Skip to main content
NCBI home page
MicrobiologyOpen logoLink to MicrobiologyOpen
. 2022 Sep 28;11(5):e1318. doi: 10.1002/mbo3.1318

Enriching captivity conditions with natural elements does not prevent the loss of wild‐like gut microbiota but shapes its compositional variation in two small mammals

Adam Koziol 1, Iñaki Odriozola 1, Lasse Nyholm 1, Aoife Leonard 1, Carlos San José 2, Joana Pauperio 3, Clara Ferreira 4, Anders J Hansen 1, Ostaizka Aizpurua 1, M Thomas P Gilbert 1,5, Antton Alberdi 1,✉
PMCID: PMC9517064  PMID: 36314753

Abstract

As continued growth in gut microbiota studies in captive and model animals elucidates the importance of their role in host biology, further pursuit of how to retain a wild‐like microbial community is becoming increasingly important to obtain representative results from captive animals. In this study, we assessed how the gut microbiota of two wild‐caught small mammals, namely Crocidura russula (Eulipotyphla, insectivore) and Apodemus sylvaticus (Rodentia, omnivore), changed when bringing them into captivity. We analyzed fecal samples of 15 A. sylvaticus and 21 C. russula, immediately after bringing them into captivity and 5 weeks later, spread over two housing treatments: a “natural” setup enriched with elements freshly collected from nature and a “laboratory” setup with sterile artificial elements. Through sequencing of the V3–V4 region of the 16S recombinant RNA gene, we found that the initial microbial diversity dropped during captivity in both species, regardless of treatment. Community composition underwent a change of similar magnitude in both species and under both treatments. However, we did observe that the temporal development of the gut microbiome took different trajectories (i.e., changed in different directions) under different treatments, particularly in C. russula, suggesting that C. russula may be more susceptible to environmental change. The results of this experiment do not support the use of microbially enriched environments to retain wild‐like microbial diversities and compositions, yet show that specific housing conditions can significantly affect the drift of microbial communities under captivity.

Keywords: 16S, captivity, diversity loss, gut‐microbiome, host‐associated microbiota, non‐model organism

We assessed how the gut microbiota of two wild‐caught small mammals, Crocidura russula (greater white‐toothed shrew, Eulipotyphla, insectivore) and Apodemus sylvaticus (wood mouse, Rodentia, omnivore), changed when bringing them into captivity and housing them under different conditions. Our results showed that gut microbial diversity was significantly reduced when bringing wild animals into captivity regardless of housing conditions. However, while diversity was lost, the temporal development of the gut microbiome was very different between treatments, particularly in C. russula.

graphic file with name MBO3-11-e1318-g003.jpg

1. INTRODUCTION

The study of host‐microbiota interactions has become integral to our understanding of animal health, ecology, and evolution (Nyholm et al., 2020). Due to the complexity of gut microbial communities, and their sensitivity to environmental factors, captivity experiments (both using laboratory and wild animals) have proven essential for detecting and measuring the detailed interactions between animals and microorganisms (Hird, 2017; Rosshart et al., 2019; Shinohara et al., 2019). Such set‐ups enable the controlling and limiting of experimental factors that may influence the measured outcome, including host genetic variation (Bonder et al., 2016), developmental stage (Arrieta et al., 2014), and social interactions (Raulo et al., 2021), as well as environmental factors such as temperature (Sepulveda & Moeller, 2020), humidity (Rosenbaum et al., 2009) and diet (Bibbò et al., 2016; Martínez‐Mota et al., 2020; Maurice et al. 2015; Morrison et al., 2020). However, simplified captive environments can also modify the microbiota in a variety of ways, by reducing its diversity in comparison to that of wild communities, or recruiting new bacteria that are not found in wild populations (Alberdi et al., 2021). Such changes can differ significantly across species (Kohl et al., 2014), and might also decouple the optimal animal‐microbiota balance dropping host fitness (Rosshart et al., 2017). Hence, adequate assessment of all experimental variables relating to gut microbiota dynamics is important, as deviations from healthy biotic states can lead to erroneous experimental outcomes (Beura et al., 2016; Kinross et al., 2011).

In light of such limitations and biases, researchers are actively seeking strategies to firstly maintain the original gut microbial communities of wild animals once they have been moved to captivity, and secondly, modify the gut microbiota of laboratory animals to resemble that of their wild counterparts (Rosshart et al., 2019). Since the diet is one of the factors that conditions gut microbial communities, attempts have been made to employ dietary interventions to achieve these goals (Martínez‐Mota et al., 2020). An alternative strategy that has been explored is the introduction of microbes through non‐dietary related environmental sources, with some studies demonstrating there can be a significant positive effect on the gut microbiome (Liu et al., 2021; Weinstein et al., 2021; Zhou et al., 2016). While the inclusion of environmental microbes in captivity experiments has been assessed to have positive outcomes, no studies have addressed this in the context of a management tool, thus making it important to assess the value of microbially enriching the environment used in captivity experiments.

To explore a new potential way to help captive animals retain wild‐like gut microbiotas, we studied whether enriching captivity housing conditions with natural elements (while maintaining diet as a constant) contributes to the retention of the original (precaptivity) gut microbial community, as proxied by fecal samples, of animals captured in the wild. We carried out our experiment on two widespread non‐model small mammals with differing evolutionary history and ecology: the European wood mouse (Apodemus sylvaticus—AS, order Rodentia, omnivorous diet) and the greater white‐toothed shrew (Crocidura russula—CR, order Eulipotyphla, insectivorous diet). The animals were kept in captivity for 5 weeks under two different treatments: a “Natural” setup containing enrichment elements freshly collected from nature, and a “Laboratory” setup containing artificial enrichment elements. We analyzed variations in the gut microbiota from various perspectives: (i) the change in alpha diversity to assess if nature‐enriched conditions contributed to maintaining wild‐like gut microbial diversity; (ii) the change in beta diversity between time points and within individuals to explore whether nature‐like conditions maintained a composition more similar to the original wild‐like community, and (iii) the interacting effects of treatment and time on bacterial community composition to explore if the community changed in different directions over time (i.e., if the microbial community took different temporal trajectories) under the contrasting captivity conditions. Using the Hill numbers framework, we calculated neutral and phylogenetic diversity and dissimilarity indices at multiple orders of diversity (Chao et al., 2014). This allowed us to disentangle the contribution of closely versus distantly related bacteria and rare versus common bacteria to the variations between time points and treatments for each host species.

2. METHODS

2.1. Animal trapping and collection

Adult AS and CR were collected across the Northern Iberian Peninsula, Europe (43.2 N, 2.2 W), from June to August 2019 (due to trapping success) over 11 field sites. Animals were trapped using Sherman traps over 3 days at each site and checked every 12 h. Traps were cleaned between locations and the baits used were a mixture of oats and tuna and a small wedge of apple. Upon successful detection, each animal was transferred into a plastic bag for species and sex identification. Maturity of the animal was confirmed with morphometrics (e.g., body weight and length) and pelage for both species, any individuals which did not meet adult criteria or which were pregnant/lactating were excluded and released. Individuals were then individually placed into a small, microisolator cage for transfer to the ZIBA animal experimentation facilities in Zarautz, Basque Country, Spain.

2.2. Processing and identification of shrews and mice

Before experimental inclusion, animals were checked for any signs of serious distress or ailment, and if deemed healthy to continue, an initial fecal sample (~50 mg) was collected upon arrival at the experimentation facility. Each animal was then anesthetized over a heated mat using 2% isoflurane, to allow the subcutaneous injection of a Mini HPT10 radio frequency identification chip (Biomark) into the nape of the neck for subsequent individual identification. Each individual was monitored for 5 minutes for any adverse effects before being transported into the corresponding housing enclosure.

2.3. Housing conditions and experimental design

Animals were cohoused with conspecifics of the same sex in groups of 4–5 individuals in 840 cm2 polycarbonate cages (Unno Type III, 38.2 × 22.0 cm). Cages were randomly assigned to two different environmental enrichment conditions, either Natural conditions (herein NC) or Laboratory conditions (herein LC). Although the cages used in both conditions contained similar three‐dimensional enrichment structures, the structures themselves were created using either natural or artificial elements, respectively (Appendix A: Figure A1). NC involved the use of natural elements freshly collected from the habitats in which the animals were trapped; specifically, the soil was used as bedding, moss was provided as nesting material, and sticks and stones were used as enrichment elements. All enrichment materials for NC were collected from natural areas of low human encroachment and hence are unlikely to consist of any human features. LC included paper and wood bedding, cotton as nesting material, and 3D‐printed plastic sticks and stones as enrichment elements. For both treatments, cages were cleaned and materials replaced or sterilized every week. In NC soil, nesting and enrichment items were replaced with fresh materials, while in LC fresh bedding and nesting materials were added and the enrichment materials cleaned and sterilized. With regard to diet, AS were fed a standard chow diet, while CR were fed a feed containing rice and chicken. Both diets were maintained unchanged across the whole experimental period. Animals were kept under a strict 12 h night and day cycle, and routine cage cleaning was performed each week (replacement of bedding and nesting material), access to food was ad libitum, and food was changed daily. Environmental conditions were kept constant with an average humidity of 70%, temperature of 22°C, and 60 revolutions of air per minute by keeping the animals in an HPP 750 LIFE climate controller chamber (Memmert).

