Climate change and tuberculosis drive non-adaptive shifts in the faecal microbiota of wild meerkats

doi:10.21203/rs.3.rs-1873485/v1

Download PDF

Article

Climate change and tuberculosis drive non-adaptive shifts in the faecal microbiota of wild meerkats

https://doi.org/10.21203/rs.3.rs-1873485/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Climate change and climate-driven increases in infectious disease threaten wildlife populations globally. Yet, their combined long-term effects on gut microbial communities remain unknown. Over the past two decades, wild meerkats inhabiting the Kalahari have experienced rapid climate change paired with increasing tuberculosis (TB) disease prevalence. Here we show that over this period the meerkat faecal microbiota has become inflated in Bacteroides and Fusobacterium, and depleted in members of Bacilli. Changes were underpinned by both rainfall patterns and TB status, with increasing taxa primarily linked to TB infection, and decreasing taxa primarily associated with drought. A decline in co-occurring clusters of Bacilli was associated with decreased survival, suggesting that shifts may not be adaptive but rather reflect a breakdown in microbial homeostasis. Our findings demonstrate that the impact of climate change and infectious diseases extends beyond hosts to gut microbial communities, potentially accelerating threats to wildlife health and survival.

Wildlife populations across the globe are under increasing pressure from climate change ¹ and the emergence or re-emergence of infectious pathogens ^2–4. The link between climate change and infectious disease was recognised over two decades ago ^5–8, and their combined impacts are felt across marine and terrestrial ecosystems ^9–11. To date, studies on these twin pressures have focused on their effect on host and parasite demography and population viability ^12–14, yet little is known about their long term impacts on host-associated microbial communities. Recent evidence suggests that gut microbiotas are affected by both warming ^15,16 and changes to host ecology driven by climate change ¹⁷, but long-term studies documenting shifts in gut microbial communities within a population are absent.

The gut microbiota is a diverse ecological community that promotes host physiological homeostasis ¹⁸, and its plasticity in response to environmental cues may allow hosts to rapidly adapt to ecological change ¹⁹. For instance, gut microbial commensals may confer protective effects from pathogens via modulation of host immunity ^20–22, or optimise physiology to changing temperatures ²³. Protective bacterial strains or groups of strains may therefore be expected to increase within the host population in response to a novel pressure, either via direct selection for beneficial bacterial strains, or via behavioural or developmental mechanisms that promote the colonisation and transmission of beneficial strains ²⁴.

Alternatively, changes to gut microbiomes may be neutral or even maladaptive to the host when fitness outcomes for hosts and commensal microbes are mismatched ²⁵. In such cases, strains that have either neutral or negative impacts on hosts may be able to increase in population size, potentially outcompeting beneficial strains in the process. For instance, experimental warmer temperatures during frog development leads to a breakdown in host-associated microbial homeostasis and increases mortality later in life ¹⁵. However, distinguishing between adaptive and maladaptive microbial responses can be challenging for observational systems. This is exemplified by shifts in human faecal microbiota composition with recent urbanisation, which confer a higher digestive efficiency for novel diets yet also appear to be linked to increased risk of metabolic disease ²⁶. Identifying whether observed changes to gut microbial communities are beneficial for hosts requires long-term information on host fitness, which in wild populations remains rare.

Meerkats (Suricata suricatta) inhabiting the South African Kalahari have experienced both climate change and an increase in tuberculosis (TB) prevalence over the past two decades ^14,27–30. Maximum temperatures have increased over the study period (Fig. 1a) at a rate five times faster than the global average ³¹, which decreases pup survival ²⁹. Mean seasonal rainfall, which is strongly linked to both fecundity and survival ³², is highly variable but has not changed over the past two decades (Fig. 1b). Nevertheless, drought years occur regularly and high maximum temperatures exacerbate the negative fitness impacts of low rainfall, with interactions between temperature and rainfall strongly shaping meerkat survival and reproductive fitness ²⁸.

Recently, high maximum temperatures have also been associated with increased risk of TB outbreaks within social groups ¹⁴. TB is caused by the bacterial pathogen Mycobacterium suricattae, a member of the M. tuberculosis species complex, and is characterised by lengthy latent infections that progress to clinical signs in about 25% of cases ^30,33,34. TB disease was first officially reported in this population in 1997, yet TB prevalence and social exposure rapidly increased hence, with up to 40% of the population showing signs of disease in in 2016 and 2017 ^{Fig. 1c; ,30}. Once clinical disease manifests, individuals rapidly lose condition and generally reach terminal stages in approximately four to seven months ^30,35.

Here we link long-term changes in both climate and TB with shifts in the meerkat faecal microbiota between 1997 and 2019, applying 1,141 faecal samples collected from 235 individuals (Fig. S1a). Feacal samples were collected from individually-recognisable meerkats as part of a long-term research program established in 1993 ³⁶. Samples used in this study were collected throughout the study period, although most were collected between 2006 and 2007, and again between 2014 and 2017 (Fig. S1a). Sampled meerkats were selected because they survived to at least one year of age and belonged to a social group that persisted at least two years. In total, our samples originate from 12% of the monitored adult (survived past the age of one year) population (Fig. S1b). We have analysed this wild meerkat faecal microbiota dataset previously to show that time of day is a strong predictor of microbiota composition, with only weak signals of identity, season and age ^37,38.

