Impact of Cage Density on Pullet Performance and Blood Parameters of Stress 1

Impact of Cage Density on Pullet Performance and Blood Parameters of Stress 1 P. H. PATTERSON,2 and H. S. SIEGEL Department of Poultry Science, The Pennsylvania State University, University Park, Pennsylvania 16802-3501 ABSTRACT The effects of cage density on pullet live performance and blood indices of stress were evaluated in two commercial White Leghorn strains housed at 38, 32, 26, and 20 birds per cage from Day 1 to 6 wk of age, and 19, 16, 13, and 10 birds per cage from 6 to 18 wk. Cage densities of 26 and 13 birds per cage represent a U.S. standard of 142 and 284 cm2 per bird that is often applied in commercial pullet rearing. Cage density treatments include confounding cage, feeder, and drinker spaces per bird as might be encountered in commercial practice when growing more pullets per cage. Body weight was significantly reduced at greater bird densities in both strains; however, one strain was affected as early as 6 wk of age, whereas in the other strain, body weight was not reduced significantly until 18 wk. Feed intake was increased by more than 13% in both strains at the lowest density treatment (20 birds per cage) from Day 1 to 2 wk but reduced by more than 9% by higher cage densities during the remainder of the study. Feed conversion (FC) ratio was similarly improved (i.e., reduced) when more pullets were housed per cage, and increased when fewer birds were placed per cage compared to the standard. Despite a significant reduction in feed intake and a corresponding loss of body weight, cage density treatments had no significant affect on hemagglutinin titers to sheep red blood cell antigen, percentage heterophils (H), lymphocytes (L), or the H:L ratio. However, pullet age and strain differences were observed for all blood parameters. Overall, treatments allowing more cage, feeder, and drinker spaces per bird than the U.S. commercial standard provided no body weight advantage, and allowed for greater feed intake and poorer FC at several ages. (Key words: White Leghorn pullet, cage density, heterophil to lymphocyte ratio, sheep red blood cell titer) INTRODUCTION It is a common rearing practice in the U.S. to house Single Comb White Leghorn (SCWL) pullets chicks in multiple deck cage systems at 142 cm2 per bird and then double the area per bird (284 cm2 per bird) at approximately 4 to 8 wk to provide more room and feeder and drinker space as pullets grow (North and Bell, 1990; and DeKalb, 1994). However, chick placements per cage are often modified for a number of reasons, including demand for more started pullets, anticipated mortality, or body weight control. Although the impact of modifying the number of pullets per cage plus or minus a few birds might seem small, the impact on blood parameters of stress has not been evaluated. Several studies have measured the impact of cage density on pullet live performance. Leeson and Summers (1984) compared the effects of providing 293 cm2 1998 Poultry Science 77:32 40 vs 586 cm2 per bird from as early as 2 wk of age in combination with various dietary energy levels as a means of stimulating growth. Irrespective of diet treatment, the more liberal density resulted in a 5 to 8% increase in feed intake; however, increased intake was not associated with increased body weight and the authors assumed that greater nutrient intake was related to a greater maintenance requirement associated with increased bird activity. Reduced pullet feed intake and 18-wk BW were demonstrated in two experiments at greater bird densities between 311 and 222 cm2 per bird when feeder and drinker space were reduced with greater bird numbers (Carey, 1987). However, Anderson and Adams (1992) observed no impact of pullet density (221, 249, 277, and 304 cm2 per bird) on 18-wk BW when cage population and feeder and drinker spaces were held constant. In a second experiment using a 2 2 2 factorial design with two densities, and feeder and drinker spaces, smaller rearing and feeder spaces Received for publication February 26, 1997. Accepted for publication August 14, 1997. 1Where trade names appear in this article, no discrimination is intended, and no endorsement by Penn State University is implied. 2To whom correspondence should be addressed: php1@psu.edu Abbreviation Key: FC = feed conversion; SCWL = Single Comb White Leghorn; H:L = heterophil to lymphocyte ratio; D S = density by strain interaction; T S = time by strain interaction. 32

