To lek or not to lek: mating strategies of male fallow deer

To lek or not to lek: mating strategies of male fallow deer

We studied the mating system of fallow deer {Dama dama) for 6 years in central Italy. Males in this population could defend territories that were either single, clumped in leks, or satellite to leks. The most highly successful males in our study were in leks. When we considered all males, there were no significant differences in average copulatory success according to territory type because many lek males did not achieve any copulations, which were seen in only a few lek territories. The variance in copulatory success, however, was much greater for leks than elsewhere. Single territories were occupied for shorter times during the rut than lek territories. Fighting among males was more frequent in the lek, even when we excluded highly successful lek males from the analysis. Chases of nonterritorial males and harem size were correlated with the number of copulations achieved by individual males, but did not vary according to territory type. Copulatory success of some individuals increased with age, but there were no age differences among males holding different types of territories. Satellite males switched to lek territoriality in the course of one rut, but switches from single territory to lek territory were rare. We suggest that males in single territories are inferior competitors that select a low-risk, lowbenefit strategy, whereas those in lek territories where no copulations were seen may be attempting to establish themselves on the lek to increase their copulatory success in future years. [Behav Ecol 3:25-31 (1992)]

W

hen two or more male mating strategies elaphus) are fertilized by dominant haremare found within the same population, holding males than previously estimated from evolutionary theory predicts that each indibehavioral observations, and males that do not vidual should choose the strategy that will give hold harems appear to be almost completely him, on average, the greatest reproductive unsuccessful (Pemberton et al., 1992). success. Choice of strategy should depend Gosling and Petrie (1990) suggested that in principally on the male's competitive ability topi {Damaliscus korrigoni) the largest males and the strategies chosen by other males (Rudefend single territories, smaller males cluster benstein, 1980). in leks, and some males on single territories Among ungulates, the evidence for alterenjoy greater mating success than most lek native mating strategies is limited. In prongmales. The situation appears to be opposite horn {Antilocapra americana), the age distri- in Uganda kob {Kobus kob), where males debution of males in a population may determine fend either single, resource-based territories whether the mating system will include terri- or lek territories (Buechner, 1961). Balmford tory defense (Byers and Kitchen, 1988). In (1990) reported that in this species the mawaterbuck {Kobus ellipsiprymnus), territorial jority of females mated on leks, and males and satellite males are found within the same defending single territories had low copulapopulation (Wirtz, 1981). One problem is the tory success. Similarly, Clutton-Brock et al. difficulty of determining paternity from be(1988) estimated that within an enclosed pophavioral observations. It is often obvious that ulation of fallow deer {Dama dama), only 5% subordinate males attempt to copulate (Hogg, of matings occurred off the lek, although re1988), but their success at fertilizing estrous source territories often attracted large numfemales is unknown. Without this informa- bers of females. Thirgood (1991), on the other tion, the relative costs and benefits of alter- hand, suggested that nonlekking males may native mating strategies can be difficult to as- have accounted for up to 34% of copulations sess. DNA fingerprinting, for' example, in one rut in his study population. suggests that more female red deer {Cervus In fallow deer, ejaculations are readily iden-

Apollonio et al. • Fallow deer mating strategies

Marco Apollonio Istituto Nazionale di Biologia della Selvaggina "A. Ghigi," Via Ca' Fornacetta 9, 40064 Ozzano Emilia (BO), Italy Marco Festa-Bianchet Large Animal Research Group, Department of Zoology, University of Cambridge, Cambridge CB2 3EJ, UK Franco Mari Via Lanino 8, 20144 Milano, Italy Stefano Mattioli Dipartimento di Biologia e Fisiologia Generali, Sezione Zoologia, Viale delle Scienze, 43100 Parma, Italy Benedetta Sarno Dipartimento di Scienze del Comportamento Animale e dell'Uomo, Via Volta 6, 56126 Pisa, Italy Centro di Faunistica Tropicale CNR, Via Romana 17, 50125 Firenze, Italy

