Bioresource Technology 40 ( 1992 ) 35-41
Grape Marc as a Source of Feedstuff after Chemical Treatments and Fermentation with Fungi C. Vaccarino, R. B. Lo Curto, M. M. Tripodo, R. Patan6 & A. Ragno Dipartimento di Chimica Organica e Biologica, Universith di Messina, Italy (Received 13 September 1990; revised version received 8 February 1991; accepted 14 February 1991 )
The results are presented of research linked to the utilization of grape marc as a substrate for SCP production, after pretreatments with Na2CO 3 and NaOH at 120°C, by Aspergillus sp., Geotrichum candidum and Trichoderma viride. Submerged fermentations were slow and gave rise to only a modest increase of crude protein (from 2"2% to about 4% nitrogen on d.m.) and invitro digestibility (from 15"0% to about 26.0%). Fermentations carried out on the liquid resulting after NaOH pretreatments gave rise to modest SCP growth, probably because of the inhibitory effects of polyphenols (hydrolyzed by NaOH treatment, but recomposed under the fermentation conditions), and of the lignin degradation products present in the liquid. Experiments suggest that a possible way for a profitable utilization of this waste material is to submit it to consecutive alkaline and acidic treatments, which gives two products, whose value as feedstuffs for ruminants should be confirmed with in-vivo alimentation assays. Key words: Grape marc, sCP, feedstuff.
INTRODUCTION Exhausted grape marc may be included among the lignoceUulosic waste materials as a candidate for SCP production by fermentation. It has almost no value, but is available in large quantifies from the alcohol distilleries, where the sugar-rich marc coming from the vats is processed. However, notwithstanding the significant content of proteins and polymeric carbohydrates, its utilization as a feedstuff is prevented by its indigestibility, as well
as by its high content of polyphenols (Aguilera, 1987). Introductory tests (Vaccarino et al., 1982a, b) showed that for utilizing grape marc as a substrate for fermentations, chemical pretreatments of the types already suggested for other lignocellulosic materials which weaken the bonds between lignin and cell-wall polysaccharides, would be necessary (Worgan, 1976; Linko, 1977; Callihan & Clemmer, 1979). Alkaline pretreatments also reduce the inhibitory effects of polyphenols (Chavan et al., 1979; Reichert etal., 1980; Lamptey et al., 1986), even if the mechanism causing the antinutritional properties of polyphenols is not yet completely clear (Muller-Harvey, 1989). A subsequent systematic research (Vaccarino et al., 1987), showed that NaOH (1%) solution at 120°C is very effective in degrading and solubilizing grape marc crude fibre, but causes a notable solubilization also of nitrogen compounds; Na2CO 3 (1%) solution at 120°C and SO 2 (1.5%) solution at 100°C have, in this order, lower effects. The same research showed also that these pretreatments are much less effective if the marc is previously dried. In this paper the results of three series of tests are presented, in which dump marc, stored by addition of antifermentative or frozen, was utilized. Pretreatments were effected with N a 2 C O 3 o r NaOH at 1200C.
METHODS First series of assays
Exhausted, dump grape marc with a water content of about 60%, with pips removed, whose composition is shown in Table 1, was stored at room temperature for 2-3 months after addition of
35 Bioresource Technology 0960-8524/92/S03.50 © 1992 Elsevier Science Publishers Ltd, England. Printed in Great Britain
C. Vaccarino, R. B. Lo Curto, M. M. Tripodo, R. PatanO, A. Ragno
Table 1. Characteristics of grape marc after alkaline treatment and after submerged fermentation with Aspergillus sp. (% on dry matter)
a Initial dump grape marc
b Dump marc stored with 6% Na2CO 3 c Dump marc stored with 10% Na2CO 3
c Marc further treated with 6% Na2CO3 f Sofids g Final meal"
b Marc further treated with 6% NaOH h Solids b i Final meal'
b Marc further treated with 6% NazCO3 a d Solids b
e Final meaF
"By autoclaving, see Methods. bArter pretreatment. ~After fermentation.
