Background Afforestation is one of the primary measures for enhancing carbon sinks in terrestrial ecosystems. However, since the large-scale afforestation efforts in Yudaokou, the effectiveness of these projects in carbon sequestration, and whether different tree species and afforestation modes lead to significant differences remains unclear. This study investigates two main tree species and two afforestation modes in Yudaokou, aiming to clarify the impact of afforestation projects on carbon storage in the sandy ecosystems of the Bashang region, North Hebei, thereby providing a theoretical basis for vegetation construction in the area.

Methods This study investigated ecosystem carbon density under different afforestation species and modes, focusing on Pinus sylvestris var. mongolica and Larix principis-rupprechtii plantations under two silvicultural modes (with and without forest-grass intercropping) in Yudaokou Forest Farm. Non-afforested lands, categorized into CKI, CKII, and CKIII based on decreasing biomass, served as controls. Biomass data were obtained through field surveys and allometric equations, while soil organic carbon was determined using the potassium dichromate external heating method. Data were analyzed with Excel 2021 and one-way ANOVA in SPSS 25. Figures were generated in Origin 2021, and significant differences between treatments were assessed using Duncan’s test (P < 0.05).

Results 1) In all plantation ecosystems, carbon density was predominantly contributed by the tree and soil layers, with the herb and litter layers accounting for relatively minor proportions. 2) The tree layer carbon density was significantly higher (P < 0.05) in non-forest-grass intercropped P. sylvestris var. mongolica (NFGP) and L. principis-rupprechtii (NFGL) plantations (36.70 t/hm² and 52.94 t/hm², respectively) than in their forest-grass intercropped counterparts (FGP (P. sylvestris var. mongolica): 19.41 t/hm²; FGL (L. principis-rupprechtii): 32.54 t/hm²). Moreover, L. principis-rupprechtii plantations under both modes had significantly greater tree layer carbon density than those of P. sylvestris var. mongolica (P < 0.05). 3) Soil carbon density was in the order: CKI (127.56 t/hm²) > NFGL (110.55 t/hm²) > FGL (79.42 t/hm²) > FGP (76.85 t/hm²) > NFGP (63.07 t/hm²) > CKII (56.06 t/hm²) > CKIII (31.78 t/hm²). No significant difference was observed between CKI and NFGL, but both were significantly higher than all other types (P < 0.05). That in CKIII was significantly lower than those of all types except CKII (P < 0.05). 4) Ecosystem carbon density decreased in the order: NFGL (178.49 t/hm²) > CKI (132.70 t/hm²) > FGL (119.48 t/hm²) > NFGP (109.88 t/hm²) > FGP (102.01 t/hm²) > CKII (59.89 t/hm²) > CKIII (34.22 t/hm²). That in NFGL was significantly higher than those of all others, while that in CKII and CKIII were significantly lower (P < 0.05). No significant differences were detected among the remaining types (P > 0.05). 5) Since the implementation of the “Three Saihanba Afforestation Projects”, an additional carbon storage of 115.54 × 10⁴ tC has been achieved in Yudaokou through artificial afforestation.

Conclusions Afforestation significantly enhances ecosystem carbon density in the sandy lands of Bashang region, North Hebei. L. principis-rupprechtii with non-forest-grass intercropping resulted in the highest carbon storage. However, on sites with better initial vegetation and higher soil organic carbon, using P. sylvestris var. mongolica or adopting forest-grass intercropping patterns may lead to a reduction in ecosystem carbon storage.