Use of Cool Perches by Broiler Chickens 1

Use of Cool Perches by Broiler Chickens 1 I. Estevez,*,2 N. Tablante, R. L. Pettit-Riley,* and L. Carr *Department of Animal and Avian Sciences; Virginia-Maryland Regional College of Veterinary Medicine; and Department of Biological Resources Engineering, University of Maryland, College Park Maryland 20742 ABSTRACT Broilers under commercial conditions might experience relatively high temperatures during summer and leg disorders year round that may be partially alleviated by providing them with access to cooled perches. It is unknown, however, how perch temperature and factors such as height and position of the perch affect perch use. Furthermore, little is known regarding gender effects. Eight thousand 1-d-old, mixed-sex broilers were exposed to three perch treatments to determine preferences for water-cooled perches over ambient temperature perches and preferences for height, location, and temperature section of the perch. The experimental treatments were as follows: 1) three cool perches 15 cm above the floor (Cool 15), 2) three ambient perches 7.5 cm off the floor (Ambient 7.5), 3) three ambient perches 15 cm high (Key words: chicken, perch, cooling system, behavior, body weight) 2002 Poultry Science 81:62 69 INTRODUCTION Broiler chickens (Gallus gallus domesticus) have been selected for rapid muscle growth and high feed conversion. This genetic selection has influenced the health of broiler chickens to the extent that leg problems affect the birds at a relatively high rate (Kestin, et al., 1992; Thorp, 1992) and may result in substantial financial losses to poultry producers, due to poor feed conversion, culling, mortality, and carcass downgrading (Morris, 1993). To reduce the impact of leg problems, different management practices such as decreasing growth rate, use of alternative lighting programs, and other options have been investigated (for review, see Hester, 1994). Other researchers have indicated that the incidence and severity of leg problems can be reduced by increasing the level of exercise of the birds (Haye and Simons, 1978; Newberry et al., 1988; Kestin et al., 1992; Elson, 1993; Reiter and Bessei, 1996, 1998a,b; Bizeray et al., 2000). 2002 Poultry Science Association, Inc. Received for publication March 19, 2001. Accepted for publication August 24, 2001. 1 Funding for this research was provided by Maryland Agriculture Experiment Station Grant #AASC-99-13. 2 To whom correspondence should be addressed: ie7@umail.umd.edu. (Ambient 15), and 4) control chambers with no perches. Total number of birds perching, their positions, and temperature section within the perch were recorded. The results indicate a strong preference for high perches as birds grow (P < 0.0001). The cooler sections of the perch were utilized more than warmer sections within the cool treatments (P < 0.05). Females showed a stronger tendency to perch than males, particularly within the cool treatment (P < 0.0001). The higher perch use could be one of the reasons for the higher eviscerated body weight found in females with access to cool perches (P < 0.05). Differences in mean body weight were not significant (P = 0.07). Potential beneficial effects of perch access in final body weights needs to be further investigated. Exercise is particularly low at high densities (Newberry and Hall, 1990; Estevez et al., 1997), and this lack of activity may be one reason for the exacerbated leg weakness found in broilers at high densities (Sørensen et al., 2000). Roosting is a natural behavior of jungle fowl (Collias and Collias, 1967) and their domesticated relatives (Blokhuis, 1984). Therefore, providing chickens with perches could be a way to encourage exercise. Use of perches with laying hens has been related to increased bone strength (Appleby and Hughes, 1990; Duncan et al., 1992). However, modern lines of broilers are typically not provided with perches in commercial flocks, mainly because at moderate and relatively high densities perch use is relatively low (LeVan et al., 2000; Pettit-Riley and Estevez, 2001). Nevertheless, it may be possible to encourage perch use if broilers can gain additional benefits by perching, such as provision of a cooler microclimate during the summer. It has been demonstrated that broiler breeder hens (Muiruri et al., 1991) and broiler chickens (Reilly et al., 1991) prefer to roost on cool perches when exposed to high environmental temperatures, resulting in improved Abbreviation Key: LESREC-PAF = Lower Eastern Shore Research and Education Center-Princess Anne Facility. 62

