J. Dairy Sci. 90:3671–3680 doi:10.3168/jds.2006-766 © American Dairy Science Association, 2007.
Sprinklers and Shade Cool Cows and Reduce Insect-Avoidance Behavior in Pasture-Based Dairy Systems P. E. Kendall,* G. A. Verkerk,† J. R. Webster,* and C. B. Tucker*‡1 *Animal Behavior and Welfare, AgResearch Ruakura Research Centre, Private Bag 3123, Hamilton, New Zealand †Dexcel Ltd., Private Bag 3221, Hamilton, New Zealand ‡Department of Animal Science, University of California, Davis 95616-8521
ABSTRACT The body temperature of dairy cows in pastoral systems during summer reaches a peak during and following the p.m. milking. Shade and sprinklers can be used separately or in combination at the milking parlor to reduce heat load. Farmers anecdotally report that the use of sprinklers reduces irritation from insects that occurs while cows are waiting for milking. Once daily, we assessed the effectiveness of short-term exposure to shade and sprinklers for cooling cows [via respiration rate and body (vaginal) temperature] and reducing insect-avoidance behaviors before the p.m. milking in a pasture-based dairy system. Head position was measured as an indicator of whether cattle were avoiding water from the sprinklers. Forty-eight HolsteinFriesian dairy cows were divided into 12 groups (4 cows per group, n = 3 groups/treatment) and were exposed to 1 of 4 treatments for 90 min before the p.m. milking: 1) shade, 2) sprinklers, 3) shade and sprinklers, or 4) uncooled control. Respiration rate was reduced by 30% with shade alone compared with controls [54 vs. 78 ± 2.3 (±SED) breaths/min, respectively]. Sprinklers alone (30 ± 2.3 breaths/min) and the combined effects of shade and sprinklers (24 ± 2.3 breaths/min) reduced the respiration rate by 60 and 67%, respectively, compared with controls. Shaded cows had lower body temperatures during the 90-min treatment period compared with controls (shade: 38.6°C; shade and sprinklers: 38.6°C; control: 38.9 ± 0.09°C). The decrease in body temperature of cows under sprinklers was more marked than for shade alone and remained lower for at least 4 h after milking (sprinklers: 38.7°C; shade and sprinklers: 38.6°C; shade: 38.9°C; control: 39.2 ± 0.10°C). The sprinkler treatment reduced the number of tail flicks (control: 12.6 vs. sprinklers: 6.6 ± 2.4 flicks/min) and hoof stamps (control: 4.4 vs. sprinkler: 2.2 ± 0.5 stamps/min). Cows exposed to sprinklers spent more time with their
Received November 16, 2006. Accepted April 9, 2007. 1 Corresponding author:
[email protected]
heads lowered compared with cows in the shaded and control treatments. The reductions in body temperature and respiration rate attributable to shade and sprinklers were greatest when the temperature-humidity index and heat-load index were ≥69 and 77, respectively, and cows benefited from cooling when these levels were exceeded. Key words: heat stress, respiration rate, body temperature, thermoregulation INTRODUCTION An increasing proportion of milk products on the global market originate from pasture-based farming systems (FAO, 2004). Consumer demand for organic milk products is growing and some organic standards, including those used by the USDA’s National Organic Program, require that cows spend no less than 120 d/yr on pasture (National Organic Standards Board, 2005). This reflects a growing perception that pasturebased systems are beneficial to dairy cattle welfare compared with barn-housed systems. A consequence of pasture-based dairy systems is that cows are exposed to extremes of weather. Hot weather, in particular, can decrease milk production, compromise reproductive performance, and impair animal welfare. The respiration rate and body temperature of lactating cows increases during acute heat exposure (Ominski et al., 2002; Spiers et al., 2004), and cows with darkcolored coats are more susceptible to heat than cows with light-colored coats (Hansen, 1990). Cows provided with access to shade while on pasture have a lower body temperature than those without shade (Davison et al., 1988). Recently, we found that dairy cows readily used shade when it was provided and produced an additional 0.5 kg of milk/d than cows without access to shade (Kendall et al., 2006). Regardless of access to shade, all cows reached a similar maximum body temperature after walking to the parlor for the p.m. milking. Cooling cows with both fans and sprinklers for 20 to 30 min after they walked to the milking parlor reduced both body temperatures and respiration rates in the short
3671
3672
KENDALL ET AL.
