Study area

The study area was the Nevado de Toluca Flora and Fauna Protection Area, in the State of Mexico, Mexico. This NPA has elevations ranging from 3,000 to 4,660 m asl. Climate is sub-humid semi-cold (CEh), with average annual temperatures ranging from -2 to 7 °C, and annual precipitation from 1,000 to 1,400 mm. Overall NPA area is 53,987 ha, divided among 61 agricultural nuclei. Within the NPA are twenty towns and a total of approximately 10,000 inhabitants¹¹. These communities have about 191,000 head of livestock of which 60 % are sheep. Grazing is done in forest and grasslands inside communal use areas¹¹. Originally a national park, NT was reclassified as a Flora and Fauna Protection Area in 2013, a change which allows people living in the NPA to maintain ownership of their land and continue some productive activities, without changing land use¹².

Producer identification and data collection

During 2015, a total of 162 questionnaires were applied to sheep producers who had been chosen by convenience sampling. This is the same method used in previous studies, although it does not allow building of a reliable sampling framework¹³. The questionnaire had three sections: (i) livestock, which recorded data on number of animals, feeding and grazing, reproduction, health and technical practices; (ii) agricultural activity, which recorded the number of crops, cultivated area and machinery; and (iii) socioeconomic characteristics, which covered head of household age and education level, income sources, family participation in labor, experience and training. Producers had no records on technical and economic information, which is common in family production systems^7,14,15. As a result, the collected data was supplemented by direct observation during visits to production units, as well as forty in-depth interviews with sheep producers held during grazing periods.

Statistical analysis

A principal component analysis (PCA) was done using 25 variables to provide an initial approach to the variables describing SPS in the study area. The PCA produced small groups of linear combinations (components or factors) which explained as much variance as possible in the original data with minimal data loss¹⁶. Parsimony of the principal components (PC) was verified with the Kaiser-Meyer-Olkin (KMO) test, and sample suitability confirmed with the Bartlett test of sphericity¹⁶. Orthogonal varimax rotation was applied to the PCs to improve interpretation^6,16.

Linear PCA combinations were introduced into the hierarchical cluster analysis (CA) to form groups of producers and characterize the SPS. Case clustering was done following Ward’s method, and the squared Euclidean distance used as a measure of similarity. A dendrogram analysis and cluster coefficient were applied to identify the number of groups⁶. Because group size was not homogeneous, differences between groups were identified with an analysis of variance (quantitative variables) by comparing Hochberg means¹⁷. Categorical variables were analyzed with contingency tables and a χ² test⁶. Statistical analysis results were triangulated with field observations and interviews. All statistical analyses were run with the SPSS ver. 22.0 program.

Sheep production system in the Nevado de Toluca

The SPS in the NT is a low-tech family-run system closely linked to agriculture, and which provides financial security, much like a previously described SPS¹⁸. Most of the sheep herds were fed by grazing, and supplemented in stables. Breeds were mainly Suffolk, Hampshire and crosses thereof, chosen for weight and ease of handling. Reproduction is continuous although females exhibit marked seasonality. Mating occurs most frequently in the summer with births between November and February. This is the coldest time of the year, which may contribute to high lamb mortality due to respiratory diseases.

Factors characterizing SPS

The PCA identified six principal components (PC) which explained 71% of the variance. Results for the KMO test (0.61) and Bartlett’s test of sphericity (P<0.001) confirmed analysis trustworthiness. Sixty percent (60%) of the analyzed variables were retained in the PC (Table 1), which coincides with previous reports of a 64% retention of variables^7,13. This suggests that characterizing these production systems may only require from 10 to 20 variables, and that production unit size (i.e. number of hectares and animals) has the highest relevance in classification¹⁹. Variables with communality values less than 0.5 were excluded from the PCA because this indicates they had low associations with the selected PC^16,17.

Principal component one (PC 1, capitalization) incorporated variables relating to system assets (Table 1), which determine the investment capacity in technical practices²⁰. The second one (PC 2, human capital) showed the inverse relationship between head of household education level, age and experience. This reflects these adults’ limited access to education, due in part to the lack of formal employment opportunities and nonexistent infrastructure for access to educational services. These are associated with the nature of the NPA and led these adults to enter productive activities at an early age²¹.

