The relationship between physical activity and leg health in the broiler chicken

British Poultry Science " %& ,) # $ * ! ie & ""+ "' %! (" ) ( ev rR ee rP Fo ! " ( w (" -. + "/ (" ! 0" " ! "$". 0" " ! " " "&&" . (" / (" ! 0" " ! "$". 0" " ! " " 1 ' ( . ) " / 2 3" ! "$" + ' . " , # ( "' ' -" " " " (!. / 2 3" ! "$" + ' . " ,# ( ' -" " " " 4 (" . ( (" / (" ! 0" " ! "$". 0" " " " "( 3 . " . + $( $. 4" , " ! ly ' On 5"!6 "' E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 1 of 26 British Poultry Science CBPS-2008-359 ed. MacLeod, Nov 2009 5 The relationship between physical activity and leg health in the broiler chicken Fo L. SHERLOCK*, T.G.M. DEMMERS, A. E. GOODSHIP1, I. D. MCCARTHY1 AND C.M. WATHES The Royal Veterinary College, Veterinary Clinical Sciences, Hatfield, Hertfordshire, ee 10 rP AL97TA and 1Institute of Orthopaedics and Musculoskeletal Science, University College London, Brockley Hill, Stanmore, Middlesex, HA7 4LP, England, UK ev rR RUNNING TITLE: ACTIVITY AND LEG HEALTH IN BROILERS 15 iew On 20 ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 *Correspondence to: Miss Louise Sherlock, The Royal Veterinary College, Veterinary Clinical Sciences, Hatfield, Hertfordshire, AL9 7TA, England. Tel: +44 (0)1707-666-333 25 Fax: +44 (0)1707-666-298 E-mail: lsherlock@rvc.ac.uk Accepted for publication 1st September 2009 E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 2 of 26 British Poultry Science Abstract 1. The relationship between the physical activity and leg health of broiler chickens was assessed on a semi-commercial scale. 2. Three batches of birds (2128 per batch) were raised under two lighting regimes during the photoperiod; either a step-wise change of light intensity alternating between 5 an illuminance of 200 and 10 lx or a constant illuminance of 10 lx. The activity of focal individuals (24 per batch) was observed at 2, 4 and 6 weeks of age, and leg health Fo assessed weekly, based on gait score, the prevalence of burns on the hock and foot pad, and angulation and rotation of the leg at the intertarsal joint. Cortical bone density and rP thickness and area moments of inertia of the mid-physis tibiotarsus were measured post 10 mortem at 6 weeks of age. ee 3. The step-wise change in light intensity did not affect overall performance, activity or leg health. rR 4. An individual bird’s activity did not affect its gait score, the prevalence of hock burn or foot pad burn, cortical density or thickness or shape of the tibiotarsus. Sex of the bird 15 ev was the only factor to affect significantly the area moment of inertia in the horizontal and vertical planes of the tibiotarsus, with females showing a lower moment of inertia iew for both. No variable had a significant effect on cortical density or thickness. Mean cortical density was low across all birds and may indicate that, when allowed to move freely as much or as little as they choose, broiler chickens do not exercise enough or do 20 On not perform the higher impact activities required to affect bone quality. 5. These findings imply that the activity of broiler chickens raised on a semi- ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 commercial scale is unaffected by step-wise changes in light intensity and that other husbandry measures are needed to raise activity and hence improve leg health. INTRODUCTION 25 Background In the UK, over 800 million chickens are produced annually (Defra, 2007), reaching a slaughter weight of over 2 kg within 6 weeks, well before sexual maturity E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 3 of 26 (approximately 18 weeks) and skeletal maturity (approximately 23-27 weeks) (Latimer, 1927; Rath et al., 2000). This fast growth rate is associated with poor leg health that can lead to lameness (Kestin et al., 2001; Sanotra et al., 2001), which has been demonstrated to be detrimental to the bird’s welfare (McGeown et al., 1999) and likely 5 to infringe FAWC’s Five Freedoms (www.fawc.org.uk). Large-scale surveys of lameness using gait scoring techniques in commercial broiler flocks have estimated that Fo about 27% of birds have an abnormal gait of sufficient severity for the birds’ welfare to be compromised (Kestin et al., 1992; Knowles et al., 2008). rP Causes of poor leg health and lameness 10 There may be many causes of lameness in a broiler chicken, with the aetiology ee generally classified as developmental, degenerative or infectious (Bradshaw et al., 2002). Developmental abnormalities include varus-valgus deformation (VVD) at the rR intertarsal joint, one of the most common leg distortions in broilers (EC, 2000), and abnormal rotation within the tibiotarsus or femur. The most widespread degenerative 15 ev disorder in commercial broilers is contact dermatitis, while infectious disorders are suggested to cause the most severe cases of lameness (Kestin et al., 1994). iew Developmental deformities may cause walking difficulties in the birds and increase the risk of injury during catching, causing downgrading or even rejection by the processor (Aviagen, 2001). The underlying cause of these conditions is poorly 20 On understood but high growth rate leading to altered load bearing, lack of activity and genetic factors are possible candidates (Duff and Thorp, 1985; Bradshaw et al., 2002). ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 British Poultry Science Contact dermatitis is a widespread problem in European broiler production (EC, 2000), with lesions occuring on areas of the body in prolonged contact with the litter, most commonly on the feet (foot pad or podo-dermatitis) but also on the hocks, 25 commonly described as hock burns, and breast muscle, particularly in heavier and lame birds (Kristensen et al., 2006b). In a survey of 28 broiler flocks in Denmark (Sanotra et al., 2001) the prevalence of podo-dermatitis was 42%, while others suggest it may be lower at 11.1% in the UK (Haslam et al., 2007). Proper litter management and the E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 4 of 26 British Poultry Science absence of pre-existing leg conditions that cause birds to sit for long periods can reduce the incidence and severity of contact dermatitis. One possible reason for poor leg health in broilers is reduced bone quality. Although the dimensions of the tibiotarsus in selected (modern) strains of broilers fed 5 ad libitum are correct for load support, the bone itself is weak. The cortical bone shows high porosity and low mineral content (Thorp and Waddington, 1997; Williams et al., Fo 2000; Corr et al., 2003a; Williams et al., 2004), which may be due to rapid bone deposition in the outer layers to increase width to support the increasing mass of the rP bird but with insufficient time to allow infilling by osteoblasts due to the birds’ rapid 10 growth rate (Williams et al., 2004), although others found that reducing growth rate did ee not improve bone quality (Leterrier et al., 1998). Welfare concerns rR Broilers from selected strains and fed ad libitum often show a number of morphological changes compared with those of relaxed (i.e. randomly bred) strains (Corr et al., 15 ev 2003a). The result of this altered morphology is a change in walking in an attempt to increase stability (Corr et al., 2003b). This altered gait is considered inefficient and iew would rapidly tire birds, giving an explanation for the low levels of activity seen in all broilers. Sound broilers spend, on average, 76% of their time lying down, but in lame 20 On birds this increases significantly to 86% (Weeks et al., 2000). Time spent walking at slaughter-weight is significantly shortened from 3.3% to 1.5% in the worst cases of ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 lameness. Sound birds also spend more time standing idle, standing preening and standing eating (Weeks et al., 2000). Poor leg health may affect the bird in several ways, with particular concern 25 given to the possibility of pain. McGeown and colleagues (1999) showed that moderately lame birds would complete an obstacle course faster when given an analgesic, suggesting that in this case birds’ walking speed was affected by pain from their gait abnormality, rather than the abnormality itself. Lame birds will also self-select E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 5 of 26 an analgesic (Danbury et al., 2000). However, others have found that administration of an analgesic does not change the walking behaviour of broilers (Corr et al., 2007); thus, the presence of pain arising from poor leg health and its role in walking ability remains contentious. Nevertheless, the differences in behaviour exhibited between sound and 5 lame birds may indicate poor welfare caused in ways other than pain. Alleviation of poor leg health through husbandry Fo Changes in husbandry such as stocking density or feeding regime can alleviate some causes of poor leg health (Su et al., 1999; Sorensen et al., 2000), as can selective breeding. 