Baby on board: olfactory cues indicate pregnancy and fetal sex in a non-human primate

Abstract

Olfactory cues play an integral, albeit underappreciated, role in mediating vertebrate social and reproductive behaviour. These cues fluctuate with the signaller's hormonal condition, coincident with and informative about relevant aspects of its reproductive state, such as pubertal onset, change in season and, in females, timing of ovulation. Although pregnancy dramatically alters a female's endocrine profiles, which can be further influenced by fetal sex, the relationship between gestation and olfactory cues is poorly understood. We therefore examined the effects of pregnancy and fetal sex on volatile genital secretions in the ring-tailed lemur (Lemur catta), a strepsirrhine primate possessing complex olfactory mechanisms of reproductive signalling. While pregnant, dams altered and dampened their expression of volatile chemicals, with compound richness being particularly reduced in dams bearing sons. These changes were comparable in magnitude with other, published chemical differences among lemurs that are salient to conspecifics. Such olfactory ‘signatures’ of pregnancy may help guide social interactions, potentially promoting mother–infant recognition, reducing intragroup conflict or counteracting behavioural mechanisms of paternity confusion; cues that also advertise fetal sex may additionally facilitate differential sex allocation.

1. Introduction

Female animals routinely broadcast information about their reproductive state. In general, the research emphasis in reproductive signalling has been on cues associated with fecundity, ovulation or mate quality; however, there are also potentially potent cues associated with gestation. The physiological changes underlying pregnancy are complex and, compared with ovulation, are more enduring. The corresponding changes to endocrine concentrations, which may guide reproductive cues of pregnancy, can be extreme. Notably, in comparison with maximum concentrations outside of pregnancy, concentrations of certain sex steroids during pregnancy can increase more than 1600% [1] and vary by fetal sex (reviewed in [2]). Moreover, pregnancy cues may serve important functions. In primates, for instance, gestational cues may reduce intragroup conflict in species that compete intensely for mates [3] or may prepare cooperatively breeding males for the energetic and behavioural changes necessary for paternal care [4]. Historically, the conveyance of reproductive information, particularly in primates, has been evaluated primarily through visual and behavioural channels (e.g. [5,6]); nonetheless, there is growing recognition of reproductive signalling via olfactory channels [7]. As in several non-primate species [8–10], olfactory cues of pregnancy occur in humans, possibly facilitating mother–infant recognition [11]; otherwise, the potential for olfactory gestational cues in primates has been ignored. Here, we investigate pregnancy effects on olfactory cues in a promiscuous, non-human primate, the ring-tailed lemur (Lemur catta).

Recently, the ring-tailed lemur has emerged as a useful model for examining the form and function of olfactory signals relevant to reproduction. Among females, which use scent to demarcate birthing sites [12] and for competitive overmarking [7], variations in odorant profiles (whether in compound type, number or relative concentration, see e.g. diversity indexes in Charpentier et al. [13]) correspond to natural [14,15] and experimentally induced [16] variations in reproductive hormones. Through behavioural bioassays, genetic profiling and chemical analyses, lemur chemosignals have been shown to convey information about sex, breeding condition and individual identity, as well as genetic quality and relatedness (reviewed in [7,16]). By contrast, dominance status within the sexes, which in lemurs can be nonlinear, non-transitive and highly variable [17], is recognizable via scent in known individuals only [18], but is not chemically encoded [18,19]. Using a within-subjects sampling design, we extend these prior studies by characterizing chemical changes in lemur genital secretions between preconception and pregnancy, while also considering the potential effects of other concurrent physiological or environmental changes, such as age, season, litter size and fetal sex.

2. Material and methods

The subjects were 12 captive, sexually mature (1.5–21 years), female ring-tailed lemurs, housed in semi-free-ranging, mixed-sex groups at the Duke Lemur Center in Durham, NC, USA. We monitored reproduction over a 6-year period (2004–2010), including during 14 pregnancies ([2]; see electronic supplementary material). To illustrate any potential olfactory–endocrine relationship, we present accompanying sex steroid concentrations in female subjects, recalculated from [2].

We collected samples of labial secretions near conception (mean ± s.e.m.: 17.3 ± 3.7 days preconception) and mid-pregnancy (mean ± s.e.m.: 88.8 ± 4.6 days of a 135-day gestation period) for analysis by gas chromatography–mass spectrometry [20]. We used richness (or number) as a proxy of chemical complexity [13], retaining only compounds that eluted at 8–43 min and comprised more than or equal to 0.01% of the chromatogram [20]. We assigned ‘fetal sex’ following [2], grouping dams carrying singleton or twin females as ‘females bearing daughters’ (FBD; n = 8) and dams carrying at least one male (singleton male, twin males or mixed-sex twins) as ‘females bearing sons’ (FBS; n = 6; electronic supplementary material).

