Microorganisms associated with animals harbour a unique set of functional traits, which is increasingly recognised as pivotal for the normal functioning of their hosts 1. Nutrition, thermoregulation and behaviour are some of the biological processes of animals that are affected by microbial communities and their activity 2. As microbial community compositions vary rapidly, microbial functions can change at a much faster rate than those encoded in animal genomes 3. These observations led many researchers to hypothesise about animal-associated microbial communities acting as boosters of acclimation and adaptation capacity of their hosts 3–7. The importance of measuring the dimensions of this contribution is not negligible, as it can help us to better predict species’ responses to anthropic pressures, such as land-use or climate change 8, as well as to implement microbiome manipulation strategies in captive breeding programs of threatened species 9. However, have we generated enough knowledge to validate or refute that hypothesis?
To address this question, we conducted a quantitative systematic review of the literature. A programmatic search of the Scopus and Web of Science publication databases yielded a non-redundant collection of 1974 articles that corresponded to our keywords. All these articles were manually screened to yield a final set of 109 studies that met our inclusion criteria. These studies were dominated by research on mammals, followed by birds, fish, reptiles and amphibians (Fig. 1a-b). The number of analysed animal individuals ranged between 2 and 758, with an average number of 77.9 ± 105.8 samples per study (Fig. 1c). Geographical distribution spanned six continents, with a clear dominance of Asia, especially China, followed by Europe (Fig. 1d). However, major geographic gaps were identified in North Africa, Central Asia and Melanesia.
The selected studies were assigned quantitative scores based on 10 criteria encompassing experimental design, methodological resolution and reproducibility. Quantitative data were then combined to obtain an overall performance score spanning 0 to 1; lowest scores indicating mere speculation on microbial contribution to host adaptability, and highest scores indicating evidence on the magnitude and mechanism of acclimation or adaptation. To reduce subjectivity, the weight of each criterion in the overall performance calculations was determined through the collective assessment of eight independent experts. The average total performance score was 0.42 ± 0.07 out of 1, with minimum and maximum values ranging between 0.25 and 0.58 (Fig. 2a). These results indicate that the field has yet to align with ideal research practices to address whether gut microorganisms confer vertebrates with enhanced acclimation and adaptation capacity. Average scores exhibited a slight yet significant increasing temporal trend of 0.9% per year (LM; β = 0.009 points/year, F = 2.196, DFnom/den=1/107, p = 0.030, Fig. 2b), driven by similarly modest yet statistically non-significant temporal increments in the three domains analysed (Supplementary Table S1, Supplementary Fig. S2). There were no differences in total scores among continents or taxa (Fig. 2c, Table S2).
One notable limitation we detected in experimental designs was the absence of microbiota manipulation, which is essential for establishing causal relationships between microbial communities and host responses 10. Inoculation of germ-free animals with microbial communities is feasible only with model species, due to the complex infrastructure required for creating and sustaining gnotobiotic animals 11. When working with wild systems, more practical alternatives include treating animals with antibiotics to deplete any microbial activity that could contribute to acclimation 12, or conducting faecal microbiota transplants between animals adapted to different conditions 13. However, none of these practices were implemented in the analysed studies. Incorporating these experimental manipulations would help mitigate the inherent constraints of correlative approaches and provide a pathway toward elucidating causal relationships between microbial communities and adaptive capacity of animals.
Similarly, methods lacked explicit measurements of fitness or related proxies, with only 5% of the studies incorporating host responses in terms of animal physiology, reproduction, or survival. This deficiency hinders the ability to quantify the extent of the microbiota's adaptive contribution to their hosts, rendering any conclusions speculative. It is worth noting that directly measuring fitness through reproductive success or survival metrics in vertebrates is challenging due to complex experimental setups and study durations. Additionally, ethical considerations and related legislation may limit the use of survival as a fitness indicator. Alternatives, such as assessing physiological parameters (e.g., calorimetry, respirometry) or behavioural metrics, can serve as valuable proxies for fitness.
Studying the functional contributions of microorganisms can provide deeper insights into the mechanisms through which microorganisms enhance vertebrates’ acclimation capacity 14. Genome-resolved metagenomics allows for the near-complete reconstruction of bacterial genomes from faecal samples 15, which can then be functionally annotated to gain insights into their metabolic capacities 16. The incorporation of RNAseq data not only provides information on functional capacities but also measures the actual functional activity of microorganisms. Maximum scores of the molecular approaches employed increased steadily throughout the years, indicating that some studies delved into the functional interactions between microbes and hosts. Nonetheless, average scores assigned to methodology exhibited little variation over the years (Supplementary Fig. S2b), due to the prevalent reliance on 16S rRNA amplicon sequencing in the majority of studies. Although cost-effective for taxonomic characterisation of microbiomes, this approach offers limited insights into the functional role of microbes in host interactions, which ultimately impedes our ability to unravel mechanistic processes.
Finally, we examined the relationship between the performance scores we measured and how researchers interpreted their findings. Among studies with an explicit hypothesis regarding microbial contributions to animal adaptation (comprising 40% of the total analysed), nearly all interpreted their results as evidence or potential evidence of adaptation (Fig. 2d). Even in the case of studies lacking an explicit hypothesis about microbial contributions to animal adaptation (60%), most researchers still interpreted their findings as indicative of potential or clear signs of adaptation facilitated by gut microbes. We did observe a positive correlation between the conclusiveness reported by the authors and the performance scores we assessed (LM; β = 0.029 points/year, F = 4.232, DFnom/den=1/107, p = 0.00, Fig. 2e), although the generally low performance scores suggest a tendency among researchers to make more assertive claims than probably warranted by their results.
Our systematic review indicates that the scientific community has not yet generated enough evidence to address the question of whether gut microorganisms confer wild vertebrates with enhanced acclimation and adaptation capacity. We identified clear limitations in the experimental and analytical approaches employed, which is probably a reflection of the complexity of the research required to tackle this scientific question. Addressing the microbe-conferred adaptability hypothesis requires a combination of expertises that include experimental study design, access to appropriate study systems and experimental facilities, capacity to measure fitness or proxies for fitness, and proficiency in complex molecular data generation and analysis. This is a mix of competences that calls for collaborative research. If we are to address this question, a shift from observational to experimental manipulation approaches is imperative, coupled with the modernisation of molecular techniques used to characterise microbiomes and host responses, leveraging the potential of multi-omic methodologies. Only then will we be able to quantify the intensity and breadth of the alleged contribution of gut microorganisms to vertebrate acclimation and adaptation capacity.