Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Towards an understanding of multimodal traits of female reproduction in chimpanzees

Towards an understanding of multimodal traits of female reproduction in chimpanzees Although primates have long been regarded as microsmatic, recent studies indicate that olfaction is an important sensory mode of primate communication, e.g., in the context of reproduction. However, large gaps remain in our understanding of primate olfactory traits, especially in the great apes. Female chimpanzees (Pan troglodytes) possess an exaggerated sexual swelling, which is an imprecise signal of fertility that is thought to serve to confuse paternity. However, some high-ranking males that copulate most frequently on the days when females are most fertile seem to have more precise information on the timing of ovulation, which suggests the existence of an olfactory fertility trait. In order to examine, and provide evidence for, fertility-related chemical information in female chimpanzees, we used gas chromatography—mass spectrometry to analyze the chemical composition of female body odor collected across the menstrual cycle during various stages of sexual swelling (97 samples of six females). The chemical composition was significantly affected by swelling stage, and eight substances were detected that were strongly related to the latter. The existence of an additional, olfactory, fertility trait may help males to fine-tune their sexual behavior or allow females to strengthen concealment of the exact timing of ovulation, and needs to be further investigated in follow-up studies. The results of our study provide much-needed evidence for the existence of an olfactory cue related to reproduction in chimpanzees, and form a basis for future studies on the interplay between visual and olfactory information on female fertility. Keywords Olfactory communication · Olfaction · Gas chromatography—mass spectrometry · Chemical composition · Menstrual cycle · Anogenital swelling Introduction Primates utilize various sensory modalities, including vision, acoustics and olfaction, for social communication, Marlen Kücklich and Susann Jänig contributed equally to this study e.g., to convey information regarding kinship [olfaction in as first authors, and Brigitte M. Weiß and Anja Widdig contributed Lemur catta (Boulet et al. 2009); vision and acoustics in equally to this study as senior authors. Macaca mulatta (Pfefferle et al. 2014a, b)] or dominance rank [vision in Mandrillus leucophaeus (Marty et al. 2009); * Marlen Kücklich marlen_kuecklich@eva.mpg.de acoustics in Macaca nigra (Neumann et al. 2010); olfac- tion in Mandrillus sphinx (Vaglio et al. 2016)]. Sensory Behavioural Ecology Research Group, Faculty modalities involved in the transfer of specific information of Life Sciences, University of Leipzig, Talstraße 33, may act either alone or as multimodal traits, whereby differ - 04103 Leipzig, Germany ent modalities transmit either redundant or complementary Primate Behavioural Ecology Research Group, Department information (Partan and Marler 1999). The fertile phase of of Human Behavior, Ecology and Culture, Max Planck Institute for Evolutionary Anthropology, Deutscher Platz 6, female primates, e.g., is frequently indicated by a female’s 04103 Leipzig, Germany behavior, such as approaching males or reducing aggression, Mass Spectrometry Research Group, Institute of Analytical by visual traits such as sexual swellings, acoustic traits such Chemistry, University of Leipzig, Linnéstraße 3, as advertisement or copulation calls, as well as olfactory 04103 Leipzig, Germany Vol.:(0123456789) 1 3 366 Primates (2022) 63:365–376 traits such as scent-marking behavior, in various combi- these are not a precise signal of fertility (Emery Thompson nations [e.g., Cebuella pygmaea (Converse et  al. 1995); 2005). The phase of increased swelling lasts for approxi- Microcebus murinus (Buesching et al. 1998); Papio ursinus mately 10–12 days (Emery and Whitten 2003), and indicates (Clarke et al. 2009); Papio anubis (Rigaill et al. 2013)]. approaching ovulation, which occurs during the period of Whereas visual and acoustic traits of female fertility maximal swelling (Deschner et al. 2004). However, maximal have been relatively well studied across primates, studies swelling persists for up to 4 days after ovulation (Desch- on olfactory traits of fertility are comparatively scarce (Drea ner et al. 2004), and is thus only an imprecise signal of the 2015). However, over the years, it has become evident that latter. This imprecise information on the actual timing of olfaction may also play an important role in the transmission ovulation allows females to confuse paternity and reduce of social information in primates (e.g., Smith and Bhatnagar the risk of infanticide by attracting a high number of mating 2004; Drea et al. 2013). Studies in several primate species partners (Wrangham 2002; Kappeler and van Schaik 2004; indicate that females convey olfactory information regarding Stumpf and Boesch 2006). Males compete intensely with their reproductive state through whole body odor, vaginal each other over their access to females (Goodall 1986; Mul- secretions, urine and feces (Drea et al. 2015). Furthermore, ler 2002), which restricts their mating opportunities [pri- olfactory inspection of the female anogenital region by males ority of access model (Altmann 1962)]. The potential for increases during the fertile period [P. ursinus (Clarke et al. males to monopolize females is further reduced since several 2009); P. anubis (Rigaill et al. 2013)], and males react with females of a group will show exaggerated sexual swellings increased exploratory behavior and sexual activity [Macaca synchronously (Goodall 1986; Ostner et al. 2008). As males mulatta (Michael et al. 1971); Galago crassicaudatus (Clark should focus their mating efforts on the days when females 1982); Saguinus oedipus (Ziegler et al. 1993); Callithrix are most fertile in order to mate with as many fertile females jacchus (Smith and Abbott 1998); Macaca arctoides (Cerda- as possible, they would benefit from using precise indicators Molina et al. 2006b)] as well as a rise in testosterone levels of female fertility, if available. Indeed, male chimpanzees and activation of brain areas linked to sexual behavior [C. alter their sexual behavior when female anogenital swellings jacchus (Ziegler et al. 2005; Snowdon et al. 2006); M. arc- increase in size and throughout the period of maximal swell- toides (Cerda-Molina et al. 2006a)] when presented with ing (Klinkova et al. 2005), and mating seems to be correlated vaginal secretions from fertile females. with the period of peak fertility of females (Deschner et al. Specific substances related to the menstrual cycle of 2004; Emery Thompson 2005). Alpha males, in particular, females have recently been identified, and males were were found to copulate more frequently during the fertile observed to distinguish between female odor from differ - period compared to the other days of the maximal swell- ent phases of the cycle in C. jacchus [an aromatic hydro- ing phase (Deschner et al. 2004). These observations sug- carbon, a heterocyclic compound, an ester and an alcohol gest that high-ranking males may benefit from additional (Kücklich et al. 2019)]. Certain substances were also found information, e.g., a multimodal cue of fertility. Rigaill to be fertility related and to sexually stimulate male Old et al. (2013) hypothesized that the visual signal in P. anu- World monkeys [fatty acids in M. mulatta (Michael et al. bis could provide a first sign of fertility to all males within 1971)], although the results could not be reproduced in a visual range, whereas high-ranking males that can approach subsequent study [M. mulatta (Goldfoot et al. 1976)]. These females closely could potentially gather more precise infor- types of fertility indicators could constitute “signals”, i.e., mation from an olfactory inspection of females with sexual traits that have specifically evolved to provide information swellings. Also, females could benefit from a more precise on fertility, or “cues”, which constitute traits that have not indicator of ovulation that is only sensed by selected, com- evolved to deliver fertility-related information to conspecif- petitive males, to obtain the best genes and protection for ics, e.g., being by-products of changes in hormone levels their infants (Kappeler and van Schaik 2004). (Wyatt 2014). Observations of increased copulation rates at the time Chimpanzees (Pan troglodytes) are prime candidates for of highest fertility in females suggest that male chim- the study of multimodal traits of fertility, as female chimpan- panzees may indeed rely on olfactory traits in a sexual zees have several traits that may be indicative of their repro- context (Emery Thompson and Wrangham 2008). Certain ductive state during their menstrual cycle, which is around fatty acids that are thought to be associated with female 36 days long (Tutin 1979; Deschner et al. 2003; Emery and fertility in rhesus macaques (Macaca mulatta) (Michael Whitten 2003). They make copulation calls, although these et al. 1971) were also found in the vaginal secretions of seem not to vary with fertility (Townsend et al. 2011), and female chimpanzees, but it remains unclear if these vary show proceptive as well as mating resistance behavior that during the menstrual cycle (Fox 1982; Matsumoto-Oda varies between their fertile and non-fertile periods (Stumpf et al. 2003). Notably, fertility-related changes in the odor and Boesch 2006). In addition, they have exaggerated sexual of female chimpanzees could occur in genital odor and/ swellings, which vary in size across their cycle, although or in the scent of the whole body (hereafter “body odor”). 1 3 Primates (2022) 63:365–376 367 Several other individual attributes have been detected in In this study, we related the chemical profiles of body the body odor of chimpanzees and other great apes [e.g., odor samples of regularly cycling female chimpanzees to indicating individual identity (Hepper and Wells 2010); their stage of sexual swelling. We investigated whether arousal level (Klailova and Lee 2014); species identity and chemical composition varies synchronously with the vis- age (Jänig et al. 2019)]. Chimpanzees sniff other individu- ual signal, and predicted a change in the former across the als (i.e., genitalia, but also other body parts), and males sexual cycle. If this prediction were true, it would suggest specifically have been observed to sniff their conspecifics that anogenital swelling and body odor potentially comprise more often than females do (Matsumoto-Oda et al. 2007; multimodal traits of fertility. For the purposes of this study, Jänig et  al. 2018). Accordingly, we expect information we used data collected as part of a larger study on various about reproductive state to be provided by the body odor traits of great ape body odor [species identity, age and sex of female chimpanzees in general, and not only by their (Jänig et al. 2019)]. However, at the time of data collection, genital odor, which may constitute a multimodal trait of we were not able to obtain hormone samples from females fertility in combination with the visual signal of fertility, in a reliable manner that would have enabled us to deter- i.e., the sexual swelling. mine the exact timing of ovulation. Hence, given the lack Given the lack of studies demonstrating the existence of of hormonal data, we cannot show whether an olfactory cue olfactory fertility traits in chimpanzees, our overarching aim exists that is a more precise indicator of fertility than the was to provide evidence, based on chemical analyses, that visual signal. Our results nonetheless provide information olfactory information exists that is related to reproduction. on whether olfaction is related to reproduction at all, and Chimpanzees live in large, multi-male/multi-female groups, thus provide a foundation for future studies that include an which split up regularly into smaller parties of variable com- investigation of hormones in this context. position (Nishida 1968; Goodall 1986). Males preferably join parties which include females with maximal sexual swellings (Matsumoto-Oda 1999). If an olfactory fertility cue exists that is even more precise than the visual signal, Materials and methods this could potentially help males to specifically join parties with females close to ovulation. Cycle-dependent olfac- Subjects tory traits could be subtle changes that arise as a result of physiological changes, such as the breakdown of hormones We conducted our study on captive chimpanzees (Pan trog- involved in the menstrual cycle, e.g., luteinizing hormone, lodytes verus) at the Wolfgang Köhler Primate Research follicle-stimulating hormone, progesterone and estradiol Centre (WKPRC) at Leipzig Zoo, Germany, in 2013. We (Nadler et al. 1985). Sexual swelling appears to be induced analyzed samples from six adult females (aged between 13 by a rise in the levels of estrogens, and detumescence is and 20 years) from two different social groups (for details on correlated with a rise in progesterone after ovulation (Gra- housing and group composition, see supplementary Online ham et al. 1972). Hence, olfactory changes could occur as Resource), which had a regular menstrual cycle, i.e., did not by-products of these hormonal changes that accordingly fol- receive contraceptives (for at least 12 months prior to the low the timing of sexual swellings, and thus act as similarly study) and were not lactating. imprecise fertility traits as the visual signal. Ovulation can be assumed to occur 1 day before the increase in the level of progesterone (Deschner et al. 2004), and is related to pre- Sexual swelling stages ovulatory peaks of luteinizing hormone and follicle-stim- ulating hormone which are preceded by an estradiol peak The sexual swelling stage of each female was routinely mid-cycle (Nadler et al. 1985). Thus, a hormone-related recorded by the zookeepers and categorized as one of the olfactory change could also occur at the time of ovulation. following: flat, increasing in size, maximal tumescence, If this were true, olfactory information on fertility would decreasing in size. Mean cycle length was 35.6 ± 9.1 days be more precise than that provided by the visual signal. On and thus within the ranges previously published for chim- the other hand, an imprecise olfactory cue could strengthen panzees [captive (Nadler et al. 1985); wild (Tutin 1979; Wal- concealment of the actual timing of ovulation, and thus sup- lis 1997; Emery Thompson 2005)]. On average, the observed port the visual signal and allow females to confuse paternity. females were detumescent for 16.7 ± 5.4 days, the swellings A more precise olfactory cue, however, would allow high- increased in size for 4.3 ± 5.5 days, remained at maximal ranking males in particular to monitor females with maximal tumescence for 12.3 ± 3.4 days, and decreased in size for swelling to pinpoint the exact time of ovulation. Either way, 2.4 ± 1.5 days, similar to data reported for wild chimpanzees the composition of female body odor would be expected, at (Deschner et al. 2003). the very least, to vary between swelling stages. 