2.4. Fecal collection

We sampled from AS (n = 15) and CR (n = 21) housed in four and five cages, respectively. The experiment was sex‐biased towards male individuals (n AS = 13, n CR = 13), due to uneven capture success. Animals were split into NC (n AS = 9, n CR = 12, cagesAS = 2, cagesCR = 3) and LC (n AS =  6, n CR = 9, cagesAS = 2, cagesCR = 2) housing treatments. Fresh feces were collected from each individual immediately upon arrival at Time point 0 (herein T0; approximately 30 min –1 h after arrival to the laboratory) and day 35 (herein T1). To do so, animals were isolated into a separate sterile housing container and upon defecation, the fecal pellets (~50 mg) were collected and stored in 500 μl of DNA/RNA shield (Zymo), left at room temperature for 1 hour, and then transferred to −20°C for long‐term storage until DNA extraction.

2.5. DNA extraction and metabarcoding

DNA was extracted using a Zymo QuickDNA Fecal/Soil Microbe 96 kit (Zymo) according to manufacturer's guidelines, eluted in 50 μl of elution buffer, and immediately stored at −20°C. This involved an initial quality check for DNA concentration using a Tapestation high sensitivity kit (Agilent). Immediately after, amplification of the V3–V4 region of the 16S recombinant RNA (rRNA) gene was performed using the primers 341F:ACTCCTACGGGAGGCAGCAG (Herlemann et al., 2011) and 806R:GGACTACHVGGGTWTCTAAT (Takai & Horikoshi, 2000) using fusion tags with unique indices for downstream identification. PCR was performed in a total volume of 50 μl consisting of 25 μl of NEB Phusion® high‐fidelity PCR master mix, 4 μl of reverse and forward fusion tag primers, 30 ng of DNA extract, and ddH20 up to 50 μl. PCR conditions consisted of an initial denaturation step of 98°C for 3 min, 30 cycles of denaturation at 98°C for 45 s, annealing at 55°C for 45 s, elongation at 72°C for 45 s, and lastly a final hold at 72°C for 7 min. After amplification, the PCR products were purified using Ampure beads (Agencourt) to remove small fragments and impurities. Samples were then quality and concentration checked by a Tapestation on a high sensitivity chip (Agilent) and pooled equimolar before sequencing 300 PE on a HiSeq. 2500 (Illumina) using services from BGI. Negative extraction controls were included throughout all stages of the process to control for cross‐contamination.

2.6. Bioinformatics and data analysis

Paired‐end reads were first demultiplexed on unique fusion tag combinations. Immediately following this we quality‐filtered the demultiplexed reads (Q > 20) using AdapterRemoval 2.3.1 (Schubert et al., 2016), and primers were removed using Cutadapt 2.10 (Martin, 2011). Low‐quality reads were removed or trimmed using the filterAndTrim function implemented in DADA2 (Callahan et al., 2016). Error pattern learning and denoising of the data set were also performed using the DADA2 algorithm using default parameters (Callahan et al., 2016). Chimera removal was then performed before the generation of an ASV table consisting of ASV read counts for each sample. Reads were abundance‐filtered across samples by a relative abundance of 0.01% to remove singletons and other reads that may exist due to sequencing or PCR artifacts. Taxonomy assignment was then performed by the naïve Bayesian classifier implemented in DADA2 against the SILVA 16S taxonomy database (v138). Alignment of ASV sequences was performed using Clustal Omega (Madeira et al., 2019) and subsequently, a phylogenetic tree was built in Iqtree (Minh et al., 2020). ASVs were filtered using the R package decontam (Davis et al., 2018) to detect relevant contaminants based on the prevalence algorithm.

2.7. Diversity and compositional modeling

Gut microbiota diversity and compositional analyses were based on the Hill numbers framework. Specifically, we computed both neutral and phylogenetic diversities of orders of diversity (q value) 0, 1, and 2 using the R package Hilldiv (Alberdi, 2019). Neutral metrics do not account for the degree of relatedness among ASVs, while phylogenetic metrics consider the phylogenetic correlations among ASVs when computing diversity. Differences between both dimensions of diversity metrics (neutral and phylogenetic) therefore provide insights into whether diversity variation is driven by phylogenetically close or distantly related taxa. The different orders of diversity assign different weights to the ASVs when computing diversity. A q value of 0 does not consider relative abundances but only the presence or absence of ASVs. At a q value of 1, ASVs are weighted according to their relative abundances. A q value of two overweighs abundant ASVs with respect to nonabundant ones. Comparisons between orders of diversity therefore yield information on how the evenness of ASV distribution within samples affects diversity estimation. Beta diversity between the two sampling time points (i.e., before and after the captivity period) was also measured in terms of Hill numbers by computing the Sørensen‐type turnover. Similarly, Sørensen‐type turnover derived from all sample pairs in the data set was used to assess the directional effect of treatment and time points in gut microbial composition (Alberdi & Gilbert, 2019; Chao et al., 2014).

Linear mixed‐effect models, as implemented in the R package nlme (Pinheiro et al., 2017), were employed to assess the change in alpha diversity and beta diversity in response to experimental treatments on the gut microbiota across all individuals. In total, eight linear mixed models (Table 1) were fitted for each combination of species (i.e., AS and CR), diversity metric (i.e., neutral and phylogenetic), and also by diversity scale (i.e., alpha and beta diversity metrics). For alpha diversity models we included as fixed explanatory variables the q value (categorical factor with three levels: “0,” “1,” and “2”), treatment (categorical factor with two levels: “natural” and “laboratory”), time (categorical factor with two levels: “T0” and “T1”) and their interactions. As several individuals were kept in each cage, and several diversity metrics were calculated from each sample, a random effect of the form “~1|Cage/Individual_ID/Sample_ID” was included in the models. Beta diversity was measured as the compositional change from T0 to T1 within each individual, hence, only treatment, q value, and their interaction were used as fixed factors and, a random effect of the form “~1|Cage/Individual_ID” was included. Linear mixed models were checked for assumptions of homoscedasticity and normality of residuals and, where assumptions were violated (e.g., alpha diversity metrics), the response variables were log‐transformed. Model complexity was reduced by dropping the nonsignificant interactions between the fixed effects using likelihood ratio tests between nested models. Regardless of their significance, all main effects as well as the random effects were retained in the models as structural parts of the experimental design.

Table 1.

Final linear mixed models

Model parameters
Model name Neutral/phylogenetic Alpha/beta Model equation
Crocidura1 Neutral Alpha diversity~time + treatment + q value, random = ~1|Cage/Mouse_ID/Sample_ID
Crocidura2 Phylogenetic Alpha diversity~time + treatment + q value, random = ~1|Cage/Mouse_ID/Sample_ID
Crocidura3 Neutral Beta dissimilarity~treatment:q value, random = ~1|Cage/Mouse_ID
Crocidura4 Phylogenetic Beta dissimilarity~treatment:q value, random = ~1|Cage/Mouse_ID
Apodemus1 Neutral Alpha diversity~time + treatment + q value, random = ~1|Cage/Mouse_ID/Sample_ID
Apodemus2 Phylogenetic Alpha diversity~time + treatment × q value, random = ~1|Cage/Mouse_ID/Sample_ID
Apodemus3 Neutral Beta dissimilarity~treatment:q value, random = ~1|Cage/Mouse_ID
Apodemus4 Phylogenetic Beta dissimilarity~treatment:q value, random = ~1|Cage/Mouse_ID
Open in a new tab

The temporal change in gut microbial composition (i.e., beta diversity between time points) may happen following independent trajectories in each individual, or directionally, following a specific trajectory across all individuals. To test the null hypothesis of no directional changes in microbiome composition from the transition of wild (T0) to day 35 (T1) we used PERMANOVA (Anderson, 2017) on pairwise dissimilarity matrices based on Sørensen‐type turnover (neutral and phylogenetic, and combining different q values) using the function adonis2 in the R package “vegan” (Oksanen et al., 2020). We fitted two PERMANOVA models per type of dissimilarity matrix, one for each host species with the form adonis2(microbiome ~  treatment × time, strata = Individual_ID). A significant treatment × time interaction would indicate that the enrichment with natural elements led to different temporal trajectories in community composition. The magnitude of effects was quantified using the adjusted R 2 and the microbiome composition visualized using NMDS (Kruskal, 1964).

