Effect of Lighting Stress on Fluctuating Asymmetry, Heterophil-to-Lymphocyte Ratio, and Tonic Immobility Duration in Eleven Breeds of Chickens

Effect of Lighting Stress on Fluctuating Asymmetry, Heterophil-to-Lymphocyte Ratio, and Tonic Immobility Duration in Eleven Breeds of Chickens J. L. Campo, 1 M. G. Gil, S. G. Dávila, and I. Muñoz Departamento de Mejora Genética Animal, Instituto Nacional de Investigación Agraria y Alimentaria, 28080 Madrid, Spain ABSTRACT The purpose of the present study was to for 16 wk. There was a significant difference between analyze the effect of a lighting stress on the fluctuating asymmetry (FA), the heterophil-to-lymphocyte ratio, and the tonic immobility duration of chickens. The experiment (440 birds) measured the FA of several traits (outer, middle, inner, and hind toe lengths and leg, wing, second primary feather, and spur lengths), the heterophil-to-lymphocyte ratio, and the tonic immobility duration in 36- wk-old hens and cocks of 8 Spanish breeds of chickens lighting treatments in both females and males on the combined FA of the 4 toes (P < 0.01) and the combined FA of toe, leg, wing, feather, and spur (in males) lengths (P < 0.05), the FA of birds housed under continuous light being greater than that of control birds. There was a significant difference (P < 0.001) for the heterophil-to-lymphocyte ratio and the tonic immobility duration between lighting treatments, the ratio being higher and the duration (Black-Barred Andaluza, Black-Red Andaluza, Black being longer in the group of birds housed under Castellana, Buff Prat, Red-Barred Vasca, Red Villafranquina, Birchen Leonesa, and Blue Leonesa), a synthetic breed (Quail Castellana), a White Leghorn population, and the e y tester line, which had been housed in continuous light (24L:0D) or in a light-dark regimen (14L:10D) continuous light. Thus, birds exposed to continuous light were more stressed and fearful than control hens. Results were consistent across the breeds and indicate that a continuous light regimen seriously negatively affects the welfare of birds. Key words: stress, light, fluctuating asymmetry, heterophil-to-lymphocyte ratio, tonic immobility 2007 Poultry Science 86:37 45 INTRODUCTION The vision of chickens evolved in natural light environments and differences between the artificial light provided in poultry houses and that required for effective vision may produce stress and poor welfare. Broilers are usually kept on a continuous or nearly continuous lighting schedule to maximize growth rate (Rozenboim et al., 1999), whereas continuous lighting could be used to increase egg production in layers (Yoo et al., 1986). In relation to animal welfare, a day and night light schedule for birds kept in captivity with a minimum dark period of 6 h is recommended (Prescott et al., 2003), because continuous light programs reduce the opportunity for sleep and sleep deprivation increases physiological stress (Blokhuis, 1983). Additionally, Gordon (1994) suggested that maximal benefit is obtained by rearing broilers under a lighting regimen of 16L:8D, because it reduces physiological stress; Morris (2004) indicated that layers held on a con- 2007 Poultry Science Association Inc. Received April 20, 2006. Accepted September 6, 2006. 1 Corresponding author: jlcampo@inia.es stant 14L:10D day will lay as many eggs as ones given step-up lighting. Although most of the reports on lighting patterns deal with their effects on production, a few of them have studied the effect of light program on 3 frequently used indicators of stress: the level of fluctuating asymmetry (FA; Parsons, 1990), the heterophil-to-lymphocyte ratio (Gross and Siegel, 1983), and the duration of tonic immobility (Gallup, 1979). Information on the effect of lighting schedule on the level of FA is generally lacking, and only 2 studies have analyzed the effect of the lighting regimen on the FA of broiler chickens, with contradictory results. Moller et al. (1999) compared the FA of 3 characters (length, width, and thickness of the metatarsus) in a fastgrowing population of chickens, using 3 different light regimes (16L:8D light-dark cycle, changing light regimen, and continuous light), and they found that chickens reared under continuous light developed larger relative asymmetry. In contrast, Stub and Vestergaard (2001) found that FA for those 3 same characters was similar in broilers reared under continuous light combined with no access to sand and in those exposed to 16L:8D with access to sand. Results on the heterophil-to-lymphocyte ratio are also scarce and contradictory. In broilers, Blair et al. (1993) 37

