2. The Landless Livestock Ruminant Production Systems

2.1. Definition of LLR systems
2.2. Description of the subsystems.
2.3. Statistics on LLR systems and production trends.
2.4. Causes and motives

2.1. Definition of LLR systems

Sere and Steinfeld (1995) have defined the world livestock production systems. The boundaries of the systems are formulated on:

(1) the source of dry matter fed;

(2) percentages of total value of output (e.g. proportion of output from ruminants, monogastrics, crops etc); and

(3) climatic criteria (or agro-ecological zone).

The relation between LLR systems and other livestock production systems is presented in Figure 1. The feed is mainly introduced from outside the farm system, thus separating decisions on feed use from those on feed production, and particularly decisions on the use of manure on fields to produce feed and cash crops. These systems are very open systems in terms of nutrient flow; the nutrients cannot be circulated within the farm system. Figure 2 presents the inputs and outputs of LLR systems. They share this feature with the landless monogastric livestock systems (LLM), the main difference is that ruminants need a fibrous ration whereas in monogastrics the feed conversion of concentrates to live weight gains is substantially more efficient (particularly in intensive chicken production).

Other farming systems			Farming systems
Mixed farming systems			Livestock systems
Grassland-based farming systems			Solely livestock systems
Landless monogastric systems			Landless livestock systems
Landless ruminant systems
	Feedlot beef	Veal	Intensive mutton	Urban dairies
Inputs:	rangelands beef herds	dairy herds	land-based rural/pastoral systems	rural mixed farms

Thus LLM systems are only competitive in situations where cheap concentrates are available and consumers are prepared to pay substantially more for quality beef than for chicken or pork. LLR systems are concentrated in a few regions of the World. Cattle LLR systems are found in a few OECD countries, Eastern Europe, CIS and near the main cities in developingcountries, while sheep LLR systems are found in West Asia and North Africa (WANA).

Livestock resources: LLR systems are part of a stratified livestock production system. Livestock in LLR systems originate from other (land-based) systems, e.g. offspring from range fed beef cattle are finished in intensive feedlots; lambs from the pastoral areas are fattened intensively; male calves from dairy breeds are fattened on milk; and high yielding dairy cows from the rural areas are sold for a final lactation in urban dairies. LLR systems are largely interlinked with other livestock and mixed production systems (see Figure 1).

Feed resources: Feeding is mainly based on good quality high energy feeds with a maximum intake of concentrates and a minimum intake of roughage. Although grains and protein supplements are the main sources of concentrate feed, depending on the system and the region, by-products from the agro-processing industry could be important. Minimum roughage requirements are around 15-30% of the ration (except for veal production).

Figure 2: Scheme of inputs and outputs in LLR systems.

INPUTS	OUTPUTS
FEED RESOURCES:	PRODUCTS:
(roughage, concentrates, byproducts)	- meat
- N, P and other minerals	- hides
- additives/vitamins	- slaughter wastes
- heavy metals	- manure
PRIMARY PRODUCTION LLR SYSTEMS
ANIMAL RESOURCES:	EMISSIONS:
animals of various age
and weight	* to the soil:
from land-based systems	- N, P, other minerals
	* to surface water:
DRUGS	- OM, N, P, other minerals
	* to the atmosphere:
FOSSIL ENERGY	- local importance: NH3
	- global importance: CO2, NOx, CH4

2.2. Description of the subsystems.

2.2.1. Beef fattening in feedlots
2.2.2. Veal production
2.2.3. Sheep fattening
2.2.4. Large-scale beef production in EE and CIS.
2.2.5. Urban dairy farming

The following subsystems and their geographical distribution are recognized within LLR systems:

(1) Feedlot fattening and finishing of beef cattle. This system is mainly concentrated in the USA and particularly in the states of Texas, Nebraska, Kansas, Colorado and California;

(2) Veal production. This type of production is mainly found in the EU and concentrated in the Netherlands, France and Italy;

(3) Intensive finishing of sheep, mainly in the Middle East, e.g. in Syria and Iran;

(4) Large-scale beef (and dairy) production in EE and CIS;

(5) Urban milk production in developing countries, e.g. cities on Indian subcontinent.

