JOURNAL OF THE SCIENTIFIC AGRICULTURAL SOCIETY OF FINLAND

187

Maataloustieteellinen Aikakauskirja
Voi. 48: 187—194, 1976

Cellulase digestion technique compared with the
in vitro digestibility of forages
Seppo Pulli
University of Helsinki, Department of Plant Husbandry, Viikki

Abstract. A red clover-grass mixture and a couch grass (E. repens) pure stand
were studied at eighteen different stages of maturity in Viikki. The analytical techniq-
ues were a cellulase digestion method and a two stage in vitro digestion technique.
A high correlation (r = .99*** for both types of stand and residual standard deviations
(RSD) of 0.9 and 1.3 respectively) was found between cellulase DM digestibility and in
vitro DM digestibility. The difference between cellulase and in vitro DM digestibility
increased as the plants advanced in maturity, because the effect of enzyme decreased,
as the plants were more matured. The relationship between cellulase DM diges-
tibility and plant protein content was very close, (r = ,99*** for the clover-grass mixture
and couch grass and RSD’s of 0.6 and 0.5 respectively). At the early stages of maturity
when the plant protein content and in vitro DM digestibility were high, the cellulase
technique digested more efficiently the clover-grass mixture than the couch grass. The
cellulase method was rapid, reproduciple and well suited to the evaluation of the plant
material studied.

7. Introduction

Forage feeding value is dependent on the levels of available energy and
protein and the palatability of the forage. The main determinant is considered
to be the amount of digestible energy per unit weight. As the feeding value of
the forage varies greatly according to the maturity, species and the environment
of the plant, the important point is to test not only the constituents, but also
the digestibility of the roughage. For this purpose, considerable efforts have
been made to develop a method which would give reliable digestibility results
without the need to resort animal trials.

Salo (1973) has reviewed the methods used in the estimation of the
digestibility. According to the review, the raw fibre has long been used as a
parameter of the digestibility (Me Meekan 1943, Griffith and Thomas 1935
and Kivimäe 1960). The techniques do not predict very well the digestibility
of feeding material. The normal acid fibre (NAF) (Griffith and Thomas
1955) and acid detergent fibre (ADF) (van Soest 1963) techniques are also not
considered adequate. The content of lignin has a little better correlation than
raw fibre with the digestibility of roughage (Armstrong et al. 1964). The
most important method used today is probably the in vitro rumen fermentation
technique. Especially the two stage technique of Tilley and Terry (1963)
is widely used to the assessment of digestibility.

https://www.c-info.fi/en/info/?token=OnoZ743sdq5S2IA4.qefel6-GRUnTDCQjXJDH0g.PM678vtGyuXY2YeIXoQC-RsMZ6r9f7NmdRPtxPS6WUbZl46xtzNDc56P8kILnpqx8s6v_COAx19dyvWasAUlciNqs-cuK7Tg3dXVateUQ8BdQ2OH4rZfv8dXSxPVEjy5dcXKatHtDRrz1Iq_6OoJoiXziVwTjDE6gw


188

An inherent disadvantage of fermentation techniques is their depen-
dence on a supply of rumen inoculum, requiring access to animals fitted with
rumen canulae and fed under controlled conditions to minimize variations in
the composition of the rumen liquor (Jones and Hayward 1973). Several
workers (Drew 1966, Troelsen and Hanel 1966, Yates and Allden 1966)
have reported variations in in vitro results due to variations between animals,
composition of diet, time of sampling relative to feeding etc.

In the technique of Tilley and Terry (1963), the first stage is a digestion
by enzymes originating from the rumen microbes. The digestion is completed
by a commercial pepsin preparation. The pepsin technique has directed atten-
tion to the use of other enzyme preparations. The use of freeze-dried enzyme
preparations as digesting agents has obvious advantages over in vitro methods,
which involve more difficult techniques and are subject to wide variations.
However, Donofer et al. (1963) in their first attempt found a poor correltion
between cellulase and in vivo digestibility. Jarringe et al. (1970) found a
better correlation between cellulase and in vivo digestibility than between in
vitro and in vivo digestibility. Guggolzt et al. (1971) also found a good
correlation between in vivo and cellulase digestibility. Abe et al. (1973) used
4, 8 and 12 hours cellulase digestion periods and found excellent correlations
between cellulase digestibility and in vivo digestibility, as teh values r =

