Effect of the feeding system on yield, chemical composition, and fatty acid profile of artisan cheese from local goat milk in northern Mexico

Location

The study was carried out in a production unit located in the ejido Ignacio Zaragoza, Municipality of Viesca, Coahuila, in the region known as Comarca Lagunera, which is located at 24° N and 104° W, at 1100 m above sea level. The climate is desert, semi-warm with cool winters; mean annual temperature is 25 °C and mean annual rainfall is 240 mm.

Animals and treatments

In this study, milk was used from 10 local "Criollo" goats (crossbreeds Alpine × Saanen × Toggenburg) at the beginning of lactation, with an average live weight of 42±2.25 kg and 2.5 births, randomly distributed into two treatments: housed system (n = 5) and extensive grazing (n = 5). The animals had an adaptation period of 14 days and an experimental period of 28 days. Goats in the housed system were kept in individual shaded pens (2×3 m) equipped with individual feeders and waterers, with water available ad libitum. They were offered 2.5 kg dry matter (DM) per goat per day of a ration based on forage and concentrate, divided into two equal meals at 08:00 and 16:00 h, formulated according to the nutritional requirements for lactating dairy goats (NRC, 2007). Rations were offered individually, and feed refusals were collected and weighed daily before the morning. Feed allowances were adjusted weekly according to each goat’s live weight to maintain consistent intake levels (Table 1). Goats in the grazing system were herded daily to native rangeland pastures and allowed to graze from 12:00 to 18:00 h, covering approximately 4-8 km day^-1. From the end of the grazing period (18:00 h) until departure the next day (12:00 h), the goats were kept in pens where they had access to water ad libitum and did not receive any additional feed supplementation.

The experiment was arranged in a completely randomized design with two feeding systems (housed vs. grazing). For milk yield and composition, the experimental unit was the individual goat (n = 5 per treatment). Measurements were taken once daily throughout the experimental period, and weekly composite samples were prepared for milk composition and cheese manufacture as described below.

Chemical composition of the concentrate and forage

Forage was sampled in weeks 1, 3, and 6, and composite samples were prepared for each species; throughout the grazing route, 300 g of each forage species. Regarding the concentrate, different samples were taken from the total mixed ration and composite samples were prepared. The forage and concentrate samples were dried and subsequently processed in a Thomas WILEY mill (Model 4, Laboratory Mill), with 1 mm mesh, for subsequent analysis.

The grazed rangeland consisted of C. virgata, C. dactylon, P. bipinnatifidum, Acacia spp., C. berlandieri, and S. elaeagnifolium. Botanical composition was estimated using quadrat sampling: 0.5-m² quadrats were systematically placed along the grazing route each week; species-level biomass was clipped at around 2 cm, simulating the animal's bite, oven-dried (60 °C, 48 h), and expressed as g DM m^-2. Sampling took place during the early dry season when the stand was mostly mature: grasses were actively tillering, and forbs and shrubs were leafy and flowering. This phenological context is important for understanding the FA profile and protein content of the forage offered. Bite-count and focal-scan observations (10-min bouts per goat; five goats per group; six observation days) revealed a higher preference for Accacia spp. and C. virgata inflorescences, and avoidance of C. dactylon. Therefore, chemical analyses focused on the species that were preferred (composite by preference).

The ingredients in the diet of the housed goats consisted of alfalfa, oats, and concentrate. The analyses of the chemical composition of the concentrate and forage samples were: dry matter (DM) (AOAC 925.45), organic matter (OM) which was calculated by difference, total protein (TP) (AOAC 991.20), ethereal extract (EE) (AOAC 989.08), ashes (AOAC 938.08) (AOAC 2005), neutral detergent fiber (NDF), and acid detergent fiber (ADF) (Van Soest et al., 1991).

