INTRODUCTION

For any country, agricultural companies play a very important role, since they are economic units that provide food for the population (^{García-Pérez, 2017}). In this sense, the agricultural company combines knowledge, information, natural, human and capital resources to produce goods of animal or vegetable origin; or services for a target market and within a profitable and sustainable operation (^{Guerra, 1998}). However, many of these agricultural companies are small or medium-sized and regardless of their productive profile. They are characterized by having limited natural resources and by constantly presenting a capital deficit to optimize production activities and little or no administration (^{Balestri et al., 2001}). Therefore, ^{Aguilar et al. (2016)}, point out that the improvement within these companies and regardless of the size of the production unit, will occur when a professional administration is started, since it is when the evolution of administrative theories to improve profitability will be considered of the companies.

On the other hand, the production of small ruminants is very important in livestock around the world, specifically dairy products of goat origin, which have an important social impact in extremely poor areas (^{Dubeuf, 2005}; ^{Escareño et al., 2012}; ^{Salinas-González et al., 2015}) ), and in certain regions they are of great interest because they are a sustainable and very profitable resource (^{Sanz et al., 2003}), since their production derives the production of food, they contribute to the employment and economic stability, generate additional value in the production units and facilitate the use of marginal lands (^{Escareño et al., 2011}); but they have the limitation that they are administered directly or indirectly by the producer (^{Cofré, 2001}) and most of the time, they lack experience in administration.

In Mexico, goat farming represents an activity that supports almost 1.5 million people and is associated with strata of the rural population with lower incomes, since it represents the livelihood for 80% of subsistence producers and takes place mainly in arid regions and semi-arid; which correspond to 60% of the national territory, where pasturelands with limited natural resources and areas with a high degree of marginalization abound (^{Guerrero-Cruz, 2010}; ^{SAGARPA, 2007}). Likewise, it is known that goat milk production represents an important source of employment for families in rural areas, since it offers job stability (^{Escareño et al., 2011}).

In this sense, Lagunera district, in northern Mexico, is considered the main goat milk production basin. However, it has been documented that although it is the main milk- producing region, there is still low productivity; mainly due to the lack of production records, no organization of producers for the production, processing and/or commercialization of milk and lack of information and technical assistance; which results in a low technological level so that, as long as it is not visualized as a company, it is difficult for goat activity to prosper as a business (^{Maldonado-Jáquez et al., 2018}) and will continue to be only a subsistence activity. However, different research efforts have been made, but these have not had a sufficient impact on the innovation of production systems (^{Salinas-González et al., 2016}), mainly due to the lack of technologies adapted to the conditions and characteristics of each producer.

Therefore, in the search for solutions to the problem presented by the producers of the Lagunera district, ^{Gómez (2007)}, the organization is suggested as an alternative to improve the profitability of the activity, since within a cooperating group and The organized organization would not be totally dependent on intermediaries, the price of milk could be negotiated and participatory processes would be optimized in which to find areas of opportunity for innovation and efficiency in the use of producer resources (^{Salinas-González et al., 2011}).

Therefore, the objective of this study was to characterize and group producers with similar socioeconomic and technological profiles in the northeast portion of the Lagunera district, in Coahuila state, Mexico; in order to design programs and strategic actions for research, innovation, technology transfer and/or training, according to their needs, This allows improving the production system, under the assumption that by transitioning to cooperative forms of production and commercialization, the adoption of technologies will increase and, consequently, the profitability of companies.

MATERIAL AND METHODS

The research was developed in the northeast portion of the Lagunera district, in Coahuila state, which includes the municipalities of Francisco I. Madero, San Pedro de las Colonias and Matamoros. The climate for the municipalities of Francisco I. Madero and San Pedro de las Colonias is a semi-warm dry subtype and the climate for the municipality of Matamoros is very dry and warm. The average annual temperature is between 20 and 24 ºC, with average annual rainfall between 200 and 400 mm (^{INAFED, 2010}).

Producers from the Cooperative for Productive Reconversion La Laguna, S.C. were taken as a case study by R.L. de C.V. recently formed (2016); which brings together 32 producers, after 14 common lands were visited and more than 300 goat farmers from the study area were invited.

