The present analysis showed that female fallow deer adopt different tactics of mate choice in a lek and that variation in the mating tactics are associated with female experience and costs: females, who were either less experienced or incurred higher travel costs more often adopted indirect forms of mate selection when compared to adult females residing near the lek. These results address a gap in the literature: only a small number of papers in fact have dealt with female mating tactics in lekking ungulates16,30,34, since females are believed to be especially constrained by the resources used for rearing and protecting offsprings14,49,50.
The large number of visits to the lek recorded in this study for some of the females was never observed before in other fallow deer leks9 (where however yearling females were not investigated). Visits usually precede the copulation date, in particular for adult females. The mean time interval between successive visits is longer than the duration of the oestrus (2 days) but shorter than the oestrus cycle (22 days51), thus multiple visits likely represent an opportunity for male evaluation rather than mating occasions, as was previously argued9.
We found that younger females remained longer in the lek, performed more visits, were observed in a higher number of territories and longer in the vicinity of bucks than adults and returned to the lek also after copulating, fully confirming hypotheses H1c,d,h (i.e. more experienced females were expected, respectively, to perform less visits to the lek, to visit a lower number of male territories, and to return to the lek less frequently after copulation). Younger (1–2 year old) females were also the only ones to be recorded copulating more than once, however our limited sample (3 out of 29 tagged females) does not allow us to generalize this observation, and besides, a previous study found that fallow deer polyandry is not related to female age in a non-lekking population37. For younger females several visits to the lek when not in oestrus and after copulation likely imply a large energetic expenditure without an immediate gain. Overall, this apparently inefficient behaviour might be a by-product of their lack of experience and thus it can act as a learning opportunity. Further, yearling females have no fawn at heel and their movements are less constrained than the ones of adult females50. Multiple visits could favour mate quality assessment9, however younger females did not arrive earlier than older ones, albeit males were already present and available to be assessed (cf. Fig. S3) confirming hypothesis H1e (i.e. younger females were expected to arrive at the lek when mating activity has started and hence not before adult females). We might therefore deduce that yearling females are mainly interested in observing the behaviour of other females at lek and in fact they were characterised by a later copulation date, as predicted by H1f (i.e. experienced females were expected to copulate earlier). Nevertheless, we found no evidence of a higher probability of copying in younger females, contrary to hypothesis H2b (i.e. we expected that less experienced females were more likely to copy the mate choice of other females). A later copulation date of younger does could be explained by a difference in reproductive physiology between yearling and adult females52. This could explain why the same has been observed in a non-lekking population of fallow deer36. Alternatively, females of different ages might respond differently to social stimuli: females can, in fact, adjust the timing of oestrus to maximise the possibility of mating with the favourite male35. As a consequence, when compared to adult females, younger females could be ready to make their mate choice only after a longer assessment of males at the lek.
Despite the time spent at the lek, it appears that younger females were not able to select the highest rank bucks, contrary to hypothesis H3 (i.e. secondary tactics could increase the precision in mate assessment, therefore females that are more likely to use them should mate with highly successful males). Probably younger females were not able to identify high-ranking bucks because of their inexperience in assessing male physical traits or in recognizing visual and olfactory marking activities that are supposed to be an important signalling of male reproductive status53. Also in a non-lekking population, yearling females were found to mate on average with bucks of lower rank than adult ones36, indicating that this inability of younger females is widespread in fallow deer populations regardless the adopted mating system. Alternatively, younger females may choose to mate with lower rank males, because they are not able to pay the potential costs associated with mating with high quality males, such as aggression from other females, or the resources required to produce offspring from large males36. We cannot completely rule out this possibility, however the energetic expenditure associated to the large number of visits at lek and the preference for successful territories, as well as the fact that not all younger females mated with lower ranking males, suggest that our result is likely due to inexperience.
Females living far away from the lek compensate for higher costs by spending less time, visiting less often and arriving later to the lek than closer females, confirming predictions H1a (i.e. females which incurred in lower travel costs were expected to perform more visits to the lek and to arrive earlier). Despite these constraints, there is no difference between farther and near females with respect to the number of visited territories and time spent in the vicinity of bucks, thus rejecting predictions H1b (i.e. closer females were also expected to visit more male territories), as well as in copulation date. Farther females appear to use aggregation more than near females, though, as assumed by hypothesis H2a. Experimental studies showed that female fallow deer are attracted un-specifically by female groups34 and these authors argued that this behaviour is probably useful to reduce harassment by immature males and that it allows females to copy the mate choice of other does3,14. However, contrary to predictions H2b-c, we did not find evidence for a higher probability of copying or territory choice in farther females, therefore aggregation is likely to represent a cheap proxy for high ranking bucks, thus allowing the narrowing of direct assessment to only the small number of males with a harem. The final outcome is that the probability of choosing a successful male is independent of distance, suggesting that farther females are more efficient than close females, thanks perhaps to the adoption of aggregation behaviour. Surprisingly, younger females, when compared to older ones, were not more likely to be observed together with other females, contrary to hypothesis H2a, confirming that yearling females are not as efficient in mate selection as adult females. We cannot exclude that younger females are more likely to be harassed by subadult males than adult does, but if this is true it would be the evidence that aggregation was not used as a tactic to avoid harassment.
