Behavior of Murrah buffaloes under influence of strangers in the milk parlor during milk control


Behavior of Murrah buffaloes under influence of strangers in the milk parlor during milk control


Análisis comportamental de búfalas Murrah bajo la influencia de personas extrañas en el Control Lechero


Maria Vanderly Andrea,¹* Ph.D, Cintia Righetti Marcondes,2 Ph.D, Adriana Bagaldo, 1 Ph.D, Kaliane Nascimento de Oliveira,¹ M.Sc, Evani de Oliveira-Strada,1 Ph.D, Emanoela Souza-Conde,³ M.Sc, Sabrina Luzia G. de Souza,¹ Ph.D.

1Universidade Federal do Recôncavo da Bahia, Centro de Ciências Ambientais e Biológicas, Rua Rui Barbosa, 720, CEP 44380.000, Cruz das Almas, Bahia, Brazil.
2Embrapa Pecuária Sudeste, Rodovia Washington Luiz, km 234, CEP 13560-970, São Carlos, São Paulo, Brazil. 3Instituto Federal de Educação, Ciências e Tecnologia Baiano Campus Teixeira de Freitas, Rodovia BR 101, Km 882, s/n, CEP 45.985-970, Teixeira de Freitas, Bahia, Brazil.


Received: February 2014; Accepted: December 2014.


Objective. The study was carried out in two farms, located in São Sebastião do Passé, Bahia, Brazil, during the three first months of lactation of 81 female buffaloes, to observe if stranger people at milk parlor could affect the milk production. Materials and methods. Four strangers were at the milking parlor during the milking time to observe the following parameters: Duration of Stay in the Milking Parlor; Milking Period; Rumination; Defecation; Urination Reactivity; Stress Level; Side Preference in the Milking Parlor ; Milk Yield, and Post Control Milk Yield. The females were categorized as 1=primiparous, 2=pluriparous, and 3=when information about calving number was not available. The data were analyzed using SAS Statistical Package. Results. During the presence of four strangers at the milking parlor, it was observed that the mean value for Duration of Stay in the Milking Parlor was 23 minutes (varying from 12 to 38 minutes). Average Milking Period was 8 minutes (varying from 4 to 13 minutes). Average Milk Yield during observation days (MY) varied from 2.30 kg to 14.70 kg, with an average value of 7.70±2.60 kg, while Mean Milk Yield after the observation days was 8.600±2.70 kg. Conclusions. As far as Post Control Milk Yield, since the animals presented increased milk yield in both farms on the days following the observations, it is possible to assert that the presence of strangers in the milk parlor during milk control reduced milk yield only during the day of observation, with no adverse effects on subsequent milk production.

Key words: Animal production, animal welfare, Bubalus bubalis (Source: CAB, USDA).


Objetivo. El estudio fue realizado en dos propiedades en el municipio de São Sebastião do Passé – Bahía, Brasil, durante los tres primeros meses de lactancia de 81 búfalas, para observar si las personas extrañas en sala de ordeño podrían afectar la producción de leche. Materiales y métodos. Cuatro personas extrañas estaban en la sala de ordeño durante el tiempo de ordeño para observar los siguientes parámetros: el Tiempo de Permanencia en la Sala de Ordeño; Tiempo de Ordeño; Rumiación; Defecación; Micción; Reactividad; Nivel de Estrés; Preferencia de Lado en la Sala de Ordeño; Producción de Leche y Producción de Leche Post – Control. Las búfalas fueron Categorizadas como 1, para las primíparas, 2 para las multíparas y 3 cuando la información del número de partos no fue obtenida en la hacienda. Los análisis fueron realizados a través del Statistical Analysis System. Resultados. Durante la presencia de cuatro extraños en la sala de ordeño, se observó que el Tiempo de Permanencia en la Sala de Ordeño medio fue de 23 minutos (variando de 12 a 38 minutos) y el Tiempo de Ordeño medio de 08 minutos (variando de 4 a 13 minutos). La producción de leche media en el día de Control Lechero varió de 2.30 kg a 14.70 kg, con media igual a 7.70±2.60 kg. La producción de leche en los días siguientes fue igual a 8.60±2.70 kg. Conclusiones. Como los animales presentaron aumento de producción lechera después del control en ambas haciendas, es posible que la presencia de personas extrañas en la sala de ordeño durante el Control Lechero mensual redujera el rendimiento apenas durante los días de observación, sin efectos adversos sobre la posterior producción de leche.