The third component (PC 3, animal health) highlighted adult animal mortality (11%) and lamb mortality (23%), both of which were higher than values reported for other SPSs²². Animal health management is therefore an area of opportunity for improving system productive and financial efficiency. The fourth component (PC 4, agriculture) associated crop diversity with family participation and use of machinery. This is characteristic of mixed production systems, in which exploitation of agricultural and livestock resources depends on working as a family¹⁹.

The fifth component (PC 5, rotational grazing) linked forest grazing with the distance traveled to grazing. In this technique animals are allowed to graze freely on a surface for short periods (about 10 min) along relatively long routes (2 to 4 km), much like the grazing circuits used in the French Mediterranean⁹. Finally, PC 6 (technical practices) associated deparasitization with weaning, which occurs at time of sale and without prior weight gain regimes that could improve producer income¹⁴. Deparasitization was implemented in 70 % of the studied herds, a rate higher than the 58% reported for other SPSs²³.

Sheep production groups in the NT

The cluster analysis identified three producer groups, differing mainly in terms of quantity of animals, cultivated agricultural area and income from sheep and agricultural. As has been done in other studies^8,24, these differences were used to classify the groups as small producers (Group 1), intermediate producers (Group 2) and capitalized producers (Group 3). Following are descriptions of the productive, socioeconomic and grazing resources use aspects of each producer group.

Sheep production

Data on the SPS show small producers (28 %) had the fewest animals, and weaned and deparasitized at lower rates, probably due to their lower training levels (Table 2). Intermediate producers (35 %) had an intermediate number of animals, but the highest percentages of deparasitization and weaning. Capitalized producers (6 %) had the largest number of animals with herds about 120 % larger than in other regions of Mexico^18,22, but 20 % smaller than herds in Europe^3,8. Despite the differences in capitalization, parturition and lamb mortality rates did not differ (P>0.05) between the three groups. Clearly in this SPS the resource base is not reflected in productivity¹⁹, probably due to lack of inadequate training.

Sheep product marketing data showed that the small and intermediate producers sold some lambs when young (4 to 5 months of age) and the remainder throughout the rest of the year, and less than half discarded unproductive animals (Table 2). This confirms financial security as one of the main functions of livestock in this system since animals are sold in response to financial need^18,23. In contrast, the larger, capitalized producers sold more lambs and waste animals because their facilities and limited labor did not allow increases in herd size. This finding suggests that herd size self-regulates and that animal load therefore remains stable, leading to improved herd productivity without increased size¹⁵, largely through use of reproductive management techniques and health treatments. Sale of wool was almost null since only about 20 % of producers sheared their sheep and these just discarded the wool due “to its low price” ($1.00/kg); the market therefore limits development of this system¹⁴.

Agricultural production

Small producers had less agricultural surface, fewer crops and less machinery. Most agricultural production was used for subsistence, and 12 % of these producers owned no farmland. Intermediate producers had greater crop diversity and more machinery use, although this was not owned by them. They tended to sow higher percentages of forage oats (Avena sativa), fava bean (Vicia faba), pea (Pisum sativum) and common bean (Phaseolus vulgaris) (P>0.05). While they sold a higher percentage of potato (Solanum tuberosum) than the small producers, a third sowed small amounts (<0.3 ha) for family use (Table 3). The capitalized producers had more cultivated area and machinery use, grew potato (S. tuberosum) on 18 % of their land and maize (Zea mays) on 26 %; in contrast, the small and intermediate producers planted S. tuberosum on only 9% of their land and Z. mays on 41 %. The proportions of crops and their commercial purpose depended on producer capitalization level. In general, all three groups grew oats on 50 % of their land. They allocated Z. mays for subsistence, and fodders and crop waste to feed horses and ruminants. These in turn provided fertilizer for crops in a complementary management system between agriculture and livestock production like that reported in other SPSs^18,22. This highlights that NT sheep producers have traditionally developed a comprehensive resource use strategy.