antifermentatives: K2S205 (potassium metabisulphite) 0.4% on dry weight o r N a e C O 3 6% and 10% on dry weight. This material was introduced into an autoclave and treated with 3-10% Na2CO 3 or 6-13% NaOH (on dry matter weight) at 120°C for 2 h; the necessary water was added to bring the dry weight in water to 7"5%. Samples of the suspended solids were separated by filtration, dried at 70°C and analyzed. The pretreated material was utilized as substrate for submerged fermentations in an LKB 1601 fermenter, of total volume 16 liters, fitted with devices for controlling and recording temperature, dissolved 02, pH stirrer rpm. Tap water was added to bring the total dry matter concentration to 4% and supplements of nitrogen and phosphorus were given by addition of urea and H3PO 4 to obtain N and P contents of 3% and 0.5% on dry matter, respectively. The pH was brought to values varying between 4.5 and 6-0, by addition of n 2 s o 4. The fermenter was then sterilized in an autoclave at 120°C for 30 min, and 400 ml of inoculum, prepared from a strain of Aspergillus sp. previously isolated from a heap of the same exhausted marc, was added. The microorganism, which, tested by HPLC (according to Francis et al., 1982) was nonaflatoxinogenic, was grown at 47°C on yeast extract (0.25%) and malt extract (0.25%).
In some cases, instead of yeast the same pretreatment liquid as for the fermentations, enriched with urea, was employed. The fermentations were carried out at 47°C, and the pH was maintained by automatic addition of NH4OH solution or phosphoric acid. During fermentation, samples were taken for analysis. At the end, the material was discharged, filtered through paper, and the solid d i e d by lyophilization, to give a 'final meal'. The following analyses on initial and pretreated marc, as well as on the final meal were carried out: crude protein by a Kjeldahl method (N x 6.25); neutral detergent fibre (NDF) according to Goering and Van Soest (1970); ash by calcination; total in-vitro digestibility by the method of Jones and Hayward (1975) as modified by Allison and Borzucki (1978). Second series of assays In this series a technique called 'surface fermentation', already used with orange peel (Vaccarino et al., 1989a) was applied. This technique consists of carrying out the fermentation in a small liquid layer above the surface of the substrate. The solid from stored and alkali-pretreated grape marc as before described, but in presence of less water (dry weight in water: 11-12%), was put
Grape marc as animal feed on to stainless-steel trays (200 x 200 x 30 mm), where the necessary water was added to bring the total dry matter content to about 80 g/liter. In these conditions a liquid layer of 3-5 mm deep on the surface of the solid was formed. Supplements of nitrogen and phosphorus were added as in submerged fermentations, then the trays were sterilized at 120°C for 30 min and inoculated with strains of Aspergillus sp. and Trichoderma viride prepared as described before. Every 3 days, from each tray, a film composed of the hyphae of the microorganism, which covered the whole liquid surface, was picked up, washed with tap water and analyzed.
