Abstract: Multi-modal communication by male mantled howler monkeys [Jones & Van Cantfort]

Jones CB, Van Cantfort TE (2007) Multimodal communication by male mantled howler monkeys ( Alouatta palliata palliata Gray) in sexual contexts: a descriptive analysis. Folia Primatol 78: 166-185.

Abstract: We analyzed continuously sampled focal and ad libitum data of male mantled howler Cmonkeys (Alouatta palliata palliata Gray) observed in random order. Males resided in two groups in a Costa Rican tropical dry forest environment (riparian habitat group: 3 adult males, 15 adult females, 402 h observation; deciduous habitat group: 2 adult males, 8 adult females, 114 h observation). Samples were limited to sexual contexts, in particular, the 60-min periods before and after each copulation observed within each group for each adult male. Time samples for each male were distributed equally before and after their own copulations. Before statistical analyses were conducted, data were corrected for differences in time sampled for males within each group. Four types of multimodal signaling were resolved: (1) audiovisual, (2) olfactory-visual, (3) olfactory-visual-tactile, and (4) tactile-gustatory. Olfactory and tactile signals were never observed in combination with auditory signals. Consistent with expectation for a Neotropical, arboreal species, audiovisual signals were the most frequently observed type of multimodal communication in both groups (riparian habitat group: n= 139; deciduous habitat group: n= 66). Our evidence strongly suggests that unimodal signals may be combined and recombined to form complex, multimodal signals. Subordinate males in each group were more likely than dominant males to emit audiovisual signals before their own copulations. Male dyads were compared to assess the relative rate of audiovisual signaling by one male before another male's copulations. On average, the subordinate male of the riparian habitat group exhibited audiovisual signals at a higher rate before his own copulations compared to the rate of audiovisual signaling by his dominant challengers. The same comparisons are not significant for males in the deciduous habitat group. The pattern of male response that we report whereby subordinates emit some complex signals at a higher rate than dominants supports the "terminal investment hypothesis" predicting that organisms should increase reproductive effort with age since, in mantled howlers, age correlates negatively with dominance rank. Additional, qualitative observations suggested that subordinates in both groups were most likely to obtain copulations when they increased rates of complex signaling and/or escalated interactions with their male challengers. Group differences were apparent, however, and we suggest factors that may account for these patterns. We assessed responses by female receivers of complex signals emitted by males in sexual contexts. In general, higher-ranking males are more attractive to females and are more successful at monopolizing them. Findings for other, less frequently displayed, multimodal signals (olfactory-visual, olfactory-visual-tactile, and tactile-gustatory) are presented and discussed. We conclude with the suggestion that howlers may be a robust model for the investigation of complex signals in Neotropical primates, including research on functionally referential communication and context-dependent syntax.

Schematic Note On Mammalian Signaling

SCHEMATIC NOTE ON MAMMALIAN SIGNALING

Table 1: Sensory modalities of mammals (“Type of Signal or Display”), with examples, including, “Feature” or “Capacity” of Signal or Display in each of six domains. Signals or Displays may be emitted or advertized to one or more Receiver, and one or more animal (non-human or human) may be actively Eavesdropping (Figure 1). Depending upon physiological and behavioral traits of species, signal and display types represent a sensory “toolbox” with potential for multimodal (“complex”) transmission of information (Fig. 1). Both unimodal and multimodal communication are subject to temporal regulation, yielding varying sensory elements subject to statistical encoding, decoding, interpretation, and “prediction”. Despite potential for regulation of signal and display features, accurate signals are characterized by “redundancy”, a property that is highly correlated with reliability of communication (Wilson 1975).

On the other hand, messages are not necessarily selected for maximal accuracy*# in a single or a few contexts or domains since it may benefit individuals to utilize a signal or display in multiple contexts and for multiple functions, leading to one or more signals optimally effective across conditions. Furthermore, highly accurate messages may be easier to escape or avoid compared to “fuzzier” ones because they may be more detectable than (optimally) inaccurate signals. Thus, it may not benefit a Signaler to optimize production of a Receiver’s “pattern detection” mechanisms.

