Int J Biometeorol (2008) 52:419429 DOI 10.

1007/s00484-007-0136-1

ORIGINAL PAPER

Environmental factors affecting feed intake of steers in different housing systems in the summer
H. Koknaroglu & Z. Otles & T. Mader & M. P. Hoffman

Received: 28 May 2007 / Revised: 14 November 2007 / Accepted: 15 November 2007 / Published online: 18 December 2007 # ISB 2007

Abstract A total of 188 yearling steers of predominantly Angus and Hereford breeds, with mean body weight of 299 kg, were used in this study, which started on 8 April and finished on 3 October, to assess the effects of environmental factors on feed intake of steers in various housing systems. Housing consisted of outside lots with access to overhead shelter, outside lots with no overhead shelter and a cold confinement building. Ad libitum corn, 2.27 kg of 35% dry matter whole plant sorghum silage and 0.68 kg of a 61% protein-vitamin-mineral supplement was offered. Feed that was not consumed was measured to determine feed intake. The temperature data were recorded by hygro-thermographs. Hourly temperatures and humidity were used to develop weather variables. Regression analysis was used and weather variables were regressed on dry matter intake (DMI). When addition of a new variable did not improve R2 more than one unit, then the number of variables in the model was truncated. Cattle in confinement had lower DMI than those in open lots and

those in open lots with access to an overhead shelter (P < 0.05). Cattle in outside lots with access to overhead shelter had similar DMI compared to those in open lots (P =0.065). Effect of heat was predominantly displayed in August in the three housing systems. In terms of explaining variation in DMI, in outside lots with access to overhead shelter, average and daytime temperatures were important factors, whereas in open lots, nocturnal, peak and average temperatures were important factors. In confinement buildings, the previous days temperature and humidity index were the most important factors explaining variation in DMI. Results show the effect of housing and weather variables on DMI in summer and when considering these results, cattle producers wishing to improve cattle feedlot performance should consider housing conditions providing less stress or more comfort. Keywords Housing . Dry matter intake . Heat stress . Temperature . Humidity

Introduction
H. Koknaroglu (*) Department of Animal Science, Suleyman Demirel University, Isparta, Turkey e-mail: hayati@ziraat.sdu.edu.tr Z. Otles Frontier Science and Technology Research Foundation, Madison, WI, USA T. Mader Department of Animal Science, University of Nebraska, Lincoln, NE, USA M. P. Hoffman Department of Animal Science, Iowa State University, Ames, IA, USA

As a rule of thermodynamics the performance of cattle depends on how much energy they consume and how much energy they spend for maintenance. Cattle as homeotherm animals live in a dynamic environment and interact with it (Hahn 1999). The environment surrounding cattle often dictates their maintenance energy requirement and their feed intake (Delfino and Mathison 1991). Thus, controlling the factors affecting dry matter intake and energy expenditure of cattle is an important means to improve the performance and efficiency of the cattle production. One way to control the effect of environment on cattle is to provide housing to minimize stress (Mitlhner et al. 2002).

420

Int J Biometeorol (2008) 52:419429

Several researchers comparing the different housing systems reported that providing overhead shelter in open lots has resulted in better performance and efficiency than open lots (Leu et al. 1977; Muhamad et al. 1983; Koknaroglu et al. 2005). However, the same researchers also found that cattle housed in confinement had a poorer performance than cattle fed in open lots, and open lots with access to overhead shelter. There has been very limited research carried out on defining the causes of difference in performance among different housing systems. Thus, the purpose of this study was to investigate the effect of environmental variables on feed intake of cattle fed in open lots, confinement and open lots with access to an overhead shelter.

Materials and methods Management and feeding The study was initiated on 18 April and was finished on 3 October, and was conducted at the Allee Experimental Farm located in the center of the northwest quarter of the state of Iowa. At the beginning of the study, the yearling steers, predominantly Angus and Hereford, were individually identified and weighed over two consecutive days. The average of these two weights was used as the starting weight (mean BW=299 kg). All steers were individually ear-tagged and, after initial weighing, the cattle were stratified by weight and color pattern and assigned to each of three types of housing facilities. Steers were weighed at 28-day intervals. Average daily gain was determined as final or intermediate weight minus initial weight divided by the number of days in the feedlot or during the period. Housing used in this study consisted of three outside lots with access to overhead shelter, three outside lots with no overhead shelter and a semi-enclosed (open-front) cold confinement building containing four lots. Open lots and open lots with access to an overhead shelter treatments had 18 steers per pen and confinement treatment had 20 steers per pen, thus a total of 188 steers were used for the experiment The outside lots were 30.5 m long 10.7 m wide, oriented in a north-south axis with a surface gradient of 4% slope to the south and providing approximately 18.1 m2 per steer. Automatic waterers, with individual metered bowls, were placed in the fence line between each pair of outside lots. The fencing on three sides of each lot consisted of five 0.95-cm stranded cables on wooden posts approximately spaced 2.7 m apart. In all instances, concrete was provided around the waterers, feed bunks, and other high-traffic areas. Wooden feed bunks were located in the fence line along the driveway and provided 0.59 m of linear feeding space per steer.

