Owens Et Al.: Protein Degradation in Red Clover and Alfalfa Silage

A key objective of forage preservation as hay or Extensive degradation of protein during fermentation of high-prosilage is to minimize quality and dry matter (DM) tein crops reduces efficiency of dietary N utilization in ruminants. losses. A major disadvantage to preserving forage as Evidence suggests that enhanced levels of fermentable carbohydrates hay is the risk of exposure to adverse weather conditions can reduce proteolysis. Our objective was to evaluate whether desince several days are usually required for drying. Much laying daily cutting time, to allow total nonstructural carbohydrates of the forage grown in humid regions of the USA, where (TNC) to accumulate, would inhibit protein degradation by way of the climate is not conducive to hay making, is preserved greater acid production in the silo. Red clover (Trifolium pratense as silage. Harvest and quality losses associated with inL.) and alfalfa (Medicago sativa L.) were harvested at 0600, 1000, clement weather are usually lower in silage because 1400, and 1800 h in 1993, 1994, and 1995 and wilted to a dry matter (DM) content of 350 g kg21 before ensiling. The level of TNC in fresh drying time is typically reduced by at least 50% comforage of both species increased throughout the day. Starch accounted pared with making hay. Quality reductions due to detrifor most of the daily change in TNC in fresh alfalfa, whereas in mental microbial activity are minimized in silage by red clover, quantitative increases in sugar and starch impacted TNC excluding oxygen from the silo and rapidly attaining a similarly. Level of TNC at initiation of ensiling did not consistently pH of 3.8 to 5.0 (Pitt, 1990). affect protein degradation during fermentation as confirmed by generPreservation of high quality legumes as silage is limally insignificant correlation coefficients. The extent of proteolysis in ited by extensive protein degradation that occurs in the silo was consistently greater in alfalfa than red clover. Silage pH many forage species during fermentation (McKersie, typically decreased and starch increased as cutting time was delayed 1985; Charmley and Thomas, 1987; Albrecht and Muck, from 0600 to 1800 h. While the extent of proteolysis was largely unaffected by inherent increases in TNC, lower silage pH and higher 1991). Of the total N (TN) in fresh forages, 75 to 90% starch concentrations indicate that silage from the afternoon cuttings is protein N (PN), and the remainder is nonprotein N may be better preserved and higher in quality. (NPN) (Oshima and McDonald, 1978). After ensiling, NPN may account for as much as 80% of TN (PapadoV.N. Owens, Plant Science Dep., South Dakota State Univ. Box poulos and McKersie, 1983; Albrecht and Muck, 1991). 2207A, Brookings, SD 57007; K.A. Albrecht and S.H. Duke, Dep. of Waldo (1985) concluded that extensive protein degradaAgronomy, 1575 Linden Dr., Univ. of Wisconsin-Madison, Madison, WI 53706; R.E. Muck, US Dairy Forage Research Center, 1925 Linden tion in the silo may result in lower dry matter intake Dr., Madison, WI 53706. This research was supported by Hatch Project by ruminants and reduce the efficiency with which N is 3270 and is a contribution of the Wisconsin Agric. Exp. Stn. Received 8 Dec. 1998. *Corresponding author ([email protected]). Abbreviations: NPN, nonprotein nitrogen; PN, protein nitrogen; TN, total nitrogen; TNC, total nonstructural carbohydrates. Published in Crop Sci. 39:1873–1880 (1999). 1874 CROP SCIENCE, VOL. 39, NOVEMBER–DECEMBER 1999 ricultural Research Station, Arlington, WI (438189N, utilized. Inefficient forage N utilization by ruminants 898219W). ‘Marathon’ red clover and ‘Dart’ alfalfa were estabalso poses potential environmental problems since exlished in the spring of 1993 on a well drained to moderately cess N is excreted in urine (Tamminga, 1992). well drained Plano silt-loam soil (fine-silty, mixed, mesic, Protein degradation in the silo is influenced by several Typic Argiudoll). Soil nutrients and pH were maintained at factors including forage species (Papadopoulos and recommended levels for alfalfa throughout the experiment. McKersie, 1983; Albrecht and Muck, 1991), pH (Brady, Eptam (S-ethyl dipropylcarbamothioate) was applied pre1961; Finley et al., 1980; Scalet et al., 1984; McKersie, plant incorporated to control weeds during establishment, and 1985), DM