Evaluación de resultados del proyecto con los criterios del Sistema Nacional

Living in groups provides a number of situations where members can gain advantages by coordinating activities and maintaining cohesion. This may require the exchange of information through some form of communication. In long term social groups, animals may develop behavioural strategies that are based on individualized relationships among members such as dominance hierarchies and roles within groups (Wilson, 2000). In these cases there might be a selective pressure for the development of identity signals and mechanisms that allow discrimination and recognition of particular individual and/or group members, from non-target individuals or groups. In fact, some studies suggest that individual recognition is an important factor in the structure of group hierarchies (Dugatkin & Earley, 2004). Long term associations among animals also provide the prior experience of particular individuals required for the development of individual-level signal recognition, based on learned familiarization (Bradbury & Vehrencamp 1998; Tibbetts & Dale, 2007).

The ability to discriminate among group members can also be advantageous when behavioural responses to signals have different consequences depending on the signaller. For example receivers can optimize the costs of responding to alarm calls given, by varying responses depending on their reliability and spatial relationship of the signaller(s) (Robinson, 1981; Ydenberg & Dill, 1986, Cheney & Seyfarth, 1988, Bachman, 1993, Kildaw, 1995). Animals can also keep track of hierarchical relationships within groups by eavesdropping on signals from interactions of

other individuals (Bergman et al., 2003).

Individual discrimination is possible when individual signal parameters have unique attributes, or when the signal parameters' variability is greater among than within individuals (Beecher, 1982; 1989). For example white-winged vampire bats (Diaemus youngi) show individual variation in the structure of social calls, that can be discriminated by the animals (Carter et al., 2008).

3.1.2. Sperm whale vocalizations

The sperm whale (Physeter macrocephalus) is a social cetacean species. This is particularly the case for females, calves and immature animals of both sexes who live in long term social units of 12 animals on average (Christal et al., 1998). These generally matrilinear units are distributed throughout subtropical and tropical waters (Rice, 1989) and their composition is largely stable over decades (Whitehead & Weilgart, 2000), albeit with occasional movements among units (Christal et al., 1998). In the Pacific Ocean these units frequently form groups with one or two other units that persist for days (Whitehead & Weilgart, 2000). Sperm whale groups often move in a coordinated fashion, spreading themselves beyond visibility range over hundreds or thousands of meters of ocean (Whitehead, 2003).

Some authors have suggested that post-menopausal females might play a special role within social units by assisting with the care of calves and acting as repositories of information that are advantageous (Gero, 2005; McAuliffe & Whitehead, 2005). Individuals within social units have preferred associates among members (Gero et al., 2008). These observations suggest that individuals might interact differently among unit members. Differential interaction among unit members primes the need for an individual discrimination system.

Sperm whales rely mostly on the emission of pulsed sounds for communication, orientation and finding prey (Jaquet et al., 2001; Whitehead, 2003; Johnson & Tyack, 2003; Madsen et al., 2002a; 2002b; Miller et al., 2004a; 2004b). These clicks are characterized by having a series of usually evenly spaced pulses of decaying amplitude (Backus & Schevill, 1966) whose inter-pulse interval (IPI) has been shown to be correlated with the whales' size (Gordon, 1991; Rhinelander & Dawson, 2004). Series of clicks are produced during foraging dives at rates of 1-2 clicks per second.

are assumed to have a communicative function (Watkins & Schevill, 1977; Whitehead & Weilgart, 1991; Schulz, 2007; Schulz et al., 2008). Sperm whale groups in the South Pacific Ocean have distinct coda dialects which are stable for periods of at least six years (Weilgart & Whitehead, 1997; Rendell & Whitehead, 2005a). Whitehead et al. (1998) found that in the Pacific coda dialect variation was correlated with mtDNA variation. This was interpreted as an indication of parallel vertical transmission of both mitochondrial haplotypes and vocal repertoires within the mostly matrilinear units. Rendell & Whitehead (2003b) also found that sperm whale units and groups in the Pacific Ocean could be aggregated into vocal clans based on their coda repertoires. Because vocal clans were often sympatric and shared most nuclear DNA haplotypes they suggested that vocal clans are the result of culturally transmitted behaviour.

The function initially proposed for codas was one of individual signatures (Watkins & Schevill, 1977; Watkins et al., 1985). Later studies which classified codas into distinct types showed evidence of coda type sharing among individuals, challenging the initial hypothesis (Whitehead & Weilgart, 1991; Rendell & Whitehead, 2004). Schulz (2007) found that most adult animals within a social unit shared the most common coda type, with the exception of the mother- calf pair whose repertoires were different from those of other unit members. Apart from the mother- calf differences, the repertoire similarities of other members did not support the idea of individually distinctiveness coda type repertoires, as most individuals share most common coda types, and produced them at similar rates. This sharing of coda repertoires suggests the function of coda repertoires to be group membership recognition, either at the unit or clan levels. This idea is further supported by the fact that social units seem preferentially to form groups with other units of their own clan (Whitehead, 2003).

Codas are not perfectly stereotyped however. Within particular coda types which might be identified statistically there is often considerable variation and it is possible that some of this variability is specific to individuals. Thus, individuals within groups might be recognisable by the way they make particular coda types rather then by the range of coda types they produce. With this in mind I hypothesise that despite similarities in coda types, variations within these could potentially carry information on animals' identity and therefore codas could have both group and individual level information. Here I test the hypothesis that variation within coda types allows statistical discrimination of individual social unit members, potentially allowing for individual identity to be communicated between members in a social unit.

3.2.