IMPACT OF CAGE DENSITY ON PULLET PERFORMANCE AND STRESS 33 reduced 18-wk BW whereas the number of birds per drinker cup had no impact. Bell (1969) demonstrated an interaction with negative affects of lower feeder and cage spaces on 15.5-wk pullet BW when varying pullet numbers 6, 8, and 10 per 61 61 cm cage and feeder space 2.5, 5.0, 6.1, and 10.1 cm per bird. Others have studied the effects of floor and cage density on hen performance and other indices of stress. Egg production and mean clutch length were significantly reduced in groups of White Leghorn hens housed on the floor at 1,236 cm2 per bird for 196 d compared with hens at 3,716 cm2 per bird for the same period (Siegel, 1959). Corticosterone concentrations were consistently higher in the serum of SCWL hens housed five per 30.5 50.8 cm cage than in birds housed at three or four per cage (Mashaly et al., 1984). Egg production tended to decline as the area per hen was reduced. Regardless of cage treatment, elevated corticosterone levels were observed until 39 d following housing, indicating that birds need at least this period of time to adapt to the stress of new environmental conditions. The heterophil:lymphocyte ratio (H:L) has been documented as an index of stress in chickens (Gross and Siegel, 1983). Although the influence of cage density on the ratio has not been previously demonstrated, other stressors that negatively impact poultry live performance often cause the H:L ratio to increase. When broiler chicks were exposed to multiple stressors in a factorial design, the main effects of ammonia, electrical shock, and heat stress significantly increased the H:L ratio compared with control birds (McFarlane and Curtis, 1989). Also, exposure to critical levels of infectious agents and toxins has been shown to elicit cellular responses. Trout et al. (1988) reported an increase in the heterophil number of 8-wk-old SCWL injected with Brucella abortus antigen. White Leghorn chicks fed 5,700 ppb dietary aflatoxin from 2 to 42 d of age had an increased H:L ratio (0.41) at 10 d compared to controls (0.23) (Ubosi et al., 1985). The stress responses to moving birds and restricting their feed was studied by Zulkifli et al. (1993) in White Plymouth Rock chickens. Transferring the birds from starter to developer batteries at 22 d of age resulted in an increased H:L ratio within 24 h. Restricting feed intake at 30 d of age to 60% of ad libitum increased the H:L ratio more than 29% for a period of 12 d compared to full-fed controls. When feed restriction was relaxed from 60 to 80% at 43 d of age, H:L ratios were initially increased at 45 d to 0.72, then declined to 0.61 and 0.57 at 50 and 55 d of age, respectively. Considerable differences exist in the literature regarding the humoral immune response to SRBC antigen as a result of various stressors. Exposure to high temperatures has been shown to depress antibody titers to a variety of antigens, including SRBC, Salmonella pullorum, and BSA (Thaxton and Siegel, 1970). Henken et al. (1983a and 1983b) reported an elevated SRBC response in 26-d-old brown egg-type pullets exposed to extreme temperatures, and feed restriction-deprivation treatments that also impacted feed intake, feed conversion (FC), BW, and growth. By contrast, no difference was observed in the antibody titers of 55-d-old dwarf or normal White Plymouth Rock chickens that consumed feed ad libitum or were restricted to 80 or 60% of normal consumption from 43 d of age (Zulkifli et al., 1993). In another experiment reported by Donker et al. (1990), chickens selected over six generations for high or low antibody titers to SRBC were immunized i.m. at 24 h after four heat stress periods of 30 min at 42 C with an intervening 30-min period at 22 C. These authors determined that heat stress had little or no effect on antibody production in either of the lines studied. Although considerable work has focused on the impact of cage density on hen performance, behavior, and measures of stress (Hester and Wilson, 1986; Anderson et al., 1989; Hester et al., 1996a,b,c) little research has addressed the effect of varying pullet cage densities on blood parameters of stress. Therefore, the purpose of this study was to measure the effect of cage densities greater than and less than the U.S. standard on pullet live performance and blood indices of stress. Cage density was adjusted by varying the number of pullets per cage with full knowledge of the confounding effects of feeder and drinker space because of the practical implications to an industry setting, in which adjusting pullet numbers would similarly modify cage, feeder, and drinker spaces. MATERIALS AND METHODS Pullet Treatments and Management Two commercial strains of SCWL pullets (DeKalb Delta and Hy-Line W-36) were reared with two nipple drinkers per 61 61 cm cage at four cage densities, which included 38, 32, 26, and 20 pullets chicks per cage providing 97.8, 116.1, 142.9, and 185.8 cm2 per bird, respectively, from Day 1 to 6 wk of age. Linear feeder space was 61 cm per cage or 1.60, 1.91, 2.35, and 3.05 cm per bird at the 38, 32, 26, and 20 birds per cage densities, respectively. At 6 wk, half the birds in each cage were moved to another cage immediately below the brooding cage, thereby doubling the number of replicate cages and reducing bird density to 19, 16, 13, and 10 birds per cage. Cage space was thus increased to 195.6, 232.3, 285.9, and 371.6 cm2 per bird and feeder space to 3.21, 3.81, 4.69, and 6.10 cm per bird from 6 to 18 wk. Cage area per bird and feeder space were reduced and the number of birds per nipple drinker were increased with increasing bird number per cage, just as would occur in a commercial setting when rearing more birds per cage. The cage density treatments of 26 and 13 birds per cage from Day 1 to 6 wk and 6 to 18 wk, respectively, correspond closely with the U.S. standard of 142 and 284 cm2 per bird for the same approximate ages. Brooding temperatures in the environmentally controlled house were maintained at 32.8 C during the 1st wk

