Previous Page Table of Contents Next Page

E. Donefer


In the early 1960s, two Canadians (R. Miller and T. Tilby) invented a procedure for “sugarcane separation”, the subsequent development of which has had a marked influence on concepts of feeding the sugarcane plant to livestock, and which may ultimately modify the operation of the conventional sugar factory. Over the past 18 years, trials with cattle and sheep based on feeding cane separation products have been conducted in many countries, primarily in the Caribbean region. Claims and counterclaims have been published about the technical and economic feasibility of livestock feeds produced by the cane separation process, and to date there has not been an adaptation of the technology to commercial livestock production systems.

Is this lack of adoption due to problems of equipment design, to the nutritive value of the feeds or to a lack of economic feasibility of intensive livestock production systems in cane-growing countries of the tropics and subtropics? This paper will attempt to review the development of cane separation as related to livestock feeds, clarify some of the misconceptions which have arisen and comment as to future perspectives.


There were two basic objectives in developing the cane separation technology, which were based on observations in the Caribbean of conventional sugar factories. It appeared that a disproportionate amount of energy was required to crush the sugarcane stalks prior to sugar extraction and that also the bagasse product resulting from crushing did not provide the sugarcane fibres in a form for optimum by-product utilization. It was proposed, therefore, to develop a system which minimized sugar factory energy requirements, reducing factory requirements for burning bagasse as a fuel with the resultant increased availability of sugarcane rind fibres providing an increased potential for producing high quality board products for construction purposes. Conventional bagasse does not make a strong fibre board, because it contains a mixture of long rind fibres and short internal fibres and pith. It is common in some countries to depith the bagasse mechanically, use the long fibres for paper or board manufacture and the pith residue for livestock feed, but this is not always cost-effective as it adds an additional processing step to the conventional factory operation, and the quality of the depithed bagasse product is very variable. The new procedure would eliminate bagasse as a sugar factory by-product.

Development of the sugarcane separation technology was based on two simple physical properties of the sugarcane stalk (stem). The first is that it is easier and requires less energy to split the stalk longitudinally than to crush it. The first step in a cane separator is the passing of the stalks at high speed over a splitter blade, dividing it down the middle into two halves (Figure 2). The second step involves the derinding of the stalk or the separation of the outer rind layer from the inner sugar-containing pith. The hard properties of the rind and soft characteristic of the inner stalk make it possible to accomplish the derinding by passing the split stalk between two special rollers. These rollers have a gap between them to allow the stalk to pass through, the outside roller moving clockwise at about six times the speed of the inner roller which is moving counterclockwise. The result of the speed and rotation differences is that the inner roller holds back the stalk allowing the outer roller to scrape and separate the soft pith from the rind. The separated rind can actually be passed through a second set of similar acting rollers which will separate the outside wax layer from the rind fibres, if cane wax is desired as a by-product.

New terminology was required to describe the product of cane separation and the commercial term “Comfith” was used by the company which developed the cane separator (Canadian Cane Equipment Ltd.) to designate the inner stalk components which consisted of fibre vascular bundles and pith, the latter in association with the naturally occurring liquid sucrose. Requiring a non-commercial term in technical publications, the term sugar-fith was introduced, the fith thus referring to the two quite different types of fibre components (fibre vascular bundles and pith) (Donefer et al., 1975). Thus the “pith” referred to in the title of this paper as the product of cane separation actually contains the sugars in addition to fibrous products, and herein will be referred to as sugar-fith. With the sugar removed the resultant product is termed fith (or commonly, pith). There has been no consistent use of cane feed terminology in the literature resulting sometimes in confusion as to the actual nature and composition of the feed being described. Figure 1 compares common and international names and numbers of sugarcane-derived feeds (Donefer and Latrille, 1980).

It is important to emphasize that the development of the separation process was designed for application to sugar factory operations and that possible uses for animal feed was a later development. The sugar-fith (or Comfith) produced in the modified sugar factory would thus be subject to sugar extraction procedures, so that there are four resulting products of the cane separation technology: rind, fith, sugar and molasses (a fifth being wax if desired). From an animal feed perspective, two feeds are possible, fith with sugar (sugar fith) or without sugar.


Essentially all experimental work has been based on sugar-fith, with its promise of a highly digestible energy content combined with the high yield potential of the sugarcane plant.