Milk yield and composition

Goats from both treatments were manually milked once a day at 06:00 h throughout the 28-day experimental period. Immediately after milking, individual milk yield (kg day^-1) was recorded using a portable digital scale. Fat-corrected milk (4 % FCM) was calculated according to NRC (2001) using the equation:

4 % FCM (kg day^-1) = 0.4 × milk yield (kg day^-1) + 15 × milk fat (kg day^-1).

For each goat, a weekly composite milk sample (100 mL) was prepared by pooling proportional aliquots from the seven daily milkings to obtain a representative sample of the week. Immediately after sampling, the milk was cooled to 4 °C and preserved with bronopol according to the manufacturer’s recommendation. Preserved samples were stored at 4 °C for no longer than 7 days before analysis. During storage, the samples were kept in sealed vials and protected from light to limit lipolysis and microbial growth.

Before analysis, the milk was gently mixed and filtered through a clean cloth to remove visible impurities that could interfere with the measurements. Milk fat, protein, and lactose contents were determined on the weekly composite samples by mid-infrared spectroscopy using an automatic milk analyzer, following the manufacturer’s instructions. All analyses were performed in duplicate, and mean values were used for statistical analysis. All samples from both feeding systems were collected, preserved, stored, and analyzed under identical conditions.

Cheese making process

The fresh cheese was made with a mixture composed of milk collected during the week, simulating the artisanal process of the region.

A total of 92.53 L^-1 of milk was processed with a total of 9 samples of fresh cheeses produced for goats in the housed system, while for the grazing goats, a total of 51.37 L^-1 was processed and 7 samples of fresh cheeses. The milk was placed in a stainless steel container of 30 L capacity to carry out pasteurization at 63 °C/30 min^-1, after which time it was allowed to cool to a temperature of 40 °C. Commercial rennet was added (coagulation strength 280 IMCU mL^-1, from Cuamex Industries, Mexico), prepared according to the supplier’s instructions, adding 10 mL per 100 L^-1 of milk. The coagulation time was 40 min at room temperature. After the coagulation period, the curd was cut into cubes of approximately 3 cm with a stainless steel knife vertically and horizontally. It was manually agitated for 30 min to gradually release the whey, then a partial draining of 3/4 was carried out and then NaCl (600 g/100 L^-1) was added, mixing thoroughly. The curds were placed in plastic molds and allowed to press under their own weight, turning them every 20 min for 1 h. Subsequently, the cheeses were covered with sterilized polyethylene paper and refrigerated for 18 to 24 h at 4 °C. The cheeses were weighed and placed in a freezer at -20 °C for their respective analyses.

Cheese yield was determined by considering the wet fraction of the cheese, which is the actual yield and was calculated according to the equation of Hu et al. (2013). Meanwhile, the one excluding this fraction is the dry matter adjusted yield and was obtained following the equation of Fenelon and Guinee (1999).

Fatty acid profile

A total of 16 cheese samples and 7 food samples were analyzed for FA analysis. Freeze-drying of cheese: 40 g of cheese was weighed into amber vials, which were frozen at -80 °C. They were then placed in a Labconco^® lyophilizer, (FreeZone 6, USA), attached to a Vacuubrand^® RZ 9 vacuum pump. Lyophilization was carried out at a vacuum point of 0.010 mBar for 48 h at -49 °C. The samples were stored under refrigeration for FA analysis.