38 CAMPO ET AL. indicated that there was no effect of lighting pattern (23L:1D or increasing from 6L:18D to 23L:1D), whereas Vo et al. (1998) found that continuous lighting increased the percentage of heterophils and decreased the percentage of lymphocytes in comparison with 16L:8D and 12L:12D. Campo and Dávila (2002) did not find significant differences in the heterophil-to-lymphocyte ratio among 3 lighting regimens (23L:1D, 14L:10D, and 18.5L:5.5D) in hens from 3 breeds. Finally, 3 studies have examined the effect of light program on the duration of tonic immobility. Sanotra et al. (2002) indicated that 2 changing light-dark schedules, with or without a reduced stocking density, decreased the duration of tonic immobility compared with continuous light in broilers. Campo and Dávila (2002) found that hens of a synthetic breed housed under a 23L:1D light program showed longer tonic immobility duration than hens housed under 14L:10D. When light regimen and access to sand were used together, Stub and Vestergaard (2001) found that they did not significantly influence the duration of tonic immobility. The purpose of the present study was to analyze the effect of a lighting stress (continuous light vs. light-dark schedule) on combined FA of multiple characters, the heterophil-to-lymphocyte ratio, and the duration of tonic immobility in 11 breeds of chickens with different genetic backgrounds. It was hypothesized that continuous light would negatively affect the well-being of birds. MATERIALS AND METHODS Hatches were made from April 5 to May 31. Birds were reared with a density of 10 birds/m 2 until 8 wk of age (sexes mixed). Artificial light was only provided during the first week of age (23L:1D). Temperature was controlled with gas heaters (33 to 35 C at the chick level during the first week, followed by a reduction of 3 C each week until 18 to 20 C was reached in the sixth week of age). Birds were reared in another all-litter house with a density of 6 birds/m 2 from 8 to 20 wk of age. The lighting regimen was 8L:16D. Birds were fed standard rearing diets containing 19% CP, 2,800 kcal of ME/kg, 1% Ca, and 0.5% available P until 8 wk of age and 15% CP, 2,700 kcal of ME/kg, 0.9% Ca, and 0.4% available P until 20 wk of age. Birds were housed at 20 wk of age (sexes separate) in pens with a raised slatted floor covering a dropping pit and straw litter on the rest of the floor; the slatted area occupied one-third of the floor. Bird density was 4 birds/m 2. Birds were fed a diet containing 16% CP, 2,700 kcal of ME/kg, 3.5% Ca, and 0.5% available P. Feed and water were supplied for ad libitum consumption. The room temperature was 16 to 20 C. Feeders, drinkers, and nest boxes were in the slatted floor area. Eight different Spanish breeds of chickens (Black- Barred Andaluza, Black-Red Andaluza, Black Castellana, Buff Prat, Red-Barred Vasca, Red Villafranquina, Birchen Leonesa, and Blue Leonesa), a synthetic breed (Quail Castellana) that originated from an F 2 cross between Black Castellana and Buff Prat (Campo and Orozco, 1986; Campo, 1991), a White Leghorn population originated by crossing 3 strains selected for egg number and egg weight (Campo and Jurado, 1982), and the e y tester line (Smyth, 1976) was used in the current study. All these breeds were maintained at the experimental station of El Encín (Madrid, Spain) in a conservation program of genetic resources started in 1975 (Campo and Orozco, 1982) and have been described by Campo (1998). Black-Barred Andaluza, Black-Red Andaluza, and Black Castellana are white shell egg breeds, whereas Buff Prat is a tinted shell egg breed and Red-Barred Vasca and Red Villafranquina are brown and dark brown shell egg breeds, respectively. Males of the Birchen Leonesa and Blue Leonesa breeds are used for fly tying. Five replicates (hatches) separated by 14 d were used in the experiment. A total of 440 birds (220 females and 220 males) was used to study the effect of a lighting stress on the fluctuating asymmetry, the heterophil-tolymphocyte ratio, and the tonic immobility duration at 36 wk of age. Within sexes, the birds were equally divided into 2 groups. Group 1 (stress) consisted of 110 birds, 10 birds of each population in 5 replicates of 2 birds, housed in continuous light (24L:0D). Group 2 (control) consisted of 110 additional birds, 10 birds of each population in 5 replicates of 2 birds, housed in a lighting regimen of 14L:10D (light from 0700 to 2100 h). Light intensity was 15 lx. The measured morphological traits were both right (RT) and left (LT) outer, middle, inner, and hind toe lengths and leg (metatarsus), wing (radius), second primary feather, and spur (in males) lengths. Only undamaged birds with intact second primary feathers were analyzed for feather length (216 and 214 instead of 220 females and 220 males, respectively). There was no molt of the primary feathers. Right and LT values of a bird were taken during the same session. All lengths were measured in millimeters using a digital calliper. Trait length was the mean of the RT and LT traits [(RT + LT)/ 2]. All traits showed normal frequency distributions. The FA for a trait, a measure of chronic stress (Moller and Swaddle, 1997), was defined by the absolute difference between sides [ RT LT ]. The unsigned value (equivalent to the mean deviation) is more appropriate for these data than the variance of RT LT sides. A series of steps (Palmer, 1994; Knierim et al., 2006) was followed before identifying exhibited asymmetry as FA (normal distribution of signed RT LT differences with a mean of 0), because there are several confounding factors that complicate the analysis of asymmetry. First, the presence of directional asymmetry (DA; normal distribution with a mean of not 0) and antisymmetry (AS; nonnormal distribution with a mean of 0) was tested by the inspection of the distribution of (RT LT). The presence of DA was tested for using t-tests. Departures from normality (e.g., AS) were assessed using skewness and kurtosis measures; AS is characterized by bimodal (or broad-peaked) distributions, tending to be platykurtic (more intermediate values than the normal distribution). Second, FA and measurement error are normally distributed about a mean