10 rP Increasing the activity of a broiler chicken has been shown to improve walking ee ability and influence bone development (Reiter and Bessei, 1995) and could be achieved by increasing the distance between feeders and drinkers (Reiter and Bessei, rR 1996) or by the addition of barriers (Bizeray et al., 2002). These practices increase the thickness and density of the cortical bone and the diameter of the tibiotarsus diaphysis 15 ev through the deposition of bone in response to the imposed stresses, as explained by Wolff’s Law (1892). Increased exercise also reduces the prevalence of abnormalities in iew the physis and physeal vasculature of bone extremities from 40 to 18% (Thorp and Duff, 1988). Another method found to be effective in encouraging activity in broiler 20 On chickens is providing step changes in light intensity (Kristensen et al., 2006a). The colour and source of artificial light can also affect activity; low-frequency fluorescent ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 British Poultry Science lighting lowers activity compared with high-frequency (Boshouwers and Nicaise, 1992) and activity is higher in red than in blue light both at the same intensity, and at higher intensities (Prayitno et al., 1997). Gait abnormalities were also reduced with red light. 25 The aim of this experiment was to investigate the relationship between physical activity and leg health of broiler chickens on a semi-commercial scale, while also exploring the effect of step changes in light intensity on individual activity within a E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 6 of 26 British Poultry Science large flock. Our hypothesis was that step-wise changes in light intensity would increase individual activity thereby improving leg health. MATERIALS AND METHODS Animals, housing and husbandry 5 Three batches of Ross 308 broilers (approximately 2128 chicks ‘as hatched’ per batch) were obtained at d-old from a commercial hatchery (P D Hooks, Bampton) and used in Fo a large scale multi-factorial experiment. On arrival, the chicks were distributed randomly between eight identical rooms, each holding 266 birds. Within each batch 40 rP sexed chicks – 20 of each sex – were marked for later identification then evenly 10 distributed between two of the rooms. At 9 d old, 6 birds of each sex from these marked ee birds in both rooms were assigned one of 12 symbols. These were sprayed with black stock marker for identification by overhead cameras. The marks were refreshed once or rR twice weekly as required. These individuals became the ‘focal’ birds for further detailed studies of behaviour. 15 ev The eight 19.6 m2 rooms were filled with wood shavings to a depth of at least 7.5 cm and warmed to a temperature of 30 oC at a relative humidity of about 70% prior iew to the chicks’ arrival. The birds were kept under normal commercial husbandry conditions until 49 d old when they weighed approximately 2.5 kg, equivalent to a stocking density of 34 kg m-2. For the first 7 d, fluorescent lighting was provided on a 20 On 23:1h cycle of light: dark with lights on at 01.00 and a dawn and dusk period each of 30 minutes. From day 7 onwards, this cycle changed to 18:6h of light: dark, with lights on ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 at 03.00 and the same dawn and dusk period. Illuminance was dependent upon the experimental treatment. Birds were provided with ad libitum access to water and were fed on a 25 standard, 4-stage commercial diet (ABN, Peterborough), provided in 4 meals per d at 03:00, 09:00, 15:00 and 20:00. Chicks were first provided with a starter crumb until dy 11, a starter pellet from d 12 to 23, grower pellets from dy 24 to 44 and a withdrawal pellet ration between d 45 and 49. The diet was formulated to provide 90% of the E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 7 of 26 normal crude protein content and the feed quantity was 5% below the normal allowance to encourage slower growth and extend the experimental period. Room temperature was reduced according to recommended guidelines (Aviagen, 2002). Humidity was maintained at approximately 70% relative humidity or 5 higher using a misting system. Animal health was monitored twice daily by a skilled stockman and severely lame birds were culled humanely. Excluding the focal birds, the Fo overall mortalities to 49 d were as follows; Batch 1: 3.7%, Batch 2: 5.5% and Batch 3: 2.3%. The higher mortality of the second batch was due to a yolk sac infection. 10 rP Experimental design ee The experiment started on d 14 and finished on d 44 when the focal birds were killed by cervical dislocation. rR The main experiment, the results of which will be reported separately, comprised a 23 multi-factorial design with three replicates (batches) of each treatment. 