Optimization of steam treatment as a method to increase in situ degradability of oil palm (Elaeis guineensis) frond in Malaysia

K Bengaly, J B Liang^*, Z A Jelan^**, Y W Ho^* and H K Ong^***

Institut d'Economie Rurale (IER), Centre Régional de Recherche Agronomique,
(CRRA) de Mopti, BP 205, Mopti, Mali.
konis6@yahoo.com^Department of Animal Science, Universiti Putra, Malaysia, 43400 UPM Serdang, Selangor, Malaysia
Institute of Bioscience, Universiti Putra, Malaysia, 43400 UPM Serdang, Selangor, Malaysia
^Malaysian Agricultural Research and Development Institute (MARDI), POBox 12301, 50774 Kuala Lumpur, Malaysia.

Abstract

The effects of steam pressure and moisture content on fibre constituents and in situ dry matter degradation of oil palm frond were investigated. Newly harvested fronds were chopped and divided into three groups: untreated, steam-treated as fresh or pre-dried to 25 - 30 % moisture before steam treatment at pressures of 10 kg cm^-2 for 20 minutes, 12.5 kg cm^-2 for 7 minutes and 15 kg cm^-2 for 4 minutes. The extent of rumen microbial utilization of fibre was evaluated by correcting dry matter degradation values for the soluble and very fine particles.

The predominant effect of steam treatment was to solubilize the hemicellulose fraction of the fronds, with the pre-drying treatment 90% hemicellulose solubilization. The decrease of hemicellulose content was related to in situ degradability parameters of fibre by the following correlation coefficients: -0.84, -0.86, -0.96 and -0.87 for the rapidly degradable, the slowly degradable, the potential degradability and the effective degradable component of fibre, respectively. Steam treatment of fresh fronds at highest pressure gave progressively decreasing utilization of remaining cell wall material by rumen microbes.

For practical implications, treatment conditions of pre-dried material to 25 - 30 % moisture and the pressure of 10 kg cm^-2 for 20 min could be used for steam-treatment of oil palm frond as feed for ruminants.

Key words: Oil palm frond, rumen degradability, steam pressure

Introduction

Oil palm frond (OPF), the aerial part of the oil palm tree is composed of two main portions: the petiole and the leaflet. It is estimated that 24.4 million metric tonnes (dry matter basis) of OPF are harvested in Malaysia annually from its 2.5 million ha of oil palm plantation (Islam 1999). The frond is used as a by-product for feeding ruminant livestock (Abu Hassan et al 1996). The petiole, which is the woody part of the frond, represents more than 70% (w / w) of the whole frond, whereas the weight of the leaflet is less than 30%. Therefore, the proportion of petiole and leaflet portions, that is determined by age of the frond at time of harvest, would be the dominant factor determining the fibre composition of a maturing OPF.

The extent of microbial degradation of specialized plant material varies generally with the stage of maturity of the plant. As plant cells mature, their cell walls thicken and the concentrations of the structural carbohydrates increase, reducing degradability. Steam treatment has been shown to be a suitable processing technique for improving the nutritive value of low-quality roughages. The treatment generally increases the amount of available energy resulting from the solubilization of the structural carbohydrates, cellulose and hemicellulose (Oji and Mowat 1978; Hart et al 1981). Klopfenstein and Bolsen (1971) reported that both rate and extent of digestion of roughages containing high levels of cell wall are improved following steam application. However, the carbohydrate solubilization that occurs during steam treatment, which has beneficial effects on the microbial degradation of roughages may also lead simultaneously to the formation of caramelization products which are known to be inhibitory to the growth of many microorganisms (Nesse et al 1977). The extent of the formation of these inhibitors has been shown to be related to the effective heating time, temperature, pressure, and the moisture content of the substrate (Goering et al 1973). At a particular steaming pressure, temperature or treatment time, the initial moisture content of the material may be of greater importance. Because water increases the rate of the reaction, and the most reactive carbohydrates being hemicellulose and soluble carbohydrates, the maximum catalytic activity of water is at 30% of the sample weight (Van Soest 1994). Steam treatment conditions vary considerably in the published literature, but most use dry materials.

The objective of this study was to examine the changes in OPF cell wall composition after treatment under different steaming conditions (moisture and pressure) and relate those changes to the pattern of in situ degradability of OPF.