LIGHT STRESS AND FLUCTUATING ASYMMETRY 39 of 0. Thus, it is essential to show that the variance in asymmetry observed among individuals is greater than the variance due to measurement error; the FA is often small and sometimes of the same magnitude as measurement error. Twenty females and 20 males were randomly chosen and measured 3 times on 3 different days. Measures were analyzed using a 2-way ANOVA (Leamy, 1984) with side (fixed) and bird (random) as main factors (1 and 19 df), their interaction (19 df), and the measurement error (80 df). Significant variation between sides indicates variation in DA, whereas a significant interaction indicates significant FA (in the absence of AS). Finally, the product-moment correlation between FA and trait length was used to determine if they were independent. If a positive relationship was found between the mean value and asymmetry of a trait, this effect would be removed by dividing the absolute asymmetry score by the trait mean, defined as the relative FA: [2 RT LT / (RT + LT)]. Relative FA for all traits had distributions that were not normal and were transformed to arcsin square root before analysis. Mean relative asymmetry was defined as the mean of the relative asymmetries of the different traits. If no size correction was needed, each RT LT value was standardized and averaged across traits to give a total score. On 2 different days, birds were tested for heterophilto-lymphocyte ratio and tonic immobility duration at 36 wk of age. Heterophil-to-lymphocyte ratio is a reliable indicator of stress in poultry (Gross and Siegel, 1983). To obtain the heterophil-to-lymphocyte ratio (on the same day as morphological traits), birds were carried to a separate room, and blood was collected immediately. Two drops of blood were taken from a small puncture in the comb of each bird, 1 drop being smeared on each of 2 glass slides. The smears were stained using May-Grünwald and Giemsa stains (Lucas and Jamroz, 1961), approximately 2 to 4 h after methyl alcohol fixation. One hundred leukocytes, including granular (heterophils, eosinophils, and basophils) and nongranular (lymphocytes and monocytes), were counted on 1 slide of each bird (the other slide was supplementary), and the heterophilto-lymphocyte ratio was calculated. Square root transformation was used before analysis. All birds were tested for tonic immobility in a separate room on the day following the blood sampling. Tonic immobility is a traditional measure of fearfulness in poultry (Gallup, 1979), and fear is considered a measure of low welfare. Tonic immobility was induced, as soon as a bird was caught, by placing the bird on its back with the head hanging in a U-shaped wooden cradle (Jones and Faure, 1981). The bird was restrained for 10 s. The observer sat in full view of the bird, about 1 m away, and fixed his eyes on the bird because of the fear-inducing properties of eye contact. If the bird remained immobile for 10 s after the experimenter removed his hands, a stopwatch was started to record latencies (s) until the bird righted itself. If the bird righted itself in <10 s, then it was considered that tonic immobility had not been induced, and the restraint procedure was repeated (3 times maximum). If the bird did not show a righting response over the 10-min test period, a maximum score of 600 s was given for righting time. Thus, tonic immobility duration ranged from 0 to 600 s. Logarithmic transformation was used before analysis. To test the differences in FA, heterophil-to-lymphocyte ratio, and tonic immobility duration between lighting treatments, a 3-way ANOVA (Sokal and Rohlf, 1981) was used with the statistical model x ijkl = +L i +B j +LB ij + r k +Lr ik +Br jk + LBr ijk + ε ijkl, where x ijkl = the analyzed measurement; = the overall mean; L i = the effect of lighting treatment (i = 1, 2); B j = the effect of breed (j = 1 to 11); r k = the effect of replicate (k =1to5);LB ij,lr ik, Br jk, and LBr ijk = the interactions; and ε ijkl = the residual (l = 1, 2). and breed were considered fixed effects, and replicates were assumed to be a random effect. Significant differences among breeds were estimated using the Student-Newman-Keuls multiple range test (Snedecor and Cochran, 1980). RESULTS Effect of Confounding Factors In females, hind toe and wing lengths exhibited significant DA (t = 3.25, P < 0.001, and t = 2.68, P < 0.01, respectively), and the RT side was consistently greater than the LT side. Signed RT LT differences were significantly leptokurtic (more small and more large values relative to the normal distribution) for feather length (g 2 = 0.62, P < 0.05). Thus, there was no evidence of AS, because bimodal (or broad-peaked) distributions tend to be platykurtic. In males, inner toe and wing lengths showed significant DA (t = 3.73, P < 0.001, and t = 9.29, P < 0.001, respectively), the LT side being consistently