Egg Quality in Furnished Cages for Laying Hens Effects of Crack Reduction Measures and Hybrid

Egg Quality in Furnished Cages for Laying Hens Effects of Crack Reduction Measures and Hybrid H. Wall 1 and R. Tauson Department of Animal Nutrition and Management, Swedish University of Agricultural Sciences, S-755 97 Uppsala, Sweden ABSTRACT Egg quality, egg production, and hen use curtains lowered the proportions of cracked eggs significantly of facilities, with respect mainly to proportion of eggs laid by softly catching and reducing the speed of the in nests and bird locations after lights-out, were studied in eggs before entering the egg cradle. The position of the furnished-cage models for six or eight birds. In these nest opening affected none of the measured parameters. cages, most of the eggs are laid in the 25 cm wide nest, Less lining in the bottom of the nest, i.e., a reduction to i.e., egg laying is concentrated to a much smaller area 30% coverage, resulted in significantly higher proportions compared with conventional cages. The study (20 to 80 of dirty eggs and lower proportions of eggs laid in nests. wk of age) used 1,296 hens of three genotypes Hy-Line Hybrid differences were found in most of the measured White, Hy-Line Brown, and Lohmann Selected Leghorn. Conventional four-hen cages were included as a reference. We separately studied the effects on the proportion nest curtains are effective measures in reducing cracks in traits. We concluded that devices like egg savers and long of eggs cracked by egg saver wires in front of the egg furnished cages, where egg laying is concentrated to a cradle, long nest curtains, and position of nest opening much smaller area than in conventional cages. Reduction (rear or front) with 30 or 100% of the nest bottom being lined with AstroTurf. Egg saver wires and long nest of the bottom lining to 30% makes nests less attractive to birds laying eggs. (Key words: egg quality, nest design, hybrid, furnished cage) 2002 Poultry Science 81:340 348 INTRODUCTION Public concern about the welfare of laying hens has resulted in minimum welfare directives in the European Union, imposing a ban on conventional cages in 2012 (European Commission, 1999). Thereafter, cages are allowed only if furnished with nests, perches, and litter baths, i.e., facilities that improve the behavioral repertoire of the birds. In Sweden, conventional cages were banned in 1999 (Statens jordbruksverk, 1997). Because no fully developed, furnished cages that had been subjected to compulsory testing were available at that time, exemptions were permitted after individual inspections of farms. Thereby the phase-out was delayed for another few years. In October 2000, the first model of a furnished cage (Bröderna Victorsson, AB) was approved by the Swedish Board of Agriculture (Jordbruksverket, 2000). When designing furnished cages, as well as other alternative housing systems, emphasis must be put on hen preferences and on practical considerations, such as pro- duction, hygiene, and management aspects. For example, it has been shown that an appropriate material and enclosure are important nest attractants and that hens prefer loose moldable materials when given a choice between different nesting materials (Appleby and McRae, 1986; Duncan and Kite, 1989; Appleby, 1990). Enclosed nests with artificial turf, however, are largely accepted when no other nesting materials are present (Abrahamsson et al., 1996; Abrahamsson and Tauson, 1997; Appleby, 1998) and prelaying behavior has been reported to be normal in nests with artificial turf (Appleby, 1998). From a hygienic point of view, the optimal nesting material would probably be one that allows manure to pass through; however, several studies have shown that nests with wire floors or plastic netting are less attractive than nests with lining (Hughes, 1993; van Niekerk and Reuvekamp, 1995; Abrahamsson et al., 1996). A study by Reed and Nicol (1992) showed that a strip of artificial grass, mounted on the rear wall of rollaway nests, encouraged nesting behavior, measured as time spent in the nest. Their findings indicate that covering 2002 Poultry Science Association, Inc. Received for publication May 14, 2001. Accepted for publication October 18, 2001. 1 To whom correspondence should be addressed: Helena.Wall@ huv.slu.se. Abbreviation Key: CO = conventional, four-hen metal cage; HYB = Hy-Line Brown; HYW = Hy-Line White (W-36); LSL = Lohmann Selected Leghorn. 340

