This research was conducted on different herds from the peasant community of Pilluni in the district of Cotaruse, province of Aymaraes, Apurímac region of Peru, located at coordinates 14°18′04″S and 73°13′52″W, with elevations between 3,900 and 5,400 m above sea level. As described by Pinares et al. (2023b), the geography is characterized by dry puna. The plateau of the area shows a slight topographic slope, with scarce water availability during the dry season, limited to springs, a few wetlands, and small lagoons. According to SENAMHI (2020), the climate is generally cold and dry, with temperatures ranging from –5°C to 10°C at night and from 11°C to 24°C during the day. During the rainy season, from November to March, rainfall increases, and occasional snowfall may occur. The annual rainfall is approximately 720 mm. Llamas are raised through extensive grazing alongside plant species such as Alchemilla pinnata, Trifolium amabile, Stipa braciphylla, Calamagrostis curvula, Calamagrostis sp., Calamagrostis vicunarum, Poa gilgiana, Paspalum pigmaeum, Bromus sp., Carex sp, Festuca huamachusensis, Poa sp., Stipa mucronata, Disticha muscoides, Muhlenbergia fastigiata y Hipochaeris taraxacoides (Pinares et al., 2023b).

We applied non-probabilistic sampling by convenience. Eighty clinically healthy Q’ara llamas with no observable physical defects were sampled from the Pilluni farming community, based on the availability and consent of 11 different owners, each with herds of approximately 25 animals. Age categories were assigned by examining the type and stage of dentition: milk teeth, two teeth, four teeth, and full mouth, as described by Pinares et al. (2023a). Fiber and skin samples were collected from the shoulder, mid-side, and rump, respectively, according to the recommendations of Aylan-Parker and McGregor (2002) and Sasahara et al. (2022). A total of 240 fiber samples were collected from 80 llamas, including 40 males and 40 females of various ages and body sites ( Figures 1A, B ).

The removal process consisted of manual dehairing of the coarse fibers from 240 samples using the tactile–visual method. This was performed on a flat surface by placing a glass plate with a black contrast background to identify white bristles and a white background for colored bristles, according to the technique described by Frank et al. (2011) and Quispe et al. (2015). Once all coarse fibers were removed from the fleece ( Figure 1C ), and using a magnifying glass (Ove brand, model 99437, China), the resulting tuft consisted only of fibers from the inner coat (“down”).

Measurements were conducted on 240 dehaired fiber (down) samples and 240 non-dehaired fiber samples. The textile properties assessed included mean fiber diameter (MFD, µm), standard deviation of MFD (SD MFD, µm), coefficient of variation of MFD (CV MFD, %), comfort factor (CF, %), and fiber curvature (CU, °/mm). These measurements were performed using an optical fiber diameter analyzer (BSC Electronics, model OFDA 2000, Australia) according to IWTO-47 procedures (IWTO-47, 2007) at the South American Camelid Research Center (CICAS La Raya), Fiber Laboratory of the National University of San Antonio Abad del Cusco. The fiber samples were transported and stored under ambient environmental conditions and measured under standard laboratory conditions of 20 ± 2°C and 65 ± 5% relative humidity.

The limited number of skin histology samples was due to the owners’ acceptance of the invasive skin biopsy method. Therefore, only 195 samples were obtained, with three subsamples per animal from the shoulder, mid-side, and rump body sites, as well as from various age groups: milk teeth, two teeth, four teeth, and full mouth. These samples were collected from 65 llamas (32 males and 33 females) out of the 80.

To characterize the primary follicle density, secondary follicle density, total follicle density, and the secondary-to-primary follicle ratio, skin was shaved using disposable stainless steel blades ( Figure 1D ), followed by local anesthesia (2% lidocaine injected subcutaneously at the margins of the sampling sites). After allowing approximately 5 min for the lidocaine to take effect, the skin was stretched using the thumb and forefinger.