15 ev This investigated the response of broilers to step-wise changes in humidity, illuminance and feed quantity. In our study, however, the only treatment under consideration was iew the step-wise change in light intensity. In each batch this treatment was applied in one room, to which the focal birds were assigned in addition to the control room. Step-wise changes in light intensity 20 On The illuminance of the control room was kept at 10 lx throughout the photoperiod. The light intensity treatment comprised step-wise changes in light intensity alternating ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 British Poultry Science between an illuminance of 10 lx and 200 lx throughout the photoperiod, as used in our previous research (Kristensen et al., 2006a). There were 4 step changes each day (i.e. dim to bright and vice versa, twice) as illustrated in Figure 1. The illuminance 25 immediately following dawn alternated each day, to balance the experimental design. Experimental procedures Figure 1 near here Activity of the focal birds E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 8 of 26 British Poultry Science Video recordings were made of the focal birds when they were 2, 4 and 6 weeks of age. Recording took place between 04:00 and 05:00, 09:00 and 10:00 and 14:00 and 15:00, covering a steady light state, one feed and 20 minutes after a step change in light intensity. This schedule also allowed time for inspection by the stockman without 5 influencing the observed birds’ activity. The activity was quantified by counting the number of gridlines crossed by an individual every 5 minutes over an hour using a grid Fo of 36 squares placed over the video image on a VDU. After initial analysis, more detailed investigation was required, so the bird’s activity during a period of 5 minutes rP immediately following a step up in illuminance (at 10.18 am) was also examined by the 10 same method when the birds were 4, 5 and 6 weeks of age. Leg health ee Measurements of leg health were taken while the birds were alive and post mortem. rR During the experimental period the focal birds were assessed weekly using gait scoring (Kestin et al., 1992; Garner et al., 2002), the prevalence and severity of any hock and 15 ev foot pad burns, and the presence or absence of leg deformity (rotation and varus-valgus deformation). To measure gait score (GS), the selected bird was herded, its movement iew observed and scored on a 6-point scale (0 - completely sound bird; 5 – unable to stand). However, in this study any birds of gait scores 4 and 5 were culled on ethical grounds. Hock burns were scored 0 - no mark, 1 - reddening, 2 – mild scabbing (<10% hock with 20 On lesion) and 3 – severe scabbing (>10% hock with lesion). Foot pad burns were scored 0 – no mark, 1 – mild lesion (<5mm) and 2 – severe lesion (>5mm) (Dawkins et al., 2004). ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Leg deformities were assessed by both varus-valgus deformation and abnormal rotation of the legs at the intertarsal joint. The bird was inverted, ventral side towards 25 handler, and held by the legs below the intertarsal joint. Varus-valgus deformation and rotation were assessed subjectively with scores assigned as follows: 0 for straight, 1 for not straight (varus - angle of >22o where legs meet; valgus - angle of >/= 30o between legs; rotation – pads face each other >15o) (Dawkins et al., 2004). The presence or E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 9 of 26 absence of crooked toes was also noted. Body weight was also measured after leg assessment prior to returning the individual to the flock. At the end of the experimental period, the focal birds were killed by cervical dislocation and both legs removed at the coxofemoral joint. They were then stored at 5 minus 20°C until analysis. The mid-physis of the tibiotarsus in both legs was scanned using a pQCT (peripheral quantitative computed tomography) scanner (StraTec XCT Fo 2000; Pforzheim, Germany) and a loop function performed for a measure of cortical bone density and thickness and area moments of inertia. Five slices, each 1 mm apart, rP were taken and the mean calculated. 10 Data analysis ee While the majority of analysis involved data from all weeks from all batches, the more detailed investigation of activity and leg health was undertaken only on the final batch rR of birds at 2, 4 and 6 weeks of age due to the time taken to collect and analyse this data. Minitab version 15 was used for all analysis. 