Abstract
Stand structural diversity, encompassing spatial and non-spatial dimensions, is a key indicator of forest carbon storage, yet its relative impacts on multiple carbon pools remain unclear. Additionally, whether structural diversity consistently influences carbon storage across overstory, understory, and soil layers is uncertain. This study examined carbon storage dynamics across 13 secondary Masson pine forests within the Fengyang Mountain Nature Reserve. Principal component analysis was used to classify the stands into three types based on their spatial and non-spatial structural diversity: Type I (high spatial and high non-spatial diversity), Type II (high spatial but low non-spatial diversity), and Type III (low spatial and low non-spatial diversity). Total carbon storage was highest in Type I, while carbon storage in the understory layers was lowest in this type. Spatial structural diversity had a stronger influence on carbon storage than non-spatial diversity, with the uniform angle index primarily affecting overstory carbon storage, and the crowding index influenced understory carbon storage. Random forest analysis identified biomass and structural diversity as major predictors of carbon storage. Partial least squares path modeling revealed that spatial structural diversity indirectly increased overstory and soil carbon storage, but reduced understory carbon storage by modulating biomass. Our results highlight that spatial structural diversity is a dominant driver of carbon storage in forest ecosystems, with contrasting effects on overstory, understory, and soil layers, underscoring its critical role in regulating forest carbon dynamics.