The Abundance of Microbial "Specific Species" Significantly Influences Downed Log Respiration

Abstract

Respiration in downed logs primarily results from microbial metabolic activities. However, the mechanisms by which environmental factors regulate these activities remain unclear. Consequently, current models for carbon release from downed logs exhibit significant deviations. Therefore, this study investigates the same forest plots at 2013 and 2022 to measure the respiration rate (Rlog), temperature, moisture, chemical element content (Clog), and microbial community structure of downed logs from Quercus aliena var. acuteserrata and Pinus tabulaeformis at various decomposition stages (DSs). The aim is to elucidate the complex causal relationships and interdependencies among factors controlling Rlog. Key findings include: 1) Rlog increases significantly with decomposition, with Q. aliena var. acuteserrata exhibiting a significantly higher average Rlog (4.12 μmolCO2·m-2·s-1) than P. tabulaeformis (3.77 μmolCO2·m-2·s-1); 2) significant differences in physicochemical properties of downed logs between 2013 and 2022, controlled by tree species and DSs, and the contents of moisture, N, P, Ca and Mg were significantly increased with the increase of DSs, while the content of K was significantly decreased (p < 0.05); 3) In both 2013 and 2022, the dominant microorganisms in downed logs at different DSs included Proteobacteria, Firmicutes, Actinobacteria (bacteria) and Acidobacteria, Bacteroidetes (fungi); 4) The species with significant abundance difference between the two years ("specific species"), including Basidiomycota, Proteobacteria, Cyanobacteria and Chlamydiota, etc, were closely correlated with Rlog, although no significant correlation was found between microbial diversity indices and Rlog; and 5) moisture emerges as the most influential factor on Rlog, followed by Ca, Mg, K, P, N contents, and temperature, particularly in later DSs (IV, V). These results suggest that future research on Rlog mechanisms should focus on microorganisms with significant temporal and spatial abundance differences in downed logs. Furthermore, carbon emission models for downed logs should consider the combined effects of moisture and temperature on Rlog, along with the mediating role of wood characteristics (species, DS, and Clog), to improve the accuracy of future Rlog predictions.

Keywords: Downed log microorganism, Downed log physicochemical property, Random forest model, Structural equation model