A subpopulation can be said to show local synchrony in breeding if births of their young are more closely simultaneous than expected compared with other subpopulations. Several hypotheses have been advanced to explain the general phenomenon of synchronized births in vertebrates and, in particular, local synchrony within subpopulations. Those hypotheses include a likely advantage of reduction of predation in birds (Jones, 1986), and one of predator confusion and saturation (Estes, 1976). Under some circumstances, mammals may be able to increase their energetic commitment to reproduction by timing the breeding season to coincide with high resource abundance (Oftedal and Gittleman, 1989)....
Canids are seasonal breeders (Ewer, 1973), and clumped births tend to be explained in terms of environmental conditions. The Ethiopian wolf (Canis simensis) is a medium-sized canid living in family packs in the Ethiopian highlands and preying exclusively on Afroalpine rodents (Sillero-Zubiri and Gottelli, 1995a). The species appears to be constrained by the distribution of prey; wolf density is closely correlated with prey abundance. Wolves appear to use all suitable habitat.
During a 4-year study in the Bale Mountains of southern Ethiopia dispersal opportunities were limited severely by habitat saturation. All males and about one-third of female offspring stayed in their natal territories, and some females eventually acquired the mother's breeding position. Dominant females monopolized breeding, and on three (possibly five) of eight occasions, they were succeeded in this role by their most dominant daughters, providing the potential for incest, and therefore, inbreeding.
Extra-pack copulations might be a mechanism to avoid inbreeding. During a short mating season each dominant female is guarded closely by resident males, but despite this, both sexes visit neighboring wolf territories to secure extra-pack copulations. Twenty-one of 30 observed copulations (70%) were between the dominant female of one pack and a male from an adjoining pack. We concluded that, under conditions where dispersal is constrained and the potential for inbreeding is high, extra-pack matings (and associated multiple paternity) result in outbreeding.
However, advantages of this system may differ markedly for the sexes. Although a fitness advantage accrues to a male with extra-pack copulations, if his mate, the dominant female, does the same, his fitness decreases. Equally, while it may be advantageous for a dominant female to seek matings with males outside her pack, she may be disadvantaged if her mate seeks such liaison to the extent that her young become a small fraction of his paternal investment. We may thus expect a conflict between sexes, particularly between the dominant male and female of each pack.
Extra-pair copulations are most common in species where females breed synchronously (Emlen, 1991). It has been argued that for both sexes, a temporal concentration of female fecundity (breeding synchrony) increases net benefits of extra-pair copulations (Stutchbury and Morton, 1995), such that breeding synchrony is an important factor promoting evolution of extrapair mating systems (Wrege and Emlen, 1991).
We assessed the evidence for such conflict between male and female Ethiopian wolves in the seeking of extra-pair copulations, which may be expressed through behavioral or physiological mechanisms. We present evidence that female wolves synchronize reproduction to monopolize male attention and thereby limit opportunities for extra-pack copulations of their mates.
MATERIALS AND METHODS
Study area.--Ethiopian wolves were studied in the central massif of Bale Mountains National Park of southern Ethiopia (7 deg N, 39 deg 45'E), the largest realm of Afroalpine habitat in the continent. The study was conducted at three different areas: the Web Valley (elevation 3,450 m), Sanetti Plateau (4,050 m), and Tullu Deemtu (3,800-4,300 m). The first two areas represented open-grassland typical of Afroalpine habitat and supported high wolf densities (ca. 1.2 wolf/km2). The latter represented another common habitat type, Helichrysum dwarf-scrub, and sustained a much lower wolf density (ca. 0.25 wolf/km6sup 2^). The three study areas were separated by < 20 km, with Tullu Deemtu only 10 km from Sanetti. There was little exchange among the three subpopulations during the study; however, genetic screening showed the three subpopulations were part of a single metapopulation....
Study population.-The population consisted of 18 packs of Ethiopian wolves ranging from 2 to 13 adults....Sixty-seven animals were marked using plastic ear tag. Nineteen animals were fitted with radio-collars under anesthesia. Radio-telemetry was used only to locate or relocate individuals during observations. In addition to the radio-collared animals, 48 individuals were identifiable by their ear tags and 64 others by coat pattern differences....Animals were habituated and easy to observe through binoculars on foot or from horseback at 20-500 m in open terrain. We collected 946 hours of observations from 427 separate focal sessions. Individual identity, group composition, behavior, and location were recorded for >5,200 independent wolf sightings. We identified dominant females using a matrix of greeting-aggression behaviors of all members of a given pack.
Wolves remained in year-round territories that were maintained by regular communal patrolling and scent marking of territory boundaries. Average group size with wolves >/-1 year old was 5.9 +/- 0.5 SE. Configuration of territorial boundaries was essentially stable during this study (Sillero-Zubiri and Gottelli, 1995b). Thus, sedentary packs and their stable range boundaries made it possible to treat groups as distinct entities across years. Dispersal into and out of the study packs was low every year, an average of 0.8 animals per pack dispersed or disappeared, while only 0.1 female on average joined a new pack.