USE OF COOL PERCHES BY BROILERS 63 bird performance (Muiruri and Harrison, 1991; Reilly et al., 1991). Previous studies have indicated that perch use by broilers varies according to the location and height of the perch (LeVan et al., 2000; Pettit-Riley and Estevez, 2001) and bird gender (Hughes and Elson, 1977). Therefore, to design an efficient perching system that maximizes perch use, it is essential to determine the optimum height and position of the perches within the pen. The objectives of this investigation were to determine the effects of perch temperature, height, and position on the frequency of perch use for male and female broiler chickens. Additionally, we wanted to determine the number of birds that, while not perching, were in physical contact with the perches. The rationale for collecting this data is that chickens that do not perch will benefit from the cooling effect of the perches by establishing body contact with them, hence allowing heat transfer from the bird to the perch. Perch contact has never been investigated. We hypothesized that broiler chickens will have a preference for access to cool perches and additionally will prefer specific perch locations within the pen. MATERIALS AND METHODS Animals and Husbandry Eight thousand 1-d-old, mixed-sex Peterson Cobb broiler chicks were obtained from a commercial hatchery and reared during the summer of 1998, at the Lower Eastern Shore Research and Education Center-Princess Anne Facility (LESREC-PAF), University of Maryland. This experimental facility has eighteen 37.21m 2 (6.1 6.1m) windowless independent chambers, each of which has a capacity for 500 birds (13.43 birds/m 2 ). The chambers were of pole and panel construction and were arranged nine on each side of the house, separated by a 1.5 m wide common hallway. Every chamber had an independent electrical distribution panel and automatic feeding, drinking, and negative pressure ventilation systems. Sixteen of these chambers were used in this experiment. The birds were raised under standard management practices and received a four-phase feeding regimen ad libitum, consisting of starter, finisher, withdrawal, and final withdrawal diet. Water was provided ad libitum via two lines of 18 nipple drinkers (total of 36 nipples per chamber). The drinker lines were placed parallel with the diagonal of the chamber, 1.2 m from the wall (Figure 1). An automatic feed line was located midway between the two drinker lines (Figure 1) and consisted of a 100-kg feed bin, seven 30-cm diameter feed pans, and one control pan. The feed was weighed and distributed to each 100-kg feed bin in a chamber by an overhead auger system running the length of the nine chambers. 3 Esterline Angus Instrument Corporation, Indianapolis, IN 46224. 4 ELKAY Manufacturing Company, Oak Brook, IL 60521. FIGURE 1. Layout of the experimental chamber with corresponding locations of the different resources, perches, feeders, and drinkers. Perch location and section of the perches are also indicated. Artificial light was provided continuously (24h/d)by four 200-watt incandescent lamps located at the quarter points along the chamber ceiling. Two 60-watt lamps were positioned along the feed line, and two 25-watt lamps were along the other diagonal line of the chamber. The floor was covered with 15 cm of sawdust litter. Temperatures were maintained between 32 and 34 C at the beginning of the rearing period and were gradually decreased every 2 to 3 d to 22 C at the end of the rearing period. However, ambient temperatures were well above the targeted temperatures, with maximum diurnal temperatures from 4 to 6 wk of age oscillating from 30.55 to 34 C, which are typical for the region during summer. Actual chamber temperatures were measured 75 cm from the litter surface by use of type T thermocouples and an Esterline Data Acquisition System (Model MRL 3 ). Experimental Design One-day-old chicks were randomly divided into 16 groups of 500 birds and assigned to one of the following treatments: 1) cool perch 15 cm off the floor (Cool 15); 2) ambient temperature perch 7.5 cm off the floor (Ambient 7.5); 3) ambient temperature perch 15 cm off the floor (Ambient 15); and 4) no perch (control). The perch treatment consisted of three horizontal perches parallel to three of the chamber walls and located 50.8 cm out from the wall. Each treatment was replicated four times and organized in a completely randomized design. The perches in all treatments were identical and consisted of three black iron pipes (4.56 m length, 3.8 cm diameter) that were held in place by wood supports with slots for positioning and securing. Cooling of the perches in the cool-perch treatment was achieved by circulating 10 C tap water through the system at 425 ml/min. Two Elkay chillers (9.5 L/h capacity; Model ER10-1B 4 ) installed at one end of the house supplied the cool water. The inlet and outlet ends of the water-cooled perches were equipped with thermocouple probes (TMTSS/

