Modeling in vitro dry matter digestibility of oat (Avena sativa) fodder based on proximate and fiber fractions

Abstract

The present investigation was conducted to evaluate the in vitro Dry Matter Digestibility (IVDMD) of oat (Avena sativa L.) fodder and to develop predictive models based on proximate and fiber fractions. Eight oat cultivars were studied at two growth stages (60 and 90 days after sowing) under a factorial randomized block design. The results revealed that IVDMD was significantly influenced by cultivar, growth stage, and their interaction (p<0.01). The overall mean IVDMD was 61.47%, with a marked decline from 65.51% at 60 DAS to 57.42% at 90 DAS, indicating reduced digestibility with advancing maturity due to increased structural carbohydrate accumulation. Among cultivars, RO-2000-19 exhibited the highest digestibility (67.07%), whereas RO-2000-16 recorded the lowest (54.23%), reflecting genotypic variation in fiber composition and degradability. Correlation analysis demonstrated that IVDMD was positively associated with crude protein (r = 0.700), ether extract (r = 0.697), and ash (r = 0.653), while it was negatively correlated with crude fibre (r = −0.756) and nitrogen free extract (r = −0.678). Furthermore, strong negative correlations were observed with detergent fiber fractions, particularly silica (r = −0.864), ADF (r = −0.801), and cellulose (r = −0.774), indicating their critical role in limiting digestibility. Regression analysis revealed that the model based on proximate constituents explained 66% of the variation in IVDMD (R² = 0.66), whereas the model based on detergent fiber fractions showed substantially higher predictive accuracy (R² = 0.85). These findings clearly indicate that fiber fractions, especially structural carbohydrates and silica, are superior predictors of forage digestibility compared to proximate composition. The study underscores the importance of early harvesting (60 DAS) for maximizing digestibility and highlights the potential of integrating chemical composition with regression modeling for rapid and reliable prediction of forage quality. The developed models can serve as effective tools for forage evaluation, ration formulation, and selection of nutritionally superior oat cultivars for enhanced livestock productivity.