EGG QUALITY IN FURNISHED CAGES FOR LAYERS 341 only one part of the wire floor bottom of the nest with, e.g., artificial turf may be sufficient to encourage nesting behavior in laying hens. Wall et al. (2002) found that nest floors partly covered (30 vs. 50%) with artificial turf resulted in lower proportions of eggs laid in the nests than in fully lined nests. However, the nests in that study had two openings, one leading in and one out of the nest, both equipped with one-way doors that made inspection by a bird more difficult. Furthermore, when the artificial turf covered only one part of the nest bottom, the hen could not see it from the entrance. Less artificial turf in the nests may improve nest hygiene as well as make cleaning between batches less time consuming. Higher proportions of cracked eggs in furnished than in conventional cages have been reported in several studies (Abrahamsson et al., 1995; Abrahamsson and Tauson, 1997; Wall et al., 2002). Accumulation of eggs in an area of about 25 cm width probably contributed to this difference. As most eggs are laid in the nest, this reduced area may increase risk of collisions in the egg cradle. Furthermore, the position of birds in the nest box, in the rear or in the front close to the egg cradle, when laying their eggs, may affect the end speed of rolling due to the increased momentum over a longer distance (Lee and Bolton, 1976). The objective of the present study was to evaluate different measures to reduce cracked eggs in three furnished-cage models and to study effects on egg quality and bird use of nests when different proportions of the nest bottom were covered with artificial turf. Conventional cages were used mainly for basic genotype studies and not for statistical comparison with furnished-cage use. Housing MATERIALS AND METHODS The study was carried out from September 1999 until November 2000; birds used were 20 wk old at the beginning of this 60-wk study. Three furnished-cage models and one conventional battery cage, the latter included as a reference, were installed in three vertical-tier batteries in the same experimental building. The furnished cages used were based on the Edinburgh Modified Cage (Appleby and Hughes, 1995; Abrahamsson et al., 1996) but with other group sizes and fully automated. The nest box was positioned at one end of the cage (at a right angle to the feed trough), and the litter bath was placed on top of the nest. Dimensions of cages and equipment are given in Table 1, and an overview of the experimental layout in Table 2. In the Hellmann furnished-cage model (Hellmann Poultry GmbH), the nesting area was separated from the cage area by a curtain made of plastic strips, through which the birds entered the nest (Figure 1). All nests were lined with brown artificial turf (AstroTurf ). The litter bath lacked a closing mechanism and was thereby available around the clock. Half of the cages (every second row) had an egg saver, i.e., a wire extending parallel to and underneath the feed trough, 55 cm from the rear partition of the cage. Hence, on the way out of cage, eggs were stopped when rolling toward the egg saver wire. This wire lifted every 15 min until 8 h after lights-on and thereafter every hour until lights-out, allowing eggs to slowly roll the last short distance to the egg cradle. In the Victorsson furnished-cage model (Bröderna Victorsson AB), a metal sheet partition separated the nest and litter bath from the cage area (Figure 2). The nest, lined with brown artificial turf, was entered through an opening at the front of the partition, at the feed trough. In the front, plastic black curtains, hanging behind the gates of the cages, enclosed the nests. In half of the cages this curtain did not reach below the manure deflector, i.e., short nest curtain, whereas the curtains in the other cages ended 1 cm above the floor, i.e., long nest curtain. A time-controlled closing mechanism of the litter bath (a door sliding sideways) enabled birds to enter the bath between certain hours. After the door closed, birds still inside could leave by pushing the door open. A perch was placed parallel to the feed trough, extending from the nest side wall to the cage partition on opposite side. A forerunner of the Victorsson cage described above, FC-6, was an experimental furnished-cage model. The closing mechanism of the litter bath and the perch location were the same as in the Victorsson furnished-cage model. The cage was designed as described by Wall et al. (2002) except for the nest openings. In the present study, the nest had one opening, positioned in the front or rear of the nest side facing the cage. The nests were lined with brown artificial turf covering 30 or 100% of the nest bottom. In nests with 30% nest lining, the turf was positioned in the upper-rear part of the nest, with the welded wire floor left uncovered in the front. CO was a conventional 4-hen metal cage. All systems had horizontal front bars and solid side partitions. The rear partitions in Victorsson and FC-6 cages comprised a metal sheet, leaving an open space under and above the sheet, whereas Hellmann cages had welded wire mesh, and conventional four-hen metal cages (CO) had horizontal metal bars at the rear. In the Hellmann cage, nests in adjacent cages were separated by wire mesh, whereas all nests in Victorsson and FC-6 cages had solid rear partitions. Birds, Rearing, Management Routines, and Feeding The trial used 504 Lohmann Selected Leghorn (LSL), 336 Hy-Line White W36+ (HYW), and 456 Hy-Line Brown (HYB) birds. The pullets were reared in conventional rearing cages and were not beak-trimmed (prohibited in Sweden). At 16 wk of age, the birds were transferred to the experimental building, where they received 10 h of light per day. The light was successively increased to 15 h at 24 wk of age and was dimmed for 6 min in the evening before lights-out at 1800 h, to imitate dusk. Light was increased over 6 min in the morning, dawn, at 0300 h.

