BREEDING AND GENETICS. Comparative Evaluation of Three Commercial Broiler Stocks in Hot Versus Temperate Climates

BREEDING AND GENETICS Comparative Evaluation of Three Commercial Broiler Stocks in Hot Versus Temperate Climates SERVET YALÇIN,* PETEK SETTAR,* SEZEN OZKAN,* and AVIGDOR CAHANER,1 *The Aegean University, Faculty of Agriculture, Izmir 35100, Turkey, and The Hebrew University of Jerusalem, Faculty of Agriculture, Rehovot 76100, Israel ABSTRACT Hot climate is a major limiting factor of broiler production in tropical and subtropical regions. The use of standard stocks in hot climates may result in large economic losses because genotypes selected in temperate climates may respond differently to the high ambient temperatures in hot regions or seasons. The summer and fall in Izmir, Turkey, provided the natural hot and temperate climates, respectively, for this study. Broiler chicks were obtained from three commercial stocks, all bred in temperate climates. Male and female chicks, 60 per pen, were housed in four pens per stock per season. Individual BW was determined at hatch, and at 4 and 7 wk of age. Feed consumption and efficiency were determined per pen. Feathering was scored at 4, 5, and 6 wk of age. Body temperature was measured twice on three birds per sex per pen, 16 h and immediately before slaughter, and feather weight was determined for each of these birds. The two seasons clearly differed in ambient temperature at the broiler house, and consequently, BW at 7 wk was significantly lower in the summer than in the fall in all stocks, with an average reduction of 23%. The season effect was largest (33.5%) on BW gain from 4 to 7 wk, along with 23 and 15% reductions in feed consumption and efficiency, respectively, during these 3 wk. A significant season by stock interaction was detected for BW gain from 0 to 4 wk and 4 to 7 wk. The three stocks exhibited similar 4- to 7-wk BW gains under the temperate fall climatic conditions, but differed significantly in the summer. These differences were not related to normal differences in feather coverage or body temperature, suggesting that standard broiler stocks must be tested in hot climates in order to find the one most suited to these conditions. (Key words: high ambient temperature, season, body temperature, growth, feathering) 1997 Poultry Science 76:921 929 INTRODUCTION Genotype by environment interaction is usually described as a situation in which different genotypes (breeds, lines, or strains) respond differently to different environments (Sheridan, 1990). Climate is one of the main environmental factors affecting poultry production. Reduced broiler performance due to high ambient temperature is well established (Leeson, 1987; Cahaner and Leenstra, 1992; Leenstra and Cahaner, 1992; Cahaner et al., 1993; Eberhart and Washburn, 1993). Hartmann (1990) and Mathur and Horst (1989) reported interactions of stocks with climatic factors, especially ambient temperature. The existence of a genotype by climate interaction means that the genotypes selected by commercial breeders as superior under optimal conditions (i.e., temperate climate) may not maintain their superiority at high ambient temperatures in hot regions or during the hot season. With the rapid development of Received for publication August 19, 1996. Accepted for publication February 17, 1997. 1To whom correspondence should be addressed. the poultry industry worldwide, especially in developing countries, importation of temperate-zone highperformance stocks to hot regions is on the rise. The use of unsuitable genotypes in hot regions results in large economic losses due to decreased growth, reduced protein gain, and higher mortality. The depression of broiler performance due to high ambient temperature cannot be fully compensated by management, especially in hot-climate developing countries limited capital is available to reduce the heat in chicken houses (Mathur and Horst, 1989; Cahaner, 1990). To achieve