A survey was carried out in order to understand the milk production systems in dairy herds under suburban conditions in the east of Mexico City. Dairy production is an activity which is carried out in the back-yard of the household. The number of animals in each household varied between 3 and 250 with the highest percentage of producers (63%) owning small herds of between 3 and 19 animals. The main breed was commercial Holstein. The feeding system was complex with a wide use of by-products and rejected vegetables from the markets. Milk production varied widely between 6 and 22 litres per cow per day. There was strong evidence that the milk was being mixed with water in a ratio of 4:1. Surplus milk was processed to make cheese, cream, jelly and custard or given to calves in large quantities to be transformed into liveweight.

Dairy production under suburban conditions in the east of the metropolis is a good example of adaptating the production system to the conditions of the city. The system has evolved in a way that has made it sustainable from the economical, social and environmental points of view. It could be used as a model with potential to be replicated in other similar environments.

Milk production in Mexico is mainly carried out with a high use of technology and inputs (Bernal and Garza 1994). Although this model of production has been broadly accepted by the majority of producers, milk production in the surrounding high population areas of the city is carried out by an equally intensive system which has particular characteristics well adapted to the suburban conditions of the city (Cortés and Losada 1993). One of the most important features is the fact that it is linked to good management and recycling of wastes. For example: (a) feed often includes rejects from the markets and food and processing industries (Losada et al 1992a) and; (b) manure from the animals is recycled for use in the nearby agricultural and horticultural industries. Thus, the model is potentially a sustainable system of dairy production even located in a highly polluted city such as Mexico.

The production of milk from dairy herds in Iztapalapa, to the east of the city (Losada et al 1992b) constitutes an interesting model, because of the large number of productive units that have survived throughout time, despite state pressure to eliminate them.

Mexico city has developed according to two main trends: (i) the North American model, with large concentrations of commercial activity and a way of life that accompanies the high levels of income; and (ii) very poor, large areas, where most of the working class live in great poverty (to support the highly developed areas). Iztapalapa is one of sixteen political delegations (Wards) into which Mexico City is divided (DDF 1985). It is located to the east of the centre surrounded by the delegations of Ixtacalco to the north, Xochimilco and Tlahuac to the south, Tlalpan and Coyoacan to the west and the State of Mexico to the east (its extreme coordinates are: N19°23'50''; S19° 17'06''; E98° 57'45'' and W99° 08'13''). The altitude is 2250 masl According to the modified Koopen System, the climate of the zone has been classified as: C(w2)(w) and BS1k which belongs to the temperate sub-humid and semi-arid zones respectively, with mean annual temperatures ranging from 15.3 to 16.6 °C. The range for total annual rainfall is 530 to 617 mm with a rainy season located between the months of May and October (García 1973). Total land area is 124 km² with 75% devoted to urban use. The population has been estimated at 1.1 million inhabitants which gives a density of 9,677/km², 30 % above the mean calculated for the city (Sánchez 1982). The population is divided into five towns and 91 settler camps. Access to services in the zone comprise: drinkable water (75%), drainage and gulleys (70%), electricity (90%), street light (70%) and surfaced roads (50%). According to mean incomes, the population has been clasified as extremely poor with the main economic activities including: manufacturing, commerce and services (Montaño and Rendón 1992). At the center of the area is located the main food storage and supply center of the city (328 ha), which receives and sells most of the products from the central and southern states of the country (CEDA 1993).

Although the presence of domestic animals in Iztapalapa is restricted by the local authorities, a high percentage of the population keep them in their backyards as an important contribution to family economy. Official estimates of the presence of domestic species in the area have shown values of around 12,000 heads of dairy cattle, 20,000 pigs and 300,000 poultry (SARH 1979).

