Skin bacterial microbiome diversity predicts lower activity levels in female, but not male, guppies, Poecilia reticulata

Abstract

While the link between the gut microbiome and host behaviour is well established, how the microbiomes of other organs correlate with behaviour remains unclear. Additionally, behaviour–microbiome correlations are likely sex-specific because of sex differences in behaviour and physiology, but this is rarely tested. Here, we tested whether the skin microbiome of the Trinidadian guppy, Poecilia reticulata, predicts fish activity level and shoaling tendency in a sex-specific manner. High-throughput sequencing revealed that the bacterial community richness on the skin (Faith's phylogenetic diversity) was correlated with both behaviours differently between males and females. Females with richer skin-associated bacterial communities spent less time actively swimming. Activity level was significantly correlated with community membership (unweighted UniFrac), with the relative abundances of 16 bacterial taxa significantly negatively correlated with activity level. We found no association between skin microbiome and behaviours among male fish. This sex-specific relationship between the skin microbiome and host behaviour may indicate sex-specific physiological interactions with the skin microbiome. More broadly, sex specificity in host–microbiome interactions could give insight into the forces shaping the microbiome and its role in the evolutionary ecology of the host.

1. Background

The effects of gut-associated microbes on host behaviour are well established [1]. For example, gut microbiome in zebrafish modulate activity level and anxiety behaviours [2]: germ-free conditions induce hyperactivity, but symbiont inoculation ameliorates this effect [3]. Host social behaviour is also modulated by the gut microbiome [4]: germ-free mice are more social [5], and gut-probiotic inoculated zebrafish display altered shoaling behaviour [6]. The microbiomes of other organs, such as the skin, could interact with animal behaviour but have received relatively little attention. For example, social cleaning goby ecotypes have more diverse skin microbiomes and a higher abundance of opportunistic pathogens than their non-cleaner counterparts [7]. Indeed, the skin microbiome may directly connect to nervous system processes: the 'brain–skin axis' [8–10], analogous to the ‘gut–brain axis’ [6,11]. Strikingly, gut and skin microbiomes often synchronously fluctuate in response to host behavioural changes [12–14]. However, previous research generally tests how different lifestyles alter the skin microbiome and not, to our knowledge, how the skin microbiome may correlate with behavioural expression within a single population.

Because of sex differences in both physiology and behaviour, it is likely that host–microbe interactions, and their effects on host behaviour, are sex-specific. Indeed, among mammals, there are sex differences in the lung, gut, skin and salivary gland microbiomes [15–19]. Very few studies have tested for sex-specific behaviour–microbiome interactions: gut-associated microbial community richness was lower in female rats with addiction phenotypes, with a specific microbe-associated with impulsive behaviours, while no significant patterns occurred in males [20] and exposure to a broad-spectrum antibiotic had sex-specific effects on aggression in hamsters [21]. While this last result could demonstrate sex-specific responses to microbiome disruption, the antibiotic used may affect males and females differently [22].

Skin microbiomes are likely particularly important in modulating host social behaviour: skin is often vital for chemical cue production [23,24]. Such cues are composed of a myriad of volatile chemicals, many of which are produced [25] and modulated by microbes in the laboratory and field [26]. For example, sex differences in the skin microbiome of tree frogs results in sex-specific chemical cues [27]. Across taxa, chemical cues can indicate the sex, age and infection status of conspecifics [28] and may be essential to species recognition, kin recognition and mate choice [29]. Thus, understanding sex-specific microbiome–behaviour correlations could have important implications for understanding host behaviour at the individual and population levels.

We used the Trinidadian guppy, Poecilia reticulata, to investigate whether activity level and shoaling tendency were associated with the skin microbiome in a sex-specific manner. Male and female guppies exhibit differences in their social, boldness, emigration and sickness behaviours [30–33]. In guppies and other taxa, activity levels are often positively associated with survival rates [34,35], metabolism [36], growth rate [37], general health [38] and reproductive success [39]. Both male [40] and female [41,42] guppies prefer more active mates. Similarly, host grouping behaviour, such as shoaling in guppies, has fitness benefits, including access to mates, foraging efficiency and defence against predators [43–47]. Therefore, guppies are an ideal system with which to test whether the skin microbiome is associated with fitness-relevant behaviours. We measured the proportion of time spent swimming—‘activity level’—and associated with a shoal—‘shoaling tendency’—and then collected the skin microbiome of the host. Our results indicate sex differences in how community richness and membership of the skin are associated with behaviour.

2. Methods

(a) . Fish origin and maintenance

We used laboratory-bred descendants of guppies collected from the Caura River, Trinidad. Guppies were housed at densities of 1–2 fish per litre in 4.5 l tanks in a recirculating system at 24 ± 1°C, on a 12 L : 12 D lighting schedule and fed daily on flake and Artemia. All conditions were kept identical for the experiment duration and conducted in five experimental batches between August and October 2017.