Our aims in this study were to 1) document shifts in the gut microbiome over the two-decade study period; 2) identify correlations between microbial composition and both climate variables and TB status; and 3) test whether observed changes to groups of co-occurring taxa increase meerkat survival. We hypothesize that if observed changes are beneficial for meerkats, taxa that are associated with increased survival in the face of both TB and climate change should become more abundant over the study period. Yet, a loss of beneficial taxa or increase of taxa associated with higher mortality would rather indicate a breakdown in microbial homeostasis.

Faecal microbiota composition has shifted over study period

Microbiome phylogenetic profiling was performed using 16S ribosomal RNA (rRNA) gene amplicon sequencing of the V4 region to generate Amplicon Sequence Variants ^{ASVs; 39}. We estimated total 16S copy number, often used as a proxy of bacterial load ^40,41 by scaling reads to an internal standard that was added to each weighed sample prior to DNA extraction.

The meerkat faecal microbiota was dominated by Clostridia (Phylum Firmicutes; 64% of reads), with Bacilli (Phylum Firmicutes) being the second most abundant class (14%), and other major classes making up 1 – 4% of reads (Fig. 2a). Faecal microbiota composition was quantified per sample (Fig. 2b), and samples were paired with data on TB status (Fig. 2b), body condition (Fig. 2c), maximum temperature (Fig. 2d), rainfall over previous fortnight (Fig. 2e), season (Fig. 2f), and year of collection (Fig. 2g). Information on social group membership, time of day samples were collected, and storage methods per sample are visualised in Fig. S2.

Microbiota shifts associated with TB and rainfall

We modelled correlates of overall composition with a distance-based redundancy analysis (dbRDA), including study period, TB status, body condition, season, and meerkat identity and additionally controlling for sample storage method (see methods) and sequencing depth. We controlled for taxa phylogeny and abundance by using weighted Unifrac, but also tested results against other measured, including unweighted Unifrac and Bray-Curtis.

As reported previously ³⁷, time of day was the strongest predictor along the first axis of variation, with abundances of the genus Clostridium sensu stricto 1 determining a sample’s position along axis 1 and being more abundant in samples taken in the morning (Fig. 3a). A sample’s position along the second axis of variation significantly correlated with TB status, body condition, study period, and rainfall, with samples with high values being more likely to be from meerkats that are TB exposed or diseased, in poor body condition, collected after 2014, or during periods of low rainfall (which generally occur during the dry season; Table S1). Samples with high values along the second axis of variation were characterised by higher relative abundances of the genera Bacteroides, Fusobacterium, Blautia, and Ruminococcus (torques group). In contrast, samples with low values were more likely to be collected before 2005, in the wet season, and from meerkats in relatively good condition, and these samples were characterised by higher relative abundances of Lactococcus, Enterococcus, and Paeniclostridium.

The constrained ordination results reflect a shift in faecal microbiome composition across the study period (Fig. 3b), although note that sampling frequency was very low before 2000. A similar pattern was observed when faecal microbiota composition is collapsed by individual (Fig. S3), with individuals born before 2006 having on average 3.6% (±5 s.d.; n = 84) Bacteroidia, and those being born after 2014 having on average 10.0% (±7 s.d.; n = 83). These trends were independent of season (Fig. S4) and social group (Fig. S5).

To investigate which specific genera increased and decreased over the study period, we applied Generalized Additive Mixed Models (GAMMs) to model abundances of the thirty most abundant genera, controlling for relevant methodological and biological variables, including storage method, sequencing run, time of day, season, meerkat identity and social group membership (Table S2). For this analysis we applied taxa counts estimated using an internal reference standard, which was spiked into weighed faecal samples prior to extraction and which acts as a reference to calculate ratios of absolute abundance.

Genera from the phylum Bacteroidetes all increased over the study period, including Bacteroides, Alloprevotella, Parabacteroides and Rikenellaceae RC9 gut group (Fig. 4a). Fusobacterium and Helicobacter, which often include opportunistic pathogens, also increased. In contrast, genera belonging to Bacilli (phylum Firmicutes), including Lactococcus, Pediococcus, and Enterococcus, tended to decrease over the study period.

We extracted GAMM model estimates to test whether taxa that increased or decreased over the study period were associated with TB exposure or disease (Fig. 4b), rainfall over the previous month (Fig. 4c), maximum temperature on the day of sample collection (Fig. 4d), and body condition (Fig. 4e). Many genera that increased over the study period were associated with both TB exposure and disease, potentially indicating latent infections produce a similar gut microbial fingerprint to TB disease. Bacteroides and Fusobacterium were particularly associated with TB. Fusobacterium, Alloprevotella, and Tyzzerella 4 all became progressively more abundant as TB progressed to a diseased state. In general, genera that decreased over the study period were not associated with TB with the exception of Lactococcus, yet were often more abundant with increased rainfall (Fig. 4c). Maximum temperature was associated with some genera, such as Bacillus, but did not explain shifts in most taxa across the study period (Fig. 4d). Many genera that increased were also associated with slightly poorer body condition, potentially as consequence of TB infection (Fig. 4e). Together, these results suggest that genera that increased over the study period are associated withTB disease and exposure, whilst genera that decreased are mostly associated with rainfall.