1 3 368 Primates (2022) 63:365–376 et al. 2016). The extract was concentrated stepwise to a vol- Sample collection and preparation ume of circa 60 µL, and 4 µL of this solution was injected into the GC (HP6890 Series GC System, Agilent, Wald- The examined individuals are well trained in participat- ing in behavioral experiments by the WKPRC. Samples of bronn, Germany, with a HP5973 MSD Mass Selective Detector in electron-impact ionization mode at 70 eV) using skin odor were collected in the morning prior to behavio- ral experiments. For the purpose of our study, individuals splitless injection, for 2 min. The GC was equipped with a J&W Fisher DB35-MS column (30 m length, 0.25 mm inner were trained by positive reinforcement (using food items) to come to the grid of the cage to participate voluntarily in diameter, 0.25 µm film; Agilent), with the inlet tempera - ture set to 250 °C. Helium was used as the carrier gas at a the sampling by presenting a self-selected body part. Clean Lilibe cotton swabs (60% cotton wool, 25% polyester micro- flow rate of 1.7 mL/min. The temperature program ran for 40.5 min, starting at 35 °C for 2 min, followed by a heating fiber, 15% polyester), heated for 30 min at 130 °C before use (Birkemeyer et al. 2016), were held with sterilized metal step of 10 °C/min until 320 °C, which was held for 10 min. The solvent delay was set to 7 min. The ion source operated tweezers and rubbed repeatedly over the skin and fur for approximately 20 s (similar to Célérier et al. 2010; Stoffel at 250 °C and the scan range was set to m/z 50–550. et al. 2015). The cotton swabs were immediately placed in pre-cleaned (washed with methanol and diethyl-ether) 4-mL Data processing glass vials (Rotilabo) and stored at − 80 °C until gas chro- matography—mass spectrometry (GC–MS) analysis. From the GC–MS data we identified peaks, their retention times (RTs) and areas (intensity) using AMDIS v. 2.65 Given that the individual chimpanzees had full control over which body part they presented, the samples included (Stein 1999). To determine substances that were repeatedly detected in our samples, we grouped consecutive RTs to one odor from the arms, belly, back and legs, but not the genitals, as these were not readily presented by the animals. Hence, RT range (see also Weiß et al. 2018) using a self-written script performed in R version 3.2.3 (R Core Team 2015). we sampled body parts that have the same types of apocrine and eccrine sweat glands (Ellis and Montagna 1962; Mon- We verified by manual inspection that mass spectral pat- terns of the peaks within a given RT range were consistent, tagna and Yun 1963) representing general body odor rather than genital odor. Body and genital odor should be similarly i.e., each RT range was assumed to reflect one substance, resulting in 152 RT ranges. In the following, we thus use the affected by hormonal changes across the menstrual cycle, as previously shown in humans (Michael et al. 1974, 1975; term “substance” rather than “RT range.” At that stage of the study, the substances had not yet been chemically identified Gildersleeve et al. 2012). Furthermore, behavioral observa- tions of both chimpanzee groups at the WKPRC revealed by using a library search, and were described only by their RT and specific patterns of signals with certain mass-to- that when males sniff females they focus on the genitals in just over half of cases and other body parts in the remaining charge-ratios (m/z). We used a non-targeted approach to find substances that cases (Jänig et al., in review). We controlled for the different body parts in our statistical analysis. Samples were not col- might be related to female fertility. We excluded 15 sub- stances from the statistical analysis which had the same or lected from the mouth, hands or feet to reduce contamination with odor from food or the environment. Furthermore, diet a higher abundance in the control than in the animal sam- ples (i.e., data on 137 of the 152 substances remained for was likely to have only a minor impact on body odor vari- ance within and between individuals, since animals were fed analysis). For all the substances which were found to be potentially related to female fertility according to the statisti- daily with similar food. In total, 97 samples [16.2 ± 2.3 samples per individual cal analysis, mass spectral comparisons using the National Institute of Standards and Technologies (NIST) Mass Spec- female covering from five to eight (mean 7.2) menstrual cycles each)] were collected from the animals (for details tral Library (NIST 14 software; NIST, Gaithersburg, MD) were conducted for their tentative identification. We report see supplementary Online Resource Table 1). In addition, 42 control samples (pure cotton swabs not rubbed over the the best library hit per substance when (1) the top library hit was consistent over samples, (2) the match had a prob- skin or fur but otherwise handled like the animal samples; one per sampling day) were collected to identify chemical ability > 80 (NIST 14 value), and (3) the tentatively iden- tified compound could be reasonably expected to elute at substances that did not originate from the chimpanzees. the given RT. Alternatively, where no library hit allowed tentative identification, we propose a structural classification Chemical analysis based on interpretation of the spectrum. The mass spectra of the substances discussed in this article are provided in We extracted the chemical substances for GC–MS analysis by adding 1.2 mL of n-hexane (Sigma Aldrich, Steinheim, supplementary Online Resource Fig. 2. Germany) to the cotton swab in a glass vial (Birkemeyer 1 3 Primates (2022) 63:365–376 369 information, see Weiß et al. (2018)]. Additionally, swelling Statistical analysis stage was fitted as a random slope within ID of the female to achieve more accurate estimates of the test predictor. First, we conducted an analysis of similarities (ANOSIM) to test whether chemical profiles of females at the same The check for normal distribution and homogeneity of the residuals by inspection of a qq-plot suggested no violation swelling stage were more similar to each other than profiles of females at different swelling stages. Following common of assumptions. Plotting the residuals against fitted values revealed a slight bottom effect, but model stability indicated practice (e.g., Stoffel et al. 2015), the ANOSIM was based on Bray–Curtis indices calculated from the log-transformed, no influential cases. We determined variance inflation factors (VIF) to check for potential collinearity (Quinn and Keough standardized intensities of substances (intensity of substance divided by the summed intensity of all 137 substances × 2002; Field 2005) calculated with the function vif of the package car (Fox and Weisberg 2011) and found no indica- 100) for each combination of sample dyads. To control for repeated measurements per female, we used a customized tion of collinearity (VIF = 1.01). max We compared the full model to a null model excluding the ANOSIM (R script written by LK) that computed P-values by permuting swelling stages within individual females only. random slope within substance by using a likelihood ratio test (LRT) (Barr et al. 2013) to determine the significance The α-level was set to 0.05 for all statistical analyses. Second, we implemented two two-tailed generalized lin- of the full model (Forstmeier and Schielzeth 2011). For the significant full model, the slope estimates of all substances ear mixed models in R v. 3.2.3 (R Core Team 2015) using the package lme4 v. 1.1.11 (Bates et al. 2015) to determine if were extracted from the model results. To determine the sub- stances whose levels changed most with swelling stage, we specific substances are associated with a particular swelling stage. Chemical datasets of animals may present several ana- compared slope estimates to the average slope estimate and focused on those substances whose absolute slope estimate lytical challenges, such as (1) substances of potential biolog- ical relevance may have very high, or very low intensities; was higher than the average absolute slope estimate + 2SDs (Weiß et al. 2018). (2) the total intensities of samples can change due to envi- ronmental factors; and (3) the intensity of certain substances (e.g., those considered contaminants, such as plasticizers) can vary strongly, and thus affect the relative composition Results of odor profiles (van den Berg et al. 2006). To overcome these problems, we fitted the two models using two differ - Chemical profiles of females at the same swelling stage were significantly more similar than profiles of females at ent responses: standardized (intensity of substance divided by the summed intensity of all 137 substances per sample different swelling stages (ANOSIM, R = 0.07, P = 0.026), although the effect was rather small when whole profiles × 100) and transformed [arcsine and log(x + 0.01)] peak areas; standardized peak areas centered (to a mean of 0) and were considered. The chemical composition of substances with log-trans- scaled (to a SD of 1). In both cases, standardization was used to correct for changing total sample intensities. In the first formed abundances was affected by swelling stage (LRT, χ = 10.77, P = 0.013). Thus, the relation of these affected approach, transformation was applied to achieve a normal distribution and to reduce the relative impact of large sub- substances to the rest of the respective odor profile changed depending on the swelling stage. This effect was most pro- stances, while differences in the relative abundances of the various substances were maintained. In the second approach, nounced for four substances (i.e., their absolute slope esti- mate was higher than the absolute slope estimate + 2SDs). centering and scaling was applied to adjust all the substances to the same size and to give all of them equal weight in the These substances were all tentatively identified as steroids (Table 1), and were least abundant in samples collected dur- analyses (van den Berg et al. 2006). The responses were vectorized from a multivariate data matrix (Jamil et al. 2013) ing the maximal swelling as well as the flat phase, and most abundant during the increase and decrease phases of swell- of samples (n = 97) and substances (n = 137). Swelling stage (dummy coded and centered) was fitted as fixed effect test ing (Fig. 1). For substances with centered and scaled abundances, predictor, and age (z-transformed) of the females was fitted as control predictor. Sample number and substance identity chemical composition was also affected by swelling stage (LRT, χ = 19.28, P < 0.001). In these cases, the substance- (ID) were included as random effects to prevent pseudo- replication as well as heteroscedastic variance due to the specific abundances changed between the profiles of dif- ferent swelling stages. The swelling stage effect was most vectorized data matrix (Jamil et al. 2013). Other random effects were the day of observation, the sampled body part, pronounced for four of the substances (i.e., their absolute slope estimate was higher than the absolute slope esti- as well as ID and group of the female. We included the ran- dom slope of the fixed effects test predictor (swelling stage) mate + 2SDs), which were different from the most affected substances of the log-transformed model. One of these within substance as the actual test predictor [for detailed 1 3 370 Primates (2022) 63:365–376 Table 1 Substances most affected by swelling stage with log-trans- classification, substance class and the largest differences between the formed abundances, including retention times (RTs), tentative identifi- absolute slope estimates of swelling stages (average absolute slope cation [National Institute of Standards and Technologies (NIST) Mass estimate ± SD, 0.096 ± 0.08) Spectral Library; NIST  14] with probability (best hit) or structural RTs Best library hit/structural classification Probability Substance class Slope estimates 42.40 Unknown steroid (cholestadiene-like) – Steroid 0.59 43.49 Cholesta-2,4-diene 94 Steroid 0.32 44.39 Cholesta-3,5-diene 94 Steroid 0.38 48.46 Cholesta-5-en-3-ol (3.beta.)-, acetate 98 Steroid 0.38 For more details, see Fig. 1 and supplementary Online Resource Table 2; for mass spectra, see supplementary Online Resource Fig. 2 Fig. 1 a–d Log-transformed intensity (%) of the four substances most influenced by swelling stage (see also Table 1). Boxplots show medians and first and third quartiles. Retention times (RTs) and standardized peak area means ± SD are as follows: a RT 42.40, 1.97 ± 2.46; b RT 43.49, 0.48 ± 0.43; c RT 44.39, 13.71 ± 8.45; d RT 48.46, 0.55 ± 0.56 substances was tentatively identified by the library search Discussion as a wax ester (Fig. 2a), whereas two substances had spec- tra with m/z patterns typical of long-chain alkylic struc- The results of the present study support the presumption tures (Fig. 2b, d), and one substance was thought to feature that olfaction is related to reproduction in chimpanzees, a phenyl substructure (Fig. 2c) based on its structural clas- as the chemical profiles of the female chimpanzees var - sification (see Table  2). Three of the substances were most ied significantly with sexual swelling stage. Initial evi- abundant in samples collected during maximal swelling dence was provided by the slightly greater similarities of compared to the other swelling stages, while one of the the whole chemical profiles of samples taken during the substances with an alkylic (sub)structure was most abun- same swelling stage compared to those taken at different dant during the flat phase and least abundant in samples swelling stages. These findings were corroborated by a collected during maximal swelling (see Fig. 2). 