The research drew on a survey conducted between April and June 2025 as part of an INIA project characterizing cattle, sheep, and camelid production systems in Ayacucho. Using a finite population formula with a 95 percent confidence level and a 10 percent margin of error, the team interviewed 96 cattle breeders, 96 sheep breeders, and 94 camelid breeders across the districts of Chuschi, Sancos, Sacsamarca, and Chipao. Of the 286 interviews collected, 175 were retained for analysis after applying two exclusion criteria: farms had to raise cattle, sheep, alpacas, and llamas together, and interviews with incomplete or irrelevant data were removed. The questionnaires covered four broad domains: sociodemographic information such as age, sex, housing, and language; socioeconomic factors including income sources and land availability; productive aspects spanning feeding, reproduction, genetics, milking, shearing, and animal health; and marketing details covering prices, supply frequency, and sales channels.

The statistical strategy combined descriptive analysis with two multivariate techniques. After summarizing the data in frequency and percentage tables using R software, the researchers selected ten categorical variables—ranging from housing material and power source to land size, irrigation, grazing system, pasture type, breeding stock origin, and veterinary care—and subjected them to multiple correspondence analysis, or MCA. This unsupervised learning method maps associations among categories in multidimensional space, so that categories appearing close together on the map are directly related. The team checked for multicollinearity using Cramer’s V coefficient and retained the first three dimensions, which together explained 20.60 percent of the total variance and exceeded Greenacre’s average inertia criterion. The MCA coordinates were then clustered with hierarchical cluster analysis using Euclidean distance and Ward’s aggregation criterion, with the number of groups determined by the elbow method and the clusters consolidated through the K-means algorithm in hierarchical principal component classification.

The descriptive results paint a vivid picture of livelihoods built almost entirely on animals. For 93.71 percent of breeders, livestock was the primary source of income: cattle supplied milk and dairy products, sheep were kept mainly for lamb sales, alpacas provided both fiber and meat, and llamas were raised predominantly for meat, with 92.29 percent of llama keepers citing meat as the main product. Cattle were raised primarily for dairy production on 85.11 percent of farms, especially cheese and quesillos, while sheep were sold live on 74.14 percent of farms. Men dominated livestock decision-making, and most breeders had been raising animals for more than 21 years. The most frequently cited social problem was lack of government support, reported by 64.29 percent of respondents, followed by migration at 31.55 percent.

Feeding and grazing practices revealed both tradition and constraint. Continuous grazing was the most common system, used on 57.71 percent of farms, particularly for sheep and camelids, while rotational grazing appeared on 26.86 percent and was more often applied to cattle. Natural pasture was the primary feed resource, with native grasslands—including bofedales, canllares, juncaceas, and tolares—mainly grazed by sheep and camelids, while cultivated pastures such as alfalfa, red clover, and ryegrass were mostly reserved for cattle. Despite the dominance of pasture-based feeding, 90.23 percent of breeders supplemented their livestock, mainly with oat hay, minerals, and vitamins during the dry season. Water scarcity loomed large: 55.17 percent of farms had no access to irrigation water, and among those that did, sources included irrigation ditches, rivers, and reservoirs.

Genetic management showed a striking pattern of unstructured breed introduction. Exotic breeds had been incorporated into 71.05 percent of cattle herds over the past five years, predominantly Brown Swiss, and into 61.97 percent of sheep flocks, with Merino favored in Chuschi and Corriedale in Sancos. Natural mating remained the predominant breeding method, although 30.77 percent of cattle breeders used artificial insemination, largely supplied through regional government projects. In sheep and camelids, reproductive biotechnologies were rarely used, a reluctance the authors attribute to traditional beliefs and past experiences with low fertilization rates. Notably, the researchers observed that exotic breeds were being introduced without any structured genetic improvement plan, a practice they warn could undermine the long-term competitiveness of the sector and risk erosion of locally adapted Creole genetics.

