A survey and case study was applied to characterise sheep production in the region of Xochimilco. The results demonstrate that the mean number of animals in the flock was 46, most of which belong to "criollo" (local) types. Animal housing was closely related with the owner’s home and the feeding regime was composed of grazing on crop by-products and native vegetation on non-conventional (unused) land. Processing of meat by steam cooking and of wool for traditional knitted garments are the main activities to supply products consumed by a strong flow of tourists from Mexico City. The non-competitive character of the system is discussed in terms of a proposal for its sustainability.

The region of Xochimilco, located to the southeast of Mexico City, comprises an area of 48, 526 hectares with a temperate climate and land use pattern consisting of areas of pasture and expanses of forest that has been traditionally used for sheep production (Sánchez 1982). Because of the close proximity to the cities of Mexico and Toluca, which consume most of the meat and manufacture much of the woollen products produced in Xochimilco, the sheep production system has survived throughout time and has surprisingly managed to preserve its traditional nature even though the growth of Mexico City has threatened to engulf the very places where production is maintained.

Important features of sheep production in the area include (i) the feeding of animals based on agricultural by-products and native grasslands, (ii) recycling of manure to agriculture, (iii) processing of meat sold as a local product linked to cultural traditions and (iv) use of wool to supply artisan industries. The characteristics of sheep production in the area have developed as a complementary system to the general agriculture of the region. Therefore, it falls under the definition of a sustainable mode of production (Smith 1992).

Although sheep production functions as an important economic resource for the local population, to date there has been a lack of research to understand its dynamics.

The objective of the research was to study sheep management and production in the region of Xochimilco, southwest of Mexico City and to evaluate its potential as a sustainable resource.

The main geographical features in Xochimilco were described in the paper of Losada et al (1995a). The procedure applied to the study included: (a) a productive evaluation to measure liveweight, duration of grazing and botanical composition of the herbage actually grazed by the animals when given access to "non conventional" areas; and (b) a survey directed to producers to understand the conditions surrounding the social, technological, economic and cultural aspects of sheep production.

The method used was based on the conventional procedure of questionnaires and interviews aimed toward the producers who work with the animals and local sellers of meat in the dominant market of the zone. The questionnaires were applied extensively in field conditions before their final application. The study area comprised seven important towns which were randomly selected. In total, 33 questionnaires were applied concerning 1,700 animals. In the second part of the research, seven flocks located in four towns were sampled weekly to ascertain liveweights, the duration of grazing and botanical composition of the actual herbage consumed in non-conventional areas. The data obtained in both procedures were analysed using a SAS program (SAS 1988) and the results were expressed as percentages and means according to conventional procedures (Daniel 1994).

Similar to other animal production systems we have already described in the study area (Losada et al 1995a), sheep housing is located within the perimeter of the producer’s home. In this respect, 79% of the owners’ homes were inhabited by one or two families most of which (56%) housed 6-7 individuals. Services in and\or close to their homes included electricity (100%), drinking water (88%), drainage and gulleys (70%) and street pavement (40%). Materials used to construct their houses were of concrete and solid brick with a significant proportion of owners (40%) reporting the use of asbestos lamina for the roof. A high proportion of sheep producers reported to be literate (91%) with a predominant level of primary education (53%) and secondary (16%) whereas a significant percentage reported no formal education (31%). Most of the work with sheep is carried out by men (72%) although women and children carry out some activities, in particular that related to grazing and housing the animals (28%). The grazing activity is often combined with the collection of wood or dry manure used as an energy source in the home. Work activities of the producers included: agriculture (58%), unskilled work (26%), commerce (9%) and the care of the household (7%). A small proportion of owners reported to be members of a rural society (9%) to assist them in obtaining governmental loans.

Data on the mean number of sheep per producer are presented in Table 1. Mean count of the flock was 46 animals with ewes showing the highest value ( 65% including young females) followed by lambs, whereas rams, as expected, occupied the lower value. The count of 46 is much higher than in the case studies as one producer owned a flock in the region of 300 thus skewing the mean value. The ratio of ram: ewes was 1 to 12. The type of animals used was mainly the "criollo" (local breed originally from Spain)(64%) or crosses with Suffolk (21%). The rest belonged to the commercial Merino breed.

