Responses of soil carbon storage to tending thinning and rapid prediction based on NIRS-BPNN in larch-birch natural secondary forests.

INTRODUCTION

Heilongjiang Province ranks third in China in forest area, with over half of its forests being natural secondary forests. The province has fertile soil and plays a crucial role in China's forest ecosystem. Nevertheless, historical indiscriminate logging has resulted in degraded forest conditions, reduced productivity, and compromised carbon sequestration capacity.

METHODS

This research focused on the larch-birch natural secondary forest in the Greater Khingan Mountains, examining the soil twelve years after thinning. Six thinning treatments (0%, 16.7%, 25.5%, 34.4%, 49.6%, 59.9%) were compared to assess soil physical and chemical properties, soil carbon storage, and to elucidate the associations between thinning intensity and soil carbon storage dynamics. A soil organic carbon (SOC) content prediction model was developed using near-infrared spectroscopy and a Back Propagation neural network (BPNN) .

RESULTS

The findings indicate that thinning intensity significantly alters soil carbon storage by affecting key parameters such as soil moisture and organic carbon content. Specifically, a 34.4% thinning intensity significantly enhanced soil moisture content (56.67% vs. 33.02% in control, p < 0.05) and SOC content (66.17 g/kg vs. 32.03 g/kg in control, p < 0.05), leading to a substantial increase in total soil carbon storage (83.80 t/ha vs. 43.99 t/ha in control, p < 0.05). The NIRS-BPNN model achieved a prediction set correlation coefficient of Rp = 0.9293 (R2p = 0.8635, RMSEP = 0.0064, RPD = 2.8), demonstrating good predictive performance.

DISCUSSION: This research elucidates the responses of soil carbon storage to tending and thinning, providing a scientific foundation for the management of secondary forests. The NIRS-BPNN model offers a novel, rapid, and non-destructive approach for the large-scale assessment of SOC content in secondary forests, holding considerable theoretical and practical significance for forest carbon monitoring and climate change mitigation.