1 3 Primates (2022) 63:365–376 371 Fig. 2 a–d Centered and scaled intensity (%) of the four substances most influenced by swelling stage (see also Table 2). Boxplots show medi- ans and first and third quartiles. RTs and standardized peak area means ± SD are as follows: a RT 32.26, 0.26 ± 1.05; b RT 34.65, 0.12 ± 0.25; c RT 40.36, 0.03 ± 0.02; d RT 48.80, 0.35 ± 0.31 Table 2 Substances most affected by swelling stages with cen- tion, substance class and the largest differences between the abso- tered and scaled abundances, including RTs, tentative identification lute slope estimates of swelling stages (average absolute slope esti- (NIST  14 library) with probability (best hit) or structural classifica- mate ± SD, 0.273 ± 0.20) RT Best library hit/structural classification Probability Substance class Slope estimates 32.26 Dodecanoic acid, isooctyl ester 91 Ester 0.87 34.65 Unknown [long-chain alkylic (sub)structure] – Unknown 0.91 40.36 Unknown phenol – Phenol 0.67 48.80 Unknown [long-chain alkylic (sub)structure] – Unknown 0.86 For more details, see Fig. 2 as well as supplementary Online Resource Table 2; for mass spectra, see supplementary Online Resource Fig. 2. For abbreviations, see Table 1 significant change in chemical composition across swell- log transformed or centered and scaled when used in the ing stages that was most pronounced with respect to eight models. In the case of log-transformed peak areas, affected substances. Thus, the results reveal that olfactory changes substances differ in contrast to the remaining chemical pro- exist that mirror changes in fertility at least as closely as file. Hence, information on the state of fertility seems to be the visual signal of sexual swelling. However, our meth- provided by the status quo of the chemical composition, and odological approach did not allow us to address whether comparisons over time are thus not necessary. When peak olfactory cues provide more precise information which areas are centered and scaled, the analysis is focused on dif- could allow males to pinpoint the time of ovulation more ferences between, instead of within, the samples (van den precisely. Hence, further studies that include hormonal Berg et al. 2006). Thus, particular substances that are found measurements are needed to unravel the temporal dynam- to be affected when peak areas are centered and scaled dif- ics of olfactory cues in the context of reproduction. fer in their levels between samples (i.e., over time or swell- The effect of swelling stage on chemical composition was ing stages). In this scenario, it seems to be the variation tested with two different approaches. Peak areas were either (or differences) over the course of the menstrual cycle that 1 3 372 Primates (2022) 63:365–376 provides information on fertility stage. A male would thus all the females and swelling stages. All of the substances need to regularly check a female’s odor over time to unravel were found for all six females as well as at all of the swell- this information. As expected, the two approaches revealed ing stages, and most of them were present in at least 80% different chemical substances that were most affected by of the samples for at least two swelling stages. However, swelling stage. two of the substances indicated by the centered and scaled Four chemical substances were found to be affected by model (an alkylic compound with an RT of 34.65 min and an swelling stage when abundances were log-transformed. All unknown phenol with an RT of 40.36 min) were only present of them were cholesterol derived, and were more abundant in a  maximum of 43% of samples per swelling stage (maxi- during the increasing and decreasing swelling stages com- mal swelling and increasing/decreasing swelling stages). If pared to the maximal swelling and flat stage. Cholesterol- substances are not always present in the chemical profile of derived substances are ubiquitous in the chemical profiles female odor, they are not very suitable for use by males as of mammals (Charpentier et al. 2012) and are precursors reliable indicators of female reproductive state. Future stud- for smaller molecules produced by bacterial degradation ies that include a higher number of samples of female odor (Ezenwa and Williams 2014). Specifically, cholesterol is the will help researchers to examine this subject in more detail. precursor for steroid hormones such as estradiol and proges- While our results indicate that body odor does indeed terone (Hu et al. 2010). Swelling of the anogenital skin is change over a female’s reproductive cycle, our methodology estrogen dependent, and detumescence occurs after a rise in did not allow us to assess whether olfactory changes are as progesterone, which inhibits the effects of estrogen (Gillman precise as the visual signal, or potentially even more pre- and Stein 1941; Graham et al. 1972; Emery and Whitten cise. In fact, in the context of an evolutionary arms race, we 2003). Tumescence of the sexual swelling occurs due to the would expect males to strive to pinpoint the exact timing of intracellular accumulation of water in the anogenital tissue, ovulation by using a fertility trait, and thus that they would which is lost during detumescence (Krohn and Zuckerman benefit from an olfactory cue that is more precise than the 1937; Clarke 1940). The consistent pattern of the four most visual cue of sexual swelling. Females, on the other hand, affected substances, i.e., that their levels were highest during would profit from generally confusing paternity by provid- tumescence and detumescence, leads to the assumption that ing the rather imprecise visual signal of fertility which could these chemical changes arise due to physiological changes be strengthened by an additional, imprecise olfactory cue. that occur during swelling. This chemical information being However, high-ranking male chimpanzees in particular have constistent with the visual signal of increasing or decreas- been observed to start copulating only in the last 3–4 days ing swelling, could enable males to recognize the beginning of maximal swelling (Deschner et al. 2004). As the duration and end of the female fertile period, allowing them to focus of maximal swelling is variable and its termination cannot their monopolization and mating efforts on females which be deduced according to the point at which it starts, such are most fertile. observations suggest that high-ranking male chimpanzees Moreover, four substances with centered and scaled abun- in particular, which are able to approach females to within a dances were found to be related to swelling stage. Similar very close distance, may rely on additional and more precise types of substances, such as long-chain fatty acid esters, sub- indicators of ovulation, i.e., presumably olfactory traits. In stances with alkylic structures, and phenols, were previously humans, odor has been proposed as a more precise indica- detected, e.g., in skin emanations [humans (Bernier et al. tor for (approaching) ovulation than visual signs (Singh and 2000)] and in genital secretions [L. catta (with the excep- Bronstad 2001; Havliček et al. 2006; Haselton and Gilder- tion of phenols) Boulet et al. 2009]. These four substances sleeve 2011; Gildersleeve et al. 2012). Relying on more than were related to the maximal swelling stage, as three of them one of the senses for the inspection of a female is likely to be were most abundant and the other one was least abundant advantageous for male chimpanzees too. The visual signal during this phase compared to the other swelling stages. could give initial information about the state of fertility of a Overall, our results suggest that the chemical composition of female at short/medium distance, which may then especially female chimpanzee odor changes over the menstrual cycle in attract high-ranking males with priority of access, allowing accordance with sexual swelling stage. Thus, body odor can them to undertake an olfactory inspection at close proximity be considered at least as a proxy of fertility similar to swell- to gain additional information (Matsumoto-Oda et al. 2007). ing stage, which has been well established as an imprecise Such olfactory information could potentially be non-redun- fertility signal. dant to the visual information (see also Higham and Hebets An olfactory trait related to reproduction is expected 2013), but our study design did not allow us to determine to be present consistently within the chemical profile of whether the olfactory cue is redundant or not. female odor to allow males to infer female fertility from it. Thus, an  important next step building on our  current Thus, we checked whether the most affected substances as results, that show that chemical composition changes over shown by the two models were continuously present across the menstrual cycle in accordance with stages of sexual 1 3 Primates (2022) 63:365–376 373 from the University of Leipzig was awarded to SJ (DFPL R00061) and swelling, would be  investigating whether chemical profiles MK (DFPL R00017). indeed provide more precise information than sexual swell- ings. To investigate this in more detail, analyses of cycle- Data availability The dataset generated and analyzed during the current related hormones are needed to determine the exact timing study is available on reasonable request. of ovulation, which could not be accomplished at the time that this study was conducted. This could be undertaken by Declarations using non-invasive hormonal measurements, as applied in previous studies to assess the timing of ovulation in non- Conflict of interest The authors declare that they have no conflict of interest. human primates (Deschner et al. 2004; Engelhardt et al. 2006; Dubuc et al. 2012; Young et al. 2013; Douglas et al. Ethical approval This study was carried out in accordance with the 2016). legal requirements of Germany and all national and institutional guide- Moreover, recently adapted sampling techniques now lines for the care and use of animals, and was approved by the ethics commission of the Department of Psychology of the Max Planck Insti- allow a wider range of chemical substances to be sampled tute for Evolutionary Anthropology and Leipzig Zoo. (Kücklich et al. 2017). The traditionally used cotton swab, which was also used in this study, primarily captures semi- Open Access This article is licensed under a Creative Commons Attri- to non-volatile substances (Birkemeyer et al. 2016). These bution 4.0 International License, which permits use, sharing, adapta- types of larger molecules are then degraded by skin bacte- tion, distribution and reproduction in any medium or format, as long ria into smaller, volatile molecules that constitute the odor as you give appropriate credit to the original author(s) and the source, profile of an individual (Ezenwa and Williams 2014) that provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are can be perceived by the main olfactory system (Dulac and included in the article's Creative Commons licence, unless indicated Torello 2003). Future studies will certainly benefit from otherwise in a credit line to the material. If material is not included in these recent methodological developments for the direct the article's Creative Commons licence and your intended use is not sampling of volatile substances, and for the correlation of permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a their abundances to individual traits of animals. copy of this licence, visit http://cr eativ ecommons. or g/licen ses/ b y/4.0/ . In conclusion, this study adds much-needed information on the role of olfaction in great apes by showing that the chemical composition of body odor is associated with a vis- ual signal of fertility in chimpanzees. Our results indicate References that olfactory cues related to reproduction exist in a great ape species with visual fertility advertisement, and provide a Altmann SA (1962) A field study of the sociobiology of rhesus mon - valuable basis for further research on more specific questions keys, Macaca mulatta. Ann NY Acad Sci 102:338–435. https:// on reproduction in chimpanzees and olfaction in great apes doi. org/ 10. 1111/j. 1749- 6632. 1962. tb136 50.x Barr DJ, Levy R, Scheepers C, Tily HJ (2013) Random effects in general. In fission–fusion societies, where group compo- structure for confirmatory hypothesis testing: keep it maximal. sition varies and visual information is short lived, olfaction J Mem Lang 68:255–278. https:// doi. org/ 10. 1016/j. jml. 2012. may allow males to fine-tune their sociosexual behavior and 11. 001 to optimize their reproductive investment. Bates D, Maechler M, Bolker B, Walker S (2015) lme4: linear mixed-effects models using Eigen and S4, R package. Version 1.0–8. http://CRAN .R-pr oject. or g/pac kag e=lme4 . Accessed 20 Supplementary Information The online version contains supplemen- Jan 2021 tary material available at https://doi. or g/10. 1007/ s10329- 022- 00995-1 . Bernier UR, Kline DL, Barnard DR et al (2000) Analysis of human skin emanations by gas chromatography/mass spectrometry. 2. Acknowledgements We thank Josep Call and the curators of the Leip- Identification of volatile compounds that are candidate attract- zig Zoo for permission to sample the chimpanzees at the WKPRC, and ants for the yellow fever mosquito (Aedes aegypti). Anal Chem are most grateful for the help of the keepers at Leipzig Zoo and the staff 72:747–756. https:// doi. org/ 10. 1021/ ac990 963k of the WKPRC. Furthermore, we also acknowledge the help of Susan Birkemeyer C, Thomsen R, Jänig S et al (2016) The sampling of Billig and Andreas Kiontke for technical assistance with the chemical body odour of primates: cotton swabs sample semi-volatiles analysis. We thank Ruth Thomsen for logistic support during sample rather than volatiles. Chem Sens 41:525–535. https:// doi. org/ collection, and Roman Wittig and the anonymous reviewers for com- 10. 1093/ chemse/ bjw056 ments on earlier versions of the manuscript. Finally, we thank the Max Boulet M, Charpentier M, Drea C (2009) Decoding an olfactory mech- Planck Institute for Evolutionary Anthropology and the University of anism of kin recognition and inbreeding avoidance in a primate. Leipzig for their logistic support. BMC Evol Biol 9:281. https:// doi. org/ 10. 1186/ 1471- 2148-9- 281 Buesching CD, Heistermann M, Hodges JK, Zimmermann E (1998) Funding Open Access funding enabled and organized by Projekt Multimodal oestrus advertisement in a small nocturnal prosim- DEAL. This work was supported by the German Research Foundation ian, Microcebus murinus. Folia Primatol 69:295–308. https:// within the Emmy-Noether program (grant no. WI 1808/3-1 awarded to doi. org/ 10. 1159/ 00005 2718 AW) and by the University of Leipzig. In addition, graduate funding 1 3 374 Primates (2022) 63:365–376 Célérier A, Huchard E, Alvergne A et  al (2010) Detective mice of Kanyawara, Kibale National Park. Int J Primatol 29:885–905. assess relatedness in baboons using olfactory cues. J Exp Biol https:// doi. org/ 10. 1007/ s10764- 008- 9286-1 213:1399–1405. https:// doi. org/ 10. 1242/ jeb. 038349 Emery MA, Whitten PL (2003) Size of sexual swellings reflects Cerda-Molina AL, Hernandez-Lopez L, Chavira R et al (2006a) Endo- ovarian function in chimpanzees (Pan troglodytes). Behav crine changes in male stumptailed macaques (Macaca arctoides) Ecol Sociobiol 54:340–351. https:// doi. or g/ 10. 1007/ as a response to odor stimulation with vaginal secretions. Horm s00265- 003- 0648-6 Behav 49:81–87. https:// doi. org/ 10. 1016/j. yhbeh. 2005. 04. 014 Engelhardt A, Heistermann M, Hodges JK et al (2006) Determi- Cerda-Molina AL, Hernandez-Lopez L, Rojas-Maya S et al (2006b) nants of male reproductive success in wild long-tailed macaques Male-induced sociosexual behavior by vaginal secretions in (Macaca fascicularis)—male monopolisation, female mate Macaca arctoides. Int J Primatol 27:791–807. https:// doi. org/ choice or post-copulatory mechanisms? Behav Ecol Sociobiol 10. 1007/ s10764- 006- 9045-0 59:740–752. https:// doi. org/ 10. 