342 WALL AND TAUSON TABLE 1. Description of housing systems Hellmann Victorsson FC-6 CO 1 Cage, width depth height (cm) 2 96 50 45 96 50 45 72 50 45 48 50 38 No. of birds per cage 8 8 6 4 Cage area per bird (cm 2 ) 2 600 600 600 600 Nest, width depth height (cm) 3 24 50 28 24 50 27 25 50 27.5... Litter bath, width depth height (cm) 24 50 21 24 50 24 25 50 23.5... Nest/litter area per bird (cm 2 ) 4 150 150 208... Perch (cm/bird) 17 12 12... 1 Conventional, four-hen metal cage. 2 Nest and litter bath excluded, height measured in the rear of the cage. 3 Height measured in nest front. 4 Nest and litter bath were of the same area. Twice each week, manure was removed with belts, and litter baths were filled by hand with sawdust. At 16 wk of age, the litter baths with closing mechanisms opened 5 h after lights-on and were then open for 4 h and 30 min. Thereafter the open period was successively increased to a maximum of 6 h and 30 min at 24 wk of age. The litter bath then opened 8 h after lights-on and closed 30 min before dark. In Hellmann cages, where the baths lacked closing mechanisms, the baths were not available to the birds (litter box turned upside down) until 24 wk of age. From then on, they were accessible around the clock. The furnished cages had automatic chain feeders, whereas hens in the CO cages were manually fed once a day. The pullets were fed a conventional grower crumbled diet during rearing. From 17 wk of age until slaughter all hens received a normal layer crumbled diet with a calculated content of 2,700 kcal/kg (11.3 MJ) metabolizable energy, 160 g crude protein, 35 g Ca, and 6 g P/kg. All hens in the Hellmann cages and half of the hens in the FC-6 cages (one of two batteries) were subjected to an unintended water deficiency during the first week after their arrival in the experimental building. Recording and Statistical Analysis of Data Production and mortality were recorded daily per replicate from 20 to 80 wk of age. Feed consumption was recorded per replicate in CO cages and as an average for each of the two batteries with Hellmann and Victorsson cages, respectively. Hens that died during the experiment were subjected to autopsy and were not replaced. Egg cradles in Victorsson, FC-6, and CO cages lacked eggcollection belts, and eggs, therefore, were always collected manually. In Hellmann cages, the cradles were equipped with belts, and eggs were collected at one end of the battery, except when exterior egg quality was measured and eggs were collected manually. The weight of eggs was recorded once weekly. The position of all eggs in the furnished cages was recorded once every fourth week from 24 wk of age, before egg collection. The location of all birds in Hellmann and Victorsson cages was recorded 1 h after lights-out at 31, 53, and 77 wk of age. On seven occasions (at 25, 31, 41, 49, 58, 66, and 72 wk of age) eggs were collected on 5 consecutive d from Hellmann, Victorsson, and CO cages and on 14 consecutive d from FC-6 cages. These eggs were analyzed for proportions of cracked and dirty eggs in a small version of a commercial egg-candling machine. Blood spots on eggshells, caused by red mites, were not included in proportions of dirty eggs. At 62 wk of age, nine eggs per replicate in Hellmann and Victorsson cages and two eggs per replicate in FC-6 cages were collected and analyzed for shape index, shell deformation, and breaking strength (Canadian Egg Shell Tester 2 ). Shell deformation was calculated from the average value of measurements on three different points across the egg equator, after a load of 1,000 g was applied on the egg (The Canadian Egg Shell Tester). Eggs for which the standard deviation of the repeated measurements exceeded 10 10 2 mm were excluded from the analysis. Live weights of the birds were recorded, and hygiene of feet and plumage was scored at 54 wk of age on all birds with one cage per replicate for Hellmann and Victorsson cages and with three cages per replicate for CO cages. The scoring system assigned values of 1 to 4 points (Tauson et al., 1984) for each trait, in which a higher score indicated cleaner conditions. Before statistical analyses, traits given in proportions (mortality, cracked and dirty eggs, egg position, and bird location) were subjected to arcsin transformation (Snede- 2 Otal Precision Company Limited, Ottawa Ontario, Canada K1G 3N3. FIGURE 1. Hellmann furnished cage. One egg is withheld by the egg-saver wire, positioned underneath the feed trough.

EGG QUALITY IN FURNISHED CAGES FOR LAYERS 343 cor and Cochran, 1989). Statistical analyses were carried out with the general linear models procedure of SAS software (SAS Institute Inc., 1996). To analyze pair-wise differences between treatments, Fisher s protected leastsignificant-difference test was used. Statistical analysis was always performed within housing system, and all main effects were considered fixed. For inspection of cracked eggs, dirty eggs, egg position, and bird location, the mean of the different proportions, recorded on each occasion, was calculated for each replicate before analysis. Traits measured once, e.g., live weight or traits accumulated before analysis (production performance) were analyzed with the following models: Hellmann and Victorsson: Y ijk = µ + a i + b j + (ab) ij + e ijk FC-6: Y ijkl CO: Y ij = µ + a i + e ij = µ + a i + b j + c k + (ab) ij + (ac) ik + (bc) jk + (abc) ijk + e ijkl where Y ijk,y ijkl,y ij = response variable; µ = overall mean; a i = effect of hybrid i; b j = effect of crack reduction measure (egg saver, nest curtain, or position of nest opening) j; c k = effect of nest bottom lining k; (ab) ij, (ac) ik, (bc) jk, and (abc) ijk = interaction effects; and e ijk,e ijkl, and e ij = random variation. For traits measured repeatedly, bird age and bird age interactions with all fixed effects were also considered, as well as a random effect of replicate. As no three- or four-factor interactions were found significant, they were excluded from the final statistical model used. Furthermore, as the objective of the study was not to evaluate age differences, two-factor interactions including age were excluded from the