INFLUENCE OF MURRAH BUFFALO BEHAVIOR IN MILKING PARLORS ON PRODUCTION CHARACTERISTICS

The present study aimed to assess the influence of different behavioral scores in milking parlors on possible variations in the production characteristics of primiparous and multiparous Murrah buffaloes. Observations of behavioral response in the milking parlor environment were made with an ethogram for subsequent evaluation of behavioral scores. Milk production and somatic cell count (SSC) were measured, in addition to physicochemical analysis of milk. Results for different behavioral scores were assessed by analysis of variance and comparison of means. The Spearman correlation was used to analyze production characteristics. Primiparous buffaloes exhibited higher mean behavioral scores and lower mean milk production when compared with multiparous buffalo cows. Despite greater reactivity, primiparous cows showed no significant differences in milking time, milk let-down time and milk let-down rate in relation to multiparous buffalo cows. In general, more reactive cows display greater mean milk production. Higher behavioral means were recorded in the first week of observation. Behavior did not influence milk composition or SSC. Mean surface skin temperature was positively correlated to total fat and solids only within primiparous buffalo cows.


INTRODUCTION
Buffalo milk production in Brazil is expanding rapidly, mainly because buffalo milk exhibits characteristics allowing greater yields in dairy products, adding value to these products and favoring commercialization when compared with cow's milk in relation to farmers' income.
The buffalo, in addition to being a source of protein with high biological value, can be an economical enterprise, primarily due to its adaptability, early maturity and longevity (PETERS et al., 2007).However, its production like the other ones produces an active and routine interface between animal and man, whose effects on the animal performance can be positive or negative (PETERS et al., 2007).Assuming that animal's productivity, efficiency and comfort are interactive ones, it becomes a positive concern of consumers and society.Thus, an alternative is to invest in the education and training of labor in order to promote a good management of animals and favoring increased productivity and efficient production systems (GONÇALEZ et al., 2008).Therefore, the study of animal behavior can contribute to understanding the animal and man interactions, since it is related to animal's reaction which is related to feeding and/or management practices, and to use them as indicators of animal welfare or stress (ALBRIGHT, 1993).
The animal welfare can be understood as their condition in relation to attempts to adapt to the environment (Broom, 1988), and behavioral measurements are equally significant in assessing it (BROOM and MOLENTO, 2007).Furthermore, an animal avoiding or strongly evading from an object or event he provides information on their feelings and, consequently, their wellbeing.The stronger the evasive reaction the more discomfort the animal shows when faced with that object or event (MOLENTO, 2007).
In addition to the notion of welfare, another widely used concept to understand human-animal interaction is temperament.This has been used to establish a link between behavior and individual behavior (ROTHBART and DERRYBARRY, 1981).Selection of docile animals need to be developed and combined with studies of adequate management techniques in order to minimize problems and negative consequences for humans and animals (LE NEINDRE et al., 1996).Specifically, buffalo cows behavior in milking parlors deserves attention, given that it can directly affect milk production and hence the profitability of this farming (SINGH et al., 2010).Considering management tools, likewise separation of buffalo mothers from their offspring, may cause stress to both and resulting in inhibition or reduced milk ejection by the cow.Observing buffalo cows' behavior during milking is of importance, since it is on this parameter that the welfare and conditioning of these animals during milking can be measured, including to evaluate their temperament characteristics.
Buffaloes and dairy cows have a similar milk ejection reflex (THOMAS, 2004) and a complete removal of milk from the udder depends on elevated oxytocin concentrations release during the entire milking (BRUCKMAIER and BLUM, 1998).Milking difficulties is associated with buffaloes and it is associated with the morphological differences of buffalo udder which suggests some peculiarity of milk letdown and it needs to be considered to improve the efficiency of machine milking (BAVA et al., 2007).Dairy cows, sheeps and goats have a larger total cisternal area and milk cisternal fraction than buffaloes cows does.The cisternal fraction in buffaloes is, on average, only 5% of total milk, decreases during lactation and increases with the age of the animal (Thomas, 2004).This author also suggests that the small volume of buffalo cistern require longer teat stimulation before teat cup attachment to guarantee oxytocin supply and a good milk letdown.Results from Bava et al. (2007) suggest that proper pre-milking stimulation and prompt cluster takeoff could improve milking efficiency, ensuring good milk letdown and protecting teat conditions.
