Braschi C., 2004.
Confronti agonistici nell'elefante marino del sud: fattori coinvolti e ruolo
delle asimmetrie tra due maschi
(Male agonistic behavior in southern elephant seals: factors and asymmetries
involved in conflict).
PhD thesis. Dipartimento di Biologia Animale e dell'Uomo. Università degli Studi
di Roma "La Sapienza", Roma, Italia.

The main focus of this thesis is the analysis of male agonistic behaviour of the
southern elephant seal, Mirounga leonina. In particular, the events which lead
up to, and the resolution of, conflict. In this species, the competition between
males is strongly connected to male reproductive success: the breeding strategy
consists of harem defence polygyny (Emlen and Oring, 1977; 3 1989; Riedman,
1990) and males compete among each other to set up a dominance hierarchy, so
that the males with the highest rank will gain the control of a harem (a group
of females) (Le Boeuf, 1974). Only a few males can actually attain complete
control of a harem and reproduce; those are the individuals with the best
agonistic capabilities (Galimberti et al., 2002). Elephant seals are large
mammals and have a high sexual dimorphism in size (Reiss, 1989). They have a
mixed life style, with two aquatic feeding phases and two land phases (breeding
and moulting). Reproduction is concentrated into a three month breeding season
and males compete among each other only during this period, while in other
phases of their life cycle they show low social interaction (Le Boeuf 1974).
Agonistic interactions between males are dyadic and asymmetric, the latter of
which depends on male physical features as well as on demographic and social
characteristics of the environment where the interaction occurs (Mc Cann, 1981;
Fabiani, 1996). All these features define the individual agonistic skills of a
male which are referred to as its Resource Holding Potential (RHP: Parker 1974).
Moreover, asymmetries between males can also depend on the interest the
individual has on resource acquisition (pay off asymmetry, Maynard Smith and
Parker, 1976). Male elephant seals mainly compete using conventional
interactions, which consist of stereotyped acoustic signals and visual display.
If agonistic conflict can not be resolved conventionally, the situation normally
leads to direct physical conflict. The presented study was conducted on Sea Lion
Island, which is a small island at the extreme south of the Falkland Islands.
The population of southern elephant seals which inhabit this island is a small
steady population, and constitutes the main breeding colony of the whole
archipelago (Galimberti and Boitani 1999). The research was part of a larger
project started in 1995 carried out by the Elephant Seal Research Group, ESRG.
Field work was carried out over the 2002 and 2003 breeding seasons; however,
data from 1995 were also included in the analyses. The main field activities
involved marking techniques for the recognition of individuals, census
techniques, measurement of animal behaviour, measurement of male structural
features (photogrammetric method), video recording and analysis. The aims of my
project were to determine the structural and functional correlates of elephant
seal agonistic behaviour, and to examine the effect of asymmetries in male
traits and the likelihood to win an interaction. Particular attention was given
to the influence of asymmetries on the probability of an individual to win a
direct fight. In fact, in spite of its low frequency, fighting is a notable
aspect of elephant seal social behaviour, both in northern (Sandegren 1976;
Haley 1994) and in southern species (Mc Cann 1981; Fabiani 1996). Therefore, a
detailed analysis of fight dynamics and resolution provides an opportunity for a
better understanding of the agonistic behaviour of this species and its impact
on male reproductive 4 success. For this purpose, I identified several
parameters as probable correlates of male RHP mainly from a biological point of
view. My results show that RHP asymmetries between two males mainly depend both
on physical characteristics (i.e., body mass, body length and age) and male
behavioural traits (i.e., aggressiveness, experience, breeding status, presence
on land and fighting record). The extent of these asymmetries does play a
particularly important role in influencing how interactions take place and their
agonistic intensity. Usually, two males solve a conflict using conventional
displays when the difference in their RHP value is clear, while they tend to
fight when their RHP values are similar. Normally, the male who wins shows a
higher RHP; however, this is not always the case, as other factors seem also to
influence the outcome of a male interaction. These results seem to follow the
theory of animal agonistic behaviour. However, both conventional and direct
interactions in this species are complex and cannot be accurately described by
general theoretical models of animal agonistic behaviour (e.g. war of attrition,
Maynard Smith, 1974; sequential assessment, Enquist and Leimar, 1987, 1990;
cumulative assessment, Payne 1998). Moreover, the resource value and the
reproductive payoffs are also important aspects when two elephant seals compete
among each other. However, both the presence and the size of a harem in the area
where the interaction takes place do not seem to have any influence on the
intensity of an agonistic conflict. The only exception is when the direct
outcome of the agonistic interaction is the immediate control of the females
(i.e., context of harem take over). Likely, in these contexts, the level of
motivation of the male assumes an important role in the resolution of a
conflict. The analysis of the effects of male asymmetries on the likelihood to
win an agonistic interaction followed a univariate approach. Once the most
important asymmetries were identified, a multivariate analysis was carried out
to investigate which of these had a central role in the resolution of a direct
fight. In fact, a multivariate approach is the only approach that can define the
relative importance of each asymmetry between two males, and to state which one
has a higher influence on fight resolution. For this purpose, a logistic
regression model was built, where male RHP correlates were the independent
variables and the victory or defeat of an aggressor male was the dependent
variable (outcome). The results of this model showed that only the body length
and the aggressiveness of the male had a significant role in fight resolution.
The results of this thesis provide new evidences of the importance of
behavioural aspects in male agonistic contexts, and a gives a better picture of
the dynamics of male agonistic interactions in elephant seals and particularly
of the main factors involved in fight resolution.