final statistical models if found not significant. RESULTS Production performance, mortality, and exterior egg quality are presented in Tables 3 and 4. As no significant differences regarding position of nest opening in FC-6 were found for any parameter, results from these treatments are not presented in tables. Production and Mortality An invasion of red mites (Dermanyssus gallinae), most likely brought into the house by the pullets, caused some losses of birds as a consequence of anemia. The outbreak could not be effectively treated because metriphonate (Neguvon vet., Bayer), the only compound licensed to use against ectoparasites of poultry, was withdrawn from the Swedish market in January 2000. Instead, the population was limited by use of regularly replaced traps made of corrugated cardboard (Nordenfors, 2000) and by spraying the cages as well as floor and walls in the bird room with various natural citrus oils. The only significant difference in mortality was in FC- 6 cages (P 0.05), in which the mortality was higher in cages with 100% AstroTurf lining (Table 4). No deaths were assigned to the unintended water deficiency, but it probably resulted in a general delay in the laying of the first egg, causing the lower laying percentage and egg mass calculated over the whole cycle in Hellmann and FC-6 cages. The average feed conversion ratio (data not shown) was 1.95 kg feed/kg egg in CO cages, in which the birds were fed manually. Corresponding figures in Victorsson and Hellmann cages were 1.97 kg feed/kg egg and 2.00 kg feed/kg egg, respectively. Neither of the cracked-egg reduction measures (egg saver, nest curtain, or position of nest opening) nor the proportion of nest bottom lining had significant effects on laying percentage or egg mass (Tables 3 and 4). In Hellmann cages, a higher egg weight was recorded in cages without egg saver (P 0.05). The only significant interaction found in production performance was a hybrid nest curtain interaction (P 0.05) for kilograms of egg per hen housed in Victorsson cages (data not shown). This interaction was caused because HYW produced significantly fewer kilograms of eggs per hen housed when housed in cages with long nest curtains, compared with all other combinations of nest curtains and hybrids. Hybrid differences in production were found in all furnished-cage models. In FC-6 cages (Table 4), the LSL were superior to HYW in laying percentage, egg weight, and egg mass (P 0.001 to 0.05). In Hellmann cages (Table 3), LSL had higher laying percentages (P 0.05) but lower TABLE 2. Description of experimental layout of housing systems, treatments, hybrids, 1 and replicates Housing Replicates per Cages per system Treatments treatment (n) replicate Hellmann Egg-saver wire 9 (5 HYB, 4 LSL) 3 No egg-saver wire 9 (5 HYB, 4 LSL) 3 Victorsson Long nest curtain 9 (3 HYB, 3 HYW, 3 LSL) 3 Short nest curtain 9 (3 HYB, 3 HYW, 3 LSL) 3 FC-6 Nest opening in front, 30% nest bottom lining 9 (5 HYW, 4 LSL) 1 Nest opening in front, 100% nest bottom lining 9 (5 HYW, 4 LSL) 1 Nest opening in rear, 30% nest bottom lining 9 (5 HYW, 4 LSL) 1 Nest opening in rear, 100 % nest bottom lining 9 (5 HYW, 4 LSL) 1 CO 2 Reference to other models 6 (2 HYB, 2 HYW, 2 LSL) 9 1 HYB = Hy-Line Brown; HYW = Hy-Line White; LSL = Lohmann Selected Leghorn. 2 Conventional, four-hen metal cage.

344 WALL AND TAUSON TABLE 3. Production performance, mortality, and exterior egg quality as influenced by method of crack reduction and hybrid in Hellmann and Victorsson furnished eight hen cages, 20 80 wk of age Hellmann Victorsson Method Method Hybrids 1 P-value 2 Hybrids 1 P-value No Short Long saver Saver HYB LSL Method Hybrids curtain curtain HYB HYW LSL Method Hybrids Laying % (hen day) 82.0 82.2 80.6 83.6 0.85 * 84.9 83.9 83.7 83.5 86.0 0.39 0.21 Egg weight, g 64.3 63.2 64.4 63.1 * ** 61.9 62.0 63.8 a 60.1 c 61.8 b 0.81 *** Egg mass, kg/hen housed 21.7 21.3 21.5 21.6 0.60 0.86 21.5 21.3 22.0 a 20.4 b 21.9 a 0.45 *** Egg mass g/hen day 52.7 52.0 51.9 52.8 0.52 0.44 52.6 52.0 53.4 a 50.2 b 53.2 a 0.48 ** Mortality, 2 % of hens housed 4.2 5.1 4.2 5.2 0.46 0.75 6.9 6.0 6.3 7.6 5.6 0.32 0.58 Cracked eggs, 2 % 11.8 4.7 7.6 9.1 *** 0.32 10.6 2.7 5.8 7.5 6.6 *** 0.20 Dirty eggs, 2 % 5.4 5.3 4.9 6.0 0.86 * 3.8 3.5 2.0 a 4.4 b 4.5 b 0.76 *** Shape index, % 72.9 72.9 73.5 72.3 0.89 ** 73.0 72.9 73.1 73.4 72.4 0.79 0.23 Shell breaking strength, kg 3.35 3.51 3.43 3.42 0.16 0.94 3.46 3.55 3.57 3.47 3.49 0.39 0.70 Shell deformation, 10 2 mm 80.9 78.2 80.4 78.7 0.42 0.60 74.3 76.4 73.7 74.4 78.0 0.32 0.20 Live weight, kg 1.81 1.77 1.94 1.64 0.26 *** 1.74 1.68 1.94 a 1.52 c 1.67 b 0.22 *** Plumage hygiene 3.5 3.3 4.0 2.9 0.15 *** 3.7 3.7 4.0 a 3.6 b 3.5 b 0.98 ** Feet hygiene 3.4 3.5 3.3 3.6 0.94 0.13 3.3 3.3 3.0 b 3.6 a 3.4 a 0.91 ** a c Values within rows and furnished cage model with different superscripts are significantly different (at least P 0.05). 1 HYB = Hy-Line Brown; HYW = Hy-Line White; LSL = Lohmann Selected Leghorn. 2 Presented as mean values instead of least-squares means because of arcsin transformation. *P 0.05; **P 0.01; ***P 0.001. egg weights (P 0.01) compared with HYB. No differences in laying percentage were found among the three hybrids when housed in Victorsson cages (Table 3), but HYB had the highest egg weight followed by LSL and HYW, all significantly different from each other (P 0.001). Furthermore, in Victorsson cages HYB and LSL showed significantly higher egg mass (kg egg/hen housed) compared with HYW (P 0.001). No differences in production were found among hybrids in CO cages (Table 4). Egg Quality The measures for crack reduction with Hellmann and Victorsson cages, i.e., egg savers and nest curtains, respectively, both lowered the proportions of cracks (P 0.001; see Table 3). No significant effect of nest opening position was shown in FC-6 cages (data not shown). The only difference among hybrids for cracked eggs was found in CO cages (Table 4), in which HYB had a higher proportion of cracked eggs than HYW or LSL (P 0.05). The propor- FIGURE 2. Victorsson furnished cage with litter bath upper right (shown as closed) and nest (lower right). Photo: Istvan Pamlényi.