Signals and displays of “solitary” compared to social mammals are not necessarily less “complex” since different selection pressures may have favored different population structures and traits. However, excepting traits related to courtship and mating, signals and displays associated with within-group coordination and control (“integration”) are expected to be more differentiated, elaborate, and “complex” among social mammals because of higher local population density and, especially, higher rates of interaction. (©Clara B. Jones, based on Goodenough et al. 2009)

	TYPE OF SIGNAL OR DISPLAY
FEATURE OR CAPACITY	Visual (horns, proboscis, sexual swellings, mimicry, “natal coats”, stotting)	Auditory (howls, echolocation, ululation, “contact calls”, “alarm calls”, language)	Chemical (pheromone, excretions, venom)	Tactile (grooming, “neck bite”, copulation, spines, defensive integuments)	Electrical (monotremes)
Effective distance	Medium	High	High	Low	Low
Localization	High	Medium	Changeable	High	High
Ability to go around obstacles	Low	Medium	Medium	Medium	High
Detection of change	High	High	Low	High	High
Complexity	High	High	Low	Changeable	Low
Durability	Changeable	Low	High	Low	Low

*Recent research shows that inaccurate signals promote phenotypic flexibility. If this is generally the case, inaccurate signalling may be one alternative response that promotes coexistence, mutualism, and/or social behavior because the response may mitigate chances of competitive exclusion or aggression [aversive responses, punishment].

#See Sultan SE, Spencer HG 2002 Am Nat 160: 271-283, p 280, column 1, paragraph 2...

Figure 1: This figure displays a Signaler-Receiver schema in optimality (cost-benefit) context as follows. The classical, Ethological, view, adopted by behavioral ecologists, holds that “the actor [Signaler] is selected to manipulate the behaviour of the reactor [Receiver]”, and communication “makes sense only in the context of an exchange of information” (Dawkins and Krebs 1978) whereby a signal or display (Table 1) is employed by a Signaler to induce a self-interested, beneficial response. A corollary of the latter perspective would be that, as a recipient of information (the “signal” or “display”), a Receiver “decides” (Hebbian decision) to respond or not, becoming a potential “Sender”, responding in a manner, presumably, biased by the original “message” but, like the original transmission, biased by self-interest. Proximate and ultimate (reproductive) benefits or losses to Signalers will be a function of statistical averages, and Signalers whose messages, on average, produce a threshold-level of benefits “propagate their genes more efficiently” than Signalers who do not. Following Dawkins and Krebs (1978), “communication results in a net average benefit to the actor”.

Fig. 1 displays differential, initial benefits and costs to Signaler and Receiver; however, it remains unclear whether or to what extent, average benefits may accrue to Receivers, particularly, where signals or displays are “dishonest” (e.g., female mimicry of male genitalia, fake orgasm by females). The present discussion assumes that the condition-dependent, “fitness optima” of Signaler and the intended Receiver (or, unintended Receiver[s]) conflict. However, theory holds that some threshold of conflict will favor costly (“exaggerated”, “complex”), honest signals and displays, maximizing the transfer of “accurate” information and minimizing likelihoods of aggression (Tinbergen 1952). In these regimes, interactions between Sender and Receiver should be, “at least statistically, predictable from…past behaviour” (Dawkins and Krebs 1978). As Dawkins and Krebs (1978), Maynard Smith and Harper (2003), and Eisenberg (1981) point out, signals and displays may derive from changes in biological information emitted by a Signaler (“cues”), including, proximate or ultimate events associated with thermoregulation (e.g., “huddling”) and excretion (urination, defecation: e.g., orgasm). Calculations of differential benefits and costs are complicated by the recognition that such calculations (“decisions”) by “ego” are estimates of future gains and losses. ©Clara B. Jones (after Bradbury and Vehrencamp 1998)