2.2.1. Beef fattening in feedlots

(mainly based on Perry, 1992).

Feedlot fattening of beef cattle is mainly done in the USA, though this practice of upgrading beef quality also exists in the EU. Furthermore, in few Latin American and African countries some finishing of cattle in feedlots is apparent, mainly to benefit from compensatory growth effects (Tacher and Jahnke, 1992; Carrillo and Schiersmann, 1992; Sosa, 1995). However, in these last-mentioned regions the fattening of beef cattle is nearly always an economic activity next to arable farming or land-based livestock production.

In the USA, the beef industry is an important segment of livestock production. The increasing population and the rising consumer buying power have together contributed to an increase in demand resulting in relatively favourable prices for beef. Consumption of beef per capita increased from 29 kg in 1946 to 50 - 56 kg in 1975 - 1980 and stabilized at around 50 kg in 1990. The practice of grain feeding of cattle has increased rapidly and over 90 % of the steers and heifers slaughtered are marketed by feedlots. Primarily as a result of the rapid increase in the demand for fed beef and increased financing from sources outside agriculture (including agri-business firms), fewer but larger feedlots have evolved with a marked geographic concentration.

Beef production is very sensitive to changes in profitability induced by changes in beef prices and consumption, cattle prices and feed cost. As most steers and heifers are slaughtered at least 3 years after their conception, a cyclical pattern of beef production ('beef cycles') results.

Generally two types of feedlot enterprises can be recognized:

(1) Farmer-operated feedlots: generally with a one-time capacity of less than 1000 head of cattle. Typically the feedlot farmers with under 1000 cattle are also involved in other farming enterprises, especially pig production and beef cows. Feedlots with a capacity of less than 200 head are often only in operation during the non-cropping season when off-season labour is available (November to May). This type of farmer-operated feedlots can be found in almost all cattle feeding regions of the USA especially in the central part of the country (Corn Belt).

(2) Large-scale feedlots (commercial feedlots): feedlots with a capacity of more than 1000 head which are in operation all year. The importance of this type of feedlot is ever increasing; in 1980 ca. 75% of the marketed feed cattle passed through this type of feedlot, compared to ca. 84% in 1990 (Perry, 1992; Fedkiw, 1992). Ownership ranges from sole proprietorships to corporate farms, including cooperatives. Most feedlots of this category purchase all or most of their feeder cattle, feed and other inputs. Feedstuffs are premixed and delivered to the cattle feed troughs by self-unloading trucks or other power equipment.

Sources of cattle. Feeder cattle are produced in almost all regions of the USA and there is a considerable inter-regional movement of feeder cattle from the place of origin to the place where they will be fattened. Some rather definite flow patterns of feeder cattle have been established, but many travel a rather circuitous route from birth to final destination.

Around two-thirds of the cattle fattened are steers, the remainder being heifers and less than 1 % are cows. This ratio changes depending on 'the beef cycle'; a situation of an expanding or a shrinking beef breeding herd. Approximately 75 % of the cattle are English breeds (Hereford, Aberdeen Angus and Shorthorn) or crossbreeds; around 10 % are dairy breeds and crosses. The initial weight depends, among other things, on the price of feed grain; low prices make it attractive to start with lighter weight cattle for a longer grain feeding period.

Basically 2 systems of beef operations can be recognized (Keener and Roller, 1994):

(1) Cow calf operation.

The calf is placed in the feedlot when about 180 kg and less than 200 days old. Feeder is fed to 475 kg during 280 days.

(2) Cow calf range operation.