Before the conquest of America, Iztapalapa was a royal village located at the east side of the five lakes, the largest of which was Mexico-Texcoco situated in the Valley of Mexico surrounding the great city of Tenochtitlan (Gonzáles 1977). The main activity was agriculture, practiced on the chinampa system (floating gardens) with a high production of vegetables, flowers, fish and edible insects. With the fall of the city to the Spanish in 1521 the lake was drained and the land divided among the conquerors (Gortari 1981). The colonial powers made significant changes to the agriculture of the region introducing seeds, plants, domestic animals and new technology (Woodrow 1980). With the new agriculture, based on the utilization of maize, wheat, barley, forage and native grass, the lands of Iztapalapa were devoted to the production of cattle in order to supply milk, meat and leather to the city (Montaño 1984). In 1854, Iztapalapa was incorporated into the Federal District, first as a prefecture, later as a municipality (1903) and thereafter as a delegation (1928)(DDF 1985).

Throughout this era the cattle production in the area mantained its functional structure without being affected substantially by the changes inland tenure that accompanied the political transformations of the country. The haciendas were transformed into smaller ranches, although owners of the haciendas maintained much of the property rights of the old system and thus continued the old ways of agricultural production. At the end of the nineteenth century, European immigrants began specializing in cattle production by means of selection for milk production, the introduction of lucerne to feed them and the creation of milking parlours (Reyes 1981). With the end of the Mexican revolution in 1920 there was a great effort by the Government to further improve the old methods of production. Holstein cattle were introduced which, in the short term, became very popular. This new method of milk production was transferred to most of the temperate regions and reached Iztapalapa and the neighboring areas, making an important contribution to milk supply for the city. With the rapid growth of Mexico City after the revolution, there was a clear trend towards urbanization in the surrounding areas of the metropolis and in 1954 Iztapalapa became an official urban area to sponsore the development of the service and manufacturing industries, thus marking the official end of the agricultural and livestock industry (Montaño 1984).

At the begining of the seventies a second ordenance from the Federal Government restricted the presence of animals in the city to delegations with low urban development. As a result of urban growth in Iztapalapa the value of land increased as a consequence of which ranches were sold to build houses and industries. Dairy cattle were settled into smaller areas within the built up zones, sharing them with family members. Towards the late seventies most of the ranches in the south of the city were moved to new milking parlours outside the Federal District but some of them found a place in Iztapalapa, though they were unlawful. At the begining of the eighties a large Metropolitan Food Supply Depot was constructed in Iztapalapa. This situation led to a high availability of vegetable and fruit rejects for which good use was found for the livestock of the area (Grande et al 1994), dairy cows occupying the most important place.

The study reported here is a part of a programme developed by the Animal Production Systems Area of the Autonomous Metropolitan University, Iztapalapa, aimed at understanding and providing scientific support to the livestock producers in the perimeter zone of the University. The project started with a survey directed at local dairy producers. Questionnaires were designed to understand the socioeconomic and technical features of suburban dairy production systems and randomly applied to the producers. Dairy stables were located visually and by means of information provided by the producers and/or neighbours. Forty six producers were interviewed covering a total of 1,200 animals. In the second part of the project a dynamic case study was developed among 12 stables over a period of eight months into which 60 cows ( five from each stable) were sampled to acertain feed intake and milk production. Samples of milk from two of those stables were analysed to determine the fat and protein contents (AOAC 1990) whereas another ten samples were randomly selected for microbiological studies (Acevedo et al 1995).

Dairy production in Iztapalapa is an activity which is carried out in the backyard of the household. According to the survey 72% of the holdings were inhabited by one family though there were reports of two (11%) and three families (13%). Services in the households included: electricity (85%), potable water (71%) and drainage and gulleys (63%). The floor of the houses is built of cement and the walls of solid bricks, whereas the roof is of corrugated iron, or is covered with bitumen, cardboard or cement. The majority of producers obtain the major portion of their incomes from the sale of milk and animals (78%).

