Behavioral Flexibility in Primates

Signaling Theory and Patterns of Branch-Breaking in Mantled Howler Monkeys (p. 87)

The pattern of BBE repetition and frequency presented above can be interpreted in terms of theoretical models, bearing in mind that the overall "signal" magnitude may depend on combining and recombining a number of display elements (i.e., flexible behavior). The problem of understanding how the receiver combines information from a sequence of BBEs in order to make an assessment of the signaler is essentially a separate issue from understanding how the signaler decides what magnitude of signal to give. Both of these topics deserve investigation in primates.

Evolutionary models show that the form of a display (in terms of number of repetitions, or changes in intensity over time within a display sequence) is predicted to depend on the way the display is assessed and interpreted by the receiver ("assessment rules"; Payne and Pagel, 1997; Parker, 1974; West- Eberhard, 2003). The present results do not contain information on "intensity" in each BBE, but they do reveal how mean numbers of repetitions differ among individuals. The modal number of BBEs per sequence is 1 BBE/1 min, and BBEs per interval never exceeded 5 (Table 6.1, column 3). This low modal number of BBEs per sequence is not consistent with the predictions of the sequential assessment model (Brockmann, 2001) in which the "signal" is the average of the display elements so far, but it is consistent with an extreme case of a signal of endurance (Payne and Pagel, 1996a), in which the signal is the cumulative sum of the elements, accounting for any variation in intensity.

The findings are also consistent with the "best-so-far" model (Payne and Pagel, 1996b), in which the signal is the intensity of the most recent element only. As Payne and Pagel's "best-so-far" model is summarized by Ord and Evans (2003, p. 1504), "When signals incur significant costs (e.g., from fatigue), assessment is more likely to be based on a cumulative measure of all displays performed." These authors continue, the model "is a cumulative function rule, together with a threshold that varies according to individual condition." The "best-sofar" model is unusual in that it is the only model that necessarily predicts that the majority of signal sequences will only contain one element, as is observed in the present data set.

Further observations of factors, such as any variation in element intensity, will be needed to resolve between the latter two models for the displays of mantled howler monkeys, and it will be necessary to systematically study the branch-break display in additional groups before confident statements can be made about its function(s).

Nonetheless, it seems reasonable to assume, tentatively, that the signal magnitude is at least in some way proportional to the number of BBE repetitions and that these quantities, taken together, can be employed as an effective measure of behavioral flexibility. It is important for future research to investigate whether survival and/or reproductive