Table 1: Composition of the flocks in the region of Xochimilco.
Type animal	Total	%	No per producer
Rams	66	4	2
Ewes	792	52	24
Female offspring
7-8 mths	198	13	6
Lambs	429	28	13
Castrated males	33	2	1
Total	1518	100	46

Housing conforms largely to the use of a pen to care for the animals during the night in the producer’s backyard (79%) whereas a minor proportion house them actually in their own home or in the field (21%). A small proportion of producers (12%) reported the use of feeders and drinkers made of wood, galvanized metal or a rejected half cut tire. Only nine percent housed the sheep in a type of shed in the corral.

In all the cases studied, manure from the animals was collected, dried and carried to the agricultural fields to be used as a source of nitrogen and organic matter. A significant proportion of producers (21%) utilized a mobile pen made of wooden planks and meshed galvanized wire to house the animals for up to 16 hr a day in those fields planned to be planted later. This system involves rotation of the pen every two days until the field is covered with a mean layer of 4 cm of manure, which afterwards is incorporated into the soil.

The sheep were grazed on the edges of roads and on banks of streams, with only a minor proportion having access to introduced pasture of Ryegrass. A significant proportion of farmers grazed their flocks on native species in the forest and on natural grasslands. Supplementary feeding included such materials as "tortillas", straw, hay and fresh lucern. Only a small percentage offered the animals commercial concentrates.

Table 2: Feed resources used for the sheep and frequency of use
Type	Frequency of use (%)
Agriculture field	36
Grazing:
Non conventional areas	28
Rye grass	16
Forest	11
Native grass	9
Supplemental feeding:
Tortilla*	41
Oat straw	24
Hay	14
Commercial concentrate	10
Maize stubble	8
Fresh lucern	3

* Mexican bread made of maize

Data on the lambing season in the area are presented in Table 3.Higher percentages of lambing were reported during the period of January to March as a response to the natural mating season during August-September. Most of the producers reported one lamb per delivery and of weaning them at six months of age. Annual values reported for rejection of animals to slaughter were of the order of 16% whereas those established for selection were 5%. The majority of producers shear once a year (58%), the rest (42%) shear biannually. Main season to shear sheep is during the period of February to May at the end of the winter season (73%) and during the period between August to September (27%). None of the producers appeared to keep records of production in any way. Identification of animals, when made, is carried out by names and brands.

Table 3: Lambing percentages of ewes during the year in Xochimilco
Period	Percentage of producers
January-March	55
July-September	27
October-December	18

Sixty five percent of farmers reported that they vaccinated their animals. Forty three percent reported cutting the tails of the lambs whilst a low percentage castrated males during the weaning period (14%). Main illnesses were: cold, pneumonia, diarrhea, hoof-rot and fasciola. Treating sick animals is carried out by the use of specific medicines for sheep (45%), home remedies (32%) and human medicines (23%). Combatting internal and external parasites is done by a significant proportion of producers (52%).

The results of the survey showed that the main objectives of sheep producers in the region of Xochimilco included: meat (45%), wool (38%) and breeding (17%). The majority of farmers interviewed calculated that a value of 30% of their total income arises from sheep production. Self consumption of meat and wool attained a mean value of 21% of production with highest values observed for meat (29%) and (14%) for wool. There is no fixed seasonal pattern for selling the animal during the year which is understandable as most of the beasts sold to slaughter are rejected adults or finished castrated rams. The sale of animals for meat consumption is carried out in all cases by guessing their liveweight in a wide variety of locations including: the house of the producer (45%), local markets (30%), local intermediaries (15%) or the slaughter house (10%). The mean value reported for wool production per animal is in the range of 3 to 3.5 kg. Although a small percentage of the fleece is sold within the area, most of it is transported to be sold in the nearby towns of: Topilejo, Gualupita, Toluca and Puebla outside the region of Xochimilco, and made into traditional garments.

