Introduction
High-latitude tundra is one of the most vulnerable ecosystems to climate warming (Myers-Smith et al., 2015), wherein permafrost thawing and changes of vegetation patterns have been observed during past decades (Schuur, Crummer, Vogel, & Mack, 2007; Sturm, Racine, & Tape, 2001; Xue et al., 2016). Given the fact that permafrost soil of high-latitude tundra contains up to half of the global soil carbon (C) storage (Tarnocai et al., 2009), warming-induced permafrost C loss to atmosphere could be one of the most severe concerns for accelerating climate warming by greenhouse gas effects (Natali et al. 2011; Schuur et al. 2009). In a warmer world, vast C may be released to the atmosphere as a result of stimulated microbial activities to accelerate permafrost C degradation (Mann et al., 2015; Zhou et al., 2012).
As the major soil C decomposers, fungi may affect the C stability of high-latitude tundra considerably (Christiansen et al., 2016). For instance, saprotrophic fungi are important regulators of soil nutrient cycling via organic C decomposition, especially lignin (Buckeridge, Banerjee, Siciliano, & Grogan, 2013; Yuste et al., 2011). In winter, fungi are the largest proportion of soil microbial biomass and predominantly affect wintertime C degradation (Buckeridge et al., 2013; Schadt, Martin, Lipson, & Schmidt, 2003). Additionally, mycorrhizal fungi form mutually beneficial associations to facilitate the growth of arctic plants, especially those warming-responsive plant species such asSalix spp., and Betula nana (Clemmensen, Michelsen, Jonasson, & Shaver, 2006), and thus can regulate C stability by affecting primary productivity. Therefore, responses of fungal communities to climate warming would be determinative for the C fate in tundra ecosystems.
Using PCR-based marker gene sequencing, several in situ field studies have demonstrated that experimental warming changes taxonomic composition and environmental group abundances of fungal communities in tundra soil, with increased diversity of saprotrophic fungi but decreased diversity of ectomycorrhizal fungi (Morgado et al., 2016; Mundra et al., 2016; Semenova et al., 2016). However, such taxonomic information provides little insight into the relationship between fungal functions and ecological processes. Functional gene analysis has been applied to quantify functional genes responsible for key ecosystem processes. The main findings of previous studies (Levy-Booth, Prescott, & Grayston, 2014; Yergeau, Kang, He, Zhou, & Kowalchuk, 2007), i.e., close relationships between functional gene abundances and ecosystem processes, have provided important insights on the roles of microbial communities in ecosystem functioning.
In addition to taxonomic composition, it is possible that fungal functional genes, serving as markers of potential functional capacities, are sensitive to climate warming. The justification includes 1) Temperature affects the syntheses of cryoprotective compounds essential for maintaining fungal growth and functions in cold environments (Weinstein, Montiel, & Johnstone, 2000); 2) Temperature sensitivity of soil heterotrophic respiration (measured as Q10) is widespread (Malcolm, Lopez-Gutierrez, Koide, & Eissenstat, 2008), implying that C degradation capacities of soil microorganisms including fungi are variable; and 3) Plant litter input would be higher under warming (Rinnan, Michelsen, Baath, & Jonasson, 2007), which may stimulate fungi-mediated C degradation.
As the change of functional gene composition can inevitably affect interactions among microbial members, it is important to elucidate complex interactions among microbial functional genes or species containing these genes in order to understand ecosystem functioning (Wu et al., 2016). However, it remains challenging due to myriad members in microbial communities and difficulties in cultivating individual species. In recent years, network analysis has been widely used to reveal potential interactions among co-existing microbial members (Banerjee et al., 2016; Banerjee, Schlaeppi, & van der Heijden, 2018; de Vries et al., 2018). Accumulating evidence shows that network properties can reflect responses of microbial communities to environmental disturbances (de Vries et al., 2018; De Vries et al., 2012), as drought treatment in grassland mesocosms led to higher connectivity and lower modularity of soil bacterial networks than fungal networks, which suggested lower stability of bacteria than fungi under drought (de Vries et al., 2018).
In the present work, we analyzed the effects of winter warming on fungal communities in the active layer of the Alaskan tundra, via 28S gene amplicon sequencing and GeoChip technologies. Through installing snow fences to increase snow cover thus winter soil temperature, we tested the following hypotheses: 1) Warming would affect fungal community composition and potential interactions; 2) Fungal C degradation capacities, represented by the abundances of C degradation genes, would be increased by warming; 3) Increased fungal C degradation capacities could explain the increase of ecosystem respiration (R eco), as recently witnessed in several studies (Liu et al., 2015; Trivedi et al., 2016; Zhao et al., 2014).