64 ESTEVEZ ET AL. 125G/12 5 ). The probes were attached to the Esterline Data Acquisition System (Model MRL 6 ) in a parallel arrangement to measure the average inlet and outlet water temperatures. Chickens were exposed to the perches from 1 d, and the cool water was turned on at 21 d. The perch parallel to the wall farthest from the hallway was designated the outside perch; the perch parallel to the wall closest to the hallway was designated the hallway perch, and the perch parallel to the front wall of the house was designated the front perch (Figure 1). Each perch was subdivided into three sections (1.52 m long) according to the expected differences in temperature within the cool perch, defined as 1) cool section (closest to the inlet of the cool water), 2) medium section, and 3) warm section (farthest from the inlet of the cool water). We did not measure the actual temperature of these three perch sections. Data Collection Instantaneous scan sampling (Martin and Bateson, 1986) was used to record the number of birds perching with respect to position of the perch within the room and temperature section (cool, medium and warm) within the perch (see Figure 1). Twenty-eight scans were taken in each room from 21 to 42 d of age. A bird was recorded as perching if it was standing or resting with both feet on the perch for more than 2 s (LeVan et al., 2000; Pettit-Riley and Estevez, 2001). If the bird was not perching but was in physical contact with the perch (e.g., resting the head or the wing on the perch) it was recorded as in contact with the perch because birds can gain some cooling benefits from this contact behavior. During the last 2 wk of the rearing period, we also recorded the total males and females perching or in contact with the perch for each section of the perch. Visual identification of males and females was based on development of secondary sexual characteristics. Mortality and culls were recorded per chamber on a daily basis, and the probable cause of death noted. Dead birds were not replaced during the experiment. Addition of new, unfamiliar birds could have created social changes in the group, which could be difficult to control and evaluate. Originally, 10 birds from each group, randomly chosen at Day 1 (it was assumed a sex ratio 50:50), were double tagged in each wing for individual recognition. The purpose was to gather information on the interindividual differences in the frequency of perch use among birds and determine its relationship to final body weight at the end of rearing. However, due to the large number of birds in the pen, and to difficulties in reading the tags from a distance (to avoid disturbing the behavior of the chickens), too few data were statistically analyzed. 5 Omega Engineering Inc., Standford, CT 06906. 6 Esterline Angus Instrument Corporation, Indianapolis, IN 46224. At the end of the rearing period (43 d), 10 birds per room were weighed. This number included all of the surviving tagged birds plus some extra (usually two to three birds) taken at random from each pen for a total of 10. Data from these 10 birds were used exclusively for analysis of live and eviscerated carcass weights. The slaughter of the birds was conducted at the LESREC- PAF facility, which has the capacity and manpower to process a maximum of 180 birds per day. The experimental protocol was approved by the Institutional Animal Care and Use Committee of the University of Maryland. Statistical Analysis Mean number of birds perching and the mean number of chickens in contact with the perches per pen were analyzed as a completely randomized design using a mixed-model repeated-measures ANOVA with age (4 to 6 wk) as the repeated measurement. Data satisfied the ANOVA requirements for normality and homogeneity of variance when observations were averaged per pen and per week of observation. No data transformation was required for the analysis. Main factors included in the model were treatment, position, and section of the perch and their interactions. A second mixed-model repeated-measure ANOVA was conducted to detect potential gender effects. Factors in the model included those mentioned earlier, plus gender. Final body weight data were analyzed using a mixed-model ANOVA. The