The care of animals is carried out by men. Schooling level included basic and superior education (20% and 37%), although a significant percentage of producers (20%) reported as being illiterate. The majority of dairy owners (73%) do not belong to any of the local societies of producers which means that each deals with the trade of animals, foods, etc, and the sale of products at their own convenience. A minority of them (27%) belong to local societies whose main function is to avoid persecution from the authorities.

The number of animals per holding showed a variation between 3 and 250 with the highest percentage of producers (63%) owning small herds of 3 to 19 animals (Table 1). Mean composition of the herd per producer in the area (Table 2) included 27 animals with 78% of cows in lactation whereas the rest were dry cows, growing calves and an adult bull. The main breed was the commercial Holstein, which is a mixture of the American and Canadian types. A small percentage of producers reported preferences for the Brown Swiss and Jersey.

Table 1: Number of animals per holding at Iztapalapa
Strata	Range	No. producers
1	3-9	18
2	10-19	11
3	20-49	9
4	50-250	8

Table 2: The mean size of the herds in Iztapalapa
Total	27
Cows in production	21
Dry cows*	3
Calves	2
Bulls	1

*Animals close to parturition

Feeding of the animals involves a complex system based on a combination of conventional and traditional foods. A list of these, together with the frecuency of their use is presented in Table 3.

Table 3 : Frequency of utilization of feeds for dairy cattle at Iztapalapa
	Proportion of producers using the feeds, %
Sundried lucerne	74
Wheat bran	52
Maize stover	52
Fruit/vegetable rejects	48
Commercial concentrates	43
Introduced grasses	30
Ground maize	33
Oat grain	11
Coconut oilmeal cake	9
Bread scraps	9
Sugarcane bagasse	7
Tortilla	7
Fermented maize dough	7
Distillery bagasse	4
Oat straw	4
Groundnut hulls	2
Soya bean meal	2

The local system is based on the use of twenty types of feed with a major use of sun-dried lucern (20% dry matter), maize stover, wheat bran and a commercial dairy cow concentrate. A high percentage of the owners (87%) also offered commercial mineral salts. The most widely used grass and legume species were kikuyo (Pennisetum clandestinum) and clover (Trifolium spp.). These are frequently grown in the public gardens of the city and propagated in other areas (on sidewalks, areas of waste land and hillsides) as a green cover crop which the producers use for grazing the animals and/or harvest as a natural crop to feed the dairy cows directly in the stables. Another very important component of the feeding system of the Iztapalapan stables are vegetable and fruit rejects originating from the main market of the city known as the Metropolitan Food Supply Depot (Table 4).

Table 4: Main portions of vegetable and fruit rejects utilized to feed cows in the stables of Iztapalapa
	Leaves	Fruit	Flower
Broccoli	*		*
Cauliflower	*
Lettuce	*
Carrot		*
Maize		*	*
Pumpkin			*
Cabbage	*
Radish	*
Turnip	*
Sugar beet	*

Some of these products have a seasonal availability, mainly carrots, pumpkins and green maize which are sold to the producers either because of poor quality for human consumption or low demand. The rest of the products are composed of leaves which are natural rejects from the market and their availability is permanent throughout the year. These rejects are given free to producers or sold at a very low price, which makes them a very profitable source of feed for local milk production systems.

Although there are no defined criteria used by the owners to feed their cows the proportion of forage in the diet occupied 76% of the total consumed (SD±13.0) whereas for the concentrate the proportion was 23% (SD±3.6). The decision to fix the forage/concentrate ratio included: price of the feeds, cost of transportation, availability, individual preferences and tradition.

A high percentage of producers (80%) sell the male calves within three days of calving to be slaughtered; the remainder are raised to be fattened and sold to the slaughter house at a mean live weight of 400 kg ( 17%) or kept as sires for local herds (3%). Female calves are retained by producers to be used as replacements or sold to other owners for the same use. Mean live weight reported for the calves at birth in the study was 33 kg (SD±9.9). Rearing is carried out using whole milk (76% of the producers) offered by means of a bucket, with the remainder utilising a mixture of whole milk and a commercial milk substitute. The amount of milk offered to the calves ranged from 3 to 6 litres per day depending on whether or not the milk will be sold for human consuption. The time to weaning was a mean of 90 days (SD±23). Once the calves are weaned the diet offered to them was the same as that given to the cows.