99,7, 99.9, 99.7 respectively and p<o.ol show.
Bailey and Jones (1971) and Jones and Bailey (1972) have used a crude

cellulase preparation from Trichoderma viride as a hydrolytic agent to study
the rate of hydrolysis of cell walls in grasses. The effect of the treatment was
relatively poor, but the correlation with in vivo digestibility satisfactory.
Later Jones and Hayward (1973) examined the use of a commercially avai-
lable enzyme from the same fungus as a digesting agent for predicting herbage
digestibility. A high correlation (r = .92***, RSD = 2.5) was found between
results from a simple one stage enzyme digestion and in vivo DM digestibility
for a range of grass apecies and varieties. A crude cellulase preparation showed
cellulase, hemicellulase and proteolytic activity.

This paper examines the use of a crude cellulase preparation as a digesting
agent compared with a determination of in vitro digestibility. The material
consists of two types of roughage, a red clover-grass mixture and couch grass
(Elytrigia repens) sampled at eighteen different stages of maturity. Other
objects of the study are the relationships between the cellulase digestibility
and plant protein content at different stages of plant maturity. The main
purpose is to test the feasibility of cellulase techniques for agronomists and
plant breeders, so that the use of animals could be avoided.

2. Experimental

2.1. Materials
Two types of roughage, a red clover grass-mixture (average of samples

during season: red clover 31 %, timothy 32 %, Festuca 24 % others 13 %)

and couch grass (E. repens) were studied at eighteen different stages of maturity



during the 1974 growing season. The first sampling date was May 23. Sampling
was completed on July 22. Four replications of the material were collected
three times a week, oven-dried at 70° C and ground to pass a 0.6 mm sieve.

2.2. Analytical techniques

In vitro dry matter digestibility was determined by the two stage technique
of Tilley and Terry (1963). The cellulase used was a preparation from
Trichoderma viride, marketed by BDH Ltd, Poole, Dorset, England. The one
stage procedure used for cellulase digestion is described by Jones and Hayward

(1973). In the procedure, 200 mg of ground sample were weighed into 25 ml
screw-capped bottles. 20 ml of buffered enzyme were added and placed in an
incubator at 40° C for 48 hours, shaking twice daily. After incubation, the
mixture was filtered through filter paper, previously dried and weighed. The
residue was washed well with water and acetone then dried overnight at 100° C
and weighed. The cellulase digestibility was calculated as the percentage
digested of the orginal 200 mg sample. Nitrogen contents of the plant material
were determined by the conventional Kjeldahl procedure. Crude protein was
calculated by using the coefficient 6.25.

3. Results

3.1. Cellulase and in vitro digestibility
The samples collected from May 23 to July 22 showed cellulase diges-

tibility values of 74.2—44.4 % for the clover-grass mixture and 76.1—40.3 %

for the pure couch grass stand (Table 1). The respective in vitro values were
82.1 and 84.9 58.0 %. At early stages of maturity when diges-
tibility was high the cellulase technique gave relatively higher values than at
later stages when digestibility was low. In consequence, when cellulase and
in vitro values are compared, the difference between the two techniques was
9.19.5 units at early stages of maturity and 17.3—18.5 units at advanced
maturity (Table 1).

Correlations between cellulase and in vitro digestibility were high in both
types of stand, as shown by the values r = .99*** for the clover-grass mixture

Table 1. Raw values of cellulase digestibility and in vitro digestibility (%) of a clover-grass
mixture and couch grass at different stages of maturity.

Digestibility Cellulase In vitro Difference
c * ass range mean range mean %-umts

Clover-grass high 74.2 66.5 70.4 82.1 77.4 79.9 9,5
medium 63.9-50.3 57.1 76.6-66.2 71.4 14.3
low 47.7-44.4 46.1 65.4-61.4 63.4 17.3

Couch grass high 76.1-69.4 72.8 84.9-78.8 81.9 9.1
medium 64.9-53.2 59.1 76.9-67.5 72.2 13.1
low 47.5-40.3 43.9 66.7-58,0 62.4 18.5

6 189



190

Table 2. Correlation of cellulase digestibility with in vitro digestibility and protein contents
of plant material.

