Maataloustieteellinen A ikakauskirja Vol. 59: 371—377, 1987 The effect of inoculants and cellulase on the fermentation and microbiological composition of grass silage. II Microbiological changes in the silages AINO KAURAMAA, JOUKO SETÄLÄ, TAUNO MOISIO and SEPPO SIVELÄ Valio Co-operative Dairies’ Association, Research and Development Department P.O. Box 176, SF-00181 Helsinki, Finland TERTTU HEIKKILÄ and MARTTI LAMPILA Agricultural Research Centre, Department of Animal Husbandry 31600 Jokioinen, Finland Abstract. Four experimental silages were made into bunker silos by adding lactic acid bacteria and cellulase enzyme. The control silages were made without any additives and with AIV II solution (80 % formic acid, 2 °/o orthophosphoric acid). The microbial compositions of the experimental silages were studied in five microbial groups; (1) total count of bacteria, (2) lactic acid bacteria, (3) Clostridia, (4) coliform bacteria and (5) fungi (yeasts and moulds). At the beginning of the fermentations enteric bacteria and lactic acid bacteria were in- creasing. Inoculation of the silage decreased the growth of enterobacteria compared with silages made with cellulase or without additives. This effect was further increased by using AIV II solution. Heterolactic fermentations started in the silages made with cellulase, inoculants and no additives, when initial reducing sugars, primarily glucose and fructose, were fermented. In good ensiling conditions only a few butyric acid forming Clostridia and small amounts of moulds and yests were found. Yeasts existed mostly in silages made with cellulase and with inoculation. Index words: Silage preservation, inoculants and enzyme, microbiological changes Introduction In the summer of 1985 four experimental silages were made from the first cut into bunker silos by adding lactic acid bacteria 2 and cellulase enzyme. The control silages were made without any additives and with AIV II solution (80 % formic acid, 2 % orthophosphoric acid, 5.2 1/ton grass). As reported by Kauramaa et ai. (1987) formic 371 JOURNAL OF AGRICULTURAL SCIENCE IN FINLAND https://www.c-info.fi/en/info/?token=EPIorpzxPRuJFp1n.qC43uTPVLn2eAiY64ccUmA.jraPkOC7PfiQVKkuV2xauLvyyzWWksovXu6VU6fHVQtK9YS3d0PFiWxr4kvd6dl2MrgaLLI9dct7ZSaZtkk5O06N0xaa0te5UQeNNv-c5JUVYi_tns17yIkHpCGf39M7c6BGTTTnFl7u513mHbxr2xzw5hEYu1JRvrkA9bRlsRObc1gO_yY78gxoxgTPf8flYZiaX1Obe5dRWypGjKtGQlWyLGpumP50DpSJmKjqRbaRVYvQhGaT1ehv2uIPIPK0n_g5MJSTotf9KzFBd-ME-LFLqH3P7WtNVEltt6_5Glg 372 acid restricted respiration and fermentation in AIV II silage. The lower contents of lactic, acetic and propionic acids, as well as am- monia-N % from total-N, temperature, and buffering capacity and the higher sugar con- tent and better digestibility in vitro stand up for this conclusion. The changes in the micro-organisms of silage were studied dur- ing ensiling. The microbial compositions of the experimental silages were studied in five microbial groups: (1) total count of bacte- ria, (2) lactic acid bacteria, (3) Clostridia, (4) coliform bacteria and (5) fungi (yeasts and moulds). Microbiological methods Samples Samples for microbiological analyses were taken by hand using plastic gloves together with the samples for chemical analyses (see Rauramaa et al. 1987). They were trans- ported to the laboratory covered with ice and analyzed 2.5 hours after sampling. The size of a sample bag was about 1 kg. Procedure 25 g of silage was weighted aseptically into a sterile plastic bag, 225 ml sterile dilute solu- tion (Ringer’s solution quarter strength) was added. The silage was homogenized by a stomacher for one minute. Further decimal dilutions were made from this homogenate and each decimal dilution was used for the incubation of media. The bacterial groups determined and the media are described be- low: Plate count, incubated aerobically