34 and reduced 1.7 C/wk through Week 3. Temperatures were further reduced 2.8 C/wk from Week 4 to 6, and then maintained at 21.1 C throughout the remainder of the study. Light was provided at an intensity of 10 lx for 22 h/ d during the 1st wk then reduced to 5 lx and 18, 16, and 10 h/d during Weeks 2, 3, and 4, respectively. During Weeks 4 to 16 light was maintained at 5 lx for 10 h/d. At 17 wk, light intensity was increased to 10 lx and 10.5 h/d, followed by 11 h at 18 wk. All chicks were precision beak trimmed (4.75-mm plate hole diameter) at 13 d of age. Feed and water were provided for ad libitum consumption throughout the study. Commercial diets containing predominantly corn and soybean meal were fed. During the starter period from Day 1 to 6 wk the diet contained 19% CP, 2,920 kcal ME/kg, 0.48% available P, and 1.0% Ca (calculated nutrient concentration) and the birds were hand fed once per day at approximately 0800 h. The grower diet fed from 7 to 12 wk contained 15% CP, 2,880 kcal ME/kg, 0.44% available P, and 1.0% Ca. A developer diet containing 13% CP, 2,860 kcal ME/kg, 0.43% available P, and 1.25% Ca was fed from 13 to 16 wk. A prelay diet containing 16% CP, 2,820 kcal ME/kg, 0.43% available P, and 2.0% Ca was fed from 17 to 18 wk of age. From 7 through 18 wk of age, pullets were hand fed twice per day at approximately 0800 and 1500 h. Average body weight was determined by groupweighing all birds in a cage on Day 1 and at Weeks 2, 6, and 12. Individual body weights were measured at 18 wk of age. Feed consumption was measured at 2, 6, 12, 16, and 18 wk from three replicate cages sharing a common feed trough within a single treatment. Dead chicks were replaced to maintain cage density treatments through Week 2; thereafter no birds were replaced. Total bird mortality for the entire 18-wk study was recorded on a daily basis as it occurred. Blood Sampling, SRBC Immunization, and Hemagglutinin Assay At 6, 12, and 17 wk of age, 1.5 ml of blood was drawn from the brachial vein of 10 pullets from each density by strain (D S) treatment that had not been previously bled or injected with SRBC, using heparinized 20-gauge needles. Heterophils, lymphocytes, and the H:L ratios were determined using the stained-slide method described by Gross and Siegel (1983). Another 10 pullets from each D S treatment that had never been bled or injected with SRBC were immunized i.v. at 6, 12, and 17 wk in the brachial vein with 0.3 ml of a 9% SRBC suspension in 0.9% saline. At 7 d postimmunization, 3.0 ml of blood was drawn from the brachial vein of each pullet, and allowed to clot at room temperature. Serum was harvested and stored at 80 C until further analysis. Hemagglutinin antibody titers to SRBC were determined using the microtiter procedure described by Wegman and Smithies (1966) expressed as the log base 2 of the reciprocal of the highest dilution showing agglutination. All animal care procedures were carried out as described in the PATTERSON AND SIEGEL protocol approved by The Pennsylvania State University Institutional Animal Care and Use Committee (95R074-0). Statistical Analysis Live performance data were analyzed by ANOVA with a factorial arrangement of the four cage densities and two strains in a completely randomized design. Cage density treatments are the practical, inclusive evaluations of the confounding factors of cage area and feeder and drinker spaces per bird realized by the simple act of placing more pullets per commercial cage. The ANOVA was computed using the GLM procedures of SAS (SAS Institute, 1994). Treatment means were separated using Tukey s multiple comparisons procedure (Steel and Torrie, 1980) at P < 0.05. Pullet blood samples were taken from different cage mates at 6, 12, and 17 wk of age, allowing the data to be analyzed as a split plot over time after arc sine square root transformation (Snedecor and Cochran, 1967). Mortality data were analyzed by two-way ANOVA as described above after arc sine square root percentage transformation. Live Performance RESULTS At 1 d of age, the Delta pullets were significantly heavier than the W-36 pullets at all cage densities (Table 1). There were no significant density effects when the chicks were randomly placed at 1 d of age; however, there was an unexplained D S interaction. At 2 wk of age, strains were significantly different, with the W-36 chicks weighing more than the Deltas. Placing 20 birds per cage negatively impacted the 2-wk body weight of Delta chicks compared to birds at either 32 or 38 per cage. The significant D S interaction suggests that strains may respond differently to varying bird density. Starting at 6 wk, and continuing to the conclusion of the experiment, the Delta pullets were significantly heavier at