Co-occurring clusters of Bacilli have a positive association with survival

We tested whether changes to faecal microbiota composition conferred survival benefits using Cox proportional hazard models that predicted survival six months after each sample was taken. We controlled for known predictors of mortality including age, social status, TB stage, and season on the date of sample collection, and included meerkat identity as a random effect. We applied two approaches to predict survival, including a co-occurrence network approach that categorised genera into groups of co-occurring clusters, as well as genus-level associations with survival.

Co-occurrence analysis identified five major groups of co-occurring clusters (Fig. 5a), with Bacteroides and Phascolarctobacterium acting as hubs that were strongly associated with each other and negatively associated with other clusters. We collapsed faecal microbial counts based on cluster membership and modelled their association with future survival. We found that the cluster made up of Lactobacillus, Enterococcus, Pediococcus, and Weisella (all belonging to the class Bacilli and all except Weisella declined over the study period) was negatively associated with mortality risk, i.e. increased abundance of this cluster predicted higher meerkat survival (hazard ratio = 0.84, p = 0.03; Fig. 5b; Table S2). No other microbial cluster predicted survival, with non-microbial factors (TB status, social status, age, season and body condition) all having stronger associations with survival (Table S2). Effects of these non-microbial factors on survival are in line with findings from previous studies on this population ^14,28,32,42.

We repeated the survival analysis with 12 genera that had either significantly increased or decreased (increased: Fusobacterium, Bacteroides, Parabacteroides, Helicobacter, Phascolarctobacterium, Alloprevotella; decreased: Lactococcus, Pediococcus, Peptoclostridium, Collinsella, Enterococcus, Slackia; Fig. 4a). There was no significant association between any genus and survival (Fig. S6), suggesting that survival effects are better predicted by groups of taxa, in particular those belonging to Bacilli, than specific bacterial genera.

Meerkats inhabiting the Kalahari region in South Africa have experienced rapid climate change and increasing TB prevalence over the past two decades. We found a marked shift in microbiota composition during the same period, with increases in members of the genera Bacteroides and Fusobacterium, and reduced members of Bacilli such as Lactococcus and Pediococcus. Both TB status and rainfall patterns were linked to microbiota shifts, whilst maximum temperatures had much weaker associations with microbiota composition. Taxa that had changed in overall abundance had smaller effects on survival compared to known drivers of mortality, such as social status and season, yet co-occurring clusters of Bacilli members, including Lactococcus, were associated with higher survival. Our findings point towards a long-term shift in the faecal microbiota in response to multiple and potentially interacting pressures faced by this population, and provide evidence that this shift may be compounding the negative effects of TB and climate change.

TB infection and rainfall were largely associated with separate suites of genera, suggesting that they are potentially impacting the gut microbiota via independent mechanisms. For example, genera that increased over the study period (e.g. Bacteroides, Parabacteroides, and Helicobacter) were associated with TB status but not rainfall, whilst genera that declined (e.g. Enterococcus and Pediococcus) showed the opposite pattern. As such, increasing taxa may be largely a response to TB exposure and infection ^{e.g. 43,44}, whilst declining taxa may be more associated with drought. Whilst drought frequency has not changed over the last century in this region ³¹, sustained droughts occurred at the very beginning of the study period (1997-1998) and again at the end of the study period (2013 – 2019), with the later drought paired with exceptionally high maximum temperatures (Fig. 2d and g). This raises the possibility that whilst drought may be the fundamental driver of gut microbiota shifts, high temperatures exacerbate these effects. The extent to which faecal microbiotas may return to ‘normal’ at the end of a drought is unknown, yet increased in Bacilli and decreases in Bacteroides at the very end of the study period (Fig. 2a) hint at a return towards a more Bacteroides-depleted microbiome. Nevertheless, the negative impacts of TB and climate change interactions on this population are projected to worsen ¹⁴, potentially driving further changes to the faecal microbiota, which in turn could be further maintained by social transmission in this highly social species ⁴⁵.

The specific mechanisms generating the observed shifts in the faecal microbiota and the direction of associations remain uncertain, and there are several potential explanations. Our results suggest that TB and rainfall together impact the faecal microbiota in a manner that may intensify the negative effects on meerkat fitness. However, initial alterations to the faecal microbiota may be caused by climate-driven changes to diet, e.g. as in polar bears ¹⁷, which may in turn increase meerkat susceptibility to TB infection via microbiome-mediated changes to immune function ^21,44,46. It is currently unclear how meerkat diet has been impacted by changing climates, yet there is widespread evidence for global changes to invertebrate communities, therefore such changes may be expected ⁴⁷. An additional factor that may contribute to changes to the faecal microbiota is microbial transmission from humans ⁴⁸, which may occur during meerkat monitoring. However, levels of human contact have been consistent across the study period, and therefore cannot not explain strong fluctuations observed. As such, whilst human transmission may be additional factor interacting with climate change and TB to influence meerkat faecal microbiotas, we consider it unlikely to be a major driver.

