Dependence of milk production of dairy sheep on climate conditions

There was proofed that in countries of moderate climate zone livestock production is changing due to global warming (Gauly et al., 2013; Gauly and Ammer, 2020). This seems to be reason for occurrence of uncommon daily temperatures and frequent season changes. Number of days in those heat-humidity index (THI) is higher than maximum value specified for wellbeing of livestock animals in European countries of moderate climate zone (above 68) has been increasing (Silanikove, Koluman-Darcan, 2015). Therefore, the influence of climate changes (mainly of heat stress) on milk production traits of dairy sheep and goats has become the object of interest to a greater extent recently (Finocchiaro et al., 2005; Marai et al., 2007; Hamzaoui et al., 2013; Salama et al., 2014; Dawood, 2017). The influence of cold stress on ewe performance occurred was shown less frequently (Ramón et al., 2016). It has been observed that breeding aimed at selection of animals resistant to climate change and/or adaptable to climate change is of an increasing importance (Ramón et al., 2016; Sánches-Molano et al., 2019). Flock management needs to be oriented in regulations aimed at minimizing of heat and/or cold stress on livestock production (Ramón et al., 2016; Gauly, Ammer, 2020). The objective of the study was to analyse influence of fixed factors (year, month and interaction year x month) Dependence of milk production of dairy sheep on climate conditions


Introduction
There was proofed that in countries of moderate climate zone livestock production is changing due to global warming (Gauly et al., 2013;Gauly and Ammer, 2020). This seems to be reason for occurrence of uncommon daily temperatures and frequent season changes. Number of days in those heat-humidity index (THI) is higher than maximum value specified for wellbeing of livestock animals in European countries of moderate climate zone (above 68) has been increasing (Silanikove, Koluman-Darcan, 2015). Therefore, the influence of climate changes (mainly of heat stress) on milk production traits of dairy sheep and goats has become the object of interest to a greater extent recently (Finocchiaro et al., 2005;Marai et al., 2007;Hamzaoui et al., 2013;Salama et al., 2014;Dawood, 2017). The influence of cold stress on ewe performance occurred was shown less frequently (Ramón et al., 2016). It has been observed that breeding aimed at selection of animals resistant to climate change and/or adaptable to climate change is of an increasing importance (Ramón et al., 2016;Sánches-Molano et al., 2019). Flock management needs to be oriented in regulations aimed at minimizing of heat and/or cold stress on livestock production (Ramón et al., 2016;Gauly, Ammer, 2020). The objective of the study was to analyse influence of fixed factors (year, month and interaction year x month) as well as influence of climate characteristics (temperature, relative humidity, total precipitation, wind speed and, mainly, THI) on milk production traits..

Biological data
The experiment was done on sheep farm Liptovská Teplička (coordinates: latitude 48° 57' 50.3" N and longitude 20° 04' 31.0" E), where dairy sheep of purebred Improved Valachian (IV) and crosses IV × LC (Lacaune) were kept. In period from 2017 to 2019, total morning milk production (TMPM), total evening milk production (TMPE), and total morning/evening milk production (TMPM + E) were measured. Number of ewes milked was also recorded on a daily basis and used for analyses of average daily milk production per ewe (ADMP). Traditionally, ewes are milked from March to October. Milking of ewes commonly depends on length of lambing season and on climate conditions. Only milk production over April to September was analysed in the present study (no all ewes were milked in March and October each year). Number of days in milk (in 2017 and 2018) was 183, whereas number of days in milk (in 2019) was 173 due to fact that milking started on April 11.

Selected climate characteristics
Climate characteristics were monitored by weather station of Slovak Academy of Sciences (supplier: firm PHYSICUS) located about 2.5 km from sheep farm in period from 2017 to 2019. Data were recorded in 10-minute intervals.
Average daily values of the following climate characteristics: air temperature -T (°C), relative humidity -RH (%), wind speed -WS (v m/s) and total precipitation -TP (mm) measured between 5 a.m. and 4 p.m. daily from April to September were used. This period was chosen due to fact that ewes were moved outside of stable (on pasture) after morning milking. Temperature-humidity index (THI) was calculated according to National Research formula (NRC, 1971), which is widely considered in countries of moderate climate zone (Gauly, Ammer, 2020).