further improvements in the world poultry industry, breeding programs need to identify genotypes that perform better in hot climates (Cahaner, 1990). Broiler stocks differing in growth rate were examined under controlled environmental conditions for their interactions with ambient temperature (Leenstra and Cahaner, 1991; Cahaner and Leenstra, 1992). At the high temperature (constant 32 C), growth after 4 wk of age was lower for birds from all stocks, as compared to their counterparts at the normal temperature (20 C); however, the magnitude of this effect varied between stocks: the fast-growing commercial broilers suffered more than 921

922 those with a lower potential for growth rate. It has been suggested that fast-growing broilers have more internally generated heat to dissipate (Chwalibog and Eggum, 1989), and consequently they may have difficulty maintaining body temperature at high ambient temperature. Such birds are expected to reduce their metabolic activity in hot environments, and, indeed, feed intake and BW gain at high temperatures were reduced more in the fast-growing stocks than in the slower-growing ones. Based on these results, Cahaner and Leenstra (1992) hypothesized that differences between stocks in their growth at high ambient temperature are related to their genetic potential for growth rate. Hence, continuous selection for increased growth rate may increase broiler sensitivity to hot climates, unless specific criteria for efficient performance at high temperatures are identified and applied in selection programs. The testing of genotypes in different climates is of considerable relevance for breeders who wish to improve the potential performance of their stocks in suboptimal environments (Hartmann, 1990). Field tests under the natural conditions prevalent during different seasons may provide more useful information pertaining to specific climates than experiments in controlledtemperature chambers (Mathur and Horst, 1994). In Turkey, where 90% of the broilers are imported from western Europe and the U.S., the summer season is characterized by high ambient temperatures. A study under natural climatic conditions was conducted to determine the effect of high (summer) and temperate (fall) natural temperatures on the performance of commercial broiler stocks bred in temperate climates, and to relate several characteristics of the stocks to their response to the climatic change from summer to fall. Climate MATERIALS AND METHODS The study consisted of two experiments in the same open-sided poultry house and under the same management, except for the different climates of the summer and fall seasons. The summer-experiment chicks hatched on July 6, 1993 and the fall experiment chicks hatched on October 11, 1993. Except for standard spot-heating by infrared lamps during the brooding period (0 to 3 wk of age), ambient temperature in the broiler house was determined by the outside climate. Temperature and relative humidity in the center of the house were recorded continuously during the experimental period. The temperatures at 0400, 0800, 1200, 1600, 2000, and 2400 h of each day were averaged by weeks, as illustrated in Figure 1. Chickens The chickens were obtained from three commercial stocks bred in the U.S. (ST1), Germany (ST2), and the U.K. YALÇIN ET AL. (ST3). Parents of ST2 and ST3 were imported directly from Germany and the U.K., whereas grandparents of ST1 were transported from U.S. to Israel and from there the day-old parent chicks were imported to Turkey. The ST1, ST2 and ST3 parents were 32, 49, and 28 wk of age, respectively, when eggs for the summer experiment were collected, and 13 wk older when eggs were collected for the fall experiment. Males were slow-feathering in ST1 and ST3, and fast-feathering in ST2; all females were fastfeathering. The three stocks used in this experiment represent 80 to 90% of the broilers grown in Turkey. Housing and Management Chicks (240) from each stock were weighed and