Address reprint requests to M. Apollonio, who is now at the Dipartimento di Scienze del Comportamento Animale e dell'Uomo, Via Volta 6, 56126 Pisa, Italy. M. Festa-Bianchet is now at the Department de Biologie, Universite de Sherbrooke, Sherbrooke, Quebec, J1K2R1, Canada Received 27 November 1990 Revised 21 March 1991 Second revision 12 May 1991 Accepted 7 June 1991 1045-224925-31/92/54.00 © 1992 International Society for Behavioral Ecology

25

tifiable, mating sites are fixed in space, and most females copulate once per breeding season (Clutton-Brock et al., 1988). Behavioral observations should therefore provide a better estimate of paternity than in species where ejaculations are not easily distinguished from nonejaculatory mounts or where multiple copulations by females are common. Understanding the alternative mating options available to males in lekking species may provide clues to the evolution of lek breeding. In fallow deer the presence within the same population of males adopting different mating strategies (Apollonio, 1990; Clutton-Brock et al., 1988; Langbein and Thirgood, 1989) offers the opportunity to test hypotheses about the reasons for the persistence of these strategies. For example, Gosling et al. (1987) suggested that lekking in topi had both high costs and high benefits so that some males may choose alternative strategies with lower costs. Long-term data are needed to test different evolutionary explanations for the persistence of alternative mating strategies (Dunbar, 1982). Here we report a 6-year study of alternative mating strategies of male fallow deer in central Italy. We first describe the different strategies, then compare the relative costs and benefits of defending single territories, lek territories, and lek satellite territories. We show that both benefits (copulations) and costs (fights, harassment by nonterritorial males) are greater on leks than on single territories, whereas lek satellites are somewhat intermediate. We suggest that a male's choice of mating strategy is affected by the female's apparent preference to mate in a few specific territories in leks.

the first and last copulation seen each year. Defense time: total daylight hours spent defending territories by individual males during the rut. Fight: aggressive interaction among adult males including one or more antler clashes. Chase: aggressive interaction where a territorial male approaches and displaces a nonterritorial male. Harem size: average number of females seen defended by a territorial male during scans at 30-min intervals. During the rut, males defended either single territories or territories on one of three leks in the study area. Two leks had satellites within 200 m but out of sight of the lek. Satellites were located along deer trails leading to the lek and occupied by up to three territorial males at a time. Males on satellites tried to intercept females coming from or going to the lek and switched between defending territories there and on the lek several times in one rut. We checked each lek and single territory at least twice during each rut to determine whether it was defended, either by sighting the defending male or by looking for fresh evidence of trampling and newly made rutting pits (Espmark and Brunner, 1974). In 1986, we checked all known single territories every 3 days for the presence of a defending male. Behavioral observations reported in this paper were conducted at lek 1 (650 h in 19851988), six single territories (359 h in 19871989), and one satellite (217 h in 1987-1988). Single territories to be observed were not selected randomly, but were those territories where we saw frequent signs of defense during periodic checks. We also observed lek 2 for 272 h in 1988 and 1989. We did not include behavioral data from lek 2 because thick vegetation did not allow us to observe the entire MATERIALS AND METHODS lek. Thus, data for lek 2 were in some ways The San Rossore estate (43° N, 10° E; 46 km2) not directly comparable with data from lek 1. in central Italy is bounded by the Tyrrhenian We used sightings of recognizable males at lek Sea to the west and the Arno and Serchio riv2, for some of the analyses in this paper. ers to the south and north, and is fenced to During observations, we identified all terthe east. It is a flat lowland with submediterritorial males by plastic ear tags (n = 33) or ranean climate and sandy soil. The vegetation by natural variation in coat color and ander includes 1661 ha of pine woods (Pinus pinea morphology (n = 46). We photographed terand P. pinaster), 1153 ha of deciduous woods ritorial males at lek 1 with an 800-mm lens to (Fraxinus spp., Populus spp., Quercus spp., and allow identification of unmarked males from Alnus glutinosa), marshes, and meadows. About year to year. Photographs were not available 7 km2 of agricultural land is fenced to exclude for single-territory males. Therefore, bethe deer. The fallow deer population probably tween-year comparisons of individual males dates from the sixteenth century or earlier include all surviving territorial males from lek (Simoni, 1910). During our study (19841, but only tagged males from elsewhere. In 1989), spring population estimates declined addition, 19 tagged males diat appeared to be from 1124 to 850 because of a winter culling territorial but were not observed for 3 h were program. identified during visits to single territories and The following definitions are used in this to the other two leks in the area. We recorded paper. Territorial male: male defending a ter- die time each male spent defending each territory for more than 3 h. Most territorial males ritory, all copulations (location and male idenwere 5 years of age or older. Rut: time between tification), and the number of females in each