The calf is weaned and put out on range at an approximate weight of 180 kg; age less than 200 days. Feeder (or stocker) is left on range till it reaches a weight of 385 kg at around 600 days. Feeder is finished in feedlot to 475 kg in around 120 days (720 days of age).

- maize and maize silage with soya bean meal and urea (Corn Belt and Lake States);
- barley, maize silage, by-products feed (e.g. apple pomace, potato waste, sugarbeet by-products); and
- maize, sorghum grain, alfalfa, straw, cottonseed hulls and molasses.

Some hormone-like growth stimulators, antibiotic feed additives and ionophores (rumen altering factors) are legalized to be included in the feed and are very commonly used. Average production parameters and ranges are given in Table 1.

Production parameter Reasonable	Range	average
Starting weight (kg)	260 - 400	340
daily gain (kg)	0.75 - 1.3	0.9
feed conversion (kg growth per kg feed DM)	6 - 11	8
fattening period (days)	90 - 280	140
final weight (kg)	410 - 550	475
vacant days (d/place/year)	10 - 100

Housing: Beef cattle in feedlots do not require shelter for protection from cold weather. However, in the Southern part of the USA more efficient growth rates will be achieved if shade is provided. There is a general trend to keep the cattle in semi-confinement in order to control waste production. Where cattle are confined and housed over slatted floors, all faeces and urine can be collected, thus eliminating the need for using bedding material. At present most of the cattle in feedlots are still kept on concrete floors, or in dry regions, on an unpaved area. Solids from manure are either collected daily and stored, or allowed to dry in the feedlot and removed periodically before spreading on fields. Collection of urine is limited to feedlots with a slatted floor.

2.2.2. Veal production

(mainly based on Toullec, 1992).

Veal production is mainly prevalent in the EU, where it was highly stimulated by subsidized milk powder used for calf rearing (as one of the ways to use the huge milk surpluses in the EU). Veal calves are exclusively given milk or milk substitutes in order to keep them at the pre-ruminant stage, thus avoiding the development of the forestomachs. Slaughtered at 2 - 5 months of age, at a live weight varying from 100 - 250 kg, they must grow rapidly (over 1 kg per day) and provide a high dressing percentage (about 60 %) a well-conformed and sufficiently fat-covered carcass with pale pinkish meat. This last-mentioned characteristic is very important from a commercial point of view (high premiums are payed by consumers for this type of meat) and can only be obtained from pre-ruminant animals on an iron deficient diet.

Two main types of veal production can be distinguished:

(1) Nursed veal calves.

This system is limited to two areas in France and include less than 12 % of veal calves slaughtered in France. The calves are mainly offspring of French beef breeds and crosses with dairy cattle. Almost all calves are born on the farm where they are reared, but sometimes a few additional calves are bought. The price of this top quality product is high (around 18 % higher than calves reared on milk substitutes). This system is decreasing because it gives a lower income than suckling calves at pasture which are sold at weaning for further fattening and red meat production.

(2) Veal calves reared on milk substitutes.

The production of milk substitute-fed veal calves is mainly localized in Western and Southwestern France, Northern Italy and Central Netherlands. Friesians with an increasing proportion of Holstein blood predominate. Most of the calves are born outside the farms where they are reared. They are usually bought from various sources at 1-3 weeks of age, Italy imports calves for rearing from The Netherlands and France. A significant number is imported into mainland Europe from the UK. Over 85 % of the production is organized under contract by dairy and non-dairy companies manufacturing milk substitutes as well as slaughterhouse companies.

Sources of calves. Male and surplus female calves are from dairy breeds, and calves from dairy cows inseminated with semen from beef breeds.

Feeding and growth. Milk substitutes fed to veal calves mainly contains milk ingredients (milk and whey powder; 70-90% of total). Starch derivates, fat substitutes (e.g. tallow, lard and saturated vegetable oils), minerals and vitamins are added in various proportions. Growth rates of an average 1.2 kg per day are common (0.7 - 0.8 kg during the first month increasing to 1.4 - 1.6 during the last part of the fattening period) with an average daily consumption of 1.8 - 2.0 kg milk substitute powder per day.