content of the crop at ensiling (Carpintero malathion (O,O-dimethyldithiophosphate of diethyl mercapet al., 1979; Muck, 1987), and temperature (Brady, 1961; tosuccinate) was used as needed to control alfalfa weevil (Hypera postica Gyll.) and potato leafhopper (Empoasca fabae Muck and Dickerson, 1988). Albrecht and Muck (1991) Harris). evaluated several legumes and detected a significant negative relationship between tannin concentration and NPN in the silo. Red clover and cicer milkvetch (AstraHarvesting and Sample Preparation galus cicer L.) consistently underwent less protein degraWilted silage was made from forage harvested on 24 Aug. dation than alfalfa during fermentation although neither 1993 (growth cycle succeeding new seeding growth), 27 May of these species contains measurable levels of tannins. 1994 (first growth cycle), and 26 June 1995 (second growth Other researchers have also noted that the increase in cycle). Plants were clipped to a 5-cm stubble height at 0600, NPN during ensiling is lower in red clover than alfalfa 1000, 1400, and 1800 h in 1993 and 1994. In 1995, forage was (Papadopoulos and McKersie, 1983; McKersie, 1985; harvested at 0600, 1000, and 1400 h on 26 June, and the Jones et al., 1995c). 1800-h cutting was taken the next day (27 June) because of cloudy and rainy weather the afternoon of 26 June. Fresh The pH of well fermented silage should rapidly fall forage from each species and cutting time was bulked, condito 5 or less to maintain forage quality and limit the tioned by crimping the stem at approximately 2.5-cm intervals, extent of protein degradation. While not completely and thoroughly mixed for sampling. Four samples in 1993 and prevented, the activity of plant proteinases decreases 1994 and three samples in 1995 (each representing a replicate linearly between pH 6 and 4 (McKersie, 1985; Jones et for statistical analyses) were randomly taken from the bulk al., 1995a). Management practices used to promote a forage of each species at each cutting time. Each of the fresh rapid pH decline during ensiling include rapid filling, forage samples was further divided into two portions: (i) apgood packing, and good sealing (Muck, 1988). In addiproximately 900 g (wet weight) was immediately placed on tion, it is critical that an anaerobic environment be wire screens elevated approximately 4 cm off the ground and achieved as quickly as possible and that sufficient suballowed to wilt in the field to a DM content of 350 g kg2; and (ii) 500 g (wet weight) was used for analysis of fresh strate be available for fermentation by lactic acid bacteherbage characteristics. A microwave oven was used to deterria. The amount of substrate necessary for successful mine rapidly the DM content of 100-g fresh forage in order fermentation depends primarily on buffering capacity to calculate the amount of water loss required to reach the (defined as the amount of acid needed to reduce forage targeted DM level. pH from 6 to 4 per unit DM) and DM content of the The 500-g fresh plant portions were immediately placed on crop at ensiling (Muck, 1988). The buffering capacity ice and taken into the lab where they were chopped into of legumes is typically higher than that of grasses (Pitt, 1-cm pieces with a paper cutter and then thoroughly mixed. 1990), and among legumes, alfalfa is usually more highly Approximately 100-g fresh forage from each sample was buffered (McDonald et al., 1991, p. 31) and has less placed in a forced air oven for 48 h at 608C for DM determinafermentable substrate (Raguse and Smith, 1966) than tion. Two 50-g subsamples were placed in separate 530-mL sterile sampling bags and immediately placed on dry ice. One red clover. Consequently, pH values tend to be lower of these subsamples was later used for determination of TN in red clover than alfalfa silage. and NPN and the other was used for sugar (primarily glucose, The amount of fermentable substrate affects the exfructose, and sucrose [Lechtenberg et al., 1971]) and starch tent of pH decline, which in turn may affect proteolysis analysis. The subsample used to determine N fractions was in the silo. Naturally available substrate for fermentakept at 2708C, and the subsample used for determination of tion varies diurnally. Total nonstructural carbohydrates sugar and starch was lyophilized and stored at 2208C until (TNC), including sugar and starch, are lowest early in analyses