Overhead shelter consisted of a large barn with the hayloft floor removed to allow for air circulation. It provided 4.64 m2 per steer and was located north of the three outside lots. Entrance to it was through doors located on its south side. The semi-enclosed (open-front) confinement building was 24.38 m long 12.19 m wide and was built of a wood framework covered with ribbed aluminum siding over a 3.8cm layer of styrofoam with vapor barrier. Three sides of the building were enclosed, leaving only the south side open. The single slope of the roof at the south eave was 5.49 m and it was 3.05 m at the north eave. There was a 3.81-m-wide alley running along the north side. It was accessible through large sliding doors at each end which could be left open during hot weather for ventilation purposes. Feed bunks were located in the north edge of the pens. Concrete fenceline feed bunks were located between the pens and the feed alley. Two of the four pens in confinement had concrete slats and two had aluminum slats over an oxidation ditch 1.12 m deep. Approximately 28 cm of feedbunk space was provided per steer. Automatic waterers were located near the open side of the building. The floor surface was either a sloping solid concrete floor with flush flumes, slots over a multiflume floor or slots over a shallow pit. Depending on the floor design, manure materials were either continuously or periodically flushed into a lagoon. Confinement building provided approximately 4.13 m2 per steeer. Space allocations and type of flooring were different for the three housing systems. However, they were similar to what is generally recommended and practical for cattle housing systems in the Midwest. Consequently, they were not considered to have a direct influence on the results of this study (Pusillo et al. 1991). Steers were offered on a daily basis ad libitum corn, 2.27 kg of 35% dry matter whole plant sorghum silage and 0.68 kg of a 61% protein-vitamin-mineral supplement. Feeds that were not consumed by steers (ort) were removed from bunks and, after drying in the oven, they were weighed to determine actual daily feed intake. Actual feed intake was calculated as feed provided minus feed not consumed. In order to determine dry matter of feed samples, feed samples sampled every 3 days were oven dried and dry matter content was determined. Because dry matter intake (DMI) increases with increasing body weight, DMI was corrected for confounding effect of body weight. The formula DMI ; kg =d 4:54 0:0125 IBW states that there is an increase of 1.25 kg in DMI for every 100 kg increase in initial weight (NRC 1996). Thus, actual DMI was adjusted by applying the NRC formula by using intermittent weights. Adjusted dry matter intake was calculated as DMIadjusted DMIactual 0:0125 actual body weight initial weight based on 28-day weights.

Int J Biometeorol (2008) 52:419429

421

Temperature and relative humidity readings The temperature data were recorded by hygro-thermographs. In the sheltered area, the hygro-thermograph was located approximately 1.8 m above ground level. The hygrothermograph in the open lot was located along the fence line of the lot about 1.2 m above the ground. In the confinement building, the hygro-thermograph was located over the pens about 3 m above the floor. Hygro-thermographs were mounted in a white painted box with louvered sides. The reason for this was to insure that no direct radiation would fall upon the instrument and

also to allow enough air flow across the instrument to minimize trapping of hot or cold air. The hygro-thermograph had an 8-day clock which powered a cylindrical drum around which a 7-day calibrated chart was mounted. This provided for a continuous ink tracing of temperature for a 7-day period. The chart had increments of 2 h which allowed 12 temperature readings, which were accurate to approximately 1F (0.55C). In order to examine effect of temperature and relative humidity readings taken at different times, variables below were developed.

Average temperature temp00 temp02 ::::::::: temp22=12; Nocturnal temperature temp22 temp00 temp02 temp04 temp06=5; Daytime temperature temp08 temp10 temp12 temp14 temp16 temp18 temp20=7; Peak temperature temp14 temp16=2;

where temp00 is the temperature reading at midnight, temp02 is the temperature reading at 0200 hours of that day and others are respective readings.
Temperature Humidity IndexTHI Average temp 0:55

0:55 RH =100 Average temp 58:8; where RH is the relative humidity as percentage. For a given day, the previous days temperature was the average temperature of the previous day. With the same approach, the previous days temperature humidity index (THI) for the given day was the THI of previous day. Relative humidity calculation RH calculated for each housing was determined using Sioux City daily RH values, maximum temperature and minimum temperature values. The environment saturation vapor (ea) pressure was calculated using Clasius Clapeyron Equation and RH from Sioux City. RH is ratio between ea/ es when expressed as a percentage. RH 100 ea es

minimum temperatures from Sioux city have lead us to find the corresponding times at open shelter. Using those times, we have corresponding temperatures to estimate es for each shelter, then respectively the daily RH values. The es (Pa) calculated (Stull 1988)   17:67 T 273:16 es 611:2 exp T 29:66 where Ta(K) is air temperature. Statistical analysis Since the data in April and October were not enough to conduct statistical analysis, data were analyzed for the months of May, June, July, August and September. Regression analysis was used to determine important variables affecting DMI and for this purpose weather variables such as hourly temperature readings and other variables were regressed on DMI. When addition of a new variable did not improve R2 more than 1 unit then the number of variables in the model was truncated. Regressions were applied separately for each month and for each housing system. In order to test for cattle performance, the data were analyzed using the General Linear Model procedure of SAS (1999) by including housing in the model and PDIFF statement was used to compare housing means for dependent variables. In order to find partial R2, standard estimates from regression analysis were obtained and then squared. After all of the squared terms were added, each squared estimate

The es (Ta) and ea are saturation and actual vapor pressure of air, respectively. The estimated es from the shelter temperature readings along with the environment air pressures from Sioux City were used to estimate the RH values for each shelter type. We have found the temperatures which are used in the RH calculations in the following manner. Maximum and

422

Int J Biometeorol (2008) 52:419429

was divided by the sum of the squared terms and thus the relative contribution of each parameter to each R2 was found.