Our finding that genera associated with higher survival declined during the study period support previous experimental findings from frogs that global change may increase mortality by promoting gut microbial dysbiosis ¹⁵. Unexpectedly, taxa that increased over the study period were not associated with higher mortality, hinting that any negative effects of microbial shifts are due to losses of health-associated taxa rather than the inflation to disease-associated taxa. Whilst speculative, such patterns may indicate that novel selection pressures could initially promote antagonistic interactions between hosts and their symbionts, although it is possible that over many host generations these interactions may once again shift to mutualism ⁴⁹. Crucially, mutualistic co-evolution, whereby one would expect adaptive changes to host control over gut microbiota composition, should be underpinned by changes to host genetics as well as fitness benefits to hosts ⁵⁰. Future studies on whether gut microbial shifts are linked to changes in host genotype over generations may elucidate whether observed gut microbial shifts are simply responses to environmental change or at least partly a result of changes to host control mechanisms.

We present evidence for a distinct shift in faecal microbiota composition in a wild population of meerkats that has faced multiple stressors over the past two decades, most notably climate change and TB. These changes may be non-adaptive or even maladaptive for meerkats because Bacilli members, which are associated with increased survival, on average decreased over the study period. Our study therefore represents the first evidence from a long-term study system that global change is not only altering the demography of wild populations, but also impacts their microbial symbionts in a way that may accelerate negative impacts of global change.

Study population and sample collection

The study population inhabits the Kalahari Desert region in South Africa (-26.96S, 21.83E). Individuals from this population are individually marked and have been monitored three to five times a week since 1993 by the Kalahari Meerkat Project ³⁶. 1,141 faecal samples from 235 individuals were analysed in this study, which were also analysed as part of Risely et al. ³⁷. 70% of samples were collected from meerkats belonging to eight social groups, whilst the remaining 30% were collected from across 34 different social groups. Faecal samples were collected from the ground immediately after a meerkat was observed defecating, and were stored next to an icepack and frozen within 8 hours. For long-term storage, samples prior to 2008 were mostly frozen at -80°C (n = 461), or, after 2008, freeze-dried and kept at room temperature (n = 648). Freeze-drying is proposed to be preferable to freezing for biological samples where cold-chain transport and storage cannot be guaranteed, because freeze-dried DNA is robust to long-term storage and temperature changes ^51–53, even when stored at room-temperature ^54–56.

We tested experimentally the effect of storage methods on quantified microbiome composition (see below for microbiome profiling methodology). Nine faecal samples from captive meerkats were either frozen or freeze-dried, storage method had negligible effects on measured faecal microbiota composition, with sample identity explaining 93% of variation in beta diversity ^{p < 0.001; Fig. S7; ,37}. In addition, we tested microbiota stability on 16 archived faecal samples originally collected from wild meerkats and freeze-dried between 2011 and 2013. The microbiome was quantified first in 2020 and again in 2022 and samples were kept at 4°C without buffer during this period. Microbial composition did not change over this two-year period, with sample identity explaining 95% in beta diversity (Fig. S8). These results indicate that the effect of sample identity overwhelms methodological effects. Moreover, reported trends are apparent in both frozen and freeze-dried samples (Fig. S9, Fig. S2), with little demarcation in microbiota composition in 2008 when samples began to be freeze-dried, and which represent the oldest freeze-dried samples (12 years old at the point of sequencing; Fig. S2).

DNA extraction with internal standard, 16S rRNA amplification and sequencing

Before DNA extraction, NAP buffer was added to all faecal samples ⁵⁷. A subsample of 0.6 ±0.05 µg (wet) was taken, and 3µl of ZymoBIOMICS Spike-in Control I (High Microbial Load) was added to each subsample prior to DNA extraction. This internal standard consists of cells belonging to Imtechella halotolerans and Allobacillus halotoleranss, two species which are rarely found in faecal microbiota communities. An internal standard allows us to quantify ratios of absolute abundance by adding a known number of cells to each sample by which to normalise microbiota counts after sequencing. This method measures 16S copy number rather than absolute abundance, but has shown to accurately reflect variation in absolute abundances when care is taken to standardize faecal sample mass ^40,41,58,59. We have shown previously with this dataset that sample identity accounts for 90% of variation in estimated bacterial load, whilst 10% is technical variation ³⁷.

The bacterial genomic DNA was extracted using the NucleoSpin 96 Soil kit (Macherey-Nagel) following the manufacturer’s instructions, and the hypervariable V4 region of the 16S rRNA gene was amplified using the primer pair 515 F (5-GTGCCAGCMGCCGCGGTAA-3) and 806 R (5-GGACTACHVGGGTWTCTAAT-3). We used the Fluidigm Access Array^TM for Illumina Sequencing Systems for indexing and adding Illumina adaptor sequences. After purification (NucleoMag® NGS Clean-up and Size Select, Macherey-Nagel) and quantification (QuantiFlour® dsDNA Systemt, Promega) of barcoded samples, the normalized pooled sample library was sequenced as paired-end run on Illumina MiSeq platform at the Institute of Evolutionary Ecology and Conservation Genomics, Ulm University. Samples were sequenced across four Illumina runs (MiSeq Reagent Kit v2, 500-cycles). Extraction and PCR negative controls were included on all runs.