greater than the RT side or vice versa. Signed RT LT differences were significantly left-skewed for outer and middle toe lengths (g 1 = 0.37, P < 0.05, and g 1 = 0.36, P < 0.05, respectively), right-skewed for inner toe length (g 1 = 0.43, P < 0.01), and leptokurtic for inner toe length and feather length (g 2 = 0.74, P < 0.05, and g 2 = 1.43, P < 0.01, respectively), with no evidence of AS. Mean square error for outer, middle, inner, and hind toe, leg, wing, and feather lengths (Table 1) represented 30.09, 50.28, 87.73, 71.15, 54.39, 104.63, and 5.00% of the interaction mean square in females and 31.07, 31.11, 17.48, 57.60, 49.40, 177.18, and 32.54% in males (13.28% for spur length). Additionally, the interaction was significant for outer, middle, inner (in males), hind (in males) toe, leg, feather, and spur (in males) lengths. Thus, measurement error was not confounded with FA (except for wing length in both sexes), and the non-da identified in the analysis was categorized as true FA. Measurement error accounted for 8.01, 8.51, 16.15, 18.87, 0.98, 13.21, and 2.19% of the total variation in females and 27.57, 24.16, 31.74, 47.69, 4.72, 10.37, and 4.22% in males (2.90% for spur length), respectively. Between sides mean square was not significant for outer (in males), middle, inner, hind toe,

40 CAMPO ET AL. Table 1. Mean squares for ANOVA on fluctuating asymmetry of various lengths measured for 20 females and 20 males in 3 sessions Source of variation Sex Outer toe Middle toe Inner toe Hind toe Leg Wing Feather Females Sides (1) 75.30** 5.13 0.07 0.03 18.45** 19.45 1.71 Birds (19) 51.42*** 59.14*** 23.71*** 10.29*** 102.50*** 101.97*** 176.58*** Interaction (19) 5.15*** 3.54* 1.63 1.04 1.82* 4.75 26.20*** Error (80) 1.55 1.78 1.43 0.74 0.99 4.97 1.31 Males 1 Sides (1) 8.42 5.14 6.64 0.55 0.49 2.34 174.39* Birds (19) 22.82* 27.30** 11.95 8.47* 102.66*** 96.64*** 486.06*** Interaction (19) 8.27*** 8.42*** 9.15*** 3.42* 3.38* 2.06 22.00*** Error (80) 2.57 2.62 1.60 1.97 1.67 3.65 7.16 1 Mean squares for spur length were 3.47, 19.16 (P < 0.001), 1.43 (P < 0.001), and 0.19, respectively. *P < 0.05; **P < 0.01; ***P < 0.001. leg (in males), wing, feather (in females), and spur (in males) lengths and provided no evidence of DA. There was a significant positive correlation coefficient between the value of absolute asymmetry and characteristic size for outer toe (0.15, P < 0.05), leg (0.23, P < 0.001), wing (0.17, P < 0.01), and feather lengths (0.20, P < 0.01), whereas the relationships for middle, inner, and hind toes and spur lengths were not significant. Relative asymmetry could, therefore, be used to control the scaling effect in the former traits. The relative asymmetry ranged from 4 to 7% of trait size for the toes and spur lengths and was <2% of trait size for leg, wing, and feather lengths. The CV of each trait were 6.8, 6.6, 6.7, 6.8, 5.1, 4.8, 5.4, and 21.4, respectively, for outer, middle, inner, and hind toe, leg, wing, feather, and spur lengths. Analysis of FA Replicates, replicate lighting treatment, replicate breed, and replicate lighting treatment breed interac- tions were not significant for any measurement, and they were pooled with the residual to give a 2-way ANOVA of lighting treatment and breed effects [x ijk = +L i +B j +LB ij + ε ijk, where x ijk = the analyzed measurement; = the overall mean; L i = the effect of lighting treatment (i = 1, 2); B j = the effect of breed (j =1to11); LB ij, = the interaction; and ε ijk = the residual (k =1to10)]. Mean values indicating the effect of the lighting treatment and the breed on relative asymmetry measurements of females are summarized in Tables 2 and 3. Lighting effect was significant for the relative asymmetry of outer toe length (P < 0.05), the combined relative asymmetry of the 4 toes (P < 0.01), wing length (P < 0.001), and the combined relative asymmetry of leg, wing, feather, and toes lengths (P < 0.05). The relative asymmetry of females housed with continuous light was larger than that of control females (except for wing length). Although there were significant lighting treatment breed interactions for the relative asymmetry of outer toe lengths, results were consistent Table 2. Mean relative asymmetry ( 100) in outer, middle, inner, and hind toe lengths and combined toes in females from 11 different breeds housed in 2 different lighting treatments (n = 220) Toe length Effect Outer Middle Inner Hind Combined 24L:0D 7.48 a 4.51 a 4.79 a 4.09 a 5.22 a 14L:10D 6.29 b 3.39 b 5.12 a 3.84 a 4.66 b SEM 0.31 0.23 0.23 0.18 0.13 Black-Barred Andaluza 6.05 bc 3.44 a 4.57 ab 3.37 b 4.36 b Black-Red Andaluza 6.28 bc 4.29 a 4.13 b 3.65 b 4.59 b Black Castellana 5.90 bc 3.68 a 5.40 ab 3.83 b 4.71 b Quail Castellana 6.68 bc 3.60 a 6.68 a 4.37 b 5.34 b Buff Prat 6.72 bc 4.28 a 4.59 ab 4.61 b 5.05 b Red-Barred Vasca 6.89 bc 3.81 a 4.44 ab 3.93 b 4.77 b Red Villafranquina 5.46 c 4.20 a 4.38 ab 3.24 b 4.32 b Birchen Leonesa 6.86 bc 3.76 a 4.41 ab 3.40 b 4.61 b Blue Leonesa 6.71 bc 4.10 a 5.38 ab 3.70 b 4.97 b White Leghorn 8.93 ab 3.67 a 4.13 b 4.06 b 5.20 b e y Tester line 9.87 a 4.76 a 6.79 a 5.85 a 6.82 a SEM 0.73 0.55 0.55 0.43 0.31 a c Means within the same effect and column with no common superscript differ (P < 0.05).

