为改善东部草原露天矿区黑黏土黏性强、透气性差、土壤养分有效性低的现况，设2个水平接种AM真菌处理:接种丛枝菌根(M)和不接种菌根的对照组(CK)。同时将黏土设置了 4 种不同配比处理:沙土(S)、黏土(N)，沙土与黏土按质量1∶1配比(S∶N=1∶1)、沙土与黏土按质量3∶1配比(S∶N=3∶1)。利用高光谱遥感技术监测在不同基质下接种丛枝菌根真菌对植物生长发育的影响，对比不同处理下典型光谱特征参数的差异，并筛选与叶绿素含量相关性高的特征参数，以此为自变量利用逐步回归和BP神经网络的建模方法反演叶绿素含量，为微生物复垦效应监测奠定理论基础。结果表明:接种AM真菌后玉米叶绿素含量较对照升高，通过对比“三边”参数可以看出接菌组的生长状况较对照组旺盛，即接种丛枝菌根可以促进植物生长;配比沙土后的黑黏土比纯黑黏土中玉米叶片叶绿素含量更高，生长状况更好，接菌条件下以1∶1的比例掺入沙土时对玉米叶绿素含量的提高效果最明显。选择相关性较好的10个光谱特征参数，采用逐步回归和BP神经网络的方法对对照和接菌分别进行叶绿素含量的模型建立，BP神经网络法显示出更好的估测能力，其对照组的模型决定系数和模型检验决定系数分别为0.860和0.896，接菌组的模型决定系数和模型检验决定系数分别为0.857和0.890，可以较好估测玉米叶片叶绿素含量。

In order to improve the present situation of strong viscosity,poor permeability and low nutrient availability of black clay in open pit mining area of eastern grassland,two levels of inoculation with AM fungi were set up:inoculation with arbuscular mycorrhizal fungi (M) and control group without inoculation (CK). At the same time,the clay was set up with four different proportions:sand (S),clay (N),sandy soil and clay with mass ratio of 1:1 ( S:N = 1:1) and mass ratio of 3 ∶ 1 (S ∶ N = 3 ∶ 1). Hyperspectral remote sensing technique was used to monitor the effects of arbuscu- lar mycorrhizal fungi inoculation on plant growth and development under different substrates. The differences of typical spectral characteristic parameters under different treatments were compared, and the characteristic parameters with high correlation with chlorophyll content were screened. Stepwise regression and BP neural network modeling method were used to invert the chlorophyll content,which laid a theoretical foundation for monitoring the effect of microbial reclamation. The results showed that the chlorophyll content of maize after inoculation with AM fungi was higher than that of the control group. The growth condition of the inoculated group was stronger than that of the control group by comparing the trilateral parameters. That is,the inoculation with arbuscular mycorrhiza could promote plant growth. The black clay after the sand ratio had higher chlorophyll content and better growth condition than pure clay. Under the condition of inoculation,when the ratio of sand to clay was 1 ∶ 1,the chlorophyll content of maize increased most obvi- ously. Selecting 10 spectral characteristic parameters with high correlations,using stepwise regression and BP neural network to establish the model of chlorophyll content for control and inoculation respectively. BP neural network meth- od showed better estimation ability. The model determination coefficient and model test determination coefficient of the control group were 0. 860 and 0. 896,respectively. The model determination system of the inoculation group was the model de-termination system. The coefficient of determination of number and model test is 0. 857 and 0. 890 respec- tively,which could be used to estimate the chlorophyll content of maize leaves.