Forage quality variation among maize inbreds: In vitro fiber digestion kinetics and prediction with NIPS

The nutritive value of forage maize (Zea mays L.) may be improved through genetic selection for increased rate of fiber digestion or decreased indigestible fiber concentration. To identify sources of genetic variation, 45 maize inbreds were evaluated for in vitro neutral detergent fiber (NDF) digestion kinetic parameters using stem internode tissue harvested at silking during 2 yr. Near infrared reflectance spectroscopy (NIRS) was also used to estimate NDF digestion kinetic parameters. Maize inbreds varied significantly in NDF concentration and digestion kinetic parameters using either conventional in vitro analysis or NIRS predictions. Using MRS predictions, inbreds varied in NDF concentration from 497 to 662 g kg(-1) dry matter (DM), rate of NDF digestion ranged from 0.037 to 0.077 h(-1), and extent of NDF digestion was 525 to 735 g kg(-1) NDF. The ranges for NIRS predicted parameters were less than those observed for the calibration data set by conventional analysis. Correspondence between conventional analysis data and PI;IRS predictions were good, except for lag time. Digestion kinetics calculated from NIRS predicted residues provided more precise predictions of lag time and fractional rate of digestion when compared with observations derived from conventional analyses, than did direct prediction of these kinetic parameters. Correlations between rate of NDF digestion and 18-h NDF digestibility (r = 0.79) or between potential extent of NDF digestion and 96-h NDF digestibility (r = 0.95) were large enough that these two fermentation intervals might substitute for conducting complete digestion kinetic studies with eight to 10 fermentation times. The substantial genetic variation among these maize inbreds shows good potential for development of silage hybrids with improved fiber digestion parameters. Year and year x genotype interactions were significant suggesting that identification of superior inbred lines will require evaluations in multiple environments.