We first addressed and dismissed any potential effects on a female's chemical richness of her age, the seasonal timing of her pregnancy and her litter size (see electronic supplementary material). We then tested for effects of reproductive condition (preconceptive versus pregnant) and fetal sex (FBD versus FBS) on her chemical richness and endocrine concentrations, using two-factor ANOVAs (univariate repeated-measures model, JMP PRO v. 11.0, SAS; electronic supplementary material). To contextualize the main effect of reproductive condition on chemical complexity, we also calculated the mean differences in chemical richness between several previously studied groups from this population ([13,14,16,19]; see electronic supplementary material). Having decided a priori to compare FBD and FBS, we resolved the significant interaction using F-tests for simple effects.

3. Results

The odorant profiles of female ring-tailed lemurs changed with pregnancy, relative to preconception, as revealed by differences in the gas chromatograms of their genital secretions (figure 1a,b): pregnancy was associated with changes in the relative proportions of the volatile compounds expressed and with a significant decrease in the total number of compounds (F_1,10 = 14.41, p < 0.01; figure 1c). The change in chemical richness (mean ± s.e.m.: 9.67 ± 6.62 compounds) was comparable in magnitude with the behaviourally salient, chemical differences observed between other groups of individuals, including those found between intact and hormonally contracepted females (table 1).

Figure 1.

Volatile chemicals varied among the genital secretions of preconceptive and pregnant ring-tailed lemurs: gas chromatograms depict variation in the relative proportions of compounds present in the labial secretions of a representative ring-tailed lemur before conception (a) and while pregnant with twin daughters (b). Numbers denote: 1, internal standard (hexachlorobenzene); 2, squalene and 3, cholesterol. Bar graphs represent differences by reproductive condition in the chemical richness (mean ± s.e.m.) derived from the chromatograms of all female subjects (c). ** denotes significance at p = 0.01.

Table 1.

Differences in the chemical richness of ring-tailed lemur genital secretions between groups or conditions for which existing behavioural data support the salience of the olfactory signal.

comparison	mean difference ± s.e.m.	source of chemical data	source of behavioural validation
individual versus individual, females	2.60 ± 0.78	Scordato et al. [19]	Scordato & Drea [18]
individual versus individual, males	3.69 ± 0.94	Scordato et al. [19]	Scordato & Drea [18]
relatively homozygous versus relatively heterozygous, males	5.28 ± 6.11	Charpentier et al. [13]	Charpentier et al. [21]
breeding season versus non-breeding season, females	17.70 ± 6.74	Boulet et al. [14]	Scordato & Drea [18]
breeding season versus non-breeding season, males	19.40 ± 3.92	Boulet et al. [14]	Scordato & Drea [18]
hormonally intact versus contracepted, females	24.08 ± 10.50	Crawford et al. [16]	Crawford et al. [16]
males versus females, non-breeding season	45.70 ± 5.26	Boulet et al. [14]	Scordato & Drea [18]
males versus females, breeding season	82.80 ± 5.76	Boulet et al. [14]	Scordato & Drea [18]

Remarkably, a dam's volatile chemical profile during pregnancy also varied with the sex of her fetus. A significant interaction between reproductive condition and fetal sex (F_1,10 = 5.21, p < 0.05) owed primarily to a greater loss of chemical richness during pregnancy in FBS than in FBD (F_1,12.34 = 5.51, p < 0.05; figure 2a). Because there was no effect of a female's eventual fetal sex on the chemical richness of her preconceptive odorants (F_1,12.34 = 0.74, p = 0.40, N.S.), the differences by fetal sex during pregnancy could not be explained by the scent signatures [19] of individual females. Instead, these findings may relate to underlying differences, by fetal sex, in the concentrations of maternal sex steroids, which increase during pregnancy (figure 2b; table 2).

Figure 2.

Mean ± s.e.m. chemical richness (a) and serum steroid concentrations (b) varied by reproductive condition and fetal sex in female ring-tailed lemurs. T, testosterone; A₄ androst-4-ene-3,17,dione; OHP, 17α-hydroxyprogesterone; E₂, 17β-oestradiol. * denotes significance at p < 0.05; ** denotes significance at p < 0.01.

Table 2.