least-squares mean test was used for all pairwise mean comparisons. All analyses were performed using statistical software from the SAS Institute Inc. (1996). Statistical significance was at P < 0.05 for all statistical tests. RESULTS General Use of Perches Perch use was different across treatments (Table 1), with more perching occurring in the cool perch treatment than in any other. This effect was influenced by the age of the birds as indicated by the treatment-byage interaction (Table 1, Figure 2). Maximum perch use was attained in the Ambient 7.5 treatment at 4 wk of age and declined over time. On the contrary, perch use within the Cool 15 and Ambient 15 treatments were lowest at 4 wk and increased with time. Differential use of the perches in the three different positions within the chamber (outside, hallway, and front), as well as use of different sections within the perch (cool, medium, and warm) were also found (Table 1). Birds used the perches located by the outside and hallway locations more often than those placed at the front of the room (Table 2). The interaction between perch treatment and perch section was significant (Table 1), with more chickens perching on the cool section of the cool perch. On the contrary, no differences between

USE OF COOL PERCHES BY BROILERS 65 TABLE 1. Model and ANOVA results for number of birds perching and number of birds in contact with the perch 1 Perching In contact with perch Source of variation df F P F P Treatment 2 (T) 2 14.77 0.001 2.59 NS Perch position 3 (P) 2 10.30 0.0001 9.74 0.0001 Perch section 4 (S) 2 14.75 0.0001 0.24 NS Age 5 (A) 2 0.28 NS 26.17 0.0001 T P 4 2.36 NS 2.33 NS T S 4 4.30 0.022 0.03 NS T A 4 45.51 0.0001 1.32 NS P S 4 0.54 NS 1.61 NS T P S 8 0.75 NS 0.08 NS A P S 8 0.84 NS 0.93 NS Error 6 274 1 Degrees of freedom were the same for both variables. 2 Perch treatment i effect, defined (i = 1to3). 3 Perch position j effect defined by the location of the perches within the pen (j = front, outside and hallway, respectively). 4 Perch Section k (k = cool, medium, warm). 5 Age n effect (n = 4to6wk). 6 Error term (Pen (Treatment)). the use of perch sections were found for the Ambient 7.5 (Figure 3). Chickens in Contact with the Perches Experimental perch treatment had no effect on the total number of birds in contact with the perches (Table 1). However, an effect of perch position was found (Table 1), with more birds having contact with the perches located in the hallway and outside positions than those at the front of the chamber (Table 3). Additionally, more chickens (Table 1) were in contact with the perches during Weeks 5 and 6 than during Week 4 (Table 4). All other main effects and factor interactions were not significant (Table 1). Effect of Gender The results of this experiment (Table 5) indicate that females used the perches more than the males (1.90 ± 0.07, mean percentage of females per perch; 1.18 ± 0.07, mean percentage of males per perch 4.56 m perch length). In addition, a gender by perch treatment interaction was found (Table 5), indicating that males and females responded differently to the perch treatments (Figure 4). Similarly, females were found to have a stronger preference for the coolest section of the perch (gender by perch section interaction, Table 5) whereas the males did not (Figure 5). The mean percentage of males and females in contact with each perch was different (Table 5), with consistently more females (1.340 ± 0.050 mean percentage per perch 4.56 m perch length) than males (1.144 ± 0.046 mean percentage per perch 4.56 m perch length) found in contact with the perches. The other factors studied in this experiment had no differential gender effect (Table 5). Effect of Perches on Body Weight A significant perch treatment-by-gender interaction was found for the eviscerated carcass weight (P = 0.02, Figure 6), with females showing higher eviscerated TABLE 2. Least-square means (± SE) for the percentage of birds per perch roosting at the different perch positions 1 FIGURE 2. Mean percentage (± SE) of birds (from the total live birds in the pen) using the total perching space available 13.68 m total perch length for the different perch treatments (cool 15, Ambient 7.5, and Ambient 15) across age. Different letters indicate significant differences within the same perch treatment across age (P < 0.05). Perch position Front Outside Hallway Percentage of birds per observation 1.701 ± 0.072 b 1.941 ± 0.076 a 2.010 ± 0.075 a a,b Means with the same superscript are not significantly different (P 0.05). 1 Means are reported per perch 4.56 m- within the chambers.