Materials and Methods

Treatments

Freshly pruned fronds were obtained from a single source at the Universiti Putra Malaysia (UPM) farm and transported to the feedmill of the Malaysian Agricultural Research and Development Institute (MARDI) where they were chopped to 3 to 5 cm length using an engine-driving chopper. The freshly chopped OPF was then divided into three batches. The first batch used as a control was oven-dried and stored for subsequent tests without any steam treatment (UT). The second batch of OPF was steam treated as it was (fresh) (FT) while the third batch was sun-dried to a moisture content of between 25 to 30 % before steam treatment (DT). There were three steam pressures: 10 kg cm^-2 for 20 min, low (L); 12.5 kg cm^-2 for 7 min, medium (M); and 15 kg cm^-2 for 4 min, high (H), resulting in a total of seven treatments, namely, UT, FT_l, FT_m, FT_h, DT_l, DT_m and DT_h. Steam treatments were done in MARDI using an autoclave (HITACHIZOSEN CO, LTD, Japan). The capacity of the autoclave was about 200 litres with a designed temperature of 214^o C. The OPF (about 40 kg) was put in a steel chamber into which steam under pressured was injected until the desired pressure was reached, and held for the required treatment time. The treated samples were air-dried and stored for subsequent tests.

Nylon bag incubation study

Three steers weighing 250 to 300 kg, each fitted with a permanent rumen cannula, were used. The animals were maintained on ad libitum access to grass hay and 2.5 kg/d of a (70%) maize: (30%) soybean meal supplement for 14 days before a 7-d measurement period. Approximately 5 g of each sample from each treatment, which had been ground through a hammer-mill fitted with a 2.5 mm screen, were weighed into each bag measuring 7.5 x 15 cm with pore size of 40 µm. The bags were incubated in the rumen for 8, 16, 24, 48, 72, and 96 h. At the end of each incubation time, the bags were retrieved and washed in cold water until clear rinse water was obtained. Two bags containing 5 g of each sample were soaked in warm water for 1 h and then washed to determine washing losses (Ørskov et al 1980). The residues in the bags were then dried in a forced-air oven for 48 h at 60^o C. The dry matter (DM) loss from bags incubated in the rumen was corrected for the soluble material and very fine particles in the samples, which may be washed out of the bag without being degraded. To do so, a sample bag was placed in a water bath at 37^o C for 24h. The soluble DM was expressed as the percentage solubility of the original DM in the nylon bag. The resulting DM disappearance was then considered to represent the extent of microbial degradation of fibre (Mould and Ørskov 1983, 84).

The degradability data were fitted to the exponential model of McDonald (1981) as modified by Dhanoa (1988) with a discrete lag phase:

P = a + b (1 - e^{-c x (t - L)} ) for t > L

where P is the potential degradability of DM at time t, a is the rapidly soluble fraction, b is the insoluble but fermentable fraction given sufficient time, c is the constant rate of degradation, and L is the lag time for the beginning of the degradation.

The model was fitted with a non-linear Marquardt procedure (SAS 1989) with the following bounds: a + b < 1; 0.01 < c < 0.05; 1 < L < 10.

The resulting constants were used to estimate the potential degradability (PD) (a + b), and the effective degradability (ED) following the equation of McDonald (1981):

ED = a + bc / (k + c) x exp (-(c + k) x L)

where k is the outflow rate assumed to be 2% h^-1 (ARC 1984).

Chemical analyses

The samples were milled through a 1 mm sieve and used for the determination of DM, ash and crude protein (CP), according to AOAC (1985). Fibre fractions comprising neutral detergent fibre (NDF), acid detergent fibre (ADF), and acid detergent lignin (ADL) were analysed using the methods of Van Soest et al (1991). Hemicellulose and cellulose were determined by the following formulae:

Hemicellulose = NDF - ADF;

Cellulose = ADF - ADL.

Statistical analyses

Effects of steam treatment on chemical composition (fibre constituents) and rumen microbial utilization were evaluated following a completely randomized design with three replicates per treatment for in situ data. The fibre constituents were determined in six replicates per treatment. Resulting data were analysed using the General Linear Models (GLM) of the SAS Institute Inc. (1989) to obtain analysis of variance. The untreated substrate was excluded before analyzing for the effect of moisture x pressure interaction. Where differences existed between treatment means, Student's t-test was used to separate them. Orthogonal polynomials were used to partition treatment sum of square into the linear (L) and quadratic (Q) components (Steel and Torrie 1980) of steam pressure and pressure x moisture interactions. Correlations were performed between fibre fractions and the degradability parameters using the PEARSON CORR procedure (SAS 1989).