all cage densities than the W-36 pullets. At 6 wk, Delta pullets at 38 and 32 birds per cage were lighter than birds housed at 20 per cage. By 12 wk, both strains had reached approximately 80% of their 18-wk body weight. Even after 6 wk, when cage area per bird was increased, the Delta pullets housed at 19 and 16 birds per cage weighed less than those at 13 or 10 per cage. Conversely, BW of the W- 36 was not significantly affected by cage density until 18 wk of age, when birds housed at 19 per cage weighed 4.5% less than those at 10 per cage. The significant D S interaction continued for all weighing periods. As both cage density and strain significantly impacted mean BW at 18 wk, other statistics of variation about the means were evaluated (Table 2). The maximum and minimum BW within each cage density increased as the mean values increased. Body weight SD values for Delta pullets were higher, ranging from 100.7 to 118.7 vs 93.6 to 97.6 for the W-36 pullets, and closely corresponded with

IMPACT OF CAGE DENSITY ON PULLET PERFORMANCE AND STRESS 35 TABLE 1. Body weight of commercial pullet strains at different ages and cage densities 1 Body weight average Density 2 birds per cage Day 1 2 wk 6 wk 12 wk 18 wk Delta W-36 Delta W-36 Delta W-36 Delta W-36 Delta W-36 38 to 19 37.9 a 35.6 bc 114 b 119 a 398 c 372 d 993 b 927 c 1,210 bc 1,161 d 32 to 16 37.8 a 36.1 b 111 b 120 a 407 bc 377 d 998 b 947 c 1,242 b 1,182 cd 26 to 13 38.0 a 35.2 c 110 bc 119 a 425 ab 373 d 1,034 a 942 c 1,329 a 1,186 cd 20 to 10 37.8 a 36.1 b 106 c 119 a 434 a 379 d 1,041 a 957 c 1,357 a 1,215 bc ANOVA (g) Probability Density (D) NS * *** *** *** Strain (S) *** *** *** *** *** D S * * ** * *** SEM 0.20 1.1 4.1 7.0 10.1 a dmeans within each age with no common superscript differ significantly. 1Average weights at 1 d, 2 wk, and 6 wk were determined from 6 cages for each density and bird strain, and from 12 cages at 12 and 18 wk. 2Cage density was reduced at 6 wk of age from 38, 32, 26, and 20 birds per cage to 19, 16, 13, and 10 birds per cage by randomly dividing birds between 2 cages in a 2-cage deck system. *P < 0.05. **P < 0.01. ***P < 0.001. the differences in body weight. However, coefficients of variation were low and similar (7 to 9%), indicating that neither cage density nor pullet strain had any meaningful affect on the SD. Furthermore, flock uniformity often utilized in commercial practice to evaluate the percentage of birds ± 10% from the mean body weight did not indicate any effect of cage density or strain, with values ranging from 74.7 to 82.2%. The uniformity goal recommended by most pullet management guides is to produce 80% or more of pullets within 10% of the mean (DeKalb, 1994; and Hy-Line, 1996). In this experiment, the goal was realized only by Delta pullets at the highest density treatment (19 birds per cage), although the uniformity results of the other treatments were within 5% of the 80% goal. As bird density remained constant until 6 wk of age, BW, gain, feed intake, and FC for the periods Day 1 to 2 wk and 2 wk to 6 wk are addressed together in Table 3. Weight gain was not affected by cage density treatments during the first 2 wk; however, the strains responded differently when the W-36 chicks gained significantly more than the Deltas. Density and strain significantly affected weight gain from 2 to 6 wk of age and there was a significant D S interaction. Cage density did not affect the gain of the W-36 pullets, whereas the higher bird densities reduced gain by as much as 15% among the Delta pullets. Feed intake for the two strains from Day 1 to 2 wk was significantly affected by cage density, with birds housed at 20 per cage eating more feed than birds at 26, 32, or 38 per cage. From 2 to 6 wk the main effects of density and strain were significant, and no D S interaction was observed. During this period, the W-36 pullets consumed 8% more feed than the Deltas (P < 0.001), and when provided with more cage and feeder space, birds at 20 per cage ate significantly more than birds at 26, 32, or 38 per cage (P < 0.001). Feed conversion from Day 1 to 2 wk was negatively affected by the more generous cage densities, with pullets housed at 20 birds per cage consuming significantly more feed with no greater gain than birds at 26, 32, and 38 per cage (P < 0.001, Table 3). In addition, strain FC means indicated the W-36 pullets were more efficient than the Delta pullets. No significant D S interaction was observed for the Day 1 to 2 wk period; however, from 2 to TABLE 2. Body weight statistics and uniformity of pullet strains at 18 wk of age and different cage densities Delta (birds per cage) W-36 (birds per cage) Statistics 10 13 16 19 10 13 16 19 Mean, g 1,357 1,329 1,242 1,210 1,215 1,186 1,182 1,161 Maximum, g 1,660 1,570 1,550 1,570 1,450 1,420 1,430 1,460 Minimum, g 1,050 1,055 660 970 1,040 910 860 930 SD 116.6 103.5 118.7 100.7 93.6 97.6 95.8 98.2 CV, % 8.60 7.78 9.53 8.32 7.71 8.23 8.07 8.47 Uniformity, 1 % 75.0 77.6 77.4 82.2 74.8 77.3 79.3 74.7 1Uniformity = percentage of the treatment population ± 10% from the mean body weight.