66 ESTEVEZ ET AL. TABLE 4. Least-square means (± SE) for the percentage of birds per perch in contact with the perch at different ages 1 Age of birds (wk) Percentage of birds per observation 4 0.945 ± 0.067 b 5 1.564 ± 0.063 a 6 1.557 ± 0.066 a a,b Means with the same superscripts are not significantly different (P 0.05). 1 Means are reported per perch 4.56 m- within the chambers. FIGURE 3. Effect of perch treatment and temperature section (cool, medium, and warm) on percentage of perch use (% of birds per perch section-; mean ± SE). Each perch section was 1.52 m long (multiply by three to get total perch use for the three available perches in the pen). a,b Different letters indicate significant differences within the same perch treatment across sections (P < 0.05). weight within the Cool 15 treatment and males in the Cool 15 and Ambient 7.5 treatments. No significant effects of treatment or treatment by gender were found for the amount of abdominal fat, (P = 0.08; mean fat content ± SE: 20.310 ± 0.93g for males; 17.950 ± 0.76 g for females). There was no significant effect of perch treatment on final body weights at 6 wk of age (P = 0.08; mean ± SE: 1,889.78 ± 31.75 g, Cool 15; 1,787.43 ± 42.47 g, Ambient 7.5; 1,798.98 ± 35.58 g, Ambient 15; and 1,763.95 ± 34.47g, control). DISCUSSION The results of this investigation show that, if given a choice, broiler chickens prefer water-cooled perches to ambient temperature perches toward the end of the rearing period. The higher general use of the cool perches in comparison to ambient temperature perch of the same height is consistent with the results reported for broiler breeder hens and broiler chickens (Reilly et al., 1991; Muiruri et al., 1991). The frequency of perch use in this experiment, particularly of the cool perches, was higher than previously reported for broilers in other perching systems at moderate densities of 11 birds/m 2 (LeVan et TABLE 3. Least-square means (± SE) for the percentage of birds per perch in contact with the perch at the different perch positions 1 Perch position Front Outside Hallway Percentage of birds per observation 1.110 ± 0.042 b 1.431 ± 0.078 a 1.532 ± 0.090 a a,b Means with the same superscript are not significantly different (P 0.05). 1 Means are reported per perch 4.56 m- within the chambers. al., 2000) and high densities of 20 birds/m 2 (Pettit-Riley and Estevez, 2001). However these percentages are still much lower than the 12 to 27% perch use obtained by Hughes and Elson (1977), depending on perch height and bird density. The large differences between the results reported by Hughes and Elson (1977) and all the other studies could be related to the faster growth rate and higher body weight of the modern broiler strains that might impair their perching ability, in comparison with the strains used in the 1970s. A clear preference was detected for low ambient perches (Ambient 7.5) vs. high perches (15 cm ambient and cool) during Week 4, which could be related to the greater difficulty in reaching a higher perch when the birds are smaller. Lower perches might be easier discoveries by birds than higher ones. By Weeks 5 and 6 there was a clear preference for the higher perches, particularly the cool perches, suggesting that once the birds reach a certain size, low perches are less attractive for chickens than higher perches (within a limited height perhaps). These results agree with the findings of LeVan et al. (2000) who reported greater use of lower perches (angled perches starting off the floor level) early during the rearing period, whereas after 3 wk of age, the percentage of birds on the horizontal perches (8.5 cm off the floor) was greater. Similarly, Pettit-Riley and Estevez (2001) indicated that broilers preferred to climb higher on angled perches when they were between 3 and 4 wk of age. The analysis of the differential use of perch positions within each chamber was conducted to test the hypothesis that the same resource might have different value to the birds according to its position within the chamber. In general, the front perches in all treatments were used less than the perches near the outside or the hallway walls. Previous studies (LeVan et al., 2000) have cited broilers favoring use of perches located in specific areas of a pen. It is possible that this preference for lateral perches was related to better ventilation at this location. Preferences for perches located in the area with the best ventilation have also been indicated in the study conducted by Pettit-Riley and Estevez (2001). Therefore, it appears that the chickens were actively seeking better ventilation by using the lateral perches while avoiding perches located at the front of the chambers. Similarly, the number of birds in contact with the outside and hallway perches in all treatments was consistently