1007/ s00265- 005- 0104-x Charpentier MJE, Barthes N, Proffit M et al (2012) Critical thinking Ezenwa VO, Williams AE (2014) Microbes and animal olfactory in the chemical ecology of mammalian communication: road- communication: where do we go from here? BioEssays 36:847– map for future studies. Funct Ecol 26:769–774. https://doi. or g/ 854. https:// doi. org/ 10. 1002/ bies. 20140 0016 10. 1111/j. 1365- 2435. 2012. 01998.x Field A (2005) Discovering statistics using SPSS. Sage, London Clark AB (1982) Scent marks as social signals in Galago crassicau- Forstmeier W, Schielzeth H (2011) Cryptic multiple hypotheses test- datus. I. Sex and reproductive status as factors in signals and ing in linear models: overestimated effect sizes and the winner’s responses. J Chem Ecol 8:1133–1151. https:// doi. org/ 10. 1007/ curse. Behav Ecol Sociobiol 65:47–55. https://doi. or g/10. 1007/ BF009 86984s00265- 010- 1038-5 Clarke RW (1940) Water distribution and sexual skin of the baboon. Fox GJ (1982) Potentials for pheromones in chimpanzee vaginal fatty Am J Physiol Legacy Content 131:325–330. https://doi. or g/10. acids. Folia Primatol 37:255–266. https:// doi. org/ 10. 1159/ 00015 1152/ ajple gacy. 1940. 131.2. 325 6036 Clarke PMR, Barrett L, Henzi SP (2009) What role do olfactory cues Fox J, Weisberg S (2011) An {R} companion to applied regression, play in chacma baboon mating? Am J Primatol 71:493–502. second edition. Sage, Thousand Oaks, CA. http:// socse rv. socsci. https:// doi. org/ 10. 1002/ ajp. 20678mcmas ter. ca/ jfox/ Books/ Compa nion. Accessed 20 Jan 2021 Converse LJ, Carlson AA, Ziegler TE, Snowdon CT (1995) Com- Gildersleeve KA, Haselton MG, Larson CM, Pillsworth EG (2012) munication of ovulatory state to mates by female pygmy mar- Body odor attractiveness as a cue of impending ovulation in mosets, Cebuella pygmaea. Anim Behav 49:615–621. https:// women: evidence from a study using hormone-confirmed ovula- doi. org/ 10. 1016/ 0003- 3472(95) 80194-4 tion. Horm Behav 61:157–166. https:// doi. org/ 10. 1016/j. yhbeh. Deschner T, Heistermann M, Hodges K, Boesch C (2003) Timing 2011. 11. 005 and probability of ovulation in relation to sex skin swelling in Gillman J, Stein HB (1941) A quantitative study of the inhibition of wild West African chimpanzees, Pan troglodytes verus. Anim estradiol benzoate by progesterone in the baboon (Papio por- Behav 66:551–560. https:// doi. org/ 10. 1006/ anbe. 2003. 2210 carius). Endocrinology 28:274–282. https:// doi. or g/ 10. 1210/ Deschner T, Heistermann M, Hodges K, Boesch C (2004) Female endo- 28-2- 274 sexual swelling size, timing of ovulation, and male behavior Goldfoot DA, Kravetz MA, Goy RW, Freeman SK (1976) Lack of effect in wild West African chimpanzees. Horm Behav 46:204–215. of vaginal lavages and aliphatic acids on ejaculatory responses in https:// doi. org/ 10. 1016/j. yhbeh. 2004. 03. 013 rhesus monkeys: behavioral and chemical analyses. Horm Behav Douglas PH, Hohmann G, Murtagh R et al (2016) Mixed messages: 7:1–27. https:// doi. org/ 10. 1016/ 0018- 506X(76) 90001-5 wild female bonobos show high variability in the timing of ovu- Goodall J (1986) The chimpanzees of Gombe. Patterns of behavior. lation in relation to sexual swelling patterns. BMC Evol Biol Belknap, Harvard University, Cambridge, MA 16:1–17. https:// doi. org/ 10. 1186/ s12862- 016- 0691-3 Graham CE, Collins DC, Robinson H, Preedy JRK (1972) Urinary Drea CM (2015) D’scent of man: a comparative survey of primate levels of estrogens and pregnanediol and plasma levels of proges- chemosignaling in relation to sex. Horm Behav 68:117–133. terone during the menstrual cycle of the chimpanzee: relationship https:// doi. org/ 10. 1016/j. yhbeh. 2014. 08. 001 to the sexual swelling. Endocrinology 91:13–24. https:// doi. org/ Drea CM, Boulet M, Delbarco-Trillo J et  al (2013) The “secret” 10. 1210/ endo- 91-1- 13 in secretions: methodological considerations in deciphering Haselton MG, Gildersleeve K (2011) Can men detect ovulation? Curr primate olfactory communication. Am J Primatol 75:621–642. Dir Psychol Sci 20:87–92. https:// doi. org/ 10. 1177/ 09637 21411 https:// doi. org/ 10. 1002/ ajp. 22143 402668 Dubuc C, Muniz L, Heistermann M et al (2012) Do males time their Havliček J, Dvorakova R, Bartoš L, Flegr J (2006) Non-advertized mate-guarding effort with the fertile phase in order to secure does not mean concealed: body odour changes across the human fertilisation in Cayo Santiago rhesus macaques? Horm Behav menstrual cycle. Ethology 112:81–90. https:// doi. org/ 10. 1111/j. 61:696–705. https:// doi. org/ 10. 1016/j. yhbeh. 2012. 03. 0031439- 0310. 2006. 01125.x Dulac C, Torello AT (2003) Molecular detection of pheromone sig- Hepper PG, Wells DL (2010) Individually identifiable body odors nals in mammals: from genes to behaviour. Nat Rev Neurosci are produced by the gorilla and discriminated by humans. Chem 4:551–562. https:// doi. org/ 10. 1038/ nrn11 40 Senses 35:263–268. https:// doi. org/ 10. 1093/ chemse/ bjq015 Ellis RA, Montagna W (1962) The skin of primates. VI. The skin of Higham JP, Hebets EA (2013) An introduction to multimodal com- the gorilla (Gorilla gorilla). Am J Phys Anthropol 20:79–93. munication. Behav Ecol Sociobiol 67:1381–1388. https://doi. or g/ https:// doi. org/ 10. 1002/ ajpa. 13302 0021010. 1007/ s00265- 013- 1590-x Emery Thompson M (2005) Reproductive endocrinology of wild Hu J, Zhang Z, Shen W-J, Azhar S (2010) Cellular cholesterol delivery, female chimpanzees (Pan troglodytes schweinfurthii): meth- intracellular processing and utilization for biosynthesis of ster- odological considerations and the role of hormones in sex and oid hormones. Nutr Metab (Lond) 7:47. https:// doi. org/ 10. 1186/ conception. Am J Primatol 67:137–158. https://doi. or g/10. 1002/ 1743- 7075-7- 47 ajp. 20174 Jamil T, Ozinga WA, Kleyer M, ter Braak CJF (2013) Selecting traits Emery Thompson M, Wrangham RW (2008) Male mating interest that explain species–environment relationships: a generalized varies with female fecundity in Pan troglodytes schweinfurthii 1 3 Primates (2022) 63:365–376 375 linear mixed model approach. J Veg Sci 24:988–1000. https:// Nishida T (1968) The social group of wild chimpanzees in the doi. org/ 10. 1111/j. 1654- 1103. 2012. 12036.x Mahali Mountains. Primates 9:167–224. https:// doi. or g/ 10. Jänig S, Weiß BM, Widdig A (2018) Comparing the sniffing behavior 1007/ BF017 30971 of great apes. Am J Primatol 80:e22872. https:// doi. org/ 10. 1002/ Ostner J, Nunn CL, Schülke O (2008) Female reproductive syn- ajp. 22872 chrony predicts skewed paternity across primates. Behav Ecol Jänig S, Weiß BM, Birkemeyer C, Widdig A (2019) Comparative 19:1150–1158. https:// doi. org/ 10. 1093/ beheco/ arn093 chemical analysis of body odor in great apes. Am J Primatol. Partan S, Marler P (1999) Communication goes multimodal. Sci- https:// doi. org/ 10. 1002/ ajp. 22976 ence 283:72–73. https://doi. or g/10. 1126/ scien ce. 283. 5406. 1272 Kappeler PM, van Schaik C (2004) Sexual selection in primates— Pfefferle D, Kazem AJN, Brockhausen RR et al (2014a) Monkeys new and comparative perspectives. Cambridge University Press, spontaneously discriminate their unfamiliar paternal kin under Cambridge natural conditions using facial cues. Curr Biol 24:1806–1810. Klailova M, Lee PC (2014) Wild western lowland gorillas signal selec-https:// doi. org/ 10. 1016/j. cub. 2014. 06. 058 tively using odor. PLoS ONE 9:e99554. https:// doi. org/ 10. 1371/ Pfefferle D, Ruiz-Lambides AV, Widdig A (2014b) Female rhesus journ al. pone. 00995 54 macaques discriminate unfamiliar paternal sisters in playback Klinkova E, Hodges JK, Fuhrmann K et al (2005) Male dominance experiments—support for acoustic phenotype matching. Proc rank, female mate choice and male mating and reproductive suc- R Soc B 281:1–9 cess in captive chimpanzees. Int J Primatol 26:357–484. https:// Quinn GP, Keough MJ (2002) Experimental design and data analysis doi. org/ 10. 1007/ s10764- 005- 2929-6 for biologists, 1st edn. Cambridge University Press, Cambridge Krohn PL, Zuckerman S (1937) Water metabolism in relation to the R Core Team (2015) A language and environment for statistical com- menstrual cycle. J Physiol 88:369–387 puting. Version 3.2.3. R Foundation for Statistical Computing, Kücklich M, Möller M, Marcillo A et al (2017) Different methods Vienna. http:// www.R- proje ct. org/. Accessed 20 Jan 2021 for volatile sampling in mammals. PLoS ONE 12:e0183440. Rigaill L, Higham JP, Lee PC et al (2013) Multimodal sexual signal- https:// doi. org/ 10. 1371/ journ al. pone. 01834 40 ing and mating behavior in olive baboons (Papio anubis). Am J Kücklich M, Weiß BM, Birkemeyer C et al (2019) Chemical cues of Primatol 75:774–787. https:// doi. org/ 10. 1002/ ajp. 22154 female fertility states in a non-human primate. Sci Rep 9:1–12. Singh D, Bronstad PM (2001) Female body odour is a potential cue https:// doi. org/ 10. 1038/ s41598- 019- 50063-w to ovulation. Proc R Soc B 268:797–801. https:// doi. or g/ 10. Marty JS, Higham JP, Gadsby EL, Ross C (2009) Dominance, col-1098/ rspb. 2001. 1589 oration, and social and sexual behavior in male drills Mandril- Smith TE, Abbott DH (1998) Behavioral discrimination between cir- lus leucophaeus. Int J Primatol 30:807–823. https://doi. or g/10. cumgenital odor from peri-ovulatory dominant and anovulatory 1007/ s10764- 009- 9382-x female common marmosets (Callithrix jacchus). Am J Primatol Matsumoto-Oda A (1999) Mahale chimpanzees: grouping patterns 46:265–284. https:// doi. org/ 10. 1002/ (SICI) 1098- 2345(1998) and cycling females. Am J Primatol 47:197–207. https:// doi. 46:4% 3c265:: AID- AJP1% 3e3.0. CO;2-D org/10. 1002/ (SICI) 1098- 2345(1999) 47:3% 3c197:: AID- AJP2% Smith TD, Bhatnagar KP (2004) Microsmatic primates: reconsid- 3e3.0. CO;2-3 ering how and when size matters. Anat Rec B 279B:24–31. Matsumoto-Oda A, Oda R, Hayashi Y et  al (2003) Vaginal fatty https:// doi. org/ 10. 1002/ ar.b. 20026 acids produced by chimpanzees during menstrual cycles. Folia Snowdon CT, Ziegler TE, Schultz-Darken NJ, Ferris CF (2006) Primatol 74:75–79. https:// doi. org/ 10. 1159/ 00007 0000 Social odours, sexual arousal and pairbonding in primates. Matsumoto-Oda A, Kutsukake N, Hosaka K, Matsusaka T (2007) Philos Trans R Soc B 361:2079–2089. https:// doi. org/ 10. 1098/ Sniffing behaviors in Mahale chimpanzees. Primates 48:81–85. rstb. 2006. 1932 https:// doi. org/ 10. 1007/ s10329- 006- 0006-1 Stein SE (1999) An integrated method for spectrum extraction and Michael RP, Keverne EB, Bonsall RW (1971) Pheromones: isolation compound identification from gas chromatography/mass spec- of male sex attractants from a female primate. Science 172:964– trometry data. J Am Soc Mass Spectrom 10:770–781. https:// 966. https:// doi. org/ 10. 1126/ scien ce. 172. 3986. 964doi. org/ 10. 1016/ S1044- 0305(99) 00047-1 Michael RP, Bonsall RW, Warner P (1974) Human vaginal secre- Stoffel MA, Caspers BA, Forcada J et al (2015) Chemical finger - tions: volatile fatty acid content. Science 186:1217–1219. prints encode mother–offspring similarity, colony membership, https:// doi. org/ 10. 1016/ 0306- 4530(75) 90007-4 relatedness, and genetic quality in fur seals. PNAS 112:E5005– Michael RP, Bonsall RW, Kutner M (1975) Volatile fatty acids, “cop- E5012. https:// doi. org/ 10. 1073/ pnas. 15060 76112 ulins”, in human vaginal secretions. Psychoneuroendocrinology Stumpf RM, Boesch C (2006) The efficacy of female choice in chim- 1:153–163. https:// doi. org/ 10. 1016/ 0306- 4530(75) 90007-4 panzees of the Tai Forest, Cote d’Ivoire. Behav Ecol Sociobiol Montagna W, Yun JS (1963) The skin of primates. XV. The skin of 60:749–765. https:// doi. org/ 10. 1007/ s00265- 006- 0219-8 the chimpanzee (Pan satyrus). Am J Phys Anthropol 21:189– Townsend SW, Deschner T, Zuberbühler K (2011) Copulation calls 203. https:// doi. org/ 10. 1002/ ajpa. 13302 10211 in female chimpanzees (Pan troglodytes schweinfurthii) convey Muller MN (2002) Agonistic relations among Kanyawara chimpan- identity but do not accurately reflect fertility. Int J Primatol zees. In: Boesch C, Hohmann G, Marchant L (eds) Behavioural 32:914–923. https:// doi. org/ 10. 1007/ s10764- 011- 9510-2 diversity in chimpanzees and bonobos. Cambridge University Tutin CEG (1979) Mating patterns and reproductive strategies in Press, pp 112–124 a community of wild chimpanzees (Pan troglodytes schwein- Nadler RD, Graham CE, Gosselin RE, Collins DC (1985) Serum furthii). Behav Ecol Sociobiol 6:29–38. https://doi. or g/10. 1007/ levels of gonadotropins and gonadal steroids, including tes-BF002 93242 tosterone, during the menstrual cycle of the chimpanzee (Pan Vaglio S, Minicozzi P, Romoli R et al (2016) Sternal gland scent- troglodytes). Am J Primatol 9:273–284. https://doi. or g/10. 1002/ marking signals sex, age, rank, and group identity in captive ajp. 13500 90404 mandrills. Chem Senses 41:177–186. https:// doi. org/ 10. 1093/ Neumann C, Assahad G, Hammerschmidt K et al (2010) Loud calls chemse/ bjv077 in male crested macaques, Macaca nigra: a signal of dominance van den Berg RA, Hoefsloot HC, Westerhuis JA et al (2006) Cen- in a tolerant species. Anim Behav 79:187–193. https:// doi. org/ tering, scaling, and transformations: improving the biological 10. 1016/j. anbeh av. 2009. 10. 026 information content of metabolomics data. BMC Genom 7:142. https:// doi. org/ 10. 1186/ 1471- 2164-7- 142 1 3 376 Primates (2022) 63:365–376 Wallis J (1997) A survey of reproductive parameters in the free- Ziegler TE, Epple G, Snowdon CT et  al (1993) Detection of the ranging chimpanzees of Gombe National Park. J Reprod Fertil chemical signals of ovulation in the cotton-top tamarin, Sagui- 109:297–307 nus oedipus. Anim Behav 45:313–322. https://doi. or g/10. 1006/ Weiß BM, Marcillo A, Manser M et al (2018) A non-invasive method anbe. 1993. 1036 for sampling the body odour of mammals. Methods Ecol Evol Ziegler TE, Schultz-Darken NJ, Scott JJ et al (2005) Neuroendocrine 9:420–429. https:// doi. org/ 10. 1111/ 2041- 210X. 12888 response to female ovulatory odors depends upon social con- Wrangham RW (2002) The cost of sexual attraction: is there a trade- dition in male common marmosets, Callithrix jacchus. Horm off in female Pan between sex appeal and received coercion? In: Behav 47:56–64. https:// doi. org/ 10. 1016/j. yhbeh. 2004. 08. 009 Boesch C, Hohmann G, Marchant L (eds) Behavioural diver- sity in chimpanzees and bonobos. Cambridge University Press, Publisher's Note Springer Nature remains neutral with regard to Cambridge, pp 204–215 jurisdictional claims in published maps and institutional affiliations. Wyatt TD (2014) Pheromones and animal behavior: chemical signals and signatures. Cambridge University Press Young C, Majolo B, Heistermann M et al (2013) Male mating behav- iour in relation to female sexual swellings, socio-sexual behav- iour and hormonal changes in wild Barbary macaques. Horm Behav 63:32–39. https:// doi. org/ 10. 1016/j. yhbeh. 2012. 11. 004 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Primates Springer Journals

Towards an understanding of multimodal traits of female reproduction in chimpanzees

Download PDF
12 pages

Loading next page...