36 PATTERSON AND SIEGEL TABLE 3. Weight gain, feed intake, and feed conversion (feed intake:weight gain) of commercial pullet strains at different ages and cage densities 1 Body weight gain Feed intake Feed conversion Density birds per cage Day 1 to 2 wk 2 to 6 wk Day 1 to 2 wk 2 to 6 wk Day 1 to 2 wk 2 to 6 wk Delta W-36 Delta W-36 Delta W-36 Delta W-36 Delta W-36 Delta W-36 (g) (g/bird) (g:g) 38 75.6 83.1 279 cd 252 d 147 b 147 b 745 (779 z ) 813 1.95 (1.86 y ) 1.77 2.67 c 3.23 b 32 73.0 84.2 296 bc 256 d 153 b 155 b 784 (826 y ) 869 2.09 (1.97 xy ) 1.84 2.64 c 3.39 ab 26 72.1 84.0 315 ab 254 d 166 b 159 b 849 (889 x ) 928 2.30 (2.09 x ) 1.89 2.70 c 3.66 a 20 68.6 83.3 327 a 259 d 191 a 221 a 884 (922 w ) 959 2.79 (2.72 w ) 2.65 2.71 c 3.70 a (72.3 z ) (83.6 y ) (816 z ) (892 y ) (2.28 y ) (2.04 z ) ANOVA Probability Density (D) NS ** *** *** *** * Strain (S) *** *** 0.0824 *** *** *** D S NS * * NS NS * SEM 1.45 5.8 4.5 10.1 0.056 0.064 a dmeans for the density by strain interaction within each age with no common superscript differ significantly. w zmeans for either cage density or strain main effects in parentheses with no common superscript differ significantly. 1n = 2. *P < 0.05. **P < 0.01. ***P < 0.001. 6 wk the significant interaction indicated higher FC for the W-36 pullets than for the Delta pullets at all densities. Although FC of the Delta pullets between 2 to 6 wk were not significantly affected by cage density, the W-36 housed at 20 and 26 birds per cage had significantly higher FC ratios than birds at 38 per cage (P < 0.001). Body weight gain, feed intake, and FC for the periods after 6 wk when cage density was reduced to 19, 16, 13, TABLE 4. Weight gain, feed intake, and feed conversion (feed intake:weight gain) conversion of commercial pullet strains at different ages and cage densities 1 Density birds per cage and 10 birds per cage are reported in Table 4. During the period from 6 to 12 wk, weight gain was significantly affected by pullet strain and cage density, whereas the D S interaction was not significant. Delta pullets gained an average of 5% more for the 6-wk period than the W-36 pullets (P < 0.001), and pullets housed at 19 per cage gained 4% less than those at 10 per cage (P < 0.05). Weight gain during the final 6 wk of the rearing period was less Body weight gain Feed intake Feed conversion 6 to 12 wk 12 to 18 wk 6 to 12 wk 12 to 18 wk 6 to 12 wk 12 to 18 wk Delta W-36 Delta W-36 Delta W-36 Delta W-36 Delta W-36 Delta W-36 (g) (g/bird) (g:g) 19 596 (574 z ) 553 217 c 234 bc 2,547 bc 2,255 d 2,300 cd 2,114 d 4.28 cd 4.08 d 10.66 a 9.06 b 16 591 (582 z ) 573 245 b 233 bc 2,688 b 2,428 cd 2,550 bc 2,357 cd 4.55 bc 4.24 cd 10.46 ab 10.17 ab 13 609 (593 yz ) 577 296 a 244 b 2,909 a 2,458 bc 2,749 b 2,603 b 4.78 b 4.26 cd 9.30 ab 10.66 a 10 607 (596 y ) 585 317 a 258 b 3,304 a 2,633 b 3,161 a 2,436 bc 5.44 a 4.50 bc 9.99 ab 9.44 ab (601 y ) (572 z ) ANOVA Probability Density (D) * *** *** *** *** NS Strain (S) *** *** *** *** *** NS D S NS *** *** *** *** ** SEM 7.1 5.8 38.9 58.8 0.083 0.344 a dmeans for the density by strain interaction within each age with no common superscript differ significantly. y zmeans for either cage density or strain main effects in parentheses with no common superscript differ significantly. 1n = 4. *P < 0.05. **P < 0.01. ***P < 0.001.

IMPACT OF CAGE DENSITY ON PULLET PERFORMANCE AND STRESS 37 than half of that realized from 6 to 12 wk as the pullets approached adolescent weight (80% of their 18 wk weight) and received diets lower in protein and energy density than the grower diet fed from 6 to 12 wk. At 10 and 13 birds per cage, the Delta pullets gained significantly more weight between 12 to 18 wk than the W-36 at the same density, or like strain pullets housed at 16 or 19 per cage. Weight gain for the W-36 pullets was not significantly affected by cage density during the 12 to 18 wk period, whereas Delta pullets at 19 per cage gained 31% less than birds housed at 10 per cage. No strain differences were observed at the higher cage densities of 16 and 19 birds. Feed intake during the 6 to 12 wk period (Table 4) was consistently greater for the Delta pullets at all density treatments than for the W-36 (P < 0.001). Cage densities of greater than 13 birds negatively affected feed intake in a linear fashion in both strains (P < 0.001). From 12 to 18 wk, a highly significant interaction dominated the strain and density main effects with regard to feed intake. Delta pullets housed at 10 per cage ate significantly more feed than any group. Intake was significantly reduced at the higher cage densities for both pullets, yet no strain differences were realized at 13, 16, or 19 birds per cage. Poorer feed conversion was observed during the 6 to 12 wk period for the Delta pullets and for birds housed at densities with fewer birds per cage (Table 4). Consistent results within each strain indicated better feed utilization with increasing bird density. However, from 12 to 18 wk neither cage density nor strain showed any significant trends, and the D S interaction, although significant, appeared to be an anomaly. Unlike the 12 to 18 wk weight gain and feed intake results, no logical pattern was observed for FC. Pullet mortality during the 18 wk rearing period was not influenced by cage treatments; however, when strains were compared, the Delta had a lower mortality rate (6.32%) than the W-36 strain (12.04%, P < 0.01). No disease or pathology