LIGHT STRESS AND FLUCTUATING ASYMMETRY 41 Table 3. Mean relative asymmetry ( 100) of leg, wing, and feather lengths and combined toes, leg, wing, and feather in females from 11 different breeds housed in 2 different lighting treatments (n = 220) Length Effect Leg Wing Feather Combined 24L:0D 1.44 a 1.49 b 1.69 a 3.71 a 14L:10D 1.34 a 2.02 a 1.71 a 3.43 b SEM 0.08 0.08 0.15 0.08 Black-Barred Andaluza 1.26 a 1.49 bc 2.36 b 3.23 b Black-Red Andaluza 1.33 a 2.32 ab 2.04 b 3.45 b Black Castellana 1.04 a 1.56 bc 1.30 b 3.25 b Quail Castellana 1.43 a 1.61 bc 2.54 b 3.87 b Buff Prat 1.30 a 2.43 a 0.70 b 3.52 b Red-barred Vasca 1.34 a 1.40 c 1.11 b 3.27 b Red Villafranquina 1.16 a 1.85 abc 1.42 b 3.10 b Birchen Leonesa 1.63 a 1.96 abc 1.35 b 3.34 b Blue Leonesa 1.78 a 1.47 bc 5.40 a 3.89 b White Leghorn 1.18 a 1.57 bc 1.23 b 3.54 b e y Tester line 1.97 a 1.61 bc 4.17 a 5.14 a SEM 0.20 0.19 0.35 0.17 a c Means within the same effect and column with no common superscript differ (P < 0.05). across the breeds, the lighting treatment being larger than the control in all the breeds (quantitative interaction). The lighting treatment breed interaction was significant and qualitative (the treatments did not keep their rankings in all breeds) for the relative asymmetry of the middle toe length, the lighting treatment being significantly larger than the control in the Black-Red Andaluza, Buff Prat, and Red Villafranquina breeds. There were significant differences among breeds for the relative asymmetry of outer (P < 0.05), inner (P < 0.01), and hind toe lengths (P < 0.01); the combined relative asymmetry of the 4 toes (P < 0.001); wing length (P < 0.01); feather length (P < 0.001); and the combined relative asymmetry of leg, wing, feather, and toe lengths (P < 0.001), the relative asymmetry being larger (except for wing length) in the e y tester line than in the other breeds. Results from males (Tables 4 and 5) indicated that there were significant effects of lighting treatment on relative asymmetry of outer toe (P < 0.001), the combined relative asymmetry of the 4 toes (P < 0.01), feather length (P < 0.01), and the combined relative asymmetry of leg, wing, feather, spur, and toe lengths (P < 0.05), with the relative asymmetry for males housed with continuous light being larger than that of control males. The lighting treatment breed interaction was significant but quantitative for the relative asymmetry of outer toe length and the combined relative asymmetry of leg, wing, feather, spur, and toe lengths, the lighting treatment being larger than the control in all the breeds. The lighting treatment breed interaction was significant and qualitative for the relative asymmetry of wing length, the lighting treatment being significantly lower than the control in the Black-Barred Andaluza, Black Castellana, Birchen Leonesa, and Buff Prat breeds. There were significant differences among breeds for the relative asymmetry of middle (P < 0.01) and inner toe lengths (P < 0.05), the combined relative asymmetry of the 4 toes (P < 0.05), spur length (P < 0.001), and the combined relative asymmetry of leg, wing, feather, spur, and toe lengths (P < 0.001), with the relative asymmetry being larger in the Buff Prat breed for the relative asymmetry of middle toe length and the combined relative asymmetry of the 4 toes, in the e y tester line for the relative asymmetry of inner toe length, and in the Red-Barred Vasca breed for the relative asymmetry of spur length, and the combined relative asymmetry of leg, wing, feather, spur, and toe lengths. Analysis of Heterophil-to-Lymphocyte Ratio and Tonic Immobility Duration Mean values indicating the effect of lighting treatment and breed on heterophil-to-lymphocyte ratio and tonic immobility duration for females are summarized in Table 6. There was a significant difference between lighting treatments for the heterophil-to-lymphocyte ratio (P < 0.001), with the ratio of females housed with continuous light being higher than that of control females. In females housed with continuous light, there was a significant (P < 0.001) increase in heterophil number, with a significant (P < 0.001) decrease in lymphocytes. Although there were significant lighting treatment breed interactions for heterophil-to-lymphocyte ratio, heterophil number, and lymphocyte number, results were consistent across the breeds, the lighting treatment being higher than the control in all the breeds (quantitative interaction) for heterophil-to-lymphocyte ratio and heterophil number and smaller for lymphocyte number. There were significant differences among breeds in terms of heterophil-to-lymphocyte ratio, heterophil number, and lymphocyte number (P < 0.001). Ratio and heterophil number were greater

42 CAMPO ET AL. Table 4. Mean relative asymmetry ( 100) in outer, middle, inner, and hind toe lengths and combined toes in males from 11 different breeds housed in 2 different lighting treatments (n = 220) Toe length Effect Outer Middle Inner Hind Combined 24L:0D 7.40 a 4.58 a 5.66 a 4.63 a 5.57 a 14L:10D 4.99 b 4.27 a 5.15 a 4.51 a 4.73 b SEM 0.28 0.23 0.27 0.23 0.13 Black-Barred Andaluza 6.21 a 4.32 ab 5.57 ab 3.67 a 4.94 ab Black-Red Andaluza 6.88 a 3.72 ab 5.01 ab 5.31 a 5.23 ab Black Castellana 5.81 a 5.24 ab 5.66 ab 4.59 a 5.33 ab Quail Castellana 5.50 a 5.81 a 6.64 ab 5.24 a 5.80 ab Buff Prat 7.57 a 5.99 a 5.28 ab 4.57 a 5.86 a Red-Barred Vasca 8.28 a 3.80 ab 5.69 ab 4.41 a 5.55 ab Red Villafranquina 5.51 a 4.45 ab 4.59 ab 4.51 a 4.76 ab Birchen Leonesa 5.38 a 4.15 ab 4.01 b 5.35 a 4.72 ab Blue Leonesa 5.25 a 3.14 b 4.82 ab 4.00 a 4.31 b White Leghorn 5.82 a 4.15 ab 5.09 ab 3.98 a 4.76 ab e y