at 30 + I°C, 72 ±2 hours (as described in FIL-IDF 100: 1980 and 109: 1982). Anaerobic plate count, as plate count, in- cubated in anaerobic jars (BBL) with An- aerocult (E. Merck, 13 289) Coliforms, incubated in VRB agar medium aerobically at 30 ± 1 °C for 24 ± 2 hours, as described in FIL-IDF 73: 1974 Enterobacteriaceae, as above with 1 % glucose added. Lactobacteriaceae, incubated anaerobically in MRS-agar medium at 30±1°C for 72 ± 2 hours Gibson’s tests inoculation with 1 ml of I—21 —2 day culture incubated anaerobically in Gib- son’s medium at 30 ± 1 °C (Gibson and Abd- El-Malek 1945). Results are expressed as a number of heterofermentative colonies from 10 colonies tested on each plate. Yeasts and moulds, incubated aerobically in yeast extract-dextrose-chloramphenicol-agar medium at 25 + I°C for 4 days, as described in FIL-IDF 94: 1980 (provisional) Butyric acid clostridial spores, incubated an- aerobically in 88-lactate broth medium. Results are expressed as MPN-values (Ber- gere et al. 1969). Results and discussion As reported by Rauramaa et al. (1987) the sugar content of raw material sward was high, over 15 % in dry matter. From the sugars two thirds were glucose, fructose, and sucrose. Also the silage making technique was good. According to the contents of carbon dioxide in the silages (Rauramaa et al. 1987) oxygen was consumed very rapidly in the fodder mass and the careful compacting and covering prevented the access of oxygen of the air into the silages. Therefore in all the experimental silages there was a good opportunity for homofermentative lactic acid fermentation. Total count of bacteria In the raw material the total leaf surface population or epiphytic microflora varied on the plate count from 106 to 108 cfu g~‘ fresh weight. The average contents of aerobic bac- teria was 4.5 X 107 cfu g_l fresh weight, and that of anaerobic bacteria 1.2 x 10s cfu g~‘ fresh weight. These figures were within the ranges expressed by Kroulik et al. (1955 a, b), Nilsson and Nilsson (1956), Gibson et al. (1958 and 1961), Langston et al. (1962) and Woolford (1984). The contents of aerobic bacteria on plate count varied during the preservation period. Most of them were apparently facultative anaerobes. According to chemical analyses there was not any fermentation in AIV II silage (Rauramaa et al. 1987) and the con- tents of microbes decreased (Fig. 1 and 2) during the first preservation day. In the con- tents of aerobic and anaerobic bacteria of every experimental silage the decrease oc- curred after four weeks from the filling of the silos. Fig. I. The changes of the microbiological contents in the experimental silages made with AIV II solution, no additives, cellulase or inoculants. 373 374 Lactic acid bacteria The contents of lactic acid bacteria in raw material varied from 102 to 105 cfu g_l fresh weight. Generally grasses are a poor source of lactic acid bacteria, often less than 102 cfu g-1 fresh weight (Stirling and Whittenbury 1963, Woolford 1984, Silley and Damoglou 1985). Lindgren et ai. (1985 a) and Ojala and Poutiainen (1986) reported contents of lactic acid bacteria in grasses from 102 to 106 cfu g 1 fresh weight. In our experiment lactic acid bacteria might have been able to grow during the delay between the starting of the microbiological analyses and the time of sampling, though the samples were covered with crushed ice. During the first preservation day lactic acid bacteria multiplied in the experimental silages excluding AIV II silage (Fig. 2) and the contents of lactic acid bacteria were simi- lar in the silages made with inoculants, cel- lulase or without additives. In all these silages was a clear decrease in the content of lactic acid bacteria just as in the total count of aerobic and anaerobic bacteria. It is possible that at that juncture lactic acid bacteria