Breeding status of all packs was monitored regularly during a 4-year field study in 1988-1992. Because all wolves were known individually, we could ascertain the maternity of all young born during the study. Dominant breeding females were identifiable from visible signs of pregnancy and birth of viable young was obvious from the behavior of the mother at the den. Birth dates were extrapolated from evidence of pregnancy, denning or emergence of young (at 3 weeks of age)....
RESULTS AND DISCUSSION
Populations of Ethiopian wolves in the Bale Mountains had a single breeding season each year. The dominant female of each pack gave birth once a year coinciding with the end of the rainy season. Only 56% of resident-dominant females (19 of 34 pack-years) bred successfully each year (i.e., young emerged from den). Young were born at the beginning of the dry season and become independent at its end, coinciding with the period of highest prey availability.
All recorded litters were born from late August to late December. Within this broad season of 'birth', packs from a given subpopulation bred synchronously, with all births within a subpopulation in a year concentrated in 2.2 weeks....In 1991, all six litters in the Web Valley were born within 1 week....
Although breeding was tightly synchronized within each subpopulation of 3-8 packs, there were clear differences among them in the onset of breeding. A significant difference in the timing of births of litters was found among the three study areas. Packs of a given subpopulation bred synchronously, but packs of separate subpopulations were often out of phase with one another, suggesting a correlation between onset of breeding and local environmental conditions.
In three breeding seasons, births at Sanetti occurred 4, 6, and 9 weeks after the subpopulation at Web, but in 1991 they occurred simultaneously. Births recorded at Tullu Deemtu in three different years occurred 1, 5, and 13 weeks before those at Sanetti....Those two areas were <20 km apart and presumably experience similar environmental conditions, although Web was drier and the rains ended earlier than at Sanetti in those years for which rainfall data were available. It was possible that initiation of breeding did not depend on environmental cues per se, but availability of rodent prey before breeding.
The pattern of local synchrony within subpopulations but asynchrony of contiguous subpopulations is unusual in vertebrates and, to our knowledge, has been reported only for rock hyraxes (Heterohyrax brucei and Procavia johnstoni) living on isolated kopjes in the Serengeti plains (Hoeck, 1982), and in white-fronted bee-eaters (Merops bullockoides) at Nakuru, Kenya (Wrege and Emlen, 1991). While no clear explanations were given for local synchrony in either study, social factors appear to be the best explanation of breeding synchrony in the Ethiopian wolf. We suggest that mating synchrony is an adaptation by breeding females to reduce the capacity of dominant males to associate monopolizing them with philandering. In the absence of rigid inhibitory environmental conditions, timing of breeding may be influenced more by social stimulation, suggesting the existence of a mechanism for synchronization of breeding locally, which possibly involves scent marking.
We argue that synchrony of estrus among dominant females in packs of a subpopulation benefits each breeding female by increasing the dominant male's paternal investment in her young. Superimposed upon that system, under conditions of inbreeding, is an advantage to the breeding female of securing extra-pack copulations. Clearly, mate guarding is not infallible, due to the dominant male being distracted by the need to sleep, eat, and engage in other social-territorial activities and the attraction of females in adjoining territories. This imperfection in mate guarding affords the female opportunities for mating with neighbors. Her liaisons at territorial borders are most likely to involve a subordinate male because, having been rejected by the dominant female of his own pack, he is free to focus on seeking such encounters. Indeed, the majority of extrapack copulations involved the dominant female of one pack in fleeting association with subordinate males of another.
Our data demonstrate that in each of 4 years there was tight local synchrony in breeding among packs in two subpopulations of Ethiopian wolves. However, while the pattern and sequence of breeding between different subpopulations varied from one year to the next, there was never any evidence of synchrony between them. The inconsistency of pattern between different areas allows us to exclude explanations that depend on a systematic environmental difference (such as elevation or rainfall). The possibility that local synchrony derives from local variation in weather, such as rainfall, is unlikely because of the proximity of the three subpopulations; furthermore, there is no support for this suggestion in the weather records. By exclusion, therefore, we conclude that it is most likely that the local breeding synchrony in Ethiopian wolves is socially mediated. Such social forces could be adaptive for dominant females in the context of the breeding system described by Sillero-Zubiri et al. (1996). In short, by synchronizing estrus between adjacent packs, females may make it difficult for males both to monopolize them and philander with their neighbors.
Figure 1: The seasonal pattern of birth and rainfall at (a) Sanetti Plateau and (b) Web Valley, Southern Ethiopia, 1984-1991.
Table 1: Seasonality of Ethiopian wolf birth dates for 37 litters in three study areas.