Complex microbial communities inhabit nearly every macro-organism. For plants, the microbes that live on and within their tissues (endophytes) can have significant effects on health, productivity, and stress tolerance. Moreover, the many microbial species that coexist within the plant host do not act in isolation, but rather interact with one another and their host in myriad ways. Understanding endophyte-plant associations requires insight into the structure of species interactions within endophytic communities.
We randomly isolated twenty microbes (ten fungal and ten bacterial) from leaves of twelve Andropogon gerardii plants, half of which had undergone sustained nutrient supplementation in the form of added Nitrogen, Phosphorous, Potassium, and trace nutrients. For each isolate, we categorized growth on 95 different resources using Biolog SF-P2 MicroPlates, and used these data to calculate niche overlap between each pair of isolates. These values were compiled to construct a resource-competition network for each community.
Results/Conclusions
There were significant differences in degree distribution when comparing bacterial to fungal isolates and control to nutrient-supplemented communities. In particular, bacterial isolates tended to have higher in-degree (other isolates substantially overlapping their niche space) while Fungi tended to have higher out-degree (substantially overlapping other isolates' niche spaces). Between treatments, Fungi interaction patterns were accentuated: increasing the number of other isolates they interacted with, while bacterial isolates were unaffected. Additionally, within Kingdom (i.e. bacterial-bacterial or fungal-fungal) interaction patterns differed from those between Kingdoms, largely reflecting the generally superior competitive abilities of fungi over bacteria. We find that isolates tend to cluster with other members of the same Kingdom especially in the control communities.
While prior research has documented functional consequences of nutrient supplementation on endophytic communities, we saw relatively few differences in network-structural metrics between treatments. Considering metrics spanning the range of local-to-global scope, most differences were consequences of the observed differences in degree distribution, suggesting a relatively minor role for higher-order-interactions in these communities.
Overall, these results show that nutrient supplementation can alter interaction networks among microbial endophytes. Further research is exploring the extent to which interaction networks reflect differences in nutrient-structured community assembly processes versus co-evolution within plants, and the implications of altered network structure for the plant host.