---

作 者 ：任建强^1，2，3;陈仲新^2，3;唐华俊^2，3

期 刊 ：应用生态学报 2006年 17卷 12期 页码：2371-2375

关键词：小麦；

摘 要 ：以黄淮海平原冬小麦主产区山东省济宁市为研究实例，利用遥感方法，采用250 m分辨率经过Savitzky-Golay滤波技术平滑处理的MODIS-NDVI遥感数据对冬小麦产量进行预测.研究选取了冬小麦关键生育期内0.2～0.8范围的旬NDVI数据，并建立了其与冬小麦产量的关系.同时，采用逐步回归方法筛选建立冬小麦关键生育期旬NDVI与冬小麦产量间关系的估产模型.利用地面实测冬小麦产量数据，对所建的估产模型进行精度检验，结果表明，估产相对误差在-3.6％～3.9％之间.表明利用Savitzky-Golay滤波技术平滑后的作物关键生育期内MODIS-NDVI遥感数据进行冬小麦估产，其方法精度较高，具有一定的可行性.

Abstract:A pot experiment with 3 levels of elemental sulfur (0, 30, and 60 mg S·kg^-1)
showed that sulfur fertilization on soybean increased the side roots number by
8.6%～33.2%, root dry weight by 6.6%～34.3%, root nodules number by 2.7%
～35.9% and dry weight by 13.0%～75.7%, chlorophyll content by 0.4～3.9
unit,and yield per plant by 7.3%～12.8%. Sulfur fertilization also increased the
amount of soil bacteria, fungi and actinomycetes and the activities of peroxidase,
urease, neutral phosphatase and polyphenoloxidase significantly. The effects of
sulfur supply differed with its application rate, and 30 mg S·kg^-1 was more
appropriate for getting high soybean yield.

---