Elsevier

Behavioural Processes

Volume 81, Issue 3, July 2009, Pages 376-382
Behavioural Processes

The relationship of adult morphology and early social signalling of the domestic dog (Canis familiaris)

https://doi.org/10.1016/j.beproc.2009.03.019Get rights and content

Abstract

Most research on dog communication has focused on either the use of lupine signals during intra-specific interactions or has studied single-breed groups, with little regard to the effects of morphological change in the dog on their communication. This oversight ignores the fact that most modern dog breeds do not resemble the wolf and thus they have lost the ability to send some signals and may encounter communication problems. Using puppies in 40 litters from 32 breeds, we investigated the relationship between the frequencies of behaviours (social signals), and the adult morphology of these dogs. Despite a high level of precision, no statistically significant relationships were found between the deviation of the adult morphology from the ancestral lupine morphotype and early social signalling in the dog. We concluded that any communication loss due to morphological deviation, is not compensated for by increasing the frequency of signalling behaviours at this age.

Introduction

Domestication alters the behaviour and morphology of animals (Fox, 1978, Price, 1984) through selection for certain physiological, morphological and behavioural traits. These changes may be the result of direct selection, such as decreased aggression in captive foxes (Trut, 1999). Alternatively, changes may occur as a consequence of selection in other areas. The selection for behavioural traits favoured by humans can also have concomitant changes in morphological traits, such as coat colour, ear shape, tail shape and body size (Belyaev, 1979, Trut, 1999, Trut et al., 2006, Trut et al., 2004). The relationship between selection for behavioural traits and morphology was demonstrated in Trut et al. (2006), who bred silver foxes (Vulpes vulpes) under three behavioural selection treatments: tameness, aggression and unselected. They found that the three populations differed markedly in their morphology, and concluded, “there is good reason to believe that the differences have a genetic basis (pp. 90)”, and that their results suggest “a common genetic basis resulting in co-segregation of behaviour with morphology (pp. 92)”. Trut et al. (2006) also found that a principle component loading for limb length, skull width and size of lumbar and cervical vertebrae was related to tame behaviour in silver foxes. Selection for tameness in silver foxes has also affected communication (increased comprehension of human gestures) by foxes (Hare et al., 2005). Thus, it seems likely then that the intra-specific communication of the foxes may also be modified. The silver fox is not the only species to show a relationship between morphology (coat colour) and behaviour; similar relationships have been found in Deer Mice (Peromyscus maniculatus) (Hayssen, 1997a, Hayssen, 1997b), Norway Rats (Rattus norvegicus) (Hayssen, 1997a) and Icelandic Sheep (Dýrmundsson and Adalsteinsson, 1980). These findings lead to the reasonable conclusion that, if the genes for morphology and behaviour co-segregate, that selection for morphological traits, as has occurred in the domestic dog due to the show ring, may lead to related behavioural changes.

The domestic dog (Canis familiaris) has undergone strong morphological selection during domestication, since it descended from the wolf approximately 100,000 years before present (Vilà et al., 1997). While domestication of the dog may have occurred at this time, it is likely that morphological divergence from the wolf occurred only 10,000–15,000 years ago (Morey, 1994, Vilà et al., 1997), with many of the present morphotypes having an even shorter history. The present pure-breed morphotypes have largely arisen through adherence to “Breed Standards” which describe the behaviour and morphology of the breeds. Many of these breeds no longer retain the purpose for which they were originally bred (Hart, 1995), and are predominantly bred for the show ring or as household pets. About 50% of owned dogs in Australia are pure-breed dogs (McHarg et al., 1995), and selection in pure-breeds has focused on morphology (see Breed Standards issued by the Australian National Kennel Council) (Australian National Kennel Council, 2009). There is evidence that this focus on morphology may have led to unexpected behavioural changes in the dog. For example, selection for different coat colours in Cocker Spaniels (Podberscek and Serpell, 1996) and Labrador Retrievers (Houpt and Willis, 2001) has led to differences in the likelihood of these dogs showing aggressive behaviour, suggesting selection for morphological traits may have unforeseen consequences on behaviour.

Rapid changes in morphology in the dog from the lupine form have limited the signals that some dog breeds are capable of sending, by modifying or removing the structures necessary for signalling, which has likely led to difficulties in intra-specific communication in the dog (McGreevy and Nicholas, 1999, Price, 1984). This rapid change in morphology from the ancestral lupine form has led some authors (Bradshaw and Brown, 1990, Goodwin et al., 1997) to suggest that visual communication may have become so unreliable that compensation through other modalities, such as vocalisations (Feddersen-Petersen, 2000) or olfaction (Bradshaw and Brown, 1990, Goodwin et al., 1997), may be necessary.

While many authors have speculated about the effect of rapid morphological change on communication, Goodwin et al. (1997) is the only study to focus on the relationship between morphological deviation from the wolf and intra-specific communication in the dog. Goodwin et al. found that the size of the signalling repertoire (the number of different wolf-like signals) of a dog breed was negatively related to the breed's level of paedomorphosis (retention of juvenile morphological traits as an adult), and that the signalling repertoire of the different dog breeds corresponded to the repertoire of wolf pups at the same level of physical (morphological) development. These findings offer good evidence that the intra-specific communication repertoire of the dog has been affected by their morphological change from the wolf. However the full consequences of this change have not been studied.