Tester line 6.21 a 3.67 ab 7.49 a 4.60 a 5.50 ab SEM 0.66 0.54 0.65 0.54 0.32 a,b Means within the same effect and column with no common superscript differ (P < 0.05). in the e y tester line, the White Leghorn population, and the Black-Barred Andaluza breed, whereas they were significantly smaller for the Birchen Leonesa breed, with the opposite being true for lymphocyte number. The tonic immobility duration was significantly longer within the group of females housed with continuous light than within the control group of females (P < 0.001; Table 6). breed interaction was significant but quantitative, indicating consistent differences between treated and control females. There were significant differences among breeds (P < 0.001), females from the White Leghorn and Black-Barred Andaluza breeds having longer tonic immobility duration, with tonic immobility of females from the Black Castellana breed being shorter. Results were similar for males (Table 7). The heterophilto-lymphocyte ratio, the heterophil number, and the tonic immobility duration were significantly greater, and the lymphocyte number was significantly smaller in males housed with continuous light than in the control males (P < 0.001). In addition, the lighting treatment breed interaction was quantitative. There were significant dif- Table 5. Mean relative asymmetry ( 100) of leg, wing, feather, and spur lengths and combined toes, leg, wing, feather, and spur in males from 11 different breeds housed in 2 different lighting treatments (n = 220) Length Effect Leg Wing Feather Spur Combined 24L:0D 1.19 a 1.26 b 2.67 a 7.17 a 4.35 a 14L:10D 1.06 a 1.74 a 1.89 b 7.70 a 3.96 b SEM 0.06 0.07 0.21 0.54 0.10 Black-Barred Andaluza 1.10 a 1.76 a 2.49 a 8.29 bc 4.21 bc Black-Red Andaluza 0.96 a 1.53 a 2.58 a 5.46 c 3.96 bc Black Castellana 1.11 a 1.66 a 1.78 a 7.23 bc 4.15 bc Quail Castellana 1.33 a 1.49 a 2.22 a 7.75 bc 4.54 abc Buff Prat 1.05 a 1.65 a 1.96 a 11.46 ab 4.98 ab Red-Barred Vasca 1.58 a 1.61 a 2.42 a 12.99 a 5.17 a Red Villafranquina 1.09 a 1.55 a 2.13 a 6.78 bc 3.83 c Birchen Leonesa 0.99 a 1.52 a 1.60 a 5.05 c 3.52 c Blue Leonesa 0.91 a 1.14 a 1.52 a 5.47 c 3.49 c White Leghorn 0.96 a 1.39 a 2.44 a 6.43 bc 3.78 c e y Tester line 1.33 a 0.96 a 4.06 a 3.91 c 4.02 bc SEM 0.15 0.17 0.51 1.27 0.24 a c Means within the same effect and column with no common superscript differ (P < 0.05).

LIGHT STRESS AND FLUCTUATING ASYMMETRY 43 Table 6. Mean heterophil:lymphocyte, heterophil number, lymphocyte number, and tonic immobility duration (s) in females from 11 different breeds housed in 2 different lighting treatments (n = 220) Heterophil: Heterophil Lymphocyte Tonic Effect lymphocyte number number immobility 24L:0D 1.13 a 49.25 a 45.87 b 428.71 a 14L:10D 0.36 b 23.49 b 69.83 a 243.29 b SEM 0.02 0.59 0.62 6.89 Black-Barred Andaluza 0.92 a 43.10 ab 53.25 cd 395.15 a Black-Red Andaluza 0.58 cd 32.00 c 62.60 a 339.00 abc Black Castellana 0.65 cd 31.90 c 55.35 bc 280.30 c Quail Castellana 0.89 ab 40.65 b 53.95 c 301.85 bc Buff Prat 0.73 bc 34.55 c 60.25 ab 349.15 abc Red-Barred Vasca 0.72 bc 34.80 c 60.15 ab 300.70 bc Red Villafranquina 0.64 cd 33.00 c 61.70 a 336.55 abc Birchen Leonesa 0.49 d 29.55 c 63.60 a 324.25 bc Blue Leonesa 0.67 cd 35.40 c 60.30 ab 357.30 ab White Leghorn 0.97 a 42.55 ab 52.85 cd 402.45 a e y Tester line 1.03 a 47.07 a 48.35 d 311.14 bc SEM 0.05 1.40 1.45 16.16 a d Means within the same effect and column with no common superscript differ (P < 0.05). ferences among breeds (P < 0.001); the heterophil-to-lymphocyte ratio and heterophil number were highest and the lymphocyte number smallest in the White Leghorn, whereas the reverse was true in the Birchen Leonesa. Tonic immobility duration was longer in the Black-Red Andaluza, Blue Leonesa, and White Leghorn breeds and shorter in the Buff Prat breed. DISCUSSION The main finding of the current study was that the lighting stress significantly affected FA, heterophil-to- lymphocyte ratio, and tonic immobility duration. The difference between continuous light (24L:0D) and the lightdark regimen (14L:10D) for the combined FA of morphological traits was, on average, almost 10%, the heterophilto-lymphocyte ratio was more than 3 as large under continuous light as compared with the light-dark schedule, and, finally, continuous light increased tonic immobility duration almost 75% compared with the light-dark treatment. These results suggest that continuous light seriously negatively affects the welfare of birds. The current study fulfilled the criteria to estimate the FA for a trait confidently (Palmer, 1994; Knierim et al., Table 7. Mean heterophil:lymphocyte, heterophil number, lymphocyte number, and tonic immobility duration (s) in males from 11 different breeds housed in 2 different lighting treatments (n = 220) Heterophil: Heterophil Lymphocyte Tonic Effect lymphocyte number number immobility 24L:0D 1.21 a 51.50 a 43.73 b 489.85 a 14L:10D 0.35 b 23.27 b 70.30 a 276.02 b SEM 0.02 0.44 0.45 8.88 Black-Barred Andaluza 0.94 b 41.20 c 54.40 c 348.55 bc Black-Red Andaluza 0.79 bcd 37.40 d 57.50 bc 466.90 a Black Castellana 0.66 de 33.65 de 60.20 ab 385.25 abc Quail Castellana 0.72 de 36.75 d 57.45 bc 387.25 abc Buff Prat 0.69 de 34.60 de 59.75 ab 307.90 c Red-Barred Vasca 0.75 cd 35.40 de 58.55 b 338.45 bc Red Villafranquina 0.69 de 31.85 e 61.85 ab 407.40 ab Birchen Leonesa 0.57 c 32.10 e 63.10 a 337.85 bc Blue Leonesa 0.91 b 44.83 ab 50.22 d 444.83 a White Leghorn 1.19 a 47.30 a 47.70 d 436.25 a e y Tester line 0.88 bc 42.12 bc 50.56 d 391.69 abc SEM 0.04 1.05 1.06 20.84 a e Means within the same effect and column with no common superscript differ (P < 0.05).