The study consisted of two phases. In the first, the 32 producers belonging to the Productive Reconversion Cooperative were interviewed, which is dedicated to the production of goat milk in the aforementioned municipalities. This study was carried out in order to stratify and characterize the socioeconomic and productive profile of the producers. Data collection was carried out through direct interviews and visits to the production units. In this study, the methods of observation, direct interview and life histories were used to gather and document information on the production and marketing chain of goat milk (^{Hernández-Sampieri et al., 2006}).

The questionnaire was developed based on the review of specialized literature and diagnostic surveys, proposed by ^{SAGARPA (2015)}. The interview was organized in 13 blocks, where 44 ordinal and nominal variables were considered. The information obtained was captured in an Excel program matrix, and later the data were analyzed by means of a multiple correspondence analysis; which was carried out using the statistical packages InfoStat (^{Balzarini et al., 2008}) and SPSS, adapting the methodology described by ^{Bernal-García et al. (2003)}, ^{Carrasco-González (2004)} and ^{Castel et al. (2003)}, in the multivariate statistical analysis.

For the multiple correspondence analysis, hard variables of socioeconomic, production and marketing characteristics were chosen; which were considered as quality variables, according to previous studies carried out by ^{Salinas-González et al. (2016)}, ^{Escareño et al. (2011)} and ^{Castel et al. (2003)}. Among those that were considered: age, education of the producer, economic dependents and family members, time dedicated to the activity, size of the herd, hours dedicated to grazing, milk production and price, among others; which were taken as the basis in the descriptive analysis of Clusters by Hierarchical Grouping (HGA), Principal Components (PCA) and Multiple Correspondence (MCA).

For this case, this type of analysis was considered, since it decreases the covariance range given by the over-parameterization, when the number of characteristics is very large; it reduces the dimensionality of the correlated characteristics and generates a reduced number of variables. Subsequently, to perform the grouping, we start from a distance matrix, containing all the distances between pairs of objects and are grouped considering the minimum distance (^{Peña-Malavera et al., 2010}; ^{Agudelo-Gómez et al., 2015}).

The second phase of the study included participatory workshops, where the previously interviewed producers were presented with the results of the surveys applied in the first phase. The workshops were intended for producers to formulate the main problems and limitations of the production system, as well as proposals for solutions. For this, the qualitative study technique Focus Group was used (^{Santiago and Roussos, 2010}). This methodology is a data collection technique that uses a semi-structured group survey, directed towards a specific topic and of which the existing knowledge is inadequate or is very complex to make inferences without considering the user or interest group (^{Escobar and Bonilla-Jiménez, 2009}).

The results obtained were classified and grouped according to the highest frequency, and starting from here, a problem tree was elaborated to help identify the main problems and their solution, through root causes and effects (^{Hernández-Hernández and Garnica-González, 2015}); where in the strategic deployment, the main problem became the strategic objective and the root causes in the strategic actions to be implemented; this was done for the producers of each cluster found.

RESULTS AND DISCUSSION

Producer groups, even those organized as is the case of the recently formed Cooperative, are made up of individuals and production units of various characteristics. By grouping the members of the subjects of technological intervention, assertive strategies can be formulated, either for the management of innovation or the basic activities of technology transfer and training. The results found show sufficient differences between subgroups (conglomerates), for which specific strategies are presented to be implemented within the cooperative, according to the subgroup of participants. Likewise, the design and implementation of certain strategic actions within the development plan of the Cooperative are discussed.

Clusters by Hierarchical Grouping

Figure 1 shows the formation of the clusters (groups), as well as the distances between them. The analysis located the producers that contained some type of similarity in clusters, synthesizing and taking advantage of the available information, and resulting in a set of information grouped according to their hierarchical level and in which 5 groups of clusters were defined.

Figure 1 Dendogram of the hierarchical cluster analysis by the Ward method 

Multiple Correspondence Analysis

In figure 2, the interactions between the PCA and HGA results are shown; PC 1 and PC 2 were taken as factors, since they are the ones with the highest variance in the collection of the data set and variables (0.2851 and 0.1804 respectively; Table 2). Here it can see the results of the profile for the producers of each group and considering that a very low percentage of the producers have received some type of training, this will serve to establish the guidelines to follow for the implementation of technology transfer schemes, as well as well as innovation and research in each group.

Figure 2 Set of PC 1 and PC 2 where they are integrated as factors. NH = Number of children; NC= Number of goats; HS = Hectares sown; LC = liters of goat milk-1 day-1; EDUC = Schooling; G1, G2, G3, G4 and G5 = Groups resulting from the HGA analysis 

Regarding the interactions observed in figure 2, the producers of group 1 are characterized by having the highest milk production in liters per day, total area and sown area; which shows an area of opportunity in the improvement of productive aspects, but also a considerable window of opportunity in terms of socioeconomic aspects.