flora changed its metabolisms in the silages made with cellulase, inoculants and no additives, because the hexose-sugars were consumed and at the same time the redox potential decreased significantly as described by Rauramaa et al. (1987), and according to Gibson’s tests heterolactic bacteria started to increase in those three silages (Fig. 2). At the end of the preservation period nearly all the lactic acid bacteria were heterofermentative in the silages made with cellulase, inoculants and no addi- tives whereas in AIV II silage only half of the lactic acid bacteria were heterofermentative (Fig. 2). Fig. 2. The changes of Ihe microbiological contents in the experimental silages made with AIV II solution, no additives, cellulase or inoculants. The amount of the added lactic acid bacte- ria was 7 X 105 cfu g _l grass in the inoculated silage. This addition was not seen in the con- tents of lactic acid bacteria, of the inoculated silage compared with those of lactic acid bac- teria in the cellulase treated and untreated silages (Fig. 2). The contents of lactic acid bacteria and also the production of lactic acid were lowest in AIV II silage (Kauramaa et al. 1987). It is clear that the addition of formic acid not only restricted the growth of lactic acid bacteria, but also selected the strains of lactic acid bacteria (Fig. 2, Gibson test). This is in agreement with McDonald, 1981. Butyric acid clostridia Relatively low amounts of Clostridia ap- peared in grass, 3—25 MPN g-1 fresh weight (Fig. 1). Clostridia are not normally included in grass microflora; they are contaminants from the soil entering the grass during har- vesting. Owing to the good harvesting tech- nique only a few clostridia spores and no butyric acid were found in the experimental silages (Kauramaa et al. 1987). Coliform bacteria Coliform bacteria belong to the family of Enterobacteriaceae. The contents of these bac- teria were 1.1—4xl0s cfu g^ 1 fresh weight in the raw material of experimental silages. During the first preservation day, reducing factors for enterobacteria were formic acid and low pH values. The acid treatment pro- duces plasmolysis in fresh fodder. It is there- fore easier to compact acid treated grass and to get it airtight during the filling of the silo. Coliform bacteria, yeasts and moulds, which cause aerobic deterioration (Lindgren et ai. 1985 b), can consume only oxygen dissolved in the effluent of silage. In our experiment the role of coliform bacteria was found to be connected with the fermentation of the first preservation days (Fig. 1). According to Kroulik et al. (1955 a, b) coliform bacteria disappear within a few days from silages, of which pH values decrease rapidly. The inoculation of lactic acid bacteria had a decreasing effect on the growth of entero- bacteria in this experiment (Fig. 1). Further, the end products of fermentations were dif- ferent qualitatively and quantitatively in the inoculated silage compared with the silage made with no additives (Kauramaa et al. 1987). During four weeks the enterobacteria incubated with or without glucose decreased below 10 cfu g-1 fresh weight in all the ex- perimental silages. In good silages coliform bacteria were not found to a greater extent than in our earlier unreported experiments. Fungi Yeasts are usually connected to initial aerobic spoilage of silages (Beck and Gross 1964, Woolford et al. 1979, Moon et al. 1980, Dellagio 1985). Such results have also been reported where many different micro- organisms have been found to be respon- sible for aerobic deterioration (Ohshima and McDonald 1978, Crawshaw et al. 1980, Woolford 1984, Lindgren et al. 1985 b). In our experiment the contents of yeasts in the raw material were an average of 3.9 x 105 cfu g-1 fresh weight and the contents of moulds were respectively 2.1 xlO4 cfu g~'. During the