USE OF COOL PERCHES BY BROILERS 67 TABLE 5. Model and ANOVA results for number of birds perching and number of birds in contact with the perch 1 Perching In contact with perch Source of variance df F P F P Treatment 2 (T) 2 15.96 0.0008 1.08 NS Gender 3 (G) 1 164.24 0.0001 8.60 0.003 Perch position 4 (P) 2 9.33 0.0001 11.02 0.0001 Perch section 5 (S) 2 13.08 0.0001 0.17 NS Age 6 (A) 1 0.14 NS 0.21 NS G T 2 10.86 0.0001 1.33 NS G A 1 0.33 NS 3.31 NS G P 2 0.56 NS 0.30 NS G S 2 4.13 0.016 0.09 NS G P S 4 0.57 NS 0.60 NS G A P 4 1.86 NS 2.18 NS G A S 4 0.49 NS 0.49 NS Error 7 408 1 Degrees of freedom are the same for both variables. 2 Perch treatment i effect, defined (i = 1to3). 3 Gender of the birds j (j = females and males). 4 Perch position k effect defined by the location of the perches within the pen (k = front, outside, and hallway, respectively). 5 Perch section n (n = cool, medium, and warm). 6 Age m effect (m = 5to6wk). 7 Error term (Pen (Treatment)). higher than the number of birds in contact with the front perch, suggesting that these birds gained additional cooling benefits at these locations. A possible explanation is that the metal perches were cooler at these locations because the higher airflow favored heat transfer from the chickens to the perches. In this study an effect of the perch section within cool perches was found, with more chickens perching at the coolest section of the perch (closest to the inlet of the cool water). This effect was not found in Ambient 7.5 perches and, although significant, was not as marked in the 15 ambient perches. This result seems to indicate that chickens perched preferentially on the coolest portion of the perches possibly to increase the cooling benefits from the conductive heat dissipater. In regards to gender effects, females consistently had greater perching frequency than males, with higher differences between genders as perch height increased. Females also showed a stronger preference for cool perches than males, as indicated by the gender-by-perch-treatment interaction. The higher percentage of perching females is in contrast with a lack of differences according to gender reported by Faure and Jones (1982a) in different strains of laying hens. However, in a second study, Faure and Jones (1982b) indicated that females perched FIGURE 4. Mean percentage (± SE) of males and females using the perches for each perch treatment. Means are given as percentage of use of any of the three perches within the same treatment (total perch use for each group will be the reported value multiplied by the number of perches in the chamber). FIGURE 5. Effect of gender on the percentage of perch use (mean ± SE) at the different perch sections. Each perch section was 1.52 m long (multiply by three to get total perch use for the three available perches in the pen).

68 ESTEVEZ ET AL. FIGURE 6. Effect of gender on the eviscerated carcass weight for each perch treatment. a,b Different letters indicate significant differences within the same gender across the different perch treatments and control (P < 0.05). more than males, in agreement with the results reported in the present study and with the gender differences reported by Hughes and Elson (1977). The differences in the perching behavior of males and females in broilers is likely to be related to the higher body weight of the males, which may impair ability to get on and off the perches. Hughes and Elson (1977) indicated a tendency for lighter birds to perch more than heavier birds, although the reason was unknown. It is very unlikely that greater use of the perches by females could be related to social factors, as level of aggression in young broilers has been reported to be very low (Estevez et al, 1997). The number of females in contact with the perches was also higher than for males. This result is somewhat surprising in that we would expect the higher male body weight to impair their ability to jump onto the perches but not impair ability to be in contact with the perches, which would still allow them to attain part of the cooling benefits. We have no explanation, at the present time, for the higher frequency of contact with the perches observed with the females. Access to cool perches might have some beneficial effects on eviscerated carcass weights and, perhaps, final body weights. In agreement with the results of performance in broilers reported by Reilly et al. (1991), females provided with access to cool perches reportedly have significantly greater perch use than females in other treatments. These females also had higher eviscerated carcass weight than birds from other treatments. However, this finding should be interpreted with caution and investigated further to assess impact of access to perches in bird performance because differences in final body weights between treatments only approached significance. In conclusion, the