differences were observed between the strains, although the majority of W-36 deaths were recorded within the first 2 wk of housing. Blood Parameters Blood measurements, like those for live performance, were similarly analyzed for the main effects of cage density, strain, and the D S interaction; however, because independent samples were drawn at 6, 12, and 17 wk from different cage mates (i.e., no bird was sampled more than once), pullet age or time could be considered in the analysis. Percentage heterophils were not affected by pullet density or strain, although a significant reduction in their level was observed over time (P < 0.001). Furthermore, the interaction between strain and time was also significant (Figure 1). The heterophil percentage had a downward trend in both stains with increasing age. The Delta pullets declined at almost twice the rate of the W-36 pullets, which led to the significant interaction. At each pullet age FIGURE 1. Heterophil percentage of commercial pullet strains at 6, 12, and 17 wk of age. Means with no common letters differ significantly (P < 0.05). significant strain differences were observed, with the Delta pullets having a higher percentage than the W-36 initially, then significantly less at 12 and 17 wk of age. Percentage lymphocyte results were similar to those for heterophils, with time and the time by strain (T S) interaction representing the only significant effects. Means at each age show the percentages were lower at 6 and 12 wk for both strains, then increase at 17 wk of age (Figure 2). Significant strain differences were observed at each sampling age, and the Delta percentage was greater at 12 and 17 wk (P < 0.01). Once more, time (P < 0.001) and the T S interaction (P < 0.05) were the only factors that significantly affected the H:L ratios (Figure 3). Each strain realized a significant fall of their respective H:L ratio with age as the result of rising percentages of lymphocytes and falling heterophils. The interaction was significant because the ratio for the Delta fell more rapidly from 31.2 to 17.4% during the 6 to 17 wk period compared to 26.4 to 20.2% for the W-36 strain. Antibody titer responses to SRBC antigen were also affected by pullet age (P < 0.001), but there were no effects of cage density or strain. Log 2 titers were 7.29, 9.44, and 8.78 at 6, 12, and 17 wk, respectively. FIGURE 2. Lymphocyte percentage of commercial pullet strains at 6, 12, and 17 wk of age. Means with no common letters differ significantly (P < 0.01).

38 FIGURE 3. Heterophil:lymphocyte (H:L) ratio percentage of commercial pullet strains at 6, 12, and 17 wk of age. Means with no common letters differ significantly (P < 0.05). Live Performance DISCUSSION Significant strain differences were evident for BW, gain, feed intake, and FC from early in the study until its conclusion at 18 wk of age. The body weights of the genetically larger Delta pullets were negatively impacted at 2 wk by cage densities of 32 to 16 birds per cage (116 to 232 cm2 per bird) or higher at 6, 12, and 18 wk, whereas the W-36 pullets were only negatively affected at 18 wk by the 38 to 19 bird per cage treatment (98 to 196 cm2 bird). However, cage density did affect feed intake and FC of the W-36 pullet early in the rearing period. Significant reductions in BW and feed intake and poorer FC induced by cage density treatments were most likely combined affects of cage and feeder spaces and the number of birds per nipple drinker. Water restriction alone could reduce feed intake and growth. These results confirm the findings of Leeson and Summers (1984), who reported that more liberal cage densities (586 cm2 bird) combined with feeder and drinker spaces resulted in a significant increase (5 to 8%) in feed intake compared to a density of 293 cm2 bird. However, under their conditions, the increased feed intake was not associated with increased body weight, which is similar to the results reported herein for W-36 pullets during the period from 6 to 12 wk; whereas in this study, the greater feed intake observed for Delta pullets at low cage density and more liberal feeder and drinker spaces coincided with heavier BW at 6, 12, and 18 wk of age. The low density treatment in this study (372 cm2 per bird) was considerably less space than that provided pullets by Leeson and Summers (1984) (586 cm2 per bird) and may not have allowed for greater bird activity to offset increased energy intake, as they had observed. Most likely, greater feed trough space at 20 to 10 and 26 to 13 birds per cage (3.0 to 6.1 cm and 2.4 to 4.7 cm per bird feed trough, respectively) and fewer birds per nipple (10 to 5 and 13 to 6.5) allowed for greater feed intake and weight gain in these treatments. The body weight recommended for Delta pullets at 18 wk (1,340 g) lies between the 10 and 13 birds per cage PATTERSON AND SIEGEL density results reported in Table 1 (DeKalb, 1994). Apparently average weight for age performance was achieved at these densities representing the U.S. standard or more space per bird; however, at 16 and 19 birds per cage, ideal weight was not reached in this study. The W-36 pullets at 18 wk (Table 1) weighed less than the 1,280 g target recommended by Hy-Line (1996). These results suggest that either the W-36 pullet has a greater cage, feeder, or drinker space requirement, or the nutrient density of the diets, or appetite of the pullets, did not allow them to reach recommended weight. Despite more birds per