A very important feature of the system is related to the particular way that the processing of the meat of the slaughtered animals influences the local culture. Different to other regions of Mexico and overseas countries, the selling of fresh meat for cooking is rare. Instead, most of the meat is steam cooked then sold as "barbecue". Main features of this process are the use of stone ovens built either above or below ground, the heating of which is carried out by burning large quantities of firewood. Agave leaves are positioned to cover the oven walls and a receptacle at the bottom is used to capture liquids from the large pieces of salted meat and offal that are placed on a grill above. The receptacle receives a mixture of herbs, spices and chickpeas and together with the fat dripping from the meat above combines to make a thick stock. After the oven is lit the meat is covered with more agave leaves and the oven is sealed with soil to allow the steam cooked process to carry on over a mean period of twelve to sixteen hrs (H Martínez 1995 personal communication). This process gives a particular texture, flavour and taste to the meat. The cooking technique is particularly suitable for sheep which naturally have a high content of collagen in their tissue rather than lambs. The traditional method of preparing the meat complements sheep production, as lambs are rarely slaughtered but are raised to maturity. The consumption of meat as "barbecue" has a strong cultural influence in maintaining sheep production in the temperate areas of Mexico and the region of Xochimilco is an important zone of Mexico City. Our research shows most of the barbecue is sold within the area especially at the market of Xochimilco which is the most influential in the region.

The composition of the studied flocks is presented in Table 4. As may be appreciated, the size of the flocks tends to be small with a mean number of 12 animals per producer. Also, with management aiming to raise lambs to maturity, the presence of males in the flocks was high as evidence of meat production being the main goal of the units. Preferred breeds are the "Criollo", and the commercial Suffolk, Hampshire and Rambouillet.

Table 4: Composition of the flocks in the case study
	Type of animal
Farm	Males*	Ewes*	Total
1	3	3	6
2	5	5	10
3	5	6	11
4	5	5	10
5	5	7	12
6	6	7	13
7	2	18	20

* includes adult and young animals

The mean live weight of the animals during the study period is presented in Table 5. Predictably rams were heavier than ewes and the same trend was observed for lambs. Grazing times were reported to be 3 (17%), 4 (17%), 6 (32%), 7 (17%) and 8 (17%) hrs per day measured as actual time spent in this activity. The botanical composition of the herbage consumed from the non- conventional areas of grazing included a wide variety of species as shown in Table 6. Thus type of vegetation is native to the region of Xochimilco and is available most of the year.

Table 5: Mean values of live weight for the sheep in Xochimilco
Animal type	Liveweight (kg)
Ram	40
Ewes	37
Male lambs	17
Female lambs	12

Table 6: Botanical composition of herbage from non-conventional grazing areas used for sheep in the region of Xochimilco
Common names:	Classification
1. Acahual	Bidens aurea
2. Cocks foot, Grama	Cynodon dactylon
3. Estafiate, Artemisa,
Ajenjo del país	Artemisa Mexicana Willo
4. Ducks tail, Little bean,
Little hen	Carnavalia Villosa Benth
5. Clover	Melilotos B. Erchart
6. Wheel barrow	Medicago Denticulata
Wind
7. Jadamao, Tea of maize
field, Acahual bronco	Bidens pilosa
8. Malva	Malvaceas hibiscus
9. Rye grass	Lolium perenne
10. Sheeps tail	Castilleja tenvifear
11. Fox tail,
Little worm	Lobelia fenestralis
12. Cow tongue	Rumer Potcher

Similar to other systems we have studied in the region (Losada et al 1995b ), the results obtained in the present study demonstrate that sheep production has evolved in close relationship with the natural resources of the zone and to cultural features which lead the producers to apply a technology which can be considered to be a model of sustainable production. Some interesting aspects supporting this suggestion are those related to the number of animals in the flock, predominant breed of sheep, the grazing habits, manure management and the processing and marketing of the meat and wool.

As it has been discussed in previous papers referring to traditional systems (Suárez and Barkin 1990), the mean size of the flocks in the area comprised 46 animals, though a lower number of animals for most of the producers appears to be dominant. The low number of animals contrasts greatly with technified sheep production which usually contains a larger number of animals within the flock (Galloway and Grant 1984). In this respect it would appear that the shortage of natural grasslands and their non-uniform distribution in the region may be the main factor limiting the number of animals. On the other hand, the type of animals preferred was the "Criollo" which seems to be best adapted to a management which favours animals capable of selecting (Identifying?) feed resources such as crop byproducts and native herbage species that dominate in non-conventional grazing areas.