Animal health emerged as one of the most pressing weaknesses. Inadequate veterinary assistance affected 79.43 percent of breeders, and in 76.57 percent of households the owners themselves handled health problems, often relying on professional advice or personal experience when assembling veterinary kits. Pneumonia was the most common disease in cattle at 41.67 percent, followed by digestive disorders such as colibacillosis, coccidiosis, and clostridiosis in calves. Sheep suffered most from pneumonia and dystomatosis, while alpaca crias were hit by colibacillosis and pneumonia. Disease timing tracked the seasons: respiratory illness and alpaca fever clustered in the cold, dry months from July to September, whereas digestive disorders peaked during the January-to-March rains. Vaccination coverage was thin—only 38.46 percent of breeders vaccinated cattle against symptomatic anthrax, and sheep and camelids were generally unvaccinated.

The multivariate analysis distilled this heterogeneity into three farm types. Group 1, accounting for 34.29 percent of farms, comprises breeders on small and large non-irrigated landholdings who live in stone houses without electricity or potable water, depend exclusively on natural rangelands, graze their animals continuously without structured management, and handle veterinary care themselves. Group 2, at 24.00 percent, consists of breeders with extensive land areas exceeding 10 hectares, mixtures of natural and cultivated pastures, gravity-fed irrigation where water is available, rotational and seasonal grazing, and breeding males sourced from other regions of the country. Group 3, the largest at 41.71 percent, includes breeders with medium-sized holdings of 3 to 10 hectares, adobe houses with electricity and drinking water, cultivated pastures under sprinkler or gravity irrigation, rotational grazing, and a greater propensity to seek professional veterinary support.

The authors argue that this typology is more than an academic exercise. Because livestock contributes roughly 40 percent of the gross value of agricultural production in Peru, and because Peru is the world’s leading producer of alpacas with an estimated 4,545,820 animals and the second-largest producer of llamas, understanding the structure of these systems is essential for policy. Ayacucho alone holds 7 percent of the national alpaca and llama population. The three groups differ sharply in their needs: the first requires urgent technical support in reproductive and farm management, the second holds substantial potential for expanding commercial production, and the third shows the greatest capacity for competitive, market-oriented farming. Tailoring interventions to these profiles, the researchers contend, is far more promising than the generalized livestock development plans that have historically overlooked such heterogeneity.

The study also documents structural barriers that cut across all three groups. More than 62 percent of breeders reported receiving no technical assistance or training at all, with the regional government’s productive projects serving as the main channel for technology transfer. Marketing remained dominated by intermediaries, who capture the largest share of profits along short production chains, and breeders avoided banks because of high interest rates. Milking and shearing were performed manually throughout the study area, with most alpacas yielding less than five pounds of fleece per animal. The authors recommend longitudinal studies that track generational turnover and the full value chain, and they suggest that the breeder classification developed here can serve as a pilot tool for local governments seeking to design targeted, sustainable interventions for the mixed livestock systems that sustain thousands of high Andean families.

Subject of Research: Characterization and typology of mixed livestock production systems among small-scale breeders in the high Andean region of Ayacucho, Peru

Article Title: Characterization of mixed livestock systems applied by small-scale breeders in the high Andean region of Ayacucho, Peru

Article References: Palomino-Guerrera, W., Guzman, H., Contreras, M., & Trillo, F. (2026). Characterization of mixed livestock systems applied by small-scale breeders in the high Andean region of Ayacucho, Peru. Pastoralism: Research, Policy and Practice, 16, Article 16335. https://doi.org/10.3389/past.2026.16335

Image Credits: AI Generated

DOI: 10.3389/past.2026.16335

Keywords: mixed livestock systems, Ayacucho, Peru, high Andean livestock, alpacas, llamas, small-scale breeders, multiple correspondence analysis, hierarchical cluster analysis, farm typology, pastoralism, livestock policy