26

Behavioral Ecology Vol. 3 No. 1

Table 1 Behavior of individual fallow deer males defending lek satellite and single territories during the rut Territory type Lek All

No. of territorial males Defense time (h) No. of copulations No. of fights No. of chases of nonterritorial males Mating rate Fighting rate Chasing rate Harem size

53

49.5 10.8 21.7 155.4 0.15 0.46 2.99 1.76

LI

L2

L3

Satellite

21 13 0 2.1 0 0

19 47 3.9

13 112

6 30 5.0 2.0

10 13 2.1 0.4

141.8 0.14 0.14 3.41 1.79

37.0 0.10 0.06 1.99 2.36

0.24 0 0.2

19.8 74

38.4 53.9 331

0.17 0.57 2.88

0.37 0.63 4.36

1.5

4.6

Single

Data from 1985 to 1989 are cumulated. Lek males are subdivided according to the proportion of copulations achieved each year (L3: >10%; L2: < .001; satellite: rs = 0.78, n = 6, p = .03; single territories: rs = .40, n = 10,/>> .1). Fighting The number of fights and the fighting rate of territorial males (Table 1) were different in the three types of territories, even when we excluded highly successful lek males (L3) (Kruskal-Wallis ANOVA, p < .003). Analyses consistently indicated that fighting was more common on the lek than on single territories (Mann-Whitney U test, p = .0001). There was

28

Harem size

1.2

1.0

Age differences and shifts between strategies

0.6

a

Figure 2 Average number of territorial male fallow deer seen defending lek satellite or single territories during morning, midday, and afternoon.

Chases of nonterritorial males

There was no difference in the number or rate of chases of nonterritorial males according to territory type (Table 1). Within each type, the number of chases was correlated with the number of copulations (lek: rs = .86, n = 14, p < .001; satellite: rs = .87, n = 5, p = .03; single territories: r, = .71, n — 8, p = .02). There was no difference in mean harem size among the three types of territories (Table 1). Within the lek, however, there were major differences in harem size between successful and unsuccessful males (Apollonio et al., 1989a, 1990). When we removed highly successful lek males (L3) from the analyses, harem size on leks was smaller than on single territories (Kruskal-Wallis ANOVA, p = .002; Mann-Whitney U tests: lek versus satellite, p = .08 (ns); lek versus single territories, p = .001). On the other hand, when we compared L3 males with satellite and single-territory males, L3 males had larger harems (Table 1; Kruskal-Wallis ANOVA, p = .0002). Harem size was correlated with the number of chases of nonterritorial males in the lek (rs = .86, n = 14, p < .001) and the satellite (rs = 1, n = 5, p < .001). A similar trend was found in single territories (r, = .60, n = 8, p = .058).

0.8 OT .2) or in single territories (rs= -.32, n = 10, p > .1).

0.4

no

u. t.

0

6-10

10-14 Time (hours)


14-18

Exact or minimum age was known for 45 marked males (32 on the lek, x = 5.7 years, 12 on single territories, x = 5.5 years, and 1 6 year old on the satellite). There were no differences in age among males holding lek or single territories (Mann-Whitney U test, p = .38). In 1987, none of seven males seen at single territories was sighted on leks or satellites. In 1988, three of five males seen in single ter-

ritories were later seen on leks (but none copulated), and in 1989, one of seven single-territory males was later seen at a lek: this male copulated once on a single territory, then moved to lek 2 and defended a territory there for 10 days, copulating once during observations. None of 32 males recognized for 2 or more years on leks was known to switch to defending single territories between years. One male defended different single territories in 2 consecutive years. Five of 6 satellite males also defended a territory in lek 1 within the same rut. This includes four of five males that copulated on the satellite; two of these also copulated on the lek. Copulatory success

