Male harm arises when traits that increase reproductive success in competition with other males also harm females as a side-effect. The extent of harm depends on male and female phenotypes, both of which can diverge between populations. Within a population, harm is inferred when increased exposure to males reduces female fitness, but studies of the divergence of male harm rarely manipulate male exposure. Here, we quantify male harm and compare its magnitude between two lab populations of Drosophila serrata that were derived from a common ancestor seven years earlier and subsequently held under conditions that minimized environmental differences. We manipulate female exposure to males in a factorial design involving all four combinations of males and females from these populations, providing insight into divergence in both sexes. Our results reveal substantial harm to females and provide stronger evidence of divergence in males than in females. Using these and other published data, we discuss conceptual issues surrounding the quantification and comparison of harm that arise because it involves a comparison of multiple quantities (e.g., female fitness under varying male exposure), and we demonstrate the increased insight that is gained by manipulating male exposure to quantify these.

We compared male harm and female resistance between the two popuations of Drosophila serrata (the '2006' and '2013' University of Ottawa populations). The experiment measured the fitness of 2006 and 2013 females when held under conditions of high or low exposure to 2006 or 2013 males. Exposure was manipulated by altering the frequency with which females were exposed to males (high = continuous exposure; low = 6h exposure every 3-4 days). All combinations of male exposure (high vs. low), male population (2006 vs. 2013), and female population (2006 vs. 2013) were performed, yielding a 2 × 2 × 2 factorial design involving eight unique combinations of treatment levels. The experiment was performed in two blocks over two consecutive generations of the populations during the summer of 2014, with 30 replicates of each of the eight treatment combinations within each block, yielding 480 replicates in total.A few individuals were lost due to experimental error, resulting in 463 replicates (average of 28.9 per treatment combination, range 24-30).

Flies for use in the assay were collected as virgins using light CO₂ anaesthesia and then stored, separately by sex, in vials in groups of 10 individuals. Two days later, females were marked (to help distinguish them from males) by dusting them with red formaldehyde-free fluorescent pigment (Brilliant Group, LLC, San Francisco, CA). We then created replicate vials with one virgin female and six virgin males per vial from the appropriate populations. In the low exposure treatment, after 6 h males were removed via aspiration and discarded, while in the high exposure treatment males remained. Females were transferred to new vials with 10 mL of fresh media every 3-4 days from this point on until the last female died. Males from the high exposure treatment were discarded at each transfer and new sexually mature males from the appropriate population were added, while in the low exposure treatment new males were placed with females for 6 h before once again being discarded. Vials were checked at least once daily for dead flies; female deaths were recorded and dead males were replaced with live ones from the appropriate population. Following transfer of females to new vials, the old vials were retained and the number of adult offspring produced per vial was counted 14 days later.

Rows are replicates (i.e. single females). Female_Last.Alive and Female_First.Dead are the days at which the female was last observed alive and first observed dead, respectively, with fractional values denoting the approximate time of day that check was performed. Num_offspring columns give the number of offspring the female produced in a given vial (i.e. females were transferred to a new vail every 3-4 days). Zero values indicate that the female was alive at least some of the time while in that vial but produced no offspring, while missing values (".") indicate that that female was dead at that point. Total_Num_Offspring is simply the sum of all these values for a given female (i.e. replicate).