Most producers detect cows in heat (estrous) by observation of animal behaviour (mounting) and/or the presence of vaginal mucus secretion. There is a clear trend among most of the owners to use natural service of the cows (87%) with only a small percentage of them using both artificial insemination (AI) and natural service. The reason given for preffering the use of natural mating is related to the effectiveness of the first method measured in terms of conception rates. For this reason the values obtained for heat repetition in small stables are quite low (12%). A secondary reason that justifies the use of a bull has been related to the fear of the cattleman of wasting heat time waiting for a technician’s assistance. According to the present research, reproductive behaviour has values within the range of 130 to 140 days for the calving/conception interval with a range of 373 to 411 days for the between calving intervals. The mean mumber of services to achieve conception was reported to be of the order of 1.6 to 1.9, with a strong indication that reproductive behaviour tends to improve as the number of animals in the stable is reduced.

The criteria chosen by the producer to select future females for production (Table 5) are based on body conformation, the level of milk production from the mother or both, whereas the breed is only taken into a consideration by a very low percentage of cattleman (4%). The reasons given for rejecting adult cows from the herd (Table 6) were: age, low level of milk production and chronic sickness.

Table 5: Criteria to select cows in the stables of Iztapalapa
Criteria	Percentage
Body conformation	39
Body conformation/milk production	36
Previous production	13
Breed	4
Other	8

Milk composition analyses showed similar low values for the fat and protein content (2.9% and 2.8%, respectively). There were indications of higher levels of fat (4.6%) for those stables with a relationship of forage/concentrate of 65/35 in contrast with values of the order of 2.8% fat for those producers who feed the cows at a ratio of 78/22 of forage and concentrate respectively (Cortés et al 1993). Bacterial counts showed the presence of both pathogenic microorganisms (Brucella and Mycobacterium ) and pollution related organisms (Escherichia, Klebsiella, Proteus, Salmonella, Shigella and Staphilococus) which were eliminated by the traditional way of boiling the milk (Acevedo et al 1995).

The major health problems were: mastitis (13%), pneumonia (17%) and foot rot (11%) which are treated by the use of specific medicines (76%), home remedies (11%) or human medicines (4%).In the case of foot rot, the producers contract hoof cutters periodically as a method of preventing the illness. A minimum of producers reported vaccinating cows against brucellosis once in their life cycle (17%) whereas most of them treat the animals for internal parasites (84%). The control of external parasites, for example insect larvae, was difficult. However, this was carried out persistently within the stable.

Milk production in the area is the result of a complex strategy which includes: production, processing and the care of the animals. Evidence from the survey suggests that the stables in the area follow a heterogeneous spatial distribution and it would appear that each unit is supported by local milk demand. For this reason most of the milk produced in the stables is sold within the unit itself (83%) and only a small percentage is sold in local markets (9%) or to intermediaries who resell the milk directly to consumers at the door step. With the exception of milk that is processed as a minor activity, most of the product is sold retail in a raw form (unpasteurized) and unbottled, that is to say straight to the consumer’s containers. The price of milk in the stables is equal to the commercial one (0.38 US$/litre). There was strong evidence that the milk was being mixed with water in a ratio of 4:1 and it was discovered that some producers make a sort of contract with the consumers to avoid water addition, provided thus the consumer agrees to pay a fifty percent higher price (0.57US$/litre) than the standard one. When total milk production is not sold, particularly for those who have a large number of animals, the stablemen process it to make cheese, cream, and popular cultured dairy products mainly for deserts such as jellies and custard. A second method often utilized by the producer for using spare milk is by means of feeding large quantities of it to calves (up to six litres/day per animal) to transform it into live weight. According to this situation the local demand supporting the unit, as mentioned before, is a full circle because the producer is not trapped into a rigid scheme of marketing.