To identify the bacterial genera most severely affected by captivity conditions in each housing condition, we analyzed the data through hierarchical modeling of species communities (HMSC) at the genus level (Warton et al., 2015), as implemented in the R package HMSC (Tikhonov et al., 2020). HMSC is a hierarchical model constructed in the generalized linear model framework using Bayesian inference. Four models were fitted, separately for each of the two species and the NC and LC. As the data were zero‐inflated, we applied a hurdle model (zero‐altered model) (Rose et al., 2006). This type of model consists of two parts, one modeling the presence‐absence of species and the other modeling abundance conditional on presence. To fit the first model, we transformed all nonzero values in the data set into one, to create a presence‐absence matrix. We applied a binomial model with a probit link function to each genus. The second model looks at abundances conditional on presences (scaled to mean zero and unit variance). We transformed zeros to missing values, and kept all nonzeros in their values, we then fitted the log‐normal model. Then, the two components of the model were fitted consecutively (Ovaskainen & Abrego, 2020). The analysis was restricted to the genera that were present in at least four samples within each treatment and host species, which resulted in 82 genera for AS NC models, 64 genera for AS LC models, 96 genera for CR NC models, and 63 genera for CR LC models. This stringent criterion was used as rare species lack adequate information for taxon‐specific modeling. As fixed explanatory variables in matrix X of HMSC, we included the categorical factor time, as well as the log‐transformed continuous variable of sequencing depth, which controlled for the variation in sequencing effort among samples. To account for the hierarchical study design, we included cage and individual ID random effects in the models. To examine whether the responses of the genera to time showed a phylogenetic signal, we included in the analysis a phylogenetic correlation matrix C among the genera, obtained as explained in the previous section. The phylogenetic signal is measured using the parameter ρ, which takes values from 0 to 1, a value of 0 meaning no phylogenetic signal in the response to time, and a value of 1 meaning a completely phylogenetically structured response to time. A significant positive (negative) association with T1 in the binomial model means that the genus has a higher (lower) probability of occurrence in T1. A significant association in the log‐normal model means that, when present, the genus is more (less) abundant in T1. The genera with a positive response to time in captivity with posterior probability of >0.9 were considered as significantly enriched in captivity. The genera with a negative response to time in captivity, with posterior probability <0.1 were considered as significantly enriched in nature. The posterior probability of >0.9 indicates that >90% of the parameter estimates of the posterior distribution are positive. The posterior probability of <0.1 indicates that <10% of the parameter estimates of the posterior distribution are positive (hence, >90% are negative). We fitted the models assuming the default priors and sampled the posterior distribution running four Markov Chain Monte Carlo (MCMC) chains, each of which was run for 37,500 iterations, of which 12,500 were discarded as burn‐in. We thinned by 100 to obtain a total of 250 posterior samples per chain and 1000 posterior samples in total. We ensured MCMC convergence by measuring the potential scale reduction factor (Tikhonov et al., 2020) for the beta parameters (measuring the response to time in captivity) and the ρ parameters (measuring phylogenetic signal in beta parameters).

3. RESULTS

We analyzed 72 fecal samples from 36 animals and four negative extraction controls to account for contamination. We generated 11,425,282 sequences (114,252 ± 42,356 per sample; mean and standard deviation, respectively) with a total of 6,570,726 (65,707 ± 21,936) sequences after quality filtering (for a full breakdown see Appendix A: Table A1). From these, 8176 unique amplicon sequence variants were generated (herein ASVs), which were assigned to 31 phyla, 68 classes, 142 orders, 226 families, and 427 genera (Figure 1). The 28 ASVs that were not assigned at least a bacterial Phylum annotation were removed from downstream analyses. Decontam detected 39 ASVs as contaminants, which were also removed from all parts of the analysis (Appendix A: Table A2). The ASV accumulation curves of all samples reached the asymptote, which confirmed sufficient sequencing depth to recover the complete microbial diversity (see Appendix A: Figure A2).

Figure 1.

Figure 1

Open in a new tab

(a) Radial tree of life of presence/absence data at the genus level across all treatments indicating community level differences between treatments for both Apodemus sylvaticus (AS) and Crocidura russula (CR). Circular rings disseminate between Phylum, T0, natural conditions, and laboratory conditions for both species. (b) Stacked bar plots of sample pairs (T0 & T1) representing relative abundance at the community composition at the phylum level for natural conditions (yellow bar) and laboratory conditions (blue bar).

3.1. Characterization of wild microbiomes across time

Wild‐caught AS harbored a microbial community consisting of 12 phyla spanning 3299 ASVs (Figure 1). The microbiota was principally dominated by Firmicutes (51.6 ± 21.5%) and Bacteroidota (36 ± 20.7%), followed by Proteobacteria (5.7 ± 7.6%). In contrast, CR harbored a gut microbiota that consisted of 30 phyla spanning 2328 ASVs. The microbial communities were dominated by Proteobacteria (62.3 ± 23.8%) and Firmicutes (30.8 ± 21%), followed by Actinobacteria (2.6 ± 10.6%). As shown by the large standard deviations around means, the initial microbiome composition was highly variable across individuals captured in the wild (Figure 2a).

Figure 2.

Figure 2

Open in a new tab

Treatment effects on neutral and phylogenetic Hill numbers calculated for alpha (A) and beta (B) diversity between treatments and time‐points. (a) Alpha diversity represented by the average difference from T0 to T1 for both Apodemus sylvaticus and Crocidura russula for each order of diversity and diversity dimension. (b) Beta diversity of data paired per individual animal for both A. sylvaticus (AS) and C. russula (CR) across the three orders of diversity and the two diversity dimensions.

3.2. Effects of captivity and housing treatments

Hill numbers were calculated for three orders of diversity (q values) and across both neutral and phylogenetic measures yielding varying numbers of effective ASVs for AS and CR (Figure 2a). We observed a reduction in total detected ASVs for AS (n = 2302, −30.2%) and CR (n = 1126, −51.6%) across time. Further, we did not detect a significant interaction between time and treatment indicating that diversity loss after time in captivity occurred similarly in individuals from either treatment (CRneutral: t 19  = −1.43, p = 0.16, CRphylogenetic: t 19  = −1.71, p = 0.10, ASneutral: t 13  = 0.11, p = 0.91, ASphylogenetic: t 13  = −0.56, p = 0.58). Similarly, treatment had no significant effect on alpha diversity measured with either neutral or phylogenetic diversity metrics (CRneutral: t 3  = −0.081, p = 0.941, CRphylogenetic: t 3  = 0.207, p = 0.847, ASneutral: t 2  = −0.274, p = 0.810, ASphylogenetic: t 2  = 0.165, p = 0.884). However, we did detect a significant reduction in the alpha diversity between both time‐points using neutral diversity metrics for AS (t 14  = −2.234, p = 0.042) and for CR (t 20  = −4.138, p < 0.001); whereas when using phylogenetic diversity metrics, the difference between both time‐points was significant for CR (t 20  = −4.861, p < 0.001), but not for AS (t 14  = −1.862, p = 0.084).

We then calculated beta diversity between time points (Figure 2b), and observed that neither treatment (ASneutral: t 2  = −0.284, p = 0.803, CRneutral: t 3  = 0.416, p = 0.706) nor q value (ASneutral q1: t 28 = −0.844, p = 0.406, ASneutral q2: t 28 = 1.557, p = 0.131, CRneutral q1: t 40 = −1.455, p = 0.154, CRneutral q2: t 40 = 0.289, p = 0.774) had a significant effect on the dissimilarity between time points for both AS and CR using neutral diversity indices. Likewise, when using phylogenetic diversity measures, we did not detect a significant effect of treatment on dissimilarity between both time points (ASphylogenetic: t 2  = −0.142, p = 0.900, CRphylogenetic: t 3  = 1.151, p = 0.333). Interestingly, however, we detected that q value had a significant effect on reducing dissimilarity between time points when using the phylogenetic diversities (Figure 2b) for both AS (ASphylogenetic q1: t 28 = −13.473, p > 0.001, ASphylogenetic q2: t 28 = −26.784, p > 0.001) and CR (CRphylogenetic q1: t 40 = −6.794, p > 0.001, CRphylogenetic q2: t 40 = −10.231, p > 0.001) with higher q values resulting in lower dissimilarity between time‐points (all linear model results can be found in Appendix A: Table A3 a‐h).

PERMANOVA analyses showed that while beta‐diversities were changing at similar rates and alpha diversities were showing similar decays, the composition between each treatment diverged after the captivity period (Figure 3). This separation was significant in CR as indicated by a significant interaction between treatment and time (Pseudo‐F 1  = 5.171, p = 0.012, R 2 = 0.07), however, the same interaction was not significant in AS (Pseudo‐F 1  = 1.297, p = 0.162, R 2 = 0.039), although some separation between treatments was visible after day 35 (Figure 3, Appendix A: Table A4).

Figure 3.

Figure 3

Open in a new tab

NMDS of community composition with 95% confidence intervals shaded in ellipses at Time point 1 (a, c): for A. sylvaticus and C. russula respectively representing the community composition of each individual immediately after being captured in the field (T0, day 1) and (b, d): 35 days later (T1) representing the community composition diverge between treatments of Natural (yellow) and Laboratory (blue).

We assessed the differential response of the most common genera to captivity in terms of probability of presence (binomial submodel of the Hurdle model) and log‐abundance conditional on the presence (lognormal submodel of the Hurdle model) (Table 2). We observed that 23% and 15% of the genera detected in AS showed a negative association with time in the binomial models in laboratory and natural conditions, respectively, whereas 8% and 9% of the genera showed positive associations (Figure 4). In the case of CR, 46% and 20% of common genera showed negative associations with time in the laboratory and natural conditions, respectively, while 25% and 30% showed positive associations. In AS abundance models, 8% and 24% of genera decreased and 6% and 1% increased in time, respectively under laboratory and natural conditions. In contrast, a predominance of negative associations was not as clear in CR abundance models: 11% and 14% of the genera showed a negative association with time whereas 8% and 14% showed positive associations, under laboratory and natural conditions. Of the few positive associations detected in the AS abundance models, we detected the proliferation of Rikenella (Phylum: Bacteriodota), Odoribacter (Phylum: Bacteriodota), Lachnoclostridium (Phylum: Firmicutes), and Candidatus_Saccharimonas ASVs which were solely detected in the LC treatment. In contrast, most positive associations in CR were with respect to genera belonging to Firmicutes and most negative associations with Bacteroidetes across both treatments. Additionally, we found some evidence that the NC housing conditions mitigated the loss of genus level diversity found at T0 which was not found in the T1 sample from the LC treatment (Figures 1a and 4), especially with CR. Loss of genera was not observed in the NC treatment which maintained all the detected genera from their initial day 1 sample (Figure 4), albeit with reduced alpha diversities and loss of ASVs. Moreover, these associations showed phylogenetic structure in both species as determined by the ρ values calculated by the HMSC model (see Table 2).

