Anaerobic gut fungal communities in marsupial hosts

Abstract

The anaerobic gut fungi (AGF) inhabit the alimentary tracts of herbivores. In contrast to placental mammals, information regarding the identity, diversity, and community structure of AGF in marsupials is extremely sparse. Here, we characterized AGF communities in 61 fecal samples from 10 marsupial species belonging to four families in the order Diprotodontia: Vombatidae (wombats), Phascolarctidae (koalas), Phalangeridae (possums), and Macropodidae (kangaroos, wallabies, and pademelons). An amplicon-based diversity survey using the D2 region of the large ribosomal subunit as a phylogenetic marker indicated that marsupial AGF communities were dominated by eight genera commonly encountered in placental herbivores (Neocallimastix, Caecomyces, Cyllamyces, Anaeromyces, Orpinomyces, Piromyces, Pecoramyces, and Khoyollomyces). Community structure analysis revealed a high level of stochasticity, and ordination approaches did not reveal a significant role for the animal host, gut type, dietary preferences, or lifestyle in structuring marsupial AGF communities. Marsupial foregut and hindgut communities displayed diversity and community structure patterns comparable to AGF communities typically encountered in placental foregut hosts while exhibiting a higher level of diversity and a distinct community structure compared to placental hindgut communities. Quantification of AGF load using quantitative PCR indicated a significantly smaller load in marsupial hosts compared to their placental counterparts. Isolation efforts were only successful from a single red kangaroo fecal sample and yielded a Khoyollomyces ramosus isolate closely related to strains previously isolated from placental hosts. Our results suggest that AGF communities in marsupials are in low abundance and show little signs of selection based on ecological and evolutionary factors.IMPORTANCEThe AGF are integral part of the microbiome of herbivores. They play a crucial role in breaking down plant biomass in hindgut and foregut fermenters. The majority of research has been conducted on the AGF community in placental mammalian hosts. However, it is important to note that many marsupial mammals are also herbivores and employ a hindgut or foregut fermentation strategy for breaking down plant biomass. So far, very little is known regarding the AGF diversity and community structure in marsupial mammals. To fill this knowledge gap, we conducted an amplicon-based diversity survey targeting AGF in 61 fecal samples from 10 marsupial species. We hypothesize that, given the distinct evolutionary history and alimentary tract architecture, novel and unique AGF communities would be encountered in marsupials. Our results indicate that marsupial AGF communities are highly stochastic, present in relatively low loads, and display community structure patterns comparable to AGF communities typically encountered in placental foregut hosts. Our results indicate that marsupial hosts harbor AGF communities; however, in contrast to the strong pattern of phylosymbiosis typically observed between AGF and placental herbivores, the identity and gut architecture appear to play a minor role in structuring AGF communities in marsupials.

Keywords: anaerobic fungi; community structure; marsupials.

Fig 1

AGF community composition in the samples studied. The phylogenetic tree showing the relationship between animals was downloaded from timetree.org and modified to include very short branch length between samples from the same animal species. Branches are color coded by family as shown in the figure legend. Tracks to the right of the tree depict the species, gut type, habitat, and nutritional type of the animals studied as shown in the figure legend. AGF genera percentage of abundances is shown to the right of the tracks, with genera with <1% relative abundance grouped in “others.”

Fig 2

(A) AGF percentage of abundance shown for all samples studied, including families, species, gut type, habitat, and nutritional type. AGF genera with <1% relative abundance are grouped in “others.” (B) Total number of AGF genera (left) and number of AGF genera with >1% relative abundance (right) identified per sample. Samples names are shown on the X-axis (names match those in Fig. 1), and samples are grouped by the animal family as depicted in the figure legend. (C) Relationship between occurrence (number of samples) and average relative abundance of each of the 85 genera encountered in this study. The number of samples in which the genera was identified is shown on the X-axis. Average percentage of abundance across samples is plotted on the Y axis in a logarithmic scale to show genera present below 1% abundance. The eight mostly abundant genera (Neocallimastix, Orpinomyces, Caecomyces, Cyllamyces, Piromyces, Khoyollomyces, Anaeromyces, and Pecoramyces) are shown with a red border. Abundance-occurrence plots are shown for all samples studied, as well as for each of the three families with >5 animals, as depicted above each figure.

Fig 3

Alpha-diversity patterns. (A) Box and whisker plots showing the distribution of Shannon diversity index for different families, species, gut types, habitats, and nutritional types of the animals studied. Results of the Kruskal-Wallis test are shown in the table to the right. (B) Box and whisker plots showing the distribution of Shannon diversity index for animals species, color coded by their gut type, in comparison with foregut and hindgut placental animal representatives. (C) Results of Dunn post hoc tests for pairwise infraclass gut-type comparisons. ****P < 0.0001. ns, not significant.

Fig 4

AGF community assembly in marsupial hosts. (A) Box and whisker plots showing the distribution of the bootstrapping results (n = 1,000) for the levels of stochasticity in AGF community assembly calculated as NST. Results compare different animal families (top row), animal species (second row), gut type (third row), habitat (fourth row), and nutritional type (fifth row). Two NSTs were calculated: the abundance-based Bray-Curtis index (left) and the incidence-based Jaccard index (right). Wilcoxon test, P value: **, 0.001< p < 0.01; ****, p < 0.0001. Details about how these results were obtained are explained in Materials and Methods. (B) The percentages of the various deterministic and stochastic processes shaping AGF community assembly of the total data set, and when subsetting for different animal families, species, gut types, habitats, and nutritional types. ns, not significant; NST, normalized stochasticity ratio.

Fig 5

Ordination plots based on AGF community structure in the studied hosts. (A) Principal coordinate analysis ordination plots based on AGF community structure were constructed using the phylogenetic similarity-based weighted Unifrac index. Samples are color coded by animal family, species, nutritional type, habitat, and gut type as shown in the legend on the right-hand side, while the shape depicts the gut type as shown in the figure legend on top. Ellipses encompassing 95% of variance are shown for each of the factors and are color coded similar to the samples. Results of PERMANOVA test for partitioning the dissimilarity among the sources of variation are shown in the table to the right. The F-statistic R² depicts the fraction of variance explained by each factor, while the P value depicts the significance of the host factor in affecting the community structure. (B) AGF community structure in marsupial hosts in comparison to placental mammals. Variance is shown for the four subcategories (foregut Marsupialia, foregut Placentalia, hindgut Marsupialia, and hindgut Placentalia). Results of PERMANOVA test for partitioning the dissimilarity are shown in the table to the right.

Fig 6

AGF load in the 43 marsupial samples examined using quantitative PCR. Boxplots showing the distribution of AGF load in the three marsupial families (A) seven marsupial species (B) and two gut types (C). (D) Comparison to AGF load in the 43 marsupial hosts to 40 placental counterparts. Boxplots in panel E show the distribution of AGF load in the 43 marsupial hosts (infraclass Marsupialia) versus the 40 placental hosts (infraclass Placentalia). **Wilcoxon t-test, P value = 0.0012.

Fig 7

Assessment of the phylogenetic position of the newly obtained isolates from a kangaroo using D1/D2 LSU as a phylogenetic marker. The tree was constructed using the maximum likelihood approach implemented in FastTree. Scale bar indicates the number of substitutions per site. Bootstrap values are shown for nodes with >70% support as gray spheres, where the size of the sphere is proportional to the bootstrap value. The four previously suggested Neocallimastigomycota families are color coded as shown in the figure legend. New isolates are identified as Khoyollomyces ramosus and are shown in red text.