Third series of assays In this series of assays, fermentation was carried out only on the liquid using Geotrichum candidum and Trichoderma viride whose characteristics are potentially better than those of Aspergillus sp. This system, which follows a concept also applied in previous research (Vaccarino et al., 1989b) is based on the fact that NaOH attacks strongly, and solubilizes, lignin and hemicelluloses, but cellulose only slightly. The objective was the utilization of the solid residue resulting after NaOH treatment as a secondchoice feedstuff for ruminants, while the SCP, grown on the more easily metabolizable substances dissolved in the liquid, should be a valuable protein-rich feed additive. Dump grape marc stored at - 1 5 ° C without preservatives was utilized. Pretreatments were effected as described for the first series of experiments, using 8-17% NaOH on marc dry matter, at the temperature of 120"(2, for 2 h. After treatment, the product was filtered under vacuum, washed with tap water and dried at 70°C, to give a 'solid residue'. This product was analyzed as before. The filtrate and wash liquids, mixed together, were utilized for fermentation. As it was noted that on acidifying the alkaline liquid before fermentation a large quantity of an insoluble precipitate was formed, the removal of this precipitate, before fermentation, was also investigated. By adding HESO 4 at 100°C, the pH was brought to 4.6, 3.7, 3.2 and 2.5 and the corresponding insoluble materials were recovered by filtration through paper, washed with tap water, dried by lyophilization and analyzed. Fermentations were carded out both on the liquids obtained directly after alkaline pretreat-
ments and on the liquids resulting from the subsequent acidification, as described before. Sterilization was effected in an autoclave at 120°C for 30 rain; in some cases the temperature was lowered to 115°C. The pH before fermentation was generally brought to 5-0 but assays at pH 4-5 and 5-5 were also carried out; the temperature was maintained at 28"C, as microorganisms Geotrichum candidum and Trichoderma viride were employed. After fermentation, the SCP was collected by filtration on paper, dried and lyophilized and analyzed as for the first series of assays. In all cases the analyses shown in the text and tables and figures refer to single assays which may be considered 'typical', because homogeneous with other assays carried out in the same conditions on the same raw material. Each figure is the average of at least three analyses.
RESULTS AND DISCUSSION First series of assays In Table 1 the results of three typical assays are presented. As the use of KaSEO5 showed some problems (development of SOa) and no special advantage in these assays, as well as in those subsequently described, Na2CO 3 as antifermentative was always employed. In line a some characteristics of initial (dried at 70°C) grape marc are reported, which differ very slightly from those of the dried marc (DEM) described previously (Vaccarino et al., 1987). Figures in lines b and c show that conservation with Na2CO 3 (6% and 10% on dry matter, respectively) does not affect notably the characteristics of the material, except for a considerable increase in ash content. However, the 'solids' (i.e. the solid materials resulting after the treatments at 120°C) whose characteristics are reported in lines d, f, and h, are more digestible, which may be attributed to the action of alkalis on the lignincellulose bonds. As far as the final meal is concerned, the figures of lines e, g and i show an increase of nitrogen content and a further slight improvement of digestibility. By comparing these values with those of the initial grape marc, there is evidence that the SCP growth by fermentation has been in all cases modest. This, notwithstanding that the abovementioned figures in Table 1 are higher than the actual, because all analyzed samples of solids were not washed and therefore also contained the
C. Vaccarino, R. B. Lo Curto, M. M. Tripodo, R. Patan~, A. Ragno
soluble products dissolved in the water which adhered to them (residual sugars, soluble proteins, mineral salts, etc.). Stronger alkaline treatments, by utilizing up to 13% NaOH on marc dry matter, caused higher solubilizations, as was expected (Vaccarino et al., 1987), but no significant improvement of the characteristics of the final meal (data not reported). Furthermore, all fermentations generally required long induction times and were completed only in at least 2-3 days and sometimes more. A first reason for these negative results may be that in submerged fermentations the enzymic activities of the microorganisms (degradation of the residual lignocellulosic material and utilization of the simple products formed) develop successively, but relative velocities vary according to operative conditions. A second reason is probably the inhibitory action of polyphenols on the enzymes. It is true that alkali treatment should have decomposed the hydrolyzable tannins present in the marc, but, as by lowering the pH this reaction is reversible, under the fermentation conditions these tannins are probably recomposed and able again to act negatively on the fermentation (McLeod, 1974). Because of the modest results described above, submerged fermentations were discontinued. Second series of assays In Fig. 1 the trend is presented of a 32 days fermentation carried out with Aspergillus sp. which is •
[3 • 3.0
2,0 40 1,5,
t,O, 20 .5,, ! 0
1"2 1"5 !'8 2"1 2"4 27 3"0 3 3 DAYS Fig. 1. Surface fermentation with Aspergillus sp., analyses of the mycelial film. o, Total in-vitro digestibility (% on dry matter); ", crude protein (% on dry matter); m, dry matter productivity (g/day per 100 g original dry matter); A, protein productivity (g/day per 100 g original dry matter).