wingbanded on the day of hatch. Each stock was represented by four replicated floor pens containing 60 birds each with males and females mixed. All pens were bedded with a wood-shavings litter and equipped with two feeders and a waterer. Chicks had ad libitum access to a commercial broiler starter diet (22% CP and 3,000 kcal ME/kg) from 0 to 28 d of age and a finishing diet (21% CP and 3,100 kcal ME/kg) from 29 to 49 d of age. A lighting program of 1 h dark:23 h light was provided. Measurements Body weight was measured for each chick at hatch (BW0), 4 wk (BW4), and 7 wk (BW7), and BW gain (BWG) was calculated from 0 to 4 wk (BWG0-4) and from 4 to 7 wk (BWG4-7). Feed efficiency (FE) was computed as the ratio of BWG to feed consumption (FC) based on pen values adjusted for mortality. These values were calculated separately for Weeks 0 to 4 (FC0-4 and FE0-4) and Weeks 4 to 7 (FC4-7 and FE4-7). For each bird, feather coverage on the back and thighs was classified as fully feathered or poorly feathered at 4, 5, and 6 wk of age; all birds were fully feathered at 7 wk of age. Twelve males and 12 females (3 per sex from each pen) were randomly selected from each stock at 7 wk of age. Rectal temperature, as a measure of body temperature (BT), was recorded twice for each of these 72 birds: at 1600 h on Day 48 (BT1) and again at 0800 h on Day 49, following 10 h of feed withdrawal (BT2). All these birds were killed by cervical dislocation immediately after recording BT2, plucked mechanically, and reweighed to calculate absolute feather weight (FW) by subtraction. The percentage of feathers was calculated relative to live BW just before slaughter. Statistical Analysis All data were subjected to two-way analyses of variance with genotype and sex as the main effects, separately for each season. The data of BWG, BW, FW, and BT were analyzed also by a three-way model with season, stock, and sex as main effects, and their interactions. Hatch weight (BW0) was included as a covariate in the models used to analyze BWG0-4, BWG4-7, and BW7, to

EVALUATING BROILER STOCKS IN HOT VS TEMPERATE CLIMATES 923 FIGURE 1. Weekly averages of ambient temperatures measured inside the broiler house six times a day, during the summer and fall seasons. correct for the differences in age of dams between stocks and seasons, which affect egg weight and BW0, and consequently may affect BWG. Pen data of FC and FE were analyzed by a model that included season and stock as main effects, and their two-way interaction. Data were analyzed using the General Linear Models (GLM) procedure of SAS (SAS Institute, 1989) and tabular results are least squares means. Frequencies of fully vs poorly feathered birds between stocks within seasons and between seasons within stocks were analyzed separately by chi-square test for each age and sex. RESULTS Daily fluctuations of the ambient temperature in the broiler house, averaged for each week, are presented in Figure 1. Brooder-house space temperatures during the first 3 wk were higher in the summer than in the fall by about 6 C at each time point. The difference between summer and fall ambient temperatures increased to about 10 C after brooding. During the last 3 wk of the experiments, ambient temperature averaged about 28 C in the summer vs 18 C in the fall, with no overlap between the average maximum daily temperature in the fall and the average minimum daily temperature in the summer. Within each season, relative humidity (RH) was reduced by 1% when ambient temperature increased by 1 C. Over the entire growth period, RH averaged 45 and 56% in the summer and fall experiments, respectively. The RH values in the summer experiment were somewhat below the optimum for broilers under heat stress (Yahav et al., 1995). It seems, therefore, that the lower performance of all birds in the summer vs fall experiments resulted from the combined effect of high temperature and low RH. Mean BW and BWG of each stock in summer and fall, and the calculated values of summer relative effect (SRE, the difference between summer and fall means, as percentage of fall), are given in Table 1. The BW0 covariate significantly affected BWG0-4, but not BWG4-7. It appears, therefore, that BW0 corrected for most of the variation, between stocks and seasons, in age-of-dam