In 1964, Canadian Cane Equipment Ltd. established a pilot plant on the Island of St. Kitts in the Caribbean in association with the local sugar factory, where a factory-scale separator was constructed. The company became interested in identifying all potential commercial products, so, in conjunction with the Department of Animal Science of McGill University, preliminary feeding trials, limited to a total of 12 cattle and covering a 90-day period, were conducted in St. kitts in 1968. Daily weight gains averaging 0.7 kg were obtained for a group of cattle fed supplemented sugar-fith, and in addition, digestibility trials were conducted with sheep fed rations of sugar-fith or fith (MSc. thesis, James, 1969).

The encouraging results from the preliminary observations in St. Kitts led to an extended experimental programme conducted in Barbados in conjunction with the Animal Nutrition Laboratory of the Barbados Ministry of Agriculture. The “Barbados Comfith Project”, initiated in late 1969 and continuing for a 4-year period, was funded by the Canadian International Development Agency (CIDA) in conjunction with the Barbados Government. This project involved the feeding of a ration predominating in sugar-fith, with cattle growth trials conducted primarily for beef production but with parallel studies with lactating cattle, sheep and swine (Donefer et al., 1975).

A major accomplishment of the Barbados project was the demonstration that high levels of animal production (cattle growth averaging 0.9 kg/day) could result from a ration containing a major proportion of sugar-fith and cane tops (together up to 80 percent of ration dry matter).

An event which had an important influence on future canefeed developments was a CIDA-sponsored seminar held in Barbados in January 1973 for representatives of Commonwealth Caribbean Countries, the major purpose of which was to present the initial results from the Barbados Comfith Project. Dr. T.R. Preston was an invited speaker at the seminar and his interest in the new separation technology was initiated. In addition to reporting the sugar-fith results presented at the seminar in various publications, Dr. Preston was able to obtain prototype cane separators for feeding trials conducted under his supervision in Mexico (Chetumal) and later in Mauritius and the Dominican Republic.

Although there have been widespread traditional practices of feeding unsupplemented sugarcane to livestock particularly under emergency conditions (i.e. drought), this usually resulted in survival or maintenance conditions for the animals. Prior to the 1970s there have also been sporadic research reports from Brazil, the USA and Mexico of feeding trials where sugarcane was a ration ingredient.

The Barbados Comfith Project thus did more than present preliminary livestock production results based on the cane separation technology; it restimulated interest in the tropical world as to the potential for using the whole sugarcane plant (as an alternative to sugar production) as a feed for intensive livestock production. The important principle demonstrated was the necessity of supplementing the energy-rich, sugarcane feed with sources of protein, minerals and vitamins to meet nutritional requirements for increased levels of production.

The next developments in the use of derinded cane (sugar-fith) were related to results of trials conducted in Chetumal, Mexico by Preston et al. (1976). Using a cane separator obtained from Barbados, their cattle rations compared two types of processed cane, sugar-fith and chopped cane, each at increasing levels of supplementation with rice polishings. They concluded that the results of the 400-head feeding trial showed no significant difference in animal performance between derinded and chopped cane. In a later report, Preston and Leng (1978) stated that “following the results of the Chetumal trials the much heralded, sophisticated and expensive ‘separation’ technology was discarded in favour of simple chopping with machines costing less than one twentieth of the price”.

The company which originally developed the separation technology, Canadian Cane Equipment Ltd., went bankrupt in mid-1974, and their patents on cane separation reverted to the Canadian Government, who had provided the company substantial development grants. Several companies have since been licensed by the Canadian Government to pursue commercial development of the equipment, but to date a sustained strategy has not been obvious.

It is with much interest and speculation on the part of those of us who have been involved in the technical research and development of cane separation technology as to why there has not been more success in its commercial adoption, particularly since many of us continue to believe that it represents a revolutionary process which could have a profound positive effect on both the sugar industry and livestock production in the humid tropics. I would suggest that the companies involved appeared to be more interested in short term profits and quick recovery of development costs rather than in the logical technical and commercial development. All cane separators manufactured have been prototypes with minimal apparent attempt to incorporate engineering principles necessary for sustained use of equipment at factory or farm level. Separator prototypes should not have been sold to far-reaching project sites (Mexico, Mauritius) without guarantees by the company for maintaining the equipment in proper condition, for there is evidence that the separators used in Mexico and Mauritius were not working properly and therefore may not have been producing a high quality and uniform feedstuff. Also, the cost charged for the prototype separators had little relationship to what the price of machines factory-produced in large quantities would have been. In addition, the equipment developers, Canadian Cane Equipment Ltd., as executing agency for the Canadian International Development Agency for the Barbados Comfith Project should have been more closely monitored by the Canadian Government agency to guarantee that commercial equipment development procedures were being carried out by the company.