Differences in the endocrine profiles of female ring-tailed lemurs in relation to reproductive condition and fetal sex. Notably, serum concentrations during gestation were significantly different between females bearing daughters (FBD) and females bearing sons (FBS) for all measured steroids.

hormone	reproductive condition	fetal sex	interaction: condition/sex	FBD versus FBS while pregnant
testosterone	F1,15.3 = 42.11 p < 0.01	F1,18.52 = 3.72 p > 0.05, N.S.	F1,15.3 = 8.58 p = 0.01	F1,17.02 = 11.56 p < 0.01
androst-4-ene-3,17,dione	F1,14.65 = 59.56 p < 0.01	F1,19.12 = 1.82 p > 0.10, N.S.	F1,14.65 = 27.98 p < 0.01	F1,16.94 = 21.41 p < 0.01
17α-hydroxyprogesterone	F1,15.11 = 63.29 p < 0.01	F1,18.61 = 4.49 p < 0.05	F1,15.11 = 8.08 p < 0.05	F1,16.97 = 12.13 p < 0.01
17β-oestradiol	F1,13.86 = 67.08 p < 0.01	F1,17.43 = 5.49 p < 0.05	F1,13.86 = 27.41 p < 0.01	F1,15.73 = 27.54 p < 0.01

4. Discussion

We show that odorant expression in lemurs is altered during pregnancy and is further differentiated in accordance with the sex of the dam's developing offspring. The chemical effects induced by pregnancy were comparable in magnitude with other differences in scent that, as revealed by behavioural bioassays, are salient to ring-tailed lemurs [16,18,21]. The degree to which compounds are lost during pregnancy and influenced by fetal sex appears to be inversely related to the female's changing sex steroid concentrations. Given the established link between sex steroids and olfactory profiles [7], these findings suggest possible endocrine involvement (or competing energy allocation) in the production of olfactory gestational cues.

Gestational cues could serve various functions across primate species, from promoting social cohesion [3] to engendering parental investment [4] or kin recognition [11]. To the extent that olfactory cues of pregnancy occur in other primates, our findings could be relevant to these existing hypotheses, as well as to theories on the functionality of multiple mating. Notably, a female's efforts to confuse males about their potential paternity (‘paternity confusion’) is commonly invoked to explain a female primate's engagement in multiple mating during pregnancy, particularly at times when there are no overt behavioural cues of pregnancy or when visual cues of pregnancy are not yet apparent [22,23]—in other words, when males presumably lack knowledge about female reproductive state. Yet, we have no information about potential scent cues of pregnancy in multiply mating species, whose males are nevertheless known to engage in olfactory investigation of vaginal secretions prior to mating [7,24,25]. We must therefore consider the possibility that olfactory cues could provide males with a means of detecting a proceptive dam's reproductive state. Discerning males could possibly thwart the female's purported tactics, either by not mating with her or by mating with her without becoming confused about paternity. If so, one might only expect multiple mating to be an effective means of paternity confusion in species in which either the female does not produce olfactory pregnancy cues or the male is unable to detect them.

To the extent that olfactory advertisement of fetal sex might also occur more broadly than in Lemur, our findings might help elucidate a potential mechanism of differential sex allocation. Under certain conditions, animals increase fitness by preferentially investing in offspring of a particular sex [26]. Although theories addressing sex allocation typically implicate mechanisms operating at conception [27] or post-parturition [28], mechanisms operating during gestation would allow animals to respond to important social or environmental changes while the ongoing costs of investment are still considerable. An expectant dam could potentially use her own olfactory cues as a self-referent to inform her investment decisions; alternately, conspecifics could use these cues as predictors of imminent sex ratios to guide their own sex-allocation strategies. Although opportunities to study a putative gestational mechanism of sex allocation will likely remain limited in this endangered, long-lived species, such limitations do not preclude investigations of potential sex-specific gestational cues or gestational sex allocation in more tractable study systems that share key life-history characteristics (e.g. [29]).

The observation that pregnancy alters scent profiles in lemurs (this study) and in humans [11] suggests that olfactory cues of pregnancy may be widespread or highly conserved among primates. Such findings highlight the importance of considering multiple sensory modalities when examining reproductive signals, even in taxa historically thought to rely relatively little on olfaction [7,30,31].

Ethics statement

Our animal procedures were approved by the Institutional Animal Care and Use Committee of Duke University under protocols A245-03-07, A232-06-07 and A171-09-06.

Data accessibility

Data can be accessed at Dryad (doi:10.5061/dryad.ps70b).

Author contribution

C.M.D. conceived of the endocrine and olfactory research program in strepsirrhines, collected the samples, and provided the endocrine data. J.C.C. and C.M.D. designed the present study. J.C.C. performed the chemical and statistical analyses and prepared the figures. J.C.C. and C.M.D. wrote the manuscript.

Funding statement

This research was funded by NSF (BCS-0409367, IOS-0719003 to C.M.D.). Analysis and write-up were supported by NSF (IOS-1021633 to C.M.D.) and by a Visiting Scholar's Fellowship from the National Evolutionary Synthesis Center (NSF EF-0423641), a Chancellor's Fellowship from the University of California, Berkeley, an NSF Graduate Research Fellowship and funding from the Research Council of Norway (to J.C.C.). DLC publication no. 1281.