Subsequently, the FA profile of the cheese was determined in the Animal Nutrition laboratory of the Livestock Program of the Colegio de Postgraduados, Campus Montecillo, Texcoco, State of Mexico. The FA extraction was performed according to the methodology of Feng et al. (2004). The FA profile was determined using the modified methylation technique of Palmquist and Jenkins (2003) and Jenkins (2010), in which FAs are in the form of methyl esters. For this purpose, 0.5 g of freeze-dried cheese sample and 0.5 g of forage and concentrate samples were used, which were placed in 50 ml culture tubes with bakelite caps. Three mL of sodium methoxide (Sigma-Aldrich catalog 403067-250) (0.5 M in methanol to protect the isomerization process of the unsaturated FAs) was added, and vortexed gently to mix. The tubes containing the cheese samples were placed in a water bath at 50 °C for 10 min and then allowed to cool for 5 min. Subsequently, 3 mL of 5 % methanolic hydrochloric acid was added, capped, and vortexed slightly. The tubes were placed in a water bath at 80 °C for 10 min and allowed to cool for 7 min. Then, 3 mL of hexane was added to dissolve and extract only the fat and 5 mL of 6 % potassium carbonate to saponify and release the FAs and neutralize the reaction. The samples were vortexed and the contents were emptied into 16.5 mL polypropylene tubes and centrifuged (Beckman J2-HS, USA), at 1233 g for 10 min. The organic phase was then extracted and placed in another polypropylene tube containing 0.5 g of sodium sulfate and 0.1 g of activated carbon to remove impurities. It was vortexed and centrifuged (Beckman J2-HS, USA), at 629 g for 10 min. The sample was then extracted and filtered through an acrodisc (Thermo Scientific, titan 44513-NN, 17 mm green filter and 0.45 μm nylon membrane; to ensure an impurity-free sample) and placed in a vial where it was stored at -20 °C until analysis by gas chromatography. FA methyl esters were determined on a Hewlett Packard 6890 chromatograph with automatic injector with a silica capillary column (100 m x 0.25 mm x 0.20 μm thick, Sp-2560, Supelco, USA) and helium as carrier gas. The column temperature was programmed at 140 °C for 2.95 minutes, increasing to 210 °C at a rate of 3 °C/min and subsequently to 235 °C at a rate of 0.7 °C/min. Run time: 62 minutes. The identification of the FAs was performed by comparing the retention times of each peak obtained from the chromatogram with a standard of 37 FA peaks from the Supelco company.

Cheese chemical composition

The determination of protein, fat, moisture, salt, and total solids in cheese was performed by near infrared transmission (NIT) that transmits light through the cheese sample with a FoodScan^TMLab analyzer (FOSS Analytical AB, Hillerød, Denmark), in the Multipurpose Laboratory of the Autonomous University of Chapingo, Texcoco, State of Mexico, using a 100 g sample of cheese.

Texture profile analysis (TPA)

Texture profile analysis (TPA) was performed at the Department of Agroindustrial Engineering in the Sensory Analysis Laboratory of the Universidad Autónoma Chapingo, Texcoco, State of Mexico. TPA was performed using the method described by Hernández-Morales et al. (2010), using a TA Xt2i texture analyzer (Stable Micro Systems; Surrey, UK) with a 5 kg load cell. The samples (25 mm diameter and height cylinders) were uniaxially compressed at 50 % strain using a 50 mm diameter acrylic disk and a pre- and pos-test head testing speed of 1 mm/s^-1. The data from the double compression allowed obtaining force/time curves from which the APT parameters were calculated: Hardness which is the peak force (F) of area one (A1) of the first compression. Adhesiveness which represents the work required to remove the disc compressing the sample and corresponds to area three (A3). Cohesiveness, defined as the property that allows the cheese to remain attached after a first compression and is calculated as the ratio of area two to area one (A2/A1). Elasticity refers to the ability of the cheese to return to its original height after a first compression and is calculated by dividing distance two by distance one (D2/D1). Finally, chewiness refers to the work required to perform a double compression and corresponds to the product of F * (A2/A1) * (D2/D1). The calculations of the APT properties were calculated as follows: Hardness = F (N), Adhesivity = A3 (N-s), Cohesivity = A2/A1 (dimensionless), Elasticity = D1/D2 (dimensionless), and Chewability = hardness x cohesivity x elasticity (N) (Torres-Salas et al., 2023) (Figure 1).