term (Valtorta and Gallardo, 2004). Furthermore, use of fans and sprinklers just prior to milking reduced body temperature for 2 to 4 h after milking compared with cows that received no cooling at the parlor (Araki et al., 1985). Collectively, these results suggest that cooling cows before the p.m. milking could be a practical and effective way to reduce the peak body temperature and respiration rate of cows in pastoral farming systems. Both shade and sprinklers are practical methods for cooling cows while they are waiting to be milked, but little is known about the effectiveness of providing a single daily short-term (e.g., 90-min) exposure to these cooling methods. We chose to explore these options, in part, because 40% of farmers in New Zealand already use sprinklers to cool their cows before the p.m. milking (Tucker et al., 2005), and guidelines are needed about the weather conditions in which sprinklers and shade should be provided. Farmers anecdotally report that sprinklers reduce irritation from insects while cows are waiting to be milked. Little is known about the behavioral responses, such as fly avoidance and other behavioral changes, associated with exposure to sprinklers. In winter, cows seek shelter, particularly when it is raining (Vandenheede et al., 1995), and when exposed to wind and rain, they are 5 times more likely to stand with their heads lowered compared with their sheltered counterparts (Tucker et al., 2007). Thus, head position may indicate whether cattle find sprinklers aversive and whether this aversion is related to weather conditions. Our objective was to assess the effects of shade and sprinklers, alone or in combination, on the physiological and behavioral responses of pasture-based dairy cows with different coat colors. Our specific interest was in the effectiveness of these cooling methods when applied only before the p.m. milking. Finally, the responses were evaluated over a range of summer conditions in a temperate climate to identify when cooling methods should be used. METHODS AND MATERIALS Experimental Design and Management of Cows All procedures were approved by the Ruakura Animal Ethics Committee in accordance with the New Zealand Animal Welfare Act 1999. All values presented in this section are mean ± SD. Forty-eight mixed-age (2 to 11 yr) Holstein-Friesian dairy cows in midlactation (175 ± 21 DIM) were used. The cows had an initial BW of 482 ± 49 kg and BCS between 3.5 and 5 on a scale of 1 to 10 (Roche et al., 2004). The coat color was determined from digital photographs of both sides of each cow using image analysis software (Scion Image version 4.0.2, Journal of Dairy Science Vol. 90 No. 8, 2007
Scion, Frederick, MD). Cows were categorized as either predominantly black (89 ± 6% black hair, mean ± SD) or both black and white (mixed; 67 ± 8% black hair). The cows were allowed to graze as a single herd at a research farm in Hamilton, New Zealand (latitude 37°47′ S, longitude 175°19′ E) in January and February (southern hemisphere summer). Each day, at approximately 1230 h, the cows were walked (590 ± 316 m, single journey) to a holding pen adjacent to the milking parlor and separated into 12 treatment groups (4 cows per group). These groups were balanced for milk production and coat color category (2 black and 2 black and white cows in each group) and were assigned 1 of 4 treatments: 1) shade, 2) sprinklers, 3) shade and sprinklers, or 4) uncooled control. Treatments were applied for approximately 90 min/d (81 ± 8 min/d) prior to milking, which occurred between 1430 and 1530 h, for 35 consecutive days. One day of observation was omitted because of heavy rainfall. During the 90-min treatment period, each group was kept in a 2.5 × 2.5 m pen. The space allowance for each group was approximately 1.25 m2/cow. This level of stocking density simulated normal conditions in holding pens before milking. Cows in the control treatment had no shade or sprinklers. The shade treatment consisted of a 2.5-m-high structure (3 × 3 m) with a double thickness of shade cloth on the roof that excluded >93% of UV light, and with 3 of the sides covered with a single layer of shade cloth. The sprinkler treatment was created with an oscillating sprinkler fixed 1.9 m above the ground (approximately 0.5 m above the cows) that delivered approximately 75 ± 74 mm of water/h (water temperature: 22 ± 1°C). Cows in the sprinkleronly treatment had no protection from the sun. All cows were treated with a fly repellent and insecticide (Blaze, Schering-Plough Animal Health Ltd., Wellington, New Zealand) twice during the course of the experiment. Sampling and Measurement Air temperature, relative humidity, wind speed, solar radiation, and rainfall were recorded at 10-min intervals using a portable weather station (Campbell Scientific, Scott Technical Instruments Ltd., Hamilton, New Zealand) located in a pastured area adjacent to the holding pens. The microclimate in each treatment area was measured with these weather stations in the days following the experiment. Temperature-humidity index (THI) and heat-load index (HLI) were used as composite measures of thermal comfort and were calculated as reported previously (Kendall et al., 2006). Body temperature was measured over 7-d periods at 10-min intervals in 3 cows in each group (2 black and white cows and 1 black cow) with a modified vaginal