Palabras clave: Bienestar animal, Bubalus bubalis, producción animal (Source: CAB, USDA).


The world buffalo population is estimated in approximately 177.240 millions of animals, distributed throughout forty-two countries and an annual increase of around 18 million animals per year for the last ten years (1). In Brazil, the buffalo herd amounts to 1.2 million animals (2). Murrah buffaloes are successfully bred and raised all over the country, and are appreciated for their high quality products.

Buffaloes have become an economically feasible alternative for the production of animal protein, mainly due to their adaptability, precocity and longevity (3-5). In terms of animal production, behavioral studies are relevant to determine the importance of behaviors such as rumination, defecation, urination and reaction to milking, which are closely related to animal welfare, to significant zootechnical characteristics such as milk yield, for example. With the 1.80% cumulative growth of the Brazilian buffalo herds in the past few years (6), the analysis of behavioral data and methods of evaluation are useful tools in selective breeding and targeted mating.

The increase of 20% in milk yield, observed after changes on how the animals were managed, is a strong evidence of the importance of animal welfare, not only in ethical terms but also in increased productivity (7,9). In fact, some producers have already adopted raising systems based on ethological factors in order to increase animal welfare and production.

Several studies on human/animal interaction have been developed to determine the factors that negatively affect animal welfare in the milking parlor (8,10). Understanding the behavior of Murrah buffaloes is relevant for stockbreeders in order to reduce the stress factors affecting welfare and productivity. At the same time, actions taken to improve animal welfare and lessen stress levels are of extreme importance in favoring the implantation of organic production systems.

Therefore, it is necessary to analyze set of behavioral elements, such as characterizing the development of an ethogram, which enables the assessment of animal welfare and its implications on productivity.

This study used an ethogram to evaluate the behavior of lactating Murrah buffaloes while exposed to the presence of strangers in the milk parlor during milk control for the first three months of lactation, and to verify the influence of people in the well-being of these animals in order to assess the variation in production performance during and after periods of milk control.


Location and production system. This study was conducted in two dairy farms located in the city Sebastião do Passé, in the state of Bahia, Brazil. Farms 1 and 2 were distant 30 km from each other. Climate in the region is rainy, hot and wet, with moderate winds and average rainfall of 1.65 mm. Both farms joined the Embrapa’s Buffalo Genetic Improvement Program (Empresa Brasileira de Pesquisa Agropecuária), established by Embrapa Eastern Amazon, in 2008. The study was based on the reality of the buffalo farms of the region rather than on experimental conditions and controlled environments.

Farm 1 consisted of 635 ha, with 535 ha being used for pasture rotation (20% of Brachiaria humidícula, 78% of Brachiaria decumbens, 2% of Brachiaria arrecta, sugar cane and Brachiaria mutica). The pastures were divided with electric fences, and water was made available at all times. Before milking, animals were transferred from pastures to an unpaved 100 m2 pre milking area without roof. The milking parlor has an automatic milking system with two lines and eight places on each side. Milk yield is automatically recorded. There were 250 lactating animals, divided into three milking groups. Calves remained with their mothers during first week of life. After this period, they only had access to the mothers during milking time, when they also received supplementary soybean meal. After milking, the females were fed a concentrate containing wheat bran, cotton seeds, soybean meal, mineral salts and barley. The amount of feed offered depended on individual milk yield. In both farms, dry cows grazed in pastures, and mineral salt was offered in feeders at all times.

Farm 2 had 600 ha, half of which were used for pasture rotation (60% of Brachiaria humidícula, 20% of Brachiaria decumbens, 1% de Pennisetum purpureum; 1% of sugar cane and 18% of overflow tolerant grass foragesEchinochloa pyramidalis and Leersia hexandra). As seen in Farm 1, pastures were divided by electric fences, and there was a river flowing throughout the property providing water to the different pastures. The corral had a roofed area of 700 m2 and an area of 500 m2 without roof. It also had a scale, cattle chute and a mechanic bucket milking system with eight places into two milking lines. The farm had 140 lactating animals divided into two milking groups.