A second important feature of the stables in Iztapalapa, associated with marketing, is related to restrictions in the physical suburban space of the unit, which has an effect on the management of the animals. More than half of the cattlemen (57%) buy adult dairy cows in middle or late pregnancy from nearby milksheds (supply areas). For this reason replacement of rejected cows by means of raising their own heifers becomes secondary. Although most of the producers sell the male calves at birth, a small percentage of them are kept for fattening and sale for meat production or reproduction. According to this situation the model works as a dual purpose system into which milk production and live weight become the main objectives. This is confirmed by the fact that most of the producers mantain the productive cows at a heavier than normal weight, so that they will get an increased return when they sell them to the slaughter house. Another support for this hypothesis is related to the lower use of artificial insemination as producers do not consider the genetic potential of the future females to be of great importance. Similarly, the selling of reject animals (including male calves at birth) is done at the stable itself (63%), though a small percentage of producers take the animals to the slaughter house to be sold. In either case, the sale is mainly done by guessing the live weight which in practice, provides a good method of coping with the absence of weigh scales.

To date, we have not been able to estimate the ratio of cost: benefit for milk production in the area. However, there are some indications suggesting that the system is profitable with respect to several factors: (1) shared expenses of living accomodation and livestock production, (2) management of the animals is mainly carried out by the family, (3) utilization of vegetable rejects and other products derived from the local human food industry to feed the animals, (4) the use of technological inputs is low, (5) milk is being sold unpasteurized and unbottled at a viable price (most of the producers add water to the milk in order to increase their profits, (6) there is not an established standard of quality for milk fat content, which allows the producer to use lower amounts of concentrate in the ration, (7) the system functions as a dual purpose one, which enables the cattleman to obtain additional income from his animals and (8) the producer does not pay any taxes on his dairy business.

Although urban development in Iztapalapa has been the major trend in the last forty years the historical features of the area have been strongly influenced by agriculture, which is reflected in the way of life of the local people (Rodríguez 1992). In this respect, milk production from dairy herds became a good example of adaptation to the conditions of the metropolis. The system has been functioning in a way which copes with sustainability from the economic, social and enviromental points of views that should be taken into consideration in order to validate it as a model with high potential to be replicated in the future. In contrast to the technified system of milk production that is carried out in the majority of the country (according to a rigid model of forage and cereal production, genetic selection of cattle, high utilization of industrial inputs and marketing of milk under pre-established conditions (Giaevar 1979), the stablemen of Iztapalapa have modeled a system which allows them to manage their animals at their own convenience. This enables them to overcome most of the problems associated with the utilization of technology to increase milk production and/or the quality standards of the product (milk) for consumption, which have proved to be barriers that decrease the profitability of milk production in the rest of the country. In this respect, it would appear that the stable as a productive unit mantains a particular system of production founded on the historical characteristics of the area, the availability of resources from the city and the needs of the range of consumers, which favours the presence of the system.

The particular way of building houses in the rural areas of the country, with large backyards to keep animals and plants, was continued despite processes of urbanization. This phenomenon allowed them to maintain some of the features of the old ways of production keeping animals instead of agriculture which would necessitate large areas of open land. A second factor that favours the presence of dairy cattle in the area is related to the a consumer preference for the consumption of rural products including milk, rather than that produced commercialy, a fact that tends to function as strong support for the local dairy systems. The high density of human population in the area has resulted in a huge availability of sub-standard products that are not of good enough quality for human consumption due to the lack of freshness of certain products from the food processing industry - such as stale bread, crackers, tortillas (mexican bread from maize), contaminated wheat flour and maize dough. The stablemen have found a market for this waste product and feed it to their dairy cows. The Metropolitan Food Supply Depot provides an estimated sixty tonnes per day of vegetable residues (CEDA 1993) which are used as an inexpensive source of forage for the diet of the local livestock.