EGG QUALITY IN FURNISHED CAGES FOR LAYERS 345 TABLE 4. Production performance, mortality and exterior egg quality parameters as influenced by percentage of nest bottom lining and hybrid in a furnished six-hen cage (FC-6) and by hybrid in a conventional four-hen cage (CO), from 20 to 80 wk of age FC-6 CO Lining (nest bottom) Hybrids 1 P-value Hybrids 1 P-value 30% 100% HYW LSL Lining Hybrid HYB HYW LSL Hybrid Laying % (hen-day) 82.5 82.4 80.6 84.2 0.95 * 83.4 81.8 84.2 0.30 Egg weight, g 61.8 62.0 60.8 62.9 0.74 *** 65.9 62.6 65.2 0.07 Egg mass, kg/hen housed 21.2 20.5 20.1 21.7 0.25 ** 22.2 20.4 23.0 0.30 Egg mass g/hen-day 51.0 51.0 49.0 53.0 0.99 *** 55.0 51.2 54.9 0.15 Mortality, 2 % of hens housed 3.8 11.7 5.0 10.4 * 0.08 11.1 8.3 0 0.28 Cracked eggs, 2 % 6.2 5.5 6.0 5.7 0.36 0.83 3.6 a 1.6 b 2.4 b * Dirty eggs, 2 % 5.9 3.9 4.6 5.3 * 0.34 5.5 b 8.8 a 8.6 a * Shape index, % 73.5 73.4 73.6 73.3 0.85 0.71............ Shell breaking strength, kg 3.49 3.48 3.40 3.57 0.98 0.42............ Deformation, 10 2 mm 74.9 78.8 79.2 74.5 0.21 0.14............ Live weight, kg.................. 2.06 a 1.56 b 1.70 b ** Plumage hygiene.................. 4.0 3.7 3.7 0.30 Feet hygiene.................. 3.7 3.6 3.9 0.08 a,b Values within rows and cage model with different superscripts are significantly different (at least P 0.05). 1 HYB = Hy-Line Brown; HYW = Hy-Line White; LSL = Lohmann Selected Leghorn. 2 Presented as mean values instead of least-squares means because of arcsin transformation. *P 0.05; **P 0.01; ***P 0.001. tions of cracked eggs increased with bird age in all housing systems (P 0.001; data not shown). Thus, at first and last recording, respectively, (i.e., at 25 and 72 wk of age) the average proportions of cracked eggs were 2.7 and 14.8% in Hellmann, 2.9 and 12.4% in Victorsson, 2.7 and 9.6% in FC-6, and 1.2 and 4.2% in CO cages. The proportion of dirty eggs was higher in cages with 30% nest bottom lining compared with 100% (P 0.05; Table 4). Significantly lower proportions of dirty eggs were observed for HYB, in all cage models in which this hybrid was present (P 0.001 to 0.05), whereas no significant difference in dirty eggs was observed between HYW and LSL. The proportions of dirty eggs were affected by bird age in the three furnished-cage models (P 0.01), but no trend in terms of higher of lower proportions of dirty eggs with increasing bird age was found. The proportion of dirty eggs was unaffected by bird age in CO cages. Shell breaking strength and shell deformation were unaffected by hybrids or cage designs. A higher shape index, indicating a more globular shape, was found with HYB than with LSL (P 0.01) but only in the Hellmann cage; see Table 3. Birds Live Weight and Hygiene Hybrid differences in bird live weight were found in all cage models in which this trait was measured, i.e., in Victorsson, Hellmann, and CO cages (P 0.001 to 0.01; see Tables 3 and 4). Method of crack reduction had no significant effect on live weight or hygiene, regarding plumage and feet, and no significant differences in hygiene were found among hybrids in CO. In Hellmann and Victorsson cages, HYB had a significantly higher plumage hygiene score, indicating cleaner plumage (P 0.001 to 0.01). There was no difference in feet hygiene between HYB and LSL in Hellmann cages, whereas HYB had a higher incidence of dirty feet compared with HYW and LSL in Victorsson cages (P 0.01). An interaction was found between hybrid and nest curtain (P 0.05) for feet hygiene in Victorsson cages. This interaction was caused mainly because HYB had fairly dirty feet (average score = 2.8) in cages with long nest curtains, whereas the opposite was true for HYW birds housed in cages with long nest curtains (average score = 3.9). Use of Facilities When recording bird locations after lights-out in Hellmann and Victorsson cages (data not shown), we noted that some groups of birds, the white genotypes especially, were easily disturbed and changed location during recording. Methods of crack reduction had no significant effect on bird location in either of the two cage models, except for a nest curtain hybrid interaction in Victorsson cages (P 0.01). This interaction was due the proportion of LSL hens spending the night in nests with long curtains being higher than in all other combinations of hybrid and nest curtain. The proportions of birds spending the night in nest boxes were 0.7 and 4.5% in HYB and LSL, respectively, when housed in Hellmann cages (P 0.01) and 1.4, 1.7, and 3.6% in HYB, HYW, and LSL, respectively, when housed in Victorsson cages (P 0.09). On average, 1.6% of the birds (no significant differences among hybrids) spent the night in, or on the edge of, litter baths in Hellmann cages (no closing device), whereas no birds were found in litter baths in Victorsson cages (with closing device). There were no significant differences in use of perches in Hellmann cages in which, on average, 81.2% roosted on the perches. In Victorsson cages, the proportion of birds on the perches was 88.1% for HYW, 79.3% for HYB, and 78.0% for LSL, with HYW being significantly different from HYB and LSL (P 0.05). The proportions