Reproductive management was similar in both farms, and the breeding season took place between the months of March and August. Timed Artificial Insemination was used from September to February. The average age for breeding varied from 18 to 25 months, and the first birthing occurred when animals were 30 to 36 months old. In both farms, all animals were vaccinated and dewormed regularly. The milking process occurred twice daily.

Management procedures such as the entry of the cows in the milking parlor, mastitis test, teat pre-dipping and drying, milking cup application and removal, post milking dipping and milk parlor exit were observed. The management in the parlor was the same for the both farms.

Data. During the morning milking, there were four strangers whose three were collecting data of milk production, and one person was observing the behavior of each buffalo at the milking line from beginning to the end of the period. Data were collected by the direct observation of 81 female buffaloes (52 in Farm 1, and 29 in Farm 2) during their stay in the milking parlor. Each animal were observed from its entrance and leaving the milking parlor. Animals were observed monthly during the first three lactating months, and the strangers arrived suddenly at the parlor to make sure that the animals would not be adapted to them. Milk weight was recorded twice daily (morning and afternoon). One week after these observations, a new milking observation was collected for further comparison with the Post Control Milk Yield (PCMY).

Females were categorized (CAT) as 1=primiparous, 2=pluriparous, and 3=when information about calving number was not available.

Continuous variables. The focal sampling and continuous collection followed a model adapted from ethogram of Paranhos da Costa and Broom (8), Hötzel (11) and Honorato (12). The behavioral variables bellows was observed:

Period in the Milking Parlor (PMP): measured as the time from entrance into the parlor until the animals were set free; Milking Period (MP): measured in minutes, from the moment of the milking cups were applied until their removal from animals; Milk Yield (MY): amount in kg of milk produced during observation days; Post Control Milk Yield (PCMY): amount in kg of milk produced one week later the first observation, without the presence of strangers in the milking parlor.

Discrete variables. The variables recorded as observed or not necessarily in the milking parlor (Rumination – RU, Defecation – DEF, Urination – UR).

Modal scores. Other measurements observed in the milking parlor:

Reactivity (REAC): defined as the actions or movements of the animal during milking (8):
1.) Yes or not leg movements;
2.) If there were legs movement:
2.1) lifted less than 15 cm above the ground;
2.2) lifted more than 15 cm above the ground.

Stress Level (SL): behavior of the animals in the milking parlor:
1.) Agitated: eyes wide-open, raised ears;
2.) Slightly agitated: watchful, staring, ear movements towards the noise source, interruption of current activity;
3.) Calm: head and ears down, drowsy;

Side Preference in the Milking Parlor (SPMP):
1.) Right side of the line;
2.) Left side of the line.

Statistical analysis. The continuous variables PMP, MP, MY and PCMY were analyzed using the SAS Statistical Package (13). To determine if Farm had a fixed effect on the discrete variables RU, DEF and UR, the GENMOD procedure of SAS was used. The modal scores for REAC (REACMO), SL (SLMO) and SPMP (SPMPMO) were considered fixed effects using the GLM procedures in analysis models that also consider Farm and CAT as fixed effects on the variables MY and PCMY.

Mean comparison tests for significant effects were performed (Tukey’s test for CAT and SL or REACMO, and t-test for Farm and SPMP). For rank correlation in both farms between MY and PCMY, the SAS PROC CORR SPEARMAN procedure was applied.

The chi-square test was used to test the following hypothesis: (alpha=0.05): H0 – rumination ratio is the same in both farms; H1- rumination ratio is not the same in both farms; H0 – defecation ratio is the same in both farms; H1- defecation ratio is not the same in both farms; H0 – urination ratio is the same in both farms; H1- urination ratio is not the same in both farms.


The mean value for the Period in Milking Parlor (PMP) was of 23 minutes (varying from 12 to 38 min), and the mean Milking Period was of 8 min (varying from 4 to 13 min). Average milk yield during observation days (MY) varied from 2.30 kg to 14.70 kg, with a mean value of 7.70±2.60 kg. Mean milk yield on the days after observations was 8.60±2.70 kg, varying from 3.70 kg to 17.30 kg (Table 1).