Although watering down the milk by the producer before the product is sold to the consumer is a phenomenon that could be considered as a disadvantage of the system from the point of view of consumer rights, in the actual situation of the country as a whole it does not necessarily equate with a serious problem. Twenty years ago the country followed a strict policy of maintaining standards of quality for milk similar to the policies applied in developed countries (Broster and Swan 1983). With the onset of the economic crisis of the 1980's, the policy tended to be less tightly controlled, and dumping the product in a watered down condition without the purchaser’s consent became normal practice. In this respect, it is possible to suggest that the Iztapalapan producers who water the milk are replicating the results of neglected government policy. A second implication of the new policy was the replacement of whole milk from the market by processing it into partially skimmed and skimmed milk, thereby adding value to the industry profits. A third change was related to the acquisition of dairy industries by the government in order to sell reconstituted, imported, powdered milk, and make it freely available to the population at low prices in exchange for political favours. This final argument in support of watering down milk is that related to quality perception, which is understood by the preference given by the consumers. As previously mentioned, there is a clear trend for the customer to prefer milk from the stables rather than the commercial product because by buying the milk direct from the stable, they enjoy the advantage of seeing that the product comes from the farm and if thereafter they wish to make into butter, they know it will produce a good end product. In addition, some consumers have reached an agreement with the stableman to pay an extra fee in order to get whole milk without any water which suggests that the consumer is cogniscent of the watering down of milk.

Although there is great variability in the range of milk production between local stables, evidence shows that production per se and cases of over production, do not constitute a problem for the stablemen. The system has found alternative ways to utilize the extra milk produced by means of processing it to make desserts or to feed calves, although the producer does not hold this as a main objective. This situation introduces another variable in the analysis of the system which can be taken into consideration to explain the ranges of variation in milk production and possibly the number of animals between different stables. Nevertheless, it is recognized that the restriction of physical space has a direct effect on the potential for increasing the number of dairy cattle. However, evidence has shown that when demand for milk increases beyond the production of the unit, the producer searches for another place to keep a larger number of cows and thereby increases milk production to fulfill the new demand. It is clear that the whole consumption of milk is the main variable that influences the number of animals in this system although the decision is also influenced by a variety of additional factors.

The logic of the stablemen in producing milk in Iztapalapa, functions in a way which suits the particular features of the city. An example of this is the number of animals in the unit, which depends to a high degree on milk demand from the local population. Whereas in the highly technified system there is a strong effort by the producer to improve the genetic potential of their cows to increase milk production per animal (Veerkamp et al 1994), the preference of the stablemen is focused on increasing the number of animals. This difference could be explained by the dynamics of the system itself, which seeks to reduce to a minimum non-lactation times in order to sustain milk products for human consumption. For this reason the producer buys adult cows from the neighbouring technified milk sheds, rather than raising them from his own stable which, in the end, seems to function well. Another advantage of this decision is that the producers avoid all of the technical implications of improving the genetic potential of dairy cattle. These include: semen, insemination and technician expeditures; care for calves and a lower resistance to illness because of the higher sensitivity of new heifers, all of which would result in a real loss of the independence of the system. This last point could explain why the stablemen prefer to use local bulls to mate the cows, as the main goal is milk production and not the newborn product.

The milk production model of the stables of Iztapalapa is one that deserves consideration for future development in other main cities of the country. To date the system has been unjustly criticized and prosecuted by the local authorities. The main reasons for this are related to claims of pollution in respect of bad odours, blockages of gulleys and drainage systems and the promotion of flies and rats associated with the presence of animals and/or feeds, in addition to risks of endemic illness for the human population related to the consumption of raw milk. Though some of these claims may be true, the owners of the stables have developed a management system focused on preventing most of these complaints. The majority of the producers maintain a good policy of cleanliness by means of washing floors and collecting manure to send to the agricultural areas which tends to reduce the presence of bad odours. They often utilize creosote to eliminate the problem of flies, while the presence of cats in the stables constitutes a good control of rodents. The risks implied in the consumption of raw milk are reduced with its consumption in a processed form, such as cream and cheese. As for milk, boiling it before consumption is on the increase which in a way resembles the pasteurization process.