An additional factor discouraging separator development was the very high prices being received for sugar, particularly in the mid-1970s. The Caribbean sugar industry was little interested in innovations which might have affected its current sales and was particularly negative about diverting any part of the sugar crop to livestock feed production. It is thus ironic that the very low sugar prices experienced in the 1980s have caused sugar factories in several countries to close and others to attempt to determine if crop and product diversification might allow profitable utilization of valuable agricultural resources.


Although the separator-produced sugar-fith contained only the internal stalk fibres and sugar, initial feeding trials in Barbados demonstrated the advantage of adding sugarcane tops to the ration due to increased voluntary intake (Donefer et al., 1975). Therefore, when comparisons were made between sugar-fith plus cane tops and chopped whole cane, the only apparent difference between the two feeds would be the rind layer removed by the separator in the case of the former. The Chetumal results (Preston et al., 1976) indicated “no significant difference in animal performance between derinded and chopped whole cane” and in a comparision of studies conducted in many countries in the Caribbean region the average cattle liveweight gains (g/day) were 755 and 736 for derinded and whole cane respectively (Preston and Leng, 1978). Perhaps more important than this slight difference was the large variability observed in the different studies, ranging from a -28 to +17 percent change in liveweight gain due to derinding. One explanation for this variation might be related to the large differences in the composition of sugarcane known to be related to season of harvest and harvesting procedure (particularly if the cane field was burnt prior to harvest to reduce the trash content).

To contrast possible options, the cane used in the Barbados Comfith project might be considered in optimum condition as it was from unburned fields but with the stalks “hand cleaned” prior to separation. A 12-month continuous sheep digestibility trial (Donefer and James, 1978) indicated a mean dry matter digestibility (%) of 69.7 ± 2.1 (SD) and mean voluntary intake (expressed as a percent of 3 kg DM intake/100 kg body weight) of 77.3 ± 8.1 (SD). The lowest values were found in the “wet season” when cane is known to be at its lowest sucrose content.

Comparison between sugar-fith and chopped cane conducted at the Sugarcane Feeds Centre (SFC)1 in Trinidad is particularly revealing as to the effect of cane variables on animal performance results.

Cattle feeding trials at the SFC have been sustained over many years involving a 500-head capacity feedlot with the chopped whole sugarcane plant as the primary ration ingredient. Table 1 summarizes cattle growth responses to whole cane harvested during the 1981 dry season in Trinidad, with an average daily gain of 0.72 kg being observed for 197 steers (Donefer et al., 1982, 1983). Immediately following the dry season, a group of 120 of the steers were fed rations comparing whole and derinded cane, with a marked growth depression (p<.001) observed for the whole cane-fed cattle (0.46 kg/day) but with no change for the cattle fed derinded cane (0.73 kg/day).

Table 2 summarizes the chemical composition of the cane and the most obvious difference between the dry and wet season harvested whole cane is the 3-fold increase in ash content, due to soil contamination under wet season harvesting conditions in Trinidad. In this climatic situation the advantage of the sugar-fith is the fact that in its preparation soil contaminated leaves (trash) have been removed prior to derinding, whereas this is not the case for the whole cane which is chopped in the condition in which it is harvested. Since climatic conditions in relation to dry season (length and extent) vary within countries and considerably between the sugarcane-growing tropical countries, these data illustrate the large effect that this variation might have on livestock feed quality and thus the applicability of results from one area to another. It also emphasizes that the results of feeding trials conducted with minimal number of animals, for short feeding periods and only covering one season, must be regarded as preliminary, to be confirmed in extended studies at experimental as well as commercial prototype levels.

1The Sugarcane Feeds Centre was initiated in 1976 as a joint project of the Government of Canada and the Government of Trinidad and Tobago, with McGill University acting as executing agency for the period 1976–81.


Until there are commercially available cane separators, it is not possible to comment on the actual cost of production of livestock feeds based on derinded cane. The success of a system based on feeding a product such as sugar-fith is also very much dependent on the utilization of the sugarcane ring for hard board production purposes; it is indeed suggested that the construction material potential of sugarcane is more valuable than the price received for its sugar content. Separators would always be more costly than choppers, but the animal feed potential of the process must be considered as a by-product of board manufacture.

At present, there is work in Jamaica on the development of an integrated sugarcane system based on sugarcane separation technology (Holmes, 1983). Under Dr. I. Sangster, Director of the Jamaican Sugar Industry Research Institute (SIRI) - Factory Technology Division, a pilot plant is in operation where a cane separator is used to produce the three major separation products, sugar (or sugar syrup), board, and livestock feed. In this system, the sugar is removed from the fith so that the sugar-free fith is primarily a low energy forage very similar in its chemical and feeding properties to conventional bagasse pith or bagacillo. Effective use of this product, not the topic of this paper, requires proper energy and other nutrient supplementation and possible use of chemical or physical treatment procedures to distrupt the ligno-cellulose structure and thus increase energy availability to ruminant species.