MP (Period in the Milking Parlor, measured as the time from entry into the parlor until the animals were set free); MP (Milking Period, measured in minutes from the moment the milking cups were applied until their removal from animals); MY (Milk Yield, amount in kg of milk obtained during observation days); PCMY (Post Control Milk Yield, values obtained one week after the first observations).

In the present study, animals could choose to be milked on the right or the left side of the milking lines. Overall, animals from Farm 1 preferred the left side of the line, while animals from Farm 2 preferred the right side of the milking line.

One hundred and fifty six observations from Farm 1 and eighty seven from Farm 2 were performed regarding the following behavioral parameters: Rumination, Urination and Defecation. The results, expressed as percentage, are shown in table 2.

According to the data, all variables presented different ratios in both farms. Chi-square tests (p<0.05) indicate that there were no equal ratios between farms for the variables urination and rumination (the expected χ2 were higher than the observed χ2 for 1d.f.), which is in agreement with the data in table 2.

Only six female buffaloes with Reactivity (REAC) index of 3 (legs lifted more than 15 cm above the ground) were observed in Farm 1. For the variable level of stress (SL), it was observed only one buffalo scoring in all observations. The same animal was classified as agitated and presented leg agitation.

According to the statistical model applied, the modal score of Stress Level (SLMO) for Milk Yield (MY) was significant (p<0.0001), i.e., farm, CAT and SLMO explained 44% of the MY variation. However, the characteristic “Farm” had no significant effect on MY (p>0.60). During milking control days, calm females produced 3 kg more than what was recorded on days after observation days. Pluriparous animals produced 1.60 kg more than the primiparous females. Older females produced 2.80 kg more milk in the absence of strangers. Therefore, the yield decreases more evident in this category (9.80 kg of milk compared to 7.00 kg during the milk control recording day).

Reactivity modal scores (REACMO) for Milk Yield (MY) were significant (p<0.0001). Farm, Category (CAT) and REACMO explained for 34% of the Milk Yield (MY) variation. However, the characteristic “Farm” had no significant effect on MY (p>0.70), and no significant differences (p>0.05) among REAC 1, 2 and 3 were observed. In terms of PCMY, the statistical model explained for 35% of the variation and was significant (p<0.0001). The effect of CAT was not significant (p>0.09). Tukey’s test indicated that females within modal score 1 (motionless legs) produced 1.70 kg more than those in modal score 3 (legs 15 cm above the ground). Farm 1 had an statistically higher production than Farm 2 (p<0.05), with an average milk yield of 2.00 kg more per animal on the days after milk control observations with the presence of strangers (Figure 1).

According to the statistical model applied, the modal score of Side Preference (SPMPMO) for Milk Yield (MY) was significant (p<0.0001), with a R-square of 21%. Farm, CAT and SPMPMO explained for 21% of the observed MY variation. In both farms, side preference (SPMPMO) had no significant effect (p>0.70) on MY during milk control days. In terms of PCMY, the model was significant (p<0.009) with a low R-square of 16%. CAT, Farm and SPMPMO were not significant (p>0.15). Side preference in the milking parlor was not a significant source of variation in milk yield. The overall rank correlation between PLO and PLP was equal to 0.76. There was a 24% change in the ranking of the animals for the two periods studied. When the analysis was done by farm, the correlation rank of Farm 1 was equal to 0.74 (or 26% change in scoring), while Farm 2 was equal to 0.58 (ie 42% change in ranking). The change in rank may be an indicative of the influence of presence of strangers in the milk parlor in the buffalo’s milk production, because there were considerable changes in the classification of buffaloes between periods.


The mean value for period in Milking Parlor (PMP) was 23 minutes (ranging from 12 to 38 minutes), and the mean of the Milking Period was 8 minutes (ranging from 4 to 13 minutes). Average milk yield during observation days (MY) ranged from 2.30 kg to 14.70 kg, with a mean value of 7.70±2.60 kg. Mean milk yield one week after the observation days was 8.60±2.70 kg, ranging from 3.70 kg to 17.30 kg, which suggests the interference of stranger in the parlor on buffaloes milk production. Peters et al (14) observed higher values for PMP, with TOR being higher in cows that were not subjected to aversive handling (PCMY = 9.6 min and MP = 7.3 min for non-aversive and DSMP = 9.2 min and MP=7.0 min for the aversive handling). According to the authors, the higher MP in animals submitted to aversive handling could not be due to the peaceful management and respect for the speed of movement of animals, which increased the time until the animals had settled in the parlor.