Microbiome bioinformatics and normalisation

All sequence reads were processed using QIIME2 version 2020.2 ⁶⁰. Sequences were merged, quality filtered, and chimeras were removed using the DADA2 pipeline ³⁹ to generate amplicon sequence variants (ASVs) ^39,61. Primers were trimmed and reads were truncated at 244 (forward) and 235 (reverse) base pairs. ASVs were assigned a taxonomy using SILVA version 132 ⁶². A tree was built using QIIME2’s fragment insertion method ⁶³. ASVs were filtered if they were not bacteria, not assigned to a phylum (as these are assumed to be spurious), or if they were classified as mitochondria or chloroplasts. We used the function decontam::isContaminant ⁶⁴ using the ‘prevalence’ method to identify sand microbes using 15 sand samples as a reference, and to remove them from the dataset. We then divided taxa counts per sample by Allobacillus halotolerans abundance per sample to quantify ratios of absolute abundance across samples ^{described in 37}. Both Allobacillus and Imtechella were then removed, and all further analysis were conducted on normalised reads. Because some samples had very high relative abundances of spike-in, we only retained samples for which read depth of the true microbiome (minus the internal reference) was over 5,000.

Sample metadata and climate data

All variables associated with each sample are described in detail in ³⁷. Here, our aim was to quantify the contributions of year, TB, rainfall, maximum temperatures, and body condition, whilst controlling for important sources of variation identified in that study. The most important predictors of taxa abundances identified were time of day, meerkat age, season, as well as sequencing depth, sequencing run, and storage. We therefore included these variables in all models (described below).

For each meerkat, dates of status changes for TB (becoming exposed and progressing to clinical TB, were assigned as described in ³⁰. In brief, individuals were considered diseased if they showed external clinical signs of TB such as characteristic submandibular lumps, and were considered exposed from the first day they were observed co-habiting with a diseased individual. Many individuals later in the study were born into groups affected by TB and therefore considered exposed at birth. Body condition was calculated as the residuals of a GAMM predicting body weight against age, month, and time of day (age being the strongest predictor of weight), applying all available weight data from the individuals included in this study. Hourly and daily weather data, including maximum temperatures, was provided by the South African Weather Service. Weather data was collected from the Van Zyl Rus weather station, approximately 25 km away from the study area. We measured time of day in reference to sunrise because this is more biologically meaningful than chronological time of day. We calculated daily sunrise times using suncalc::getSunlightTimes ⁶⁵. We categorised season into wet (October to April) and dry (May to September).

Statistical analysis of the faecal microbiota

We modelled overall composition (beta diversity) using constrained ordinations on a weighted Unifrac distance matrix and applying the function vegan::capscale. We excluded rare taxa represented by under 100 reads, and collapsed ASVs by genus, because we were interested in which genera were driving associations. Nevertheless, we repeated analyses at the ASV level and found little difference in results and interpretation. We included time of day, TB stage (unexposed, exposed, diseased), body condition (high – top quartile, low – bottom quartile, and average – between top and bottom quartile), rainfall over the previous month, study period (<2000, 2000-2004, 2005-2009, 2010-2014, and >2015), storage (frozen, freeze-dried), sequencing depth, and meerkat ID in models. We did not add maximum temperature because this is closely correlated with rainfall and season, with mean changes over the study period being relatively small compared to within-year variation. Variables of interest were categorised into factors rather than included as continuous variables because some association with the microbiome are non-linear, and to be able to calculate centroids per group. Statistical analysis of constrained ordinations was done using vegan::anova.cca. We repeated this process using unweighted Unifrac and Bray-Curtis, both at genus and ASV level.

To model genus-level abundances, we applied General Additive Mixed Models (GAMMs) to model the abundances of the 30 most abundant genera ³⁷, using the function mgcv::bam. Rare genera were included by collapsing any remaining genera that were modelled under ‘rare taxa’. To model yearly trends (Fig. 4a), we explicitly modelled year with a smoothing factor and included time of day, month, age, sequencing depth, storage, sequencing run, overall bacterial load, meerkat ID (random effect) and social group (random effect) as co-variables. To model the effects of TB, maximum temperature, and body condition, we repeated the model but included these additional variables, and included year as a random effect.

Co-occurrence networks

To identify clusters of co-occurring genera, we examined co-occurrence associations between the 30 most common genera using the package NetCoMi ⁶⁶. Because network associations are strongly dependent on bacterial abundance or sequencing depth (e.g. if bacterial load is high, both taxa 1 and taxa 2 will have high abundances even though they are negatively associated, leading to false positive associations), we accounted for overall bacterial load by CLR transforming microbiome data for both co-occurrence networks and survival analyses. Only associations with a Spearman’s correlation of over 0.2 were retained. Co-occurring clusters were identified using the NetCoMi::netAnalyze function using the fast and greedy method.

