Behavioral variation in mantled howler monkeys is structured by ecology and demography, not by individual signature.

Abstract

How behavioral variation is partitioned among intrinsic attributes, extrinsic conditions, and idiosyncratic individual signature has consequences for the targets of selection and the mechanisms by which behavior evolves. Once the structured sources of behavioral variation have been accounted for, any variation that remains and is consistently attached to particular individuals constitutes an individual signature. Contemporary frameworks recognize three axes along which such a signature can manifest: differences among individuals in mean expression (the classical personality axis, e.g., some individuals consistently more active than others); in plasticity (the reaction-norm slope, i.e., individuals differing in how strongly they adjust behavior to environmental change); and in predictability (residual intra-individual variability, i.e., individuals differing in how variable their behavior is from one occasion to the next). Whether an individual signature exists on any axis in energy-minimizing species, whose adaptive strategy may favor flexibility over consistency, has not been formally tested. We used 6486 focal observations on 17 mantled howler monkeys (Alouatta palliata) collected over 10 years to test for individual signature on all three axes after accounting for sex, reproductive state, age, group, year, food availability, temperature, and time of day. No axis showed a detectable signature: repeatability estimates were near zero (R = 0.008-0.034) with confidence intervals encompassing zero; random-regression models revealed no individual × environment interactions in responses to food availability; and residual intra-individual variability showed no among-individual signal once boundary-induced mean-variance coupling in proportional data was removed by logit transformation. Variance attributable to month and year exceeded that of individual identity by factors of four and 1.4, respectively. These convergent null results suggest that behavioral variation in this population is decomposable into structured ecological and demographic sources. Behavioral flexibility, not individual consistency, is the population-level strategy.