Cows under stress showed resistance to enter the milking parlor, hampering the work of the milkier and, consequently, presenting a reduction in milk production (15,16). Milk production affected the PCMY and MP significantly. Therefore, the time spent in the milking parlor and the milking time interfered at the efficiency of the milking process.

It is known that performance is affected by external stimuli, and that animals can perceive specific situations as pleasant or unpleasant in relation to previous experiences (17-19). The preferences of which side to be milked may be related to the way the spaces are used in both sides of the milking lines (passageway for calves and humans, for example). A research carried out by Marcondes et al (20), in Farm 1 with the observation of 248 animals, showed that 54.3% of the females chose the same side of the milking line. In a study performed by Grasso et al (19), focusing on the relation between entry of bovines in the milking parlor and milk yield, the authors observed a negative correlation for multiparous females and a positive correlation for primiparous females. However, Alves (21) observed no difference in milk production between different categories of females.

Although both Farms belonged to the same owner, daily management differed in terms of feeding and human-animal interactions. The presence of strangers in Farm 2 during Milk Control (MC) days does not affect animal comfort, as suggested by the data in table 2. High incidence of defecation, urination and reactive animals, with a higher frequency of negative actions in the milking parlor, may have reflected impaired animal welfare (9).

The low reactivity scores in categories 1 and 2 suggested that the behavioral interference was neutral, with the exception of Milk Production during observation days. Stress Level 1 (agitated) in all observations was found in only one female, which also presented leg movement. Such behavior can indicate that improvements in the routine procedures of the farm may be required (8). As reported by Lewis and Hurnik (22), previous negative experiences may increase animal reactivity. Harmonic human-animal interactions reduced stress reactions and reactivity to humans, thus facilitating handling (23-25) without impairing daily production.

The sporadic presence of strangers in the milking lines does not seem to affect animal behavior and is not followed by a reduction in milk yield (22). The presence of known humans during the control procedures might contribute to reduce the negative effects of the presence of strangers on animal behavior (10). Our results also showed that the presence of known people during Milk Control (MC) does not contribute to reduce the negative effects of the presence of strangers on MY. It is well known that bovines like routine and, since they tend to have a good memory, they can discriminate humans involved in the interactions by reacting in specific ways (23), such as inhibiting milk ejection (26), and reducing daily milk yield (21).

Changes in the daily routine of the animals may cause discomfort and interfere with their daily production (14). This may be evidenced by changes in the ranking of the MY, and PCMY animals. The milk production of buffaloes on a farm may increase as changes are made in the management to meet the behavioral needs of animals (7,24).

In conclusion the results of the current study of the behavior of Murrah females in Bahia, Brazil, showed a reduction of milk yield on milk control days due to the presence of strangers in the milking parlor. The negative effects, however, subsided as soon as the observation was concluded and the presence of strangers in the milking parlor was removed, therefore, not affecting the subsequent milk yield of the animals.


The authors thank the farmer Urbano Antônio de Souza Filho for providing the animals and support research.


1. Sethi RK. Buffalo improvement program in India. Rev Vet 2010; 21(1):5.

2. Ramos AA, Malhado CHM, Carneiro PLS, Azêvedo DMMR, Gonçalves HC. Caracterização fenotípica e genética da produção de leite e do intervalo entre partos em bubalinos da raça Murrah. Pesq Agropec Bras 2006; 41(8):1261-1267.

3. Jorge AM. Produção de carne bubalina. Rev Bras de Reprod Anim 2005; 29(2):84-95.

4. Cassiano LAP, Mariante NA. Parâmetros genéticos das características produtivas e reprodutivas de búfalos na Amazônia. Pesq Agropec Bras 2004; 39(5):451-457.

5. Garcia AR. Influência de fatores ambientais sobre as características reprodutivas de búfalos do rio (Bubalus bubalis). Rev de Ciênc Agrár 2006; 45:1-13.