The findings of Goodwin et al. (1997) show that dog breeds (or morphotypes) with different levels of neoteny/paedomorphosis have different sized signalling repertoires. Those dogs with higher levels of paedomorphosis would have smaller repertoires as they have developed fewer of the wolf signals on which dog communication is based. This would lead to a disparity between the communicative abilities of different dogs with different levels of paedomorphosis. Evidence of the possible effects of these differences in communicative ability was shown in Wickens (1993), who found that adults of different breeds showed different levels of visual signalling, with the paedomorphic Cavalier King Charles Spaniel (CKCS) showing little visual signalling, and instead, social status of the dogs was shown by the dogs displacing individuals from a resource, or allowing themselves to be displaced from the resource. The effects of deviation in a single morphological feature, tail length, were studied by Leaver and Reimchen (2008). They found that dogs were less likely to approach a life-size dog replica with a short tail than a replica with a long tail. They also found that larger dogs were more likely to approach the replica than were smaller dogs. The authors attributed these findings to the inability of the dogs to understand the “intentions” of the model with a short tail. The model used was similar to a Labrador Retriever, however, the effects of other morphological deviations of the model from the wolf, such as ear type and snout length, were not studied. This research does suggest that dogs with a shortened tail may have a smaller communication repertoire, than those with longer (full) tails.

A consequence of the above effect of morphological deviation may be that dog breeds with smaller communicational repertoires (more paedomorphic) may attempt to compensate by altering their communication, such as by developing new signals (Bradshaw and Brown, 1990, Goodwin et al., 1997), or altering the frequency of the signals that are still available but are not disguised by morphological deviation from the lupine form, such as was seen in the CKCS (Wickens, 1993). These compensations may be genetic, arising from the selection for morphological features, or learned through interactions with the dam (Slabbert and Rasa, 1997) and littermates. These compensations may appear at, and be confined to any stage during the development of the dogs from puppyhood to adulthood through the effects of development and experience. This study examines any possible compensation that may appear during the early homogenous litter environment, when access to individuals of similar morphology is the greatest, and exposure to other morphotypes is unlikely.

Dogs learn and refine the signals, and other behaviours, that regulate social interactions from 2.5 to 16 weeks (Fox, 1969, Frank and Frank, 1982). The period to 16 weeks of age is split into two under modern breeding practices (Pfaffenberger, 1963). The dogs spend the first eight weeks with their dam and littermates, and generally have no experience with other morphotypes. After eight weeks the dogs are generally rehomed, where they have less interactions with similar morphotypes (dam and littermates) and are more likely to be exposed to other new morphotypes. The differences between these two periods imply that they represent different study environments. It is only during the earlier period that any learned compensations, such as morphotype-specific signals and changes in signal frequency, are likely to be formed through development or learning of signals from littermates and the dam (Slabbert and Rasa, 1997). It is also only during this period that the puppies are not exposed to other morphotypes, and therefore their behaviour can be studied without the confounding effects of inter-morphotype experience.

Therefore, the early social signalling of 40 litters from 32 breeds ranging widely in morphology was examined. The relationships between the breed morphology and the frequency of social signals, that are not disguised by morphological deviation from the lupine form, were investigated in the first 8 weeks of life. The large variation in morphology between the 32 breeds allowed for high precision in estimating the relationships, by offering observations over the entire range of dog morphologies. This greatly increased the statistical power of the study.

Section snippets

Subjects

The study subjects were 40 litters from 32 dog breeds reflecting a wide range of dog morphotypes (Table 1). The litters were aged 5 weeks (±3 days) at the beginning of the study, and were sourced from the breeding populations of various professional organisations and hobby/show breeders in Victoria, Australia. This age range is within the range used in the pioneering studies of dog social behaviour by both Scott and Fuller (1971) and Fox (1978). All litters were housed with their respective

Results

The range of morphology scores for adult signalling structures was 2 (Malamute) to 14 (Old English Sheepdog), and there were breeds represented throughout this range (Fig. 1).

Discussion

The current study investigated the relationship between the deviation of adult morphology from the ancestral wolf morphotype and early social signalling in the domestic dog with high precision. Despite this high level of precision, no statistically significant relationships were found between morphological deviation and early social signalling in the dog. This suggests that any morphology-related signalling repertoire restriction found by Goodwin et al. (1997) is not being compensated for at

Acknowledgements

The authors would like to thank Petcare Information and Advisory Service for their help in funding this research. The assistance of Guide Dogs of Victoria, Australian Customs Service and the Victorian Police Dog Squad and other breeders was gratefully appreciated. We would also like to thank the staff of the Animal Welfare Science Centre for all their assistance. Keven Kerswell would like to thank A.K., S.K. and Y.K. for their support.

References (36)

  • J. Endler

    Some general comments on the evolution and design of animal communication systems

    Philos. Trans. R. Soc. Lond. Ser. B: Biol. Sci.

    (1993)
  • J.A. Endler

    Signal, signal conditions, and the direction of evolution

    Am. Nat.

    (1992)
  • J.A. Endler

    Evolutionary implications of interaction between animal signal and the environment

  • D.U. Feddersen-Petersen

    Vocalisation of European wolves (Canis lupus lupus L.) and various dog breeds (Canis lupus F. fam.)

    Arch. Tierzucht.

    (2000)
  • M.W. Fox

    Behavioral effects of rearing dogs with cats during the “critical period of socialization”

    Behaviour

    (1969)
  • M.W. Fox

    Behaviour of Wolves, Dogs and Related Canids

    (1971)
  • M.W. Fox

    The Dog: Its Domestication and Behaviour

    (1978)
  • B.L. Hart

    Analysing breed and gender differences in behaviour

  • Cited by (8)

    View all citing articles on Scopus
    1

    Present address: School of Biological Sciences, Faculty of Science, Monash University, Melbourne, Victoria 3800, Australia.

    View full text