Exploring the heights: Impact of altitude on dairy milk composition

Summary The aim of this study was to investigate the effect of altitude on bulk milk quality of farms located at different altitudes. Altitude had a notable effect on bulk milk composition, which can be explained by alterations in feed quality as well as physiological alteration in dairy cows because of adaptation to challenging environmental factors at higher altitudes. The results of the study should provide new insights into a little-studied production factor for ensuring animal well-being as well as product quality in dairy farms located at different altitudes.

V arious factors are known to influence milk composition and its physiochemical properties, including animal-related factors such as animal health, breed and genetics, lactation stage, parity (Jílek et al., 2006;Li et al., 2014;Stocco et al., 2017), as well as environmental and housing conditions, feeding and nutrition, and season (Hofstetter et al., 2014;O'Callaghan et al., 2016).Despite extensive research on several management and environmental factors, the studies investigating the effect of altitude on milk composition are few (e.g., Leiber et al., 2006;Saha et al., 2019;Mohammed et al., 2022).Altitude plays a significant role in dairy production, particularly in mountainous regions, where unique topographical circumstances and traditional seasonal management practices like summer transhumance on alpine pastures are prevalent (Battaglini et al., 2014).Various environmental parameters directly or indirectly related to differences in altitude can exert an impact on milk performance and composition through their influence on dairy cow physiology (Qiao et al., 2013;Saha et al., 2019).Directly related parameters include atmospheric pressure, temperature, and turbidity, while indirectly related factors encompass solar radiation, moisture, wind, season length, feed quantity and quality, and geology (Körner, 2007;Peratoner et al., 2010;Neary et al., 2013).Some studies have explored the relationship between altitude and milk properties in different lactating animals, including camels, goats, and sheep, demonstrating significant effects on conductivity, ash content, specific gravity, calcium, and potassium (Mohammed et al., 2022).Additionally, research conducted on dairy cows exposed to higher altitudes during summer transhumance showed an increase in milk fat content and a decrease in protein content compared with cows in lowland areas (Leiber et al., 2006;Saha et al., 2019).Nevertheless, the effect of altitude on dairy cows in mountain dairy farming is still relatively unexplored.To address this research gap, the present study aims to evaluate the effects of altitude, season of sampling, as well as housing and management conditions on gross milk component properties considering bulk milk samples from mountain dairy farms.
A total of 32 farms agreed to voluntarily participate in this project.Information and detailed descriptions of farm management factors, including housing systems (freestall or tiestall), grazing practice, flooring type, and cow breed were obtained from on-farm audits considering the protocol developed and published by Katzenberger et al. (2020).Data of bulk milk analysis of the farms were retrieved from the database of the Breeders Association of Bolzano province (Bolzano, Italy) and the South Tyrolean Dairy Association (Bolzano, Italy).Furthermore, farms were classified according to the altitude of the geographical location of the farm buildings into 3 altitude categories (A = 600-900 m above sea level [a.s.l.], B = 901-1,200 m a.s.l., and C = 1,201-1,500 m a.s.l.).The pastures where dairy cows had access were located near the farms.Therefore, the altitude of the graphical location of a farm corresponded to one of its pastures.
The statistical analysis involved a total of 5,680 bulk milk samples from 32 farms.The selected on-farm factors were altitude, housing system, season of sampling, and grazing.To account for the repeated observations of bulk milk samples from assessed farms and potential correlation between cows within the same