Survival analyses

To test the effects of faecal microbiome composition on future survival, we fitted mixed-effect Cox proportional hazard models using the function coxme::coxme ⁶⁷, with meerkat ID as a random effect. The model predicted whether a meerkat was alive or dead six month after each faecal sample was taken (dead n = 316, alive n = 791). For all models we controlled for meerkat age, season, social status, and TB status. We chose six months because over shorter periods sample sizes for dead vs alive were very unequal, whilst over longer periods the predictive power of models was lower (model AIC 6 months: 343; model AIC 1 year: 387).

We tested for the effect of microbial features on survival in two ways: we first collapsed microbiome data by network cluster membership and entered these into the survival model as CLR-transformed counts. Second, because, we specifically wanted to test whether abundances of genera that were either increasing or decreasing were associated with future survival, we selected 12 genera to model, including six that were increasing the most and six that were decreasing the most, and included counts of these 12 genera in the model.

Acknowledgments

We are grateful to the Kalahari Research Trust and the Kalahari Meerkat Project for access to facilities and habituated animals in the Kuruman River Reserve, South Africa. This paper has relied on records of individual identities and/or life histories maintained by the Kalahari Meerkat Project and collected by scientists and volunteers. We thank the Northern Cape Conservation Service for permission to conduct fieldwork, and the South African Weather Service (SAWS) for providing weather data. We thank Ben Danzer for facilitating with sample collation and storage and Ulrike Stehle for contributing to lab work.

Funding

German Research Foundation DFG SO 428/15-1 (SS); European Research Council 294494 (THCB); European Research Council 742808 (THCB); Human Frontier Science; Program RGP0051/2017 (THCB); University of Zurich (MBM); MAVA Foundation KRP 16026 (MBM, THCB).

Author contributions

Conceptualization: AR, SS, NMK, DWS; Formal analysis: AR, NMK; Investigation: AR, KW; Resources: SS, MBM, THCB; Data Curation: MBM, THCB; Writing - Original Draft: AR; Writing - Review & Editing: AR, NMK, DWS, NMK, SS, MBM; Project administration: SS; Funding acquisition: SS, AR, MBM, THCB.

Competing interests

Authors declare no competing interests.

Data and materials availability

All sequences and processed data used in this study are available to download at ⁶⁸. Sequences are additionally stored under NCBI BioProject PRJNA764180. R code can be downloaded at https://github.com/Riselya/Meerkat-microbiome-climate-change.

1. Bellard, C., Bertelsmeier, C., Leadley, P., Thuiller, W. & Courchamp, F. Impacts of climate change on the future of biodiversity. Ecology letters 15, 365–377 (2012).

2. Daszak, P., Cunningham, A. A. & Hyatt, A. D. Emerging infectious diseases of wildlife--threats to biodiversity and human health. Science 287, 443 (2000).

3. Jones, K. E. et al. Global trends in emerging infectious diseases. Nature 451, 990 (2008).

4. Tompkins, D. M., Carver, S., Jones, M. E., Krkošek, M. & Skerratt, L. F. Emerging infectious diseases of wildlife: a critical perspective. Trends in parasitology 31, 149–159 (2015).

5. Harvell, C. D. et al. Emerging marine diseases--climate links and anthropogenic factors. Science 285, 1505–1510 (1999).

6. Harvell, C. D. et al. Climate warming and disease risks for terrestrial and marine biota. Science 296, 2158–2162 (2002).

7. Altizer, S., Ostfeld, R. S., Johnson, P. T., Kutz, S. & Harvell, C. D. Climate change and infectious diseases: from evidence to a predictive framework. Science 341, 514–519 (2013).

8. Baker, R. E. et al. Infectious disease in an era of global change. Nature Reviews Microbiology 20, 193–205 (2022).

9. Maynard, J. et al. Projections of climate conditions that increase coral disease susceptibility and pathogen abundance and virulence. Nature Climate Change 5, 688–694 (2015).

10. Kock, R. A. et al. Saigas on the brink: Multidisciplinary analysis of the factors influencing mass mortality events. Science advances 4, eaao2314 (2018).

11. Kafle, P. et al. Range expansion of muskox lungworms track rapid arctic warming: implications for geographic colonization under climate forcing. Scientific reports 10, 1–14 (2020).

12. Cohen, J. M., Sauer, E. L., Santiago, O., Spencer, S. & Rohr, J. R. Divergent impacts of warming weather on wildlife disease risk across climates. Science 370, eabb1702 (2020).

13. Ros, A. et al. Current and projected impacts of the parasite Tetracapsuloides bryosalmonae (causative to proliferative kidney disease) on Central European salmonid populations under predicted climate change. Freshwater Biology 66, 1182–1199 (2021).

14. Paniw, M. et al. Higher temperature extremes exacerbate negative disease effects in a social mammal. Nature Climate Change 1–7 (2022).

15. Greenspan, S. E. et al. Warming drives ecological community changes linked to host-associated microbiome dysbiosis. Nature Climate Change 10, 1057–1061 (2020).

16. Sepulveda, J. & Moeller, A. H. The effects of temperature on animal gut microbiomes. Frontiers in microbiology 11, (2020).