15 ev The relationship between gait score and weight was determined using a nonparametric correlation coefficient (Spearman’s). The differences in weight between iew sexes and treatments were examined in a General Linear Model (GLM). Chi squared analysis was used to compare the distribution of gait score and other leg health measurements between treatments. Where expected frequencies of cells were less than 20 5, Fisher’s exact test was used. On Hourly activity data were transformed by square root to conform to the ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 British Poultry Science assumptions of a parametric test. A GLM was then used to analyse the activity data against age, time of day and light treatment and the interactions therein. The activity data obtained over the 5 minutes immediately following a step up 25 in light intensity could not be transformed fully so were analysed using the nonparametric Kruskal Wallis test for activity against treatment and activity against age. A two-way ANOVA was then used with logged activity data to check interactions E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps British Poultry Science between treatment and age, although attention was paid to statistical significance (P) to ensure agreement with the results of the non-parametric tests. A GLM was used to analyse transformed daily activity data against weight, age, sex, gait score and hock burn score. PQCT data from both legs were then analysed 5 using paired t-tests to determine whether the density and thickness of the cortical bone or the moments of inertia of the tibiotarsus differed significantly between the right and Fo left. A multivariate GLM was then performed of cortical density and thickness and both horizontal and vertical moments of inertia against gait score, sex, mean activity and rP final weight. 10 RESULTS ee Light treatment effects on leg health and production Figure 2 shows the relationship found between gait score and mean weight across the rR growing period for all batches and ages (2 to 6 weeks). Gait score and weight were significantly correlated (Spearman’s r = 0.7, P = 0.01); heavier birds had a poorer gait. 15 ev Mean weights differed between sexes (F = 45.8, P < 0.005) at slaughter at 6 weeks; males - 2.28 ± 0.38 kg and females - 2.04 ± 0.22 kg. Batch also affected weight (F = 19.9, P < 0.005). iew Figure 2 near here Light treatment did not affect mean weekly weight (P = 0.736) and had no significant effect on the overall numbers of sound (GS 0) or lame (GS 1-3) birds neither 20 On across individual gait scores (χ2 = 0.668, P = 0.881) nor when grouped into sound vs. lame categories (χ2 = 0.445, P = 0.505). The majority (68.6%, n = 70) of birds displayed ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 10 of 26 only slight gait impairment (GS 0 or 1) in both light treatments at week 6. Data from the final week of study from all batches (n = 70) showed that neither the prevalence nor severity of any leg health measure was affected by light treatment. 25 Table 1 shows two examples: rotation of the tibiotarsus (Fisher’s exact test: P = 0.149) and hock burns (χ2 = 5.5, P = 0.064), where the latter could be considered a trend for the step change treatment to reduce the prevalence of the highest severity of hock burns. Table 1 near here E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 11 of 26 Light treatment effects on activity The GLM analysis showed that light treatment did not affect the mean activity of focal birds at any age (n = 24, F = 12.3, P = 0.069), although this could be considered a trend with the control light treatment producing higher activities than the step change. There 5 was no interaction between light treatment and age or time. Activity decreased with age in both light treatments (Figure 3). Time and age interacted significantly to affect Fo activity levels in both light treatments (Figure 4; F = 25.5, P = 0.004); as age increased, the difference in activity between times decreased. Table 2 shows the average hourly rP activity (represented by grid lines crossed) at each age across the focal birds. 10 Figures 3, 4 near here Given the trend for activity of the focal birds to be higher in the control room ee than with a step change in light intensity, the period (5 minutes) immediately following a step up in illuminance was examined to test the hypothesis that step change birds were rR more active during this period then became less active than the control birds at other times (i.e. those used in the original study). Treatment did not affect activity (P = 15 ev 0.572), while age remained highly significant (F = 10.71, P < 0.001) and there was Table 2 near here again no significant interaction between the two factors. Individual activity and leg health iew The GLM analysis showed that none of sex, gait score, hock burn score or weight affected daily activity of focal birds. The only significant result was found with age (F 20 On = 5.97, P = 0.004); i.e. younger birds were more active than older birds. Data from bird 11 were removed as this individual had a large influence, being ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 British Poultry Science an anomaly outwith the rest of the data due to the bird’s low weight. Table 3 shows the descriptive data for cortical density and area moments of inertia in each leg as well as the final weight of the birds. The statistical analysis demonstrated