Mutual Information between Streamwater Microbiome and Hydrology

This record includes software and data used to compute shared information between streamwater microbial taxa and hydrologic metrics for the article:

URycki, D. R., Bassiouni, M., Good, S. P., Crump, B. C., & Li, B. (2022). The streamwater microbiome encodes hydrologic data across scales. Science of The Total Environment, 157911. https://doi.org/10.1016/j.scitotenv.2022.157911

Abstract: Many fundamental questions in hydrology remain unanswered due to the limited information that can be extracted from existing data sources. Microbial communities constitute a novel type of environmental data, as they are comprised of many thousands of taxonomically and functionally diverse groups known to respond to both biotic and abiotic environmental factors. As such, these microscale communities reflect a range of macroscale conditions and characteristics, some of which also drive hydrologic regimes. Here, we assess the extent to which streamwater microbial communities (as characterized by 16S gene amplicon sequence abundance) encode information about catchment hydrology across scales. We analyzed 64 summer streamwater DNA samples collected from subcatchments within the Willamette, Deschutes, and John Day river basins in Oregon, USA, which range 0.03–29,000 km2 in area and 343–2334 mm/year of precipitation. We applied information theory to quantify the breadth and depth of information about common hydrologic metrics encoded within microbial taxa. Of the 256 microbial taxa that spanned all three watersheds, we found 9.6 % (24.5/256) of taxa, on average, shared information with a given hydrologic metric, with a median 15.6 % (range = 12.4–49.2 %) reduction in uncertainty of that metric based on knowledge of the microbial biogeography. All of the hydrologic metrics we assessed, including daily discharge at different time lags, mean monthly discharge, and seasonal high and low flow durations were encoded within the microbial community. Summer microbial taxa shared the most information with winter mean flows. Our study demonstrates quantifiable relationships between streamwater microbial taxa and hydrologic metrics at different scales, likely resulting from the integration of multiple overlapping drivers of each. Streamwater microbial communities are rich sources of information that may contribute fresh insight to unresolved hydrologic questions.

This record can also be found on Github: https://github.com/uryckid/Microbial-Mutual-Information