results of this investigation show that use of perches by broiler chickens varies according to the height, temperature, and position of the perches. It is also indicated that females tended to use perches more frequently than males, which could be related to lower female body weight. ACKNOWLEDGMENTS We thank P. Okelo, C. Wabeck, and N. Zimmermann for their help at different phases of the experiment and with data recording; V. Byerd, K. Armentrout, and J. Martin for their help with the care of the birds; and E. Russek-Cohen for her advice on the statistical analysis. We also thank the Maryland Agricultural Experiment Station for funding this experiment and supporting R. Pettit (Grant No. AASC-99-13). This research contributed to the NE-127 Regional Research Project. REFERENCES Appleby, M. C., and B. O. Hughes, 1990. Cages modified with perches and nests for the improvement of bird welfare. World s Poult. Sci. J. 46:38 40. Bizeray, D., C. Leterrier, P. Constantin, M. Picard, and J. M. Faure, 2000. Early locomotor behaviour in genetic stocks of chickens with different growth rates. Appl. Anim. Behav. Sci. 68:231 242. Blokhuis, H. J., 1984. Rest in poultry. Appl. Anim. Behav. Sci. 12:289 303. Collias, N. E., and E. C. Collias, 1967. A field study of the Red Jungle Fowl in North-Central India. Condor 69:360 386. Duncan, E. T., M. C. Appleby, and B. O. Hughes, 1992. Effect of perches in laying cages on welfare and production of hens. Br. Poult. Sci. 33:25 35. Elson, A., 1993. Housing systems for broilers. Pages 177-184 in: Proceedings of the IV European Symposium on Poultry Welfare. C. J. Savory and B. O. Hughes, ed. Edinburgh, UK. Estevez, I., R. C. Newberry, and L. Arias de Reyna, 1997. Broiler chickens: A tolerant social system? Etología 5:19 29. Faure, J. M., and R. B. Jones, 1982a. Effects of sex, strain and type of perch on perching behaviour in the domestic fowl. Appl. Anim. Ethol. 8:281 293. Faure, J. M., and R. B. Jones, 1982b. Effects of age, access and time of day on perching behaviour in the domestic fowl. Appl. Anim. Ethol. 8:375 364. Haye, U., and P. C. M. Simons, 1978. Twisted legs in broilers. Br. Poult. Sci. 19:549 557. Hester P. Y., 1994. The role of environment and management on leg abnormalities in meat-type fowl. Poultry Sci. 73:904 915. Hughes, B. O., and H. A. Elson, 1977. The use of perches by broilers in floor pens. Br. Poult. Sci. 18:715 722. Kestin, S. C., T. G. Knowles, A. E. Tinch, and N. G. Gregory, 1992. Prevalence of leg weakness in broiler chickens and its relationship with genotype. Vet. Rec. 131:190 194. LeVan, N. F., I. Estevez, and W. R. Stricklin, 2000. Use of horizontal and angled perches by broiler chickens. Appl. Anim. Behav. Sci. 65:349 365. Martin, P., and P. Bateson, 1986. Measuring Behaviour. An Introductory Guide. Cambridge University Press, Cambridge, UK. Morris, M. P., 1993. National survey of leg problems. Broiler Ind. (May):20 24. Muiruri, H. K., and P. C. Harrison, 1991. Effect of roost temperature on performance of chickens in hot ambient environments. Poultry Sci. 70:2253 2258. Muiruri, H. K., P. C. Harrison, H. W. Gonyou, 1991. The use of water-cooled roosts by hens for thermoregulation. Appl. Anim. Behav. Sci. 28:333 339.

USE OF COOL PERCHES BY BROILERS 69 Newberry, R. C., and J. W. Hall, 1990. Use of pen space by broiler chickens: Effects of age and pen size. Appl. Anim. Behav. Sci. 25:125 136. Newberry, R. C., J. R. Hunt, and E. E. Gardiner, 1988. Influence of light intensity on behavior and performance of broiler chickens. Poultry Sci. 67:1020 1025. Pettit-Riley, R. L., and I. Estevez, 2001. Effects of density on perching behavior of broiler chickens. Appl. Anim. Behav. Sci. 71:127 140. Reilly, M. W., K. W. Koelkebeck, and P. C. Harrison, 1991. Performance evaluation of heat-stressed commercial broilers provided water-cooled floor perches. Poultry Sci. 70:1699 1703. Reiter, K., and W. Bessei, 1996. Effect of the distance between feeder and drinker on behaviour and leg disorders of broilers. Proceedings of the 30th International Congress of Applied Ethology. I. J. H. Duncan, T. M. Widowski, and D. B. Haley, ed. Guelph, Canada. Reiter, K., and W. Bessei, 1998a. Possibilities to reduce leg disorders in broilers and turkeys: Review. Arch. Geflugelk. 62:145 149. Reiter, K., and W. Bessei, 1998b. Effect of locomotor activity on bone development and leg disorders in broilers. Arch. Geflugelk. 62:247 253. SAS Institute, 1996. SAS. User s Guide: Statistics. SAS Institute Inc., Cary, NC. Sørensen, P., G. Su, and S. C. Kestin, 2000. Effects of age and stocking density on leg weakness in broiler chickens. Poultry Sci. 79:864 870. Thorp, B. H., 1992. Abnormalities in the growth of leg bones. Pages 3-17 in: Bone Biology and Skeletal Disorders in Poultry. Poultry Science Symposium. Vol. 23. C. C. Whitehead, ed. Carfax Publishing Company, Oxfordshire, U.K.