Table 2.

HMSC results indicating the specific model used as well as the phylogenetic signal (ρ value) in species responses to time in captivity; higher ρ values indicate a higher phylogenetic signal and the 90% credible intervals not overlapping zero indicate strong evidence for its significance

Species Treatment Model ρ Value 90% credible interval
Apodemus sylvaticus NC Binomial 0.68 [0, 0.98]
Apodemus sylvaticus LC Binomial 0.22 [0, 0.89]
Apodemus sylvaticus NC Lognormal 0.79 [0.08, 0.99]
Apodemus sylvaticus LC Lognormal 0.81 [0, 0.99]
Crocidura russula NC Binomial 0.9 [0.64, 1]
Crocidura russula LC Binomial 0.66 [0, 0.96]
Crocidura russula NC Lognormal 0.76 [0.08, 0.98]
Crocidura russula LC Lognormal 0.59 [0, 0.97]
Open in a new tab

Figure 4.

Figure 4

Open in a new tab

HMSC analysis showing the response of the most common genera to time in captivity, for (a) Apodemus sylvaticus and (b) Crocidura russula. Statistical support is provided in columns for both the presence‐absence/occurrence (binomial model) and abundance conditional on presence (lognormal model) models for natural conditions and laboratory conditions. Statistical support of >0.9 (red bars) was considered a significant increase of an ASV after 35 days; statistical support of <0.1 (blue bars) represents a significant decrease of an ASV; statistical support between 0.1 and 0.9 (white bars) was considered as not significantly affected by time in captivity (grey bars) represent depleted ASVs which were lost after 35 days. See the methods section for more details on how statistical support is interpreted in the Bayesian context.

4. DISCUSSION

Continued work in assessing the role of the gut microbiota on host fitness has demonstrated that the maintenance of a biologically optimal microbiota offers many benefits, not only to the host but also to the representativeness and translatability of research (Hauffe & Barelli, 2019; Hird, 2017). In the search for useful management practices to retain wild‐like gut microbial communities in animals taken into captivity, we measured the impact of two housing treatments in two small mammals and quantified several features of their associated microbial communities. We observed that increased exposure to environmental microorganisms did not significantly prevent gut microbiota diversity loss when compared to conventional experimental housing setups. Instead, our results showed that both treatments resulted in a reduction and restructuring of the gut microbiota community. However, the microbial composition changed in different directions between treatments in CR, indicating that microbial trajectories may be influenced by environmental factors in a species‐specific manner.

The relevance of environmental microbes is an important facet to consider when measuring host fitness as microbiota dysbiosis has been associated with factors that might reduce research translatability, such as immune function (Fujimura et al., 2014; Schuijt et al., 2016; Zhang et al., 2021), metabolism (Fan & Pedersen, 2021; Raymann & Moran, 2018; Sommer et al., 2016) and behavior (Davidson et al., 2020; Raulo et al., 2021; Singh et al., 2019). However, in our experiment, housing animals in a (semi) sterile environment or an environment enriched with natural elements had a similar effect on microbial alpha diversity patterns. This observation is in contrast with previous studies, which demonstrated that higher diversity in the environment does yield higher complexity in the host subsystem (Sbihi et al., 2019; Zhao et al., 2020; Zhou et al., 2016). Contrasting our studies, the main difference is likely due to the developmental stage of the analyzed animals, and the associated maturity of their gut microbial communities (Beura et al., 2016; Liu et al., 2021; Sbihi et al., 2019; Zhou et al., 2016). Our study included adults, which were shown to host a diverse microbial community before being introduced into captivity. In doing so, the starting gut composition was significantly more diverse compared to early‐life conspecifics (Nemergut et al., 2013). As such, the ecological niches within the adult gut‐microbiomes were likely already occupied (Langille et al., 2014; Turnbaugh et al., 2009) leading to competitive exclusion from the existing bacterial community (Baumgartner et al., 2021; Zmora et al., 2018). Thus, the lability of juvenile gut microbiomes may further promote the uptake of passively acquired environmental bacteria (Liu et al., 2021), while such a mode of acquisition seems to be negligible in adulthood, as suggested by the overall diversity loss in our experiment.

Despite the limited effect of the tested treatments at mitigating diversity loss, we did observe significant gut microbiota variation between time points in both species. Regarding alpha diversity, the observed patterns were somehow alike in the two species, even though the microbial communities associated with wild animals were radically different between AS and CR. Previous studies have demonstrated that captivity itself can significantly alter the gut microbiome of many species and may be due to a myriad of reasons, including access to nutrients, and changes in ambient temperatures/humidity (Sepulveda & Moeller 2020; Nicholls et al., 2016; Rosenbaum et al., 2009), diet (Bibbò et al., 2016; Martínez‐Mota et al., 2020; Morrison et al., 2020), which are all readily manipulated when entering captivity conditions and may also be confounded by evolutionary histories as species' responses to captivity can be host‐specific (Alberdi et al., 2021; Weinstein et al., 2021). Moreover, environmental complexity found in wild environments is not only a constituent of changes to the soil and physical surroundings but additionally to the many available bacteria in changing diets and water sources that were not tested in this experiment (Nyholm et al., 2022). As such, changing these conditions can lead to a wide range of responses, with studies reporting both significant microbial diversity increases and decreases in host responses to captivity (see [Alberdi et al., 2021] for further discussion).

Unlike overall diversity, the compositional response of the gut microbiotas to captivity and experimental treatments differed between species, which suggests that environmental access to microbes may have some effect in influencing the trajectory of the gut microbiome across different species. On the one hand, regardless of the treatment, we observed a significantly higher microbiota turnover in CR than in AS. By layering the diversity metrics at different orders of diversity with the phylogenetic information of each ASV, we were able to detect that in AS the main phylogenetic groups remained stable (Figure 2b). This indicates that when only using neutral diversity metrics, studies may omit valuable information as to whether the community changes are biologically meaningful signals or not. On the other hand, the CR gut microbiome demonstrated a clear interaction between time and treatment and separation of microbial composition based on housing conditions, which was not detected in AS. The lack of information on the microbial communities present in the natural elements prevented us to ascertain whether the observed variation was directly produced by the acquisition of environmental bacteria. However, the differences observed between AS and CR indicate that the response to the environment is likely to be highly specific‐specific and that many host‐specific factors may impact microbial sensitivity to environmental changes. We explored each host's response to the environment using HMSC, finding significant changes in the gut microbiota of specific genera. In most cases, genus‐level associations with time were either neutral or negative suggesting that the genera which responded to time in captivity were more likely to respond negatively, that is, decreasing abundances. Despite this, we did observe that some taxa proliferated under captive conditions. In AS, while only speculative, the proliferation of Odoribacter (Hiippala et al., 2020) has been known to increase propionate production, while additionally Rikinella has been known to increase lipid metabolism and energy regulation (Gálvez‐Ontiveros et al., 2020), potentially leading to positive effects on the gut health of AS. The severe drop in compositional turnover when increasing the order of diversity in phylogenetic metrics indicated that the replacement of ASVs mostly stemmed from closely related ASVs, rather than distant taxonomic groups. In contrast, we detected a highly significant, strong correlation with phylogeny in the CR data suggesting phylogenetically related genera were responding in the same way. Principally, both the presence and abundance of Proteobacteria genera significantly reduced, while bacteria within Firmicutes significantly increased in relative abundance (Figure 4b). A similar trend was found in Suncus murinus (Family: Soricidae), where domesticated individuals showed a significant reduction in the prevalence of Proteobacteria with the replacement of Firmicutes (Shinohara et al., 2019). These changes were hypothesized to have an important role in lactic acid fermentation and digestion of novel food types (Shinohara et al., 2019).

5. CONCLUSIONS

Our study showed that introducing natural elements into captivity conditions did not mitigate diversity loss in either species. Hence, considering the additional logistical burden, for example, time/resources spent in collecting housing materials, cleaning, and movement of materials compared to conventional materials which can be readily acquired, our results do not support the use of microbially enriched environments to retain wild‐like microbiotas in captivity. However, we observed that the natural elements triggered different compositional changes in different host species. Thus, implementing appropriate experimental caution through the use of pilot studies may be important when determining the suitability of microbially enriched environments for different species (Teijlingen & Hundley, 2002). Ultimately, our study shows that enriching captivity housing conditions with natural elements can shape the trajectories of microbiota variation and that this can happen in a species‐specific manner.

AUTHOR CONTRIBUTIONS

Adam Koziol: Conceptualization (lead); formal analysis (equal); investigation (lead); methodology (lead); visualization (equal); writing–original draft (lead); Writing–review & editing (lead). Iñaki Odriozola: Data curation (equal); formal analysis (equal); methodology (supporting); supervision (equal); validation (equal); visualization (equal); writing–original draft (supporting); writing–review & editing (supporting). Lasse Nyholm: Investigation (supporting); methodology (supporting); validation (supporting); writing–original draft (supporting); writing–review & editing (supporting). Aoife Leonard: Data curation (supporting); formal analysis (supporting); validation (supporting); visualization (supporting); writing–review & editing (supporting). Carlos San José: Conceptualization (supporting); investigation (supporting); methodology (supporting); project administration (supporting); supervision (supporting); writing–review & editing (supporting). Joana Pauperio: Conceptualization (supporting); investigation (supporting); methodology (supporting); writing–review & editing (supporting). Clara Ferreira: Conceptualization (supporting); data curation (supporting); investigation (supporting); methodology (supporting); resources (supporting); writing–review & editing (supporting). Anders J. Hansen: Conceptualization (supporting); funding acquisition (supporting); methodology (supporting); project administration (supporting); supervision (supporting). Ostaizka Aizpurua: Conceptualization (supporting); formal analysis (supporting); investigation (supporting); methodology (supporting); project administration (supporting); supervision (supporting); writing–review & editing (supporting). M. Thomas P. Gilbert: Conceptualization (equal); funding acquisition (supporting); project administration (supporting); supervision (supporting); writing–original draft (supporting); writing–review & editing (supporting). Antton Alberdi: Conceptualization (lead); formal analysis (equal); funding acquisition (lead); investigation (supporting); methodology (lead); resources (equal); visualization (supporting); writing–original draft (supporting); writing–review & editing (equal).

