Effects of bulk density, precision processing and processing index on in vitro ruminal fermentation of dry-rolled barley grain

A study using a batch culture technique was conducted to evaluate the effects of bulk density (BD, g/L), processing method (PM; control vs. precision processing [PP]: sieving grains into large versus small kernels and rolling based on kernel size), and processing index (PI; BD after rolling/BD before rolling) on kinetics of gas production, dry matter degradability (DMD), molar proportions and total short chain fatty acids (SCFA) of dry rolled barley grain. The study was arranged in a 2 × 2 × 2 factorial design. Gas production and DMD were measured at 3, 6, 12 and 24 h of incubation using rumen fluid from 3 fistulated beef heifers fed diet containing 700 g barley silage and 300 g of a barley grain-based concentrate. We hypothesized that incorporating other factors with PI would help improve prediction of the feeding value of processed barley grain. Barley samples were collected monthly from ten different feedlots in Southern Alberta for one year. Samples were ranked according to their BD into low (580 g/L) and high (680 g/L) BD which were subjected to different processing (control vs. PP) before being dry-rolled with extent of processing expressed as PI of 0.75 or 0.85 ± 0.03. The dry-rolled samples used in the study were not subjected to further grinding. Cumulative gas production (GP; mL/g organic matter) was fitted to an exponential model GP = B(1 − exp − c × [t − lag]). Processing method × PI interactions (P<0.01) were observed for asymptotic gas volume, rate, lag time and absolute initial gas produced during the first hour. In addition, a PM × BD interaction (P<0.05) was noted for asymptotic gas volume. Data from DMD at different times of incubation were fitted to a model DMD = a + b(1 − e−c(t−L)). Effective degradability coefficients ranged from 0.18 to 0.26. Only PI had an effect (P<0.05) on the rate of DMD. Total SCFA values ranged from 30.2 to 40.1 mmol/L. Apart from PM and PI interaction (P<0.05) on C2:C3, no other interaction was observed after 24 h of incubation. Regression results showed that BD and PM were better for predicting rate of DMD than PI. This is consistent with our hypothesis that using only PI as a measurement for extensity of grain processing may not be reliable when barley grain kernels are highly variable in size.