preservation period the contents of yeasts and moulds reduced in all silages (Fig. 1). The highest levels of yeasts were in the inoculated and cellulase treated silages, the averages being 2.6 X 104 and 5.0 X 104 cfu g-1 fresh weight. These silages had also the highest concentration of ethanol as reported by Kau- ramaa et ai. (1987). In previous studies most of the yeasts were found in the prewilted, formic acid-treated silage by Henderson et al. (1972), and slightly higher yeast and mould counts in the laboratory experiment for the formic acid-treated materials were reported by Lindgren et ai. (1985 a). In our experiment the contents of moulds decreased their level being 102 cfu r 1 fresh weight. All the above 375 mentioned contents of yeasts and moulds in the experimental silages were low and the effect of these micro-organisms on fermenta- tion remained slight. In conclusion all the experimental silages were organoleptically good, but according to the chemical analyses there occurred some differences which were in accord with the microbiological analyses. At thebeginning of the fermentations enteric bacteria were in- creasing. Inoculation of the silage decreased the growth of enterobacteria compared with silages made with cellulase or without addi- tives. This effect was further increased by using AIV II solution. Heterolactic fermenta- tions started in the silages made with cellulase, inoculants and no additives, when initial re- ducing sugars, primarily glucose and fructose, were fermented. In the present good ensiling conditions only a few butyric acid forming clostridia and small amounts of moulds and yeasts were found. Yeasts existed mostly in silages made with cellulase and with inocula- tion. AcknowledgemcnCs. The authors would like to thank the staff of the Agricultural Research Centre and the chemical and microbiological groups in the Research and Development Department of Valio for their technical assistance. References Beck, TH. & Gross, F. 1964. Ursachen der unterschied- lichen Haltbarkeit von Gärfutter. Das Wirtschafsteigene Futter. 10; 298—312. Beroere, J-L., Rousseaux, P., Ducruct, P., Mocquot, G-, Hermier, J., Gouet, P. & Zelter, S.Z., 1969. Experimantation sur la Fabrikation d’Emmental et de Comete avec du lait de vaches norries ä I’ensilage A.LV. I. Aspects technologiques Extrait du »Bulletin Tech- nique dTnformation». 239: 1—37. Crawshaw, R., Thorne, D.M. & Llewelyn, R.H. 1980. The effects of formicand propionic acids in the aerobic deterioration of grass silage in laboratory units. J. Sci. Food Agric. 31: 685—694. Dellaclio, F. 1985. Lactic acid bacteria in silage fer- mentation. Microbiologie Aliments Nutrition: 3: 91 104. Gibson, T. & Abd-El-Malek, Y. 1945. The formation of carbon dioxide by lactic acid bacteria and Bacillus licheniformis and a culture method for detecting the process. J. of Dairy Res. 14: 35—44. Gibson, T., Stirling, A.C., Keddie, R.C. & Rosen- berger, R.F. 1958. Bacteriological changes in silages made at controlled temperatures. J. Gen. Microbiol. 19: 112—129. Gibson, T., Stirling, A.C., Keddie, R.C. & Rosen Berger, R.F. 1961. Bacteriological changes in silage as affected by laceration of the fresh grass. J. Appi. Bact. 24 (1): 60—70. Henderson, A.R., McDonald, P. & Woolford, M.K. 1972. Chemical changes and losses during the ensilage of wilted grass treated with formic acid. J. Sci. Fd. Agric. 23: 1079—1087. Kroulik, J.T., Burckley, L.A. & Wiseman, H. G. 1955 a. The microbial populations of the green plant and of the cut forage prior to ensiling. J. Dairy Sci. 38: 256—262. Kroulik, J.T., Burckley, L.A., Gordon, C.H., Wise- man, H.G. & Melin, C.G. 1955 b. Microbial activities in alfalfa and orchard grass ensiled under certain con- ditions in experimental silos. J. Dairy Sci. 38: 263—272. Langston, C.W., Bouma, C. & Conner, R.M. 1962. Chemical and bacteriological changes in grass silage during the early stage of fermentation. 