Objective  Microbial biomass carbon and nitrogen (MBC and MBN) in soil at the man-made forests on northern China hilly lands were studied.

  Method  Three man-made forests that contained solely Robinia pseudoacacia, Quercus variabilis or Platyclodus orientalis trees at the Xiaolangdi Ecosystem Positioning Station of the Yellow River in Jiyuan, Henan were selected for the study. Soil samples in the layers of undecomposed litter (L), semi-decomposed/humic litter (F/H), and litter-free 0–10 cm and 10–20 cm depths were collected to determine the MBC, MBN, MBC/MBN ratio, and physiochemical properties. Correlations of the soil microbial biomass (SMB) with different tree species and soils were analyzed.

  Result  (1) The highest total carbon (TC) in the soil that contained plant litter at the 3 forests was found in the F/H layer, and the greatest total nitrogen (TN) in the L layer. The TC and TN of the forests ranked in the order of F/H layer＞L layer＞0–10 cm and 10–20 cm soils with no fallen wastes (P＜0.05). DOC was higher in the layers with litter than without, but opposite on DON . The DOC in the littered soil was significantly higher at the Q. variabilis forest than the R. pseudoacacia forest, and no significant difference between the two and the P. orientalis forest (P＞0.05). (2) In the soil layers with tree litter, MBC was higher in the F/H than in the L layer with a ranking of R. pseudoacacia＞Q. variabilis＞P. orientalis among the 3 forests. In the two layers of litter-free mineral soil, MBC was higher in the depth of 0–10 cm than in the deeper layer. It ranked among the different forests as Q. variabilis＞R. pseudoacacia＞P. orientalis with no significant difference between the Q. variabilis and R. pseudoacacia forests (P＞0.05). The MBNs of both the littered and litter-free soil at the forests were Q. variabilis＞R. pseudoacacia＞P. orientalis, while Q. variabilis generated the highest MBN that was significantly different from P. orientalis (P＜0.05). (3) The MBC/MBN ratios of the L and F/H layers did not differ significantly, whereas those of the mineral soils tended to increase with the depth. The 0–10 cm litter-free soil did not differ significantly on the ratio among the forests (P＞0.05). However, that of the 10–20 cm soil at the R. pseudoacacia forests were significantly higher than that at the P. orientalis forest (P＞0.05), but no significant difference between them and that at the Q. variabilis forest (P＜0.05). (4) Significant correlations existed between the MBC and the TC and TN at the 3 forests. The correlation coefficient of 0.959 between MBC and TC of Q. variabilis was the highest among all. The C/N ratios of R. pseudoacacia and Q. variabilis negatively, while the C/N of P. orientalis positively, correlated with TC and TN, with correlation coefficients of 0.512 and 0.524, respectively.

  Conclusion  Conducive to environmental carbon and nitrogen cycling, Q. variabilis appeared to be a prudent choice of plant for the ecological restoration in the low mountainous areas in northern China.