 
/lp/springer-journals/towards-an-understanding-of-multimodal-traits-of-female-reproduction-oeLmuh0v14

References (92)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2022
ISSN
0032-8332
eISSN
1610-7365
DOI
10.1007/s10329-022-00995-1
Publisher site
See Article on Publisher Site

Abstract

Although primates have long been regarded as microsmatic, recent studies indicate that olfaction is an important sensory mode of primate communication, e.g., in the context of reproduction. However, large gaps remain in our understanding of primate olfactory traits, especially in the great apes. Female chimpanzees (Pan troglodytes) possess an exaggerated sexual swelling, which is an imprecise signal of fertility that is thought to serve to confuse paternity. However, some high-ranking males that copulate most frequently on the days when females are most fertile seem to have more precise information on the timing of ovulation, which suggests the existence of an olfactory fertility trait. In order to examine, and provide evidence for, fertility-related chemical information in female chimpanzees, we used gas chromatography—mass spectrometry to analyze the chemical composition of female body odor collected across the menstrual cycle during various stages of sexual swelling (97 samples of six females). The chemical composition was significantly affected by swelling stage, and eight substances were detected that were strongly related to the latter. The existence of an additional, olfactory, fertility trait may help males to fine-tune their sexual behavior or allow females to strengthen concealment of the exact timing of ovulation, and needs to be further investigated in follow-up studies. The results of our study provide much-needed evidence for the existence of an olfactory cue related to reproduction in chimpanzees, and form a basis for future studies on the interplay between visual and olfactory information on female fertility. Keywords Olfactory communication · Olfaction · Gas chromatography—mass spectrometry · Chemical composition · Menstrual cycle · Anogenital swelling Introduction Primates utilize various sensory modalities, including vision, acoustics and olfaction, for social communication, Marlen Kücklich and Susann Jänig contributed equally to this study e.g., to convey information regarding kinship [olfaction in as first authors, and Brigitte M. Weiß and Anja Widdig contributed Lemur catta (Boulet et al. 2009); vision and acoustics in equally to this study as senior authors. Macaca mulatta (Pfefferle et al. 2014a, b)] or dominance rank [vision in Mandrillus leucophaeus (Marty et al. 2009); * Marlen Kücklich marlen_kuecklich@eva.mpg.de acoustics in Macaca nigra (Neumann et al. 2010); olfac- tion in Mandrillus sphinx (Vaglio et al. 2016)]. Sensory Behavioural Ecology Research Group, Faculty modalities involved in the transfer of specific information of Life Sciences, University of Leipzig, Talstraße 33, may act either alone or as multimodal traits, whereby differ - 04103 Leipzig, Germany ent modalities transmit either redundant or complementary Primate Behavioural Ecology Research Group, Department information (Partan and Marler 1999). The fertile phase of of Human Behavior, Ecology and Culture, Max Planck Institute for Evolutionary Anthropology, Deutscher Platz 6, female primates, e.g., is frequently indicated by a female’s 04103 Leipzig, Germany behavior, such as approaching males or reducing aggression, Mass Spectrometry Research Group, Institute of Analytical by visual traits such as sexual swellings, acoustic traits such Chemistry, University of Leipzig, Linnéstraße 3, as advertisement or copulation calls, as well as olfactory 04103 Leipzig, Germany Vol.:(0123456789) 1 3 366 Primates (2022) 63:365–376 traits such as scent-marking behavior, in various combi- these are not a precise signal of fertility (Emery Thompson nations [e.g., Cebuella pygmaea (Converse et  al. 1995); 2005). The phase of increased swelling lasts for approxi- Microcebus murinus (Buesching et al. 1998); Papio ursinus mately 10–12 days (Emery and Whitten 2003), and indicates (Clarke et al. 2009); Papio anubis (Rigaill et al. 2013)]. approaching ovulation, which occurs during the period of Whereas visual and acoustic traits of female fertility maximal swelling (Deschner et al. 2004). However, maximal have been relatively well studied across primates, studies swelling persists for up to 4 days after ovulation (Desch- on olfactory traits of fertility are comparatively scarce (Drea ner et al. 2004), and is thus only an imprecise signal of the 2015). However, over the years, it has become evident that latter. This imprecise information on the actual timing of olfaction may also play an important role in the transmission ovulation allows females to confuse paternity and reduce of social information in primates (e.g., Smith and Bhatnagar the risk of infanticide by attracting a high number of mating 2004; Drea et al. 2013). Studies in several primate species partners (Wrangham 2002; Kappeler and van Schaik 2004; indicate that females convey olfactory information regarding Stumpf and Boesch 2006). Males compete intensely with their reproductive state through whole body odor, vaginal each other over their access to females (Goodall 1986; Mul- secretions, urine and feces (Drea et al. 2015). Furthermore, ler 2002), which restricts their mating opportunities [pri- olfactory inspection of the female anogenital region by males ority of access model (Altmann 1962)]. The potential for increases during the fertile period [P. ursinus (Clarke et al. males to monopolize females is further reduced since several 2009); P. anubis (Rigaill et al. 2013)], and males react with females of a group will show exaggerated sexual swellings increased exploratory behavior and sexual activity [Macaca synchronously (Goodall 1986; Ostner et al. 2008). As males mulatta (Michael et al. 1971); Galago crassicaudatus (Clark should focus their mating efforts on the days when females 1982); Saguinus oedipus (Ziegler et al. 1993); Callithrix are most fertile in order to mate with as many fertile females jacchus (Smith and Abbott 1998); Macaca arctoides (Cerda- as possible, they would benefit from using precise indicators Molina et al. 2006b)] as well as a rise in testosterone levels of female fertility, if available. Indeed, male chimpanzees and activation of brain areas linked to sexual behavior [C. alter their sexual behavior when female anogenital swellings jacchus (Ziegler et al. 2005; Snowdon et al. 2006); M. arc- increase in size and throughout the period of maximal swell- toides (Cerda-Molina et al. 2006a)] when presented with ing (Klinkova et al. 2005), and mating seems to be correlated vaginal secretions from fertile females. with the period of peak fertility of females (Deschner et al. Specific substances related to the menstrual cycle of 2004; Emery Thompson 2005). Alpha males, in particular, females have recently been identified, and males were were found to copulate more frequently during the fertile observed to distinguish between female odor from differ - period compared to the other days of the maximal swell- ent phases of the cycle in C. jacchus [an aromatic hydro- ing phase (Deschner et al. 2004). These observations sug- carbon, a heterocyclic compound, an ester and an alcohol gest that high-ranking males may benefit from additional (Kücklich et al. 2019)]. Certain substances were also found information, e.g., a multimodal cue of fertility. Rigaill to be fertility related and to sexually stimulate male Old et al. (2013) hypothesized that the visual signal in P. anu- World monkeys [fatty acids in M. mulatta (Michael et al. bis could provide a first sign of fertility to all males within 1971)], although the results could not be reproduced in a visual range, whereas high-ranking males that can approach subsequent study [M. mulatta (Goldfoot et al. 1976)]. These females closely could potentially gather more precise infor- types of fertility indicators could constitute “signals”, i.e., mation from an olfactory inspection of females with sexual traits that have specifically evolved to provide information swellings. Also, females could benefit from a more precise on fertility, or “cues”, which constitute traits that have not indicator of ovulation that is only sensed by selected, com- evolved to deliver fertility-related information to conspecif- petitive males, to obtain the best genes and protection for ics, e.g., being by-products of changes in hormone levels their infants (Kappeler and van Schaik 2004). (Wyatt 2014). Observations of increased copulation rates at the time Chimpanzees (Pan troglodytes) are prime candidates for of highest fertility in females suggest that male chim- the study of multimodal traits of fertility, as female chimpan- panzees may indeed rely on olfactory traits in a sexual zees have several traits that may be indicative of their repro- context (Emery Thompson and Wrangham 2008). Certain ductive state during their menstrual cycle, which is around fatty acids that are thought to be associated with female 36 days long (Tutin 1979; Deschner et al. 2003; Emery and fertility in rhesus macaques (Macaca mulatta) (Michael Whitten 2003). They make copulation calls, although these et al. 1971) were also found in the vaginal secretions of seem not to vary with fertility (Townsend et al. 2011), and female chimpanzees, but it remains unclear if these vary show proceptive as well as mating resistance behavior that during the menstrual cycle (Fox 1982; Matsumoto-Oda varies between their fertile and non-fertile periods (Stumpf et al. 2003). Notably, fertility-related changes in the odor and Boesch 2006). In addition, they have exaggerated sexual of female chimpanzees could occur in genital odor and/ swellings, which vary in size across their cycle, although or in the scent of the whole body (hereafter “body odor”). 1 3 Primates (2022) 63:365–376 367 Several other individual attributes have been detected in In this study, we related the chemical profiles of body the body odor of chimpanzees and other great apes [e.g., odor samples of regularly cycling female chimpanzees to indicating individual identity (Hepper and Wells 2010); their stage of sexual swelling. We investigated whether arousal level (Klailova and Lee 2014); species identity and chemical composition varies synchronously with the vis- age (Jänig et al. 2019)]. Chimpanzees sniff other individu- ual signal, and predicted a change in the former across the als (i.e., genitalia, but also other body parts), and males sexual cycle. If this prediction were true, it would suggest specifically have been observed to sniff their conspecifics that anogenital swelling and body odor potentially comprise more often than females do (Matsumoto-Oda et al. 2007; multimodal traits of fertility. For the purposes of this study, Jänig et  al. 2018). Accordingly, we expect information we used data collected as part of a larger study on various about reproductive state to be provided by the body odor traits of great ape body odor [species identity, age and sex of female chimpanzees in general, and not only by their (Jänig et al. 2019)]. However, at the time of data collection, genital odor, which may constitute a multimodal trait of we were not able to obtain hormone samples from females fertility in combination with the visual signal of fertility, in a reliable manner that would have enabled us to deter- i.e., the sexual swelling. mine the exact timing of ovulation. Hence, given the lack Given the lack of studies demonstrating the existence of of hormonal data, we cannot show whether an olfactory cue olfactory fertility traits in chimpanzees, our overarching aim exists that is a more precise indicator of fertility than the was to provide evidence, based on chemical analyses, that visual signal. Our results nonetheless provide information olfactory information exists that is related to reproduction. on whether olfaction is related to reproduction at all, and Chimpanzees live in large, multi-male/multi-female groups, thus provide a foundation for future studies that include an which split up regularly into smaller parties of variable com- investigation of hormones in this context. position (Nishida 1968; Goodall 1986). Males preferably join parties which include females with maximal sexual swellings (Matsumoto-Oda 1999). If an olfactory fertility cue exists that is even more precise than the visual signal, Materials and methods this could potentially help males to specifically join parties with females close to ovulation. Cycle-dependent olfac- Subjects tory traits could be subtle changes that arise as a result of physiological changes, such as the breakdown of hormones We conducted our study on captive chimpanzees (Pan trog- involved in the menstrual cycle, e.g., luteinizing hormone, lodytes verus) at the Wolfgang Köhler Primate Research follicle-stimulating hormone, progesterone and estradiol Centre (WKPRC) at Leipzig Zoo, Germany, in 2013. We (Nadler et al. 1985). Sexual swelling appears to be induced analyzed samples from six adult females (aged between 13 by a rise in the levels of estrogens, and detumescence is and 20 years) from two different social groups (for details on correlated with a rise in progesterone after ovulation (Gra- housing and group composition, see supplementary Online ham et al. 1972). Hence, olfactory changes could occur as Resource), which had a regular menstrual cycle, i.e., did not by-products of these hormonal changes that accordingly fol- receive contraceptives (for at least 12 months prior to the low the timing of sexual swellings, and thus act as similarly study) and were not lactating. imprecise fertility traits as the visual signal. Ovulation can be assumed to occur 1 day before the increase in the level of progesterone (Deschner et al. 2004), and is related to pre- Sexual swelling stages ovulatory peaks of luteinizing hormone and follicle-stim- ulating hormone which are preceded by an estradiol peak The sexual swelling stage of each female was routinely mid-cycle (Nadler et al. 1985). Thus, a hormone-related recorded by the zookeepers and categorized as one of the olfactory change could also occur at the time of ovulation. following: flat, increasing in size, maximal tumescence, If this were true, olfactory information on fertility would decreasing in size. Mean cycle length was 35.6 ± 9.1 days be more precise than that provided by the visual signal. On and thus within the ranges previously published for chim- the other hand, an imprecise olfactory cue could strengthen panzees [captive (Nadler et al. 1985); wild (Tutin 1979; Wal- concealment of the actual timing of ovulation, and thus sup- lis 1997; Emery Thompson 2005)]. On average, the observed port the visual signal and allow females to confuse paternity. females were detumescent for 16.7 ± 5.4 days, the swellings A more precise olfactory cue, however, would allow high- increased in size for 4.3 ± 5.5 days, remained at maximal ranking males in particular to monitor females with maximal tumescence for 12.3 ± 3.4 days, and decreased in size for swelling to pinpoint the exact time of ovulation. Either way, 2.4 ± 1.5 days, similar to data reported for wild chimpanzees the composition of female body odor would be expected, at (Deschner et al. 2003). the very least, to vary between swelling stages. 