Premium
Predictors of Alfalfa Forage Quality: Validation with Field Data
Author(s) -
Sanderson Matt A.
Publication year - 1992
Publication title -
crop science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.76
H-Index - 147
eISSN - 1435-0653
pISSN - 0011-183X
DOI - 10.2135/cropsci1992.0011183x003200010049x
Subject(s) - forage , neutral detergent fiber , growing degree day , calibration , medicago sativa , mathematics , biology , mean squared prediction error , statistics , zoology , agronomy , sowing
The mean stage system for describing alfalfa ( Medicago sativa L.) morphological development and quality has potential as a management tool. Published equations for predicting alfalfa leaf, stem, and herbage quality have not been tested widely. Study objectives were to (i) test published equations for predicting quality from mean stage weight (MSW) or count (MSC) or growing degree days (GDD, base = 5°C) with forage quality data for alfalfa stems, leaves, and herbage, and (ii) test a published equation for converting MSC values to MSW. Two published data sets from Iowa ( n = 83 and1 6) and two unpublished data sets from Texas ( n = 830 and 164) were used. Predicted values of neutral‐ and acid‐detergent fiber (NDF and ADF), acid‐detergent lignin (ADL), in vitro true digestibility (IVTD), and crude protein (CP) for leaves, stems, and herbage were regressed on observed values. In most instances, intercepts and slopes were significantly different from zero and one, respectively, for values from any predictor regressed on actual values, indicating that the prediction equations were biased; however, the prediction error frequently was less than the calibration error. Values of NDF and ADF in alfalfa stems predicted from GDD equations for actual values better than did predicted values from MSC and MSW equations for the Iowa data sets. For the Texas data sets, GDD was inaccurate and imprecise in predicting NDF or IVTD of herbage. Values of MSW predicted from MSC fit observed values well. Biases in the prediction equations indicate that calibration for specific geographic or environmental regions and frequent recalibration are necessary.