Estimation of milk production in hair ewes by two methods of measurement


Estimación de la producción de leche en ovejas de pelo mediante dos métodos de medición


Irina Peniche G,1* M.Sc, Luis Sarmiento F,1 Ph.D, Ronald Santos R,1 Ph.D.

1Autonomous University of Yucatan, Faculty of Veterinary Medicine and Animal Science, Department of Animal Nutrition, Carr. Mérida-X’matkuil km 15.5, Apdo. 4-116 Itzimná, CP 97100, Mérida, Yucatán, México.


Recibido: June 2014; Accepted: December 2014.


Objective. The objective of the current study was to compare two methods to estimate daily milk production in crossbred hair ewes. Materials and methods. Eight multiparous, crossbred (Pelibuey x Katahdin) lactating hair ewes were used in a completely randomized design with repeated measurements, for 56 days. Ewes were fed ad libitum with a diet based on concentrate. Milk production was estimated twice a week for each ewe by both methods, suckled-hand (SH) and weigh-suckle-weigh (WSW) method. Results. Milk production (p>0.05) means were 1031.9±95.6 and 1119.0±95.6 g/day/ewe for SH and WSW method, respectively. Conclusions. No significant differences (p>0.05) were found between methods for milk yield.

Key words: Hand milking, lambs, weighing (Source: AGROVOC).


Objetivo. El objetivo del presente trabajo fue comparar dos métodos para estimar la producción diaria de leche en ovejas de pelo cruzadas. Materiales y métodos. Se utilizaron ocho ovejas multíparas, cruzadas (Pelibuey x Katahdin) y lactantes, en un diseño completamente al azar, con medidas repetidas, durante 56 días. Las ovejas fueron alimentadas ad libitum con una dieta basada en concentrado. La producción de leche se estimó dos veces a la semana a cada oveja, mediante ambos métodos, el doble pesaje del cordero-ordeño manual de la ovejas (DPC-OMO) y el doble pesaje del cordero (DPC). Resultados. La producción de leche estimada (p>0.05) fue de 1031.9±95.6 y 1119.0±95.6 g/día/ovejas para el DPC-OMO y DPC, respectivamente. Conclusión. No se encontraron diferencias significativas (p>0.05) entre métodos de estimación de la producción de leche.

Palabras clave: Corderos, ordeño manual, pesaje (Fuente: AGROVOC).


Daily milk intake is the most important factor in determining lamb growth rate; therefore survival and potential growth of lambs depend on lactation of the dam. Due to the influence of milk yield on weaning weight and to the large variation in milking ability among the genotypes available in tropical areas, the estimation of milk production by lactating ewes contributes with useful information for the implementation of management strategies for ewes and their lambs. Direct measurement of milk production in non-dairy tropical sheep is difficult. For these reasons direct (e.g. hand or mechanic milking) and indirect (e.g. weigh-suckle-weigh method, udder dimensions) methods are commonly used to estimate milk production, however, these methods have demonstrated inaccurate results (1,2).

One of the most cited methods for measuring milk production is the weigh-suckle-weigh (WSW) method (3,4). Some authors describe that use of the WSW method can generate underestimation of milk production at the beginning of lactation (1,5), mainly in ewes with single lambing. For this reason, other authors modified the WSW method (6), by adding a hand milking after second weighing of the lamb, in an attempt to obtain a valid estimation of milk production. The evaluation or comparison of methods to estimate milk yield in non-dairy tropical sheep has been little documented; however some authors reported milk production of 1029 g/day in Pelibuey ewes at the first 20 days post-lambing by WSW method (7). The objective of the current study was to compare two methods to estimate daily milk production in crossbred hair ewes.


Study site. The study was conducted at the Faculty of Veterinary Medicine and Animal Science, Autonomous University of Yucatan (20°58’N and 89°37’O) Yucatan, Mexico. The climatic conditions in the study area correspond to the AW0 classification, characterized as warm and subhumid with a single rainy season between May and October, annual rainfall from 800 to 1000 mm, annual temperature from 26 to 28°C and relative humidity between 65-85% (8).