Results and discussion Performance variables of steers are given in Table 1. Since purpose of the study was to examine the effect of environmental factors on daily DMI, only DMI will be examined in this section. Cattle in confinement had lower DMI than those in open lots and those in open lots with access to an overhead shelter (P <0.05). Cattle in open lots with access to an overhead shelter had similar DMI compared to those in open lots (P =0.065). In terms of DMI in different housings, similar findings were reported by Leu et al. (1977), Muhamad et al. (1983) and Koknaroglu et al. (2005), who found that cattle fed in confinement consumed significantly less daily dry matter than those in open lots or open lots with access to overhead shelter In these studies, cattle in open lots or open lots with access to overhead shelter had similar daily DMI. Even though previous studies reported differences in terms of DMI in different housings they did not investigate why this difference occurred. The purpose of this study was to investigate and identify effect of environmental variables in different housings on daily DMI of cattle. Thus, the feeding period, which started at the end of April and finished at the beginning of October, was divided into months and the effects of environmental variables in different months and different housings were examined. Temperature variations in different housings are presented in Fig. 1. In general, the lowest temperature was observed at 0600 hours in three housings and temperature started increasing until 1600 hours when it started decreasing, thus making 1600 hours the hottest hour of the day (Fig. 1). Temperature readings showed that in shelter and confinement July was the hottest and May was the coolest month, whereas in open lot August was hottest and May was lowest month (Table 2). In the first month of the study (May), dry matter intake of cattle in open lot with access to an overhead shelter was 7.59 kg/day (Table 3) and was affected by daytime temperature which constituted 65.84% of the total variation.
Table 1 DMI and performance of steers (SE) throughout the experiment in different housings

The exact reason for this is not known, although the adaptation period might have played a role in this. Probably, the acclimation and the temperature variability could explain these results. One observation with these cattle was that temperature at 1000, 1400 and 2000 hours contributed to the explanation of total variation. Temperature at 1000 hours coincided with the abrupt increase in temperature, whereas temperature at 2000 hours coincided with a decrease in temperature (Fig. 1a). A behavior study of cattle in Iowa conducted at the same experimental station by Hoffman and Self (1973) showed that in summer cattle preferred lying under shelter between 0900 hours and 1800 hours and preferred lying outside between 1800 hours and 0900 hours. Thus, at 2000 hours, cattle might have left the overhead shelter and gone to the open lot where ambient temperature was lower. The effect of outside temperature clearly illustrates this because in the open lot abrupt temperature increase starts after 0800 hours and the temperature starts decreasing after 1800 hours (Fig. 1c). Dry matter intake of cattle in open lot with access to overhead shelter was highest in June (8.08 kg/day; Table 3). Average temperature and daytime temperatures were the two most important variables explaining variation (65.33%) in June. Legates et al. (1991) reported that correlations of ambient temperature with body temperatures and respiratory rates were highest, those for radiation were next, followed by vapor pressure and air movement. It seems that cattle during this period did not feel any heat stress because temperature exceeding 25C was only at 1600 hours. Effects of heat stress on performance are mainly a result of a decrease in DMI, and research showed that depressive effect of ambient temperature is displayed when it exceeds 25C (Morrison 1983; Hahn et al. 1992; Hahn 1995). Using radiotelemetry measurement of body temperatures of feedlot steers, Lefcourt and Adams (1996) found that the threshold for increase in daily maximum body temperatures was 25.6 and 20.6C for maximum ambient temperature and mean ambient temperature, respectively. Other important variables affecting DMI in June were the previous days THI and the previous days temperature (Table 3). Brown-Brandl et al. (2005) found that the current days response was related to the previous days. Hahn et al. (1987) reported a lasting effect of

Variables Initial weight (kg) Final weight (kg) Days on feed Daily DMI (kg/day) Average daily gain (kg/day) Feed efficiency (kg feed/kg gain)

Open lot 2990.08 4955.67b 158 7.170.18a 1.240.04b 5.800.14ab

Shelter 2990.08 5175.67a 158 7.38 0.21a 1.370.04a 5.380.14a

Confinement 2990.07 4794.91b 158 6.690.21b 1.130.03b 5.960.12b

ab

Means with different superscripts are significantly different (P <0.05)

Int J Biometeorol (2008) 52:419429 Fig. 1 Temperature variations for months in shelter (a), open lot (b) and confinement (c)
30 25 Temperature, C
o

423

20 15 10 5 0 0 2 May 4 6 June 8 10 12 14 August 16 18 20 22 Hours July

September

30 25 Temprature, C
o

20 15 10 5 0 0 2 May
30 25

6 June

10 July

12

14 August

16

18

20

22

Hours September

Temperature, C

20 15 10 5 0 0 2 May 4 6 June 8 10 12 14 August 16 18 20 22 Hours July

September

temperature and found that current responses were affected by temperatures for up to 60 h preceding. In July, average temperature was the sole contributor to the explanation of variation followed by daytime temperature and both contributing 95% of the variation. July was the hottest month of the year for cattle in open lot with access to overhead shelter. It should be noted that DMI of these cattle decreased in this month compared to DMI in June. This result would be expected because THI exceeded 75 and this might have had an effect on DMI. We did not measure respiration rate, though results show that respiration rate is the most appropriate indicator of thermal stress to monitor because it is consistently affected in all THI categories, and there is little or no lag associated with it. (Brown-Brandl et al. 2004). Experiments showed that ambient temperature is the major atmospheric influence on body temperatures and respiratory rate under most hot conditions in the southeast-

ern US, with solar radiation, vapor pressure, and air movement generally following in order of importance (Seath and Miller 1946; Shrode et al. 1960). In August, DMI of cattle in open lot with access to an overhead shelter decreased compared to DMI in July (7.85 vs 7.27 kg/day; Table 3). Daytime temperature was the most effective variable affecting DMI. Additional to this variable, the previous days temperature and peevious days THI, and temperature at 1800 and 2200 hours also contributed to explaining variation in DMI. Weather conditions for these cattle in this month were not as severe as in July. THI was lower than 74 and there were no temperature readings exceeding 25C. The exact reason for lower DMI in this month is not known, though higher nocturnal temperature in the open lot might have affected heat dissipation from the cattle. Body temperature is determined by heat input from metabolic heat production