CONFLICT OF INTEREST

None declared.

ETHICS STATEMENT

All animal captures and captivity experiments were approved by the Regional Government of Gipuzkoa under license codes PRO‐AE‐SS‐206 and PRO‐AE‐SS‐168. All experiments were performed following the agreed‐upon guidelines and regulations.

ACKNOWLEDGMENTS

This study was funded by the Lundbeckfonden grant R250‐2017‐1351 awarded to AA and the Danish National Research Foundation grant DNRF143 awarded to MTPG. We would also like to thank everyone at the Center of Evolutionary Hologenomics at the University of Copenhagen for their continued help and support. In particular, Nuno Martins, Linett Rasmussen, Marta Ciucani, Sofia Marcos, Christina Lyngaard, and Sarah Mak. We are also grateful to the people associated with the ZIBA Experimentation Center, in particular to Martxel Aizpurua for his continuous support, and students Andoni Aguirrezabala and Lander Olasagasti for their help in the fieldwork. This project has received funding from the European Union's Horizon 2020 research and innovation program under the Marie Skłodowska‐Curie grant agreement No 801199. We would also like to thank the University of Maryland Center for Environmental Science (http://ian.umces.edu/symbols/) for the use of their vector images. This study was performed following the ARRIVE guidelines, however, due to variation in capture success of wild individuals we were unable to have even sample sizes across treatments and species.

APPENDIX A.

See Figures A1 and A2 and Tables A1, A2, A3, A4.

Figure A1.

Figure A1

Open in a new tab

Examples of housing conditions (a) natural conditions and (b) laboratory conditions

Figure A2.

Figure A2

Open in a new tab

Species accumulation curves across all samples. All samples reached species asymptote.

Table A1.

Quality filtering of raw sequence reads for each sample from initial reads to final ASVs

Sample ID Number of initial reads Average forward read length Average reverse read length Primer‐trimmed reads Filtered reads Dereplicated reads ASVs before chimera filtering ASVs after chimera filtering Reads represented by ASVs
M10P11 96,790 295 275 96,760 89,468 37,739 5751 716 45,162
M10P51 91,760 296 277 91,728 85,274 32,003 3833 437 48,835
M11P11 106,689 295 270 106,655 99,491 36,867 3750 493 62,167
M11P51 88,216 293 271 88,186 78,282 31,903 3964 380 42,999
M12P11 96,046 297 274 96,030 91,282 37,884 5063 649 46,713
M12P51 87,990 297 276 87,967 83,048 29,080 3819 441 45,975
M13P11 117,929 298 283 117,888 112,196 30,464 6521 509 63,966
M13P51 118,706 291 285 118,636 100,215 26,278 6288 291 50,967
M14P11 80,167 292 284 80,153 72,480 23,798 4192 459 39,174
M14P51 138,320 294 285 138,252 126,374 39,586 3690 519 81,349
M15P11 170,094 292 269 169,999 145,340 67,001 5637 633 73,759
M15P51 144,609 299 270 144,572 137,022 54,434 4957 479 78,276
M1P11 99,687 295 258 99,650 92,718 42,457 3144 414 45,971
M1P51 95,687 293 260 95,655 86,242 34,400 4601 434 46,630
M2P11 98,285 298 281 98,257 91,386 31,780 5562 512 48,261
M2P51 126,529 292 270 126,432 106,270 38,399 5108 527 55,070
M3P11 155,555 292 259 155,460 134,782 58,773 4303 654 75,001
M3P51 195,103 299 258 195,059 185,803 73,456 7481 540 96,636
M4P11 171,195 292 268 171,140 148,959 30,904 5403 361 92,218
M4P51 126,073 295 268 126,031 116,108 45,108 5863 486 58,233
M5P11 184,764 298 269 184,716 175,335 68,721 12584 678 86,756
M5P51 99,766 298 271 99,732 94,799 28,252 3766 325 63,784
M6P11 125,702 296 269 125,655 117,567 53,559 4439 752 61,612
M6P51 162,691 297 271 162,632 154,614 35,780 4770 492 97,196
M7P11 105,769 298 271 105,725 100,303 39,491 3219 623 59,877
M7P51 136,513 299 273 136,456 130,339 32,911 7216 405 78,354
M8P11 173,262 297 270 173,210 164,496 65,348 9259 678 86,913
M8P51 88,098 295 270 88,062 82,147 32,280 4994 354 41,132
M9P11 95,276 292 277 95,233 80,071 36,244 2136 627 45,794
M9P51 104,545 293 277 104,519 91,983 29,223 3406 421 55,966
NC11 10,623 295 287 10,616 9933 1344 79 77 9722
NC21 2261 296 286 2260 2141 331 35 32 2057
NC31 2458 297 287 2456 2281 405 38 36 2182
NC41 2636 298 286 2636 2482 429 52 52 2363
S10P11 98,029 299 273 98,011 93,532 16,021 770 353 81,163
S10P51 99,296 298 269 99,279 94,523 15,421 2096 117 71,240
S11P11 109,643 296 284 109,625 104,145 14,293 1100 123 87,173
S11P51 88,572 297 284 88,563 83,145 15,096 1814 138 64,941
S12P11 200,073 297 272 200,029 189,851 61,947 13,792 711 98,834
S12P51 75,606 298 288 75,582 71,158 11,862 982 81 56,764
S13P11 123,606 295 271 123,562 115,738 24,895 4895 296 77,626
S13P51 132,630 298 287 132,600 124,585 30,115 5390 291 74,186
S14P11 206,199 291 284 206,100 174,331 45,232 10,733 482 95,321
S14P51 79,919 292 288 79,905 70,793 9971 573 57 62,209
S15P11 251,040 295 271 250,978 235,252 57,311 14,865 545 133,440
S15P51 81,859 294 287 81,837 75,538 9764 594 79 67,201
S16P11 132,078 293 272 132,056 122,795 19,900 2258 208 97,032
S16P51 93,866 295 285 93,837 87,783 19,051 2207 149 64,667
S17P11 158,046 296 285 158,007 148,823 39,037 7275 448 89,183
S17P51 147,398 297 287 147,359 139,227 35,712 6737 333 81,508
S18P11 142,135 292 272 142,072 123,741 25,602 5623 296 78,791
S18P51 66,957 299 266 66,942 63,922 8673 779 90 55,636
S19P11 125,308 299 283 125,279 120,485 20,238 2821 256 96,684
S19P51 97,326 298 284 97,280 90,831 15,987 2763 189 67,037
S21P11 155,087 299 258 155,047 148,938 34,548 7973 317 90,833
S21P51 181,482 298 259 181,432 173,139 40,657 7781 260 102,983
S22P11 99,774 291 282 99,683 80,449 13,890 2940 250 52,211
S22P51 134,685 292 284 134,630 117103 14474 3279 138 82610
S23P11 158,933 297 259 158,888 151,108 35,063 7768 276 89,170
S23P51 75,563 296 262 75,527 71,096 13,675 1652 84 53,711
S24P11 127,873 294 282 127,819 117,532 16,927 4691 277 81,625
S24P51 115,895 295 283 115,832 105,360 8103 718 86 95,863
S25P11 129,221 296 282 129,171 120,091 15,379 3513 230 86,775
S25P51 112,446 297 284 112,399 104,109 6841 451 57 95,201
S2P11 191,410 298 289 191,359 181,346 50,008 10,071 550 98,420
S2P51 101,890 291 288 101,842 85,542 14,052 1706 101 66,295
S3P11 76,301 293 290 76,274 69,598 15,010 851 454 58,783
S3P51 92,025 294 289 92,005 83,943 19,487 1835 344 63,776
S4P11 89,161 296 288 89,149 83,445 19,491 1517 432 64,439
S4P51 84,771 297 290 84,750 79,243 13,958 1863 82 58,796
S5P11 81,843 298 289 81,829 76,822 11,537 810 103 66,896
S5P51 85,355 298 286 85,342 79,864 19,654 2223 223 55,377
S6P11 122,178 291 283 122,116 100,890 20,432 2838 240 73,663
S6P51 98,257 292 286 98,241 86,641 24,916 2417 366 61,319
S9P11 132,306 294 283 132,271 119,276 40,263 4381 700 55,538
S9P51 80,321 295 285 80,304 73,721 14,730 1684 125 56,233
Open in a new tab

Table A2.