A specially designed 8 mm diameter biopsy punch (KaiMedical, Japan) was used to obtain skin biopsies, according to the procedures recommended by Frank et al. (1989, 2006), Sasahara et al. (2022), and Jones (2023). The biopsies were wrapped in sterile gauze together with Canson tracing paper labeled with each animal’s identification. The three subsamples from each animal were carefully placed together in a single jar containing 10% formaldehyde solution (Sasahara et al., 2022). The samples were subsequently transported in rigid technopor boxes at refrigeration temperature to avoid direct exposure to light for at least 72 h after sampling. Finally, the samples were stored and processed under standard conditions in the Histology Laboratory at the National University of Altiplano in the Puno region of Peru.

For the identification of primary and secondary follicles per mm², the skin samples were processed using the standard technique of embedding in solid paraffin, and 5–7 μm sections were collected on slides and stained with Masson trichrome (Lusky et al., 2006; Sasahara et al., 2022). The image sections were captured using a digital camera connected to a light microscope with LAS EZ 3.4 DVD 272 software (Leica Microsystems brand, DM750 model, Germany). Follicle characteristics were observed at 10X magnification with a resolution of 1920 × 1080 pixels. The most appropriate area for follicle density counting was selected, and the primary and secondary follicles were identified and carefully recorded. These images were stored on a computer and then multiplied by 0.84 (Antonini et al., 2004), the correction coefficient calculated after contraction of the skin sample due to fixation and dehydration procedures (Lusky et al., 2006).

The normality of the data was verified using the Shapiro–Wilk normality test. Since some response variables of fiber quality and follicle parameters did not conform to normality, arcsine transformations were applied. Analysis of variance (Type I ANOVA, without assuming equal variances) was performed separately for each fixed factor and its levels: age (milk teeth, two teeth, four teeth, and full mouth) and body site (shoulder, mid-side, and rump). A Tukey test (α = 0.05) was used for multiple comparisons of means. In addition, fiber quality between dehaired and non-dehaired samples, as well as the effect of sex (female or male), was compared using paired and independent Student’s t-tests (α = 0.05), respectively. The nonparametric Spearman rank correlation coefficient was calculated for all fiber quality variables and follicle parameters without transformation (α = 0.05). Statistical analyses were performed using R software, version 4.2.3 (R Core Team, 2023).

This section provides a brief description of the main textile characteristics of raw Q’ara llama fiber and the quality of the dehaired fiber after the manual dehairing process. Age category was found to influence the quality of both non-dehaired and dehaired fiber, while sampling site had a greater influence on the homogeneity of fiber diameter ( Table 1 ).

Dehairing contributed to absolute differences in each fiber characteristic of Q’ara llama fiber. MFD, along with its SD and CV, decreased by –1.91 µm, 1.04 µm, and 3.05%, respectively, while CF and CU increased by 4.79% and 3.49°/mm, respectively (p ≤ 0.01). Manual dehairing improved the fiber quality of Q’ara llamas raised in the Apurímac region by reducing fiber fineness and homogeneity and increasing both comfort factor and fiber curvature ( Table 2 ). Therefore, it is recommended that this type of fiber be marketed only after dehairing to ensure quality-based payment.

Table 3 shows the mean values of primary follicle density (2.33 ± 0.04), secondary follicle density (14.49 ± 0.32), total follicle density (16.80 ± 0.36), and the secondary-to-primary follicle ratio (6.28 ± 0.08). Primary follicle density varies by age category, body site, and sex. However, secondary follicle density is affected only by age (p < 0.05). All follicle parameters were significantly affected by age category (p < 0.001).

This illustration shows hair follicles forming either as a compound follicle group (CFG) or a simple follicle group (SFG). The SFG consists of secondary follicles, as seen in young llamas ( Figure 2A ). The CFG consists of primary follicles surrounded by secondary follicles, as seen in adult llamas, especially older ones ( Figures 2B–D ). A decreasing trend in follicle density is observed, which becomes more pronounced with age.

All fiber quality characteristics were correlated with follicle parameters in Q’ara llama (p < 0.05), except for fiber curvature and the secondary-to-primary follicle ratio ( Table 4 ). There was a moderate correlation (0.52) between total follicle density and the secondary-to-primary follicle ratio. Mean fiber diameter was negatively correlated with fiber curvature, total follicle density, and the secondary-to-primary follicle ratio. However, comfort factor was moderately correlated with fiber curvature (0.49), and slightly correlated with both total follicle density and the secondary-to-primary follicle ratio.