Exploring the heights: Impact of altitude on dairy milk composition
Mousaab Alrhmoun, Thomas Zanon,* Katja Katzenberger, Louis Holighaus, and Matthias Gauly farm, the study employed a linear mixed-effects model using SAS software v. 9.4 (SAS Institute Inc., Cary, NC).Multiple explanatory variables that could potentially influence milk composition were included in the model as fixed effects.The fixed effects were altitude (categorical, 3 levels: A, B, C), season of sampling (categorical, 4 levels: 1 = summer (June to August), 2 = autumn (September to November), 3 = winter (December to February), and 4 = spring (March to May), grazing practice (categorical, 2 levels: yes and no), and housing system (categorical: tiestall and freestall).
The bulk milk records were collected over a span of 4 yr, with multiple observations obtained from each farm.The use of random intercepts allows for the consideration of the variability between farms, effectively capturing the farm-specific effects on milk composition.By accounting for these random effects, the model acknowledges the inherent clustering of milk samples within farms and ensures that the estimates of the fixed effects (e.g., altitude, season, grazing, housing system) are not biased by within-farm correlations.
Multiple comparisons of least squares means were conducted for the effects of altitude, season, grazing practice, and year using post hoc Bonferroni's test (P < 0.05).Investigated farms were classified regarding the altitude of their geographical location, housing system, and pasture use.For instance, 10 farms belonged to altitude category A and B and 12 farms belonged to altitude category C. Regarding housing system, 15 farms used tiestalls and 17 farms had freestalls.The latter is explainable by the fact that because of topographical and structural limitations, old-fashioned tiestall housing is still very common in mountain areas despite being criticized for impairing animal welfare (Beaver et al., 2021).Further, 19 farms enabled their cows to have access to pasture, whereas 13 farms did not use pasture.
In Table 1, descriptive statistics of investigated milk components are summarized.The mean values for fat, protein, and lactose were 3.99%, 3.47%, and 4.73%, respectively (Table 1).The mean for casein was 2.70% (Table 1).Free fatty acids and SCS averaged 1.09% and 3.27 units, respectively.Further, MUN had a mean value of 22.05 mg/100 mL (Table 1).Finally, the means for acetone and pH were 0.62 mmol/L and 6.60, respectively (Table2).The coefficient of variation ranged from 1.53% (pH) to 27.30% (MUN).The second most variable trait beside MUN was SCS with a coefficient of variation (CV) of 24.14% (Table 1).The average milk chemical composition in this study was comparable to previous investigations on bulk milk samples reported by Wittwer et al. (1999) and Saha et al. (2019).The mean and CV of urea content resembled values observed in Arunvipas et al. (2004) and Vallas et al. (2010).Urea variability in milk can be attributed to factors such as different feeds, the quality of rations, feeding procedures in dairy herds, as well as nonnutritional factors including herd productivity variations, seasonal and parity variations, and the timing of milk sampling after feeding, as noted by Rajala-Schultz and Saville (2003), Arunvipas et al. (2004), Vallas et al. (2010), and Zanon et al. (2021a).
Results of the ANOVA of bulk milk samples are shown in season, grazing, and housing system) were significant (P < 0.05) in explaining the variation of bulk milk from investigated farms.Altitude had the most significant (P < 0.0001) impact on fat, casein, and transformed SCS, whereas season was important in explaining the variability in protein and MUN of the milk.Grazing had a limited effect on most of the traits, except for lactose.The effect of housing system had some effect on fat and acetone (Table 2).The coefficient of determination (R 2 ) ranged from 0.12 for FFA to 0.81 for fat, indicating that the effects included in the analysis accounted for a substantial proportion of the variation observed in the studied traits (Table 2).The root mean squared error values provide an estimate of the average deviation between the observed values and the predicted values based on the effects included in the model.In Table 3, least squares means of compositions and traits of bulk milk samples of investigated effects are summarized.Fat, protein, casein, FFA, as well as SCS were notably higher in farms belonging to altitude category C while lactose content was lower compared with farms belonging to altitude category A and B (Table 3).These findings partly align with a previous study conducted by Saha et al. (2019), which reported higher fat content in milk from cows exposed to transhumance on alpine pastures above 1,200 m a.s.l.compared with those that remained in the farm at the valleys below 800 m a.s.l.The same study, however, reported lower protein contents in milk of cows at higher altitude, which is the opposite of what we observed in our analysis (Table 3).Similarly, Zemp et al. (1989) reported higher fat but lower protein and lactose contents in milk from dairy cows exposed to higher altitudes during summer pasturing.The higher fat content in milk from cows exposed to higher altitudes could be related to a higher lipolysis rate caused by an insufficient energy intake due to limited feed intake, which results in a negative energy balance and a higher body fat mobilization as commonly observed in high-yielding dairy cows at the beginning of lactation (Churakov et al., 2021).For example, Zhang et al. (2022) observed a lower DMI in heifers farmed at high altitudes, which could explain the energy and nutrient deficiencies and the resulting alterations in milk composition (e.g., fat) in the present study and body conformation observed in previous assessments (Zemp et al., 1989;Saha et al., 2019;Zhang et al., 2022).In this regard, cattle breed was shown to have a relevant effect on body fat mobilization as Zendri et al. (2016) observed breed-specific differences in variation of BCS of dairy cows exposed to higher altitudes during transhumance.The authors described a higher body reserve mobilization and decline in milk production in specialized highyielding dairy breeds such as