924 YALÇIN ET AL. TABLE 1. ANOVA and least squares means of BW at hatch (BW0), BW gain (BWG) from 0 to 4 and 4 to 7 wk, BW at 7 wk (BW7), and the calculated values of summer relative effect (SRE) 1 of females (F) and males (M) from three broiler stocks reared during the summer and fall (n = 120 per season per stock per sex) Summer Fall SRE Variable BW0 BWG0 4 BWG4 7 BW7 BW0 BWG0 4 BWG4 7 BW7 BW0 BWG0 4 BWG4 7 BW7 (g) (%) Stock 2 ST1 44.9 b 964 a 901 b 1,910 a 46.6 a 1,009 a 1,350 2,409 3.6*** 4.4*** 33.3*** 20.7*** ST2 45.6 a 910 b 863 c 1,815 b 45.3 b 977 b 1,345 2,363 0.7 6.8*** 35.8*** 23.2*** ST3 40.9 c 781 c 945 a 1,758 c 41.9 c 944 c 1,364 2,343 2.6*** 17.3*** 31.4*** 24.9*** Sex F 43.4 y 840 y 818 y 1,700 y 44.6 x 912 y 1,212 y 2,167 y 2.7*** 7.9*** 32.5*** 21.6*** M 44.2 x 930 x 981 x 1,954 x 44.6 x 1,042 x 1,484 x 2,576 x 0.9 10.7*** 33.9*** 24.1*** Source of variation Probabilities from ANOVA within each season Probabilities from ANOVA of both seasons Stock <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 NS NS <0.001 <0.001 0.003 <0.001 Sex <0.001 <0.001 <0.001 <0.001 NS <0.001 <0.001 <0.001 0.032 <0.001 <0.001 <0.001 BW0... <0.001 NS NS... 0.014 NS 0.080... <0.001 0.091 <0.001 Stock sex NS NS NS NS NS NS NS NS NS NS NS NS Season <0.001 <0.001 <0.001 <0.001 Season stock <0.001 <0.001 0.048 0.035 Season sex 0.044 0.002 <0.001 <0.001 Season stock sex NS NS NS NS a cstock means within variable with no common superscript differ significantly (P < 0.05). x,ysex means within variable with no common superscript differ significantly (P < 0.05). 1SRE = Nonanalyzable expression of the summer effect on stock or sex means: the difference between least squares means in summer and fall, as percentage of fall mean. 2Commercial broiler stocks bred in the U.S. (ST1), Germany (ST2), and the U.K. (ST3). **The differences between stock or sex means in summer and fall were significant at P < 0.01. ***The differences between stock or sex means in summer and fall were significant at P < 0.001.

EVALUATING BROILER STOCKS IN HOT VS TEMPERATE CLIMATES 925 TABLE 2. ANOVA and least squares means of feed consumption and feed efficiency, 1 and calculated values of summer relative effect (SRE) 2 of three broiler stocks reared during summer and fall (four pens per season/stock, sexes mixed in pens) Feed consumption (FC) Feed efficiency (FE) FC0 4 (0 to 4 wk) FC4 7 (4 to 7 wk) FE0 4 (0 to 4 wk) FE4 7 (4 to 7 wk) Stock 3 Summer Fall SRE Summer Fall SRE Summer Fall SRE Summer Fall SRE (g) (%) (g) (%) (g:g) (%) (g:g) (%) ST1 1,508 a 1,501 a +0.5 2,244 a 2,910 a 22.9*** 0.634 a 0.666 a 4.8* 0.407 a 0.476 a 14.5** ST2 1,513 a 1,490 a +1.5 2,224 a 2,882 a 22.8*** 0.610 b 0.650 ab 6.2* 0.376 b 0.463 b 18.8*** ST3 1,381 b 1,437 b 3.9 2,254 a 2,932 a 23.1*** 0.548 c 0.637 b 13.9*** 0.403 a 0.456 b 11.6** Mean 1,467 1,476 0.6 2,240 2,908 22.9*** 0.590 0.651 9.4*** 0.395 0.464 14.9*** Source of variation Probabilities Season NS 3 <0.001 <0.001 <0.001 Stock <0.001 NS <0.001 0.038 Season stock NS NS <0.001 NS a cstock means within variable with no common superscript differ significantly (P < 0.05). 1Feed efficiency = weight gain/feed consumption. 2SRE = Nonanalyzable expression of the summer effect on stock means: the difference between least squares means in summer and fall, as percentage of fall mean. 3Commercial broiler stocks bred in the U.S. (ST1), Germany (ST2), and the U.K. (ST3). *The differences between stock means in summer and fall were significant at P < 0.05. **The differences between stock means in summer and fall were significant at P < 0.01. ***The differences between stock means in summer and fall were significant at P < 0.001.