346 WALL AND TAUSON of birds in nests and on perches increased with bird age in Victorsson cages (P 0.05), reaching 3.4% in the nest and 82.5% on the perches at 77 wk of age. In Hellmann cages, bird location at night was unaffected by age. The proportions of eggs laid in nests and litter baths (data not shown) were not significantly affected by method of crack reduction (egg saver, nest curtain, or position of nest opening) or by hybrid. On average, the proportions of eggs laid in nests in Hellmann and Victorsson cages, respectively, were 88.0 and 94.8%, and the proportions in litter baths were 1.0 and 0.09%. The proportion of artificial turf in nests significantly affected the proportion of eggs laid in nests (P 0.05); on average 84.8% were in nests with 30% artificial turf covering, and 95.4% were in nests with 100% turf covering. The proportion of eggs in litter baths in FC-6 cages was not affected by proportion of nest bottom lining and was, on average, 0.30%. In all three furnished-cage models, the proportions of eggs laid in nest boxes increased with bird age (P 0.001 to 0.01), from 66.9 to 96.2% in Hellmann cages, from 86.8 to 96.3% in Victorsson cages, and from 82.1 to 90.4% in FC-6 cages. The first recording was at 24 wk of age, and the last was at 79 wk of age. In Victorsson and FC-6 cages, significant effects of bird age were found also in proportions of eggs laid in litter baths (P 0.001), but no trend in terms of higher or lower proportions with increasing age was observed. DISCUSSION On the whole, the mortality in the present study was higher than reported earlier by Abrahamsson and Tauson (1997) and Wall et al. (2002). Even if anemia was the primary cause of death only in a few hens according to the autopsy protocols, it is likely that the invasion of red mites affected the general health status of all birds. The significantly higher mortality in cages with 100% AstroTurf compared with 30%, is difficult to explain. Besides the delay in laying of the first egg, birds subjected to the unintended water deficiency seemed to recover quickly, and no long-term negative effects were detected. The significantly higher egg weight found in in Hellmann cages without an egg saver cannot be explained and is assumed to be a random effect. In CO cages, the hybrids were ranked similarly to those in the furnishedcage models, in regards to production parameters. Hence, too few replicates were probably the reason for differences in production not being significant in CO cages. In CO, LSL produced a considerably high egg mass per hen housed, compared with that in the furnished-cage models. This result can be partly explained by the higher egg weight but also by the very low mortality found only with LSL in CO cages. In nests with long nest curtains (the Victorsson cage), eggs stopped rolling when they reached the curtain or they rolled into the cradle but with reduced speed. Eggs that stopped at the curtain started to move again when cage floor vibrated due to bird movements in the cage or when touched by another egg or a hen. In Hellmann cages, the egg-saver wire stopped all eggs gently; when the wire was lifted, the eggs rolled slowly down to the egg cradle. Hence, both devices were very effective in lowering the proportions of cracked eggs in the furnished cages. However, high levels of cracked eggs have been reported in furnished cages for large groups of hens with egg savers (van Niekerk and Reuvekamp, 1999). Their results have shown that even if measures to reduce cracked eggs are used, the design of the cage and the nest, especially, are important. A disadvantage of long nest curtains and egg savers is that the eggs are retained for some time within reach of the hens. However, in the present study, egg eating was a problem in only one cage. The problem occurred during the second half of the study in a cage with an egg saver. Based on the present study, no recommendations regarding the location of nest opening can be given, as no difference was found in egg quality, the birds use of nests, or any other parameter measured. However, a practical consideration is that inspection from the aisle is easier if the nest opening is closer to the front of the cage than the rear. The increase in proportion of cracked eggs with bird age was expected, as eggs become larger and egg shells thinner with increasing age. The significantly higher proportion of cracked eggs in HYB when housed in CO cages cannot be explained. Eggs with high shape indexes, i.e., more globular in shape, might have a higher risk of becoming cracked when rolling down to the egg cradle, due to higher acceleration (Wall et al., 2002). In the present study, however, the shape index was higher for HYB eggs than for LSL eggs from Hellman cages, but no difference was found among the hybrids for proportion of cracked eggs. This result illustrates that the problem with cracked eggs in furnished cages may be multifactorial and affected by bird behavior, eggshell characteristics, and nest design. It is likely that