424 Table 2 Weather variables (SD) for housings by months Housings Months May Shelter Average temperature (C) Nocturnal temperature (C) THI Previous days THI RH (%) Daytime temperature (C) Peak temperature (C) Previous days average temperature (C) Open lot Average temperature (C) Nocturnal temperature (C) THI Previous days THI RH (%) Daytime temperature (C) Peak temperature (C) Previous days average temperature (C) Confinement Average temperature (C) Nocturnal temperature (C) THI Previous days THI RH (%) Daytime temperature (C) Peak temperature (C) Previous days average temperature (C) June July

Int J Biometeorol (2008) 52:419429

August

September

16.434.58 14.634.09 61.8510.80 61.5511.03 56.8012.68 17.715.15 19.496.05 16.324.67 15.894.87 12.914.26 60.4811.15 60.1611.37 49.5112.45 18.025.68 20.196.62 15.774.96 15.765.57 13.664.12 62.0213.34 61.6513.62 54.6012.12 18.477.80 19.396.06 16.325.68

20.852.97 18.162.87 72.467.24 71.757.50 56.9610.96 22.773.42 24.884.11 20.563.07 20.653.21 16.892.81 71.517.54 70.737.67 49.0112.34 23.323.99 25.725.07 20.323.25 20.183.17 17.312.94 70.817.55 70.097.69 55.8011.06 22.243.71 24.374.55 19.893.21

21.842.99 19.783.03 75.137.39 75.397.43 61.7611.12 23.313.16 25.073.53 21.953.01 21.593.13 18.143.56 74.127.59 74.397.70 54.8213.57 24.043.32 26.243.92 21.713.20 21.563.00 19.323.10 74.467.33 74.657.39 61.9510.73 23.183.20 24.953.72 21.673.04

21.174.13 19.373.62 73.089.67 73.209.63 59.4211.50 22.474.65 24.385.25 21.184.12 21.784.38 18.763.66 74.2110.05 74.389.99 53.2313.23 23.935.15 26.355.95 21.834.36 21.263.99 19.133.36 73.249.22 73.519.11 59.2911.71 22.784.68 24.845.57 21.333.94

16.464.27 14.424.62 61.919.99 62.5210.10 56.6117.70 17.924.51 20.175.00 16.704.32 16.724.39 13.975.04 62.4110.03 63.1710.18 50.5315.30 18.694.62 21.325.35 16.964.33 16.374.17 14.204.39 61.619.70 62.109.77 56.3314.98 17.914.56 20.305.25 16.534.18

Average temperature Aaverage values of temperatures measured at 0000, 0200, 0400, 0600, 0800, 1000, 1200, 1400, 1600, 1800, 2000 and 2200 hours. Nocturnal temperature Aaverage values of temperatures measured at 2200, 0000, 0200, 0400 and 0600 hours THI Temperature humidity index: Average temp 0:55 0:55 RH =100 Average temp 58:8; where RH is the relative humidity as percentage Previous days THI For a given day, it is the average temperature of the previous day Daytime temperature Average values of temperatures measured at 0800, 1000, 1200, 1400, 1600, 1800 and 2000 hours Peak temperature Average values of temperatures measured at 1400 and 1600 hours Previous days average temperature For a given day, this is the average temperature of the previous day

and solar radiation and by heat output through evaporative and nonevaporative avenues. When heat loss does not attain heat gain, heat is stored, thus causing an increase in body temperature (Brosh et al. 1998). Cool nighttime temperatures help animals dissipate heat gained during the day (Bianca 1961; Kabuga 1992). If an animal does not get enough cooling during nighttime hours, body temperature of the animal would not return to normal levels resulting in a higher peak the following day (Brown-Brandl et al. 2005). Another reason for cattle in August to have lower DMI could be acclimation of cattle. Blackshaw and Blackshaw (1994) reported that, following exposure to heat, cattle appear to acclimatize within 27 weeks. Temperature at 1800 hours in the overhead shelter corresponds with the highest temperature in August. The

reason for this temperature to affect DMI could be the heat load imposed on the cattle. Another observation is that temperature at 2200 hours contributed to explanation of variation. The reason for this could be the lag between ambient temperature and body temperature. Data-dependent system time series analyses of tympanic temperature records indicated the effect of ambient conditions of previous 24 h on current tympanic temperature (Parkhurst and Hahn 1987). Thus, the effect of higher temperature at 1800 hours might have been displayed at 2200 hours. Lowest dry matter intake of cattle in open lot with access to an overhead shelter occurred in September (6.04 kg/day; Table 3). The exact reason for this is not known because it seems that there is no factor that would cause heat stress on cattle. Daytime temperature and current day THI had