Residual
, Mean value Range of Coefficient .

. ,Material ° standard Regression
equatiofor x x r r 2deviation

Correlation of cellulase digestibility (x) with in vitro digestibility (y)
Clover-grass 59.1 74.2-44.4 .99*** 98.6
Couch grass 59.5 76.1-40.3 .99»*» 98.0

0.9 y = 0.64 x + 34.07
1.3 y = 0.71 x + 30.69

Correlation of cellulase digestibility (x) with protein content (y)
Clover-grass 60.5 74.2-44.4 .99»** 98.9
Couch grass 61.0 76.1—40.3 .99*** 99.6

0.6 y = 0.50 x - 9,52
0.5 y = 0.57 x - 12.65

and r = .99*** for couch grass (Table 2). The residual standard deviations
(RSD) were 0.9 and 1.3 resepctively.

Regression lines (Fig. 1) show a slight difference in slopes for the two types
of material studied. At a high level of in vitro digestibility, the cellulase
digestibility values for couch grass are lower than those for the clover-grass
mixture. This result may be due to the higher protein content of couch grass,
as pointed out later. In Fig. 1, the dotted line represents the regression obtained
by Jones and Hayward (1973) between cellulase and in vivo digestibility.
Their regression line agrees very well with the results obtained at Viikki.

3.2. Cellulase digestibility and protein content

The cellulase digestibility follows closely plant protein content as the crop
stand advances in maturity. Table 2 shows a good correlation between the
cellulase digestibility and the protein content of the material studied. The
values obtained were r = .99*** for the clover-grass mixture and r = .99***
for couch grass. The RSD values were 0.6 and 0.5 respectively. As mentioned
earlier, at the same level of protein content the digestibility of couch grass
was lower than the digestibility of the clover-grass mixture. (F'ig. 2).

3.3. Development of the digestibility during the growing season

The use of the cellulase digestion technique and the in vitro method were
studied at eighteen different times during the growing season. The results
(Fig. 3) show a cubic effect of maturity upon digestibility for both of the
materials studied. The two types of stand showed very similar trends. The
regression equations between time (x) and DM digestibility (y) were as follows:

Grass-clover: Cellulase y = 73.7 + 0.0801 x 0.0901 x 2 + 0.0031 x 3
In vitro y = 91.9 - 0.0354 x - 0.0128 x 2 + 0.00014 x 3

Couch grass: Cellulase y = 75.1 + 0.1601 x 0.0031 x 2 4- 0.00038 x 3
In vitro y = 84.1 4- 0.0445 x - 0.0172 x 2 4- 0.00016 x 3

Cellulase and in vintro values for the clover-grass mixture and cohuch
grass (Fig. 3) were identica at the beginning of Jyly and represented cellulase



191

and in vitro DM values of 50 % and 67.5 % respectively. This digestibility
level represents roughly a 15 % protein content of the plant material as
indicated in Fig. 2.

Fig 1. Regression between cellulase and in vitro DM digestibility for clover-grass mixture
and couch grass compared with regression between cellulase and in vivo DM digestibility
obtained by Jones and Hayward for various grasses.

Fig 2. Regression between cellulase DM digestibility and plant DM protein content for
clover-grass mixture and for couch grass.



4. Discussion
In the study, the commercially available cellulase preparation was found to

be useful. In the digestion the buffer-enzyme mixture worked uniformly
throughout the series of tests. The use of a 48 hour digestion period proved
to be satisfactory. Shorter periods of digestion have also been used successfully
(Abe at al 1973). The filtration of the digested material was made with filter
paper instead of crucibles as recommended by Jones and Hayward (1973).
The use of filter paper needs special care in drying and weighing the paper.
The use of acetone as an agent did not cause any difficulties although some
loss of filter paper (1.5—2.0 mg/paper) was found.