44 CAMPO ET AL. 2006) without the effect of confounding factors. The type of asymmetry was determined, an assessment was made of measurement error, and the relationship between the amount of asymmetry and trait size was controlled. Additionally, combined FA estimates of multiple traits were used; the effects of lighting treatment, breed, and its interaction were estimated using the ANOVA and Student- Newman-Keuls multiple-range procedures; and a large sample size was used. Although there were significant effects of the lighting stress on the FA for outer toe and feather (in males) lengths, these traits should be excluded from the analysis in females and males, respectively, because they showed significant DA in the ANOVA (Leamy, 1984). Wing length should be excluded, too, because it showed significant DA using the t-test in both females and males, and, additionally, estimates of FA were confounded with measurement error. Then, there were only 2 single traits with significant lighting effect on their FA (i.e., middle toe length in females and outer toe length in males), although in the former the effect of light only was significant in 3 breeds. However, there were significant effects in both females and males of the lighting stress on the combined FA of the 4 toes and the combined FA of all lengths, indicating that FA is a valid indicator of environmental stress (Parsons, 1990), especially if the analysis is based on the composite FA of multiple traits (Leung et al., 2000). In all the cases, the FA of birds housed with continuous light was larger than that of control birds. This result confirmed the finding of Moller et al. (1999) that broiler chickens reared under continuous light developed larger mean relative FA for length, width, and thickness of the metatarsus (4.7%) than chickens reared under 16L:8D or changing light regimes (3.6 and 3.3%, respectively), although these authors found a significant effect on the FA for 2 of the 3 single traits (length and thickness of the metatarsus; 2.4 vs. 1.8% and 6.8 vs. 4.8%, respectively). On the contrary, the finding in the current study disagrees with the result reported by Stub and Vestergaard (2001), who did not find any significant effect on the degree of FA in the chickens reared under the continuous light-no sand and the 16L:8D-sand conditions, although the effect of each factor could not be separated. The difference in heterophil-to-lymphocyte ratio among birds housed with the 2 lighting regimes was significant in both sexes. Birds housed with continuous light showed higher ratios than those housed with 14L:10D, birds in the former group having significant heterophilia and lymphopenia. Thus, it is suggested that a continuous lighting schedule was associated with an increased stress response and affected bird welfare. This result agrees with that of Vo et al. (1998), who reported that heterophil-to-lymphocyte ratio was higher in the continuous light group of birds. However, the result in the current study disagrees with those of Blair et al. (1993) and Campo and Dávila (2002), who reported that heterophil-to-lymphocyte ratio was unaffected by a nearly continuous lighting schedule (23L:1D). There were significant differences between lighting treatments in both sexes for the tonic immobility duration. Tonic immobility of birds housed under 24L:0D was longer than that of birds housed under 14L:10D, suggesting that continuous lighting treatment caused greater fearfulness. This finding agrees with results reported by other authors, who found an association between lighting treatment and fearfulness. Sanotra et al. (2002) reported that changing light-dark schedules, in combination or not with reduced stocking density, had a decreasing effect on fearfulness in broiler chickens in comparison with a continuous light regimen. Similarly, Campo and Dávila (2002) reported a decreasing effect on fearfulness in a synthetic breed for a 14L:10D light treatment in comparison with a nearly continuous lighting schedule (23L:1D). On the contrary, the result disagrees with the finding by Stub and Vestergaard (2001), who did not indicate any significant effect associated with a continuous light treatment in broiler chickens in comparison with a 16L:8D light treatment. In conclusion, results of the current study indicated that adult chickens from several diverse populations housed with continuous light have increased combined relative FA value of several traits, the heterophil-to-lymphocyte ratio, and the tonic immobility duration and are more stressful and fearful than birds housed under a light-dark program (14L:10D). It is suggested that a continuous light regimen causes severe negative welfare consequences. REFERENCES Blair, R., R. C. Newberry, and E. E. Gardiner. 