the lower proportions of dirty eggs found for HYB in the cage models tested arose because the person operating the candling machine had greater difficulty in detecting dirt on brown-shelled eggs than on white. The general levels of dirty eggs were higher in the present study than in earlier studies on egg quality in furnished cages (Abrahamsson and Tauson, 1997; Wall et al., 2002). However, what is perceived as a dirty egg varies with each person, and, therefore, comparisons among studies should be made with caution. It is more important that the average proportions in several models of furnished cages, e.g., Victorsson, in fact were lower than in the CO cages, which differs from reports on the older models used by Abrahamsson and Tauson (1995). This finding illustrates that furnished cages might have improved over time in this respect. The curtain strips that separate the nest from the cage area in Hellmann cages enabled birds to lie with one part of the body, generally the front part, inside the nest with the rest of the body outside. This position occurred, e.g., when the nesting area was occupied by one or several other hens and resulted in some eggs being laid close to,

EGG QUALITY IN FURNISHED CAGES FOR LAYERS 347 but outside, the nest. These eggs would probably have been laid inside the nest if the nest wall had been rigid instead of flexible. Hence, if a method of crack reduction is present only outside and in front of the nest area, e.g., such as the long nest curtain in Victorsson cages, eggs laid outside the nest might have a higher risk of getting cracked. Some egg producers repeatedly run the egg belt a short distance during the period when most eggs are laid to reduce the risk of eggs colliding in the egg cradle. However, as the method has not been scientifically evaluated, its effectiveness is not known. Overall, the proportions of eggs laid in nests were high, especially at the end of the study, and the proportions of eggs in litter baths were very low. In Hellmann cages, the proportion of eggs in the baths was higher than in the other furnished-cage models, most likely due to the lack of a closing mechanism. However, the proportion was also moderate in the Hellmann cages, possibly due to the litter boxes being closed until 24 wk of age, when the birds had become accustomed to laying eggs in nests. Partly covering of the nest bottom with artificial turf was not successful. The lower attraction to such nests concurs with findings reported by Wall et al. (2002). However, in the present study the decrease in proportion of eggs laid in nests was quite moderate. Whereas Wall et al. (2002) found no effect of nest bottom lining on proportion of dirty eggs, the proportion in the present study was significantly higher with less AstroTurf lining, in contrast to what was expected. Dirty nest linings are most likely caused by droppings from hens staying in the nest for a longer period, e.g., during the night. Given that staying overnight is a rare event, nest hygiene seems to be at least as good in nests with full covering with AstroTurf as in nests with less AstroTurf. It is possible that the manipulation of nesting materials by birds during the nesting activity may remove dirt and, thereby, contributes to good hygiene. It is interesting to note, however, that the higher proportions of dirty eggs in nests with less lining agrees with the higher proportion of dirty eggs in the CO cages, in which all eggs are laid on the wire floor. The use of perches after dark was slightly lower than reported in several earlier studies (Abrahamsson et al., 1996; Abrahamsson and Tauson, 1997; Appleby, 1998) but is in agreement with levels found by Wall et al. (2002). In the present study, it was observed that birds were not always settled in their locations during the recordings; the presence of red mites might have contributed to a general restlessness and, in turn, lower proportions of birds on perches. For the HYW, the proportion of birds on perches at night was higher than for the HYB and LSL; HYW was also the hybrid with the lowest live weight. Lower live weights require less perch space to be occupied per hen. However, lack of perch space was probably not the reason for the lower use of perches with the other two hybrids, as the difference in live weight between HYW and LSL birds was quite moderate. Few birds were found in litter baths after dark in Hellmann cages, despite the baths being open around the clock. It has been shown that hens prefer elevated perches, when able to choose, and struggle to avoid end positions on the perch (Olsson and Keeling, 2000). In the present study, roosting close to other hens in the cage was probably more important than roosting at the most elevated place in the cage, i.e., the litter bath. In Victorsson cages, the increased use of perches after dark as the birds aged agrees with results published by Wall et al. (2002). The reasons for HYW hens, when housed in cages with long nest curtains, having significantly lower egg masses and cleaner feet are not known. The significantly higher plumage