could be completed. Wilted forage samples were the morning and increase through mid to late afternoon taken into the lab, chopped into 1-cm pieces, and thoroughly (Holt and Hilst, 1969; Lechtenberg et al., 1971). We mixed. Two 50-g subsamples (one for TN and NPN and one hypothesized that forage legumes harvested later in the for sugar and starch) were treated in the same manner as fresh forage. A separate 100-g subsample was dried in a forced-air day would contain greater levels of available substrate oven for 48 h at 608C for DM determination. for fermentation, thus reducing the extent of protein degradation in the silo. Therefore, our objective was to determine whether delaying cutting time during the day, Ensiling to enhance TNC accumulation, would inhibit protein Duplicate 50-g subsamples of wilted forage were ensiled degradation as a result of greater acid production in using methods described previously (Owens et al., 1999). the silo. Briefly, forage was ensiled in 100-mL polypropylene centrifuge tubes after inoculating with appropriate lactic acid bacteMATERIALS AND METHODS ria. Silos, with gas traps attached, were placed in a 308C water bath for 35 d and then stored at 2208C until further chemical Plant material used in this research was harvested in 1993, 1994, and 1995 at the University of Wisconsin Arlington Aganalyses were performed. OWENS ET AL.: PROTEIN DEGRADATION IN RED CLOVER AND ALFALFA SILAGE 1875 added to each tube and incubated at 558C for 24 h. At the Characterization of Fresh Forage, Wilted Forage, conclusion of the incubation period the samples were centriand Wilted Silage fuged at 16 000 3 g for 10 min. Glucose released from starch Ten grams of the frozen fresh forage, wilted forage, and hydrolysis was determined with glucose Trinder reagent wilted silage were diluted to 100 g with distilled water and (Sigma Chemical Diagnostic kit No. 315). Absorbance was macerated for 30 s at high speed in a Waring blender. The pH read at 505 nm and the results compared to glucose standards of the homogenate was measured immediately after blending, containing 0, 5, 10, 20, 40, 60, and 80 mg glucose mL2 distilled then the homogenate was prepared for analysis of NPN and deionized water. Starch was calculated by multiplying the glufermentation products (Owens et al., 1999). Two 20-mL alicose concentration by 0.9. quots were dispensed into separate 50-mL polypropylene centrifuge tubes. Five milliliters of 25% (w/v) trichloroacetic acid Statistical Analysis (TCA) was added to one of the tubes to precipitate protein from the solution. Tubes with and without TCA were centriWhen duplicate samples were evaluated, the average of the fuged and the supernatant decanted into 20-mL scintillation two values was used for statistical analysis. The experiment vials and stored at 2208C. The solution to which TCA had was conducted in a completely randomized design with four been added (TCA extract) was evaluated for NPN concentrareplicates in 1993 and 1994 and three replicates in 1995. Nution using the micro-Kjeldahl procedure of Bremner and merous year 3 treatment interactions and other higher order Breitenbeck (1983). The solution to which TCA had not been interactions were detected when data were analyzed across added (water extract) was used for determination of fermentayears. Some of these interactions resulted in changes in order tion products by high pressure liquid chromatography acof treatments; therefore, data were analyzed by individual cording to the method of Muck and Dickerson (1988). years. Analysis of variance was used to test statistical significance of species, cutting time of day (CT), and species 3 CT The remaining unmacerated material was weighed and interaction using the general linear model (GLM) of SAS dried in a forced-air oven at 608C for DM determinations in (SAS Institute, Cary, NC). Species and cutting time were 1993. In 1994 and 1995, unmacerated fresh forage, wilted fortreated as fixed effects in the analysis of variance model. age, and wilted silage were lyophilized for DM determination Means were separated by Fisher’s protected least significant due to concerns with the loss or transformation of carbohydifference (LSD P 5 0.05) when F-tests were determined to drate or N components during drying at 608C (Raguse and be significant. The CORR procedure of SAS (SAS Institute, Smith, 1965). We suspected that some NH3 was lost