Socioeconomic aspects

The families (5 members) of small producers participated less in agricultural activities because at least one member took non-agricultural jobs outside the NPA, consequently reducing dependence on natural resources². Intermediate producers had larger families (6 members) and these were responsible for a larger share of agricultural activities. However, they also had more income sources since different members could service the livestock after other activities (e.g. housework, jobs, or school) (Table 4). The present data support engagement in multiple activities as a strategy for increasing income and ensuring the flow of financial resources to the household¹⁰. The capitalized producers were younger, had finished elementary school and their families (5 members) covered all livestock care needs. These results coincide with other studies done in rural Mexico²⁵, but contrast with results for European countries where 30% of producers have a high school or university education²⁰ and family members contribute from 33 to 74% of labor^8,13.

Income

Incomes from agriculture and livestock did not differ (P>0.05) between the small and intermediate producers (Table 4). The ANOVA detected no differences due to high intragroup variability. This phenomenon has been observed in studies including socioeconomic aspects^6,13. In addition, income flow from sheep was not continuous because producers kept animals for sale later when they needed money, leading to productive cycles exhibiting various high and low sales periods; for example, only 16% of the studied households sold animals during the study period. Income from sheep was highest among the capitalized producers, although sale of sheep provided only a third of overall household income. The NT SPS is clearly a complement to family income^14,24, in addition to employment outside the NPA, remittances and government subsidies.

Use of grazing resources

Most (97%) of the surveyed producers grazed their herds in the grazing-only or semi-stabled modalities (grazing-stabled). Of these, 58% grazed in the forest, although this proportion increased among the capitalized producers, who used grazing-only to avoid raising costs from feed purchases²². The small and intermediate producers used semi-stabled (Table 5), because their herds required less feed volume, allowing them to reduce grazing and channel their labor into higher-income activities².

Forest grazing circuits were frequently changed and differed between producers, who modified their routes year-round based on their perception of vegetation availability^5,9. This suggests that pressure on grazing resources is regulated by apparent fodder availability, previous grazing cycles and in situ agroecological characteristics. Management of this SPS may therefore be compatible with conservation efforts if animal load is adjusted by developing methodologies appropriate to rotational grazing and adequate knowledge is shared occurs between producers.

Sheep herds are grazed in the NT forest mainly in the dry season (March to May) when pasture quantity and quality decreases. During the rainy season (June to October) the preference is for grasslands containing the genera Vulpia, Nassella, Trisetum, Muhlenbergia, Potentilla²⁶, and forest grazing decreases. After local crops are harvested (November to February) field stubble and roadsides are grazed. Producers in this SPS are clearly adapting to year-round resource availability^22,26. Their experience in grazing area use could be incorporated into zoning plans within the new Flora and Fauna Protection Area designation, essentially integrating local knowledge into management plan design²⁷.

Feeding

Most producers (88%) provided feed concentrate (commercial and empirically-processed homemade mixtures) to their herd, confirming that traditional grazing management is transforming into a semi-stabled system, with day-grazing followed by stabling and supplementary feed in the evening. This arises from producer interest in intensifying production^1,5 and adaptation strategies responding to restrictions on natural resource use⁴. Some producers (22 %) were intensifying their strategy by stabling weaned lambs to promote weight gain and consequently higher sales prices (Table 5). However, producers need to be careful that intensification does not undermine profitability due to the need for input purchase³.

The technological transition observed in the NT has been reported for different production systems^5,7,14. It also suggests that use of grazing resources in NPA should be incorporated into more complex feeding strategies that include commercial feed and cultivated fodder. Grazing-based systems that adequately integrate their resources can be financially efficient and environmentally friendly^3,15, since grazing can contribute to maintaining biodiversity¹⁰ and avoids accumulation of combustible material^13,28.

The main feeds used in the study area were hayed oats and corn stover (Table 5); both are low cost because they are by-products of crops grown by producers. A very few producers fed their sheep mixtures of wheat bran or soybean (13 %) or potato waste (2 %), although use of this resource has not been documented. Feed management practices in the NT are similar to those used in other SPS^22,23. They also approximate traditional agrosilvopastoral management in which resources are used in an integrated manner, although different plant strata do not necessarily share the same space²⁷. This differs from agroecological designs in which trees are combined with different vegetation strata to provide animal feed²⁸.