(b) . Behavioural experiment and microbiome sampling

Our data come from the ‘pre-infection’ time point in the experiment presented in Jog et al. [32] and Stephenson [48]: we collected swabs for microbiome analysis from a subset of these fish. To assay behaviours, we used a central glass aquarium (test aquarium), flanked on both short sides by smaller aquaria (stimulus aquaria), all with 5 cm water depth. The stimulus aquaria were lit from above by 32 cm LED strip lights (350 lm, 5 W, 4000 k, MeRox® Technics). The test aquarium had a black Perspex lid, to which was fastened a GoPro camera (Hero4 Black Edition; GoPro Inc., San Mateo, CA, USA). We placed a test fish (of either sex: n = 18 females; 19 males) in an acetate cylinder in the centre of the test tank for a 5 min settling period. We then remotely set the GoPro to record, lifted the settling cylinder using a pulley and recorded a 10 min test period. For a subset of trials, we placed a shoal of three females in one of the stimulus tanks (side chosen haphazardly [32,48]). The short sides of the test aquarium were one-way mirrors, such that test fish could see stimulus fish, but not vice versa. The experimental set-up was enclosed in a blackout curtain to ensure even lighting and prevent disturbance. The test aquarium was scrubbed with 70% ethanol, rinsed and refilled between trials. We quantified ‘shoaling tendency’ as the proportion of trial time the test fish spent within the third of the test tank closest to the stimulus tank and ‘activity level’ as the proportion of time the test fish actively swam throughout the 10 min. One of four observers, blind to the treatment, recorded the activity levels and shoaling levels from the videos using JWatcher™ 1.0 (www.jwatcher.ucla.edu; [49]).

(c) . Skin microbiome collection

Immediately following the behaviour trial, test fish were moved to an individual 1 l glass tank for identification purposes. Each of these tanks was sprayed with 70% ethanol, scrubbed, rinsed and filled with clean water before a fish was added. The day after the behaviour trials, fish skin microbiome samples were collected, following established protocols [50,51]. We anaesthetized individual fish (0.02% tricaine methanesulfonate; PHARMAQ Ltd, Fordingbridge, UK) and washed them twice with two separate 20 ml aliquots of sterile water to ensure that the sample included primarily skin-associated microbes rather than transient microbes [51,52]. We then swabbed the body surface with a sterile cotton swab for 30 s. Between fish, all glassware was sterilized with 70% ethanol. This methodology may not have removed all of the DNA of dead or transient bacteria (doing so is rarely possible [52,53]), but any contamination and resulting homogenization of the microbiome samples across fish would have reduced our probability of detecting a correlation with behaviour. The swabs were stored at −20°C until DNA extraction.

(d) . Microbiome inventory

We used QIAamp PowerFecal DNA Isolation Kit (Qiagen, Hilden, Germany) to extract DNA, following the manufacturer's protocol. To control for contaminants, we included four blank samples in extractions and sequencing [53]. Extracted DNA was shipped on dry ice to the DNA Services Facility at the University of Illinois at Chicago (Chicago, IL, USA). All library preparation, amplification and sequencing were performed on the Illumina Miseq platform (V4 region of the bacterial 16S rRNA gene). We filtered the raw sequences for quality, removed chimeric reads, merged forward and reverse reads and assigned the remaining reads to amplicon sequence variant (ASVs) using the DADA2 pipeline in QIIME2 2019.7 [54,55]. We trimmed each read to 220 base pairs, removed all singleton reads and used FastTree to build a phylogenetic tree [56], using the SILVA database classifier release 138 to assign taxonomy [57]. We removed all non-bacterial reads (archaea, chloroplast or mitochondria) and any ASVs detected in control blank samples prior to analysis [53,58].

(e) . Data analysis

First, we calculated Faith's phylogenetic distance (Faith's PD) as a metric for skin microbial alpha diversity for each sample and compared Faith's PD between female and male guppies using a Kruskal–Wallis test using QIIME2.

To test for correlations between behaviour and microbiome alpha diversity, we used activity level or shoaling tendency as the response variable in two generalized linear mixed models (glmm) in the glmmTMB package, v. 1.1.2.3 (Beta error family and identity logit function within R v.4.1.0 [59]; see electronic supplementary material). As fixed main effects, we included test fish sex and length (controlling for sex differences), body condition (scaled mass index [60]), Faith's PD, trial date and time of day, and whether there was a shoal present. We also included the two-way interaction between sex and Faith's PD. We used visreg [61] and ggplot2 [62] for plotting.

