PS 31-118 - Plant growth stage and soil moisture modify nitrogen cycle gene co-occurrence networks in East Africa

Background/Question/Methods

Soil nitrogen fertility in tropical soils is limited by a broad range of environmental and edaphic factors, including weathering and erosion from extreme weather events. Climate-smart interventions to conserve soil resources and limit greenhouse gas emissions from agriculture often focus on cropping systems that build organic matter through perennial root systems. Microbial nitrogen (N) cycling pathways are largely responsible for producing forms of N that are available for plant uptake or lost from the system as gas or leachate. The temporal dynamics of microbial N pathways in rainfed tropical agroecosystems are not well defined, even though they are critical to understanding the potential impact of proposed conservation strategies. Our objectives were to 1) determine the extent to which climate-smart perennial cropping systems affect N cycle functional gene abundance and 2) characterize temporal changes in N cycle gene co-occurrence across a seasonal gradient. Soil samples (n=334) were collected from two replicated field trials in Rwanda consisting of perennial forage grasses grown with or without legumes between September 2020 and March 2021. Total genomic DNA was extracted from fresh soil samples and used to quantify twenty N cycle functional gene targets on a high-throughput qPCR platform.

Results/Conclusions

Neither location nor plant treatment impacted the structure of microbial N functional communities, thus, we combined data from both locations and separated by soil collection date to analyze gene co-occurrence networks across time. N cycle gene co-occurrence networks constructed using Spearman’s correlation coefficients (rho >0.75 and p< 0.01) had the greatest density in the dry season. During similar stages of plant growth (anthesis or early growth following harvest), network density, or gene co-occurrence, decreased by 52.5% in the rainy season compared to the dry season. Changes to network density were not linear, with network density varying by up to 272% within the same season depending on plant growth stage (anthesis or early growth). We expect ongoing analyses to identify ‘keystone’ functional genes and co-occurrence relationships that remained consistent across sampling date. Ultimately, this work will enhance our understanding of how climate-smart interventions may better target microbial populations of interest. Our results also imply that resource scarcity in the dry season (e.g., soil moisture and soluble N) strengthens the co-occurrence of distinct functional groups involved in the N cycle. Facilitation of N products and niche-sharing may therefore be important in explaining seasonal shifts in N cycling community dynamics in tropical agroecosystems.