Slowing Wind Increases Soil Carbon Emissions in a Fragmented Subtropical Forest: a Study Combining Field and Model Experiments

Hao-Hao Wu,Ling-Ling Zhang,Ai-Min Liao,Wu,Xiao-Ming Lai,Hai-Xia Zhang,Ye Xia,Yang Cao,Zi-Chun Zhu,Zhen-Ming Ji,Cong-Sheng Fu
DOI: https://doi.org/10.1007/s11104-024-06916-z
IF: 4.993
2024-01-01
Plant and Soil
Abstract:Worldwide wind speed decline (i.e., global terrestrial stilling) and rapid forest fragmentation have been widely documented. The internal hydrothermal conditions within fragmented forests are particularly susceptible to wind speed changes, potentially influencing carbon emissions through soil respiration. However, the qualitative and quantitative effects of the wind speed decline on soil carbon emissions in fragmented forests, as well as corresponding underlying mechanisms, remain highly uncertain. We comprehensively investigated the influences of wind speed changes on soil respiration in a fragmented subtropical forest in Eastern China, based on field experiments and model experimental simulations using the Community Land Model version 5 (CLM5). Wind speed decreased by 0.09 m s−1 decade−1 from late 1950s to early 2020s at the fragmented forest site and resulted in an increase in soil respiration of 4.14 g C m−2 decade−1, consisting of a 5.83 g C m−2 decade−1 increase in heterotrophic respiration and a 1.85 g C m−2 decade−1 decrease in autotrophic respiration. Although CLM5 accurately simulated seasonal soil respiration dynamics, it clearly underestimated the wind speed decline-induced increases in soil respiration. Observations indicated that the negative impact of wind speed on soil carbon emissions was driven by the desiccating microclimate created by dry external air brought in by the wind, while CLM5 proposed wind-induced cooling caused by increased latent heat loss, i.e., primarily canopy transpiration. This study quantitatively evaluates for the first time the impact of declining wind speed on soil respiration in fragmented forests of eastern China, providing valuable insights for future model improvements.
What problem does this paper attempt to address?