Int J Biometeorol (2008) 52:419429 Table 3 Important variables affecting DMI, partial R2 of each variable for shelter housing by months and DMI by months Shelter May Times of temperature readings (hours) 0000 0200 0400 0600 0800 1000 1200 1400 1600 1800 2000 2200 1.88 1.02 0.61 0.26 4.32 6.00 0.58 0.14 5.80 0.08 28.05 4.36 1.37 0.14 41.06 1.53 7.590.62 62.36 9.80 0.03 16.84 7.34 8.080.36 68.71 0.49 58.25 0.33 0.07 0.12 3.47 0.16 7.850.57 64.71 6.40 5.00 15.33 2.58 7.80 1.13 June July August

425

September 0.94 5.01

0.55

1.46

3.90 9.37 8.12

Average temperature (C) Nocturnal temperature (C) THI Previous days THI RH (%) Daytime temperature (C) Peak temperature (C) Previous days average temperature (C) DMI (kg/day) R2 See Table 2 for explanation of weather variables

4.03 22.11 4.67 7.270.91 43.67

13.90 3.69 13.77 4.54 6.041.58 88.94

considerable contributions to total R2. Among single temperature readings, temperatures at 1800, 1200, 1400, 2200, 0200, 1000 and 2000 hours were important ones contributing to total R2. Sudden temperature increase in the shelter started at 1000 hours and the highest temperature was attained at 1600 hours. Thus, a prolonging effect of temperature on body temperature might have played a role in this. When the general trend among months was considered for cattle in open lot with access to an overhead shelter, it could be seen that average temperature and daytime temperature are most important factors explaining variation in DMI. These results are in agreement with Legates et al. (1991), who reported that the highest correlation of body temperature and respiratory rates was with ambient temperature. Dry matter intake of cattle in open lot was 7.47 kg/day in May (Table 4). Average temperature and peak temperature were the two most important factors explaining the variation in DMI (Table 4). Other important factors were temperature at 1600 hours and nocturnal temperature. Considering the highest temperature occurred at 1600 hours, and temperature at 1600 hours is part of peak temperature, their combined effect shows that peak temperature is the most important factor affecting variation in DMI of cattle in open lot in May. Nocturnal temperature also had an effect on DMI and this shows the effect of heat dissipation on

thermoregulation of cattle. Compared to May, DMI of cattle in open lot in June increased a little (7.65 kg/day; Table 4). In June, nocturnal temperature had the highest effect on DMI variation (Table 4). The second biggest contributor was temperature at midnight, although that temperature at midnight is a part of nocturnal temperature, thus their combined effect clearly shows the effect of heat dissipation on cattle performance. Brody et al. (1955) reported that even though daily maximum temperatures exceeds 40C for a few hours each day, cool nights that help stored heat to be radiated to the surroundings allows dairy cattle to produce at near optimal levels. Experiments by Givens et al. (1966) suggested that relatively high daytime temperatures had no effect on performance of cattle when temperature at night was low. Peak and average temperatures also explained some variation (Table 4). Peak and average temperatures for June were close to the threshold temperatures that causes increase in body temperature reported by Lefcourt and Adams (1996). The general trend seen in June is also observed in July except average temperature did not affect DMI (Table 4). In July, combining the effect of temperature at 0200 and 0400 hours with nocturnal temperature and temperature at 1400 hours with peak temperature leads us to conclude that nocturnal temperature and peak temperature are the most important factors affecting DMI of cattle in open lot. Compared to DMI in July, a decrease in DMI of

426

Int J Biometeorol (2008) 52:419429

Table 4 Important variables affecting DMI, partial R2 of each variable for open housing by months and DMI by months Open lot May Times of temperature readings (hours) 0000 0200 0400 0600 0800 1000 1200 1400 1600 1800 2000 2200 1.58 June 11.86 3.31 8.18 0.35 1.70 0.63 0.53 0.13 0.07 4.34 6.03 1.32 0.57 0.64 17.23 5.48 1.82 2.97 13.05 16.32 July August 2.45 September 15.41 24.99 9.71 10.88 0.12 4.55 0.16 0.32 5.85 40.29 5.05 0.63

Average temperature (C) Nocturnal temperature (C) THI Previous days THI RH (%) Daytime temperature (C) Peak temperature (C) Previous days average temperature (C) DMI (kg/day) R2 See Table 2 for explanation of weather variables

42.29 3.62 0.32 4.52 7.18 2.49 7.710.62 71.94 5.25 0.12 0.08

14.59 7.470.62 47.79

9.32 7.650.33 75.86

4.25 6.991.28 45.391

5.911.47 71.59

cattle in open lot in August was observed (7.71 vs 6.99 kg/ day; Table 4). An interesting observation with DMI of cattle in open lot in August is that the contribution of hourly temperature started at 0800 hours and in an increasing fashion it lasted until 1400 hours (Table 4). These values coincided with the fashionable increase in ambient temperature (Fig. 1b). Daytime time temperature and previous days temperature also had an effect on DMI. This would be expected because average temperature exceeded 20.6C which is reported as the threshold temperature for maximum body temperature (Lefcourt and Adams 1996). The previous days temperature might have an effect on the current days DMI because nocturnal temperature in August was the highest of all months and the ability of cattle to dissipate heat during the night where temperature is coolest might have had an effect on the next days performance. Research showed that lower nighttime temperatures can help cattle tolerate higher ambient temperatures during the day (Mendel et al. 1971; Wiersma and Stott 1974; Scott et al. 1983). Dry matter intake of cattle in open lot was lowest in September (5.91 kg/day; Table 4). The exact reason for this is not known, but a lasting effect of heat stress could be the reason (Blackshaw and Blackshaw 1994). Among variables affecting DMI in open lot, temperature readings comprising nocturnal temperature