Results and Discussion

As heating temperature in the steel chamber was fluctuating between 120 and 165^oC for pressure 10 kg cm^-2, from160 to 175^oC for pressure 12.5 kg cm^-2 and from 170 to 180^oC for pressure 15 kg cm^-2 within the required treatment time, the real heat conditions under which the samples were treated may not be actually defined.

Steam treatment significantly affected all the fibre constituents of OPF and the biological parameters (except rate of degradation and lag time, P > 0.05) (Table 1). Most of the variations among the mean concentration of fibre constituents were associated with pre-dried vs. fresh, steam treatment. The finding of increased hydrolysis of the hemicellulose fraction at low moisture content (30%) is contrary to the hypothesis of(Van Soest, 1994) of increased reactivity of hemicellulose with increasing moisture level , the greatest degree of hydrolysis being achieved when OPF was pre-dried (Table 1).

Table 1 Fibre constituents and in situ degradability (g / kg DM) of untreated and steam-treated oil palm frond
Treatment^a	Wash^b	Fibre constituents^c					In situ degradability^*
	loss	NDF	ADF	HEM	CEL	ADL	a	b	*c(h^-1)*	PD	ED	*Lt (h)*
Untreated	113	701	502	198	424	79	111	207	0.038	317	213	3.2
Fresh treated
L	208	612	563	50	393	170	314	236	0.032	550	434	3.5
M	369	608	531	77	394	136	319	239	0.029	558	445	2.8
H	397	616	549	67	406	144	284	268	0.046	552	415	5.8
Pre-dried treated
L	352	598	580	18	410	171	272	294	0.035	566	391	7.8
M	362	588	570	18	392	178	293	312	0.027	606	421	6.8
H	429	564	565	0	366	199	296	304	0.035	600	451	4.2
SEM^d		8.94	7.66	11.74	13.64	13.17	0.92	0.99	0.007	1.78	2.43	3.56
Significance^**
Ut vs. all^e		<0.01	<0.01	<0.01	0.05	<0.01	<0.01	<0.01	0.50	<0.01	<0.01	0.48
F vs. D		<0.01	<0.01	<0.01	0.46	0.01	<0.01	<0.01	0.44	<0.01	0.48	0.29
P_L		0.11	0.07	0.97	0.27	0.95	0.64	0.01	0.21	0.18	0.25	0.78
P_Lx M		0.04	0.91	0.14	0.05	0.05	<0.01	0.16	0.19	0.23	0.04	0.26
P_NL		0.96	0.04	0.17	0.99	0.23	0.02	0.96	0.06	0.20	0.49	0.81
P_NL x M		0.39	0.10	0.66	0.69	0.57	0.40	0.06	0.78	0.49	0.51	0.55
^aUntreated = no steaming; fresh or pre-dried oil palm frond steam-treated at pressure L, 10; M, 12.5; H, 15 kg cm^-2.^bWashing loss = 0 hr. incubation.^cNDF = neutral detergent fibre; ADF = acid detergent fibre; HEM = hemicellulose; CEL = cellulose; ADL = acid detergent lignin. a = intercept; b = slowly degradable fraction; c = rate of degradation; PD (a+b) = potential degradability; ED = effective degradability; Lt = lag time.^dStandard error of the mean based on three and six observations for in situ data and fibre constituents, respectively.^eFor contrast, see definition in the text. ^*Numbers are probability values.

The ADF and lignin contents showed an adverse tendency for NDF and hemicellulose. At the pre-drying conditions and the highest steam pressure, ADF content was slightly higher than NDF. Thus hemicellulose content was considered to be zero. Non-enzymatic browning reaction may have occurred, resulting in increased ADF and lignin contents (Oji and Mowat, 1978). Similar results were observed by Takigawa (1987) in softwoods and hardwoods and by Castro et al. (1993) in wheat straw.