not different from those of assays with Trichoderma viride. The graph shows that the quality of SCP improves during fermentation (crude protein content rises from about 19% to 23% and invitro digestibility from about 43% to 48%). But the dry matter and protein productivity decrease quickly with time and this outweighs all advantages presented by this system (simplicity, possibility of carrying out the fermentation without sophisticated devices for avoiding contamination, etc.). This conclusion is the same as that drawn after similar assays carried out on orange peel with Trichoderma viride and Geotrichum candidum (Vaccarino et al., 1989a), and may probably be explained by the increasing concentration of catabolites in the limited water available for microorganism growth. Third series of assays NaOH pretreatments confirmed the results of previous work (Vaccarino et al., 1987). As shown by the graph of Fig. 2 (which deals with assays carried out on frozen dump marc of average quality), on increasing the ratio NaOH: initial marc dry matter, the weight of solid residue decreases rapidly, due to solubilization of hemicelluloses and lignin, but the in-vitro digestibility becomes higher, due to the action of NaOH on lignin-cell wall bonds as well as on polyphenols, which are partially hydrolyzed and dissolved. If, for instance, the NaOH addition is limited to 10% of initial dry matter a solid residue containing about 42-5% crude cellulose, 26-0% lignin, 7.5% crude protein and with 27-0% in-vitro digestibility is obtained. By comparing these characteristics with those of the initial marc, whose digestibility was 15.2%, we may conclude that this product is potentially a better feedstuff for ruminants, as was expected. As far as fermentation is concerned, all tests showed that Trichoderma viride was more suitable than Geotrichum candidum, probably because it metabolizes to a certain extent the degradation products of lignin and is more resistant to polyphenols. Nevertheless, with Trichoderma viride we generally noted only a modest yield of good quality SCP. In some cases, at the end of fermentation, we collected a very high weight of solid material: up to 15 g per 100 g initial grape marc dry matter, with a significant nitrogen content (5.0-5.6%). But the in-vitro digestibility was very low: 6.0-7-1% (which was increased to 10-7-11.4% by a sub-
Grape marc as animal feed
Fig. 3. Acidicprecipitate (see Methods) characteristics: *, precipitate weight (g/100 g original grape dry matter); % crude protein content (% on dry weight); e, in-vitro digestibility(% on dry matter).
g/|0Of[ d . m
.... ~ .... 1"o . . . . 1~" Fig. 2. Characteristicsof solid residues after NaOH treatments by autoclaving(see Methods). o, Residue %, weighton originaldry matter; o, residue in-vitro digestibility(% on dry matter); ", residue cellulose content (% on dry matter); K, residue lignincontent (% on dry matter); D, residue protein content (% on dry matter).
sequent 2 h heating at 120°C). A further investigation showed that during the sterilization at 120°C which preceded fermentation, a fair quantity of insoluble material separated from the liquid and contaminated the SCE By removing this material by filtration before fermentation, a SCP with about 50% digestibility was obtained; but the crude protein content was only 20-22% and the yield extremely low. The material removed on filtration showed an in-vitro digestibility higher than expected, 20-21%, which is in contrast with the above mentioned low digestibility of the mixture with SCE This could perhaps be explained by supposing that this material contains some digestible compounds that during fermentation are utilized for the microorganism metabolism or dissolved in water, so that a large quantity of a practically indigestible residue remains mixed with the final SCP, whose production is in reality low. The 5.0-5.6% nitrogen content found in the mixture