Figure 1. Texture profile analysis (TPA) curve. Calculations of the textural properties: hardness = F (N), adhesiveness = A3 (N s), cohesiveness = A2/A1 (dimensionless), elasticity = D1/D2 (dimensionless), and chewiness = hardness x cohesiveness x elasticity (N). Taken from Torres-Salas et al. (2023)

Statistical analysis

The data obtained from the variables protein, fat, moisture, salt, total solids, actual and adjusted cheese yield, and textural parameters, were analyzed in a completely randomized design in a repeated measures arrangement with the MIXED procedure of the SAS v.9.4 statistical software. The Bartlett test was performed to determine the homogeneity of variances of the data. The Bayesian information criteria of Schwartz and Akaike were used to determine the most appropriate covariance structure. The FA profile information was analyzed using a one-factor ANOVA with the GLM procedure and the comparison of means was performed with Tukey's test. The general structure of the model is as follows:

Y_ijkl=µ + R_i(j) + T_j + S_k + T_j*S_k + E_ijkl

Where:

Y_ijkl: Actual and adjusted yield and content of chemical components (fat, protein, moisture, salt, and total solids).

µ: Constant characterizing the population

R_i(j): Random effect of the i-th cheese nested within treatment (i=1,2)

T_j: Fixed effect of the jth treatment (j=1, 2)

S_k: Fixed effect of the k-th week of treatment (k=1...4)

T_j*S_k: Treatment*week interaction effect.

E_ijkl: random error, which was assumed for all components normally distributed with zero mean and common variance.

Milk

Milk yield, fat-corrected milk, and its chemical composition were different between treatments (p<0.05) (Table 3).

Milk and fat-corrected milk production increased by 96.767 % and 89.164 %, respectively, in housed goats compared to grazing goats (p<0.05). This increase is attributed to concentrate consumption, which increases the production of propionate in the rumen, converting it into glucose in the liver and transforming it into lactose in the mammary gland (Hills et al., 2015), generating greater water movement to the mammary secretory cells, and consequently greater milk volume (Otaru et al., 2020). Overall, the higher milk and fat-corrected milk yields observed in housed goats indicate that the stall-feeding system provided a more stable energy supply than the native rangeland available to grazing goats. The concentrate-based ration likely supported greater dry matter intake and a more favorable energy balance, which in turn enhanced mammary secretory activity. Similar patterns have been reported in goat dairy, where housed or mixed systems (stall-feeding plus limited pasture) increased milk volume and fat-corrected milk compared with animals relying exclusively on grazing (Granados-Rivera et al., 2022). These results suggest that, under the semi-arid conditions of the Comarca Lagunera, housing with a controlled ration can be an effective strategy to maximize milk yield in local Criollo goats, although at the cost of higher feed inputs.

In the case of milk production in grazing, this decreases due to the energy expenditure caused by the displacement during the search for forage (Lachica and Aguilera, 2005), which is also associated with the low availability and quality of forage that generates an energy and protein deficit, causing the goats to use their body reserves (Goetsch, 2019), thus leading the proteins of the animal's muscle fibers to be mobilized from the synthesis of milk components decreasing milk production (Ramírez-Rivera et al., 2019).

Table 3. Production and chemical composition of milk from local housed and grazing goats

^abDifferent letters between columns indicate a difference (p<0.05). SEM - standard error of the mean

Milk fat concentration was 4.330 % higher in grazing (p=0.027). This result is due to fiber intake. In this sense, the forages present a higher concentration of galactolipids, specifically linoleic acid (C18:2 c9, c12) and α-linolenic acid (C18:3) (D'Urso et al., 2008) which are hydrolyzed to glycerol, FA, and galactose, the latter is rapidly fermented and transformed into volatile fatty acids (VFA). Thus, it can be inferred that the increase in milk fat production in grazing goats was due, in part, to the greater production of acetic acid, a key precursor to the synthesis of milk fat (Seymour et al., 2005).

Protein was higher in milk from housed goats (p<0.001), with an increase of 14.482 % with respect to grazing goats (p<0.05). The carbohydrates in the feed consumed by the animals are easily fermentable, so there is greater production of energy, propionic acid, and microbial protein (Ángeles-Hernández et al., 2020). This, in turn, leads to changes in the animal’s body signals, increasing the concentration of insulin in the blood, which generates changes in the mammary gland favoring the concentration of protein in the milk of housed goats (Jenkins and McGuire, 2006). On the other hand, dietary protein in animals includes nitrogen (N), present as true protein, which is broken down into amino acids and ammonia and used by rumen microorganisms for the synthesis of microbial proteins necessary for protein synthesis in milk (Lu et al., 2019).