This study aimed to assess the influence of different behavioral reactions in the milking parlor on possible changes in the milk production characteristics of primiparous and multiparous Murrah buffaloes.

MATERIAL AND METHODS
The experiment was conducted at Tapuio Agropecuária Ltd., in the county of Taipu in the Agreste region of Rio Grande do Norte state.It was carried out over a period of six weeks, from September 2 to October 8, 2010.Data collection was taken on two consecutive days in each week during afternoon milking.
Thirty Murrah buffalo cows were used, 15 primiparous and 15 multiparous.Days in lactation at the start of the experiment varied from 11 to 64 (30,93 ± 13,59) days and 3 to 32 (15,33 ± 9,38) days for primiparous and multiparous females, respectively.Milking interval was 10-14 hours, with the first milking at 5am and the second at 3pm; milking were made in a double 20 herringbone milking parlor with low level vacuum line.One month before their expected calving date first calving heifers were introduced into the milking parlor to adapt to milking procedures.Lactating animals were managed in a pasture of Brachiaria brizantha on a rotational grazing.After birth, calves suckled directly from their dams for seven consecutive days when they were separated from cows, permanently.During milking, each cow received a concentrated mixture of cottonseed, sunflower meal, urea and mineral salt at a rate of concentrate to milk of 1:3.
Prestimulation of cows to evoke milk ejection is very important to perform in dairy buffaloes since the milk volume in the udder cistern is very low compared to cows and goats.It was not used hormonal stimulation in buffaloes in this study.Knowing this fact, milking management consisted of the animals kept for resting in a shaded corral with natural ventilation and access to water before entering into the milking parlor.Once inside milking parlor pre-dipping, drying the teats, attaching milking cluster, removal of milking cluster, postdipping and exit of the animals.
Behavioral observations in the milking parlor were performed twice a week by two trained observers during the six consecutive postpartum weeks, with events recorded on a previously compiled ethogram.The milking process was divided into pre-milking, consisting of animal entry and pre-dipping; milking, from attaching milking teat cups until milk flow stop; and post-milking, corresponding to the removal of milking teat cups , post-dipping and animals leaving the milking parlor.Behavioral responses were recorded throughout these three stages, as shown in Table 1.
In accordance with records of behavioral events for primiparous and multiparous cows in the milking parlor, a behavioral score (BS) was defined consisting of four classes regarding This was quantified individually, once a week, during morning and afternoon milking by a scale meter attached to each milking unit.The milk let-down was observed from the time the teats of the cows were attached to the milk cups until the appearance of the first milk flow in the suspension cup.The rate of milk flow was also evaluated by dividing the total milk yield by the total time spent in milking.
In the second and sixth weeks of the experiment, individual cow milk samples were collected during morning milking for analysis of composition and somatic cell count.Samples were taken by a collecting cup attached to each milking unit, homogenized, and transferred to 40 mL flasks containing Bronopol  .Following identification, they were sent to the Laboratory of the Management Program for Dairy Herds in the Northeast, in the Department of Animal Sciences at State Rural Federal University of Pernambuco.Analyses of total fat, protein and solids were conducted by mid infrared absorption using a Bentley 2000  instrument, in addition to somatic cell count (SCC) by flow cytometry with a Somacount 300 -Bentley  .
Skin surface temperature of the animals was measured once a week in four different regions of the body in order to establish individual mean temperature.
Measurement was accomplished with an Instrutherm ® digital infrared thermometer, positioned as close to the animal as possible, in areas close to skin of head, foreleg, udder and back, with the average of the four measurements being obtained.