The MFD in non-dehaired fiber was 25.95 µm, influenced only by age, while sex and sampling site were not significant, similar to the reports of Pinares et al. (2014); Quispe-Ccasa et al. (2020), and Apaza et al. (2022). Our result of 24.04 µm for non-dehaired fiber is physiologically and nutritionally consistent with the 25.83 µm reported by Celis and Panez (2025) for Q’ara llamas from Cerro de Pasco and the 25.47 µm reported by Quispe-Ccasa et al. (2020) for Ch’aku llamas from CICAS La Raya of Cusco. It is also consistent with the 24.72 µm reported by Poma and Ayala (2022) in Bolivian Q’ara llamas. However, Paucar (2021) reported a value of 27.93 µm in Q’ara llamas from Quimsachata in Puno, Peru. This variation could be explained by population and nutritional management, especially selection criteria for fiber or meat.

The comfort factor in non-dehaired fiber was 81.51% ( Table 2 ), similar to the 84.88% reported by Poma and Ayala (2022) and higher than the 70.62% reported by Paucar (2021), with no significant difference by sex. However, comfort decreased by 15.02% from the milk teeth stage (89.68%) to the full mouth stage (74.66%). Our result is similar to the decreasing trend of 79.76% in llamas >5 years old found by Quispe (2016) and Quispe-Ccasa et al. (2020). The standard deviation and coefficient of variation of MFD in raw fiber did not show significant differences by sex, but age had an important effect, as these characteristics decreased with increasing age.

Dehairing leads to structural changes in the textile raw material, as the longer and straighter (objectionable) fibers are eliminated. This directly affects spinning and dyeing processes. Greater fiber diameter variation has been observed in double-coated fleeces (Coates and Ayerza, 2004; Frank et al., 2012; Apaza et al., 2022; Quispe et al., 2023; Hick et al., 2024). Similarly, MFD decreases between 2.50 and 4.21 µm (Pinares et al., 2014; Paucar, 2021; Poma and Ayala, 2022), and CV MFD decreases between 2.51% and 9.97% (Pinares et al., 2014; Paucar, 2021; Quispe et al., 2023). However, in dehaired fiber, CF increases between 6.64% and 16.13%, and CU increases between 4.75 and 5.48°/mm (Pinares et al., 2014; Paucar, 2021; Poma and Ayala, 2022). This may be due to a guard hair yield of only 16 ± 7.4% in Q’ara llamas (Quispe et al., 2023), although all reported values are slightly higher than those observed in our study. Selecting for improved fiber quality would likely be a more effective approach than relying solely on dehairing.

Fine fiber is economically important due to its value in the textile industry; however, the total economic value also depends on follicle population density. Higher density results in greater individual fleece weight, allowing breeders to earn more income from fiber sales (Pallotti et al., 2020). In this regard, developing sustainable fiber production systems using llamas is a viable option in the puna. The greater ability of llamas to adapt to very poor soils and their larger body size provide them a significant advantage over alpacas. Their fine mean fiber diameter (less than 23 μm, if guard hairs are excluded) allows their fleeces to be sold at a premium price (Coates and Ayerza, 2004; Stemmer et al., 2005). Switching to llama production would increase farmers’ incomes and enhance the sustainability of agriculture in the Apurímac region of Peru. Baby and young llamas produce valuable dehaired fiber at first shearing, which sells at high prices, making llama production more profitable. Producers could raise llamas for both meat and fiber to boost their income (Coates and Ayerza, 2004; Antonini, 2010).

The mean TFD was 16.80 follicles per mm². However, Frank et al. (2001) reported a value of 24.36 in double-coated Argentinean llamas. The primary follicle density (PFD) found in this study was 2.33 follicles per mm², comparable to the values of 4.25, 3.77, and 2.11 reported by Frank et al. (2001, 2006) and Curasma and Rodriguez (2016) in Q’ara llamas. Secondary follicle density was 14.49 follicles per mm², consistent with the values of 14.24 and 16.92 reported by Curasma and Rodriguez (2016) and Frank et al. (2006), respectively.