from Cary, NC) was used to generate Pearson correlation coeffiseveral of the silage samples during drying at 608C in 1993. cients to evaluate the relationship of sugar and TNC concenTo correct the data from these samples, values for TN used trations of wilted forage with silage NPN levels and silage pH in statistical analyses were calculated by combining the meavalues, and silage pH values with silage NPN concentrations. surements for TN (measured from oven-dried silage) and NH3–N (determined from the aqueous extract). In 1994 and 1995, this problem was avoided by lyophilizing the remainder RESULTS AND DISCUSSION of the frozen material. Oven-dried samples from 1993 and lyophilized samples from 1994 and 1995 were ground to pass General Characteristics of Fresh Forage, a 0.85-mm (20 mesh) screen with an intermediate ThomasWilted Forage, and Wilted Silage Wiley mill. Total N was measured on these samples by the Forage was cut when the stage of maturity was typical micro-Kjeldahl method of Bremner and Breitenbeck (1983). of what would be harvested and preserved as silage in The oven-dried or lyophilized silage samples were also used the North Central USA. Red clover was harvested at for sugar and starch analysis. Sugar and starch concentrations about 33% bloom stage in 1993, at early bud stage in in fresh forage were determined from the subsamples that had been lyophilized and stored at 2208C. 1994, and at 50% bloom stage in 1995. Alfalfa was harvested at early-bud stage in 1993, at late-bud stage in 1994, and at 10% bloom stage in 1995. Total N for fresh TNC Analysis red clover ranged from 27 to 35 g kg21 DM and for Sugar Extraction and Measurement. Sugar and starch were alfalfa ranged from 29 to 49 g kg21 DM, and was lowest determined by the methods described by Rong et al. (1996). in 1995 because both species were harvested at more Approximately 45-mg lyophilized and ground tissue was advanced stages of maturity (Tables 1 to 3). The concenrinsed four times with 80% (v/v) ethanol (1.0 mL ethanol tration of TN in both species tended to decrease slightly per rinse), centrifuged, and the supernatant decanted into a during the day, in agreement with findings by Younggraduated test tube. Anthrone reagent was added to an aliquot berg et al. (1972) with alfalfa. The decrease in TN is containing up to 200-mg free sugar, boiled for 8 min, and then likely attributed to dilution of N compounds during cooled to room temperature in a cold water bath. Absorbance at 625 nm was determined with a Shimadzu UV-1201 spectrodaytime fixation of carbon into nonstructural carbohyphotometer attached to a ASC-5 auto sample changer (Shidrates. Loss of TNC during wilting caused the level of madzu Corporation, Kyoto, Japan). Total sugars were exTN in wilted forage to be more similar across cutting pressed as glucose equivalents based on standards containing times within each harvest, although a trend of decreasing 0, 5, 10, 50, 100, 150, and 200 mg glucose mL2 80% ethanol. TN with later cutting times was still present (Tables 1 Starch Hydrolysis and Measurement The tube containing to 3). Dry matter losses associated with respiration and the ethanol-insoluble pellet was placed in a 558C oven for 24 fermentation resulted in slightly higher levels of TN in h to evaporate any residual ethanol before starch hydrolysis. silage. The observed changes are comparable to those Distilled deionized water (0.5 mL) was added to each tube reported by Albrecht and Muck (1991) for unwilted red and starch gelatinized by boiling samples for 10 min. After clover and alfalfa silage. cooling to room temperature, 1 U amyloglucosidase (Sigma The DM content of fresh forage was lowest at 0600 h Chemical product A3514) and 40 U a-amylase (Sigma Chemical product A2643) in 200 mM acetate buffer (pH 5.0) were and tended to increase to a maximum by early after1876 CROP SCIENCE, VOL. 39, NOVEMBER–DECEMBER 1999 Table 1. Means for dry matter (DM) content, total N (TN), nonprotein N (NPN), sugar (primarily glucose, fructose, and sucrose), starch, total nonstructural carbohydrate (TNC), and pH in fresh and wilted forage and wilted silage harvested 24 Aug. 1993. Fresh forage Wilted forage Wilted silage Cutting Time (CT) DM TN NPN Sugar† Starch TNC pH DM TN NPN Sugar Starch TNC pH DM TN NPN Sugar Starch TNC pH g kg21 g kg21 g kg21 g kg21 DM g kg21 g kg21 g kg21 g kg21 DM g kg21 g kg21 g kg21 g kg21 DM DM TN DM TN DM TN