ASVs detected by decontam—all ASVs were removed

ASVs Kingdom Phylum Class Order Family Genus
ASV_38 Bacteria Proteobacteria Gammaproteobacteria Vibrionales Vibrionaceae Vibrio
ASV_126 Bacteria Proteobacteria Gammaproteobacteria Burkholderiales Comamonadaceae Delftia
ASV_486 Bacteria Firmicutes Clostridia Lachnospirales Lachnospiraceae Cellulosilyticum
ASV_803 Bacteria Proteobacteria Gammaproteobacteria Enterobacterales Morganellaceae Morganella
ASV_829 Bacteria Firmicutes Clostridia Clostridiales Clostridiaceae Clostridium_sensu_stricto_13
ASV_875 Bacteria Firmicutes Clostridia Lachnospirales Lachnospiraceae Herbinix
ASV_1023 Bacteria Actinobacteriota Actinobacteria Corynebacteriales Dietziaceae Dietzia
ASV_1207 Bacteria Firmicutes Clostridia Clostridiales Clostridiaceae Clostridium_sensu_stricto_13
ASV_1318 Bacteria Proteobacteria Gammaproteobacteria Pasteurellales Pasteurellaceae NA
ASV_1370 Bacteria Firmicutes Clostridia Lachnospirales Lachnospiraceae Epulopiscium
ASV_2014 Bacteria Firmicutes Clostridia Lachnospirales Lachnospiraceae Anaerosporobacter
ASV_2087 Bacteria Firmicutes Clostridia Lachnospirales Lachnospiraceae Lachnospiraceae_NK4A136_group
ASV_2470 Bacteria Firmicutes Clostridia Clostridiales Clostridiaceae Clostridium_sensu_stricto_1
ASV_2502 Bacteria Proteobacteria Gammaproteobacteria Burkholderiales Comamonadaceae Aquabacterium
ASV_2556 Bacteria Firmicutes Clostridia Lachnospirales Lachnospiraceae Lachnospiraceae_NK4A136_group
ASV_2913 Bacteria Actinobacteriota Actinobacteria Corynebacteriales Corynebacteriaceae Corynebacterium
ASV_2959 Bacteria Firmicutes Bacilli Staphylococcales Staphylococcaceae Jeotgalicoccus
ASV_3004 Bacteria Firmicutes Clostridia Lachnospirales Lachnospiraceae Lachnospiraceae_NK4A136_group
ASV_3034 Bacteria Proteobacteria Gammaproteobacteria Pseudomonadales Moraxellaceae Acinetobacter
ASV_3077 Bacteria Firmicutes Bacilli Erysipelotrichales Erysipelotrichaceae Turicibacter
ASV_3220 Bacteria Firmicutes Bacilli Staphylococcales Gemellaceae Gemella
ASV_3388 Bacteria Proteobacteria Gammaproteobacteria Burkholderiales Oxalobacteraceae Noviherbaspirillum
ASV_3691 Bacteria Proteobacteria Alphaproteobacteria Rhizobiales Rhizobiaceae Allorhizobium‐Neorhizobium‐Pararhizobium‐Rhizobium
ASV_3732 Bacteria Firmicutes Bacilli Lactobacillales Aerococcaceae Aerococcus
ASV_4125 Bacteria Proteobacteria Gammaproteobacteria Pseudomonadales Pseudomonadaceae Pseudomonas
ASV_4126 Bacteria Proteobacteria Alphaproteobacteria Sphingomonadales Sphingomonadaceae Sphingomonas
ASV_4238 Bacteria Proteobacteria Gammaproteobacteria Burkholderiales Oxalobacteraceae Noviherbaspirillum
ASV_4771 Bacteria Actinobacteriota Actinobacteria Micrococcales Micrococcaceae Renibacterium
ASV_4846 Bacteria Proteobacteria Gammaproteobacteria Burkholderiales Oxalobacteraceae Massilia
ASV_4930 Bacteria Proteobacteria Gammaproteobacteria Burkholderiales Comamonadaceae Acidovorax
ASV_5194 Bacteria Proteobacteria Alphaproteobacteria Azospirillales Azospirillaceae Azospirillum
ASV_5676 Bacteria Proteobacteria Gammaproteobacteria Pseudomonadales Moraxellaceae Enhydrobacter
ASV_5975 Bacteria Firmicutes Bacilli Staphylococcales Staphylococcaceae Staphylococcus
ASV_6063 Bacteria Firmicutes Clostridia Clostridia_vadinBB60_group NA NA
ASV_6312 Bacteria Proteobacteria Alphaproteobacteria Azospirillales Azospirillaceae NA
ASV_6315 Bacteria Proteobacteria Alphaproteobacteria Caulobacterales Caulobacteraceae Brevundimonas
ASV_6507 Bacteria Bacteroidota Bacteroidia Flavobacteriales Weeksellaceae Chryseobacterium
ASV_7211 Bacteria Proteobacteria Gammaproteobacteria Enterobacterales Enterobacteriaceae NA
ASV_7452 Bacteria Firmicutes Clostridia Lachnospirales Lachnospiraceae NA
Open in a new tab

Table A3.

a–h: Linear mixed model results for each model

Value Std. error df t Value p Value
a. Crocidura neutral—beta diversity
(Intercept) 0.8689446 0.04057842 40 21.390725 0.000
TreatmentLC 0.0242597 0.05860281 3 0.415690 0.7056
qvalueq. 1 −0.037708 0.02566214 40 −1.454951 0.1535
qvalueq. 2 0.0070877 0.02566214 40 0.289183 0.7739
b. Crocidura phylogenetic—beta diversity
(Intercept) 0.8420088 0.03970625 40 21.205950 0
TreatmentLC 0.0582521 0.05058904 3 1.151476 0.333
qvalueq. 1 −0.2577344 0.03793803 40 −6.793564 0
qvalueq. 2 −0.3881623 0.03793803 40 −10.231485 0
c. Apodemus neutral—beta diversity
(Intercept) 0.7585985 0.04716536 28 16.083807 0
TreatmentLC −0.0192982 0.06797048 2 −0.283920 0.8032
qvalueq. 1 −0.0283791 0.03361145 28 −0.844327 0.4056
qvalueq. 2 0.0523375 0.03361145 28 1.557134 0.1307
d. Apodemus phylogenetic—beta diversity
(Intercept) 0.7407412 0.02745100 28 26.984119 0
TreatmentLC −0.0053725 0.03792140 2 −0.141674 0.9003
qvalueq. 1 −0.3116469 0.02313081 28 −13.473241 0
qvalueq. 2 −0.6195252 0.02313081 28 −26.783553 0
e. Crocidura neutral—alpha diversity
(Intercept) 5.394585 0.19493481 82 27.67379 0
FirstorSecondT2 −0.822103 0.19865859 20 −4.13827 0.0005
TreatmentLC −0.020447 0.25226285 3 −0.08105 0.9405
qvalueq. 1 −2.451099 0.06864411 82 −35.70734 0
qvalueq. 2 −3.293441 0.06864411 82 −47.97850 0
f. Crocidura phylogenetic—alpha diversity
(Intercept) 5.032623 0.16111364 80 31.23648 0
FirstorSecondT2 −0.826753 0.17007061 20 −4.86123 0.0001
TreatmentLC 0.042234 0.20433123 3 0.20669 0.8471
qvalueq. 1 −2.534690 0.09074075 80 −27.93331 0
qvalueq. 2 −3.475121 0.09074075 80 −38.29725 0
FirstorSecondT2:qvalueq. 1phy −0.092760 0.12832680 80 −0.72284 0.4719
FirstorSecondT2:qvalueq. 2phy 0.194257 0.1279426 80 1.4698 0.1455
g. Apodemus neutral—alpha diversity
(Intercept) 6.160106 0.15415349 58 39.96086 0
FirstorSecondT2 −0.399795 0.17894850 14 −2.23413 0.0423
TreatmentLC −0.050041 0.18263856 2 −0.27399 0.8098
qvalueq. 1 −1.501942 0.08511615 58 −17.64579 0
qvalueq. 2 −2.534178 0.08511615 58 −29.77318 0
h. Apodemus phylogenetic—alpha diversity
(Intercept) 5.487686 0.08767164 56 62.59363 0
FirstorSecondT2 −0.206698 0.11099538 14 −1.86222 0.0837
TreatmentLC 0.016139 0.09767066 2 0.16524 0.8839
qvalueq. 1 −2.397517 0.06886942 56 −34.81251 0
qvalueq. 2 −4.107089 0.06891415 56 −59.60574 0
FirstorSecondT2:qvalueq. 1phy −0.056658 0.09739606 56 −0.59338 0.5553
FirstorSecondT2:qvalueq. 2phy 0.176851 0.09739606 56 1.81579 0.0748
Open in a new tab

Table A4.