Donefer, 1978 E. and James, L.A. Digestibility and intake of derinded sugarcane (sugar-fith) based rations by Barbados Black Belly Sheep. Report of Regional Livestock Meeting, Barbados.

Donefer, E. and Latrille, L. 1980 Description of sugarcane feeds: Nomenclature and nutritional information. Standardization of analytical methodology for feeds. Proceedings of a Workshop held in Ottawa, Ontario. March 12–14, 1979. IDRC-134e International Development Research Centre, Ottawa, Ont.

Donefer, 1975 E., James, L.A. and Laurie, C.K. Use of a sugarcane-derived feedstuff for livestock. Proc. III World Conf. Animal Production, Melbourne, Australia, 1973, pp. 563–566, Sydney Univ. Press.

Donefer, 1982 E., Charles, K. and Neckles, F.A. Effect of season on the growth of cattle fed sugarcane-based ration. Journal of Animal Science 55 (Supplement 1) 182 (Abstract).

Donefer, E., Brunton, P.D. and Neckles, F.A. 1983 Intensive beef production system using sugarcane-based ration. Proceedings of the Fifth World Conference on Animal Production. Vol. II, 845–846, Tokyo, Japan.

Holmes, 1983 J. Cane separation technology. A potential for progress in the sugarcane industry. Sugar Techy News on Sugar Technology. Vol. 5, No. 2.

Preston, 1978 T.R. and Leng, R.A. Sugarcane as cattle feed. Part I: Nutritional constraints and perspectives. World Anim. Rev., 27: 7–12.

Preston, 1976 T.R., Carcano, C., Alvarez, F.J. and Gutierrez, D.G. Rice polishings as a supplement in a sugarcane diet. Effect of level of rice polishings and of processing the sugarcane by derinding or chopping. Tropical Animal Production, 11:150–162.

Figure 1: International and common names for sugarcane-derived feeds

Figure 1

Figure 2: The cane separation process (Holmes, 1983)

Figure 2
Table 1. Sugarcane Feeds Centre: cattle feeding trial (1981) - Summary of results
Steers, no.1978040
Weight, initial, kg201243246
final, kg240282307
ADG, kg0.720.460.73
DM intake, % LW   
Cane1.55 (61)1.60 (60)1.89 (64)
Supplement1.00 ( 39)1.07 ( 40)1.08 ( 36)
Total2.55 (100)2.67 (100)2.97 (100)

Table 2. Sugarcane Feeds Centre: summary of chemical analysis of cane, % DM
Crude protein1.93.22.3
  15.0 (last month) 

E. Donefer

El equipo técnico de separación de la caña rompe el tallo y separa la corteza exterior de la médula contenedora del azúcar y los hacecillos vasculares de fibra. Se desarrolló este proceso a fin de reducir las necesidades de energía de las fábricas de azúcar eliminando el desmenuzamiento y a fin de utilizar las fibras de la corteza en la fabricación de tableros de partículas muy resistentes. Ensayos realizados en San Cristóbal y Barbados demostraron que el componente de médula contenedora de azúcar podía representar una parte importante de la ración de un rumiante, lográndose un aumento de peso de 900 g/día como promedio. Ensayos ulteriores en muchos países en los que se compararon los resultados de la alimentación con caña descortezada y con caña integral picada indicaron sólo pequeñas diferencias en cuanto al aumento de peso del ganado, por lo cual se prefirió el equipo picador menos costoso. Estudios realizados en Trinidad (Centro de piensos a base de caña de azúcar) dieron resultados mucho mejores para la caña descortezada que para la caña picada (730 frente a 460 g/día) debido a la contaminación del suelo que limitaba la utilización de la caña integral en la estación de las lluvias.

La tecnología de separación de la caña sirvió para estimular el interés mundial en la utilización de este cultivo en la alimentación del ganado y no en la producción industrial de azúcar.

Las perspectivas en cuanto a la adopción de la technología de separación de la caña dependen de la disponibilidad en el mercado de equipo de separación eficiente y parece viable cuando los tableros de corteza son el principal producto y los piensos un subproducto poco costoso. La separación industrial producirá corteza y azúcar, y la médula a la que se ha extraído el azúcar podrá utilizarse como forraje de baja calidad (de composición análoga al bagazo o bagacillo) en la alimentación de los rumiantes, sea en forma no tratada o tratada (con álcalis o al vapor).

Previous Page Top of Page Next Page