Data analyzed were submitted to a normality test using the proc univariate and proc means procedures of SAS (2002).All data showed normal distribution, except milk flow rate; milk let-down time and milking duration, which were transferred to 10-logarithmic transformation.Analysis of variance (ANOVA) was performed for the variables: behavior, milk production, milk letdown rate and duration of milking considering the effects of parity and week of lactation, applying the GLM procedure (SAS, 2002).Means were, standard deviation, compared using Tukey at 5.0%.The means of the variables somatic cell count and milk production and composition to assess the effects of behavior and surface temperature on primiparous and multiparous cows were analyzed and compared using Pearson's correlation for the second and sixth weeks of lactation.Data of somatic cell count were transferred to a 10-logarithmic transformation.

RESULTS AND DISCUSSION
Means for behavior observed in the milking parlor and productive characteristics are displayed in Table 2  and 3. Parity influenced behavior in the milking parlor (P<0.05).Primiparous females exhibited higher mean behavioral scores when compared with multiparous cows, likely because multiparous buffaloes are more accustomed to the parlor environment than their primiparous counterparts.These results corroborate those of Cavallina et al. (2008), who evaluated the behavioral response of Murrah buffaloes submitted to mechanical milking.Higher frequencies of agitated behavior including kicks, defecation and urination and knocking off the milking cluster for attaching or detaching were more frequent among primiparous cows.Thus, the greater the occurrence of these behaviors, the higher was the mean behavioral score for the individual temperament.De Rosa et al. (2003) found that frequency of kicking is an indicator of agitation and aggressiveness, while Das and Das (2004) confirmed that urination frequency is also strictly related to temperament, as well as being an indirect measurement of fear.
In addition to displaying greater mean behavioral scores, primiparous buffaloes also produced less milk, emphasizing the score's negative effect on production (Table 3).This corroborates Oliveira et al. (2010), who evaluated Murrah buffalo behavior in the presence of strangers and recorded reduced milk production among animals that defecated and urinated in the milking parlor.Moreover, Peters et al. (2007).The authors observed that low reactivity during milking, primarily regarding the frequency of defecation and urination, resulted in higher milk production and quality levels.Oliveira et al. (2010) also showed that milk production varied significantly when milking occurred in the presence or absence of strangers, with greater production when animals were familiar with the people present.
Although primiparous cows were more reactive and produced less milk, milking time, milk let-down and milk letdown rates did not vary in comparison with multiparous females (P>0.05).These findings agree with those of Fruchi et al. (2009), who also recorded superior milk production among multiparous buffaloes, but found that parity did not influence milking duration.It is important to note that higher milk production among multiparous cows may not be attributed to behavior, but also a result of physiological factors favoring productive performance when females reach maturity (COBUCI et al. 2000).
According to Bruckmaier et al. (1993) the central milk ejection where oxytocin inhibition from hipophyse occurs with frequency during various emotional stresses.Milking in unknown ambient resulted in milk ejection inhibition mainly in primiparous cows when first time milked (BRUCKMAIER et al., 1992) since it is a new experience for these cows; but, a gradual transferring of these animals to unknown ambient make the oxytocin release to reach a normal level (BRUCKMAIER et al., 1996).
When evaluating behavioral scores separately and in conjunction with production characteristics, we found that the most reactive buffaloes, that is, those with higher scores, displayed greater mean milk production.However, Nayak and Mishra (1984), when comparing docile, agitated, nervous and aggressive cows, reported that docile females were easier to milk, handle and manage and produced more high-quality milk in relation to "nervous" or aggressive buffaloes.They also observed that docile females exhibited higher consumption of the concentrate, had lower time to reach milk let-down, shorter milking duration, larger milk flow rates and greater percentages of fat in their milk.