As an adaptation of the system to the sub-urban conditions of Xochimilco an interesting aspect to further discuss is that related to the source of pastures to feed sheep. Because most of the arable lands in the region are cropped according to a seasonal regime, crop byproducts tend to follow the seasonal pattern with higher availability of grazing after the crop season in October and November (INEGI 1994). To compensate for this effect throughout the rest of the year most of the grazing is confined to what we have called "non- conventional area" (Cortés et al 1992) defined as areas covered with natural vegetation (including mixtures of good quality grass and legumes) that are not specially cultivated for use by ruminants.

According to our research, non-conventional areas for sheep grazing include a wide variety of places: roadside verges, stream banksides, public gardens, wasteland and other similar locations which are extensively used by the producers to graze their animals as an important but inexpensive way of providing nutritious feed that in the conditions of large cities is usually wasted. The botanical composition (see Table 6) shows the presence of twelve species which supply plentiful biomass to the animals. During the autumn period when pasture grazing is reduced the producer feeds the animals with a wide range of local products such as tortilla, hay and maize stubble which allows them to support the animals until the next forage season.

Evidence of the mobile pens for manuring agricultural fields has been stated by a number of authors as being in use in other regions of Mexico (Nahed and Parra1989) particularly in the temperate high land areas of Chiapas that possess similar conditions of small flocks, mainly of the Criollo breed. Here the use of the mobile pen is on land near the household, later to be planted with vegetables and flowers (Pérezgrovas 1990). In the case of Xochimilco the use of pens is focused on cereal production, particularly maize, to be sold as fresh corn-cob or on introduced species of grass (mainly Rye grass).

It is clear that in traditional systems the link between agriculture and animal production becomes an important feature for maintaining diversity a fact which is rarely known in technified systems where there is a clear trend to favour monocultural processes. Due to reduced urban development, and therefore the presence of green areas and forest together with prehispanic remains and towns possessing markets and local foods, the area has developed a strong flow of tourists from the city during weekends. In this respect the processing of meat from sheep as barbecue and sale of wool garments provide an important attraction for the visitors. Demand for barbecue in the region is high enough to be the principal way of supporting sheep production. However, often the supply of sheep meat does not satisfy the demand. According to some data (DDF 1985), a mean value of 2000 slaughtered sheep per week has been reported as being processed in the local abattoir, most of which come from areas neighbouring Xochimilco. There is a similar overdemand for wool products. Though the region lacks a singular large retail outlet, the town of Gualupita in the neighbouring southeast manufactures a wide variety of wool garments in a cottage style industry such as: jorongos (a sort of poncho), sweaters and caps (Sánchez 1982). It is clear that the main products from sheep production (meat and wool) have found an important place to sustain tourism, allowing the producers to create a defined system based upon the selective use of natural resources.

The features described of the sheep farming system indicate that it is a non-competitive model of production. In addition, the intensive use of manure as a source of nitrogen and organic matter for certain crops in the zone, and a low use of external inputs to produce the sheep, results in a system which can be classified as "non-polluting" (Ghimire 1992). Though the reported contribution of sheep production to family incomes was of the order of 30%, which on its own would be a limiting factor within the sustainability concept, it has to be stressed that sheep production forms part of a complex family model comprising several species of domestic animals (Rivera et al 1993). Furthermore, as mentioned earlier, most of the owners obtain their incomes from agriculture and other work (unskilled, commerce and so on) thus the production of sheep is complementary.

Concerning those aspects related to sustainable animal welfare, brought to prominence in recent years (Fraser and Broom 1990), it is worth mentioning that the use of small flocks tends to ensure better care of the animals and therefore also satisfies the welfare requirement. However, there is some evidence that animals are mistreated during slaughter in the abattoir, a feature that ought to change in the near future.

The authors wish to thank the sheep producers in the region of Xochimilco for information and allowing us to measure some parameters of their flocks; the authorities of the Universidad Autonoma Metropolitana (Autonomous Metropolitan University) for facilities given to the research; and the students of the Animal Production option who helped with the field work.

Received 21 August 1996