1 3 368 Primates (2022) 63:365–376 et al. 2016). The extract was concentrated stepwise to a vol- Sample collection and preparation ume of circa 60 µL, and 4 µL of this solution was injected into the GC (HP6890 Series GC System, Agilent, Wald- The examined individuals are well trained in participat- ing in behavioral experiments by the WKPRC. Samples of bronn, Germany, with a HP5973 MSD Mass Selective Detector in electron-impact ionization mode at 70 eV) using skin odor were collected in the morning prior to behavio- ral experiments. For the purpose of our study, individuals splitless injection, for 2 min. The GC was equipped with a J&W Fisher DB35-MS column (30 m length, 0.25 mm inner were trained by positive reinforcement (using food items) to come to the grid of the cage to participate voluntarily in diameter, 0.25 µm film; Agilent), with the inlet tempera - ture set to 250 °C. Helium was used as the carrier gas at a the sampling by presenting a self-selected body part. Clean Lilibe cotton swabs (60% cotton wool, 25% polyester micro- flow rate of 1.7 mL/min. The temperature program ran for 40.5 min, starting at 35 °C for 2 min, followed by a heating fiber, 15% polyester), heated for 30 min at 130 °C before use (Birkemeyer et al. 2016), were held with sterilized metal step of 10 °C/min until 320 °C, which was held for 10 min. The solvent delay was set to 7 min. The ion source operated tweezers and rubbed repeatedly over the skin and fur for approximately 20 s (similar to Célérier et al. 2010; Stoffel at 250 °C and the scan range was set to m/z 50–550. et al. 2015). The cotton swabs were immediately placed in pre-cleaned (washed with methanol and diethyl-ether) 4-mL Data processing glass vials (Rotilabo) and stored at − 80 °C until gas chro- matography—mass spectrometry (GC–MS) analysis. From the GC–MS data we identified peaks, their retention times (RTs) and areas (intensity) using AMDIS v. 2.65 Given that the individual chimpanzees had full control over which body part they presented, the samples included (Stein 1999). To determine substances that were repeatedly detected in our samples, we grouped consecutive RTs to one odor from the arms, belly, back and legs, but not the genitals, as these were not readily presented by the animals. Hence, RT range (see also Weiß et al. 2018) using a self-written script performed in R version 3.2.3 (R Core Team 2015). we sampled body parts that have the same types of apocrine and eccrine sweat glands (Ellis and Montagna 1962; Mon- We verified by manual inspection that mass spectral pat- terns of the peaks within a given RT range were consistent, tagna and Yun 1963) representing general body odor rather than genital odor. Body and genital odor should be similarly i.e., each RT range was assumed to reflect one substance, resulting in 152 RT ranges. In the following, we thus use the affected by hormonal changes across the menstrual cycle, as previously shown in humans (Michael et al. 1974, 1975; term “substance” rather than “RT range.” At that stage of the study, the substances had not yet been chemically identified Gildersleeve et al. 2012). Furthermore, behavioral observa- tions of both chimpanzee groups at the WKPRC revealed by using a library search, and were described only by their RT and specific patterns of signals with certain mass-to- that when males sniff females they focus on the genitals in just over half of cases and other body parts in the remaining charge-ratios (m/z). We used a non-targeted approach to find substances that cases (Jänig et al., in review). We controlled for the different body parts in our statistical analysis. Samples were not col- might be related to female fertility. We excluded 15 sub- stances from the statistical analysis which had the same or lected from the mouth, hands or feet to reduce contamination with odor from food or the environment. Furthermore, diet a higher abundance in the control than in the animal sam- ples (i.e., data on 137 of the 152 substances remained for was likely to have only a minor impact on body odor vari- ance within and between individuals, since animals were fed analysis). For all the substances which were found to be potentially related to female fertility according to the statisti- daily with similar food. In total, 97 samples [16.2 ± 2.3 samples per individual cal analysis, mass spectral comparisons using the National Institute of Standards and Technologies (NIST) Mass Spec- female covering from five to eight (mean 7.2) menstrual cycles each)] were collected from the animals (for details tral Library (NIST 14 software; NIST, Gaithersburg, MD) were conducted for their tentative identification. We report see supplementary Online Resource Table 1). In addition, 42 control samples (pure cotton swabs not rubbed over the the best library hit per substance when (1) the top library hit was consistent over samples, (2) the match had a prob- skin or fur but otherwise handled like the animal samples; one per sampling day) were collected to identify chemical ability > 80 (NIST 14 value), and (3) the tentatively iden- tified compound could be reasonably expected to elute at substances that did not originate from the chimpanzees. the given RT. Alternatively, where no library hit allowed tentative identification, we propose a structural classification Chemical analysis based on interpretation of the spectrum. The mass spectra of the substances discussed in this article are provided in We extracted the chemical substances for GC–MS analysis by adding 1.2 mL of n-hexane (Sigma Aldrich, Steinheim, supplementary Online Resource Fig. 2. Germany) to the cotton swab in a glass vial (Birkemeyer 1 3 Primates (2022) 63:365–376 369 information, see Weiß et al. (2018)]. Additionally, swelling Statistical analysis stage was fitted as a random slope within ID of the female to achieve more accurate estimates of the test predictor. First, we conducted an analysis of similarities (ANOSIM) to test whether chemical profiles of females at the same The check for normal distribution and homogeneity of the residuals by inspection of a qq-plot suggested no violation swelling stage were more similar to each other than profiles of females at different swelling stages. Following common of assumptions. Plotting the residuals against fitted values revealed a slight bottom effect, but model stability indicated practice (e.g., Stoffel et al. 2015), the ANOSIM was based on Bray–Curtis indices calculated from the log-transformed, no influential cases. We determined variance inflation factors (VIF) to check for potential collinearity (Quinn and Keough standardized intensities of substances (intensity of substance divided by the summed intensity of all 137 substances × 2002; Field 2005) calculated with the function vif of the package car (Fox and Weisberg 2011) and found no indica- 100) for each combination of sample dyads. To control for repeated measurements per female, we used a customized tion of collinearity (VIF = 1.01). max We compared the full model to a null model excluding the ANOSIM (R script written by LK) that computed P-values by permuting swelling stages within individual females only. random slope within substance by using a likelihood ratio test (LRT) (Barr et al. 2013) to determine the significance The α-level was set to 0.05 for all statistical analyses. Second, we implemented two two-tailed generalized lin- of the full model (Forstmeier and Schielzeth 2011). For the significant full model, the slope estimates of all substances ear mixed models in R v. 3.2.3 (R Core Team 2015) using the package lme4 v. 1.1.11 (Bates et al. 2015) to determine if were extracted from the model results. To determine the sub- stances whose levels changed most with swelling stage, we specific substances are associated with a particular swelling stage. Chemical datasets of animals may present several ana- compared slope estimates to the average slope estimate and focused on those substances whose absolute slope estimate lytical challenges, such as (1) substances of potential biolog- ical relevance may have very high, or very low intensities; was higher than the average absolute slope estimate + 2SDs (Weiß et al. 2018). (2) the total intensities of samples can change due to envi- ronmental factors; and (3) the intensity of certain substances (e.g., those considered contaminants, such as plasticizers) can vary strongly, and thus affect the relative composition Results of odor profiles (van den Berg et al. 2006). To overcome these problems, we fitted the two models using two differ - Chemical profiles of females at the same swelling stage were significantly more similar than profiles of females at ent responses: standardized (intensity of substance divided by the summed intensity of all 137 substances per sample different swelling stages (ANOSIM, R = 0.07, P = 0.026), although the effect was rather small when whole profiles × 100) and transformed [arcsine and log(x + 0.01)] peak areas; standardized peak areas centered (to a mean of 0) and were considered. The chemical composition of substances with log-trans- scaled (to a SD of 1). In both cases, standardization was used to correct for changing total sample intensities. In the first formed abundances was affected by swelling stage (LRT, χ = 10.77, P = 0.013). Thus, the relation of these affected approach, transformation was applied to achieve a normal distribution and to reduce the relative impact of large sub- substances to the rest of the respective odor profile changed depending on the swelling stage. This effect was most pro- stances, while differences in the relative abundances of the various substances were maintained. In the second approach, nounced for four substances (i.e., their absolute slope esti- mate was higher than the absolute slope estimate + 2SDs). centering and scaling was applied to adjust all the substances to the same size and to give all of them equal weight in the These substances were all tentatively identified as steroids (Table 1), and were least abundant in samples collected dur- analyses (van den Berg et al. 2006). The responses were vectorized from a multivariate data matrix (Jamil et al. 2013) ing the maximal swelling as well as the flat phase, and most abundant during the increase and decrease phases of swell- of samples (n = 97) and substances (n = 137). Swelling stage (dummy coded and centered) was fitted as fixed effect test ing (Fig. 1). For substances with centered and scaled abundances, predictor, and age (z-transformed) of the females was fitted as control predictor. Sample number and substance identity chemical composition was also affected by swelling stage (LRT, χ = 19.28, P < 0.001). In these cases, the substance- (ID) were included as random effects to prevent pseudo- replication as well as heteroscedastic variance due to the specific abundances changed between the profiles of dif- ferent swelling stages. The swelling stage effect was most vectorized data matrix (Jamil et al. 2013). Other random effects were the day of observation, the sampled body part, pronounced for four of the substances (i.e., their absolute slope estimate was higher than the absolute slope esti- as well as ID and group of the female. We included the ran- dom slope of the fixed effects test predictor (swelling stage) mate + 2SDs), which were different from the most affected substances of the log-transformed model. One of these within substance as the actual test predictor [for detailed 1 3 370 Primates (2022) 63:365–376 Table 1 Substances most affected by swelling stage with log-trans- classification, substance class and the largest differences between the formed abundances, including retention times (RTs), tentative identifi- absolute slope estimates of swelling stages (average absolute slope cation [National Institute of Standards and Technologies (NIST) Mass estimate ± SD, 0.096 ± 0.08) Spectral Library; NIST  14] with probability (best hit) or structural RTs Best library hit/structural classification Probability Substance class Slope estimates 42.40 Unknown steroid (cholestadiene-like) – Steroid 0.59 43.49 Cholesta-2,4-diene 94 Steroid 0.32 44.39 Cholesta-3,5-diene 94 Steroid 0.38 48.46 Cholesta-5-en-3-ol (3.beta.)-, acetate 98 Steroid 0.38 For more details, see Fig. 1 and supplementary Online Resource Table 2; for mass spectra, see supplementary Online Resource Fig. 2 Fig. 1 a–d Log-transformed intensity (%) of the four substances most influenced by swelling stage (see also Table 1). Boxplots show medians and first and third quartiles. Retention times (RTs) and standardized peak area means ± SD are as follows: a RT 42.40, 1.97 ± 2.46; b RT 43.49, 0.48 ± 0.43; c RT 44.39, 13.71 ± 8.45; d RT 48.46, 0.55 ± 0.56 substances was tentatively identified by the library search Discussion as a wax ester (Fig. 2a), whereas two substances had spec- tra with m/z patterns typical of long-chain alkylic struc- The results of the present study support the presumption tures (Fig. 2b, d), and one substance was thought to feature that olfaction is related to reproduction in chimpanzees, a phenyl substructure (Fig. 2c) based on its structural clas- as the chemical profiles of the female chimpanzees var - sification (see Table  2). Three of the substances were most ied significantly with sexual swelling stage. Initial evi- abundant in samples collected during maximal swelling dence was provided by the slightly greater similarities of compared to the other swelling stages, while one of the the whole chemical profiles of samples taken during the substances with an alkylic (sub)structure was most abun- same swelling stage compared to those taken at different dant during the flat phase and least abundant in samples swelling stages. These findings were corroborated by a collected during maximal swelling (see Fig. 2). 