Technical note: Effects of filter bags on neutral detergent fiber recovery and fiber digestion in vitro.

Filter bags facilitate the measurement of amylase-treated neutral detergent fiber (aNDF) and in vitro (IV) undigested aNDF (uNDF) by eliminating the transfer of residues from beakers into filtration crucibles. The objectives of this study were to (1) determine effects of filter bags on recovery of aNDF and (2) evaluate effects of filter bags on IV uNDF. For study 1, 6 samples each of grass hay (GR), alfalfa (AL), and corn silage (CS) were selected. Large standard deviations (SD) of ash-free aNDF (aNDFom) for samples in each forage type indicated compositional diversity (15.1, 7.45, and 12.9% of DM for GR, AL, and CS, respectively), and starch SD for CS was 16.4% of DM. Samples were weighed into Berzelius beakers or filter bags [25-µm pores (F57) or 6-µm pores (F58); Ankom Technology, Macedon, NY] for measurement of aNDF and aNDFom. All samples were extracted with neutral detergent, thermostable α-amylase, and sodium sulfite, and then soaked in boiling water and then acetone. Residues from beakers were filtered through a sea sand-covered GF/D filter (Whatman, Marlborough, MA) in Gooch crucibles (CR). Filter bags were extracted in a pressurized chamber at 100°C. The aNDF values did not differ between F57 and CR, but F58 was greater than CR for CS and AL. For GR, F58 was greater than CR for aNDFom. For study 2, diverse samples with large SD of aNDFom (20.7, 7.45, and 12.9% of DM for GR, AL, and CS, respectively) were weighed as loose powder into medium bottles (LS) or F57 bags, which were weighted to prevent floating. Blended ruminal fluid from 3 steers fed a 30% aNDFom diet was used as inoculum. Three samples of 1 forage type were randomly assigned to 1 of 6 IV runs using both treatments (LS and F57), and 3 bottles of each sample-treatment combination were removed after 12 h and 2 were removed after 120 h to measure uNDF. For LS, residues were extracted as in study 1 for CR. For F57, bags were rinsed in cold water and extracted as described in study 1. After 12 h, uNDF of F57 was greater than LS in CS, AL, and overall types. Ash-free uNDF (uNDFom) after 12 h of F57 was greater than LS in CS and overall types. After 120 h, F57 was greater than LS for uNDF of CS, but no differences were detected for uNDFom. The SD of uNDFom, but not uNDF, was higher after 12 and 120 h for F57 compared with LS. From 6 to 96 h, overall gas production of F57 was less than LS, and F57 was less than LS for CS from 3 to 96 h. Overall, LS gave greater maximum and faster rates of gas production than F57, as did AL and CS, but lag did not vary. Results indicate that filter bags affected aNDF and aNDFom measurement and inhibited fermentation for some materials.