Fungal microbiomes are determined by host phylogeny and exhibit widespread associations with the bacterial microbiome

Abstract

<jats:title>ABSTRACT</jats:title><jats:p>Interactions between hosts and their resident microbial communities are a fundamental component of fitness for both agents. Despite a recent proliferation of research on interactions between animals and their associated bacterial communities, comparative evidence from fungal communities is lacking, especially in natural populations. This disparity means knowledge of host-microbe interactions is biased towards the bacterial microbiome. Using samples from 49 species from eight metazoan classes, we demonstrate that the ecological distance between both fungal and bacterial components of the microbiome shift in tandem with host phylogenetic distance. Though so-called phylosymbiosis has been shown in bacterial communities, we extend previous knowledge by demonstrating that the magnitude of shifts in fungal and bacterial community structure across host phylogeny are correlated. These data are indicative of coordinated recruitment by hosts for specific suites of microbes, and potentially selection for bacterial-fungal interactions across a broad taxonomic range of host species. Using co-occurrence networks comprising both microbial groups, we illustrate that fungi form a critical component of microbial interaction networks, and that the strength and frequency of such interactions vary across host taxonomic groups. Collectively these data indicate fungal microbiomes may play a key role in host fitness and suggest an urgent need to study multiple agents of the animal microbiome to accurately determine the strength and ecological significance of host-microbe interactions.</jats:p>

Leveraging Experimental Strategies to Capture Different Dimensions of Microbial Interactions

“…Different mathematical approaches can be used for network inference and microbial community modeling, their challenges and applications have been reviewed in ( Dohlman and Shen, 2019 ; Matchado et al, 2021 ; Zaramela et al, 2021 ). Multiple cross-sectional and longitudinal studies used co-occurrence or co-abundance to generate association networks from a variety of environments (e.g., Faust et al, 2015 ; Ke et al, 2019 ; Zotta et al, 2019 ; Chen et al, 2020a ; Harrison et al, 2020 ; Moran-Ramos et al, 2020 ). A meta-network encompassing samples at the earth scale has even been constructed ( Ma et al, 2020 ).…”