scores, indicating a cleaner plumage on HYB in the furnished cages, were most likely due to dirt in the plumage being harder to detect on brown birds. In CO cages, all hybrids had rather good plumage hygiene, probably because there were no obstacles, e.g., perches, to cause areas where dirt could accumulate. Hybrid differences regarding foot hygiene in furnished cages have been reported earlier by Tauson and Abrahamsson (1996), who found dirtier feet on ISA- Brown birds compared with LSL birds. In the present study, the cause of dirtier feet found on HYB birds when housed in Victorsson cages is not known. It is possible that HYB, being a heavier hybrid, had lower locomotive activity, which in turn caused a dirtier floor; however, locomotive activity and cage hygiene were not recorded. In conclusion, in most furnished-cage models, egg laying is concentrated to a very narrow area, as compared with conventional cages without nests. Egg savers and long nest curtains proved to be effective measures in reducing cracks in such models, whereas there were no differences due to location of the opening, i.e., front vs. rear of nest. Thus, adding these devices to the furnishedcage model implied similar proportions of cracks to conventional cages. Reduction of the proportion of AstroTurf lining in nests cannot be recommended, as no benefit in egg quality was obtained, and bird use of the nest was decreased. ACKNOWLEDGMENTS The Swedish Farmers Foundation for Agricultural Research and the Swedish Board of Agriculture are thanked for financial support. REFERENCES Abrahamsson, P., and R. Tauson, 1997. Effects of group size on performance, health and birds use of facilities in furnished cages for laying hens. Acta Agric. Scand. Sect. A Anim. Sci. 47:254 260. Abrahamsson, P., R. Tauson, and M. C. Appleby, 1995. Performance of four hybrids of laying hens in modified and conventional cages. Acta Agric. Scand. Sect. A Anim. Sci. 45:286 296. Abrahamsson, P., R. Tauson, and M. C. Appleby, 1996. Behaviour, health and integument of four hybrids of laying hens in modified and conventional cages. Br. Poult. Sci. 37:521 540. Appleby, M. C., 1990. Behaviour of laying hens in cages with nest sites. Br. Poult. Sci. 31:71 80. Appleby, M. C., 1998. The Edinburgh Modified Cage: Effects of group size and space allowance on brown laying hens. J. Appl. Poult. Res. 7:152 161.

348 WALL AND TAUSON Appleby, M. C., and B. O. Hughes, 1995. The Edinburgh modified cage for laying hens. Br. Poult. Sci. 36:707 718. Appleby, M. C., and H. E. McRae, 1986. The individual nest box as a super-stimulus for domestic hens. Appl. Anim. Behav. Sci. 15:169 176. Duncan, I. J. H., and V. G. Kite, 1989. Nest site selection and nest-building behaviour in domestic fowl. Anim. Behav. 37:215 231. European Commission, 1999. Council Directive 1999/74/EC of 19 July 1999 laying down minimum standards for the protection of laying hens. Official Journal L 203, 03/08/1999. Hughes, B. O., 1993. Choice between artificial turf and wire floor as nest sites in individually caged laying hens. Appl. Anim. Behav. Sci. 36:327 335. Jordbruksverket, 2000. Ny teknikprovning av Trivsel-buren för åtta värphöns. Jordbruksverket, 551 82 Jönköping, Sweden. (In Swedish). Lee, D. J. W., and W. Bolton, 1976. Battery cage shape: The laying performance of medium- and light-body weight strains of hens. Br. Poult. Sci. 17:321 326. Nordenfors, H., 2000. Epidemiology and control of the poultry red mite, Dermanyssus gallinae. Ph.D. Thesis. Department of Veterinary Microbiology, Swedish University of Agricultural Sciences, Uppsala, Sweden. Olsson, I. A. S., and L. J. Keeling, 2000. Night-time roosting in laying hens and the effect of thwarting access to perches. Appl. Anim. Behav. Sci. 68:243 256. Reed, H. J., and C. J. Nicol, 1992. Effects of nest linings, pecking strips and partitioning on nest use and behaviour in modified battery cages. Br. Poult. Sci. 33:719 727. SAS Institute Inc., 1996. System for Windows, Release 6. SAS Institute Inc., Cary, NC. Snedecor, G. W., and W. G. Cochran, 1989. Statistical Methods. 8th ed. Iowa State University Press, Ames, IA. Statens jordbruksverk, 1997. Statens jordbruksverks författningssamling. SJVFS 1997:71, L 106. Statens jordbruksverk, 551 82 Jönköping, Sweden. (In Swedish). Tauson, R., and P. Abrahamsson, 1996. Foot and keel bone disorders in laying hens. Effects of artificial perch material and hybrid. Acta Agric. Scand. Sect. A Anim. Sci. 46:239 246. Tauson, R., T. Ambrosen, and K. Elwinger, 1984. Evaluation of procedures for scoring the integument of laying hens Independent scoring of plumage condition. Acta Agric. Scand. 34:400 408. van Niekerk, T. G. C. M., and B. F. J. Reuvekamp, 1995. Improving the welfare of laying hens in cages. Misset World Poult. 11:38 41. van Niekerk, T., and B. Reuvekamp, 1999. Legehennen in großn Gruppenkäfigen. Heutiges System nicht praxisreif. DGS Magazin 26:18 22. Wall, H., R. Tauson, and K. Elwinger, 2002. Effect of nest design, passages and hybrid on use of nest and production performance of layers in furnished cages. Poultry Sci. 81:333 339.