The milk substitute ration is highly digestible and the transformation in edible products is very efficient, however, the milk substitute is an expensive feed. The development of the production of milk substitutes has resulted from the growing surpluses of skim milk and whey powder in the EU Milk substitutes play an important role in the regularization of the dairy market.

The use of anabolic agents to improve nitrogen retention, live weight gain (LWG) and feed conversion rate (FCR) are strictly not permitted in the EU but in reality are used frequently.

Average production performance of veal calves in the main EU countries is in table 2.

Table 2 . Production performance of veal calves in the EU in 19872 . Production performance of veal calves in the EU in 1987 (Toulec, 1992).

	Fattening	LWG	FCR	Carcass
Country	period (days)	(kg/day)	(kg DM/kg)	weight (kg)
France	120	1.2	1.45	113
Netherlands	180	1.1	1.60 1	144
Italy	150	1.2	1.50	130

¹: FCR is estimated at 1.85 in 1992 (WUMM, 1994), probably, as a result of higher final weight.

2.2.3. Sheep fattening

(mainly based on Qureshi, 1987).

Sheep production in WANA is mainly based on the traditional pastoral system making use of the vast desert ranges and natural pastures. Following migratory or semi-sedentary grazing practices, the pastoralists and their livestock trek through the harsh terrain to produce milk and meat from resources that would not have been used otherwise. Milk production is important in traditional pastoral sheep production. With the rising demand for sheep and goat meat the pressure on the rangeland has increased with increased risks of overgrazing. Purchase of lambs for fattening on the available feedstuffs on a farm has been a common practice in the region (e.g. Maarse and Idris, 1988). Large-scale feedlot enterprises based on purchased sheep and purchased feed have been stimulated by rising meat prices and/or by government development programmes (e.g. in Syria). In Iran, for example, intensive sheep fattening has increased because of: (1) the low productivity of the land-based system; and (2) the provision of subsidized barley to increase the level of self-sufficiency in mutton. Prevention of further degradation of rangelands is often the justifiable objective of the barley subsidy.

Galal (1986) reports that in Egypt fattening operations become particularly active 2 to 4 months before Eid Al-Adha. Rams weighing 20 to 40 kg are fattened to over 50 kg. In each feedlot, sheep numbers range from as few as 2 to over 100. Sheep are sold at local markets, to butchers or to urban centres for religious festivals.

There are clear peaks in demand for mutton during the muslim religious festivals, and as the dates of the festivals change over the years, so do the fattening periods.

Sources of animals. Male lambs, surplus female lambs but also older (cull) animals are purchased from pastoralist herds and small farmers for further intensive fattening over a period of 2 to 4 months.

Feeding and growth. Feeding is based on feed grain (often subsidized), supplemented with by-products and cut forage or straw. The proportion of roughage although often not well described, is known to be generally very low (10-30%). Little data is available on the performance of sheep in feedlots. The data presented in Table 3 refer to trials at research stations, which may be different from practical situations.

Production parameter	Range	Reasonable average
Age at beginning (days)	45 - 90	60
Weight at beginning (kg)	15 - 30	21
Fattening period (days)	56 - 130	100
Daily gain (kg)	0.12 - 0.35	0.21
Feed conversion	3.1 - 8.7	5
Final weight (kg)	38 - 54	42
Dressing %	42 - 50	47

Sources: Galal, 1986; Harb, 1986; Al-Haboby and Ali, 1994; Elicin and Ertugrul, 1994; Economides, 1994; Özcan et al, 1994.

2.2.4. Large-scale beef production in EE and CIS.

Information is mainly based on de Haan et al. (1992) and Mudahar et al. (1992).

