Virus diversity and activity is driven by snowmelt and host dynamics in a high-altitude watershed soil ecosystem

Abstract

Viruses, including phages, impact nearly all organisms on Earth, including microbial communities and their associated biogeochemical processes. In soils, highly diverse viral communities have been identified, with a global distribution seemingly driven by multiple biotic and abiotic factors, especially soil temperature and moisture. However, our current understanding of the stability of soil viral communities across time, and their response to strong seasonal change in environmental parameters remains limited. Here, we investigated the diversity and activity of environmental DNA and RNA viruses, including phages, across dynamics seasonal changes in a snow-dominated mountainous watershed by examining paired metagenomes and metatranscriptomes. We identified a large number of DNA and RNA viruses taxonomically divergent from existing environmental viruses, including a significant proportion of RNA viruses target fungal hosts and a large and unsuspected diversity of positive single-stranded RNA phages (Leviviricetes), highlighting the under-characterization of the global soil virosphere. Among these, we were able to distinguish subsets of active phages which changed across seasons, consistent with a "seed-bank" viral community structure in which new phage activity, for example replication and host lysis, is sequentially triggered by changes in environmental conditions. Zooming in at the population level, we further identified virus-host dynamics matching two existing ecological models: "Kill-The-Winner" which proposes that lytic phages are actively infecting abundant bacteria, and "Piggyback-The-Persistent" which argues that when the host is growing slowly it is more beneficial to remain in a lysogenic state. The former was associated with summer months of high and rapid microbial activity, and the latter to winter months of limited and slow host growth. Taken together, these results suggest that the high diversity of viruses in soils is likely associated with a broad range of host interaction types each adapted to specific host ecological strategies and environmental conditions. Moving forward, while as our understanding of how environmental and host factors drive viral activity in soil ecosystems progresses, integrating these viral impacts in complex natural microbiome models will be key to accurately predict ecosystem biogeochemistry.