cage and less feeder and drinker access in this study, body weight uniformity at 18 wk for the higher bird densities were not adversely affected (Table 2). Indeed, all treatments lie within 5% of the 80% uniformity goal. Blood Parameters Despite a significant reduction in feed intakes (27 and 13%) and corresponding losses in body weight (11 and 4.5%) during the 12- to 18-wk period, neither Delta or W- 36 pullets reared at 19 vs 10 birds per cage exhibited significant differences in percentage heterophils, lymphocytes, or the H:L ratio (P > 0.05). Apparently the impact of stress on the H:L ratio can be transient, as the duration of the H:L ratio response reported by Zulkifli et al. (1993) lasted only 12 d when birds were restricted to 60% of ad libitum at 30 d of age. Heterophil:lymphocyte ratios declined from 0.72 to 0.57 at 45 and 55 d, respectively, when feed-restricted birds were released from 60 to 80% restriction at 43 d. Similarly, Ubosi et al. (1985) observed an almost twofold increase in the H:L ratio of Leghorn chicks, measured within the acute period of 8 d after feeding 5,700 ppb aflatoxin. Although no blood indices of stress were associated with cage density in this study, Delta pullets housed at 38, 32, 26, and 20 birds per cage to 6 wk had H:L ratios almost twofold higher than those measured at 17 wk when they were housed at 19, 16, 13, and 10 birds per cage (Figure 3). Although the H:L ratio results are less dramatic for the W- 36 pullets, a significant 6.2% difference coinciding with the Day 1 to 6 wk rearing densities does not rule out a cage density affect including cage, feeder, and drinker spaces per bird. The significant T S interactions reported in Figures 1, 2, and 3 indicate that either pullet age or the stepping down of cage density over time impacted the percentage heterophils, lymphocytes, and the H:L ratio. Because cage density effects within each sampling time did not impact any of the blood parameters, we believe pullet age was the more likely factor influencing heterophils, lymphocytes, and the H:L ratio results in Figures 1, 2, and 3. Hester et al. (1996a) also reported no significant effect of cage density on the percentage heterophils, lymphocytes, or the H:L ratio when pullets at 17 wk were placed in single- or multiple-bird cages and bled at 18 wk of age. Like the strain differences reported herein, these authors observed a significantly lower H:L ratio for a

IMPACT OF CAGE DENSITY ON PULLET PERFORMANCE AND STRESS 39 White Leghorn line selected for improved livability and productivity in multiple-bird cages than in either an unselected control or a commercial line. The high density cage treatments reported herein for the Delta pullets from 2 to 6 wk reduced growth rate and feed intake, yet no significant effect of cage density was realized when challenged with SRBC antigen at 6 wk (P > 0.05). Again at 12 to 18 wk, weight gain and feed intake were significantly decreased (31 and 27%, Delta; and 9 and 13%, W-36, respectively) when the 10 and 19 bird densities were compared (P < 0.001) yet no effect of housing density on SRBC antibody titers were observed at 17 wk. Perhaps the chronic duration of the housing treatments from Day 1 to 6 wk at 38, 32, 26, 20 birds per cage and 6 to 18 wk at 19, 16, 13, and 10 birds per cage, respectively, masked any response to SRBC antigen challenge at 6, 12, and 17 wk. Possibly the more acute stressors of hot or cold environmental temperatures, and feed restriction with feed deprivation before antigen injection reported by Henken et al. (1983a,b) are necessary to elicit a change in the SRBC response. According to Hester et al. (1996c) cage density treatments alone had no effect on SRBC antibody levels in hens at 33 wk of age; however, hens housed singularly experienced immunosuppression when exposed to 0 C for 72 h relative to hens housed 12 per cage. At 44 wk of age, the hens housed 12 per cage had significantly lower SRBC antibody titers than single hens per cage following high environmental temperatures of 38 C for 3 h, suggesting that the multiple stressors of cage density, including cage, feeder, and drinker spaces, and environmental temperature will impact hens differently. Pullet age was the only significant factor in this experiment influencing antibody response to SRBC. McCorkle and Glick (1980) reported similar findings in a closed flock of New Hampshire chickens housed in batteries until 6 wk of age that were then moved to floor pens. Titers to SRBC in their study at 4, 12, 25, 52, and 104 wk were 6.83C, 8.59AB, 9.57A, 8.85ABC, and 7.13BC, respectively (P < 0.01). The pullet titers reported herein showed a similar pattern increasing with maturity from 6 to 12 wk (7.29C to 9.44A) then declining at 17 wk to 8.78B (P < 0.001). In summary, increasing bird density by three and six pullets per cage vs reducing bird number by three from an industry standard providing 143 (Day 1 to 6 wk) and 286 (6 to 18 wk) cm2 per bird had measurable effects on pullet live performance. Although the impact of cage, feeder, and drinker spaces can not be distinguished, body weight was significantly reduced at the greater bird densities, whereas reducing by three the number of birds per cage had no benefit. Feed intake and the FC ratio were both reduced when more birds were housed per cage. When three fewer birds were placed per cage, intake and FC were increased compared to the industry standard cage density. Mortality, body