were important factors explaining variation in DMI (Table 4). Other variables were current days THI and temperature at 1400 hours. When considering all months, results show that nocturnal temperature and peak temperature are the two important variables affecting DMI of cattle in open lot. This shows the heat load cattle face in open lot due to solar radiation and the effectiveness of nocturnal temperature to dissipate heat to bring body temperature back to normal range. Cattle in confinement in May had DMI of 7.26 kg/day (Table 5). Important variables explaining variation in DMI in order of importance were average temperature, previous days THI, previous days temperature, THI of current day and nocturnal temperature. Combined together, these variables comprised near 93% of total R2 (Table 5). This shows the importance of temperature and humidity along with nocturnal temperature on cattle in confinement. DMI of confinement cattle in June increased a little (7.38 kg/day; Table 5). Previous days temperature, previous days THI, current days THI, daytime and nocturnal temperatures were important variables explaining variation in DMI. Peak temperature, daytime temperature and THI were lower than the level that would cause stress in cattle. The reason for the previous days temperature and THI to affect cattle could be the confinement absorbing heat during the day and

Int J Biometeorol (2008) 52:419429 Table 5 Important variables affecting DMI, partial R2 of each variable for confinement housing by months and DMI by months Confinement May Time of temperature readings (hours) 0000 0200 0400 0600 0800 1000 1200 1400 1600 1800 2000 2200 June July August

427

September 1.33

2.03 1.27 2.00 0.84 1.19 1.36 6.64 0.49 0.26 0.62 1.03 1.34 18.51 7.05 8.55 15.45 3.17 3.60 8.89 16.16 5.13 14.44 7.260.53 68.88 16.40 7.360.19 53.68 5.23 28.92 7.380.32 72.55 1.07 25.22 6.270.80 72.99 1.26 9.27 21.11 10.85 28.71 0.98 1.33 11.44 0.01 22.58 14.16 5.0971.35 55.89 1.01 0.96 0.20 1.89

Average temperature (C) Nocturnal temperature (C) THI Previous days THI RH (%) Daytime temperature (C) Peak temperature (C) Previous days average temperature (C) DMI (kg/day) R2 See Table 2 for explanation of weather variables

radiating heat during the night, and thus interfering with dissipation of heat by the cattle during night. Dry matter intake of cattle in confinement in July was similar to that in June (7.38 kg/day; Table 5). Previous days temperature and previous days THI were the two most important variables explaining variation in DMI and their partial contribution to total R2 was higher than that in June (Table 5). Current days THI and peak temperature were other important variables. In July, average temperature exceeded 20.6C and peak temperature approached the threshold temperature for stress (25C) and THI exceeded 74 Even though these values were higher than in June, it did not affect cattle DMI. This shows the adaptability of cattle when stress factors are imposed on them. There was a drastic decrease in DMI of cattle in confinement in August (6.27 kg/day). Compared to the DMI in July, it was more than 1 kg (1.11 kg/day). This observation is not peculiar to cattle in confinement because cattle in open lot and cattle in open lot with access to an overhead shelter also had lower DMI in August. However, the decrease in DMI in confinement cattle was more pronounced. Previous days THI and previous days temperature were the two most important variables explaining variation in DMI (Table 5). Compared to July, the previous days THI in August had more effect in explaining variation in DMI. Even though temperature and THI in August was lower than those in July, the decrease in

DMI in August could be the extended effect of heat stress on cattle (Blackshaw and Blackshaw 1994). Other important variables were nocturnal temperature and temperature at 0600 hours. Temperature at 0600 hours corresponded to the lowest temperature of the day (Fig. 1c) and, combined with nocturnal temperature, this shows the effect of heat dissipation at night. Among all months, cattle in confinement had lowest DMI in September (5.09 kg/day; Table 5). Daytime temperature, previous days temperature and previous days THI were important variables explaining variation in DMI (Table 5). The exact reason for this is not known because temperature and THI that would cause stress in cattle is not apparent. When considering all months, results show that previous days THI and previous days temperature along with current days THI are important variables affecting DMI of cattle in confinement. This shows the ineffectiveness of confinement dissipating heat gained during the day and the carry-over effect of this heat load being displayed in next days cattle performance. When comparing cattle performance in different housings, the environment the cattle are exposed to should be considered. The environment to which cattle in open lots were exposed was different from that of the open lots with access to an overhead shelter and from the environment of the cattle in confinement. Body temperature is determined by heat input from metabolic heat production and solar