When we considered all the territorial males, there was no difference in the estimated number of copulations achieved in the three territory types (Kruskal-Wallis ANOVA, p = . 17; Table 1). There were also no differences in mating rate (p = .35). Limiting the analyses to males that were seen copulating did not result in significant differences among territory types. The samples of copulating males on single territories (n = 3, x = 7 copulations) and on the satellite (n = 5, x = 6 copulations), however, were small. There was no significant difference in the proportion of territorial males that were seen copulating in leks and single territories (Fisher's Exact test/) = .12). The variance in estimated copulatory success differed greatly among territory types, being greatest for leks (404.37) and smallest on single territories (19.43). This difference was highly significant (two-tailed/? < .01) according to an F^ test (F 3,9 = 20.81), although it must be noted that the skewed distribution of copulation violated the assumption of normality required for parametric tests. Age and experience seemed to be related to copulatory success on the lek: of 25 newcomers, only 2 were able to acquire the status of L3 in their first season on the lek. On the contrary, 8 out of 10 males that successfully occupied a territory for more than 1 year finished their reproductive activity, and presumably their life, as L3. DISCUSSION Compared to the defense of a single territory, lek territoriality in fallow deer appears to be a high benefit-high cost strategy, which only a few males can afford. The costs, mainly the high rate of fighting, are incurred by all males that join the lek, irrespective of their mating success, but only a few lekking males obtain copulations. The satellite strategy apparently combines low costs and high benefits, and five

of six males that held satellite territories copulated at least once. It appears, however, that the satellite strategy is dependent on the presence of a lek, and most males switch between satellite territoriality and lek territoriality. It is therefore difficult to evaluate the suitability of the satellite strategy independently of the lek: satellite territoriality seems more a complement than an alternative to lekking. Nevertheless, it is surprising that satellite territoriality was not adopted by more males. Possibly, holders of the satellite were able to prevent the establishment of neighbors. Another nonexclusive explanation is that other males may simply not have been aware of the reproductive value of this site. The satellite attracted more nonterritorial males than single territories, probably because of its proximity to the lek. Nonterritorial males likely concentrate in the area near the lek because of the high number of estrous females there, as suggested by the correlation between harem size and frequency of chases of nonterritorial males in all territory types. Defense time varied greatly between leks and single territories. Highly successful lek males could defend a lek territory for up to 11 consecutive days, leaving it for only a few minutes each day (Apollonio et al., 1989a). Single territories, on the other hand, were seldom defended for an entire day, and males held them for only a few consecutive days. We found a positive relationship between copulatory success and defense time in the satellite and in the lek, but not in single territories. Given the apparently much lower cost of defending single territories, one would expect greater defense time there than on the leks, but we found the reverse. We suggest that males left single territories after a short time not because they were unable to defend them any longer but because they were unlikely to achieve copulations by staying there. We have no data suggesting that males simply left one territory to defend another one. Defense of a single territory appears to combine low cost (few fights) with low benefits, and it may be considered a risk-averse strategy, possibly chosen by inferior competitors with no chance of success on the lek. For Uganda kob (Leuthold, 1966), the defense time of males in single territories is longer than that of lek males. For kob and topi (Gosling and Petrie, 1990), the alternative to lek territory is a large, resource-based single territory. Males can defend these for long times because they can feed in them. Fallow deer single territories in our study area are small and contain little or no food. Therefore, their defense entails a cost in term of lost foraging time. Fallow deer differ from other lekking ungulates in that, with few exceptions (Clut-