The observed DM degradability (DMD) values are shown in Fig. 1. The DT samples had a greater rate of degradation than the FT samples at the latest incubation times. After correction for the soluble and very fine particles, DMD for DT was againhigher at late incubation times. This pattern of increased DMD may be an indication of increased susceptibility to microbial attack of remaining cell wall material of the pre-dried samples . It is also consistent with the strong negative correlations between NDF and hemicellulose content and the potential degradability (PD) of corrected DM (Table 2).

Figure 1. Observed (filled symbols) and corrected (open symbols) dry matter degradability of untreated(■ □),
fresh steamed (● ○) and pre-dried, steamed (▲ Δ) oil palm frond.

Furthermore, the positive association between the degradability of the remaining cell wall with lignin content (Table 2) would support the possibility of depolymerization and solublization of lignin (Saddler et al 1982; Brownell and Saddler 1987; Wong et al 1988).

Table 2. Correlation coefficients between fibre fractions and kinetics of digestion
	*SF^x*	b	PD	ED	c	Lt
NDF	-0.871^*	-0.828^*	-0.969^***	-0.916^**	0.327	-0.384
ADF	0.698	0.826^*	0.843^*	0.704	-0.227	0.699
ADL	0.793^*	0.823^*	0.911^**	0.835^*	-0.319	0.428
Hemicellulose	-0.838^*	-0.861^*	-0.958^***	-0.868^**	0.301	-0.528
Cellulose	0.687	0.559	-0.726	-0.769^*	0.356	0.110
^xSF, soluble fraction corresponding to the intercept (a) of the above equation; PD, potential degradability; ED, effective degradability; c, rate constant; Lt, lag time.^* P<0.05; ^ P<0.01; ^* P<0.001; test the null hypothesis that the coefficient or intercept equals zero.

The response to steam pressure treatment (Figure 2) was influenced by the moisture content of OPF at all incubation times, but not at 48 h (P>0.10). From 8 to 24 h of incubation, DMD increased with increasing steam pressure with the pre-dried treatment, but at a decreasing rate (P<0.05) with the fresh treatment. This trend was observed up to the latest periods of incubation, and the greatest response was obtained when the pre-dried samples were treated at the highest steam pressure.

The reason for this is unclear. The hydrolysis products of hemicellulose may have been further decomposed forming caramelization products, which are known to be inhibitory to the growth of many microorganisms (Nesse et al 1977). The rate of the reaction was probably faster in the fresh samples. Unfortunately it was not possible to measure the presence of these compounds in the present study. Another possible explanation may be that the complete removal of hemicellulose under the pre-drying conditions (Table 1) also led to increased accessibility to cellulose.


Figure 2a. In sacco dry matter degradability at 8h of fresh steam-treated and pre-dried steam-treated oil palm frond at steam pressures of 10, 12.5 and 15 kg cm^-2.	Figure 2b. In sacco dry matter degradability at 16h of fresh steam-treated and pre-dried steam-treated oil palm frond at steam pressures of 10, 12.5 and 15 kg cm^-2.	Figure 2c. In sacco dry matter degradability at 24h of fresh steam-treated and pre-dried steam-treated oil palm frond at steam pressures of 10, 12.5 and 15 kg cm^-2

Figure 2d. In sacco dry matter degradability at 48h of fresh steam-treated and pre-dried steam-treated oil palm frond at steam pressures of 10, 12.5 and 15 kg cm^-2	Figure 2e. In sacco dry matter degradability at 72h of fresh steam-treated and pre-dried steam-treated oil palm frond at steam pressures of 10, 12.5 and 15 kg cm^-2

The results presented in this study would suggest that the initial moisture level of the material should be taken into consideration when steam-treated crop residues are evaluated, since results of in situ degradability using treated, fresh materials may underestimate their nutritive value.

Conclusions

In conclusion, steam treatment of oil palm frond pre-dried to 25-30% moisture is capable of solubilizing 90% of the hemicellulose fraction and improving rumen microbial utilization of the remaining cell wall material. However, the release of caramelization products due to level of moisture, stage of maturity of the tree and probably the actual management practices in the oil palm plantation may be major factors affecting advantages of processing whole oil palm frond. These factors should be evaluated in further studies. The common practice, which consists of using completely dry materials before steam treatment may not be necessary. Where other types of crop residues (eg: maize stover, rice straw, etc.) are available, the material should be treated shortly after harvesting, before the soluble and potentially degradable carbohydrates have declined.