PERMANOVA results for A. sylvaticus and C. russula

Main_effects df SumOfSqs R 2 F Pr(>F)
A. sylvaticus
Treatment 1 0.756 0.083 2.853 0
First or second 1 1.117 0.123 4.214 0
Treatment: first or second 1 0.344 0.038 1.297 0.162
Residual 26 6.891 0.757 NA NA
Total 29 9.108 1 NA NA
C. russula
Treatment 1 1.035 0.083 5.235 0
First or second 1 2.918 0.234 14.76 0
Treatment: first or second 1 1.022 0.082 5.171 0.012
Residual 38 7.513 0.602 NA NA
Total 41 12.488 1 NA NA
Open in a new tab

Koziol, A. , Odriozola, I. , Nyholm, L. , Leonard, A. , San José, C. , Pauperio, J. , Ferreira, C. , Hansen, A. J. , Aizpurua, O. , Gilbert, M. T. P. , & Alberdi, A. (2022). Enriching captivity conditions with natural elements does not prevent the loss of wild‐like gut microbiota but shapes its compositional variation in two small mammals. MicrobiologyOpen, 11, e1318. 10.1002/mbo3.1318

DATA AVAILABILITY STATEMENT

Sequence data are publicly available under the ENA project accession number PRJEB48838: https://www.ebi.ac.uk/ena/browser/view/PRJEB48838

