The influence of sample preparation of forage crops and silages on recovery of soluble and non-structural carbohydrates and their predictions by Fourier transform mid-IR transmission spectroscopy

A total of 24 grass–legume and 12 whole-crop samples, i.e., 18 crops and 18 silages were analyzed for non-structural carbohydrates and used to study the effects of: (i) four preparation–extraction methods and two solvents on recovery of water soluble sugars [Su(s)] in the grass–legume samples, (ii) three preparation methods of the whole crops on recovery of acetate buffer Su(s), soluble [St(s)] and insoluble starch [St(i)] and non-structural carbohydrates (NSC; soluble sugars and total starch). The preparation methods consisted of oven drying at 60 °C (ODR), freeze drying (FDR) and no drying of the fresh-frozen materials (FF). The latter preparation was divided into a hot (standard method) and a cold water extraction of the grass–legume samples. A Fourier transform-(mid) infrared (FTIR) spectrometer was calibrated against all extracts and dry matter was determined both by drying and by Karl Fisher titration. Dry matter contents were higher for Karl Fisher titration (P<0.001) but FDR and ODR did not differ. The grass–legume ODR preparations had 19 and 29% lower Su(s) contents compared with the FF preparations for crops and silages, respectively. There were no differences among FDR and the two FF variants, and results from acetate and water extractions were similar for this group of samples. In whole-crop samples, values were 7% lower values for the ODR preparation compared to FDR and FF (P<0.05) in Su(s), with no effects of sample preparation on total starch P>0.05). The proportion of starch that was soluble, i.e., non-centrifugeable in the acetate buffer, ranged from 0.06 to 0.85 for the FF preparation, but for the FDR preparation, solublization was less than half of the FF preparation (P<0.05). Mean FTIR prediction error for NSC was lowest for the acetate extracts and highest for the combined set, but both were below 0.12, using 13 to 19 wavelengths. Fresh, freeze dried or cold water extracts are recommended for analysis of soluble sugars in forage crops and silages. Acetate buffer or water will extract similar amounts of sugars and are both recommended for both grass–legume and whole-crop samples, but large proportions of the starch may exist in a “soluble” form. FTIR analysis appears to be a promising alternative to reduce analytical costs.