weight, uniformity, and blood indices of stress were not impacted by the cage density treatments. However, there were significant strain differences observed in both the measures of live performance and immune response. ACKNOWLEDGMENTS The authors acknowledge and appreciate the donation of chicks by DeKalb Poultry Research, Inc. and Hy- Line International, and donations of feed by Wenger Feeds. Thanks are extended to Eric Lorenz, Bonnie Ford, and Nuket Acar for their technical and statistical assistance with the study. This research was also partially supported by the Purina Mills, Inc., Research Fellowship Program. REFERENCES Anderson, K. E., and A. W. Adams, 1992. Effect of rearing density and feeder and waterer spaces on the productivity and fearful behavior of layers. Poultry Sci. 71:53 58. Anderson, K. E., A. W. Adams, and J. V. Craig, 1989. Behavioral adaptation of floor-reared White Leghorn pullets to different cage densities and cage shapes during the initial settling-in period. Poultry Sci. 68:70 78. Bell D., 1969. Crowding in cage rearing affects pullet weights. Pacific Poultryman, January, 18, 19, 28. Carey, J. B., 1987. Effects of pullet-stocking density on performance of laying hens. Poultry Sci. 66:1283 1287. DeKalb, 1994. Growing program. Pages 4 10 in: DeKalb Delta Pullet and Layer Management Guide. 3rd ed. DeKalb Poultry Research, Inc., DeKalb, IL. Donker, R. A., M.G.B. Nieuwland, and A. J. van der Zijpp, 1990. Heat-stress influences on antibody production in chicken lines selected for high and low immune responsiveness. Poultry Sci. 69:599 607. Gross, W. B., and H. S. Siegel, 1983. Evaluation of the heterophil/lymphocyte ratio as a measure of stress in chickens. Avian Dis. 27:972 979. Henken, A. M., A.M.J. Groote Schaarsberg, and M.G.B. Nieuwland, 1983a. The effect of environmental temperature on immune response and metabolism of the young chicken. 3. Effect of environmental temperature on the humoral immune response following injection of sheep red blood cells. Poultry Sci. 62:51 58. Henken, A. M., A.M.J. Groote Schaarsberg, and W. van der Hel, 1983b. The effect of environmental temperature on immune response and metabolism of the young chicken. 4. Effect of environmental temperature on some aspects of energy and protein metabolism. Poultry Sci. 62:59 67. Hester, P. Y., and E. K. Wilson, 1986. Performance of White Leghorn hens in response to cage density and the introduction of cage mates. Poultry Sci. 65:2029 2033. Hester, P. Y., W. M. Muir, J. V. Craig, and J. L. Albright, 1996a. Group selection for adaptation to multiple-hen cages: Hematology and adrenal function. Poultry Sci. 75: 1295 1307. Hester, P. Y., W. M. Muir, J. V. Craig, and J. L. Albright, 1996b. Group selection for adaptation to multiple-hen cages: Production traits during heat and cold exposures. Poultry Sci. 75:1308 1314. Hester, P. Y., W. M. Muir, and J. V. Craig, 1996c. Group selection for adaptation to multiple-hen cages: Humoral immune response. Poultry Sci. 75:1315 1320. Hy-Line, 1996. Monitoring body weight. Page 10 in: Hy-Line Variety W-36 Management Guide. Hy-Line International, West Des Moines, IA.

40 PATTERSON AND SIEGEL Leeson, S., and J. D. Summers, 1984. Effects of cage density and diet energy concentration on the performance of growing Leghorn pullets subjected to early induced maturity. Poultry Sci. 63:875 882. Mashaly, M. M., M. L. Webb, S. L. Youtz, W. B. Roush, and H. B. Graves, 1984. Changes in serum corticosterone concentration of laying hens as a response to increased population density. Poultry Sci. 63:2271 2274. McCorkle, F., and B. Glick, 1980. The effect of aging on immune competence in the chicken: Antibody-mediated immunity. Poultry Sci. 59:669 672. McFarlane, J. M., and S. E. Curtis, 1989. Multiple concurrent stressors in chicks. 3. Effects on plasma corticosterone and the heterophil:lymphocyte ratio. Poultry Sci. 68:522 527. North, M. O., and D. D. Bell, 1990. Cage management. Pages 297 303 in: Commercial Chicken Production Manual. 4th ed. Van Nostrand Reinhold, New York, NY. Ubosi, C. O., W. B. Gross, P. B. Hamilton, M. Ehrich, and P. B. Siegel, 1985. Aflatoxin effects in White Leghorn chickens selected for response to sheep erythrocyte antigen. 2. Serological and organ characteristics. Poultry Sci. 64: 1071 1076. SAS Institute, 1994. SAS/STAT User s Guide: Statistics. Version 6.08. SAS Institute, Inc., Cary, NC. Siegel, H. S., 1959. Egg production characteristics and adrenal function in White Leghorns confined at different floor space levels. Poultry Sci. 38:893 898. Snedecor, G. W., and W. G. Cochran, 1967. Statistical Methods. 6th ed. The Iowa State University Press, Ames, IA. Steel, R.G.D., and J. H. Torrie, 1980. Principles and Procedures of Statistics. McGraw-Hill Book Co., Inc., New York, NY. Thaxton, P., and H. S. Siegel, 1970. Immunodepression in young chickens by high environmental temperature. Poultry Sci. 49:202 205. Trout, J. M., M. M. Mashaly, and H. S. Siegel, 1988. Research Note: Changes in heterophils following antigen injection in immature male chickens. Poultry Sci. 67:1775 1777. Wegman, T. G., and O. Smithies, 1966. A simple hemagglutination system requiring small amounts of red cells and antibodies. Transfusion 6:67 73. Zulkifli, I., E. A. Dunnington, W. B. Gross, A. S. Larsen, A. Martin, and P. B. Siegel, 1993. Responses of dwarf and normal chickens to feed restriction, Eimeria tenella infection, and sheep red blood cell antigen. Poultry Sci. 72: 1630 1640.