Validation of a Technique for Estimating Alfalfa (Medicago sativa) Biomass from Canopy Volume  

作　　者：Christopher G. Misar Lan Xu Arvid Boe Roger N. Gates Patricia S. Johnson Andrew E. Olson Christopher G. Misar;Lan Xu;Arvid Boe;Roger N. Gates;Patricia S. Johnson;Andrew E. Olson(Northern Crop Science Laboratory, USDA-ARS, Fargo, ND, USA;Department of Natural Resource Management, South Dakota State University, Brookings, SD, USA;Department of Plant Science, South Dakota State University, Brookings, SD, USA;Department of Natural Resource Management, WRAC, South Dakota State University, Rapid City, SD, USA)

机构地区：[1]Northern Crop Science Laboratory, USDA-ARS, Fargo, ND, USA [2]Department of Natural Resource Management, South Dakota State University, Brookings, SD, USA [3]Department of Plant Science, South Dakota State University, Brookings, SD, USA [4]Department of Natural Resource Management, WRAC, South Dakota State University, Rapid City, SD, USA

出　　处：《American Journal of Plant Sciences》2016年第1期238-245,共8页美国植物学期刊（英文）

摘　　要：Determining biomass production of individual alfalfa (Medicago sativa L.) plants in space planted evaluation studies is generally not feasible. Clipping plants is time consuming, expensive, and often not possible if the plants are subjected to grazing. A regression function (B&#8242 = 0.72558 + 0.11638 × V&#8242) was developed from spaced plants growing on rangeland in northwestern South Dakota near Buffalo to nondestructively estimate individual plant biomass (B) from canopy volume (V). However, external validation is necessary to effectively apply the model to other environments. In the summer of 2015, new data to validate the model were collected from spaced plants near Brookings, South Dakota. Canopy volume and clipped plant biomass were obtained from ten alfalfa populations varying in genetic background, growth habit, and growth stage. Fitted models for the model-building and validation data sets had similar estimated regression coefficients and attributes. Mean squared prediction errors (MSPR) were similar to or smaller than error mean square (MSE) of the model-building regression model, indicating reasonable predictive ability. Validation results indicated that the model reliably estimated biomass of plants in another environment. However, the technique should not be utilized where individual plants are not easily distinguished, such as alfalfa monocultures. Estimating biomass from canopy volume values that are extrapolations (>2.077 × 10⁶ cm³) of the model-building data set is not recommended.Determining biomass production of individual alfalfa (Medicago sativa L.) plants in space planted evaluation studies is generally not feasible. Clipping plants is time consuming, expensive, and often not possible if the plants are subjected to grazing. A regression function (B&#8242 = 0.72558 + 0.11638 × V&#8242) was developed from spaced plants growing on rangeland in northwestern South Dakota near Buffalo to nondestructively estimate individual plant biomass (B) from canopy volume (V). However, external validation is necessary to effectively apply the model to other environments. In the summer of 2015, new data to validate the model were collected from spaced plants near Brookings, South Dakota. Canopy volume and clipped plant biomass were obtained from ten alfalfa populations varying in genetic background, growth habit, and growth stage. Fitted models for the model-building and validation data sets had similar estimated regression coefficients and attributes. Mean squared prediction errors (MSPR) were similar to or smaller than error mean square (MSE) of the model-building regression model, indicating reasonable predictive ability. Validation results indicated that the model reliably estimated biomass of plants in another environment. However, the technique should not be utilized where individual plants are not easily distinguished, such as alfalfa monocultures. Estimating biomass from canopy volume values that are extrapolations (>2.077 × 10⁶ cm³) of the model-building data set is not recommended.

关 键 词：Forage Production Forage Yield LUCERNE PHYTOMASS Predictive Ability 

分 类 号：S54[农业科学—作物学]