Large-scale beef (and dairy) production used to be entirely in the hands of the state and cooperative sector. In Eastern Europe the average size of state farms varies between 500 and 7000 ha and in the cooperative sector between 100 and 4500 ha. Herd size varied from around 200 head in former Yugoslavia to over 1000 head in Rumania. In the CIS, the state and collective sector with over 52,000 farms owned almost 80 % of the cattle population. The average collective farm size in the Russian Federation is around 6,600 ha (of which around 4,000 ha is crop land) and 1,900 head of cattle. State farms are around 9,000 ha (50 % planted) and around 2,100 head of cattle. More than 90 % of state and collective farms raise cattle; most dairy cattle are not grazed or only grazed during the short growing season. Little use is made of existing topographical potential for the stratification of the production system (e.g. raising young dairy stock in the hills and mountains). Fattening is mostly done on the same farm where young stock is also reared.

Feeding efficiency is low; livestock productivity is only about 50 - 60 % of Western European levels. This is partly due to low quality of the roughage, imbalanced feeding by over-using grain, and cereal by-products (ca. 40-50% of the ration), resulting in protein shortages. In the CIS for example, only 13 % of the feed requirements are met from pastures and a further 21 % from roughage (hay, silage and straw). Considering the region with such vast areas suitable for pasture, this is an extremely low percentage..

On the other hand, inefficiencies of livestock production in EE and the CIS are sometimes exaggerated. Mudahar et al. (1992) for instance, states a feed conversion ratio (in oat units) of 12.1 compared to an FCR of 5.2 in Germany. However, as theoretical requirements per kilogram LWG are between 6.6 and 9.1 oat units for animals of 200 and 400 kg LW respectively, it would seem that the values of Germany refer to kilograms starch equivalents (which is roughly equal to 0.6 oat units!).

There is no evidence that there is a significant number of livestock farms without land; only few specialized landless beef fattening farms near big cities have been installed (Dmitriev, 1991), but most feedlots are part of farms which also have land (Dohy and Bodó, 1992). With the exception of these few specialized beef farms, sufficient land for manure application is most probably available, but due to lack of economic incentives (low fertilizer prices) as well as large farm size, the distribution of the manure on the land is a problem. Probably the worst legacy of past collective policies is the massive environmental contamination caused by unsound farming practices. The large (mega) livestock farms are sources of soil, water and air pollution. This pollution is exacerbated by imbalanced feeding resulting in a high nutrient excretion per kilogram production. Collectivization and mechanization has focused much attention on the use of grain and neglect of fodder crop and grassland production. However, ruminant livestock density in EE and the CIS hardly ever exceeds 10 LU per ha and a sizeable proportion of farm income comes from non-livestock activities. Strictly speaking, this type of production does not come under the LLR production system, and will therefore get minimal attention in this report.

In recent years, the livestock sector in EE and the CIS has contracted (e.g. by ca. 20 % in EE) following removal of subsidies, increased prices of inputs and a reduced purchasing power of the population. Emphasis in livestock production is shifting from the large collective units to smaller units in the private sector. In the future more emphasis is necessary on fodder and pasture-based production and more balanced supplementary concentrate feeding. Grazing and a more even manure distribution over the available land will reduce environmental problems.

2.2.5. Urban dairy farming

Though (peri-)urban milk production is prevalent in nearly all developing countries, it is mainly important in South Asia particularly in India and Pakistan. An attractive market within the large urban population for such a perishable product like milk is one of the reasons that an urban production system has developed: it considerably shortens the distance between the milking animal and the consumer, thereby reducing the risk of spoilage of the milk and the marketing cost (Maki-Hokkonen, 1994). Another major reason for the development of this system is the relatively low price of high quality by-products, compared to the price of roughage and milk (Schiere and Nell, 1993).