should be mainly attributable not to the SCP proteins, but to those (probably scarcely digestible) of the precipitate, and to other nitrogen compounds. By limiting the sterilization temperature to 115°C, no separation of insoluble material was noted, but the fermentation gave rise to, again, only a modest quantity of SCP. From the foregoing, the conclusion may be reached that the liquid obtained after NaOH pretreatment, notwithstanding the large quantity of dissolved substances, is not suited for SCP development. This was also confirmed by the fact that fermentations were always characterized by long delay in starting, a short-time exponential phase and low oxygen utilization. In order to have some information on the above mentioned dissolved substances, the alkaline liquid was submitted to acidic treatments, as described in Methods. On adding H2SO 4 at 110°C, a precipitate was formed, which was called 'acidic precipitate' to distinguish it from the 'thermic precipitate' previously described. Figure 3, which deals with exhausted grape marc stored for about 3 weeks in a heap before freezing for storage, shows that on lowering the pH the weight of this acidic precipitate increases, but its digestibility decreases. As the crude protein content is almost constant (19-20%), it may be concluded that the proteins contained in the liquid become progressively insoluble with acidification. These proteins are 78-80% insoluble in hot water and
C. Vaccarino, R. B. L o Curto, M. M. Tripodo, R. Patan~, A. Ragno
may be therefore considered mainly 'true' proteins. If, instead of being carried out at 100°C the acidic treatment at pH 5 was carded out at 120°C, the precipitated material increased notably, obviously because of the formation of a thermic precipitate, which added to the small quantity of acidic precipitate. The digestibility and the nitrogen content of this mixture were only slightly lower than in the case of treatment at 100°C. It is noteworthy that all figures concerning weight, protein content and digestibility of acidic precipitate (as well as those of solid residue obtained by NaOH treatment) vary considerably in relation to the nature and conservation state of grape marc, and are presumably dependent on the extent of the lignification process. So, by utilizing freshly collected grape marc at pH 2.5 an acidic precipitate with 32.7% in-vitro digestibility was obtained. It may be assumed that the main component of the acidic precipitate is alkali-lignin, a well known product of lignin degradation, which is still a polymeric compound (molecular weight: about 1000), solubilized by NaOH but insoluble in acidic solution. This alkali-lignin, which is recovered as a byproduct of the Kraft cellulose process, has some industrial uses (Szmant & Harris, 1986), but, in the present case, the acidic precipitate, mainly on account of its protein content, could be a valuable feedstuff for ruminants. In fact, its low digestibility could be only apparent, because it was measured with a test (Allison & Borzucki, 1978) carried out at a low pH, at which the product is insoluble, while it dissolves again if the pH is increased to the values of ruminant digestion (6"0-6-5). Even the presence of a large quantity of active polyphenols is not necessarily a problem: the formation of complexes with hydrolyzable tannins may preserve proteins from ruminal degradation and allow their subsequent utilization in the stomach, where the pH is lower and the complexes are broken (McLeod, 1974). Only in-vivo tests with ruminants could give an answer to these questions. The possibility of carrying out fermentation by utilizing as a substrate the liquid obtained after removing the acidic precipitate, was also investigated. The result, as was expected, was the same as fermentations carried out after separation of the thermic precipitate: an SCP of acceptable quality in very low yield. From all the foregoing the conclusion may be reached that the liquid obtained from exhausted
grape marc after a strong alkaline treatment (which partially solubilizes lignin, hemiceUuloses and other polymeric substances) does not seem suitable for fungal SCP growth (at least with the species Trichoderma viride and Geotrichum candidum ). This may be explained by the fact that alkalilignin and other degradation products of NaOH treatments are not good enough nutrients for these microorganisms, as well as by the presence in the liquid of polyphenols not hydrolyzed by NaOH and which are recomposed on lowering the pH. On the other hand, the removal (by acidification and/or heating above 100°C) of a large quantity of the NaOH-dissolved materials gives rise to a liquid which is too poor in nutrients for growing SCP. As far as the problem of a profitable utilization of this waste material is concerned, a possible way is probably the utilization as feedstuff of both solid residue and acidic precipitate obtained by successive alkaline and acidic treatments, as described. In-vivo tests on ruminants, which require large quantifies of the product, are needed to confirm this suggestion.