926 YALÇIN ET AL. TABLE 3. Chi-square tests of the percentage of poorly feathered females and males, between seasons within stock and between stocks within season (about 100 birds in each season-stock-sex subclass) Females Males Age Stock 1 Summer Fall Chi-square 2 Summer Fall Chi-square 2 (%) (%) 4 wk ST1 47 81 13.8*** 90 99 10.1** ST2 52 91 40.1*** 59 93 34.3*** ST3 46 69 10.9** 88 100 17.4*** Chi-square 3 3.6 19.2*** 39.6*** 8.9* 5 wk ST1 7 20 7.4** 57 66 1.7 ST2 21 16 0.8 29 24 0.6 ST3 14 10 0.8 67 60 1.1 Chi-square 3 11.2** 4.1 35.6*** 49.5** 6 wk ST1 4 0 5.6* 33 13 11.6*** ST2 1 0 1.4 6 0 8.5** ST3 3 0 4.2* 24 17 1.5 Chi-square 3 2.8... 24.8*** 21.7*** 1Commercial broiler stocks bred in the U.S. (ST1), Germany (ST2), and the U.K. (ST3). 2Chi-square tests between seasons within stocks. 3Chi-square tests between stocks within season. *Difference in percentage of poorly feathered birds between stocks or seasons was significant at P < 0.05. **Difference in percentage of poorly feathered birds between stocks or seasons was significant at P < 0.01. ***Difference in percentage of poorly feathered birds between stocks or seasons was significant at P < 0.001. effect on the performance of broilers. Although BW and BWG were lower in summer than in fall in all three stocks, season by stock and season by sex interactions were significant for these traits (Table 1). In the fall, ST1 birds had the highest BWG0-4, ST3 birds had the lowest and ST2 birds were intermediate. The SRE for BWG0-4 was more severe in ST3 than in the ST1 and ST2, hence stocks ranked the same in summer but BWG0-4 of ST3 birds was markedly lower than that of the other two stocks (Table 1). In the fall, BWG4-7 was similar for all stocks, but in the summer it was higher in ST3 than in ST1, with ST2 exhibiting the lowest value. In all stocks, the reduction in BWG due to high summer temperatures (SRE) was much greater in the last 3 wk (about 34%) than in the first 4 wk (about 9%). Stocks differed significantly in BW7 during the summer but not in the fall. The SRE in BW7 was 20.7, 23.2 and 24.9% in ST1, ST2, and ST3 broilers, respectively. The stock by sex interaction was not significant for BW measurements at any age. In all stocks, SRE for BWG and BW7 was lower in females than in males, and the differences between sexes were reduced in the summer, resulting in a significant season by sex interaction. Feed consumption from hatch to 4 wk (FC0-4) was not affected by season, but summer FC4-7 was reduced in all stocks by about 23% as compared to FC4-7 in the fall (Table 2). Feed efficiency from 0 to 4 wk (FE0-4) was lower by 4.8, 6.2, and 13.9% in, respectively, ST1, ST2, and ST3 birds grown in summer vs fall. The season effect on FE4-7 was larger; it was reduced under summer conditions by 14.5, 18.8, and 11.6% in ST1, ST2, and ST3 birds, respectively. There was a significant season by stock interaction for FE0-4, but not for FE4-7. In the fall, ST1 birds had higher FE4-7 than ST2 and ST3 birds, but in the summer, FE4-7 of ST3 birds was as high as that of ST1, and FE4-7 of ST2 birds was significantly lower. The proportions of poorly feathered birds in each stock during summer and fall are given in Table 3. Data are presented for 4, 5, and 6 wk of age, because at 7 wk all birds were fully feathered. Among the males of the slow-feathering stocks (ST1 and ST3) there were more poorly feathered birds than among ST2 males at 4, 5, and 6 wk of age in both seasons. At 4 wk of age, the proportion of poor feathering was lower in summer than in fall by 23 to 39% in fast-feathering birds (all females and ST2 males) and by about 9 to 12% in slowfeathering males (ST1 and ST3). At 5 and 6 wk of age, the differences between seasons within stock and sex were smaller and mostly nonsignificant. There was no significant difference in BW or BWG between poorly and fully feathered birds (data not shown). Seven-week FW, expressed as a percentage of BW, was higher in females than males in both seasons and all stocks (Table 4). The three stocks had similar FW in the summer but differed in the fall. Fall FW of ST3 was lower than in the summer by a relative difference of 15%. The FW was also lower in the fall than in the summer in ST2 (by 7%), but ST1 broilers had similar FW in both seasons (Table 4). The ST3 birds had higher body temperatures (BT1 and BT2 in the fall, BT1 in the summer) than ST1 and ST2 birds; only BT2 in the summer was equal in all three stocks (Table 4). In ST1 and ST2, BT1 was higher in summer than in fall by about 0.6 C, whereas BT2 was similar for the two seasons. In ST3, BT1 was similar in both seasons, but BT2 was lower in summer than in fall. BT2 was lower than BT1 in all cases, apparently due to the lower ambient temperature at 0800 than at 1600 h and the lower metabolic rate of feed-deprived birds. The BT1-

EVALUATING BROILER STOCKS IN HOT VS TEMPERATE CLIMATES 927 TABLE 4. Least squares means of feather weight (FW, % of BW) and of body temperatures on-feed (BT1) and off-feed (BT2) at 7 wk of age of females (F) and males (M) from three broiler stocks reared during the summer and fall seasons (n = 12) in each season-stock-sex subclass) Summer Fall Summer fall Variable FW BT1 BT2 BT1-BT2 FW BT1 BT2 BT1-BT2 FW BT1 BT2 BT1-BT2 (%) (C) (%) (C) (%) (C) Stock 1 ST1 5.11 a 41.77 b 41.00 a 0.77 b 5.05 a 41.16 b 41.01 b 0.15 a +0.06 +0.61*** 0.01 +0.62*** ST2 5.07 a 41.86 b 41.05 a 0.80 b 4.72 a 41.29 b 41.06 b 0.23 a +0.35* +0.57*** 0.01 +0.57*** ST3 5.06 a 42.08 a 41.05 a 1.05 a 4.28 b 41.99 a 41.79 a 0.20 a +0.78*** +0.09 0.75*** +0.85*** Sex F 5.36 x 41.88 x 41.04 x 0.81 x 4.92 x 41.46 x 41.31 x 0.15 x +0.44** +0.42*** 0.27** +0.66*** M 4.80 y 41.93 x 41.03 x 0.94 x 4.45 y 41.49 x 41.26 x 0.23 x +0.35* +0.44*** 0.23* +0.71*** Source of variation Probabilities from ANOVA within each season Probabilities from ANOVA of both seasons Stock NS 2 0.031 NS 0.067 0.001 <0.001 <0.001 NS 0.014 <0.001 <0.001 NS Sex <0.001 NS NS NS 0.006 NS NS NS <0.001 NS NS NS Stock sex NS NS NS NS NS NS NS NS NS NS NS 0.037 Season 0.001 <0.001 <0.001 <0.001 Season stock 0.007 0.001 <0.001 NS Season sex NS NS NS NS Season stock sex NS NS NS NS a cstock means within variable with no common superscript differ significantly (P < 0.05). x,ysex means within variable with no common superscript differ significantly (P < 0.05). 1Commercial broiler stocks bred in the U.S. (ST1), Germany (ST2), and the U.K. (ST3). *Differences between stock or sex means in summer and fall were significant at P < 0.05. **Differences between stock or sex means in summer and fall were significant at P < 0.01. ***Differences between stock or sex means in summer and fall were significant at P < 0.001.