Skin follicle parameters were not influenced by sex or sampling site, except for primary follicle density ( Table 3 ). The effect of sex was not taken into account in the statistical analysis by Antonini et al. (2004) and Pallotti et al. (2020). However, all skin follicle parameters decreased in older llamas. TFD was influenced by age, with the highest value of 21.01 in the milk teeth category and the lowest value of 13.87 in the full mouth category ( Figure 2 , p < 0.001). In addition, fiber density decreased with increasing age, with a strong negative linear correlation of –0.57 between age and fiber density (Flores et al., 2023). This mode of development is also present in secondary follicle density. While primary follicle density showed slight variation, it decreased at the rump site ( Table 3 ). On the other hand, total follicle density has been found to decrease with age, although it is not affected by type of shearing, locality, or fleece type (Frank et al., 2006). This may be due to skin stretching caused by animal growth (Curasma and Rodriguez, 2016). The genetic correlation between hair duct density and fiber density per mm² in alpacas was 0.87 ± 0.06 (Flores et al., 2023).

The secondary-to-primary follicle ratio of 6.28 was slightly higher than the 4.89 reported in Ch’aku llamas by Antonini et al. (2004) and the 5.35 ± 1.44 in Q’ara llamas reported by Lusky et al. (2006), who also reported 5.45 in Th’ampulli llamas, with no significant differences. However, these values are lower than those reported in Huacaya alpacas, with S/P ratios of 8.62 and 8.51 in shorn and unshorn animals, respectively, at 16 months of age (Pallotti et al., 2020). In a detailed study, Atlee et al. (1997) reported an S/P ratio of 9 in llamas. The S/P ratio decreased as the age category increased, consistent with the patterns described by Curasma and Rodriguez (2016), although Lusky et al. (2006) found no changes with age in young animals.

Our findings indicate that the density of primary follicles was higher on the mid-side than on the rump. Fiber quality and production depend on skin properties (Antonini et al., 2004). Consistent with our results, total skin volume was higher in the femoral area and lower in the shoulder area. The total follicle group volume was also higher in the femoral area, resulting in increased fiber production (Sasahara et al., 2022). Follicle density is related to skin thickness. The epidermis is thick on the heads and necks of guanacos, llamas, and crosses. Llamas have a very thick dermis, with marked differences between the superficial and deep layers (Atlee et al., 1997; Carpio, 1991; Lacolla et al., 2010). This may be unfavorably related to follicle density and fiber diameter.

The relationship between MFD and TFD showed a negative correlation (–0.17), lower than the values of –0.35 to –0.47 found in Chilean colored alpacas by Crossley et al. (2014), and –0.8 reported by Carpio (1991). This indicates that as follicle density increases, mean fiber diameter decreases. This result was expected, as the negative correlation between MFD and TFD has already been established in Argentinean llamas with different fleece types (Frank et al., 1989). This could be partly explained by the decrease in TFD that occurs as the animal grows, while the inflection points of MFD increase (Frank et al., 2006). In this study, TFD and the S/P ratio showed a positive correlation (0.52), similar to the 0.38 reported in alpacas by Ferguson et al. (2012).

Between MFD and the S/P ratio, we found a negative correlation (–0.18), which is lower than –0.3 reported by Oruna (2016) and –0.44 reported by Ferguson et al. (2012). Badajoz (2007) reported stronger negative correlations of –0.73 and –0.75 for Suri and Huacaya alpacas, respectively. This is favorable, as it indicates that a higher secondary-to-primary follicle ratio is associated with a smaller fiber diameter. The main histological characteristic of fine fibers is the greater number of secondary follicles, as it has been widely demonstrated that animals with smaller fiber diameters have more secondary follicles (Atlee et al., 1997; Moore et al., 2015).

Although the S/P ratio is the main parameter used to define skin follicle structure in both types of llamas, it may be a subjective classification due to variation in traits within the same genetic population (Lusky et al., 2006; Antonini, 2010). Due to the limited sample size and the study’s focus on a single breed, future research should examine environmental effects on fiber properties and the influence of age from birth to adulthood. These studies could help inform breeding programs, genetic correlations, and marker-based selection strategies.