Odors play a key role in mammalian communication, conveying information about identity, sex, social status, reproductive state, group membership and/or territorial boundaries 12,15,71. We sought to understand mechanisms of olfactory communication in brown bears by investigating the chemical components of skin-borne secretions used in scent marking. Although our sample size is relatively small (n = 12 individuals), to our knowledge we are the first to examine the chemical composition of brown bear cutaneous secretions beyond pedal scents and anticipate this as a starting point for further inquiry.
Chemical profiles code for age
We found strong evidence of age-dependent variation in VOC profiles of coastal brown bears. Age class was a significant predictor of the average number of compounds and chemical profiles varied according to age in composition and structure. Contrary to our prediction on age-dependent sexual dimorphism, age did not require grouping with sex to be a significant predictor of compound variation. Age-related differences in chemical profiles have also been found in giant pandas 27, red deer (Cervus elaphus) 72,73, Eurasian otters (Lutra lutra) 24,74, mandrills (Mandrillus sphinx) 56, and white rhinos (Ceratotherium simum) 75. In addition, behavioural indications of age signals in excretions and secretions have also been found in black rhinos (Diceros bicornis) 76 and giant pandas 77.
Age-related differences in chemical profiles have been shown to be hormonally regulated in mammals 78, which with physiological maturity, affect the chemical constituents of secreted scent 24,56. In this study, skin-borne secretions of young bears had a higher number of compounds than mature bears, and one specific compound contributed significantly to age-related differences in chemical profiles. Investigations into the influence of symbiotic microbes on mammalian odours have found similar variation in bacterial communities by age class. Endocrine signalling and microbiota have a bi-directional relationship, where hormones can drive shifts in microbial composition or microbes can transform hormones as well, thus altering VOC composition (reviewed by 79,80). This dynamic relationship can lead to differences in microbial composition related to age and reproductive cycle, thus driving changes in chemical composition associated with each (reviewed by 81). This phenomenon has been observed in striped hyenas (Hyaena hyaena), where scent pouch composition and structure varies by age-class, with adults possessing a core suite of microbiota and younger animals showing more variation 82. Similarly, the microbiota within anal glands of meerkats (Suricata suricatta) varied between males and females only after individuals were sexually mature, suggesting that physiological maturity was a catalyst for these changes 83. Age determination from chemical signals could have fitness advantages for both young and mature bears. With some variation, brown bears reach sexual maturity at ~ 5 years of age for both males and females 84. However, older male bears have higher reproductive success 85, females select larger males, and age is highly correlated with body size for males 86. Likewise, females between nine and twenty years old produce the most offspring 87. Age determination could function in mate selection and competitor assessment for mature bears, and may reduce risk to younger bears through honest signalling (sensu 88). For example, in Asian elephants, the musth of young males varies from mature males in composition, and is thought to convey a non-threatening chemical message of naivety to avoid conflict with older males 11. A signal of subordinance (low competitive ability) would also be advantageous for brown bears, as a solitary carnivore that exhibits breeding and foraging aggregations, and female matrilineal assemblages 89.
Brown bears in this study were divided into age classes based on the demographics of the dataset; the difference between the eldest bear in the class ‘young’ and the youngest in the class ‘mature’ was eight years. This ensured a representative division according to life stage, and we believe reduced error in assignment concerning young adult bears (5–6 years of age). Comparative studies should be aware of the potential for varying results based on age classification and between age groups in the analysis of age-related chemical signals (sensu 82).
Chemical profiles code for individuality within sex
Chemical signals of individuality are well documented in terrestrial mammals, including mustelids 66,90, rodents 91–93, primates 56,94 and carnivores 28,63,95. In line with these studies, our results suggest that individual identity is coded in both the composition (digital coding) and structure (analog coding) of brown bear chemical profiles. While we found no unique compounds between individual bears, our prediction on profiles coding individual identity was supported: compound incidence and abundance varied by individuals within sex and when individuals were grouped by age-sex class. Individual compound uniqueness also varied according to individual bears. Mammalian chemical profiles are known to be complex mixtures 96 of a variety of chemical components that vary in incidence and abundance, resulting in distinctive odours 28. Adding additional complexity, we found that compound uniqueness showed interindividual variation, with some compounds expressed in all samples from one individual, but absent from another. However, we did not find consistent compound uniqueness specific to each individual in the study.
