Phylogenetic review of tonal sound production in whales in relation to sociality

Abstract

Background: It is widely held that in toothed whales, high frequency tonal sounds called 'whistles' evolved in association with 'sociality' because in delphinids they are used in a social context. Recently, whistles were hypothesized to be an evolutionary innovation of social dolphins (the 'dolphin hypothesis'). However, both 'whistles' and 'sociality' are broad concepts each representing a conglomerate of characters. Many non-delphinids, whether solitary or social, produce tonal sounds that share most of the acoustic characteristics of delphinid whistles. Furthermore, hypotheses of character correlation are best tested in a phylogenetic context, which has hitherto not been done. Here we summarize data from over 300 studies on cetacean tonal sounds and social structure and phylogenetically test existing hypotheses on their co-evolution.

Results: Whistles are 'complex' tonal sounds of toothed whales that demark a more inclusive clade than the social dolphins. Whistles are also used by some riverine species that live in simple societies, and have been lost twice within the social delphinoids, all observations that are inconsistent with the dolphin hypothesis as stated. However, cetacean tonal sounds and sociality are intertwined: (1) increased tonal sound modulation significantly correlates with group size and social structure; (2) changes in tonal sound complexity are significantly concentrated on social branches. Also, duration and minimum frequency correlate as do group size and mean minimum frequency.

Conclusion: Studying the evolutionary correlation of broad concepts, rather than that of their component characters, is fraught with difficulty, while limits of available data restrict the detail in which component character correlations can be analyzed in this case. Our results support the hypothesis that sociality influences the evolution of tonal sound complexity. The level of social and whistle complexity are correlated, suggesting that complex tonal sounds play an important role in social communication. Minimum frequency is higher in species with large groups, and correlates negatively with duration, which may reflect the increased distances over which non-social species communicate. Our findings are generally stable across a range of alternative phylogenies. Our study points to key species where future studies would be particularly valuable for enriching our understanding of the interplay of acoustic communication and sociality.

Figure 1

Optimizations of tonal sounds (a) and whistles (b) versus sociality using the broad concept approach [see additional file 1]. A brief glance at the black branches (indicating presence of tonal sounds/whistles and 'complex' sociality) on each side does not suggest detailed correspondence of acoustic structure with sociality. In other words whistles have a different phylogenetic distribution than does complex sociality etc, indicating that their co-evolutionary history (if any) may be more complicated than previously thought.

Figure 2

Optimization of group size in Cetacea (using natural log). Dark purple and blue colored branches indicate small groups and demark most of the 'basal' whales. More brightly colored (green, yellow and red) indicate larger groups. The phylogeny suggests gradual increase in group size in the lineage leading to Delphinidae, with independent evolution of huge groups (red) in several lineages and some reversals to smaller groups (e.g. Cephalorynchus hectori).

Figure 3

Optimizations of social structure as a four state character (a) leaving polymorphic species as such, (b) lowest social state, (c) highest social state. All analyses were done using the highest social state optimizations (see Methods).

Figure 4

Optimization of Cetacean tonal sound standard acoustic parameters (using natural log). Dark colors (purple and blue) indicate low values, while brighter colors (green, yellow, red) indicate higher values.

Figure 5

Regression analysis between independent contrasts of mean group size and mean number of inflection points. One conspicuous outlier (arrow) represents a contrast including the killer whale (Orcinus orca) which forms relatively small social groups but produces highly modulated whistles. It has been proposed that the killer whale uses whistles in a manner different from any other delphinid to indicate motivational state. That multiple factors are at work shaping tonal sounds in cetaceans may obscure and make difficult to discover true co-evolutinary histories of characters. Accordingly when O. orca is removed from the analysis the regression between the two characters becomes stronger.