The fact that far and near females mate on average with bucks of same rank, suggests a mechanism for the origin of multiple arenas which are present in some populations9,54. Regarding the increased costs of a longer distance between the home range and the lek, these can be buffered by modifications in female behaviour. As we have shown in this paper, almost the whole population can congregate in a single lek, but we expect that when farther females are no longer able to compensate travel costs they are more likely to congregate in a different but nearer arena. We believe that in the case of Castelporziano, we are at the limit of a size where a single lek can be present; indeed the farthest females of our samples were in a distance class of 8.5 km and not every year were able to attain the lek. Since no other arenas have been identified in our study area (probably because the studied lek is able to attract the large majority of individuals) it is likely that females who did not reach the arena mated with males adopting strategies other than lekking (e.g. territory holding). Living far from lek may be detrimental for females (if we assume that comparing numerous males leads to a better mate choice) but this disadvantage could be counterbalanced by the benefits of philopatry55. In relatively larger areas, the presence of multiple arenas could be more advantageous. The critical distance threshold, though, is not expected to be the same in every study area since travel costs depend in a complex way on topography, vegetation type, physical barriers (roads, channels), presence of predators, hunting and disturbance. We also expect that variation in the condition of the study area may change, even largely, the trade-offs we have investigated.
Mate choice by females based on the selection of successful territories is widespread in lekking species (fallow deer15, Uganda kob40, blackbuck Antilope cervicapra56, black grouse27, great snipe Gallinago media57). This is because the location of a territory within the lek can be an honest signal of male quality, enabling less costly mate-sampling and potentially more accurate mate choice than direct female mate assessment27. A novel result in our study is that younger females were more likely to select a successful territory (territory choice) than older females confirming hypothesis H2c. Yearlings could use this tactic to increase the probability of mating with a high quality buck but this study has shown that such an approach is on average unsuccessful, probably because of the rapid turnover of bucks on the better territories. Thus, the increase in male reproductive variance in the post-Peak period (confirming a larger consensus of female choice later in the season, prediction H1g) can be explained by the adoption of territory choice by yearlings, which mate later than adult females.
Contrary to hypothesis H2b we were unable to detect a different amount of copying in younger or in farther females. This observation had been unexpected because previous studies25,41,58 have suggested that copying can be an effective tactic for cost-reducing and increasing precision. This finding concords with34 who found aggregation, but not copying, in female fallow deer under experimental conditions (oestrus females were offered the choice between small paddocks containing males with groups of females and males without females, or males that they had seen mating and males that they had not seen to mate). The review by Vakirtzis59 evidences that the presence of copying is variable in lekking species. Pruett-Jones58 used a theoretical game to show that it is convenient for females to copy each other’s mating decisions only when the costs of searching are not negligible. Maybe this is not the case at Castelporziano, an undisturbed area with no obstacles to animals’ movement. Gibson and Höglund25 argued that the benefits of imitation increase if the individuals that mate first are the ones with the greatest experience, and this is why we have originally hypothesized a more frequent copying in younger females. Losey et al.41 have shown that the benefit of copying increases with the number of “peeks” at lek and indeed younger females stay longer at the lek than adults. However, younger females can be deterred from copying mate choice by a rapid male turnover in territories. On the other hand, farther females cannot spend enough time in the lek to get enough peeks. A further argument which can explain the lack of differences in copying between near and farther females is males’ sperm depletion, more likely in bucks who have mated earlier25. Insemination failure is a problem especially for farther females, which would incur higher costs than near ones if obliged to come back to the lek later.
Our results suggest, but do not prove, that adult females residing near the lek probably perform a direct choice more than farther females. Indeed they stayed longer, visited more often and arrived earlier as predicted by H1. We can explain this pattern assuming that direct choice is especially complex and requires a careful assessment of both physical features and behaviour of bucks. In other species, direct choice is used more than indirect tactics. This is the case of the great snipe57 and of the non-lekking pied flycatcher Ficedula hypoleuca60. Since fallow deer exhibit a strong plasticity in the mating system61 which also depends on population density62 we can expect that direct choice may vary among populations and be more frequent in low density populations.
In general, we cannot say to what extent our results can be applied to other lekking populations of fallow deer, since relying on only one study area can introduce biases due to, e.g., geographical characteristics or management of the population, which could affect both individual spatial behaviour and demography. However, the strong behavioural difference between young and adult females found here is probably present in different populations as well, despite differences in mating system, ecology, and management.
This study demonstrated the condition-dependent variation in female mating decisions in an ungulate lek, a crucial issue to improve the understanding of mammalian mating systems63 and already shown in a wide range of taxa39. Our results can thus contribute to further clarifying the basis of the co-evolution of mating strategies in both sexes and, eventually, the evolution of leks. A main limitation of this study was that we were unable to account for the constraints determined by male behaviour on female tactics (for instance, we did not considered the effect of herding on female choice), an argument which could be the subject of further investigations.
Source: Ecology - nature.com