Animals (mainly buffaloes) fresh in milk or just before calving are purchased from the rural areas and transported to the cities where they are kept under zero-grazing conditions. All feed is purchased, and at the end of lactation they are generally slaughtered. Unit size varies from around 10 to 100 lactating animals. Originally the units were spread all over the city, but governments have attempted (with fluctuating success) to concentrate these units in so-called 'colonies' in the peri-urban areas, mainly for sanitary reasons. Examples of this are the Landhi Colony in Karachi and the Aarey Milk Colony of Bombay. However, the ever expanding cities have again slowly absorbed such colonies into the urban area, and the inevitable growth of the cattle/buffalo population has resulted in overcrowding and is an environmental and sanitary hazard. Sometimes the peri-urban vegetable growers form a ready market for the solid manure, sometimes disposal of manure is a problem. Urine and waste water is disposed of through the public sewage system or simply seeps into the ground water or the surface water. Van de Berg (1990) believes that the colonies rarely helped to solve the problem of the urban dairies. Many animals remained in the cities, legal measures were not fully implemented and the hygiene and manure disposal problems continued in the new colonies with high concentrations of livestock.

2.3. Statistics on LLR systems and production trends.

2.3.1. Beef fattening in feedlots in the USA.
2.3.2. Veal production
2.3.3. Sheep fattening in WANA.
2.3.4. Large-scale beef production in EE and CIS
2.3.5. Urban dairy farming

The statistical base for LLR systems is very weak. Total livestock population and production data are available per country, but separate data on the different production systems is not distinguishable. An additional problem is that the animals in LLR systems are bred and reared in other livestock systems so that only part of the total production of those animals can be attributed to LLR systems. The turnover in some subsystems within LLR systems is high (e.g. in feedlots, fattening cycles of around 140 days) and, thus, annual population data do not adequately reflect total production estimates.

2.3.1. Beef fattening in feedlots in the USA.

Apart from the 'beef cycles', production and consumption has remained rather constant since the mid-seventies. This plateauing of beef consumption is owing to: (1) income stagnation and population growth; (2) reduction of the family income spent on meat; (3) increased resistance to eating beef for health reasons; and (4) competition from cheaper poultry and pork. Poultry and pork production are expected to increase at sustained rate. These products will benefit most from the downward trend in cereal prices because they convert cereal into meat more efficiently than do ruminants (adapted from Jarrige and Auriol, 1992).

In Canada and the USA the continuing shift towards specialized beef herds and an increased number of grain-fed animals, resulted in 1994 in a higher average carcass weight and higher total production (FAO-Food Outlook, 1995).

Current estimates are that ca. 26 - 27 million head of cattle are fattened annually in the USA in approximately 10 million feedlot places. Fox (1994) assumes a production of 235 kg carcass per feedlot place. Another approach is ca. 1 Kg LWG per feedlot place per day or around 355 Kg per year equivalent to ca. 215 Kg carcass weight. Therefore, calculations are based on 10 million places with a production of 3.55 million tons LW¹. Around 84 % of the animals are fattened in large-scale landless feedlots. Around 75 % of the production is realized in the states Texas, Nebraska, Kansas, Colorado and California.

¹Following Sere and Steinfeld (1995) approximately 1.66 million mt of beef is produced in these 10 million feedlot places, or 166 kg per place per annum. This is an unlikely low level of production, possibly did they confuse edible beef production (retail weight) with carcass weight.

2.3.2. Veal production

Sere and Steinfeld (1995) estimated the veal production in OECD countries at 781,000 ton, representing less than 5.1 million heads slaughtered. This is lower than the number of calves slaughtered in the EU (Table 4), while also in the USA and Canada some 1.1 million veal calves are slaughtered (Groeneveld, 1991). On the other hand, other types of calf production (e.g. nursed calves) are included in the EU statistics. Calculations are, therefore, based on 6 million veal calves annually.

Table 4. Slaughter of calves in the EU 1992 - 19934. Slaughter of calves in the EU 1992 - 1993 (EUROSTAT, 1994).