Lactose production was higher in the milk of goats in the housed system (p<0.001) with respect to those in grazing. In this regard, concentrated feeds generate higher proportions of propionic acid in the rumen (Maldini and Allen, 2019), which serves as the substrate for the production of glucose (Kittivachra et al. 2007), the main precursor of lactose synthesis by mammary gland cells (Otaru et al., 2020), consequently increasing lactose production in the milk of housed goats.

It is interesting to note that there were significant interactions over time in milk fat concentration, which decreased during the experimental period. Regarding protein and lactose, they showed an increase for both treatments (p<0.05) (Figure 2). This behavior is associated with a negative correlation between milk production and its components, i.e. as milk production increases, protein and lactose also increase, while fat content decreases due to a dilution effect (Morand-Fehr et al., 2007).

Figure 2. Concentration of fat, protein and lactose in milk from local goats in housed and grazing systems during the 4-week experimental period (weekly measurements, weeks 1-4)

Cheese

The chemical composition and yield of the cheese were different between treatments (p<0.05) (Table 4). The concentration of fat, protein, salt, and total solids, showed an increase of 5.662 %, 23.503 %, 23.529 % and 9.784 %, respectively, in cheeses from housed goats with respect to grazing goats (p<0.05). The above is related to the caseins and fat molecules that were transferred from milk to cheese during the manufacturing process (Cuchillo et al., 2010), which begins with the hydrolysis of κ-casein due to enzymatic action, this destabilizes the casein micelles, which then aggregate to form a gel or curd network that primarily traps fat and protein globules. Once the curd has formed, the syneresis process begins, where the casein network contracts, trapping the fat and protein within the matrix. These fats and proteins are then transferred to the cheese, while much of the lactose and minerals are lost in the whey during the drainage process (Ribak. 2014). Zervas and Tsiplakou (2011) reported that cheese quality depends to a large extent on the composition of the milk, mainly the fat and protein content. Regarding the fat component, it was significantly higher in cheese from housed goats. This result is due to the fact that the casein network traps the milk fat, having a higher fat recovery in the curd (Pazzola et al., 2019).

The cheeses for both production systems presented moisture contents higher than 50 %, classifying them as soft cheeses, in which the basket molds are stacked for self-pressing, in addition to being characterized by rennet action and slow draining (Jiménez-Maroto and Ibáñez et al., 2023). According to Mexican regulations (NOM-223-SCFI/SAGARPA-2018 and NOM-121-SSA1-1994), the cheeses analyzed are classified as fresh cheeses, which are characterized by high moisture contents, unripened, lacking thin rind, short shelf life, and ready for consumption immediately after production.

Table 4. Chemical composition of cheese from local housed and grazing goats

^abDifferent letters between columns indicate a difference (p<0.05). SEM - standard error of the mean.

Interactions over time were found in cheese chemical components (p<0.05) (Figure 3). Fat concentration had an increase until week three with a decrease at week four. Protein and total solids concentrations remained stable while salt dereased during the experimental period.

Figure 3. Concentration of fat, protein, salt, and total solids in cheese from local goats in housed and grazing systems during the 4-week experimental period (weekly measurements, weeks 1-4)

Cheese yield over time was higher in housed goats (p<0.001), with an increase of 1.136 % higher than those on grazing (Figure 4). This increase is attributed to the fact that whey proteins that form complexes with κ-casein form a gel that traps fat and are retained in the cheese curd which contributes to an increase in yield (Rangel-Ortega et al., 2024). In this research, a higher protein content was found in the milk of housed goats, which probably contributed to the increase in cheese yield. In this regard, Fekadu et al. (2005) reported that there is a positive correlation between fat concentration and cheese yield. This means that the higher the fat and protein concentration, the higher the cheese yield and vice versa.

