Mother Vocal Recognition in Antarctic Fur Seal Arctocephalus gazella Pups: A Two-Step Process

Abstract

In otariids, mother's recognition by pups is essential to their survival since females nurse exclusively their own young and can be very aggressive towards non-kin. Antarctic fur seal, Arctocephalus gazella, come ashore to breed and form dense colonies. During the 4-month lactation period, females alternate foraging trips at sea with suckling period ashore. On each return to the colony, females and pups first use vocalizations to find each other among several hundred conspecifics and olfaction is used as a final check. Such vocal identification has to be highly efficient. In this present study, we investigated the components of the individual vocal signature used by pups to identify their mothers by performing playback experiments on pups with synthetic signals. We thus tested the efficiency of this individual vocal signature by performing propagation tests and by testing pups at different playback distances. Pups use both amplitude and frequency modulations to identify their mother's voice, as well as the energy spectrum. Propagation tests showed that frequency modulations propagated reliably up to 64m, whereas amplitude modulations and spectral content greatly were highly degraded for distances over 8m. Playback on pups at different distances suggested that the individual identification is a two-step process: at long range, pups identified first the frequency modulation pattern of their mother's calls, and other components of the vocal signature at closer range. The individual vocal recognition system developed by Antarctic fur seals is well adapted to face the main constraint of finding kin in a crowd.

Fig 1. Experimental signals modified in the temporal domain.

From left to right: natural control, synthetic control, wfm (no frequency modulation), and wam (no amplitude modulation).

Fig 2. Experimental signals modified in the frequency domain.

From left to right: synthetic control, l300 (linear shift: entire frequency spectrum shifted up by 300 Hz), a300 (additional shift: fundamental frequency shifted up by 300 Hz).

Fig 3. Behavioural scores (PC1) obtained with the natural controls, synthetic controls and the different experimental signals.

Abbreviations: wam (no amplitude modulation); wfm (no frequency modulation); rev (time-reversed signal); l75,l175, 300 (linear shifts: entire frequency spectrum shifted up by 75, 150 and 300 Hz respectively); a75, a150, a300 (additional shift: fundamental frequency shifted up by 75, 150 and 300 Hz respectively). Boxplots indicate medians (square markers) and 25–75 percentiles, and whiskers show min-max values of PC scores. P-values resulting from the comparison of responses between the synthetic controls (Mean Synth) and the different experimental signals are indicated as * for p < 0.05, ** for p < 0.01 and *** for p < 0.001.

Fig 4. Number of pups within a group responding to female calls at 3 different distances.

When the distance decreases, the number of responding pups decreased too. For all tested distances, the filial pup of the female chosen for the playback always responded.