Country	Calves				Carcass
	Slaughtered (000 heads)		Weight (000 tons)		Carcass weight kg per animal
	'92	'93	'92	'93	'92	'93
Belgium	376	379	59	61	157	161
Germany	552	526	67	66	121	125
France	2376	2205	289	272	121	123
Italy	1514	1419	207	194	137	137
Netherlands	1197	1174	184	187	153	159
Other states	308	258	33	31	107	120
Total EU	6324	5962	839	811	133	136

The production of milk substitute fed veal calves is based on low prices for skimmed milk. With the present quota system to reduce milk production in the EU the production of veal calves has also reduced and a further decline is foreseen. Furthermore, public opposition to the production methods to produce white veal through iron deficient milk substitutes is growing. There is also particular antipathy towards the individual housing system for veal calves. Large-scale public protests in the UK have resulted in a (temporary ?) halt to the export of bobby calves for veal production to mainland Europe.

Group housing is becoming increasingly common in the Netherlands (13 % in 1991) and Germany (25 % in 1991) and it is functioning well. In Italy and France scarcely any veal calves are housed in groups. In the UK group housing for veal calves is common, but the total veal production is limited. Research has also been done on individual housing, concerning the minimum box size from an ethological point of view and the limited supply of roughage. Most of these changes are unattractive for veal producers, not because productivity will decrease (e.g. inclusion of some roughage might even increase daily growth rates; Van der Braak and Mol, 1991) but because the costs will substantially increase. These trends are more in line with the needs of the animals and the demands of society. Policies on animal welfare are largely based on the results of this research and are mainly focused on improving the health of calf group-housing systems. Prospects for the veal industry are determined by necessary investments in the environment and animal welfare. As this will reduce the profitability of the sector, a smaller scale production in the future is foreseen (Woudstra, 1991).

2.3.3. Sheep fattening in WANA.

Sere and Steinfeld (1995) provide the only current estimates of the number of sheep fattened in feedlot situations. They estimate that on average 10 % of the sheep are fattened in feedlots, producing 15 % of the mutton in the area. Based on this estimate 10 million sheep are fattened annually in WANA, producing 100,000 tons of mutton, mainly concentrated in Iran (37%), Syria (21%) and Algeria (14%). These 10 million sheep are only fattened for ca. 70-120 days. There are indications that in some countries a higher percentage of the sheep are fattened (Kamalzadeh, 1995). However, it is not clear if the data refer to peak periods or to annual averages.

Little is known about intensive sheep fattening trends in WANA. Sere and Steinfeld (1995) estimate annual growth rates for the period 1982 - 1992 at almost 10%. Increasing incomes and population growth has led to an increase in the demand for mutton. The demand for mutton has been met in the past by animals from the pastoral areas. Fear of overgrazing and degradation as a result of increased stocking rates has led to promotion of intensification through supplement. Barley is the main feed supplement. Feed subsidies are common even though nearly all WANA countries are net importers of barley. In Jordan, for example, the government feed subsidy is 35% (Maurer, 1994). The main effects of these subsidies are a growing independence of sheep farming from rangelands, a higher degree of market orientation and a further sedentarisation of sheep farmers. In most WANA countries these barley subsidies are or have been reduced, but the price ratio feed/mutton remains attractive for intensive fattening.

2.3.4. Large-scale beef production in EE and CIS

Accurate estimates of livestock numbers in EE and the CIS are scarce, and the proportion of ruminants in landless livestock system is completely unknown. Sere and Steinfeld (1995) estimated that 40 % of the livestock population would be part of LLR systems, producing also 40% of the beef. This estimate is rather weak as no dairy farming is included while growth rates are assumed to be similar to land based livestock production. This estimate assumes LLR systems to be far more important than data from other sources suggest. According to Mudakar et al. 1992, the cattle population in CIS, by far the most important region, consists for 3 % of beef breeds and 37 % of dual purpose cattle. It is likely that not all animals are kept in LLR systems. Also, solely in the CIS, it has been indicated that specialized beef fattening farms were producing less than 4% of all beef (Dmitriev, 1991), while part of these farms were not landless or did not have sufficient land in the near surroundings.