Although milk production demonstrated variations between different cows' behavior, there was no significant difference (P>0.05) between milking time and milk flow rate.Gonzaga and Lorenzo (2007) mention that the greater the milk production, the higher the milk flow rate is, therefore, more time is required for milking once they were positively correlated.The fact that milk letdown rate and milking time did not vary across the behavioral scores may be due to adequate animal handling in milking parlor.Table 4 presents behavioral scores means in the milking parlor and milk production characteristics of Murrah buffaloes in the first six weeks of lactation, showing increased milk production as lactation progresses.In the first third of lactation, milk production increases to a peak, declining as the lactation stage advances.The milk flow rate, milking time and behavior did not vary over the six weeks.
Regarding to behavioral score, the mean in the first week was expected to be higher than in the sixth, since cows were separated from their calves, management practice that influenced milk production for that week, which was the lowest in the study period.It was expected as lactation advanced would contribute to reducing behavioral scores due to cows' adaptation to the milking parlor environment.However, this did not occur, since buffaloes produced well even when displaying high behavioral scores.This may be explained by the good milking practices carried out in farm management practices.
When exposed to stressors, dairy cattle exhibit disruptions in milk ejection (Van Reenen et al., 2002).According to Negrão and Marnet (2003), during the first time milking, particularly among primiparous buffaloes cows secret and release catecholamine and a failure in releasing of oxytocin occurs with a decrease in blood flow to the mammary glands.
Tables 5 and 6 show correlations regarding behavioral score, mean surface skin temperature and production characteristics of primiparous buffaloes in the second week of lactation.For multiparous cows, correlations were made between the same variables, however no significant difference was found.Regardless of calving order, behavior was not significantly correlated (P>0.05) with the other variables, likely due to the short study period or small number of observations.These results corroborate those of Peters (2008), who reported that stress caused by inadequate management practices did not influence milk composition or SSC.It is possible that the short duration of treatment, where the stressor is the milking parlor environment, was not sufficient to alter behavioral scores.
With respect to mean surface temperature, a positive correlation was recorded for fat (P<0.05) and total solids (P<0.01) in primiparous females.These findings may be partially attributed to the fact that primiparous cows have lower milk production and, as a result, their milk contains higher concentrations of nutrients.
These findings contrast those of Peters et al. (2007), who observed that low reactivity during milking and reduced defecation and urination resulted in higher milk production and superior quality milk.Oliveira et al. (2010) also emphasize that the presence of strangers in the parlor environment has a significant negative impact on milk production.Moreover, Porcionato et al. (2005) reported that initial milking are stressful, while Tancin et al. (2001) state that modifying the environment provides favorable behavioral changes during the milking period.
Tables 7 and 8 depict correlations between behavior, temperature and production characteristics of primiparous and multiparous females in the sixth week of lactation.As in the second week, there was no significant correlation between behavior and milk composition, indicating that under these study conditions, regardless of calving order, behavioral score did not affect this production characteristic (P>0.05).Among multiparous cows, a negative correlation (P<0.05) was observed with milk production, that is, the higher the surface body temperature, the lower the animal's milk production.Breuer et al. (2000) found that negative interactions resulted in greater reactivity during milking, although milk production remained unaffected.However, this outcome was not corroborated by Hemsworth et al. (2000), who observed that negative interaction during milking showed significant negative correlation with milk production and fat and protein content.
Peters (2008) points out that negative attitude, such as shouting and hitting, may result in higher stress levels among the animals, compromising milk production and quality since it favors the accumulation of residual milk.The author also recorded no correlation between animal welfare indicators and the occurrence of mastitis, corroborated by findings in the present study.Nevertheless, milkers displaying aggression towards the animals may be negligent with regard to hygiene during milking, predisposing the herd to increased mastitis.However, Rosa (2004) argues that positive attitudes by themselves are not sufficient to improve behavioral indicators and animal wellbeing.

Primiparous
buffaloes exhibit greater restlessness during milking than their multiparous counterparts.
Behavior did not influence milk composition or SSC.Mean surface temperature is positively correlated to fat percentage and total solids in primiparous buffaloes, and negatively correlated with milk production in multiparous females.