428

Int J Biometeorol (2008) 52:419429

radiation and by heat output through evaporative and nonevaporative avenues. When heat loss does not attain heat gain, heat is stored, thus causing an increase in body temperature (Brosh et al. 1998). Thus, providing shade to decrease thermal radiation in hot weather would be a means to help cattle maintain body temperature (Blackshaw and Blackshaw 1994; Mader et al. 1997). Shade changes the radiation balance of an animal but does not affect air temperature or humidity (Hahn et al. 1970; Buffington et al. 1981; Esmay 1982). By providing shade, heat load of cattle could be reduced by more than 30% (Bond et al. 1967). Yamamoto et al. (1994) reported that solar radiation as measured by black globe temperature contributed substantially more to the heat load of animals than did dry bulb temperature, and there was no effect of solar radiation on the rate of heat production or heart rate of the heifers. Yamamoto et al. (1994) and Whittow (1962) showed that in cattle skin temperature exposed to solar radiation was higher than that of the non-exposed side. When Coleman et al. (1996) measured black globe temperature above and below shade netting they found that morning readings were reduced by more than 3C, and afternoon temperatures were reduced by nearly 7C directly under the shade netting; readings remained relatively stable as distances under the shade netting increased and shade reduced temperatures both inside the hutches and in the outer exercise areas during both years. Valtorta et al. (1997) reported that black globe temperature was higher in nonshaded pens compared to shaded pens, and rectal temperature and respiration rates were higher in non-shaded pens. The effect of shade was demonstrated in another experiment by Pereira et al. (2001) where they kept two breeds of heifers under direct solar radiation between 1200 and 1400 hours and, in the following 2 h, animals were kept in shade. They found that 80 min after changing to the shade, Alentejana breed heifers body temperature completely recovered the initial body temperature whereas Limousine breed heifers did not completely recover their initial temperature. Cattle in open lot with access to an overhead shelter had higher DMI because they did not get exposure to the sun and thus were not exposed to the heat load by solar radiation, and since they preferred lying outside during nighttime hours when the temperature was lower they were able to dissipate the heat they produced during the day. Regulatory mechanisms responsible for body heat dissipation are chiefly adjustments in behavior, in blood circulation, and evaporative heat loss (Griffin 2001). At high temperatures, evaporative cooling is the principal mechanism for heat dissipation in cattle. It is influenced by humidity and wind speed and by physiological factors such as respiration rate, and density and activity of sweat glands (Blackshaw and Blackshaw 1994). At upper thermal-critical

temperature, a 454-kg animal can have 141.75 and 454 g/ h of water as maximum respiratory evaporation and maximum skin evaporation, respectively, while relative humidity and airflow are closely interrelated to maximal evaporation (Griffin 2001). McLean and Calvert (1972) found that heat loss by evaporation amounted to 18% of the total heat loss at 15C and to 84% at 35C, and heat loss by respiratory evaporation amounted to 54% of the total evaporative heat loss at 15C and to 38% at 35C. They also found that heat loss by evaporation, respiratory evaporation and skin evaporation decreased as humidity increased at 35C. These results show the effectiveness of humidity on evaporative cooling at high temperatures. Cattle in confinement had lower DMI than other cattle due to the effect of humidity as displayed by THI, and by not being able to leave the confinement which stored the heat and dissipated heat at night, thus causing a decrease in heat dissipation and relieving stress. Thus, the previous days THI and temperatures were important variables affecting DMI of confinement cattle.

Conclusions Research showed that performance of cattle is affected by the housing provided. Even though cattle were fed for the same duration and started on feed at the same weight, cattle in open lots with access to an overhead shelter were 38 kg heavier than those in confinement. This proves that providing a shelter is beneficial to the producers. Cattle in open lots with access to an overhead shelter and in open lots were able to dissipate the heat stored during the day at night and were thus able to alleviate the effect of heat stress. However, cattle in confinement were affected by the previous days conditions as the building stores the heat during the day and dissipates it during the night and the next day. Research was conducted by examining months separately and showed that in different months in different housings different variables affected feed intake. Results of the study could be used by producers or extension services to give advice to farmers.

References
Bianca W (1961) Heat tolerance in cattle. Intl J Biometeorol 5:526 Blackshaw JK, Blackshaw AW (1994) Heat stress in cattle and the effect of shade on production and behaviour: a review. Aust J Exp Agric 34:285295 Bond TE, Kelly CF, Morrison SR, Pereira N (1967) Solar, atmospheric, and terrestrial radiation received by shaded and unshaded animals. Trans Am Soc Agric Eng 10:622627 Brody S, Ragsdale AC, Yeck RG, Worstell D (1955) Milk production, feed and water consumption and body weight of Jersey and