6. Bernardes O. Bubalinocultura no Brasil: situação e importância econômica. Rev Bras Reprod Anim 2007; 31:293-298.

7. Dawkins, MS. A user’s guide to animal welfare Science. Trends Ecol Evol 2006; 21:77-82.

8. Paranhos Da Costa MJR, Broom DM. Consistency of side choice in the milking parlour by Holstein-Friesian cows and its relationship with their reactivity and milk yield. Appl Anim Behav Sci 2001; 70(3):177-186.

9. Hanna D, Sneddon IA, Beattie VE. The relationship between the stock person’s personality and attitudes and the productivity of dairy cows. Animal 2009; 3:737–743.

10. Hötzel MJ, Gomes CCM, Machado Filho LCP. Comportamento de vacas leiteiras submetidas a um manejo aversivo. Biotemas 2009; 22:135-140.

11. Hötzel MJ, Machado Filho LCP, Yunes MC, Silveira MCAC. An aversive milker causes fear, but does not influence milk yield of Holstein cows. Braz J Anim Sci 2005; 34:1278-1284.

12. Honorato LA, Hötzel MJ. A interação humano-animal e o uso de homeopatia no manejo sanitário de rebanhos leiteiros em pequenas propriedades no Sul do Brasil. Rev Bras de Agroec 2012; 7(2):77-86.

13. SAS. Statistical Analysis System - SAS Version 9.0, 13th ed. Cary: Statistical Analysis System Institute; 2004.

14. Peters MDP, Barbosa Silveira ID, Pinheiro Machado Filho LC, Machado AA, Pereira LMR. Manejo aversivo em bovinos leiteiros e efeitos no bem-estar, comportamento e aspectos produtivos. Arch Zootec 2010; 59(227):435-442.

15. Breuer K, Hemsworth PH, Coleman GJ. The effect of positive or negative handling on the behavioural and physiological responses of non lactating heifers. Appl Anim Behav Sci 2003; 84(1):3-22.

16. Costa JHC, Hötzel MJ, Longo C, Balcão LF. A survey of management practices that influence production and welfare of dairy cattle on family farms in southern Brazil. J Dairy Res 2013; 96(1):307-317.

17. Broom DM, Molento CFM. Bem-estar animal: conceito e questões relacionadas – revisão. Arch Vet Sci 2004: 9:1-11.

18. Molento CFM. Bem-estar animal: qual é a novidade? Acta Sci Vet 2007; 35(2):224-226.

19. Grasso F, De Rosa G, Napolitano F, Di Francia A, Bordi A. Entrance order and side preference of dairy cows in the milking parlour. Ital J Anim Sci 2007; 6:187-194.

20. Marcondes CR, Andréa MV, Oliveira KN, Oliveira RS, Macedo Junior CM, Cerqueira RS et Al. Side Preference in the Milking Parlor Among Murrah Buffaloes in the Recôncavo of Bahia, Brazil. Rev Vet 2010; 21(1):598-599.

21. Alves BG, Silva TH, Igarasi MS. Manejo de ordenha. PUBVET 2013: 7(6): Art. 1514. URL Disponible en:

22. Lewis NJ, Hurnik J F. The effect of some common management practices on the ease of handling of dairy cows. Appl Anim Behav Sci 1998; 58:213-220.

23. Honorato LA, Hötzel MJ, Gomes CCM, Silveira IDB, Pinheiro Machado Filho LC. Particularidades relevantes da interação humano-animal para o bem-estar e produtividade de vacas leiteiras. Cienc Rural 2012; 42(2):332-339.

24. Watters RD, Schuring N, Erb HN, Schukken TH, Galton DM. The effect of premilking udder preparation on Holstein cows milked 3 times daily. J Dairy Sci 2012; 95:1170-1176.

25. Waiblinger S, Mulleder C, Schmied C, Dembele I. Assessing the animals’ relationship to humans in tied dairy cows: between-experimenter repeatability of measuring avoidance reactions. Anim Welf 2007; (16):143-146.

26. Breazile, JE. The Physiology of Stress and Its Relationship to mechanisms of Disease and therapeutics. Vet Clin N Am: Food Anim Prac 1988; 4:441–480.

Métricas de artículo

Cargando métricas ...

Metrics powered by PLOS ALM

Enlaces refback

  • No hay ningún enlace refback.

Copyright (c) 2016 Revista MVZ Córdoba