that there was no 25 difference between the right and left legs of the birds for any value; cortical bone density (P = 0.461), thickness (P = 0.232) horizontal plane area moment of inertia (P = 0.319), vertical plane area moment of inertia (P = 0.977). Therefore, for subsequent analysis only data from the right legs were used. Females showed lower moments of E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps British Poultry Science inertia (lower resistance to bending) for both horizontal (F = 5.0, P = 0.04) and vertical (F = 13.8, P = 0.002) planes of the tibiotarsus; females: 142.1 ± 24.7 and 163.1 ± 26.2 mm4 and males: 198.3 ± 46.4 and 233.0 ± 35.2 mm4. Final weight also increased the horizontal plane moment of inertia (F = 15.7, P = 0.001), but this is likely to be 5 somewhat confounded by sex (females weighed less than males at slaughter age, 6 weeks). No variable had a significant effect on cortical density or thickness. The R2 Table 3 near here Fo value was high at 0.98. DISCUSSION rP Effects of step-wise changes in light intensity on leg health and production 10 The correlation seen between gait score and weight was expected. Other studies have ee reported a strong positive correlation between gait score and body weight (Kestin et al., 1992; Sorensen et al., 1999; Vestergaard and Sanotra, 1999; Kestin et al., 2001; rR Kristensen et al., 2006a), although this is sometimes confounded when severely lame birds (scores 4 and 5) are included since an inability to feed causes a loss in body 15 ev weight. At a slaughter age of 6 weeks, a broiler chicken’s skeleton is not mature, so heavier birds put more strain on soft bones and less fully formed joints. Complete iew ossification does not occur in the chicken until around 25 weeks (Latimer, 1927; Rath et al., 2000). The rapid growth rate of modern strains of broilers leads to less dense and poorly mineralised cortical bone (Williams et al., 2000; Williams et al., 2004), which 20 On could increase the risk of fractures and leg deformations (varus-valgus or rotation). Heavier birds may also tire faster, either from their sheer weight or from an altered morphology, so are reluctant to walk. ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 12 of 26 The light treatment had no significant effect on overall gait score although there was a trend for the number of birds with the highest severity of hock burn to be reduced 25 with the step-wise changes in light intensity. This may have been due to a reduction in time spent sitting or a drier litter with reduced ammonia content. A study on turkeys found that the incidence of leg abnormalities, as observed by gait and deformities at the intertarsal joint, was significantly reduced in high intensity light regimes (Hester et al., E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 13 of 26 1983). However, Kristensen and colleagues showed that broiler leg health was unaffected by light intensity (Kristensen et al., 2006b). Overall, our light treatment had no significant effect on the leg health of broilers. Activity and light treatment 5 There was insufficient evidence to support the original hypothesis that the step-wise changes in light intensity increases the activity of individual birds. On the basis of Fo previous work (Kristensen et al., 2006a), we had expected activity to be greater when the step-wise changes were implemented, but lighting treatment had no significant rP effect on the activity of individuals; indeed there was a trend for control birds to be 10 more active than those in the step change lighting treatment. This is contrary to the ee results of Kristensen et al. who found that a similar step-wise change in light intensity resulted in greater activity of broilers during the periods of bright intensity (Kristensen rR et al., 2006a), but is similar to a study in which male turkeys were less active during the period of higher light intensity (Hester et al., 1987). There are several explanations for 15 ev these conflicting findings. Firstly Kristensen et al. (2006) only used 4 broiler chickens in each group, compared to much larger flock of 266 birds (of which 12 were the focal iew birds). There is some evidence of greater jostling between individuals at higher stocking densities (Dawkins et al., 2004), which may have affected the activity of the birds preventing similar results to be seen, and broilers have also been observed to walk 20 On longer distances at lower densities (Lewis and Hurnik, 1990). Secondly, it may be the case that birds response with a spurt of activity to the intensity change per se rather than ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 British Poultry Science the absolute intensity. This was tested by observing activity