29

ton-Brock et al., 1988), the alternative to a lek territory is not a large, resource-based single territory, but a small display site along female movement routes. Leks in some antelope may originate in areas of high overlap of resource territories (Gosling, 1986). In fallow deer, leks appear to be the result of clumping of small display territories in areas of high female traffic. Males located their territories along female movement routes and were more likely to defend them when females were more likely to traverse them. For example, the satellite was along the deer trail leading from a feeding area through the satellite to lek 1, then on to a daytime resting area. In the morning most females first crossed the satellite and then the lek; in the evening their direction was reversed. This may explain why the satellite was defended mostly in the morning: at that time a female that had entered estrus could be met on the satellite before she reached the lek, whereas in the evening estrous females could stop in the lek before reaching the satellite. Our data revealed major difference in the frequency of male-male combat among different territory types. All males on the lek, irrespective of their copulatory success, were engaged in more fights (both in terms of frequency and of absolute number) than males in satellite or single territories. Fights are dangerous: we recorded three fatalities and several injuries during our study. Fighting success does not correlate with copulatory success (Apollonio et al., 1989a), and males do not scale the probability of fighting to the potential fitness rewards of different lek territories (Festa-Bianchet et al., 1990). A high frequency of fighting could simply be the inevitable consequences of up to 12 territorial males concentrated in less than 1 ha. On the lek, fights occur at about the same rate in territories where no copulations are seen and in territories where numerous copulations are seen, thus joining the lek always involves a considerable cost for a male. The same applies to kob and topi leks (Floody and Arnold, 1975; Gosling and Petrie, 1990), and in the latter species unsuccessful lek males may even fight more often than successful lek males. Gosling et al. (1987) found that in topi, fighting-related horn damage was more common among lek males than among males holding single territories. Thirgood (1991) saw no fights in fallow deer single territories in the New Forest, England, during 78 h of observations but reported frequent fights on a lek. Surprisingly, the average copulatory success of territorial males did not vary with territory type. Although it is clear that in this system there is no "average" male, our data may be biased in favor of single territories

30


because we observed those where we saw the greatest activity. We likely observed the most successful single territories in the study area, and we watched these territories mostly during dawn and dusk activity peaks. The copulatory success of most males on single territories may have been lower than our estimate. There were major differences in copulatory success within the lek. We have shown (Apollonio et al., 1989a, 1990) that a few territories in the lek are preferred by females, and no copulations are seen in most other territories. There may be little gain for a male joining the lek unless he can obtain one of the successful territories. Because of the apparent preference of females for only certain territories, most male fallow deer appear to have little opportunity to adopt alternative mating strategies and can expect little or no reproductive success regardless of whether they can defend an unsuccessful lek territory or a single territory. We suggest that joining the lek cannot be simply considered one strategy whereby an "average" male could expect some "mean fitness returns." The extremely high variance in copulatory success of lek males suggests that the fitness returns that can be expected by a male entering a lek are unpredictable. The consequences of joining the lek vary greatly according to which territory the male is able to obtain. We do not suggest, however, that territory location is the only determinant of reproductive success of lekking fallow deer: phenotypic differences among males also appear to play a role (Apollonio et al., 1989a; CluttonBrocket al., 1989). Males that joined the lek and occupied unsuccessful territories experienced frequent fighting and loss of foraging time. Within a single breeding season these males may have been better off by holding single territories or giving up reproduction altogether. They may have joined the lek to obtain high reproductive success in future years. Of 10 highly successful males observed during our study (L3 males), only 2 were first seen in the year when they were highly successful. We saw all the others in the same lek for 1 or 2 years before the year in which they became highly successful. As individual males aged, their copulatory success improved (Apollonio et al., 1989a). Interactions among unsuccessful males on the lek may affect their reproductive success in future years by establishing individual relationships. This suggestion is supported by the increase in fighting frequency that followed an apparent disruption of an established hierarchy caused by the removal of most successful lek males following the 1986 rut (Apollonio et al., 1989b). Choice of territorial strategy may depend

on body size and body condition. There is great variation in physical characteristics of males in the San Rossore population: weights of 18-month-old males range from 35 to 52 kg (Apollonio M, unpublished data). Size-related specializations in different mating strategies were reported for male kob by Leuthold (1966) and for male topi by Montfort-Braham (1975) and Gosling and Petrie (1990). Finally, we should point out that our study population is free from predation on adult deer, the largest predator being foxes (Vulpes vulpes). It is still not clear if it is more dangerous for a male, in the presence of predators, to defend a territory on a lek or an isolated territory (Balmford, 1990; Gosling and Petrie, 1990), but it is reasonable to expect that the strenuous activity in the lek may make successful males more vulnerable to predation after the rut. We are grateful to E. Bruno, L. Burrini, M. Locati, G. Rasola, and M. Riva for field assistance and to the gamekeepers of the San Rossore Presidential Estate for logistic help. Our research was partially supported by the Istituto Nazionale di Biologia della Selvaggina. M.F.B. was supported by a Canada-UK NATO Science Fellowship.