11. Bacterio- logical changes. J. Dairy Sci. 45: 618—624. Lindgren, S., Pettersson, K., Jonsson, A., Linowall, P. & Kaspersson, A. 1985 a. Silage inoculation. Swedish J. agric. Res. 15: 9—lB. Lindgren, S., Pettersson, K., Kaspersson, A., Jonsson, A. & Lincvai l, P. 1985 b. Microbial dynamics during aerobic deterioration of silages. J. Sci. Food Agric. 36: 765—774. McDonald, P. 1981. The Biochemistry of silage. Chi- chester, John Wiley & Sons. 226 p. Moon, N.J., Ely, L.O. & Sudweeks, E.M. 1980. Aerobic deterioration of wheat, lucerne and maize silages prepared with Lactobacillus acidophilus and a Candida spp. J. Appi. Bact. 49; 75—87. Nilsson, G. & Nilsson, P.E. 1956. The microflora on the surface of some fodder plants at different stages of maturity. Archiv fur Microbiologie. 24: 412—422. Ohshima, M, & McDonald, P. 1978. A review of the changes in nitrogenous compounds of herbage during ensilage. J. Sci. Fd. Agric. 29; 497—505. Ojala, R. & Poutiainen, E. 1986. The use of lactoacid bacteria preparate in grass silage making. Helsingin yli- opisto. Kotieläintieteen laitoksen tiedote nro 8: I—3B. 376 Rauramaa, A., Setälä, J., Moisio, T., Heikkilä, T. & Lampila, M. 1987. The effect of inoculants and cel- lulase on the fermentation and microbiological com- position of grass silage. Part 1. Biochemical changes in the silages. J. Sciest. Agric. Soc. Finl. 59: 361—370. SitEY, P. & Damoglou, A.P. 1985. The effect of three commercial silage additives on numbers of lactobacilli entering the silo at the onset of fermentation. FEMS Microbiology Letters 30: 107—110. Stirling, A.C. & Whittenbury, R. 1963. Sources of the lactic acid bacteria occurring in silage. J. Appi. Bad. 26 (81): 86—90. Woolford, M.K. 1984. The silage fermentation. New York. 350 p. Woolford, M.K., Honig, H. & Fenlon, S.J. 1979. Un- tersuchungen liber den aerobem Abbau in Silage mit einer Labormetode. Teil 3: Microbiologische, physika- lische und chemische Veränderungen vvährend des aeroben Abbau in frischer und angewelkter Grasssilage. Das Wirtschafteigene Putter 25 (2/3): 158—177. Ms received June 17, 1987 SELOSTUS Maitohappobakteeri- ja sellulaasientsyymi- lisäyksen vaikutus nurmisäilörehun käymiseen ja mikrobiologiseen koostumukseen. II Mikrobiologiset muutokset säilörehussa Aino Kauramaa, Jouko Setälä, Tauno Moisio ja Seppo Sivelä Valio Meijerien Keskusosuusliike Tutkimus- ja tuotekehittelyosasto PL 176, 00181 Helsinki Terttu Heikkilä ja Martti Lampila Maatalouden Tutkimuskeskus Kotieläinhoidon Tutkimuslaitos 31600 Jokioinen Kesällä 1985 valmistettiin koiranheinä-timoteinurmesta (1 : 2) neljä koesäilörehua Maatalouden Tutkimuskeskuk- sen Lintupajun koetilalla. Säilöntäaineet olivat: maito- happobakteerit,sellulaasientsyymi ja AIV 11 -liuos. Yk- si rehu valmistettiin ilman säilöntäainetta. Säilönnän aikana käymisiä seurattiin kairaamalla ote- tuista näytteistä sekä kemiallisten analyysien että mikro- biologisten määritysten avulla. Rehunäytteistä määritet- tiin viisi mikrobiryhmää: (I) bakteerien kokonaismäärä, (2) maitohappobakteerit, (3) voihappobakteerien itiöt, (4) koliryhmän bakteerit ja (5) homeet ja hiivat. Mikrobitulokset osoittavat, että säilönnän alkuvaihees- sa lisääntyvät lähinnä koliryhmän bakteerit ja maitohap- pobakteerit. Maitohappobakteerien lisäys rehumassaan vaikutti vähentävästi koliryhmän bakteerien kasvuun ver- rattaessa tätä rehua ilman säilöntäainetta ja sellulaasili- säyksellä valmistettuun rehuun. AIV 11 -liuoslisäys hil- litsi kaikkia käymisiä ja suosi homofermentatiivisia mai- tohappobakteereita. Hyvistä säilöntäolosuhteista johtuen rehuissa oli vain vähän voihappobakteerien itiöitä, hii- voja ja homeita. Eniten hiivoja esiintyi sellulaasilla ja mai- tohappobakteerilisäyksellä valmistetuissa rehuissa. 377