In vitro rumen gas production and starch degradation of starch-based feeds depend on mean particle size

Our objective was to model the effect of mean particle size (mPS) on in vitro rumen starch 
degradation (IVSD) and the kinetics of gas production for different starch-based feeds. For 
each feed, two batches of the same grains were separately processed through two different 
mills (cutter or rotor speed mills), with or without different screens to achieve a wide range of 
mPS (0.32 to 3.31 mm for corn meals; 0.19 to 2.81 mm for barley meals; 0.16 to 2.13 mm for 
wheat meals; 0.28 to 2.32 mm for oat meals; 0.21 to 2.36 mm for rye meals; 0.40 to 1.79 for 
sorghum meals; 0.26 to 4.71 mm for pea meals; and 0.25 to 4.53 mm for faba meals). The 
IVSD data and gas production kinetics, obtained by fitting to a single-pool exponential model, 
were analyzed using a completely randomized design, in which the main tested effect was mPS 
(n = 6 for all tested meals, except n = 7 for corn meals and n = 5 for sorghum meals). Rumen 
inocula were collected from 2 fistulated Holstein dairy cows that were fed a total mixed ration 
consisting of 16.2% crude protein, 28.5% starch, and 35.0% neutral detergent fiber on a dry 
matter basis. The IVSD, evaluated after 7 h of rumen incubation, decreased linearly with 
increasing mPS for corn, barley, wheat, rye, pea, and faba meals, and decreased quadratically 
with increasing mPS for the other meals. The y-axis intercept for 7h IVSD was below 90% 
starch for corn, barley, and rye feeds, but greater than 90% for the other tested feeds. The mPS 
adjustment factors for the rate of rumen starch degradation varied widely among the different 
tested feeds. There was a linear decrease in starch degradation with increasing mPS for barley, 
wheat, rye, and pea meals, while a quadratic decrease in starch degradation for the other tested 
meals. Further, there was a linear decrease in the rate of gas production (kd) with increasing 
mPS in each tested feed, except for pea meal, which had a quadratic relationship. For each 1 
mm increase in mPS, the kd was adjusted by -0.009 h-1 for corn, -0.011 h-1 for barley, -0.008 h- 
1 for wheat, and -0.006 h-1 for faba; while numerically greater adjustments were needed for oat (-0.022 h-1), rye (-0.017 h-1), and sorghum (-0.014 h-1). These mPS adjustment factors could be 
used to modify the starch-based feed energy values as a function of mean particle size, although in vivo validation is required.