The enzyme digestion technique produced results highly correlated with
those of the in vitro digestion technique, which is widely used in Finland.
As mentioned earlier, several studies revealed good correlations between cellulase
digestibility and in vivo digestibility and prove the method reliable. As in
vivo and in vitro digestibility values are practically identical, the present
results support earlier findings. It is suggested that studies on the feasibility
of using enzyme preparations for estimating the digestibility of legumes
would be wortwhile.

As the cellulase technique has many advantages over the in vitro method,
the use of the new method is recommended for agronomists and plant breeders,
especially when rumen liquor is not available.

Acknowledgements. The author is grateful to Professor Esko Poutiainen and Dr. Maija-Liisa
Salo, Dept, of animal Husbandry, for their interest in this work and for providing laboratory
facilities for the tests.

Fig. 3. Cellulase and in vitro digestibility of clover-grass mixture and couch grass at different
stages of maturity during the growing season.

192



193

REFERENCES

Abe, A., Horii, S. & Kameoka, K. 1972. Development and application of cellulase hydrolysis
for predicting digestibility of roughage I. Digestion of cell-well carbohydrates by
cellulase. Jap. J. Zootechn. Sei. 43:141 145, Ref. Nutr. Abstr: Rev. 43:4047—49.

Armstrong, D. G., Alexander, R. H. &Me Gowan, M. 1964. The use of in vitro diges-
tibilities of dried grasses for the prediction of their energy values for ruminants. Proc.
Nutr. Soc. 22; XXVI-XXVIII.

Bailey, R. W. & Jones, D. I. H. 1971. Pasture quality and ruminant nutrition. 3. Hydrolysis
of rye-grass structural carbohydrates with carbohydrases in relation to rumen digestion.
N. Z. J. Agric. Res. 14:847-857.

Donofer, E., Niemann, P. J., Crampton, E. W. & Lloyd, G. E. J. 1963. Dry matter
disappearance by enzyme and aqueous solution to predict the nutritive value of foreages.
J. Dairy Sci, 46. 965-970.

Drew, K. R. 1966. The in vitro prediction of herbage digestibility. Proc. N. Z. Soc. Anim.
Prod. 26:52-70.

Griffith, G. ap. & Thomas, D. C. 1955. The use of normal acid fibre and A.O.A.C. fibre
determinations for the estimation of herbage digestibility. Agric. Progr. 30; 124 128.

Guggozl
, J. Saunders, R. M.. Kohler, G. O. & Kloppenstein, T. J. 1971. Enzymatic

evaluation of processes for improving agricultural wastes for ruminant feeds. J. Anim.
Sci. 33: 167-170.

Jarrige, P., Thivand, P. & Demarquilly, C. 1970. Development of a cellulolytic enzyme
digestion for predicting the nutritive value of forages. Proc. XI Int. Grassl. Congr.
Queensland, Australia, pp. 762 766.

Jones, D. I, H. & Bailey, R. W. 1972. The hydrolysis of cell-wall polysaccharides from freeze-
dried and oven-dried herbage by rumen and mould carbohydrases. J. Sci. Food Agric.
23:609-614.

Jones, D. I. H. &M, V. Hayward, 1973. A cellulase digestion technique for predicting the
dry matter digestibility of grasses. J. Sci. Fd. Agric. 24: 1419 1426.

Kivimae, A. 1960. Estimation of the digestibility of grassland crops from their chemical
composition. Proc. 8. Intern. Grassl. Congr. 466 470.

Mc Meekan, C. P. 1943. A note on the relationship between crude fibre and the digestibility
of organic matter. N. Z. J. Sci. Technol. Sect. A 25: 152 156.

Salo, Maija-Liisa 1973. Analytisk bestämning av vallväxternas fordervärde. N. J. F:s
symposium: Foraedling for kvalitet hos foderplanter, Okt. 1973.

Soest, P. J. van. 1963. Use of detergents in the analysis of fibrous feeds. I Preparation fo
fiber residues of low nitrogen content. 11. A rapid method for the determination of
fiber and lignin. A.O.A.C. 46: 825 835.

Tilley, J. M. A. & Terry, R. A. 1963. A two-stage technique for the in vitro digestion of
forage crops. J. Br. Grassl. Soc. 18: 104 111.