1993. Effects of lighting pattern and dietary tryptophan supplementation on growth and mortality in broilers. Poult. Sci. 72:495 502. Blokhuis, H. J. 1983. The relevance of sleep in poultry. World s Poult. Sci. J. 39:33 37. Campo, J. L. 1991. Use of the sex-linked barring (B) gene for chick sexing on an eumelanotic columbian background. Poult. Sci. 70:1469 1473. Campo, J. L. 1998. Conservation and genetical study of Spanish chicken breeds (28). Pages 155 158 in Proc. 6th World Congr. Genet. Appl. Livestock Prod., Armidale, Australia. Univ. New England, Armindale, New South Wales, Australia. Campo, J. L., and S. G. Dávila. 2002. Effect of photoperiod on heterophil to lymphocyte ratio and tonic immobility duration of chickens. Poult. Sci. 81:1637 1639. Campo, J. L., and J. J. Jurado. 1982. Evaluation of multiple trait selection in strains of layers (8). Pages 869 874 in Proc. 2nd World Congr. Genet. Appl. Livestock Prod., Madrid, Spain. Editorial Garsi, Madrid, Spain. Campo, J. L., and F. Orozco. 1982. Conservation and genetical study of Spanish chicken breeds (6). Pages 88 93 in Proc. 2nd World Congr. Genet. Appl. Livestock Prod., Madrid, Spain. Editorial Garsi, Madrid, Spain. Campo, J. L., and F. Orozco. 1986. Genetic basis of the Melanotic Prat phenotype. Br. Poult. Sci. 27:361 367. Gallup, G. G. 1979. Tonic immobility as a measure of fear in domestic fowl. Anim. Behav. 27:316 317. Gordon, S. H. 1994. Effects of daylength and increasing daylength programmes on broiler welfare and performance. World s Poult. Sci. J. 50:269 282. Gross, W. B., and H. S. Siegel. 1983. Evaluation of the heterophil to lymphocyte ratio as a measure of stress in chickens. Avian Dis. 27:972 979.

LIGHT STRESS AND FLUCTUATING ASYMMETRY 45 Jones, R. B., and J. M. Faure. 1981. Sex and strain comparisons of tonic immobility ( righting time ) in the domestic fowl and the effects of various methods of induction. Behav. Proc. 6:47 55. Knierim, U., S. van Dongen, B. Forkman, F. A. M. Tuyttens, M. Spinka, J. L. Campo, and G. E. Weissengruber. 2006. Fluctuating asymmetry as animal welfare indicator? A review of methodology and validity. Physiol. Behav. (In press) Leamy, L. 1984. Morphometric studies in inbred and hybrid house mice. V. Directional and fluctuating asymmetry. Am. Nat. 123:579 593. Leung, B., M. R. Forbes, and D. Houle. 2000. Fluctuating asymmetry as a bioindicator of stress: Comparing efficacy of analyses involving multiple traits. Am. Nat. 155:101 115. Lucas, A. M., and C. Jamroz. 1961. Atlas of Avian Hematology. Agriculture Monograph 25. USDA, Washington, DC. Moller, A. P., G. S. Sanotra, and K. S. Vestergaard. 1999. Developmental instability and light regime in chickens (Gallus gallus). Appl. Anim. Behav. Sci. 62:57 71. Moller, A. P., and J. P. Swaddle. 1997. Asymmetry, Developmental Stability, and Evolution. Oxford Univ. Press, Oxford, UK. Morris, T. R. 2004. Environmental control for layers. World s Poult. Sci. J. 60:163 175. Palmer, A. R. 1994. Fluctuating asymmetry analyses: A primer. Pages 335 364 in Developmental Instabilty: Its Origins and Evolutionary Implications. T. A. Markow, ed. Kluwer Acad., Dordrecht, the Netherlands. Parsons, P. A. 1990. Fluctuating asymmetry. An epigenetic measure of stress. Biol. Rev. Camb. Philos. Soc. 65:131 145. Prescott, N. B., C. M. Wathes, and J. R. Jarvis. 2003. Light, vision, and the welfare of poultry. Anim. Welf. 12:269 288. Rozenboim, I., B. Robinzon, and A. Rosenstrauch. 1999. Effect of light source and regimen on growing broilers. Br. Poult. Sci. 40:452 457. Sanotra, G. S., J. D. Lund, and K. S. Vestergaard. 2002. Influence of light-dark schedule and stocking density on behaviour, risk of leg problems, and occurrence of chronic fear in broilers. Br. Poult. Sci. 43:344 354. Smyth, J. R. 1976. The inheritance of melanic pigmentation in the fowl. Pages 1 28 in Proc. 25th Natl. Poult. Breed. Roundtable, Kansas City, MO. Univ. Missouri, Kansas City. Snedecor, G. W., and W. G. Cochran. 1980. Statistical Methods. 7th ed. Iowa State Univ. Press, Ames. Sokal, R. R., and F. J. Rohlf. 1981. Biometry. Freeman and Co., London, UK. Stub, C., and K. S. Vestergaard. 2001. Influence of zinc bacitracin, light regimen, and dustbathing on the health and welfare of broiler chickens. Br. Poult. Sci. 42:564 568. Vo, K. V., C. Burgess, N. A. Adefope, C. Catlin, and T. Wakefield. 1998. Effects of lighting regime and feed restriction for various durations on growth performance, leg abnormalities and hemo-stress response of broilers. Poult. Sci. 77(Suppl. 1):3. (Abstr.) Yoo, B. H., B. L. Sheldon, and R. N. Podger. 1986. Analyses of oviposition times and intervals in a wide range of layer flocks under normal and continuous lighting regimes. Br. Poult. Sci. 27:267 288.