its behaviour while moving around the arena was then observed over an additional 5-minute interval. The time (in seconds) the individual spent inactive or ‘frozen,’ as opposed to actively moving around, was measured over this period. The data for emergence time and time spent frozen were then inverted to provide final scores for ‘boldness’ and ‘activity’, respectively.

Ejaculates were extracted under anaesthesia after the last behavioural trial. Sperm velcoity (VCL; curvilinear velocity) was measured using computer-assisted sperm analysis, and sperm number was estimated using an Improved Neubauer hemocytometer.

The readme file contains an explanation of each of the variables in the dataset and (if applicable) their measurement units.