Acknowledgements

This work accounts for part of the PhD. Thesis submitted by the first author to the Universiti Putra Malaysia. We thank the Ministry of Science, Technology and Environment, Malaysia for financial support, the Malaysian Agricultural Research and Development Institute for providing access to, and assistance with the steaming machine. Thanks are also due to one anonymous referee for his helpful comments throughout this work.

References

Abu Hassan O, Ishida M, Oshio S and Ahmad T J 1996 Utilization of oil palm trunks and fronds as feed for ruminants. In: Ho Y W, Vidyadaran M K and Sánchez M D (Eds), Proceedings of the MSAP / FAO First International Symposium on the Integration of Livestock to Oil Palm Production. Kuala Lumpur. Paper 16 MSAP/FAO127.

AOAC 1985 Association of Official Analytical Chemists. 12^th Edition AOAC, Washington, D.C. 1930 p.

ARC 1984 Agricultural Research Council. The Nutrient Requirement for Ruminant Livestock. Suppl. No. 1. Commonwealth Agricultural Bureaux. Slough.

Brownell H H and Saddler J N 1987 Steam pretreatment of lignocellulosic material for enhanced enzymatic hydrolysis. Biotechnology Bioengineering. 29: 228-235.

Castro F B, Hotten P M and Ørskov E R 1993 The potential of dilute-acid hydrolysis as a treatment for improving the nutritional quality of industrial lignocellulosic by-products. Animal Feed Science and Technology 42, 39-53.

Dhanoa M S 1988 On the analysis of dracon bag data for low degradability feeds. Grass Forage Science 43, 441-444.

Goering H K, Van Soest P J and Hemken R W 1973 Relative susceptibility of forages to heat damage as affected by moisture, temperature, and pH. Journal of Dairy Science 56, 137-143.

Hart M R, Walker Jr H G, Graham R P, Hanni P J, Brown A H and Kohler G O 1981 Steam treatment of crop residues for increased ruminant digestibility. I. Effects of process parameters. Journal of Animal Science Vol. 51 No.2.

Islam M 1999 Characterisation and utilization of oil palm (Elaeis guineensis) frond as fibrous feed for ruminants. PhD. Dissertation, Universiti Putra Malaysia.

Klopfenstein T J and Bolsen K K 1971 High temperature pressure treated crop residues. Journal of Animal Science 33, 290 (abstract).

McDonald I M 1981 A revised model for the estimation of protein degradability in rumen. Journal of Agricultural Science Cambridge 96, 251-252.

Mould F L and Ørskov E R 1983 / 84 Manipulation of rumen fluid pH and its influence on cellulolysis in sacco, dry matter degradability and the rumen microflora of sheep offered either hay or concentrate. Animal Feed Science and Technology 10, 1-14.

Nesse N, Wallick J and Harper M J 1977 Pretreatment of cellulosic wastes to increase enzyme reactivity. Biotechnology Bioengineering 19, 323-336.

Oji U I and Mowat D N 1978 Nutritive value of steam treated cornstover. Canadian Journal of Animal Science 58, 177-181.

Ørskov E R, Hovell F D DeB and Mould F L 1980 The use of nylon bag technique for the evaluation of feedstuffs. Tropical Animal Production 5, 195-213.

Saddler J N, Brownell H H, Clermont L P and Levitin N 1982 Enzymatic hydrolysis of cellulose and various pretreated wood fractions. Biotechnology Bioengineering 24, 1389-1402.

SAS 1989 SAS User's Guide: Statistics. Version 6.0. SAS Institute Inc. Cary NC.

Steel R G D and Torrie J H 1980 Principles and procedures of statistics. 2nd ed. McGraw-Hill Book Co. Inc, New York, NY.

Takigawa A 1987 Feeding value of steamed wood and explosively depressurized wood. JARQ. 204, 282-291.

Van Soest P J, Robertson J B and Lewis B A 1991 Methods for dietary fibre, neutral detergent fibre, and nonstructural polysaccharides in relation to animal nutrition. Journal of Dairy Science 74, 3583-3597.

Van Soest P J 1994 Nutritional ecology of the ruminant. 2nd ed. Cornell University Press, Ithaca, NY.

Wong K K Y, Deverell K F, Mackie K L, Clark T A and Donalson L A 1988 The relationship between fibre porosity and cellulose digestibility in steam-exploded Pinus radiata. Biotechnology Bioengineering 31, 447-456.