On the other hand, group 2 producers have the highest educational level, which can be considered as an opportunity area for a greater incorporation of technologies in technical-productive aspects. The above coincides with the report by ^{Hundal et al. (2016)}, where they indicate that producers with a certain degree of training improve their knowledge of the activity significantly faster than other producers. Group 3 producers present a window of opportunity in productive aspects, in addition to being one of the oldest groups and obtaining the highest price per liter of milk. Therefore, in this group it would be relatively easy to apply a technology transfer program, since they lack training in all the links of the production chain. In this regard, ^{Chipasha et al. (2017)}, group producers, that have few production skills, little access to the market and infrastructure and a high incidence of diseases, and point out them as the group with the greatest production limitations and perfectly defined areas for improvement.

Regarding the producers of group 4, these are characterized by being of advanced age, having the largest number of dependents, more experience in goat activity and greater roots in the locality. On the other hand, the producers that comprise group 5 have a greater number of children, a greater number of goats and male goats, with more hours devoted to grazing; so they are mostly related to productive aspects, but the number of children is also included; which can be considered as an advantage to establish some technology transfer scheme, since they can have access to a greater amount of labor, compared to the rest of the groups.

Strategies for the development of producers

Goat farming is an important option for the maintenance of low-income rural producers, and it is the main livestock economic activity in the desert and semi-desert areas of Mexico (^{Alva-Pérez et al., 218}); Therefore, Table 5 shows the strategies suggested to promote the development of producers, which are proposed based on the results obtained in the YMCA (Figure 1), the MCA (Figure 2), and the analysis of cause- effect of the problem tree (Table 4). For each group or conglomerate of producers within the Cooperative, a strategic objective and actions to be implemented are proposed to solve the main problem of each conglomerate (table 5). Likewise, each objective and strategic action was designed according to the perspective and variables to be improved according to the specificities of the group of producers in each conglomerate. Therefore, within the strategic plan, innovation, training and / or technology transfer actions were defined; which are the product of the demand of the producers themselves and are complemented by previous experiences reported in the scientific literature.

This information coincides with the existing literature for small producers in Mexico. In this regard, ^{Cárdenas-Bejarano et al. (2016}), point out that the rate of adoption of technologies increases as producers organize; since there is a better flow of information within the groups and they indicate that to better understand the technology adoption processes, it is necessary to consider the analysis of the structure of the group to which they belong; as well as the existing interrelationships without forgetting the socioeconomic context. On the other hand, some other authors refer some suggestions and/or recommendations to solve problems similar to those found in the groups of producers in this study; For example, there are reports (^{Tanwar et al., 2011}; ^{Fonseka et al.,2018b)}, which suggest developing food conservation schemes and conservation of outstanding genetic material through selection, as well as the interaction between the different actors of the production chain to facilitate the flow of market information and the development of infrastructure that improves the marketing system for goat milk (^{Chipasha et al.,2017}); o the training of members and organization to improve the productivity and health of the production units (^{Hundal et al., 2016}; ^{Kumar et al., 2015}; ^{Sandhu et al., 2018}); as well as access to credit and extension services as triggers for the success of companies dedicated to the production of goats (^{Ifeanyichukwe et al., 2018}). Furthermore, if one considers what ^{Barrera-Perales et al. (2018)}), some aspects that play in favor of the Mexican semi-arid producers, is that the use of exclusively family labor, low technological level, limited infrastructure and dependence on the pasture; They are factors that contribute to the profitability and continuity of the production units, since large investments are not required to operate, and this in turn helps to alleviate the limiting effects of the current production system.

CONCLUSIONS

The results found suggest that rigorously delimiting and discriminating producers according to their similarities, as well as the identification of limitations in the goat production system, from the producers' point of view, allows them to be stratified and grouped so that they can be clearly identified the critical points of attention for the establishment of research, innovation and technology transfer schemes, according to their reality. This will allow in the immediate future that the definition of strategies and public policies to strengthen goat production systems, both intensive and extensive in the arid and semi-arid zones of Mexico, have an efficient impact; and goat farming in the north of our country is consolidated as an activity that generates jobs in marginal areas.