The ability to discriminate individuals based on their chemical profile would provide considerable benefits for species that are able to retain this information for later use, and modify their behaviour in such a way that reduces risk and/or provides a benefit 92. The presence of individual scent signatures does not necessitate their use in individual recognition, nor that they have evolved for this function. Nonetheless, if individual recognition contributes to fitness, individually distinctive odours should be used by other animals to gain information on conspecifics and their use as a chemical signal would evolve. Individuals that advertised their distinctive odours (e.g., through scent marking) would then have a competitive or reproductive advantage, especially if their odour also contains coded information on age and sex. Bears are thought to possess the ability to recognise previous mates 97, kin 89,98, and other conspecifics 99,100, and scent marking has been proposed as a method of signalling dominance and mate attraction 9,35. Indeed, Morehouse et al. 101 recently found a positive relationship between tree rubbing and reproductive success for both male and female brown bears, and Hansen et al. 100 found that familiarity between female bears was important for home range settlement, which they suggest is facilitated in part by scent cues. Although brown bears are not a gregarious species, they do have some ecological traits (e.g., breeding and feeding aggregations, overlapping home ranges) that lead to conspecific interactions and result in a social hierarchy. Chemical signals that encode individuality could facilitate individual recognition within these social contexts.
Our finding that profile composition and structure varied according to individuals nested in sex, also indicates that some of the variability may be due to differences between sexes. We found no sex-specific differences in average number of compounds, but compound uniqueness varied by both body site and age when grouped by sex. Contrary to our prediction on mature male profile variance, mature male bears were the only age-sex class where composition and structure did not vary significantly between individuals, and mature male individual profiles showed less variance than mature females and other age-sex classes. Mature male brown bears are considered to be the dominant age-sex class based on behavioural observations 102,103, and have been shown to engage more in scent marking via tree rubbing than other age-sex classes, including frequency and time investment 45,46. As chemical profiles in brown bears appear to encode a signal for age (see above), which for other species conveys an honest signal of competitive ability 11, perhaps a mature male chemical signature - a signal of high competitive ability - is also present here. We are unable to develop this hypothesis further due to low sample size of mature males in our study (n = 2 individuals, 12 samples), however behavioural evidence indicates adult male-to-male signalling via scent marking in some populations 35,45 and evidence from spotted hyena (Crocuta crocuta) indicates that the scent profile of high-ranking females contains an ‘olfactory badge of status’ 104. Further work is needed to test this hypothesis on a larger sample size of adult male brown bears.
Body site variation and compound uniqueness
We found no body site-specific differences in the average number of compounds, and no significant variation in profile composition and structure between body sites, or when body sites were tested within individuals and grouped by age-sex class.
Our results indicated that compound uniqueness varied by body site and age within sex, and samples from the dorsal (hump) region of mature males showed the highest level of compound uniqueness across males; giving partial support to our prediction of increased variation of hump samples compared to those taken from other body sites. In behavioural studies, dorsal (back) rubbing is a core marking posture, particularly for adult males 46,105,106. Histological and histochemical analyses of the hump area of male bears have shown that sebaceous glands are enlarged and produce more oily secretion prior to and during the breeding season for intact males, influenced by testosterone concentration 41. Similar was found for apocrine glands during the breeding season, indicating that both sebaceous and apocrine glands in the back of male bears vary by season and reproductive status 40. As our samples were collected towards the end of- (July) and post-breeding season (August/September), chemical profiles may vary to those during the breeding season, especially in relation to body sites of male bears. Future studies should compare these results to samples taken during the breeding season, for both sexes. Histological and behavioural analyses in combination with the results presented here, provides a body of evidence that supports back rubbing as a focal method of chemical signalling for male brown bears.
It remains unclear why bears rub different body sites against trees (and other objects). Giant pandas are thought to mark urine using the handstand posture to deposit scent higher on trees, which communicates size and therefore competitive ability 3. Our results indicate that while different body site secretions may contain different compounds, very rarely are they unique to that body site across either individuals or age-sex classes. In addition, the lack of variation across body sites within chemical profiles poses the question: why are brown bears using multiple body sites when marking with cutaneous secretions? One explanation that we propose is that these glands are relatively small, compared to highly specialised scent glands such as anal glands or sacs, and produce a smaller volume of secretion compared to urine or faeces excretion. Indeed, Alberts 107 proposed that by marking on elevated surfaces, signalling animals increase the active space of the scent mark and that for certain species, hair may facilitate the distribution of scent over a larger area. Therefore, marking with multiple body sites may increase the surface area of the overall scent mark, but not necessarily provide different coded information for brown bears, at least outside of the breeding season.
The surface may also influence the body site used, e.g., pedal marking the substrate 38 compared to, or in combination with, tree marking. We suggest further analyses of brown bear cutaneous scent focus on hump and pedal body sites only, to reduce potential redundancy. An exception to this could be examining the flank secretions of female bears (Results) 46 and the hump secretions of males during the breeding season.
We explored signal content and found that VOC profiles coded for age and individual identity, when individuals were grouped by sex. A suite of compounds varied in both composition (digital coding) and structure (analog coding) within chemical profiles, according to these attributes. Individual compounds were rarely unique to specific ages, sexes, individuals or body sites, instead we found their presence to vary by body site and age for both males and females. In combination with existing histological, chemical and behavioural analyses, our results indicate the presence of encoded chemical signals in cutaneous glandular secretions of brown bears. This study addresses an ecological knowledge gap for bears which can contribute towards further understanding of mating systems and social behaviour.