REFERENCES Apollonio M, 1990. Lekking in fallow deer: just a matter of density? Ethol Ecol Evol 2:291-294. Apollonio M, Festa-Bianchet M, Mari F, 1989a. Correlates of copulatory success in a fallow deer lek. Behav Ecol Sociobiol 25:89-97. Apollonio M, Festa-Bianchet M, Mari F, 1989b. Effects of removal of successful males in a fallow deer lek. Ethology 83:320-325. Apollonio M, Festa-Bianchet M, Mari F, Riva M, 1990. Site-specific asymmetries in male copulatory success in a fallow deer lek. Anim Behav 39:205-212. Balmford AP, 1990. Lekking in Uganda kob (Ph.D. dissertation). Cambridge: University of Cambridge. Buechner HK, 1961. Territorial behaviour in Uganda kob. Science 133:698-699. Byers JA, Kitchen CW, 1988. Mating system shift in a pronghorn population. Behav Ecol Sociobiol 22:355— 360. Clutton-Brock TH, Green D, Hiraiwa-Hasegawa M, Albon SD, 1988. Passing the buck: resource defence, lek breeding and mate choice in fallow deer. Behav Ecol Sociobiol 23:281-296. Clutton-Brock TH, Hiraiwa-Hasegawa M, Robertson A, 1989. Mate choice on fallow deer leks. Nature 340: 436-465. Dunbar RIM, 1982. Intraspecific variation in mating strategy. In: Perspectives in ethology (Bateson PPG, Klopfer PH, eds). New York: Plenum; 385-431. Espmark Y, Brunner W, 1974. Observations on rutting behaviour in fallow deer, Dama dama (Linn, 1758). Saugetierkundl Mitt 22:135-142. Festa-Bianchet M, Apollonio M, Mari F, Rasola G, 1990. Aggression among lekking male fallow deer: territory effects and relationship with copulatory success. Ethology 85:236-246. Floody DR, Arnold AP, 1975. Uganda kob (Aderwta kob Ihomasi) territoriality and the spatial distribution of sexual and agonistic behaviours at a territorial ground. Z Tierpsychol 37:192-212.

Gosling LM, 1986. The evolution of mating strategies in male antelopes. In: Ecological aspects of social evolution (Rubenstein DI, Wrangham RW, eds). Princeton: Princeton University Press; 244-281. Gosling LM, Petrie M, 1990. Lekking in topi: a consequence of satellite behaviour by small males at hots pots. Anim Behav 40:272-287. Gosling LM, Petrie M, Rainy ME, 1987. Lekking in topi: a high-cost, high benefit strategy. Anim Behav 35:616— 618. Hogg JT, 1988. Copulatory tactics in relation to sperm competition in Rocky Mountain bighorn sheep. Behav Ecol Sociobiol 22:49-59. Langbein J, Thirgood SJ, 1989. Variation in mating system of fallow deer (Dama dama) in relation to ecology. Ethology 83:195-214. Leuthold W, 1966. Variation in territorial behaviour of Uganda Kob, Aderwta kob thomasi (Newmann 1896). Behaviour 27:214-257. Montfort-Braham N, 1975. Variations dans la structure sociale du topi, Damaliscus komgum Ogilby, au Pare National de l'Akagera, Rwanda. Z Tierpsychol 39:332364. Pemberton JM, Albon SD, Guinness FE, Clutton-Brock TH, and Dover GA, 1992. Behavioral estimates of male mating success tested by DNA fingerprinting in a polygynous mammal. Behav Ecol 3:66—75. Rubenstein DI, 1980. On the evolution of alternative mating strategies. In: Limits to action: the allocation of individual behaviour (Stadden JER, ed). New York: Academic Press; 1-44. Simoni D, 1910. San rossore nella storia. Florence: Olski Press. Thirgood SJ, 1991. Alternative mating strategies and reproductive success in fallow deer. Behaviour 116:1—10. Wirtz P, 1981. Territorial defence and territory take-over by satellite males in the waterbuck Kobus ellipsiprymnus (Bovidae). Eehav Ecol Sociobiol 8:161-162.


31