It has been argued before (section 2.2.4 ) that, strictly speaking, the LLR system hardly exists in EE and the CIS, but that their large-scale livestock units could have comparable problems. Data on the proportion of livestock in the private, the state and collective sector could give an indication of the extent of the problems. In the CIS in 1990, for example, 79.6 % of beef cattle were kept at state and collective farms producing 74.5% of the total beef production (Mudahar et al., 1992). However, the data is incomplete and varies from country to country, partly due to recent political developments. Moreover, these developments also make the assumption on the less concentrated nature of private farms untenable, as in some EE and CIS countries privatized state farms have not been abolished but continue to produce in a similar way as the former state farms.

2.3.5. Urban dairy farming

No statistical information is available on urban dairy farming. Nestel (1984) estimates around 1 dairy cow for every 10 citizens in urban areas in India, with city milkers commonly owning 10 to 20 cows. State Governments are attempting to shift these urban dairy units to rural areas, the Aarey Milk colony of Bombay (containing 15,000 buffaloes and 1000 crossbred cows) is an example of this. Maki-Hokkonen (1994) estimates 260,000 dairy animals in Karachi in Pakistan producing daily 1.7 million litres of milk. The average herd size is around 55 animals; within the Landhi colony the herd size is around 100 head. The human population of Karachi is around 6 million. Sere and Steinfeld (1995) did not include this production system in their estimates, because they assumed that hardly any manure problems exist. However, as concluded earlier, manure problems do exist partly because the lower DM content of the manure as a result of high concentrate rations makes this manure unsuitable for use as fuel. Moreover, little land is available for manure application in the surroundings of these farms due to their location in peri-urban areas.

2.4. Causes and motives

Though LLR systems are very heterogeneous, most subsystems seem to have one feature in common: a high livestock product/concentrate price ratio. In the USA, for example, live weight/maize price ratio was ca. 13-14 in 1992. In the EU beef fattening on high concentrate rations is less common probably due to a lower price ratio of 9-11. In most developing countries this ratio is much lower (2-3) making feedlots generally uneconomic (Simpson, 1988), though they exist in some countries, mainly in (small) areas with a good supply of cheap concentrates but sometimes also due to political incentives (Tacher and Jahnke, 1992).

The major exception are the LLR systems in the CIS and EE where price ratios were less favourable (e.g. 10.9 in the CIS in 1990), though this picture is very confusing as manufactured feed was subsidized. These less favourable price ratios did not prohibit high concentrate feed rations, partly because political incentives were overruling economic logic: state and collective farms were not driven by the need for profitable livestock production as farm gate price of cattle in Russia in 1992 covered only 60-73% of the cost of production (Mudahar et al., 1992). Hence these farms incurred major debts: highly problematic in the economic adjustment period (van der Graaf et al., 1990). It is envisaged that the practice of high(er) grain utilization for feed will gradually disappear in the near future mainly due to privatization of large collective farms and economic adjustments.

In the case of veal production and sheep fattening, subsidies on milk and barley play a crucial role. An extreme example seems to be Algeria where the mutton/barley price ratio increased from 30 in 1970 to 66 in 1987. Also in other WANA countries barley subsidies in combination with a free market for mutton has increased mutton/barley price ratio considerably (Treacher, 1993).

Favourable price ratios are as a result of low concentrate prices and in many cases also the result of the high premiums paid for the specific quality meat produced by the LLR subsystems. In developing countries usually no premiums are paid, so LLR systems in most of them are not economically viable. Some Latin American countries are the exception where beef fattening occurs and in many cases meat prices are hardly affected by meat quality all because of the low concentrate price. Direct or indirect subsidies on concentrate (grain or milk) are likely to diminish, thus increasing feed costs and decreasing the relative advantage of LLR systems.