Statistical analysis
Covariance analysis was used to analyse the influence of selected climate characteristics on TMPM, TMPE, TMPM + E and ADMP. The following fixed factors: year (2017,2018,2019), month (April, May, June, July, August, September) and interacion year × month, and covariates: air temperature, total precipitation, wind speed, relative humidity and THI index (instead of T and RH). Pearson coefficients were used to assess phenotype and residual correlations between milk traists and climate characteristics (data measured in same days). Residual correlations were assessed from estimates of residua provided by covariance analysis. Statistical program SASv9.2 (procedures GLM and CORR) were employed. Scheffe multiple-range tests were used to asses significance of differences between individual levels of analysed fixed effects: P <0.05 (+), P <0.01 (++), P <0.001 (+++).

Results and discussion
Basic statistics of milk production traits and selected climate characteristics are given in Table 1. In analysed population, which ranged from 138 to 429 heads, ADMP per ewe was equal to 0.487 l, in best days with milk production above 700 ml occurred. These values agreed with milk production data published for Improved Valachian breed (Breeding Services of SR). Wide ranges between minimal and maximal values of selected climate characteristics were found. For instance, extremly low temperatures (under 0 °C) and high temperatures (up to 27 °C) were recorded. On average, relative humidity was above 95%, total precipitation and wind speed were 1.55 mm and 1.77 m/s (Table 1), respectively.
Covariance analysis of milk traits which included four climate characteristics: T, RH, WS, TP, showed that month had highest importance (P <0.001). The interaction year x month also had highly significant influence (P <0.001). When ADMP estimated for same months across respective years was compared, significant differences between April 2017 (0.607 l), April 2018 (0.553 l) and April 2019 (0.438 l) were found, for example. Similarlly, significant differences in ADMP between May months and September months across respective years were also found. Temperature had significant influence on TMPM + E (P <0.01) and ADMP (P <0.05), respectively. Wind speed and RH had not significant influence on milk traits, total precipitation had significant influence on TMPM (P <0.05). http://www.acta.fapz.uniag.sk © Slovak University of Agriculture in Nitra Faculty of Agrobiology and Food Resources 1, 2, 3 -total milk production (morning, evening and morning + evening, l), 4 -average daily milk production per ewe (l); the same explanation for Tables 2 and 3; THI5 -temperature-humidity index according to NRC (1971) Alternative covariance analysis, which considered three climate characteristics: THI, total precipitation and wind speed (Table 2), showed highest significant influence of THI mainly on TMPM (P <0.001) and ADMP (P <0.01), respectively. Total precipitation had important influence on TMPM (P <0.01) and TPMM + E (P <0.05), respectively. Wind speed did not show significant influence on milk traits. Coefficients of determination (R 2 ) were high, i.e. between 0.675 and 0.878. Taking into account fixed effects (year, month, interaction year x month), residual correlation between ADMP and THI was equal to -0.166 (P <0.001), i.e. ADMP was decreasing along with increasing THI. Finocchiaro et al. (2005) found that in dairy sheep in the Mediterranean region, daily milk and also fat-plus protein yield decreased with increasing THI index. When relationship between ADMP on one hand and T, TP, RH, WS on the other hand was analysed, residual correlation was equal to +0.133 (P <0.01) between ADMP and T, and equal to -0.091 (P <0.05) between ADMP and RH, respectively. These findings indicated that relationships between milk production and climate characteristics may not be linear. Similarly to study of Ramón et al. (2016), climate characteristics highly influenced milk production on respective day of measuerement or day before day of measurement. When days in milk were assigned to three groups according to THI values (Table 3), milk production of ewes assigned to group cold stress was significatly lower in three of four analysed climate characteristics (P <0.05) than milk production of ewes milked on days without heat stress, i.e. during temperature neutral zone (THI >40.0 and THI <= 68.0). Between ewes assigned to group medium heat stress and to group without heat stress, the only difference in TMPE was found significant (P <0.05). No days with THI >75.0 occurred. Similarly, Ramón et al. (2016) reported a higher decrease of milk production due to cold stress than due to heat stress.

Conclusions
The results of this study show that milk production traits of dairy sheep may be significantly influenced by climate also in moderate climatic zone. In analysed population, the higher influence of cold stress than that of heat stress was found. However, ewes mostly produced milk without cold stress and/or without heat stress (86% of days in milk). Milk production in days of medium heat stress was only slightly lower than in days of without heat stress. From aspect of ewe wellbeing, minimizing of both cold stress and heat stress is needed.