1 3 Primates (2022) 63:365–376 371 Fig. 2 a–d Centered and scaled intensity (%) of the four substances most influenced by swelling stage (see also Table 2). Boxplots show medi- ans and first and third quartiles. RTs and standardized peak area means ± SD are as follows: a RT 32.26, 0.26 ± 1.05; b RT 34.65, 0.12 ± 0.25; c RT 40.36, 0.03 ± 0.02; d RT 48.80, 0.35 ± 0.31 Table 2 Substances most affected by swelling stages with cen- tion, substance class and the largest differences between the abso- tered and scaled abundances, including RTs, tentative identification lute slope estimates of swelling stages (average absolute slope esti- (NIST  14 library) with probability (best hit) or structural classifica- mate ± SD, 0.273 ± 0.20) RT Best library hit/structural classification Probability Substance class Slope estimates 32.26 Dodecanoic acid, isooctyl ester 91 Ester 0.87 34.65 Unknown [long-chain alkylic (sub)structure] – Unknown 0.91 40.36 Unknown phenol – Phenol 0.67 48.80 Unknown [long-chain alkylic (sub)structure] – Unknown 0.86 For more details, see Fig. 2 as well as supplementary Online Resource Table 2; for mass spectra, see supplementary Online Resource Fig. 2. For abbreviations, see Table 1 significant change in chemical composition across swell- log transformed or centered and scaled when used in the ing stages that was most pronounced with respect to eight models. In the case of log-transformed peak areas, affected substances. Thus, the results reveal that olfactory changes substances differ in contrast to the remaining chemical pro- exist that mirror changes in fertility at least as closely as file. Hence, information on the state of fertility seems to be the visual signal of sexual swelling. However, our meth- provided by the status quo of the chemical composition, and odological approach did not allow us to address whether comparisons over time are thus not necessary. When peak olfactory cues provide more precise information which areas are centered and scaled, the analysis is focused on dif- could allow males to pinpoint the time of ovulation more ferences between, instead of within, the samples (van den precisely. Hence, further studies that include hormonal Berg et al. 2006). Thus, particular substances that are found measurements are needed to unravel the temporal dynam- to be affected when peak areas are centered and scaled dif- ics of olfactory cues in the context of reproduction. fer in their levels between samples (i.e., over time or swell- The effect of swelling stage on chemical composition was ing stages). In this scenario, it seems to be the variation tested with two different approaches. Peak areas were either (or differences) over the course of the menstrual cycle that 1 3 372 Primates (2022) 63:365–376 provides information on fertility stage. A male would thus all the females and swelling stages. All of the substances need to regularly check a female’s odor over time to unravel were found for all six females as well as at all of the swell- this information. As expected, the two approaches revealed ing stages, and most of them were present in at least 80% different chemical substances that were most affected by of the samples for at least two swelling stages. However, swelling stage. two of the substances indicated by the centered and scaled Four chemical substances were found to be affected by model (an alkylic compound with an RT of 34.65 min and an swelling stage when abundances were log-transformed. All unknown phenol with an RT of 40.36 min) were only present of them were cholesterol derived, and were more abundant in a  maximum of 43% of samples per swelling stage (maxi- during the increasing and decreasing swelling stages com- mal swelling and increasing/decreasing swelling stages). If pared to the maximal swelling and flat stage. Cholesterol- substances are not always present in the chemical profile of derived substances are ubiquitous in the chemical profiles female odor, they are not very suitable for use by males as of mammals (Charpentier et al. 2012) and are precursors reliable indicators of female reproductive state. Future stud- for smaller molecules produced by bacterial degradation ies that include a higher number of samples of female odor (Ezenwa and Williams 2014). Specifically, cholesterol is the will help researchers to examine this subject in more detail. precursor for steroid hormones such as estradiol and proges- While our results indicate that body odor does indeed terone (Hu et al. 2010). Swelling of the anogenital skin is change over a female’s reproductive cycle, our methodology estrogen dependent, and detumescence occurs after a rise in did not allow us to assess whether olfactory changes are as progesterone, which inhibits the effects of estrogen (Gillman precise as the visual signal, or potentially even more pre- and Stein 1941; Graham et al. 1972; Emery and Whitten cise. In fact, in the context of an evolutionary arms race, we 2003). Tumescence of the sexual swelling occurs due to the would expect males to strive to pinpoint the exact timing of intracellular accumulation of water in the anogenital tissue, ovulation by using a fertility trait, and thus that they would which is lost during detumescence (Krohn and Zuckerman benefit from an olfactory cue that is more precise than the 1937; Clarke 1940). The consistent pattern of the four most visual cue of sexual swelling. Females, on the other hand, affected substances, i.e., that their levels were highest during would profit from generally confusing paternity by provid- tumescence and detumescence, leads to the assumption that ing the rather imprecise visual signal of fertility which could these chemical changes arise due to physiological changes be strengthened by an additional, imprecise olfactory cue. that occur during swelling. This chemical information being However, high-ranking male chimpanzees in particular have constistent with the visual signal of increasing or decreas- been observed to start copulating only in the last 3–4 days ing swelling, could enable males to recognize the beginning of maximal swelling (Deschner et al. 2004). As the duration and end of the female fertile period, allowing them to focus of maximal swelling is variable and its termination cannot their monopolization and mating efforts on females which be deduced according to the point at which it starts, such are most fertile. observations suggest that high-ranking male chimpanzees Moreover, four substances with centered and scaled abun- in particular, which are able to approach females to within a dances were found to be related to swelling stage. Similar very close distance, may rely on additional and more precise types of substances, such as long-chain fatty acid esters, sub- indicators of ovulation, i.e., presumably olfactory traits. In stances with alkylic structures, and phenols, were previously humans, odor has been proposed as a more precise indica- detected, e.g., in skin emanations [humans (Bernier et al. tor for (approaching) ovulation than visual signs (Singh and 2000)] and in genital secretions [L. catta (with the excep- Bronstad 2001; Havliček et al. 2006; Haselton and Gilder- tion of phenols) Boulet et al. 2009]. These four substances sleeve 2011; Gildersleeve et al. 2012). Relying on more than were related to the maximal swelling stage, as three of them one of the senses for the inspection of a female is likely to be were most abundant and the other one was least abundant advantageous for male chimpanzees too. The visual signal during this phase compared to the other swelling stages. could give initial information about the state of fertility of a Overall, our results suggest that the chemical composition of female at short/medium distance, which may then especially female chimpanzee odor changes over the menstrual cycle in attract high-ranking males with priority of access, allowing accordance with sexual swelling stage. Thus, body odor can them to undertake an olfactory inspection at close proximity be considered at least as a proxy of fertility similar to swell- to gain additional information (Matsumoto-Oda et al. 2007). ing stage, which has been well established as an imprecise Such olfactory information could potentially be non-redun- fertility signal. dant to the visual information (see also Higham and Hebets An olfactory trait related to reproduction is expected 2013), but our study design did not allow us to determine to be present consistently within the chemical profile of whether the olfactory cue is redundant or not. female odor to allow males to infer female fertility from it. Thus, an  important next step building on our  current Thus, we checked whether the most affected substances as results, that show that chemical composition changes over shown by the two models were continuously present across the menstrual cycle in accordance with stages of sexual 1 3 Primates (2022) 63:365–376 373 from the University of Leipzig was awarded to SJ (DFPL R00061) and swelling, would be  investigating whether chemical profiles MK (DFPL R00017). indeed provide more precise information than sexual swell- ings. To investigate this in more detail, analyses of cycle- Data availability The dataset generated and analyzed during the current related hormones are needed to determine the exact timing study is available on reasonable request. of ovulation, which could not be accomplished at the time that this study was conducted. This could be undertaken by Declarations using non-invasive hormonal measurements, as applied in previous studies to assess the timing of ovulation in non- Conflict of interest The authors declare that they have no conflict of interest. human primates (Deschner et al. 2004; Engelhardt et al. 2006; Dubuc et al. 2012; Young et al. 2013; Douglas et al. Ethical approval This study was carried out in accordance with the 2016). legal requirements of Germany and all national and institutional guide- Moreover, recently adapted sampling techniques now lines for the care and use of animals, and was approved by the ethics commission of the Department of Psychology of the Max Planck Insti- allow a wider range of chemical substances to be sampled tute for Evolutionary Anthropology and Leipzig Zoo. (Kücklich et al. 2017). The traditionally used cotton swab, which was also used in this study, primarily captures semi- Open Access This article is licensed under a Creative Commons Attri- to non-volatile substances (Birkemeyer et al. 2016). These bution 4.0 International License, which permits use, sharing, adapta- types of larger molecules are then degraded by skin bacte- tion, distribution and reproduction in any medium or format, as long ria into smaller, volatile molecules that constitute the odor as you give appropriate credit to the original author(s) and the source, profile of an individual (Ezenwa and Williams 2014) that provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are can be perceived by the main olfactory system (Dulac and included in the article's Creative Commons licence, unless indicated Torello 2003). Future studies will certainly benefit from otherwise in a credit line to the material. If material is not included in these recent methodological developments for the direct the article's Creative Commons licence and your intended use is not sampling of volatile substances, and for the correlation of permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a their abundances to individual traits of animals. copy of this licence, visit http://cr eativ ecommons. or g/licen ses/ b y/4.0/ . In conclusion, this study adds much-needed information on the role of olfaction in great apes by showing that the chemical composition of body odor is associated with a vis- ual signal of fertility in chimpanzees. Our results indicate References that olfactory cues related to reproduction exist in a great ape species with visual fertility advertisement, and provide a Altmann SA (1962) A field study of the sociobiology of rhesus mon - valuable basis for further research on more specific questions keys, Macaca mulatta. Ann NY Acad Sci 102:338–435. https:// on reproduction in chimpanzees and olfaction in great apes doi. org/ 10. 1111/j. 1749- 6632. 1962. tb136 50.x Barr DJ, Levy R, Scheepers C, Tily HJ (2013) Random effects in general. In fission–fusion societies, where group compo- structure for confirmatory hypothesis testing: keep it maximal. sition varies and visual information is short lived, olfaction J Mem Lang 68:255–278. https:// doi. org/ 10. 1016/j. jml. 2012. may allow males to fine-tune their sociosexual behavior and 11. 001 to optimize their reproductive investment. Bates D, Maechler M, Bolker B, Walker S (2015) lme4: linear mixed-effects models using Eigen and S4, R package. Version 1.0–8. http://CRAN .R-pr oject. or g/pac kag e=lme4 . Accessed 20 Supplementary Information The online version contains supplemen- Jan 2021 tary material available at https://doi. or g/10. 1007/ s10329- 022- 00995-1 . Bernier UR, Kline DL, Barnard DR et al (2000) Analysis of human skin emanations by gas chromatography/mass spectrometry. 2. Acknowledgements We thank Josep Call and the curators of the Leip- Identification of volatile compounds that are candidate attract- zig Zoo for permission to sample the chimpanzees at the WKPRC, and ants for the yellow fever mosquito (Aedes aegypti). Anal Chem are most grateful for the help of the keepers at Leipzig Zoo and the staff 72:747–756. https:// doi. org/ 10. 1021/ ac990 963k of the WKPRC. Furthermore, we also acknowledge the help of Susan Birkemeyer C, Thomsen R, Jänig S et al (2016) The sampling of Billig and Andreas Kiontke for technical assistance with the chemical body odour of primates: cotton swabs sample semi-volatiles analysis. We thank Ruth Thomsen for logistic support during sample rather than volatiles. Chem Sens 41:525–535. https:// doi. org/ collection, and Roman Wittig and the anonymous reviewers for com- 10. 1093/ chemse/ bjw056 ments on earlier versions of the manuscript. Finally, we thank the Max Boulet M, Charpentier M, Drea C (2009) Decoding an olfactory mech- Planck Institute for Evolutionary Anthropology and the University of anism of kin recognition and inbreeding avoidance in a primate. Leipzig for their logistic support. BMC Evol Biol 9:281. https:// doi. org/ 10. 1186/ 1471- 2148-9- 281 Buesching CD, Heistermann M, Hodges JK, Zimmermann E (1998) Funding Open Access funding enabled and organized by Projekt Multimodal oestrus advertisement in a small nocturnal prosim- DEAL. This work was supported by the German Research Foundation ian, Microcebus murinus. Folia Primatol 69:295–308. https:// within the Emmy-Noether program (grant no. WI 1808/3-1 awarded to doi. org/ 10. 1159/ 00005 2718 AW) and by the University of Leipzig. In addition, graduate funding 1 3 374 Primates (2022) 63:365–376 Célérier A, Huchard E, Alvergne A et  al (2010) Detective mice of Kanyawara, Kibale National Park. Int J Primatol 29:885–905. assess relatedness in baboons using olfactory cues. J Exp Biol https:// doi. org/ 10. 1007/ s10764- 008- 9286-1 213:1399–1405. https:// doi. org/ 10. 1242/ jeb. 038349 Emery MA, Whitten PL (2003) Size of sexual swellings reflects Cerda-Molina AL, Hernandez-Lopez L, Chavira R et al (2006a) Endo- ovarian function in chimpanzees (Pan troglodytes). Behav crine changes in male stumptailed macaques (Macaca arctoides) Ecol Sociobiol 54:340–351. https:// doi. or g/ 10. 1007/ as a response to odor stimulation with vaginal secretions. Horm s00265- 003- 0648-6 Behav 49:81–87. https:// doi. org/ 10. 1016/j. yhbeh. 2005. 04. 014 Engelhardt A, Heistermann M, Hodges JK et al (2006) Determi- Cerda-Molina AL, Hernandez-Lopez L, Rojas-Maya S et al (2006b) nants of male reproductive success in wild long-tailed macaques Male-induced sociosexual behavior by vaginal secretions in (Macaca fascicularis)—male monopolisation, female mate Macaca arctoides. Int J Primatol 27:791–807. https:// doi. org/ choice or post-copulatory mechanisms? Behav Ecol Sociobiol 10. 1007/ s10764- 006- 9045-0 59:740–752. https:// doi. org/ 10. 