According to the present research, the main pressure to eliminate the stables from the area comes from the construction industry seeking new places to build houses or flats. However, no protection is given against this sort of pressure by the official authorities for various reasons. The unlawful character of the stables allows some officials to obtain money in under-hand ways from the local stablemen. Reasons for mantaining the system are related to pollution aspects. Most of the reject food which is used by the producer to feed dairy cattle would became garbage and thus increase public expediture in the disposal process. Another factor in favour of sustaining the system is because it produces incomes for the family and offers viable employment for a significant percentage of the local population which, as noted before, has been classified as very poor. The last, and perhaps the most important reason of all, is related to the lack of an alternative model to be offered to the stablemen. In this respect it would be worthwhile mentioning that the reallocation of the stables from the south of the city to the new milking parlours in the neighbouring areas led to a big failure. The reason for this was obviously related to the adoption of the technified system which trapped the producers in the hands of middlemen thus reducing their incomes sustantially.

The apparent restrictions of physical space in the stable led us to think about the problems of animal welfare, a subject which has recently been incorporated within discussions of sustainability (Fraser and Broom 1990 ). Although some stables occupy a relatively large area most are of a medium size which compels producers to utilize the space more intensively. This situation suggests the need to understand more about animal comfort related to space availability. Preliminary studies carried out by our research team have shown a clear trend for cows to display more agressive behaviour (fighting, pushing and butting) in smaller areas compared with more friendly social behaviour in larger areas such as licking (Vieyra et al 1994). Female behaviour would be an interesting subject to deal with in the Iztapalapan stables as there is good evidence that space per animal in our area has a mean value of 10 m², contrasting with a recommended value of 40 m² (Nakanishi 1992) in order to guarantee normal behaviour.

Our objetives which were to increase the efficiency of the system have been difficult to achieve. The main limitation is the producer himself and his reluctance to introduce changes. Another factor is related to our lack of experience in dealing with a novel system that appears to be in harmony with its working environment. Although our interviews with the stablemen have shown them to be cooperative in describing the system, to date they are not prepared to introduce even small changes. Two extreme examples include: the recording of reproductive behaviour and/or daily production from the cows and the building of fermentation tanks for manure to reduce odours and obtain methane gas before residues are sent to the fields. The stablemens’ reticence is again related to the particular features of the system. In the first case the producers do not appear to need to record the behaviour because of the small size of the herd and the short time that the cows are in the stable. Again, with the unlawful condition of the system the producers have lost the sense of security as landowners which results in a negative attitude to change. This complicates the issue because the local authority is impelled to recognise the stables as unlawful, thus establishing a vicious circle. For this reason we are now attempting participative research with the producers in order to understand their decisions and thereaf ter any posible changes to increase the efficiency of the system.

The authors wish to thank the producers of milk from the stables in Iztapalapa for their collaboration by giving information and allowing us to measure some parameters of their systems; the authorities of the Universidad Autónoma Metropolitana ( Autonomous Metropolitan University) for the facilities given to do the research; Mr Mike Neale a British student sparing time with us for manuscript corrections and offering suggestions for its improve ment.

Acevedo A, Arellano A, Preciado A, Ruíz G, Losada H, Cortés J, Vieyra J, Grande D, Arias L, Zamudio A y López M 1995 Análisis microbiológico de leches broncas en los establos de Iztapalapa. Memorias del Segundo Congreso Internacional de Investigación en Sistemas de Producción Agropecuaria. Universidad Autónoma Metropolitana Iztapalapa y Universidad Autónoma del Estado de México. pp. 12

Received 1 July 1996