Objective This study aims to quantify the belowground biomass (BGB) of trees and shrubs in various permafrost forest types, clarify their spatial distribution patterns, and analyze interspecific differences in BGB and their responses to stand age, topography, and permafrost type.

Method Based on field quadrat survey data, tree BGB was estimated using allometric equations combined with the root-to-shoot ratio method, whereas shrub BGB was determined via direct excavation. The BGB characteristics of trees and shrubs across different quadrats were statistically analyzed.

Result (1) The mean tree BGB in the permafrost region of the Da Xing'an Mountains was 6.10 t/ha, while the mean measured shrub BGB was 0.64 t/ha. (2) Tree and shrub BGB exhibited divergent responses to stand age: Tree BGB increased significantly with age class, following the order mature forest > near-mature forest > middle-aged forest > young forest (P < 0.001), whereas shrub biomass was not significantly affected by plot age group (P = 0.33). (3) Slope gradient significantly influenced both tree and shrub BGB, albeit with contrasting patterns: Tree BGB followed the order steep slope > gentle slope > flat slope (P < 0.001), while shrub BGB followed the order gentle slope > steep slope > flat slope (P < 0.001). (4) Permafrost type had no significant overall effect on tree BGB, but the effect was species-dependent, with both larch and birch showing significantly lower values in the sparse island permafrost zone. In contrast, shrub BGB decreased significantly with permafrost degradation (P < 0.001), specifically in the order: predominantly continuous permafrost zone > extensive but discontinuous permafrost zone > sparse isolated permafrost zone.

Conclusion BGB distribution is primarily influenced by slope, stand age, and permafrost type, with marked differences in the response patterns of trees and shrubs. Notably, heterogeneity in root-to-shoot ratio sources across species and methodological inconsistencies between tree and shrub biomass estimation may introduce uncertainties, warranting future validation and standardization of localized root-to-shoot equations. This study addresses critical gaps in belowground biomass research and provides scientific support for regional carbon stock estimation and the implementation of China's "Dual Carbon" strategy.