1007/ s00265- 005- 0104-x Charpentier MJE, Barthes N, Proffit M et al (2012) Critical thinking Ezenwa VO, Williams AE (2014) Microbes and animal olfactory in the chemical ecology of mammalian communication: road- communication: where do we go from here? BioEssays 36:847– map for future studies. Funct Ecol 26:769–774. https://doi. or g/ 854. https:// doi. org/ 10. 1002/ bies. 20140 0016 10. 1111/j. 1365- 2435. 2012. 01998.x Field A (2005) Discovering statistics using SPSS. Sage, London Clark AB (1982) Scent marks as social signals in Galago crassicau- Forstmeier W, Schielzeth H (2011) Cryptic multiple hypotheses test- datus. I. Sex and reproductive status as factors in signals and ing in linear models: overestimated effect sizes and the winner’s responses. J Chem Ecol 8:1133–1151. https:// doi. org/ 10. 1007/ curse. Behav Ecol Sociobiol 65:47–55. https://doi. or g/10. 1007/ BF009 86984s00265- 010- 1038-5 Clarke RW (1940) Water distribution and sexual skin of the baboon. Fox GJ (1982) Potentials for pheromones in chimpanzee vaginal fatty Am J Physiol Legacy Content 131:325–330. https://doi. or g/10. acids. Folia Primatol 37:255–266. https:// doi. org/ 10. 1159/ 00015 1152/ ajple gacy. 1940. 131.2. 325 6036 Clarke PMR, Barrett L, Henzi SP (2009) What role do olfactory cues Fox J, Weisberg S (2011) An {R} companion to applied regression, play in chacma baboon mating? Am J Primatol 71:493–502. second edition. Sage, Thousand Oaks, CA. http:// socse rv. socsci. https:// doi. org/ 10. 1002/ ajp. 20678mcmas ter. ca/ jfox/ Books/ Compa nion. Accessed 20 Jan 2021 Converse LJ, Carlson AA, Ziegler TE, Snowdon CT (1995) Com- Gildersleeve KA, Haselton MG, Larson CM, Pillsworth EG (2012) munication of ovulatory state to mates by female pygmy mar- Body odor attractiveness as a cue of impending ovulation in mosets, Cebuella pygmaea. Anim Behav 49:615–621. https:// women: evidence from a study using hormone-confirmed ovula- doi. org/ 10. 1016/ 0003- 3472(95) 80194-4 tion. Horm Behav 61:157–166. https:// doi. org/ 10. 1016/j. yhbeh. Deschner T, Heistermann M, Hodges K, Boesch C (2003) Timing 2011. 11. 005 and probability of ovulation in relation to sex skin swelling in Gillman J, Stein HB (1941) A quantitative study of the inhibition of wild West African chimpanzees, Pan troglodytes verus. Anim estradiol benzoate by progesterone in the baboon (Papio por- Behav 66:551–560. https:// doi. org/ 10. 1006/ anbe. 2003. 2210 carius). Endocrinology 28:274–282. https:// doi. or g/ 10. 1210/ Deschner T, Heistermann M, Hodges K, Boesch C (2004) Female endo- 28-2- 274 sexual swelling size, timing of ovulation, and male behavior Goldfoot DA, Kravetz MA, Goy RW, Freeman SK (1976) Lack of effect in wild West African chimpanzees. Horm Behav 46:204–215. of vaginal lavages and aliphatic acids on ejaculatory responses in https:// doi. org/ 10. 1016/j. yhbeh. 2004. 03. 013 rhesus monkeys: behavioral and chemical analyses. Horm Behav Douglas PH, Hohmann G, Murtagh R et al (2016) Mixed messages: 7:1–27. https:// doi. org/ 10. 1016/ 0018- 506X(76) 90001-5 wild female bonobos show high variability in the timing of ovu- Goodall J (1986) The chimpanzees of Gombe. Patterns of behavior. lation in relation to sexual swelling patterns. BMC Evol Biol Belknap, Harvard University, Cambridge, MA 16:1–17. https:// doi. org/ 10. 1186/ s12862- 016- 0691-3 Graham CE, Collins DC, Robinson H, Preedy JRK (1972) Urinary Drea CM (2015) D’scent of man: a comparative survey of primate levels of estrogens and pregnanediol and plasma levels of proges- chemosignaling in relation to sex. Horm Behav 68:117–133. terone during the menstrual cycle of the chimpanzee: relationship https:// doi. org/ 10. 1016/j. yhbeh. 2014. 08. 001 to the sexual swelling. Endocrinology 91:13–24. https:// doi. org/ Drea CM, Boulet M, Delbarco-Trillo J et  al (2013) The “secret” 10. 1210/ endo- 91-1- 13 in secretions: methodological considerations in deciphering Haselton MG, Gildersleeve K (2011) Can men detect ovulation? Curr primate olfactory communication. Am J Primatol 75:621–642. Dir Psychol Sci 20:87–92. https:// doi. org/ 10. 1177/ 09637 21411 https:// doi. org/ 10. 1002/ ajp. 22143 402668 Dubuc C, Muniz L, Heistermann M et al (2012) Do males time their Havliček J, Dvorakova R, Bartoš L, Flegr J (2006) Non-advertized mate-guarding effort with the fertile phase in order to secure does not mean concealed: body odour changes across the human fertilisation in Cayo Santiago rhesus macaques? Horm Behav menstrual cycle. Ethology 112:81–90. https:// doi. org/ 10. 1111/j. 61:696–705. https:// doi. org/ 10. 1016/j. yhbeh. 2012. 03. 0031439- 0310. 2006. 01125.x Dulac C, Torello AT (2003) Molecular detection of pheromone sig- Hepper PG, Wells DL (2010) Individually identifiable body odors nals in mammals: from genes to behaviour. Nat Rev Neurosci are produced by the gorilla and discriminated by humans. Chem 4:551–562. https:// doi. org/ 10. 1038/ nrn11 40 Senses 35:263–268. https:// doi. org/ 10. 1093/ chemse/ bjq015 Ellis RA, Montagna W (1962) The skin of primates. VI. The skin of Higham JP, Hebets EA (2013) An introduction to multimodal com- the gorilla (Gorilla gorilla). Am J Phys Anthropol 20:79–93. munication. Behav Ecol Sociobiol 67:1381–1388. https://doi. or g/ https:// doi. org/ 10. 1002/ ajpa. 13302 0021010. 1007/ s00265- 013- 1590-x Emery Thompson M (2005) Reproductive endocrinology of wild Hu J, Zhang Z, Shen W-J, Azhar S (2010) Cellular cholesterol delivery, female chimpanzees (Pan troglodytes schweinfurthii): meth- intracellular processing and utilization for biosynthesis of ster- odological considerations and the role of hormones in sex and oid hormones. Nutr Metab (Lond) 7:47. https:// doi. org/ 10. 1186/ conception. Am J Primatol 67:137–158. https://doi. or g/10. 1002/ 1743- 7075-7- 47 ajp. 20174 Jamil T, Ozinga WA, Kleyer M, ter Braak CJF (2013) Selecting traits Emery Thompson M, Wrangham RW (2008) Male mating interest that explain species–environment relationships: a generalized varies with female fecundity in Pan troglodytes schweinfurthii 1 3 Primates (2022) 63:365–376 375 linear mixed model approach. J Veg Sci 24:988–1000. https:// Nishida T (1968) The social group of wild chimpanzees in the doi. org/ 10. 1111/j. 1654- 1103. 2012. 12036.x Mahali Mountains. Primates 9:167–224. https:// doi. or g/ 10. Jänig S, Weiß BM, Widdig A (2018) Comparing the sniffing behavior 1007/ BF017 30971 of great apes. Am J Primatol 80:e22872. https:// doi. org/ 10. 1002/ Ostner J, Nunn CL, Schülke O (2008) Female reproductive syn- ajp. 22872 chrony predicts skewed paternity across primates. Behav Ecol Jänig S, Weiß BM, Birkemeyer C, Widdig A (2019) Comparative 19:1150–1158. https:// doi. org/ 10. 1093/ beheco/ arn093 chemical analysis of body odor in great apes. Am J Primatol. Partan S, Marler P (1999) Communication goes multimodal. Sci- https:// doi. org/ 10. 1002/ ajp. 22976 ence 283:72–73. https://doi. or g/10. 1126/ scien ce. 283. 5406. 1272 Kappeler PM, van Schaik C (2004) Sexual selection in primates— Pfefferle D, Kazem AJN, Brockhausen RR et al (2014a) Monkeys new and comparative perspectives. Cambridge University Press, spontaneously discriminate their unfamiliar paternal kin under Cambridge natural conditions using facial cues. Curr Biol 24:1806–1810. Klailova M, Lee PC (2014) Wild western lowland gorillas signal selec-https:// doi. org/ 10. 1016/j. cub. 2014. 06. 058 tively using odor. PLoS ONE 9:e99554. https:// doi. org/ 10. 1371/ Pfefferle D, Ruiz-Lambides AV, Widdig A (2014b) Female rhesus journ al. pone. 00995 54 macaques discriminate unfamiliar paternal sisters in playback Klinkova E, Hodges JK, Fuhrmann K et al (2005) Male dominance experiments—support for acoustic phenotype matching. Proc rank, female mate choice and male mating and reproductive suc- R Soc B 281:1–9 cess in captive chimpanzees. Int J Primatol 26:357–484. https:// Quinn GP, Keough MJ (2002) Experimental design and data analysis doi. org/ 10. 1007/ s10764- 005- 2929-6 for biologists, 1st edn. Cambridge University Press, Cambridge Krohn PL, Zuckerman S (1937) Water metabolism in relation to the R Core Team (2015) A language and environment for statistical com- menstrual cycle. J Physiol 88:369–387 puting. Version 3.2.3. R Foundation for Statistical Computing, Kücklich M, Möller M, Marcillo A et al (2017) Different methods Vienna. http:// www.R- proje ct. org/. Accessed 20 Jan 2021 for volatile sampling in mammals. PLoS ONE 12:e0183440. Rigaill L, Higham JP, Lee PC et al (2013) Multimodal sexual signal- https:// doi. org/ 10. 1371/ journ al. pone. 01834 40 ing and mating behavior in olive baboons (Papio anubis). Am J Kücklich M, Weiß BM, Birkemeyer C et al (2019) Chemical cues of Primatol 75:774–787. https:// doi. org/ 10. 1002/ ajp. 22154 female fertility states in a non-human primate. Sci Rep 9:1–12. Singh D, Bronstad PM (2001) Female body odour is a potential cue https:// doi. org/ 10. 1038/ s41598- 019- 50063-w to ovulation. Proc R Soc B 268:797–801. https:// doi. or g/ 10. Marty JS, Higham JP, Gadsby EL, Ross C (2009) Dominance, col-1098/ rspb. 2001. 1589 oration, and social and sexual behavior in male drills Mandril- Smith TE, Abbott DH (1998) Behavioral discrimination between cir- lus leucophaeus. Int J Primatol 30:807–823. https://doi. or g/10. cumgenital odor from peri-ovulatory dominant and anovulatory 1007/ s10764- 009- 9382-x female common marmosets (Callithrix jacchus). Am J Primatol Matsumoto-Oda A (1999) Mahale chimpanzees: grouping patterns 46:265–284. https:// doi. org/ 10. 1002/ (SICI) 1098- 2345(1998) and cycling females. Am J Primatol 47:197–207. https:// doi. 46:4% 3c265:: AID- AJP1% 3e3.0. CO;2-D org/10. 1002/ (SICI) 1098- 2345(1999) 47:3% 3c197:: AID- AJP2% Smith TD, Bhatnagar KP (2004) Microsmatic primates: reconsid- 3e3.0. CO;2-3 ering how and when size matters. Anat Rec B 279B:24–31. Matsumoto-Oda A, Oda R, Hayashi Y et  al (2003) Vaginal fatty https:// doi. org/ 10. 1002/ ar.b. 20026 acids produced by chimpanzees during menstrual cycles. Folia Snowdon CT, Ziegler TE, Schultz-Darken NJ, Ferris CF (2006) Primatol 74:75–79. https:// doi. org/ 10. 1159/ 00007 0000 Social odours, sexual arousal and pairbonding in primates. Matsumoto-Oda A, Kutsukake N, Hosaka K, Matsusaka T (2007) Philos Trans R Soc B 361:2079–2089. https:// doi. org/ 10. 1098/ Sniffing behaviors in Mahale chimpanzees. Primates 48:81–85. rstb. 2006. 1932 https:// doi. org/ 10. 1007/ s10329- 006- 0006-1 Stein SE (1999) An integrated method for spectrum extraction and Michael RP, Keverne EB, Bonsall RW (1971) Pheromones: isolation compound identification from gas chromatography/mass spec- of male sex attractants from a female primate. Science 172:964– trometry data. J Am Soc Mass Spectrom 10:770–781. https:// 966. https:// doi. org/ 10. 1126/ scien ce. 172. 3986. 964doi. org/ 10. 1016/ S1044- 0305(99) 00047-1 Michael RP, Bonsall RW, Warner P (1974) Human vaginal secre- Stoffel MA, Caspers BA, Forcada J et al (2015) Chemical finger - tions: volatile fatty acid content. Science 186:1217–1219. prints encode mother–offspring similarity, colony membership, https:// doi. org/ 10. 1016/ 0306- 4530(75) 90007-4 relatedness, and genetic quality in fur seals. PNAS 112:E5005– Michael RP, Bonsall RW, Kutner M (1975) Volatile fatty acids, “cop- E5012. https:// doi. org/ 10. 1073/ pnas. 15060 76112 ulins”, in human vaginal secretions. Psychoneuroendocrinology Stumpf RM, Boesch C (2006) The efficacy of female choice in chim- 1:153–163. https:// doi. org/ 10. 1016/ 0306- 4530(75) 90007-4 panzees of the Tai Forest, Cote d’Ivoire. Behav Ecol Sociobiol Montagna W, Yun JS (1963) The skin of primates. XV. The skin of 60:749–765. https:// doi. org/ 10. 1007/ s00265- 006- 0219-8 the chimpanzee (Pan satyrus). Am J Phys Anthropol 21:189– Townsend SW, Deschner T, Zuberbühler K (2011) Copulation calls 203. https:// doi. org/ 10. 1002/ ajpa. 13302 10211 in female chimpanzees (Pan troglodytes schweinfurthii) convey Muller MN (2002) Agonistic relations among Kanyawara chimpan- identity but do not accurately reflect fertility. Int J Primatol zees. In: Boesch C, Hohmann G, Marchant L (eds) Behavioural 32:914–923. https:// doi. org/ 10. 1007/ s10764- 011- 9510-2 diversity in chimpanzees and bonobos. Cambridge University Tutin CEG (1979) Mating patterns and reproductive strategies in Press, pp 112–124 a community of wild chimpanzees (Pan troglodytes schwein- Nadler RD, Graham CE, Gosselin RE, Collins DC (1985) Serum furthii). Behav Ecol Sociobiol 6:29–38. https://doi. or g/10. 1007/ levels of gonadotropins and gonadal steroids, including tes-BF002 93242 tosterone, during the menstrual cycle of the chimpanzee (Pan Vaglio S, Minicozzi P, Romoli R et al (2016) Sternal gland scent- troglodytes). Am J Primatol 9:273–284. https://doi. or g/10. 1002/ marking signals sex, age, rank, and group identity in captive ajp. 13500 90404 mandrills. Chem Senses 41:177–186. https:// doi. org/ 10. 1093/ Neumann C, Assahad G, Hammerschmidt K et al (2010) Loud calls chemse/ bjv077 in male crested macaques, Macaca nigra: a signal of dominance van den Berg RA, Hoefsloot HC, Westerhuis JA et al (2006) Cen- in a tolerant species. Anim Behav 79:187–193. https:// doi. org/ tering, scaling, and transformations: improving the biological 10. 1016/j. anbeh av. 2009. 10. 026 information content of metabolomics data. BMC Genom 7:142. https:// doi. org/ 10. 1186/ 1471- 2164-7- 142 1 3 376 Primates (2022) 63:365–376 Wallis J (1997) A survey of reproductive parameters in the free- Ziegler TE, Epple G, Snowdon CT et  al (1993) Detection of the ranging chimpanzees of Gombe National Park. J Reprod Fertil chemical signals of ovulation in the cotton-top tamarin, Sagui- 109:297–307 nus oedipus. Anim Behav 45:313–322. https://doi. or g/10. 1006/ Weiß BM, Marcillo A, Manser M et al (2018) A non-invasive method anbe. 1993. 1036 for sampling the body odour of mammals. Methods Ecol Evol Ziegler TE, Schultz-Darken NJ, Scott JJ et al (2005) Neuroendocrine 9:420–429. https:// doi. org/ 10. 1111/ 2041- 210X. 12888 response to female ovulatory odors depends upon social con- Wrangham RW (2002) The cost of sexual attraction: is there a trade- dition in male common marmosets, Callithrix jacchus. Horm off in female Pan between sex appeal and received coercion? In: Behav 47:56–64. https:// doi. org/ 10. 1016/j. yhbeh. 2004. 08. 009 Boesch C, Hohmann G, Marchant L (eds) Behavioural diver- sity in chimpanzees and bonobos. Cambridge University Press, Publisher's Note Springer Nature remains neutral with regard to Cambridge, pp 204–215 jurisdictional claims in published maps and institutional affiliations. Wyatt TD (2014) Pheromones and animal behavior: chemical signals and signatures. Cambridge University Press Young C, Majolo B, Heistermann M et al (2013) Male mating behav- iour in relation to female sexual swellings, socio-sexual behav- iour and hormonal changes in wild Barbary macaques. Horm Behav 63:32–39. https:// doi. org/ 10. 1016/j. yhbeh. 2012. 11. 004 1 3

Journal

Primates – Springer Journals

Published: Jul 1, 2022

Keywords: Olfactory communication; Olfaction; Gas chromatography—mass spectrometry; Chemical composition; Menstrual cycle; Anogenital swelling

There are no references for this article.