Troelsen, J. E. & Hanel, D. J. 1966. Ruminant digestion in vitro as affected by inoculum
donor, collection day and fermentation tim. Can. J, Anim. Sci. 46: 149 156.

Yates, N. G. & Allden, W, G. 1956. A study of herbage digestibility using an in vitro fer-
mentation technique. Proc. Aust. Soc. Anim. Prod. VI: 340 349.

MS received January 6, 1976



194

Selostus

Sellulaasi- ja in vitro-sulavuuden vertailua eri kasvuasteen nurmirehuilla

Seppo Pulli
Helsingin Yliopisto, Kasvinviljelytieteen laitos, Viikki

Tutkimuksessa verrattiin kemiallista sellulaasi-sulavuutta biologis-kemialliseen in vitro-
sulavuuteen apila-heinä sekanurmella ja juolavehnällä. Kasvustot edustivat vuoden 1974
kasvukautta ja näytteet 18 eri kehitysvaihetta kasvukauden alusta heinäkuun loppuun. Yksi-
vaiheinen sellulaasisulatus tehtiin Jones & Hayward (1973) menetelmällä ja kaksivaiheinen
in vitro sulatus Tilley & Terry (1963) metodiikalla. Raakaproteiinipitoisuus määritettiin
Kjeldahl-menetelmällä.

Sellulaasi- ja in vitro sulavuuden välillä vallitsi kummallakin kasvustotyypillä kiinteä
lineaarinen suhde, kuten korrelaatiokerroin r = .99*** osoittaa. Sellulaasi-entsyymi osoitti
suhteellisesti enemmän aktiivisuutta kasvin aikaisessa kehitysvaiheessa kuin myöhemmässä
kehitysvaiheessa, joten ero sellulaasi- ja in vitro sulavuuden välillä suureni kasvin vanhetessa.
Sellulaasisulavuudcn ja in vitro-sulavuuden välisen regressiosuoran suunta oli lähes yhteneväi-
nen Jones & Hayward (1973) saaman sellulaasi- ja in vivo sulavuuden välisen suoran
kanssa.

Sellulaasisulavuudcn ja proteiinipitoisuuden välillä vallitsi kiinteä korrelaatio (r = .99***)
molemmilla kasvustotyypeillä, mikä merkitsee sitä, että proteiinipitoisuutta voidaan pitää
tietyssä määrin myös kasvuston sellulaasi sulavuuden mittana. Lisätutkimuksia kuitenkin
tarvitaan mm. eri ravinnetasojen ja erilaisten streessitekijöiden, kuten kuivuuden vaikutuksen
selvittämiseksi.

Kun tutkituilla kasvustotyypeillä proteiinipitoisuus ja in vitro sulavuus olivat korkeita,
sellulaasi tehosi heinä-apila seokseen paremmin kuin juolavehnään. Tämä saattaa johtua tiet-
tyjen vaikeasti sulavien yhdisteiden kuten ksylaanien runsaasta olemassaolosta heinillä.
Sulavuuden muutokset kasvin kehitysvaiheen myötä edustivat kolmannen asteen regressio-
suoraa molemmilla kasvustotyypeillä.

Tutkimuksessa testattu sellulaasi-metodi osoittautui nopeaksi, luotettavaksi ja helppo-
käyttöiseksi. Menetelmän etuna in vitro-menetelmään verrattuna on se, että fistelieläimiä ei
tarvita ja menetelmä soveltuu käytettäväksi vaatimattomissakin laboratorio-olosuhteissa.

Useat ulkomaiset tutkimukset osoittavat läheisen yhteyden sellulaasisulavuudcn ja in
vivo-sulavuuden välillä. Koska in vivo- ja in vitro-sulavuudet ovat lähes yhtenäiset, tämän
tutkimuksen tulokset tukevat aikaisempia havaintoja.

Suoritettujen tutkimusten pohjalta voidaan sellulaasimenetelmää suositella kasvinviljely-
tutkimuksiin. Erikoisen käyttökelpoinen se on koeasemaolosuhteissa. Menetelmää voidaan
suositella myös nurmikasvien jalostajille, koska sillä saadaan nopeasti ja pienistä näytteistä
suhteellisen luotettavaa tietoa.