REFERENCES

  1. Alberdi, A. (2019). Hilldiv: An R package for integral analysis of diversity based on hill numbers. BioRxiv, 5456. 10.1101/545665 [DOI] [Google Scholar]
  2. Alberdi, A. , & Gilbert, M. T. P. (2019). A guide to the application of hill numbers to DNA based diversity analyses. Molecular Ecology Resources, 19(4), 804–817. [DOI] [PubMed] [Google Scholar]
  3. Alberdi, A. , Martin, G. , & Aizpurua, O. 2021. Captivity systematically alters the composition yet not the diversity of vertebrate gut microbiomes. Scientific Reports, 11, 226600. [Google Scholar]
  4. Anderson, M. J. (2017). Permutational Multivariate Analysis of Variance (PERMANOVA). Wiley StatsRef: Statistics Refence Online. 1–15. [Google Scholar]
  5. Arrieta, M.‐C. , Stiemsma, L. T. , Amenyogbe, N. , Brown, E. M. , & Finlay, B. (2014). The intestinal microbiome in early life: Health and disease. Frontiers in Immunology, 5(September), 427. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Baumgartner, M. , Pfrunder‐Cardozo, K. R. , & Hall, A. R. (2021). Microbial community composition interacts with local abiotic conditions to drive colonization resistance in human gut microbiome samples. Proceedings. Biological Sciences/The Royal Society, 288(1947), 20203106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Beura, L. K. , Hamilton, S. E. , Bi, K. , Schenkel, J. M. , Odumade, O. A. , Casey, K. A. , Thompson, E. A. , Fraser, K. A. , Rosato, P. C. , Filali‐Mouhim, A. , Sekaly, R. P. , Jenkins, M. K. , Vezys, V. , Haining, W. N. , Jameson, S. C. , & Masopust, D. (2016). Normalizing the environment recapitulates adult human immune traits in laboratory mice. Nature, 532(7600), 512–516. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bibbò, S. , Ianiro, G. , Giorgio, V. , Scaldaferri, F. , Masucci, L. , Gasbarrini, A. , & Cammarota, G. (2016). The role of diet on gut microbiota composition. European Review for Medical and Pharmacological Sciences, 20(22), 4742–4749. [PubMed] [Google Scholar]
  9. Bonder, M. J. , Kurilshikov, A. , Tigchelaar, E. F. , Mujagic, Z. , Imhann, F. , Vich Vila, A. , & Deelen, P. (2016). The effect of host genetics on the gut microbiome. Nature Genetics, 48(11), 1407–1412. [DOI] [PubMed] [Google Scholar]
  10. Callahan, B. J. , McMurdie, P. J. , Rosen, M. J. , Han, A. W. , Johnson, A. J. A. , & Holmes, S. P. (2016). DADA2: High‐resolution sample inference from illumina amplicon data. Nature Methods, 13(7), 581–583. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Chao, A. , Chiu, C.‐H. , & Jost, L. (2014). Unifying species diversity, phylogenetic diversity, functional diversity, and related similarity and differentiation measures through hill numbers. Annual Review of Ecology, Evolution, and Systematics, 45, 297–324. 10.1146/annurev-ecolsys-120213-091540 [DOI] [Google Scholar]
  12. Davidson, G. L. , Raulo, A. , & Knowles, S. C. L. (2020). Identifying microbiome‐mediated behaviour in wild vertebrates. Trends in Ecology & Evolution, 35(11), 972–980. [DOI] [PubMed] [Google Scholar]
  13. Davis, N. M. , Proctor, D. M. , Holmes, S. P. , Relman, D. A. , & Callahan, B. J. (2018). Simple statistical identification and removal of contaminant sequences in marker‐gene and metagenomics data. Microbiome, 6(1), 226. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Fan, Y. , & Pedersen, O. (2021). Gut microbiota in human metabolic health and disease. Nature Reviews Microbiology, 19(1), 55–71. [DOI] [PubMed] [Google Scholar]
  15. Fujimura, K. E. , Demoor, T. , Rauch, M. , Faruqi, A. A. , Jang, S. , Johnson, C. C. , Boushey, H. A. , Zoratti, E. , Ownby, D. , Lukacs, N. W. , & Lynch, S. V. (2014). House dust exposure mediates gut microbiome lactobacillus enrichment and airway immune defense against allergens and virus infection. Proceedings of the National Academy of Sciences of the United States of America, 111(2), 805–810. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gálvez‐Ontiveros, Y. , Páez, S. , Monteagudo, C. , & Rivas, A. (2020). Endocrine disruptors in food: Impact on gut microbiota and metabolic diseases. Nutrients, 12(4), 1158. 10.3390/nu12041158 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hauffe, H. C. , & Barelli, C. (2019). Conserve the germs: The gut microbiota and adaptive potential. Conservation Genetics, 20(1), 19–27. [Google Scholar]
  18. Herlemann, D. P. R. , Labrenz, M. , Jürgens, K. , Bertilsson, S. , Waniek, J. J. , & Andersson, A. F. (2011). Transitions in bacterial communities along the 2000 Km salinity gradient of the Baltic Sea. The ISME journal, 5(10), 1571–1579. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hiippala, K. , Barreto, G. , Burrello, C. , Diaz‐Basabe, A. , Suutarinen, M. , Kainulainen, V. , & Bowers, J. R. (2020). Novel Odoribacter splanchnicus strain and its outer membrane vesicles exert immunoregulatory effects in vitro. Frontiers in Microbiology, 11(November), 575455. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hird, S. M. (2017). Evolutionary biology needs wild microbiomes. Frontiers in Microbiology, 8(April), 725. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kinross, J. M. , Darzi, A. W. , & Nicholson, J. K. (2011). Gut microbiome‐host interactions in health and disease. Genome Medicine, 3(3), 14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kohl, K. D. , Skopec, M. M. , & Dearing, M. D. (2014). Captivity results in disparate loss of gut microbial diversity in closely related hosts. Conservation Physiology, 2(1), cou009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kruskal, J. B. (1964). Nonmetric multidimensional scaling: A numerical method. Psychometrika, 29(2), 115–129. [Google Scholar]
  24. Langille, M. G. , Meehan, C. J. , Koenig, J. E. , Dhanani, A. S. , Rose, R. A. , Howlett, S. E. , & Beiko, R. G. (2014). Microbial shifts in the aging mouse gut. Microbiome, 2(1), 50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Liu, W. , Sun, Z. , Ma, C. , Zhang, J. , Ma, C. , Zhao, Y. , Wei, H. , Huang, S. , & Zhang, H. (2021). Exposure to soil environments during earlier life stages is distinguishable in the gut microbiome of adult mice. Gut Microbes, 13(1), 1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Madeira, F. , Park, Y. M. , Lee, J. , Buso, N. , Gur, T. , Madhusoodanan, N. , Basutkar, P. , Tivey, A. , Potter, S. C. , Finn, R. D. , & Lopez, R. (2019). “The EMBL‐EBI search and sequence analysis tools APIs in 2019. Nucleic Acids Research, 47(W1), W636–W641. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Martin, M. (2011). Cutadapt removes adapter sequences from high‐throughput sequencing reads. EMBnet. Journal, 17(1), 10–12. [Google Scholar]
  28. Martínez‐Mota, R. , Kohl, K. D. , Orr, T. J. , & Denise Dearing, M. (2020). Natural diets promote retention of the native gut microbiota in captive rodents. The ISME journal, 14(1), 67–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Maurice, C. F. , Knowles, S. C. L. , Ladau, J. , Pollard, K. S. , Fenton, A. , Pedersen, A. B. , & Turnbaugh., P. J. (2015). Marked seasonal variation in the wild mouse gut microbiota. The ISME journal, 9(11), 2423–2434. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Minh, B. Q. , Schmidt, H. A. , Chernomor, O. , Schrempf, D. , Woodhams, M. D. , von Haeseler, A. , & Lanfear, R. (2020). IQ‐TREE 2: New models and efficient methods for phylogenetic inference in the genomic era. Molecular Biology and Evolution, 37(5), 1530–1534. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Morrison, K. E. , Jašarević, E. , Howard, C. D. , & Bale, T. L. (2020). It's the fiber, not the fat: Significant effects of dietary challenge on the gut microbiome. Microbiome, 8(1), 15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Nemergut, D. R. , Schmidt, S. K. , Fukami, T. , O'neill, S. P. , Bilinski, T. M. , Stanish, L. F. , Knelman, J. E. , Darcy, J. L. , Lynch, R. C. , Wickey, P. , & Ferrenberg, S. (2013). Patterns and processes of microbial community assembly. Microbiology and Molecular Biology Reviews: MMBR, 77(3), 342–356. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Nicholls, H. T. , Krisko, T. I. , LeClair, K. B. , Banks, A. S. , & Cohen, D. E. (2016). Regulation of adaptive thermogenesis by the gut microbiome. The FASEB Journal, 30(1_suppl), 854.2. [Google Scholar]
  34. Nyholm, L. , Koziol, A. , Marcos, S. , Bolt Botnen, A. , Aizpurua, O. , Gopalakrishnan, S. , Limborg, M. T. , Gilbert, M. T. P. , & Alberdi, A. (2020). Holo‐omics: Integrated host‐microbiota multi‐omics for basic and applied biological research. I Science, 23(8), 101414. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Nyholm, L. , Odriozola, I. , Bideguren, G. M. , Aizpurua, O. , & Alberdi, A. (2022). Gut microbiota differences between paired intestinal wall and digesta samples in three small species of fish. PeerJ (Corta Madera, CA and London), 10(February), e12992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Oksanen, J. , Blanchet, F. G. , Friendly, M. , Kindt, R. , Legendre, P. , McGlinn, D. , Minchin, P. R. , O'Hara, R. B , Solymos, P. , Stevens, M. H. H. , Szoecs, E. , Wagner, H. , Barbour, M. , Bedward, M. , Bolker, B. , Borcard, D. , Carvalho, G. , Chirico, M. , De Caceres, M. , … Weedon, J. (2020). Vegan: Community ecology package. https://CRAN.R-project.org/package=vegan
  37. Ovaskainen, O. , & Abrego, N. (2020). Joint species distribution modelling: With applications in R. Cambridge University Press. [Google Scholar]
  38. Pinheiro, J. , Bates, D. , DebRoy, S. , Sarkar, D. , Heisterkamp, S. , Van Willigen, B. , & Maintainer, R. (2017). Package ‘nlme.’. Linear and Nonlinear Mixed Effects Models, Version, 3(1), 1– 159. http://cran.rapporter.net/web/packages/nlme/nlme.pdf [Google Scholar]
  39. Raulo, A. , Allen, B. E. , Troitsky, T. , Husby, A. , Firth, J. A. , Coulson, T. , & Knowles, S. C. L. (2021). Social networks strongly predict the gut microbiota of wild Mice. The ISME Journal, 15, 2601–2613. 10.1038/s41396-021-00949-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Raymann, K. , & Moran, N. A. (2018). The role of the gut microbiome in health and disease of adult honey bee workers. Current Opinion in Insect Science, 26(April), 97–104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Rose, C. E. , Martin, S. W. , Wannemuehler, K. A. , & Plikaytis, B. D. (2006). On the use of Zero‐Inflated and hurdle models for modeling vaccine adverse event count data. Journal of Biopharmaceutical Statistics, 16(4), 463–481. [DOI] [PubMed] [Google Scholar]
  42. Rosenbaum, M. D. , VandeWoude, S. , & Johnson, T. E. (2009). “Effects of cage‐change frequency and bedding volume on mice and their microenvironment. Journal of the American Association for Laboratory Animal Science: JAALAS, 48(6), 763–773. [PMC free article] [PubMed] [Google Scholar]
  43. Rosshart, S. P. , Herz, J. , Vassallo, B. G. , Hunter, A. , Wall, M. K. , Badger, J. H. , McCulloch, J. A. , Anastasakis, D. G. , Sarshad, A. A. , Leonardi, I. , Collins, N. , Blatter, J. A. , Han, S. J. , Tamoutounour, S. , Potapova, S. , Foster St Claire, M. B. , Yuan, W. , Sen, S. K. , Dreier, M. S. , … Rehermann, B. (2019). Laboratory mice born to wild mice have natural microbiota and model human immune responses. Science, 365(6452), eaaw4361. 10.1126/science.aaw4361 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Rosshart, S. P. , Vassallo, B. G. , Angeletti, D. , & Hutchinson, D. S. (2017). Wild mouse gut microbiota promotes host fitness and improves disease resistance. Cell, 171, 1015–1028. https://www.sciencedirect.com/science/article/pii/S0092867417310656 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Sbihi, H. , Ct Boutin, R. , Cutler, C. , Suen, M. , Finlay, B. B. , & Turvey, S. E. (2019). Thinking bigger: How early‐life environmental exposures shape the gut microbiome and influence the development of asthma and allergic disease. Allergy, 74(11), 2103–2115. [DOI] [PubMed] [Google Scholar]
  46. Schubert, M. , Lindgreen, S. , & Orlando, L. (2016). AdapterRemoval v2: Rapid adapter trimming, identification, and read merging. BMC Research Notes, 9(February), 88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Schuijt, T. J. , Lankelma, J. M. , Scicluna, B. P. , de Sousa e Melo, F. , Roelofs, J. J. T. H. , de Boer, J. D. , & Hoogendijk, A. J. (2016). The gut microbiota plays a protective role in the host defence against pneumococcal pneumonia. Gut, 65(4), 575–583. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Sepulveda, J. , & Moeller, A. H. (2020). The effects of temperature on animal gut microbiomes. Frontiers in Microbiology, 11, 384. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Shinohara, A. , Nohara, M. , Kondo, Y. , Jogahara, T. , Nagura‐Kato, G. A. , Izawa, M. , & Koshimoto, C. (2019). Comparison of the gut microbiotas of laboratory and wild asian house shrews (Suncus murinus) based on cloned 16S rRNA sequences. Experimental Animals/Japanese Association for Laboratory Animal Science, 68(4), 531–539. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Singh, A. , Faber‐Hammond, J. J. , O'Rourke, C. F. , & Renn, S. C. P. (2019). Gut microbial diversity increases with social rank in the African Cichlid Fish, Astatotilapia Burtoni . Animal Behaviour, 152(June), 79–91. [Google Scholar]
  51. Sommer, F. , Ståhlman, M. , Ilkayeva, O. , Arnemo, J. M. , Kindberg, J. , Josefsson, J. , Newgard, C. B. , Fröbert, O. , & Bäckhed, F. (2016). The gut microbiota modulates energy metabolism in the hibernating brown bear Ursus Arctos . Cell Reports, 14(7), 1655–1661. [DOI] [PubMed] [Google Scholar]
  52. Takai, K. , & Horikoshi, K. (2000). Rapid detection and quantification of members of the archaeal community by quantitative PCR using fluorogenic probes. Applied and Environmental Microbiology, 66(11), 5066–5072. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. van Teijlingen, E. , & Hundley, V. (2002). The importance of pilot studies. Nursing Standard: Official Newspaper of the Royal College of Nursing, 16(40), 33–36. [DOI] [PubMed] [Google Scholar]
  54. Tikhonov, G. , Opedal, Ø. H. , Abrego, N. , Lehikoinen, A. , de Jonge, M. M. J. , Oksanen, J. , & Ovaskainen, O. (2020). Joint species distribution modelling with the R‐Package hmsc. Methods in Ecology and Evolution/British Ecological Society, 11(3), 442–447. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Turnbaugh, P. J. , Hamady, M. , Yatsunenko, T. , Cantarel, B. L. , Duncan, A. , Ley, R. E. , Sogin, M. L. , Jones, W. J. , Roe, B. A. , Affourtit, J. P. , Egholm, M. , Henrissat, B. , Heath, A. C. , Knight, R. , & Gordon, J. I. (2009). A core gut microbiome in obese and lean twins. Nature, 457(7228), 480–484. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Warton, D. I. , Blanchet, F. G. , O'Hara, R. B. , Ovaskainen, O. , Taskinen, S. , Walker, S. C. , & Hui, F. K. C. (2015). So many variables: Joint modeling in community ecology. Trends in Ecology & Evolution, 30(12), 766–779. [DOI] [PubMed] [Google Scholar]
  57. Weinstein, S. B. , Martínez‐Mota, R. , Stapleton, T. E. , Klure, D. M. , Greenhalgh, R. , Orr, T. J. , Dale, C. , Kohl, K. D. , & Dearing, M. D. (2021). Microbiome stability and structure is governed by host phylogeny over diet and geography in woodrats (Neotoma spp.). Proceedings of the National Academy of Sciences of the United States of America, 118(47), e2108787118. 10.1073/pnas.2108787118 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Zhang, C. , Franklin, C. L. , & Ericsson, A. C. (2021). Consideration of gut microbiome in murine models of diseases. Microorganisms, 9(5), 1062. 10.3390/microorganisms9051062 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Zhao, R. , Symonds, J. E. , Walker, S. P. , Steiner, K. , Carter, C. G. , Bowman, J. P. , & Nowak, B. F. (2020). Salinity and fish age affect the gut microbiota of farmed chinook salmon (Oncorhynchus Tshawytscha). Aquaculture, 528(November), 735539. [Google Scholar]
  60. Zhou, D. , Zhang, H. , Bai, Z. , Zhang, A. , Bai, F. , Luo, X. , Hou, Y. , Ding, X. , Sun, B. , Sun, X. , Ma, N. , Wang, C. , Dai, X. , & Lu, Z. (2016). Exposure to soil, house dust and decaying plants increases gut microbial diversity and decreases serum immunoglobulin E levels in BALB/c mice. Environmental Microbiology, 18(5), 1326–1337. [DOI] [PubMed] [Google Scholar]
  61. Zmora, N. , Zilberman‐Schapira, G. , Suez, J. , Mor, U. , Dori‐Bachash, M. , Bashiardes, S. , Kotler, E. , Zur, M. , Regev‐Lehavi, D. , Brik, R. B. , Federici, S. , Cohen, Y. , Linevsky, R. , Rothschild, D. , Moor, A. E. , Ben‐Moshe, S. , Harmelin, A. , Itzkovitz, S. , Maharshak, N. , … Elinav, E. (2018). Personalized gut mucosal colonization resistance to empiric probiotics is associated with unique host and microbiome features. Cell, 174(6), 1388–1405.e21. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Sequence data are publicly available under the ENA project accession number PRJEB48838: https://www.ebi.ac.uk/ena/browser/view/PRJEB48838

Articles from MicrobiologyOpen are provided here courtesy of Wiley

RESOURCES