3673
RESPONSE TO COOLING IN GRAZING DAIRY COWS
controlled internal drug release insert (InterAg, Hamilton, New Zealand) fitted with a microprocessor-controlled data logger (Minilog TX, Vemco Ltd., Shad Bay, Nova Scotia, Canada). Thus, throughout this experiment, the term “body temperature” refers to vaginal temperature. Temperature loggers were inserted into the vaginal cavity on a 7-d in and 3-d out cycle to minimize irritation. This schedule resulted in 25 d of body temperature measurements in total. Respiration rate (flank movements/30 s) was recorded before treatments were applied in a subset of cows (n = 24, same cows on all days after the walk to the parlor) and in all cows after the 90-min treatment period. The frequency of tail flicks (defined as the number of times the tail crossed the midline of the leg) and hoof stamps (defined as the number of times any hoof was lifted off the ground) were recorded continuously for 30 s approximately every 12 min for 60 min during the 90-min treatment period, for a total of 150 s of observations/cow per day. Head position was recorded before and after each of these 30-s sessions using instantaneous scan sampling (Martin and Bateson, 1993) and was categorized as either low (chin below the brisket), high (poll above the wither), or middle (between low and high). Over the course of the experiment, 11 observers recorded respiration rate and behavioral traits. Interobserver agreement was calculated as the percentage of agreement and was 96% for respiration rate, 95% for tail flicks, 83% for hoof stamps, and 93% for head position. Although respiration rate and behavioral traits were measured over 30-s periods, the results are expressed on a per minute basis to facilitate comparison with other studies. Finally, daily milk production was measured at the a.m. and p.m. milkings, and a pooled milk sample was collected twice weekly to determine the percentages of fat, CP, and lactose using an infrared milk analyzer (FT120; Foss Electric, Hillerød, Denmark). Milk production from the a.m. milking was considered separately, because longer term (after the p.m. milking) effects of the treatment were more likely to be detected at that time. Somatic cell count was measured on the same twice-weekly samples using an automated cell counter (Fossomatic 5000; Foss Electric). Statistical Analyses The effects of shade and sprinklers were tested with a 2 × 2 design (4 treatments of shade, sprinklers, a combination of shade and sprinklers, and uncooled controls). The group served as the experimental unit. Data were averaged across the collection period for each treatment (25 d for body temperature; 34 d for behavior, milk production, and respiration rate). For body tem-
perature, data were divided into four 2-h time blocks (1050 to 1240, 1250 to 1440, 1450 to 1640, and 1650 to 1840 h) for analysis of treatment differences, and 3-h time blocks (0010 to 0300, 0310 to 0600, . . . 2110 to 0000 h) to look at daily differences in body temperature among treatment groups. Somatic cell counts were log10 transformed to normalize the distribution. The effect of treatment (shade, sprinklers, shade by sprinkler interaction, 1 df each) was tested against the group term (8 df) and the effects of coat color category (percentage of black or mixed; interactions of coat color × shade, coat color × sprinklers, and coat color × shade × sprinklers, 1 df each) were tested against the cow term (32 df) in either GenStat version 8.1 (Payne, et al., 2005) or SAS version 9.1 (SAS Institute, 1999). For all measures of milk production and composition, a covariate was used to correct for lactational performance before the start of the experiment, and inclusion of these covariates reduced the error to 7 df for the group term and 31 df for the cow term. The influence of weather on the response to treatment was assessed by using daily information from each cow to calculate the slope of the relationship between the response variable (body temperature, respiration rate, and behavior) and THI and HLI variables during the 90-min treatment period. The slopes were compared in the same model as described above. The effect of weather on milk production was explored by calculating the partial regression slopes of the relationship between daily milk yield (defined as the p.m. and following a.m. milk sample) for each cow and the THI or HLI for the day of the p.m. milking. For this analysis, weather information from the 24 h before milking was used. A linear adjustment was made for the effect of time (days) to allow for the declining milk production and treatment differences analyzed by the model described above. The influence of weather on body temperature and respiration rate responses to the treatments were investigated by visually inspecting Figure 1 to identify the threshold values for THI and HLI levels when sprinklers, shade, or both should be used (intersection between regression lines). Regression lines intersected only in the body temperature figure; thus, the model described above was repeated using only body temperature and data from days when THI was