Int J Biometeorol (2008) 52:419429 Holstein cows in relation to several diurnal temperatureyrhythms. Univ Mo Res Bull 578 Brosh A, Aharoni Y, Degen AA, Wright D, Young BA (1998) Effects of solar radiation, dietary energy, and time of feeding on thermoregulatory responses and energy balance in cattle in a hot environment. J Anim Sci 76:26712677 Brown-Brandl TM, Eigenberg RA, Nienaber JA, Hahn GL (2004) Indicators of heat stress in shaded and non-shaded feedlot cattle. American Society of Agricultural Engineers Annual Meeting. Paper number 044037 Brown-Brandl TM, Eigenberg RA, Hahn GL, Nienaber JA, Mader TL, Spiers DE, Parkhurst AM (2005) Analyses of thermoregulatory responses of feeder cattle exposed to simulated heat waves. Int J Biometeorol 49:285296 Buffington DE, Collazo-Arocho A, Canton GH, Pitt D, Thatcher WW, Collier RJ (1981) Black globe-humidity index (BGHI) as comfort equation for dairy cows. Trans Am Soc Agric Eng 24:711714 Coleman DA, Moss BR, McCaskey TA (1996) Supplemental shade for dairy calves reared in commercial calf hutches in a southern. J Dairy Sci 79:20382043 Delfino JG, Mathison GW (1991) Effects of cold environment and intake level on the energetic efficiency of feedlot steers. J Anim Sci 69:45774587 Esmay ML (1982) Principles of animal environment. AVI Publishing, Westport, Conn. Givens RL, Garret WN, Bond TE (1966) The California environment and the use of shade for feedlots. 6th Ann Calif Feeders Day Report Griffin D (2001) How health status and environment influence the feedng behavior of feedlot cattle. Intermountain Nutrition Conference, pp 117124 Hahn GL (1995) Environmental influences on feed intake and performance of feedlot cattle. In: Intake by Feedlot Cattle, Oklahoma Agricultural Experimental Station, P-942. Oklahoma State University, Stillwater, pp 207224 Hahn GL (1999) Dynamic responses of cattle to thermal heat loads. J Anim Sci 77(Suppl. 2):1020 Hahn GL, McQuigg JD, Decker WL (1970) Probability of occurrence of cloud cover during summer daytime hours. ASAE Paper 70-826. American Society of Agricultural Engineers, St. Joseph, Mich. Hahn GL, Nienaber JA, Klemcke HG (1987) Tympanic temperature rhythms as a means of assessing stress in cattle. Reports of the CIGR Sect. II Seminar on Latest Developments in Livestock Housing. SP-6-87. American Society of Agricultural Engineers, St. Joseph, Mich., pp 358364 Hahn GL, Chen YR, Nienaber JA, Eigenberg RA, Parkhurst AM (1992) Characterizing animal stress through fractal analysis of thermoregulatory responses. J Therm Biol 17:115120 Hoffman MP, Self HL (1973) Behavioral traits of feedlot steers in Iowa. J Anim Sci 37:14381445 Kabuga JD (1992) The influence of thermal conditions on rectal temperature, respiration rate and pulse rate of lactating HolsteinFriesian cows in the humid tropics. Intl J Biometeorol 36:146150 Koknaroglu H, Loy DD, Wilson DE, Hoffman MP, Lawrence JD (2005) Factors affecting beef cattle performance and profitability. Prof Anim Sci 21:286296 Lefcourt AM, Adams WR (1996) Radiotelemetry measurement of body temperatures of feedlot steers during summer. J Anim Sci 74:26332640 Legates JE, Farthing BR, Casady RB, Barrada MS (1991) Body temperature and respiratory rate of lactating dairy cattle under field and chamber conditions. J Dairy Sci 74:24912500

429 Leu BM, Hoffman MP, Self HL (1977) Comparison of confinement, shelter and no shelter for finishing yearling steers. J Anim Sci 44:717721 Mader TL, Fell LR, McPhee MJ (1997) Behavior response of nonBrahman cattle to shade in commercial feedlots. Proc Fifth Int Livestock Environ Symp, American Society of Agricultural Engineers, St. Joseph, Mich., pp 795802 McLean JA, Calvert DT (1972) Influence of air humidity on the partition of heat exchanges of cattle. J Agric Sci Camb 78: 303307 Mendel VE, Morrison SR, Bond TE, Lofgreen GP (1971) Duration of heat exposure and performance of beef cattle. J Anim Sci 33:850 Mitlhner FM, Galyean ML, McGlone JJ (2002) Shade effects on performance, carcass traits, physiology, and behavior of heatstressed feedlot heifers. J Anim Sci 80:20432050 Morrison SR (1983) Ruminant heat stress: effect on production and means of alleviation. J Anim Sci 57:15941601 Muhamad YB, Hoffman MP, Self HL (1983) Influence of different ratios of corn and corn silage, housing systems and seasons on the performance of feedlot steers. J Anim Sci 56:747 NRC (1996) Nutrient requirements for beef cattle (7th edn). National Academy Press, Washington, DC Parkhurst AM, Hahn GL (1987) Data dependent systems time series analysis of cattle tympanic temperature rhythms. ASAE Paper MCR 87131. American Society of Agricultural Engineers, St. Joseph, Mich. Pereira A, Baccari F Jr, Titto E, Almeida A (2001) Effect of shade after exposure to direct solar radiation on body temperature dynamic responses in Alentejana and Limousine beef cattle. Livestock environment. VI. Proc 6th Int Symp, Louisville, Kentucky, USA, pp 455460 Pusillo GM, Hoffman MP, Self HL (1991) Effect of placing cattle on feed at two-month intervals and housing on feedlot performance and carcass grades. J Anim Sci 69:442450 SAS (1999) The general linear model procedure. User s guide: Statistics version 8. SAS, Cary, N.C. Scott IM, Johnson HD, Hahn GL (1983) Effect of programmed diurnal temperature cycles on plasma thyroxine level body temperature and feed intake of Holstein dairy cows. Int J Biochem 82:47 Seath DM, Miller GD (1946) The relative importance of high temperature and high humidity as factors influencing respiration rate, body temperature, and pulse rate of dairy cows. J Dairy Sci 29:465 Shrode RR, Quui FR, Rupel IW, Leighton RE (1960) Variation in rectal temperature, respiration rate, and pulse rate of dairy cattle as related to variation in four environmental variables. J Dairy Sci 43:1235 Stull R (1988) An introduction to boundary layer meteorology. Kluwer, Amsterdam Valtorta SE, Leva PE, Gallardo MR (1997) Evaluation of different shades to improve dairy cattle well-being in Argentina. Int J Biometeorol 41:6567 Whittow GC (1962) The significance of the extremities of the ox (Bos taurus) in the thermoregulation. J Agric Sci Camb 58:109120 Wiersma F, Stott GH (1974) Dairy cow body temperature in a hot climate. Trans Am Soc Agric Eng 17:745 Yamamoto S, Young BA, Purwanto BP, Nakamasu F, Matsumoto T (1994) Effect of solar radiation on the heat load of dairy heifers. Aust J Agric Res 45:17411749