COS 94-5 - Patterns of microbial abundance and diversity relating to discontinuity in a non-perennial Kansas stream

Abstract: Non-perennial streams, flowing waters that dry ephemerally or intermittently, are widespread but understudied. Climate change is expected to increase global intermittent stream miles, making non-perennial streams increasingly important landscape features, with biological functions that impact downstream water quality. We seek to understand the consequences of intermittency on diversity and function of aquatic microbial communities, to inform understanding of how changing flow regimes will affect aquatic ecology and biogeochemistry. We sampled microbial communities at 50 locations within a single watershed at Kings Creek, near Manhattan, Kansas, USA, over one week in June 2021. Using quantitative PCR and Illumina metabarcoding, we examined microbial diversity and density in four microhabitats: sediments, epilithic biofilms, leaf litter, and the water column. We expected that connected locations would receive greater transport of nutrients from the landscape, which would support larger populations of microbes; and considered that transport of microorganisms from the landscape could either increase dispersal of terrestrial populations, increasing in-stream diversity, or that flow could homogenize populations among non-isolated locations, decreasing diversity. To measure connectivity, a particular challenge, we used a new method of analyzing directed acyclic graphs to estimate the level of network discontinuity at each sampling location: The “communication distance” metric. Communication distance (CD) is analogous to the resistance to transport of materials between locations in a stream network due to surface flow. A location with longer CD is less connected.

CD was correlated with stream microbial diversity and abundance; however, different microhabitats and taxa showed different trends. CD and bacterial abundance correlation coefficients (R2) ranged between 0.51 to 0.13, while fungal relationships were weaker (R2 between 0.21 to 0.02), and correlations were positive for all microhabitats except biofilms; i.e., only biofilm populations responded to flow connectivity in the manner predicted. The R2 of Shannon diversity and CD ranged between 0.28 to 0.00. Diversity and connectivity were positively related in sediments, but negatively related in waters; i.e., connectivity tended to promote diversity in stream waters, but reduce it in sediments. This is consistent with concurrent dispersal of terrestrial-derived microbial taxa in waters and flow homogenization of diversity in sediments. However, further analysis is needed to fully characterize the forces driving microbiome dynamics in non-perennial streams. Next steps include considerations of other connectivity metrics, evaluation of these patterns following temporal connectivity changes, and making more direct links with nutrient and material flux and residence time.