To investigate correlates of skin microbiome bacterial community membership, we calculated the unweighted UniFrac distance matrix in Qiime2, imported it into Primer v7, and built a distance-based linear model (DISTLM; [63]). This model included the same predictor variables as the glmm. We used permutation to assess statistical significance. We visualized the fitted DISTLM models in multidimensional space, using the distance-based redundancy analysis in PRIMER v7 (see electronic supplementary material).

Finally, to test for associations between individual bacterial taxa and guppy activity level and shoaling tendency, we used MaAsLin2 [64] and added trial date as a fixed effect to control for differences caused by experimental date. This program first uses boosting in a univariate pre-screen to identify taxa (arcsine square-root transformed relative abundances) potentially associated with a focal variable and then uses linear mixed-effects models to test for statistically significant associations using the Benjamini–Hochberg false discovery rate method.

3. Results

(a) . Sex-specific skin microbiomes

We found no significant difference between the sexes in Faith's PD (figure 1a; Kruskal–Wallis pairwise test: H = 2.04, q = 0.153); however, there was a significant sex difference in community membership (figure 1b; DISTLM model marginal test: pseudo-F = 1.37, p = 0.033).

Figure 1.

The diversity of the skin microbiome bacterial community (Faith's PD) did not significantly differ between male and female guppies (a), but bacterial community membership did (b) Principal coordinates plot of unweighted UniFrac distances between samples). Box plots give the median (dark line), interquartile range (box), values within 1.5 x the interquartile range (whiskers) and outliers (points).

(b) . Activity level

The correlation between skin microbial alpha diversity and activity level differed significantly between female and male guppies (figure 2a; interaction between sex and Faith's PD, likelihood ratio test; χ² = 10.54, p = 0.001). As a post hoc test to evaluate whether this correlation was significantly different from 0 among either sex, we subset the data by sex and re-ran the model. We found that female guppies with higher community richness spent less time actively swimming (likelihood ratio test, χ² = 15.91, p < 0.0001), but there was no such pattern among males (likelihood ratio test, χ² < 0.001, p = 0.980).

Figure 2.

The proportion of time guppies spent swimming (a) and associating with a shoal (b) depended on an interaction between the diversity of the skin microbiome (Faith's PD) and host sex. Points are back-transformed partial residuals from the models, and curves and shading give the model fits and 95% confidence bands.

The best DISTLM model explained 21.7% of the variation in bacterial community and showed that, as well as fish sex, community membership covaried significantly with the date on which the trial was conducted (pseudo-F = 1.88, p = 0.001), as has been shown previously [65–67]. Among females, community membership covaried with activity level (pseudo-F = 1.42, p = 0.034). The relative abundances of 16 bacterial taxa in the female guppy skin microbiota were negatively correlated with activity level table 1: 38% of these 16 were proteobacteria, 12.5% were Acidobacteria and 12.5% Bacteroidota.

Table 1.

The correlation coefficient, s.e. and statistical significance (q-value using the Benjamini–Hochberg correction for multiple testing) of correlations between female guppy activity level and the relative abundance of bacterial features.

bacterial feature	coefficient	s.e.	q-value
p__Acidobacteriota.c__Vicinamibacteria.o__Vicinamibacterales.f__uncultured.g__uncultured	−0.093	0.012	0.000
p__Bacteroidota.c__Bacteroidia.o__Sphingobacteriales.f__env.OPS_17.g__env.OPS_17	−0.047	0.008	0.003
p__Chloroflexi.c__Chloroflexia.o__Chloroflexales.f__Roseiflexaceae.g__uncultured	−0.049	0.009	0.007
p__Desulfobacterota.c__Desulfuromonadia.o__PB19.f__PB19.g__PB19	−0.058	0.011	0.008
p__Proteobacteria.c__Gammaproteobacteria.o__Burkholderiales.f__Nitrosomonadaceae.g__Ellin6067	−0.096	0.019	0.012
p__Planctomycetota.c__OM190.o__OM190.f__OM190.g__OM190	−0.102	0.023	0.027
p__Proteobacteria.c__Gammaproteobacteria.o__Burkholderiales.f__Methylophilaceae.g__Methylotenera	−0.121	0.028	0.027
p__Proteobacteria.c__Gammaproteobacteria.o__CCD24.f__CCD24.g__CCD24	−0.044	0.010	0.027
p__Proteobacteria.c__Alphaproteobacteria.o__Caulobacterales.f__Hyphomonadaceae.g__SWB02	−0.043	0.010	0.027
p__Proteobacteria.c__Alphaproteobacteria.o__Rhizobiales.f__Rhizobiaceae.g__Mesorhizobium	−0.092	0.022	0.027
p__Patescibacteria.c__Saccharimonadia.o__Saccharimonadales.f__Saccharimonadales.g__Saccharimonadales	−0.029	0.007	0.029
p__Nitrospirota.c__Nitrospiria.o__Nitrospirales.f__Nitrospiraceae.g__Nitrospira	−0.097	0.023	0.030
p__Proteobacteria.c__Alphaproteobacteria.o__Rickettsiales.f__Rickettsiaceae.g__uncultured	−0.062	0.016	0.030
p__MBNT15.c__MBNT15.o__MBNT15.f__MBNT15.g__MBNT15	−0.039	0.009	0.030
p__Acidobacteriota.c__Vicinamibacteria.o__Vicinamibacterales.f__Vicinamibacteraceae.g__Vicinamibacteraceae	−0.087	0.022	0.036
p__Bacteroidota.c__Bacteroidia.o__Chitinophagales.f__Chitinophagaceae.g__Sediminibacterium	−0.106	0.028	0.047

