Effects of spontaneous heating on forage protein fractions and in situ disappearance kinetics of crude protein for alfalfa-orchardgrass hays packaged in large round bales

During 2006 and 2007, forages from 3 individual hay harvests were used to assess the effects of spontaneous heating on concentrations of crude protein (CP), neutral detergent insoluble CP (NDICP), acid detergent insoluble CP (ADICP), and in situ disappearance kinetics of CP and NDICP for large round bales of mixed alfalfa (Medicago sativa L.) and orchardgrass (Dactylis glomerata L.). Over the 3 harvests, 96 large round bales were made at preset bale diameters of 0.9, 1.2, or 1.5 m and at moisture concentrations ranging from 9.3 to 46.6%. Internal bale temperatures were monitored daily during an outdoor storage period. The change in concentrations of NDICP (poststorage-prestorage) increased with heating degree days (HDD) > 30 degrees C in a relationship best explained with a nonlinear model {Y = 24.9 - [22.7 x (e(-0.000010 x x x x))]; R(2) = 0.892} that became asymptotic at + 24.9 percentage units of CP, thereby indicating that NDICP increases rapidly within bales that heat spontaneously. When maximum internal bale temperature (MAX) was used as the independent variable, the best regression model was quadratic and the coefficient of determination was still relatively high (R(2) = 0.716). The change in concentrations of ADICP (poststorage - prestorage; Delta ADICP) also increased with HDD and was best fitted to a nonlinear model {Y = 14.9 - [15.7 x (e-(0.0000019 x x x x))]} with a very high coefficient of determination (R(2) = 0.934). A similar quartic response was observed for the regression of Delta ADICP on MAX (R(2) = 0.975). Increases in Delta ADICP as a result of heating (HDD or MAX) were paralleled by concurrent increases in hemicellulose at relatively low increments of heating, but the inverse relationship was observed as hemicelluloses likely became reactive and concentrations decreased in more severely heated hays. Changes in ruminal disappearance rate of CP were best fitted to cubic models for regressions on both HDD (R(2) = 0.939) and MAX (R(2) = 0.876); these changes represented an approximate 50% rate reduction in severely heated hays relative to prestorage controls. Within ranges of heating most commonly encountered under field conditions, changes in rumen-degradable protein decreased in a primarily linear relationship with HDD or MAX. However, the mean change in rumen-degradable protein for the 4 most severely heated hays was only - 2.6 percentage units of CP, which represents a minimal reduction from prestorage controls and is far less than the maximum of -7.9 percentage units of CP observed with less-severe heating. Interpretation of these results was complicated by poor recovery of NDICP from our most severely heated hays following machine rinsing of 0-h Dacron bags; theoretically, and by definition, this unrecovered pool of NDICP is assumed to be entirely degradable in the rumen. It remains unclear whether these responses could be corroborated in vivo or by other analytical techniques, or whether the magnitude of HDD or MAX for our most severely heated hays exceeds the reliable limits for estimating RDP via in situ methodology.