17. Watson, S. E. et al. Global change-driven use of onshore habitat impacts polar bear faecal microbiota. The ISME journal 13, 2916–2926 (2019).

18. Sommer, F. & Bäckhed, F. The gut microbiota—masters of host development and physiology. Nature Reviews Microbiology 11, 227–238 (2013).

19. Alberdi, A., Aizpurua, O., Bohmann, K., Zepeda-Mendoza, M. L. & Gilbert, M. T. P. Do vertebrate gut metagenomes confer rapid ecological adaptation? Trends in Ecology & Evolution 31, 689–699 (2016).

20. Jiménez, R. R. & Sommer, S. The amphibian microbiome: natural range of variation, pathogenic dysbiosis, and role in conservation. Biodiversity and Conservation 26, 763–786 (2017).

21. Harris, E. V., de Roode, J. C. & Gerardo, N. M. Diet–microbiome–disease: Investigating diet’s influence on infectious disease resistance through alteration of the gut microbiome. PLoS pathogens 15, e1007891 (2019).

22. Fleischer, R. et al. Interaction between MHC diversity and constitution, gut microbiota and Astrovirus infections in a neotropical bat. Molecular Ecology (2022).

23. Chevalier, C. et al. Gut microbiota orchestrates energy homeostasis during cold. Cell 163, 1360–1374 (2015).

24. Moeller, A. H. & Sanders, J. G. Roles of the gut microbiota in the adaptive evolution of mammalian species. Philosophical Transactions of the Royal Society B 375, 20190597 (2020).

25. Carmody, R. N., Sarkar, A. & Reese, A. T. Gut microbiota through an evolutionary lens. Science 372, 462–463 (2021).

26. Sonnenburg, J. L. & Sonnenburg, E. D. Vulnerability of the industrialized microbiota. Science 366, eaaw9255 (2019).

27. Parsons, S. D., Drewe, J. A., van Pittius, N. C. G., Warren, R. M. & Van Helden, P. D. Novel cause of tuberculosis in meerkats, South Africa. Emerging infectious diseases 19, 2004 (2013).

28. Paniw, M., Maag, N., Cozzi, G., Clutton-Brock, T. & Ozgul, A. Life history responses of meerkats to seasonal changes in extreme environments. Science 363, 631–635 (2019).

29. Van de Ven, T. M., Fuller, A. & Clutton‐Brock, T. H. Effects of climate change on pup growth and survival in a cooperative mammal, the meerkat. Functional Ecology 34, 194–202 (2020).

30. Müller-Klein, N. et al. Two decades of tuberculosis surveillance reveal disease spread, high levels of exposure and mortality, and marked variation in disease progression in wild meerkats. (2022).

31. Pattinson, N. B. et al. Collapse of Breeding Success in Desert-Dwelling Hornbills Evident Within a Single Decade. Frontiers in Ecology and Evolution 434 (2022).

32. Groenewoud, F. & Clutton‐Brock, T. Meerkat helpers buffer the detrimental effects of adverse environmental conditions on fecundity, growth and survival. Journal of Animal Ecology 90, 641–652 (2021).

33. Drewe, J. A., Eames, K. T. D., Madden, J. R. & Pearce, G. P. Integrating contact network structure into tuberculosis epidemiology in meerkats in South Africa: Implications for control. Preventive Veterinary Medicine 101, 113–120 (2011).

34. Donadio, J. et al. Characterizing Tuberculosis progression in wild meerkats (Suricata suricatta) from fecal samples and clinical signs. The Journal of Wildlife Diseases 58, 309–321 (2022).

35. Patterson, S. J. et al. Combining analytical approaches and multiple sources of information to improve interpretation of diagnostic test results for tuberculosis in wild meerkats. Animals 11, 3453 (2021).

36. Clutton-Brock, T. H. & Manser, M. Meerkats: cooperative breeding in the Kalahari. Cooperative breeding in vertebrates 294, 317 (2016).

37. Risely, A., Wilhelm, K., Clutton-Brock, T., Manser, M. B. & Sommer, S. Diurnal oscillations in gut bacterial load and composition eclipse seasonal and lifetime dynamics in wild meerkats. Nature communications 12, 1–12 (2021).

38. Risely, A. et al. Gut microbiota repeatability is contingent on temporal scale and age in wild meerkats. (2022).

39. Callahan, B. J. et al. DADA2: High-resolution sample inference from Illumina amplicon data. Nature Methods 13, 581–583 (2016).

40. Stämmler, F. et al. Adjusting microbiome profiles for differences in microbial load by spike-in bacteria. Microbiome 4, 28 (2016).

41. Tourlousse, D. M. et al. Synthetic spike-in standards for high-throughput 16S rRNA gene amplicon sequencing. Nucleic Acids Res 45, e23 (2017).

42. Thorley, J. et al. Sex‐independent senescence in a cooperatively breeding mammal. Journal of Animal Ecology 89, 1080–1093 (2020).

43. Hong, B.-Y. et al. Microbiome changes during tuberculosis and antituberculous therapy. Clinical microbiology reviews 29, 915–926 (2016).

