Contrasting seasonal dynamics and drivers of soil extracellular enzyme activities in two temperate forest-steppe ecotones

Soil extracellular enzyme activities (EEAs) are crucial in ecosystem nutrient cycling. Their seasonal dynamics, particularly beyond the growing season and across different regions, are poorly understood. This study compared the seasonal patterns and drivers of EEAs in two temperate forest-steppe ecotones with contrasting resource availability (resource-poor Saihanba vs. resource-rich Saihanwula).

Seasonal soil sampling was conducted across 23 sites. Potential activities of carbon (β-glucosidase, BG), nitrogen (β-1,4-N-acetylglucosaminidase, NAG), and phosphorus (acid phosphatase, AP) cycling enzymes were measured. Soil properties, microbial biomass, and community composition (via PLFAs) were analyzed. Drivers were identified using ANOVA and structural equation modeling (SEM).

Integrating winter data revealed divergent seasonal enzyme dynamics between two regions. Contrary to expectations, environmental filtering overrode vegetation type in structuring enzyme profiles. Compared to growing-season levels, winter BG and AP activities remained unexpectedly high in the resource-poor Saihanba region, but declined significantly in resource-rich Saihanwula. SEM and correlation analyses indicated divergent regulatory mechanisms: abiotic factors (particularly dissolved organic carbon , dissolved nitrogen and moisture) primarily drove enzyme activities in Saihanba, whereas biotic factors (microbial biomas carbon: microbial biomass nitrogen and Fungi :Bacteria ratio) dominated in Saihanwula.

This distinct winter pattern suggests a mechanism where microbes in nutrient-poor sites reallocate nitrogen from biomass to extracellular enzyme pools to alleviate non-growing season nutrient limitations. These findings highlight environmental filtering's role and emphasize that ignoring winter metabolic processes may cause an underestimation of annual soil organic matter decomposition in temperate ecosystems.