(c) . Shoaling tendency

The correlation between skin microbial alpha diversity and shoaling differed significantly between female and male guppies (figure 2b; interaction between sex and Faith's PD, likelihood ratio test; χ² = 7.55, p = 0.006). A post hoc test in which we subset the data by sex and re-ran the model revealed that the correlation was not significantly different from 0 among either sex (see electronic supplementary material). Additionally, community membership was not associated with shoaling (see electronic supplementary material).

4. Discussion

We found that the guppy sexes differed significantly in community membership but not overall community richness (Faith's PD) of their skin microbiomes (figure 1). The correlation between skin Faith's PD and both behaviours also differed between the sexes (figure 2): male behaviour did not correlate with microbiome alpha (Faith's PD) or beta diversity metrics, but females with higher skin microbial species richness were significantly less active. We found that community membership was significantly correlated with female activity level, suggesting that specific bacterial taxa may be responsible for this result. Among females, community membership was also associated with activity level: we found 16 bacterial features that were negatively correlated with activity level (table 1). We focus our discussion on activity level because the results for this behaviour were more robust.

One explanation for our results is that the skin microbiome influences fish behaviour. The skin microbiome may be particularly relevant to fish as they are inhabited by many opportunistic pathogens [7,68,69] that could manipulate host immune function and energetics. Dysbiosis, the imbalance or shift in the naturally present microbiome compared to a healthy individual, has been shown to impact activity level [70,71]. Higher microbiome diversity could therefore reflect infection, to which reduced activity level is a common behavioural response [72]. Indeed, Sphingobacteriales.f_env.OPS bacteria is an opportunistic pathogen of zebrafish gills [73]. Alternatively, these microbes could be affecting the host through many different physiological pathways such as hormone production, olfactory modulation and inflammation, which could influence the fish's activity level without active disease [74].

Alternatively, it is possible that both behaviour and the microbiome are affected by some third factor we did not measure. For example, we co-housed fish from birth until adulthood, and group-level factors may have affected the behaviour and microbiome of our test fish [75]. We also found that both behaviour and the microbiome (as has been shown previously [65–67]) changed over the weeks during which we collected data (see electronic supplementary material).

Finally, fish behaviour, or the physiological correlates of that behaviour, may influence the skin microbiome. This explanation is potentially supported by the sex-specificity of our results: the sexes differed in their behavioural expression (figure 2; electronic supplementary material), and we found significant correlations between behaviour and the microbiome only among females. While there was a mild community membership difference between males and females, we did not find any specific bacterial taxa that differed significantly between the sexes, or differences in alpha diversity. The guppy sexes therefore have similar skin microbiomes, but may interact with them differently. For example, that the pattern of sex differences in behavioural responses to infection with Gyrodactylus spp. appears opposite to the results presented here [32,48] raises the possibility that the sexes differentially invest in immune responses to micro- and macroparasites [48,76–78]. Thus, while assigning causation requires further investigation, our results indicate that skin microbiomes show important, sex-specific correlations with fitness-enhancing behaviours.

Ethics

This work was performed with ethical approval from the Veterinäramt of Zürich (licence no. ZH177/15).

Data accessibility

The raw experimental data, description of the data and analysis code are available from the Dryad Digital Repository (https://doi.org/10.5061/dryad.34tmpg4nm) [79] and the electronic supplementary material [80]. Raw sequencing data are available from NCBI (accession number PRJNA838496).

Authors' contributions

R.D.K.: data curation, formal analysis, visualization, writing—original draft and writing—review and editing; K.D.K.: funding acquisition, methodology, resources, supervision and writing—review and editing; J.F.S.: conceptualization, data curation, funding acquisition, investigation, methodology, project administration, resources, supervision and writing—review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

The University of Pittsburgh and the Center for Adaptation to a Changing Environment (ACE) at ETH Zürich funded this research.