928 BT2 difference in the fall was low (about 0.2 C) and similar in all stocks, whereas in the summer it was about 0.8 C in ST1 and ST2, and 1.05 C in ST3, leading to a significant season by stock interaction (Table 4). DISCUSSION The two seasons clearly differed in ambient temperature. Even during the brooding period, fall ambient temperatures in the open-sided broiler house were lower by about 6 C at each daily time point, but chicks in each pen were heated by infrared lamps up to 3 wk of age. The difference between summer and fall ambient temperatures increased to 10 C between 3 and 7 wk, averaging about 28 C in the summer vs 18 C in the fall (Figure 1). In addition, RH values in the summer were somewhat below the optimum for broilers under heat exposure (Yahav et al., 1995), suggesting that the summer effect combined high temperature and low RH. The recommended ambient temperatures for broilers are gradually reduced from 32 C at Day 1 to about 24 C at 4 wk; hence, summer heat exposure in the present study probably started at 3 wk and was indisputable from 4 wk of age onward. Indeed, BW was significantly lower in summer than in fall in all three stocks (mean reduction of 23%), but this effect was substantially greater in BWG from 4 to 7 wk (33.5%) than from 0 to 4 wk (9.5%). Similar results have been obtained in controlled-temperature experiments (Cahaner and Leenstra, 1992; Cahaner et al., 1993). The summer heat exposure was also characterized by having a greater effect on males than on females, as has been found in controlled experiments by Osman et al. (1989) and Cahaner and Leenstra (1992). Body temperature of the normally fed broilers (BT1) increased with ambient temperature, i.e., it was higher in summer than in fall. This effect was not significant in ST3; its average BT1 was highest among the stocks in the summer and quite similar in the fall. It is hypothesized that the high summer BT1 values were due to the high ambient temperature, which reduced the dissipation rate of heat generated when feed is digested and metabolized. This hypothesis is supported by the much lower body temperature of feed-deprived birds (BT2 vs BT1) in summer, but not in fall. Although ambient temperature was lower when BT2 was measured, it cannot be the only cause of the high BT1-BT2 values in the summer, because this difference (about 0.9 C) was larger than the difference between BT1 in summer and fall (averaged 0.43 C). Teeter et al. (1987) also demonstrated that shortterm feed deprivation lowers body temperature. It is possible that broilers under summer heat exposure counteract their increasing body temperature by reducing feed intake. Indeed, FC from 4 to 7 wk was 23% lower in summer than in fall. This reduction was probably the main reason for the lower BWG4-7 in summer vs fall (by an average 33.5%). The BWG4-7 could also have been reduced due to the higher energy expended for thermoregulation in the summer relative YALÇIN ET AL. to fall, as reflected by the difference in FC (0.395 vs 0.464 in summer and fall, respectively). Stocks differed significantly in BWG0-4 with the same ranking in both seasons and, therefore, it appears that these differences could not be associated to climate or to age of parent stock. However, the use of BW0 as a covariate may not account for all the potential confounding of age-of-dam and climate effects. The three commercial broiler stocks used in the present study exhibited very similar BWG from 4 to 7 wk in the temperate fall climate, but they differed significantly in BWG4-7 in the summer s high ambient temperature. Such differences could be attributed to differences in feather coverage, as suggested by studies with naked neck broilers and with lines selected divergently for fast or slow development of feather coverage (Lou et al., 1992; Ajang et al., 1993; Cahaner et al., 1993). In the present study, differences between stocks in the proportion of poorly feathered birds was not associated with their response to summer climate, and feathering classification was not associated with BW or BWG within season or age. Moreover, although the three stocks differed in BWG4-7 in the summer, they had very similar FW (as percentage of BW). In the fall, ST3 broilers exhibited lower FW, possibly related to their significantly higher BT1 and BT2. In the summer, the ranking of stocks in terms of BT1 or BT1-BT2 was not related to their BWG4-7. It is possible, however, that greater differences in feather coverage, and consequently in thermoregulation, as between normal and naked neck birds, may result in more substantial differences in broiler growth under high ambient temperatures (Cahaner et al., 1993, 1994). The climate during the fall experiment was quite similar to that of the sites in the U.S., Germany, and the U.K., where the three commercial broiler