REFERENCES Aguilera, J. E (1987). Improvement of olive cake and grape by-products for animal nutrition. In Degradation of Lignocellulosic in Ruminants and in Industrial processes, ed. J. M. Van Der Meer, B. A. Rijkens & M. P. Ferranti. Elsevier Applied Science, London. Allison, M. & Borzucki, R. (1978). Cellulose methods for the efficient digestion of grasses and brassicas. Journal of the Science of Food and Agriculture, 29, 293-7. Callihan, C. D. & Clemmer, J. E. (1979). Economic Microbiology, Vol. 4, ed. A. H. Rose. Academic Press, London, pp. 271-88. Chavan, J. K., Kadam, S. S., Ghonsikar, C. P. & Salunkhe, D. K. (1979). Removal of tannins and improvement of in vitro protein digestibility of sorghum seeds by soaking in alkali. Journal of Food Science, 44, 1319-21. Francis, O. J., Lipinski, L. J., Gaul, J. A. & Campbell, A. D. (1982). High pressure liquid chromatographic determination of aflatoxins in peanut butter using a silica gel-packed flow cell for fluorescence detection. Journal of A OA C, 65, 673-8. Goering, H. K. & Van Soest, P. J. (1970). Forage Fibre Analyses. USDA Agriculture Handbook No. 379. Jones, D. H. J. & Hayward, M. V. (1975). The effect of pepsin pretreatment of herbage on prediction of dry matter digestibility. Journal of the Science of Food and Agriculture, 26, 711-18. Lamptey, J., Moo-Young, M. & Robinson, C. (1986). Pretreatment of lignocellulosies for bioconversion. In Biotechnology and Renewable Energy, ed. M. Moo-Young. Elsevier Applied Science, London.
Grape marc as animal feed
Linko, M. (1977). Advances in Biochemical Engineering, Vol. 5, ed. T. K. Ghose, A. Fieehter & N. Blakebrough. Springer-Verlag, Berlin, pp. 25-48. McLeod, M. M. (1974). Plant tannins. Their role in forage quality. Nutrition Abstract and Reviews, 44 (11 ) 803-15. Muller-Harvey, I. (1989). Identification and importance of polyphenolic compounds in crop residues. In Physicochemical Characterization of Plant Residues for Industrial and Feed Use, ed. A. Chesson & E. R. Orskow. Elsevier Applied Science, London, pp. 88-101. Reichert, R. D., Fleming, S. E. & Sabwato, D. J. (1980). Tannin deactivation and nutritional improvement of sorghum by anaerobic storage of H20, HCI or NaOHtreated grain. Journal of Agriculture and Food Chemistry, 28, 824-9. Szmant, H. & Harris, J. (1986). Industrial Utilization of Renewable Resources. Technomic Publications Co. Lancaster, Basel, pp. 150-7. Vaccarino, C., Lo Curto, R., Tripodo, M. M., De Gregorio, A., Bellocco, E. & Laganh, G. (1982a). Produzione di SCP da vinacce pretrattate con SO2. Nota 2 Atti della Societ~
ltaliana di Scienze Fisiche, Matematiche e Naturali, 28, 15-22. Vaccarino, C., Lo Curto, R., Leuzzi, U., Tripodo, M. M., De Gregorio, A. & Cimino, G. (1982b). SCP production from SO2 pretreated grape marc. Note 1. Annali di Microbiologia, 32, 61-9. Vaccarino, C., Lo Curto, R., Tripodo, M. M., Bellocco, E., Laganh, G. & Patant, R. (1987). Effect of SO2, NaOH and Na2CO 3 pretreatments on the degradability and cellulose digestibility of grape marc. Biological Wastes, 20, 79-88. Vaccarino, C., Lo Curto, R., Tripodo, M. M., Patant, R., Laganh, G. & Schacter, S. (1989a). SCP from orange peel by fermentation with fungi -- submerged and surface fermentations. Biological Wastes, 29, 279-87. Vaccarino, C., Lo Curto, 1L, Tripodo, M. M., Patant, R., Laganh, G. & Ragno, A. (1989b). SCP from orange peel by fermentation with fungi -- acid-treated peel. Biological Wastes, 30, 1-10. Worgan, J. T. (1976). Foodfrom Waste, ed. G. C. Birch, K. J. Parker & J. T. Worgan. Applied Science Publishers, London, pp. 23-41.