Brown Swiss and Holstein Friesian.This highlights a higher difficulty of specialized dairy breeds in adapting to the harsh environment present on highland summer pastures compared with local dual-purpose breeds Alpine Grey and Rendena (Zendri et al., 2016).Since we considered bulk milk and not individual milk samples, we could not disentangle possible differences in milk to respective breeds as investigated farms mainly reared multibreed herds.Furthermore, animals exposed to high altitudes show some additional physiological adaptation such as an increase in the number of red blood cells and the amount of hemoglobin for acclimatization to low oxygen environment at high altitudes (Neary et al., 2013;Neary et al., 2015;Kong et al., 2021).Such changes in blood serum composition could influence blood enzyme activity as well as other blood biochemical parameters such as glucose, triglycerides, or cholesterol.This could influ-ence milk quality as Andjelić et al. (2022) observed significant (P < 0.05) correlations between milk and blood enzymes as well as blood and milk biochemical parameters.Moreover, the higher fat content could also be related to a difference feed composition and quality.For instance, a higher crude fiber content in roughage is well described for alpine forage (pasture and hay) (Pittarello et al., 2018;Ineichen et al., 2019;Pornaro et al., 2019).However, Zhang et al. (2022) showed also that digestibility of nutrients was higher in cows reared at high altitudes as a physiological consequence of the limited feed intake, availability, or both.This could explain the higher protein content in milk from cows reared at higher altitudes as feed nitrogen, according to Zhang et al. (2022), is used more efficiently.In fact, thanks to the urea cycle via rumino-hepatic circulation ruminants can reuse urea nitrogen produced in the digestive system for producing microbial protein (Getahun et al., 2019).The trend toward higher SCS in farms located at a higher altitude might be more related to management and housing factors.In fact, beside breed-specific effects as demonstrated by Zanon et al. (2021a), SCS is influenced by milking and postmilking hygiene (Chassagne et al., 2005), animal-related parameters like lactation stage and parity (Zanon et al., 2020b), as well as environmental effects (Lambertz et al., 2014) and feeding (Dufour et al., 2011).Notwithstanding, Friedrich and Wiener (2020) reported in their literature review that the harsh environment at high altitudes (e.g., increased solar radiation, cold winds, colder temperatures) can trigger the metabolic rate of animals.In addition, high altitude can cause pulmonary hypertension, which puts pressure on the organism.This might result in an increased susceptibility against some pathogen-related diseases such as mammary infections.The higher SCS in farms located at higher altitudes in the present study could explain the lower lactose content as Costa et al. (2019) described that lactose is lost through the bloodstream during mammary infections due to disrupted blood-milk barrier and changed permeability of alveolar epithelium.The lower MUN in farms of altitude category C could highlight a possible nitrogen deficiency of the cows as Ineichen et al. (2019) described a lower ruminally degradable protein in hay from alpine pastures due to higher lignin and cellulose content.Albeit significant (P < 0.05), differences in acetone and pH among farm altitude categories were negligible (Table 3).Further, fat, protein, casein, FFA contents were lower in summer, which is in line with results reported in previous investigations (e.g., Zanon et al., 2020a;Lu et al., 2021), whereas SCS and MUN were lower in autumn (Table 3).Lactose as acetone TA and pH showed little variation throughout the year (Table 3).The effect of grazing showed limited differences on bulk milk composition traits (Table 3), which is in contrast with results found in literature (e.g., Hofstetter et al., 2014;O'Callaghan et al., 2016;Zanon et al., 2023b).This might be explained by the limited grass feed intake of dairy cows in investigated farms as most of the forage intake occurred in the stables.However, the latter cannot be discussed further as specific information regarding specific feed ratio composition was not available.Finally, fat, FFA, SCS, and acetone tended to be higher in tiestalls (Table 3).In this regard, however, it is important to consider that the reported differences are more likely be related to different feeding management, feed quality, as well as genetic aspects and overall farm conditions present in respective housing systems as to the housing system per se.Nevertheless, the effect on housing system and facility design is already well described in having a notable impact on animal health and welfare (Popescu et al., 2014;Zanon et al., 2021bZanon et al., , 2023a)), which has also an effect on milk composition (Weaver et al., 2016).
Our results show a significant (P < 0.05) influence of altitude on bulk milk quality parameters such as fat, protein, casein, and SCC.In addition to possible variation in feed quality due to changing environmental influences related to differences in altitude and breed effect, which were not considered in the present study, it was discussed that physiological changes also could have a notable influence on bulk milk composition.Therefore, our results indicate the special needs of dairy cows under such harsh production conditions at high altitudes.Those needs should be considered in farm management (e.g., feeding management, breed decision) so that animal health and the associated animal welfare as well as the productivity of traditional small-scale mountain dairy farms can be guaranteed.

Table 2 .
Overall, the results indicate that all considered effects (altitude, 140 Alrhmoun et al. | Effect of altitude on milk composition

Table 1 .
Descriptive statistics of compositions of bulk milk samples(n = 5,680)

Table 2 .
Results from ANOVA for composition traits of bulk milk samples 1 1 df: level of category; FFA = free fatty acid; SS = type III sum of squares; RMSE = root mean squared error.

Table 3 .
Least squares means of compositions and traits of bulk milk samples across altitude (3 categories), season, grazing practice, and housing system 1 Trait