Figure 4. Real and dry matter-adjusted cheese yield from local goats in housed and grazing systems during the 4-week experimental period (weekly measurements, weeks 1-4)

The main ingredient of the diet of grazing goats is forages, so greater amounts of unsaturated FA are transferred to dairy products such as cheese (Delgadillo-Puga et al., 2021). In this sense, the FA profile of the cheese was different between production systems (p<0.05) (Table 4). The cheese FA with the highest concentrations in both systems were capric (C10:0), lauric (C12:0), myristic (C14:0), palmitic (C16:0), stearic (C18:0), and oleic (C18:1c-9), which represent 84.991% in housed goats and 83.553 % in grazing.

The concentration of SFA was lower in cheeses from grazing goats (p<0.001). According to Serrapica et al. (2020), this result is attributed to the fact that with forage-based diets there is a reduced synthesis of short- and medium-chain SFA in the mammary gland, due to the high levels of PUFA in the fresh forage diet that compete with SFA in de novo synthesis for esterification in the mammary gland.

MUFA had an increase of 34.183 % in cheese from grazing goats (p<0.001), with respect to those from housed goats, among which cis-10-heptadecanoic (C17:1), oleic (C18:1 c9), and elaidic (C18:1 t9) acids stand out. To this respect, grazing can improve the environment for rumen bacteria, leading to an increase in odd-chain FA like C17:1. Grazing increases microbial diversity and abundance, and the types of bacteria present are responsible for synthesizing these FA, which are then incorporated into milk. At the same time, for PUFA, C18:3 and CLA were higher for grazing goats (p<0.001). The presence of PUFA in cheeses, especially those with potentially favorable effects on human health, depends mainly on the FA composition of the milk used (Dauber et al. 2021). Grazing goats fed native forages produce milk richer in MUFA and PUFA, especially C18:1 c9, CLA, and C18:3 (Coppa et al., 2011). The above because forages contain more α-linolenic acid, a precursor of vaccenic acid (VA) in the rumen, which is subsequently converted to rumenic acid (RA) in the mammary gland through Δ9-desaturase (Renes et al., 2020), thus generating a higher concentration of CLA in milk from grazing goats.

Table 5. Fatty acid profile (g/100 g of total FA) of cheese from local goats in housed and grazing systems

^abDifferent letters between columns indicate a difference (p<0.05). SEM - standard error of the mean. UN - Unidentified, ¹CLA - Conjugated linoleic acid; ²SFA - sum of saturated fatty acids; ³MUFA - sum of monounsaturated fatty acids; ⁴PUFA - sum of polyunsaturated fatty acids.

The above is indicative that the nutritional properties of forages consumed by goats are transferred to milk and in turn to dairy products such as cheese, positively changing the FA composition in terms of PUFA (Serrapica et al., 2020). The same pattern was observed by Dauber et al. (2021), for FA in milk and cheese from goats supplemented with chia seeds, further confirming that these FA are stable during cheese manufacture.

There were interactions between treatment and time with some FA related to human health (p<0.05), among which acid caprylic (C8:0), C10:0, C18:1 c9, and C18:3 stand out. In the case of C8:0 and C10:0, the interaction was similar with a slight increase at the end of the experimental period, being higher in milk from goats in housed. Meanwhile C18:1 c9, and C18:3, decreased except for C18:3, which had a slight increase at the end of the experiment, with higher concentration in milk from goats in grazing.

The analysis of texture profile (ATP) of the cheese was different between treatments (p<0.05) (Table 5). In the case of hardness, it was lower for cheeses from grazing goats (p<0.05), with respect to those from housed goats. This decrease is attributed to the type of forage that influences the composition of milk FAs, which plays a key role in cheese texture (Martin et al., 2005; Coppa et al., 2011). Pasture lipids have a higher proportion of unsaturated FA which have a lower melting point than saturated FA, producing a more fluid fat and consequently softer cheeses with less hardness (Serrapica et al., 2020). In this regard, Coppa et al. (2011) found a lower hardness in cheeses from grass-fed cows compared to concentrate-fed cows, arguing the relationship between C16:0 and C18:1 cis-9 acids, which are the main saturated and unsaturated FA in milk that have high and low melting points, respectively.

