Socionomy of SLI elephant seals
Demography and social system
The social system of elephant seals populations is shaped by two factor: land breeding during a short breeding season and strong tendency of females to aggregate when on land. These two factors are common in land breeding pinnipeds and result in an high level of polygyny. Two additional factors should be considered when dealing with elephant seals: the low mobility of females on land and enormous sexual dimorphism. All these factors result in a mating system based on harem formation: females aggregate in large groups, each one defended by a single male (the alpha male or harem master) with exclusive (or almost exclusive) access to breeding females of the harem. This mating system permits the achievement of the highest level of polygyny recorded in mammals.
A large (more than one hundred females) elephant seals harem (the harem holder is in the middle of the harem; one beta is between females on the right side)
The stability of harems in both space and time is the result of the strong tendency of females to aggregate and of the reduction of movements after parturition. The grouping of females is an autocatalytic process: the increase in size of an harem makes it much more visible and more attractive for females, and that favors the increase in size. The percentage of females in harems is higher than 90% for the most of the breeding season and isolated pupping is very unusual (a single case in 1996, with death of the pup; four cases, with three alive weanlings in 1997). Hence, breeding in an harem is an almost obliged component of female breeding strategies in this species. The structure of the mating systems depends on local demography, which defines the basal level of polygyny, and on the intensity and result of intermale competition, which defines the realized level of polygyny. We call the sum of demographic and social factors socionomy of the population.
Stability of the social system
There was a low change of the breeding males system across breeding seasons. About 65-70 % of the breeding males of the 1996 and 1997 breeding season have been marked as resident males during the previous season (34.0 % were breeding males during the previous season). Sea Lion Island is a small island with a small local population of elephant seals and preliminary information show that there is very few breeding of elephant seals in the rest of the Falklands: so breeding males should have few opportunities to roam around, and this should in turn determines a greater stability of the social system.
Spatial distribution of harems
On Sea Lion Island the settlement of females is uneven and the distribution of harem is irregular: all the breeding activities happen on the sandy beaches at the eastern tip of the island. The restriction of breeding to sandy beaches could depend upon the easy of access, because their modest slope facilitate the haul-out of pregnant females. The distribution of harems shows a little variation between breeding seasons on large scale, apart from two aspects: creation of small new harems in previously unused areas and hundred meters scale variation of the initial site of the harem. The fidelity of individual females to pupping sites is very high on large scale but may be quite low on a finer one: in elephant seals there is a strong relationship between prepartum movements and male interference and so the settlement of females depends not only on preferences at arrival but also on interaction with males in arrival phase. Movements of both males and females (before parturition) are limited between breeding zones but quite common between harems. Fidelity is high at zone level (about 70% of females tagged came back to the same zone for breeding) but is lower at harem level (about 39% of females breeding in the same harem, about 55% of females breeding in another harem already present in the previous season and about 6% of females breeding in new harems). Fidelity to breeding site of individual females across seasons is high on kilometer scale but quite low at hundred meter scale: any case it's quite remarkable that more than one third of females came back to exactly the same harem for breeding.
Timing of harems formation
Different harems have different timing of formation and disappearance. The median day of formation was day 20 to 23 (third week of September) in different seasons; the median lifetime of harems was 53 to 57 days. There is a general relationship between the date of formation of an harem and its size, with an early beginning for larger harems, but this relationship is not an absolute one, and also late forming harem may achieve a mid range size. The variation in the timing of different harems has a strong effect on male mating strategies: males excluded from the control of harems at the beginning of the season have additional opportunity to get control of late harems. These means that the most powerful males will get control of the early harems, which are usually but not always the larger ones, but later in the season new opportunities will be available for less powerful males. Sometimes late harems will grow up to a large size giving a good breeding opportunity to secondary males. This process reduces the strength of the link between male phenotypic qualities and access to females, reducing the opportunity for sexual selection to operate.
Harem size
The single most important component of socionomy is harem size, which measures the tendency of females to breed in an aggregated social system. We define harem size as the maximum number of females counted in an harem during the daily census; it represents about 90% of the number of females that actually breed in the harem. Median harem size varied between 31 and 35 females in different seasons. The mean harem size recorded at Sea Lion is smaller than the ones recorded at Punta Delgada (Valdes Peninsula), and also much smaller than the mean harem size reported for South Georgia and the majority of other population of southern elephant seals, but is larger than the median harem size of 11 females calculated for the whole Valdes Peninsula population. The intraseasonal variation of the harem size is large (5-125 females): this is the first indicator of the huge variation in breeding performance between males who are able to maintain control of a harem. This variation in size of harems is one of the main source of variance in mating success, and hence of potential for sexual selection.
Social and behavioural effects of harems structure
The first correlate of harem structure should be the level of competition: the mating system of elephant seals is an overcompetitive one, with the most of the breeding males excluded from direct control of females. These peripheral males usually assemble around the larger harems, and so we expect a positive relationship between competition pressure and harem size. In fact an index of potential level of competition calculated from the daily number of secondary males associated to each harems has a strong correlation with harem size. This relationship is almost linear (82% of variance explained by the linear component): the linearity of this relationship is important because it implies a gradual increase in costs of defense with the increase of benefits due to access to a larger number of breeding females. The competitive interaction rate (interactions per observation period) has a wide variation among harems, with median values ranging from 1 to 18. There is a strong positive relationship between median interaction rate and harem size, but the relationship is not linear: for harems up to about 60 females of size the interaction rate is low and grows slowly, while for larger harem the increase in interaction rate with size of the harem is sharp. This two speed process implies an increase of cost of defense more than proportional for very large harems: only large, adult, experienced males are able to manage this level of competition pressure.