Animals and experimental design. In this trial eight multiparous hair ewes (Pelibuey x Katahdin) with single lamb (five females and three males) and average body weight (BW) of 42.93±9.93 kg were used in a completely randomized design with repeated measurements. Animals were individually housed (2.5x2.0 m) with their lambs in pens equipped with feed and water containers. The experimental period was the first 56 days of lactation, considering the first seven days as an adaptation period. Ewes were fed ad libitum with a diet based on concentrate, 10.25% crude protein (CP) and 12.32 MJ/kg dry matter (DM) of metabolizable energy (ME), balanced according to NRC (9) requirements. Pens and feed containers were arranged so that the lambs have no access to the dam’s feed. Ewes were dewormed after lambing by an intramuscularly application of Levamisole (Aranda Laboratories, Queretaro, Qro., Mexico).

Milk production was estimated twice a week for each ewe by both methods, suckled-hand (SH) method (6) and weigh-suckle-weigh (WSW) method (10), from the second to the eighth week of lactation.

Milk estimation through SH method. Once a week, lambs were separated from their dams for 12 h beginning at 20:00 h. The day after lambs were weighed with a digital weighing scale (5 g/kg of precision) at 8:00 h and at that moment they were allowed to suckle the ewe for 10 minutes, then lambs were weighed again immediately to estimate the amount of milk suckled. Additionally, after suckling, the ewes were hand milked soon after to get the milk left in their udders and this quantity of milk was recorded. In order to obtain the residual milk in dams an intramuscular injection of three IU of synthetic oxytocin (Aranda Laboratories, Queretaro, Qro., Mexico) was used. During the separation period lambs were located in pens allowing sight and smell contact with their dams and avoiding suckling. After milk measurement lambs were kept with their dams the rest of the day. Daily milk production was calculated by summing up the milk suckled by lambs (difference between lamb’s weights) and milk obtained by hand milking. Then, the result of this calculation was multiplying by two, with the aim to extrapolate the estimation of the milk production in 24 h period.

Milk estimation through WSW method. Once a week, lambs were separated at 3:00 a.m. from their dams for three hours and after that, they were allowed to suckle for 10 minutes. Then, lambs were separated again from the ewes for three hours. During the separation time lambs were located in pens allowing sight and smell contact with their dams and avoiding suckling. After this second separation period (9:00 h), lambs were weighed and then they were allowed to suckle the ewe for 10 minutes, after that they immediately were weighed again. The difference between pre and post-suckling weights was defined as milk consumption and consequently milk production. This recording was made over the 12 h period (6:00, 9:00, 12:00, 15:00 and 18:00 h). Then, the daily milk production was calculated by summing up the estimations recorded and multiplying the result by two.

Ethics aspects. The present trial was approved by the Bioethic Committee at the Faculty of Veterinary Medicine and Animal Science of the Autonomous University of Yucatan.

Statistical analysis. Repeated measurements for estimations of weekly milk production was conducted using MIXED procedure of SAS (11). The fixed effects were the methods of estimation of milk production and the random effects were the animals. Additionally, orthogonal regression analysis was used to verify if both methods provide comparable milk production measurements. Orthogonal regression is useful when both axis, the predictor (x) and answer (y), have estimation error. Accordingly, the WSW method was considered the predictor for the SH method in the milk production estimation. The error variance ratio (σ SH/ σ WSW) was estimated and included in the orthogonal regression analysis using Minitab (12).


Mean milk production of hair ewes measured using two different methods are presented in table 1. No significant differences (p>0.05) were found between methods for estimate the milk production. Daily estimation of milk production showed similar coefficient of variation (0.35 for SH and 0.34 for WSW method).

Table 1

Milk production between weeks (Figure 1) was different (p≤0.05). At the seventh post-lambing week there was significant difference (p≤0.05) in the estimation of milk production (1043 and 1497 g/day for SH and WSW, respectively) according to the method used.

Figure 1

The orthogonal regression (Table 2), shows no significant difference (p>0.05) between methods in the estimation of milk production coefficient. The confidence intervals (95%) for the intercept (a) and the slope (b), included 0 (p>0.05) and 1 (p≤0.001), respectively. Moreover, the error-variance ratio (SH/WSW) between methods around 1 suggests equality in the estimation by both methods.

Table 2


The estimation of milk production by the methods used in this study was similar to the results reported in Pelibuey (1029 g/day) and Karadi ewes (1223 g/day) by different authors (6,7) and higher than the results described (603-750 g/day) in West African ewes (4,13).

The maximum milk production for the SH method in the fourth post-lambing week is in agreement with other studies (6,14) where the production peak were reached between the third and fifth week of lactation. In the normal curve of lactation, close to the third week, a peak of lactation is expected (6). Furthermore, some authors have reported that milk production on non-dairy sheep maintains a maximum level during the first four or five weeks of lactation (2).