immediately after a step change in intensity, but there was no significant difference between treatments. Individual activity and leg health 25 Rath and colleagues have surmised that bone density can be used as an index of bone strength (Rath et al., 2000), since tibiotarsus breaking strength and bone density are significantly correlated (Frost and Roland, 1991). Therefore in our study cortical density was used as a proxy measure for bone strength. There was insufficient evidence E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 14 of 26 British Poultry Science to support the original hypothesis that increased activity promotes good leg health, as measured in terms of gait, the prevalence of hock burn, cortical bone density and thickness and the moments of inertia of the tibiotarsus. This is contrary to the findings of Reiter et al. (1995), who showed that active birds had thicker and denser cortical 5 bone in the tibiotarsus compared with less active controls. However, the former had been trained to run 100 m in 20 minutes daily on a treadmill, whereas the controls moved freely around their home pen (a 1 x 1m enclosure with 8 birds per m2). The Fo distance travelled by these trained birds was mostly higher than the hourly average rP activity of the focal birds in our experiment, which ranged from approximately 140.6 m 10 at two weeks of age to 46.1 m at 6 weeks (estimated from the average distance between ee grid lines and the number crossed). It is a strong possibility that the activity of our broilers, even in those considered more active, was insufficient to affect cortical density rR or shape of the tibiotarsus. An alternative explanation is that treadmill-based exercise imposes a greater dynamic load than normal walking, enlarging the circumference of 15 ev the tibiotarsus; however this would not explain the increased bone density. Reiter and colleagues reported that by training the fast growing strain of birds, the reduction in iew activity normally seen at three weeks of age was delayed until 6 weeks. While the final liveweight was unaffected by exercise, perhaps the high growth rate typical of broilers was delayed enough to allow sufficient infilling by osteoblasts, a process suggested by 20 On Williams et al. (2004). The mean cortical densities of the birds in this study are higher than those reported by Reiter et al., but this can be explained by the location of ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 scanning used in the latter study, namely at the proximal end of the tibiotarsus rather than at the mid point used here. However, both measures of cortical density are much lower than those found in comparable studies of slower growing broiler breeders (1160 25 mg/cc at a liveweight of 0.99 kg at five weeks of age, and 1250 mg/cc at 2.5kg at 15 weeks of age (Rath et al., 2000)). The cortical density of the tibiotarsus of adult layers is also much higher at 1350mg/cc (Zhang and Coon, 1997). The significance of sex on the shape of the tibiotarsus may be related to the change in width of the bone in E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 15 of 26 response to increased loading, i.e. mass, in the males. Our results indicate that the fast growing, young broiler chickens used in meat production have weak legs for their body size that, despite an increased width to reduce bending, may nonetheless put them at a greater risk of leg fractures and leg deformations. 5 Conclusions In this study, carried out in a semi-commercial setting in contrast to other studies of its Fo kind, a novel lighting regime of step-wise changes in light intensity did not affect performance, leg health or activity of broiler chickens. There was no association rP between an individual’s activity and its leg health, measured in terms of gait, the 10 prevalence of hock burn, tibiotarsus cortical bone density and thickness and shape. ee Although this latter finding is counter intuitive, one plausible explanation is that modern, fast growing strains of broiler chickens do not take sufficient exercise to rR produce a difference in bone quality when given the choice. ACKNOWLEDGEMENTS 15 ev Funding was provided by the BBSRC. Support was provided by Professor Claire Wathes, John Lowe and staff from the RVC’s Biological Services Unit and Professor Goodship’s research group. iew REFERENCES 20 AVIAGEN (2001) Leg health in broilers. Ross Tech. Newbridge, Scotland, Aviagen. On AVIAGEN (2002) Broiler management manual. Ross. Newbridge, Scotland, Aviagen. 25 ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 British Poultry Science BIZERAY, D., ESTEVEZ, I., LETERRIER, C. & FAURE, J. M. (2002) Influence of increased environmental complexity on leg condition, performance, and level of fearfulness in broilers. Poultry Science, 81, 767-773. 