The ratio of palmitic acid/oleic acid has been closely positively related to cheese hardness (Martin et al., 2005). The increased hardness in the cheeses from housed goats in this study can be attributed to the significantly higher concentration of palmitic acid, which has a higher melting point than UFA, resulting in a more solid fat at room temperature. In fact, palmitic acid correlates significantly and positively with hardness and chewiness attributes (O'Callaghan et al., 2017). This same behavior was found by Rangel-Ortega et al. (2024), who, in addition, argued that a high protein content is an indicator of hardness, which coincides with what was found in this research for protein content in milk.

Table 6. Texture profile analysis of cheese from local housed and grazing goats

^abDifferent letters between columns indicate a difference (p<0.05). SEM - standard error of the mean

Chewiness had the same behavior as for hardness (p<0.05), which is consistent, since both variables are closely related positively. Therefore, greater hardness corresponds to greater chewiness. This response agrees with the results obtained by Rangel-Ortega et al. (2024), who obtained similar values for cheeses from housed goats (9.92 N).

The cohesiveness and elasticity were higher for cheeses from goats in a housed system (p<0.05) with respect to those from grazing. This result is attributed to the acidity of the cheese which influences the pH. When the pH turns out to be lower than the isoelectric point (4.4-5.7), caseins acquire a negative charge, repelling protein aggregates. This leads to a greater water absorption capacity in the cheese, resulting in greater elasticity (Rangel-Ortega et al., 2024). As elasticity increases, the resistance to deformation of the cheese increases due to the flexibility of the internal bonds, thus increasing the cohesiveness of cheeses from housed goats (Martin et al., 2005). Adhesiveness showed no significant differences between treatments (p>0.05).

The hardness and cohesiveness of cheese from local goats in a housed and grazing system, over time, are shown in Figure 5. The hardness of cheese from housed goats showed a slight increase at the end of the experimental period (p<0.05) and the cohesiveness followed a similar trend throughout the experiment. In this regard, O'Callaghan et al. (2017) reported that palmitic acid correlates significantly and positively with hardness attributes due to a higher melting point than UFA, resulting in a more solid fat.

Figure 5. Hardness and cohesiveness of local goat cheeses in housed and grazing systems during the 4-week experimental period (weekly measurements, weeks 1-4)

It is relevant to state that this study is limited by a small sample size, which reduces statistical power and generalizability, potentially missing longer-term trends, and a lack of testing for seasonal effects, which could obscure important cyclical variations. These limitations mean that the findings may not be representative or robust and could be influenced by specific time-of-year factors, necessitating future research with a larger sample and a longer follow-up period to capture these effects, which is the aim of our research team to work on.

Study limitations and future research

This study has some limitations that should be considered when interpreting the results. First, the experiment was conducted with a relatively small number of local "Criollo" goats from a single herd, which may limit the generalization of our findings to other breeds or production systems. Second, the experimental period covered only four weeks at the beginning of lactation, so the long-term effects of the feeding systems across the entire lactation remain unknown. Third, the grazing conditions and botanical composition of the native rangeland are specific to the semi-arid environment and season of the Comarca Lagunera, and may differ from those in other regions. In addition, we did not directly measure feed intake and nutrient digestibility, nor did we assess rumen fermentation parameters or the sensory and consumer acceptance of the cheeses, which restricts our ability to fully explain the underlying mechanisms and to link compositional changes to perceived product quality. Future studies including larger herds, different breeds, longer experimental periods and a more detailed characterization of intake, rumen metabolism, sensory attributes and economic performance would help to confirm and extend the present findings.