پيش بيني تغييرپذيري كربن آلي خاك و تعيين اثرات متغيرهاي فيزيكي و مديريتي بر آن در يك حوضه نيمه‌خشك با كاربري ديم با استفاده از تكنيك‌هاي تحليل چندمتغيره تفكيك اعتباري(CDA)

Soil Organic Carbon Variability Prediction and Determination of Physical and Management Variables Impacts in a Semi-arid Rainfed Watershed Using Multivariate Canonical Discriminate Analysis (CDA) Techniques

Protection of soil organic carbon (SOC), as most important soil quality indicator, is the main factor in sustainable agriculture and soil ecosystem conservation. SOC distribution is mainly affected by soil management, status. This research was conducted to investigate the effects of physical and management variables on SOC variations and to quantify the relative importance of these variables on SOC distribution in a rainfed watershed by use of canonical discriminate analysis (CDA) and stepwise discriminate analysis techniques. SOC quantities in soil sampling points were classified in four quality categories as: poor, low, medium and high. Then the effects of 30 physical and management variables on prediction of SOC classes were evaluated. Results indicated that among predicting models with physical exploratory variables, model with soil characteristics as independent variables including TNV, SP, gravel, clay and sand was able to reasonably predict optimum SOC class. But all models with management exploratory variable were able to predict optimum SOC class by use of first linear combination of canonical functions at aO.OOOl. Model M5 showed highest canonical correlation with the first linear combination. All the variable combinations in significant models were able to predict poor and low SOC classes, precisely. Only M5 model had highest ability to distinguish high class of SOC. Among management variables, tillage system scenario and its components had highest impacts on SOC variability in this rainfed watershed. Stepwise discriminate analysis was able to distinguish the effects of winter fallow system on SOC status improvement. Keywords: Soil organic carbon, Canonical discriminate analysis, Stepwise discriminate analysis, Classification