On the contrary, the peak of milk production with the WSW method disagrees with the literature as it reaches the maximum level lately; at seventh week normally milk production decreases (15). It could be partially related to the fact that methods used to measure milk production have the potential to interfere with the natural behaviour of the dam and her young. Besides, it is considered that in early lactation the WSW method can generate underestimation of milk production when a single lamb is used, as in this study, considering that in non-dairy ewes there is surplus milk in the udder at the beginning of lactation which small lambs are not able to ingest (1).

The results of the orthogonal regression for both methods showed that the estimation of milk production with SH method is comparable with the measurement by WSW method.

Since estimation of the weekly milk production did not show significant difference between methods and both methods showed a similar coefficient of variation, it is concluded that both methods can be used in hair ewes; however, the authors of the present study consider that the SH method requires less animal management, which support the use of that under experimental conditions.


The first author would like to express her gratitude to the International Center for Development and Decent Work (ICDD-Germany) and to the Consejo Nacional de Ciencia y Tecnologia (CONACyT-México) for the PhD scholarships awarded.


1. Van der Linden DS, Lopez-Villalobos N, Kenyon PR, Thorstensen E, Jenkinson CMC, Peterson SW, Blair HT. Comparison of four techniques to estimate milk production in singleton-rearing non-dairy ewes. Small Rum Res 2010; 90(1-3): 18-26.

2. Ünal N, Atasoy F, Akçapinar H, Koçak S, Yakan A, Erol H, Uğurlu, M. Milk yield measured by oxytocin plus hand milking and weigh-suckle-weigh methods in ewes originating from local crossbred in Turkey. Rev Méd Vét 2007; 158(6): 320–325.

3. Nezamidousta N, Kominakisb A, Safari A. Use of Wood’s model to analyze the effects of milking methods on lactation curve in sheep. Small Rum Res 2013; 113(1): 195-204.

4 Espinoza JC. Producción y composición de la leche y composición corporal de la borrega Pelibuey lactante, alimentada con tres niveles de energía metabolizable. [Ph.D Thesis]. Merida: University of Yucatan; 2014.

5. Aboul-Naga AM, El-Shobokshy AS, Moustafa MA. Milk production from subtropical non-dairy sheep 2. Method of measuring. J Agric Sci 1981; 97(02): 303-308.

6. Baker IA, Dosky KN, Alkass JE. Milk yield and composition of Karadi ewes with the special reference to the method of evacuation. J Duhok Univ 2009; 12(1): 210-215.

7. Pavón M, Fuentes JL, Lima T, Albuernes R, Efremov A, Perón N. Estudio de la producción de leche de la oveja Pelibuey, Pelibuey x Suffolk y Pelibuey x Corriedale y el crecimiento del nacimiento al destete de sus crías. Rev Cub Reprod Anim 1987, 13(1): 39-49.

8. Instituto Nacional de Estadistica y Geografia (INEGI). Anuario estadistico de los Estados Unidos Mexicanos 2010. Aguascalientes, Ags, MX: INEGI; 2010.

9. National Research Council (NRC). Nutrient requirements of small ruminants sheep, goats, cervids, and new world camelids. Washington, USA: The national academic press; 2007.

10 Poulton SG, Ashton WM. Studies on ewe’s milk. V. The effect of high cereal diets on ewes and on the yield of milk and milk constituents. J Agric Sci 1972; 78 (02): 203-213.

11. SAS. Release 9.2. Cary (NC): SAS Institute, Inc. 2009.

12. Minitab. Release 16. Philadelphia (PA): Minitab, Inc. 2009.

13. Ünal N, Akçapinar H, Atasoy F, Yakan A, Uğurlu M. Milk yield and milking traits measured with different methods in Bafra sheep. Rev Méd Vét 2008; 159(10):497-501.

14. Neville TL, Meyer AM, Reyaz A, Borowicz PB, Redmer DA, Reynolds LP, Caton JS, Vonnahme KA. Mammary gland growth and vascularity at parturition and during lactation in primiparous ewes fed differing levels of selenium and nutritional plane during gestation. J Anim Sci Biotech 2013; 4(6):1-7.

15. Ochoa MA, Torres G, Ochoa AE, Vega L, Mandeville PB. Milk yield and composition of Rambouillet ewes under intensive management. Small Rum Res 2002; 43(3): 269-274.