44. Dumas, A. et al. The host microbiota contributes to early protection against lung colonization by Mycobacterium tuberculosis. Frontiers in Immunology 9, 2656 (2018).

45. Sarkar, A. et al. Microbial transmission in animal social networks and the social microbiome. Nature ecology & evolution 4, 1020–1035 (2020).

46. Alavi, S. et al. Interpersonal gut microbiome variation drives susceptibility and resistance to cholera infection. Cell 181, 1533-1546. e13 (2020).

47. Halsch, C. A. et al. Insects and recent climate change. Proceedings of the national academy of sciences 118, e2002543117 (2021).

48. Clayton, J. B. et al. Captivity humanizes the primate microbiome. Proceedings of the National Academy of Sciences 113, 10376–10381 (2016).

49. Johnson, C. A. et al. Coevolutionary transitions from antagonism to mutualism explained by the Co-Opted Antagonist Hypothesis. Nature communications 12, 1–11 (2021).

50. Lambert, M. R., Brans, K. I., Des Roches, S., Donihue, C. M. & Diamond, S. E. Adaptive evolution in cities: progress and misconceptions. Trends in Ecology & Evolution 36, 239–257 (2021).

51. Blekhman, R. et al. Common methods for fecal sample storage in field studies yield consistent signatures of individual identity in microbiome sequencing data. Scientific reports 6, 1–5 (2016).

52. Shen, W., Wu, D., Qiu, W. & Yi, X. Evaluation of freeze-drying for quantification of the microbiome and metabolome in neonatal faecal samples. Medicine in Microecology 8, 100044 (2021).

53. Bensch, H. M., Tolf, C., Waldenström, J., Lundin, D. & Zöttl, M. Freeze-drying can replace cold-chains for transport and storage of fecal microbiome samples. PeerJ 10, e13095 (2022).

54. Wu, Y. et al. Lyophilization is suitable for storage and shipment of fresh tissue samples without altering RNA and protein levels stored at room temperature. Amino Acids 43, 1383–1388 (2012).

55. Mareninov, S. et al. Lyophilized brain tumor specimens can be used for histologic, nucleic acid, and protein analyses after 1 year of room temperature storage. Journal of neuro-oncology 113, 365–373 (2013).

56. Molnar, A. et al. Lyophilization and homogenization of biological samples improves reproducibility and reduces standard deviation in molecular biology techniques. Amino acids 53, 917–928 (2021).

57. Menke, S., Gillingham, M. A., Wilhelm, K. & Sommer, S. Home-made cost effective preservation buffer is a better alternative to commercial preservation methods for microbiome research. Frontiers in microbiology 8, 102 (2017).

58. Hardwick, S. A., Deveson, I. W. & Mercer, T. R. Reference standards for next-generation sequencing. Nature Reviews Genetics 18, 473 (2017).

59. Lin, Y., Gifford, S., Ducklow, H., Schofield, O. & Cassar, N. Towards quantitative microbiome community profiling using internal standards. Applied and environmental microbiology 85, (2019).

60. Bolyen, E. et al. QIIME 2: Reproducible, interactive, scalable, and extensible microbiome data science. (2018).

61. Callahan, B. J., McMurdie, P. J. & Holmes, S. P. Exact sequence variants should replace operational taxonomic units in marker-gene data analysis. The ISME Journal 11, 2639–2643 (2017).

62. Pruesse, E. et al. SILVA: a comprehensive online resource for quality checked and aligned ribosomal RNA sequence data compatible with ARB. Nucleic acids research 35, 7188–7196 (2007).

63. Janssen, S. et al. Phylogenetic placement of exact amplicon sequences improves associations with clinical information. Msystems 3, (2018).

64. Davis, N. M., Proctor, D. M., Holmes, S. P., Relman, D. A. & Callahan, B. J. Simple statistical identification and removal of contaminant sequences in marker-gene and metagenomics data. Microbiome 6, 1–14 (2018).

65. Agafonkin, V. & Thieurmel, B. suncalc: Compute sun position, sunlight phases, moon position, and lunar phase. R package version 0.3. (2017).

66. Peschel, S., Müller, C. L., von Mutius, E., Boulesteix, A.-L. & Depner, M. NetCoMi: network construction and comparison for microbiome data in R. Briefings in bioinformatics 22, bbaa290 (2021).

67. Therneau, T. M. coxme: Mixed Effects Cox Models. R package version 2.2-17. (2022).

68. Risely, A., Wilhelm, K., Clutton-Brock, T. H., Manser, M. & Sommer, S. Data and code for: Diurnal oscillations in gut bacterial load and composition eclipse seasonal and lifetime dynamics in wild meerkats, Suricata suricatta. Version 1.0. Zenodo (2021) doi:10.5281/zenodo.5337076.

There is NO Competing Interest.

Supplementarymaterials.docx
Supplementary figures and tables

Download PDF

Version 1

posted

You are reading this latest preprint version

Climate change and tuberculosis drive non-adaptive shifts in the faecal microbiota of wild meerkats

Status:

Version 1

Abstract

Figures

Introduction

Results

Discussion

Methods

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1