stocks had been bred. Therefore, the similar growth of the three stocks under these conditions could be expected. Less expected were the differences in growth in a hot climate, which could not be associated to the traits measured here. These results support the conclusion that stock evaluation in different climates is of considerable consequence for producers as well as breeders who wish to develop stocks with improved performance in other than optimal environments. Moreover, the higher BW7 of ST1 than the other stocks in the summer of Turkey could be attributed to the rearing of its grandparent flock under similar climate in Israel, as compare to direct importation of ST2 and ST3 from Germany and the U.K. Although this advantage of ST1, after one generation only, could be a mere coincidence, it was reported that a broiler stock that was selected in India for several generations performed better than its U.S.-bred counterpart, under the local hot climate (Singh, 1992). REFERENCES Ajang, O. A., S. Prijono, and W. K. Smith, 1993. Effect of dietary protein content on growth and body composition

EVALUATING BROILER STOCKS IN HOT VS TEMPERATE CLIMATES 929 of fast and slow feathering broiler chickens. Br. Poult. Sci. 34:73 91. Cahaner, A., 1990. Genotype by environment interactions in poultry. Pages 16:13 20 in: Proceedings 4th World Congress on Genetics Applied to Livestock Production. Vol. 16. Edinburgh, U.K. Cahaner, A., and F. Leenstra, 1992. Effects of high temperature on growth and efficiency of male and female broilers from lines selected for high weight gain, favorable feed conversion and high or low fat content. Poultry Sci. 71: 1237 1250. Cahaner, A., N. Deeb, and M. Gutman, 1993. Effect of plumage reducing naked-neck (Na) gene on the performance of fastgrowing broilers at normal and high ambient temperatures. Poultry Sci. 72:767 775. Cahaner, A., R. Yunis, and N. Deeb, 1994. Genetics of feathering and heat tolerance in broilers. Pages 67 70 in: Proceedings 9th European Poultry Conference. Vol. 2. Glasgow, U.K. Chwalibog, A., and B. O. Eggum, 1989. Effect of temperature and performance, heat production, evaporative heat loss and body composition in chickens. Arch. Geflügelkd. 53: 179 184. Eberhart, D. E., and K. W. Washburn, 1993. Assessing the effects of the naked neck gene on chronic heat stress resistance in two genetic populations. Poultry Sci. 72: 1385 1390. Hartmann, W., 1990. Implications of genotype environment interactions in animal breeding: genotype-location interactions in poultry. World s Poult. Sci. J. 46:197 210. Leenstra, F., and A. Cahaner, 1991. Genotype by environment interactions using fast-growing, lean or fat broiler chickens, originated from the Netherlands and Israel, raised at normal or low temperature. Poultry Sci. 70:2028 2039. Leenstra, F., and A. Cahaner, 1992. Effects of low, normal and high temperatures on slaughter yield of broilers from lines selected for high weight gain, favorable feed conversion and high or low fat content. Poultry Sci. 71:1994 2006. Leeson, S., 1987. Nutritional consideration of poultry during heat stress. World s Poult. Sci. J. 42:69 81. Lou, M. L., O. K. Quoi, and W. K. Smith, 1992. Effects of naked neck gene and feather growth rate on broiler in two temperatures. Page 62 in: Proceedings 19th World s Poultry Congress. Vol. 2. Amsterdam, The Netherlands. Mathur, P. K., and P. Horst, 1989. Temperature stress and tropical location as factors for genotype environment interactions. Pages 159 174 in: Genotype Environment Interactions in Poultry Production. P. Merat, ed. INRA, Paris, France. Mathur, P. K., and P. Horst, 1994. Genotype by environment interactions in laying hens based on relationship between breeding values of sires in temperate and tropical environments. Poultry Sci. 73:1777 1784. Osman, A.M.A., E. S. Tawfik, F. W. Klein, and W. Hebeler, 1989. Effect of environmental temperature on growth, carcass traits and meat quality of broilers of both sexes and different ages. Arch. Geflügelkd. 53:158 175. SAS Institute, 1989. SAS User s Guide: Statistics. SAS Institute Inc., Cary, NC. Sheridan A. K., 1990. Genotype environment interaction. Pages 897 912 in: Poultry Breeding and Genetics. R. D. Crawford, ed. Elsevier, Amsterdam, The Netherlands. Singh, H., 1992. Selection for adaptability to sub-optimal conditions. Pages 597 599 in: Proceedings 19th World s Poultry Congress. Vol. 2. Amsterdam, The Netherlands. Teeter, R. G., M. O. Smith, S. Sangray, and F. B. Mather, 1987. Effects of feed intake and fasting duration upon body temperature and survival of thermostressed broilers. Nutr. Rep. Int. 35:531 537. Yahav, S., S. Goldfeld, I. Plavnik, and S. Hurwitz, 1995. Physiological response of chickens and turkeys to relative humidity during exposure to high ambient temperature. J. Therm. Biol. 20:245 253.