30 BOSHOUWERS, F. M. G. & NICAISE, E. (1992) Responses of broiler chickens to high-frequency and low-frequency fluorescent light. British Poultry Science, 33, 711-717. E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 16 of 26 British Poultry Science BRADSHAW, R. H., KIRKDEN, R. D. & BROOM, D. M. (2002) A review of the aetiology and pathology of leg weakness in broilers in relation to welfare. Avian and Poultry Biology Reviews, 13, 45-103. 5 10 CORR, S. A., GENTLE, M. J., MCCORQUODALE, C. C. & BENNETT, D. (2003b) The effect of morphology on walking ability in the modern broiler: A gait analysis study. Animal Welfare, 12, 159-171. CORR, S. A., MCCORQUODALE, C., MCDONALD, J., GENTLE, M. & MCGOVERN, R. (2007) A force plate study of avian gait. Journal of Biomechanics, 40, 2037-2043. rP 15 CORR, S. A., GENTLE, M. J., MCCORQUODALE, C. C. & BENNETT, D. (2003a) The effect of morphology on the musculoskeletal system of the modern broiler. Animal Welfare, 12, 145-157. Fo DANBURY, T. C., WEEKS, C. A., WATERMAN-PEARSON, A. E., KESTIN, S. C. & CHAMBERS, J. P. (2000) Self-selection of the analgesic drug carprofen by lame broiler chickens. Veterinary Record, 146, 307-311. 20 ee DAWKINS, M. S., DONNELLY, C. A. & JONES, T. A. (2004) Chicken welfare is influenced more by housing conditions than by stocking density. Nature, 427, 342-344. 25 rR DEFRA (2007) Poultry and poultry meat statistics notice. London, Department for Environment, Food and Rural Affairs. ev 30 DUFF, S. R. & THORP, B. H. (1985) Abnormal angulation/torsion of the pelvic appendicular skeleton in broiler fowl: Morphological and radiological findings. Research in Veterinary Science, 39, 313-319. iew EC (2000) The welfare of chickens kept for meat production (broilers). Report of the scientific committee on animal health and animal welfare. Brussels, European Commission. 35 On FROST, T. J. & ROLAND, D. A., SR. (1991) Research note: Current methods used in determination and evaluation of tibia strength: A correlation study involving birds fed various levels of cholecalciferol. Poultry Science, 70, 1640-1643. 40 ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 GARNER, J. P., FALCONE, C., WAKENELL, P., MARTIN, M. & MENCH, J. A. (2002) Reliability and validity of a modified gait scoring system and its use in assessing tibial dyschondroplasia in broilers. British Poultry Science, 43, 355-363. 45 HASLAM, S. M., KNOWLES, T. G., BROWN, S. N., WILKINS, L. J., KESTIN, S. C., WARRISS, P. D. & NICOL, C. J. (2007) Factors affecting the prevalence of foot pad dermatitis, hock burn and breast burn in broiler chicken. British Poultry Science, 48, 264 - 275. 50 E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 17 of 26 HESTER, P. Y., ELKIN, R. G. & KLINGENSMITH, P. M. (1983) Effects of high intensity step-up and low intensity step-down lighting programs on the incidence of leg abnormalities in turkeys. Poultry Science, 62, 887896. 5 HESTER, P. Y., SUTTON, A. L. & ELKIN, R. G. (1987) Effect of light intensity, litter source, and litter management on the incidence of leg abnormalities and performance of male turkeys. Poultry Science, 66, 666-675. 10 KESTIN, S. C., ADAMS, S. J. M. & GREGORY, N. G. (1994) Leg weakness in broiler chickens, a review of studies using gait scoring. Proceedings of the 9th European Poultry Conference. Glasgow, UK, WPSA. rR 25 KESTIN, S. C., KNOWLES, T. G., TINCH, A. E. & GREGORY, N. G. (1992) Prevalence of leg weakness in broiler chickens and its relationship with genotype. Veterinary Record, 131, 190-194. ee 20 KESTIN, S. C., GORDON, S., SU, G. & SORENSEN, P. (2001) Relationships in broiler chickens between lameness, liveweight, growth rate and age. Veterinary Record, 148, 195-197. rP 15 Fo KNOWLES, T. G., KESTIN, S. C., HASLAM, S. M., BROWN, S. N., GREEN, L. E., BUTTERWORTH, A., POPE, S. J., PFEIFFER, D. & NICOL, C. J. (2008) Leg disorders in broiler chickens: Prevalence, risk factors and prevention. PLoS ONE, 3, e1545. ev 30 KRISTENSEN, H. H., PERRY, G. C., PRESCOTT, N. B., LADEWIG, J., ERSBOLL, A. K. & WATHES, C. M. (2006b) Leg health and performance of broiler chickens reared in different light environments. British Poultry Science, 47, 257-263. On 35 KRISTENSEN, H. H., AERTS, J. M., LEROY, T., WATHES, C. M. & BERCKMANS, D. (2006a) Modelling the dynamic activity of broiler chickens in response to step-wise changes in light intensity. Applied Animal Behaviour Science, 101, 125-143. iew LATIMER, H. B. (1927) Postnatal growth of the chicken skeleton. American Journal of Anatomy, 40, 1-57. 40 ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 British Poultry Science LETERRIER, C., ROSE, N., CONSTANTIN, P. & NYS, Y. (1998) Reducing growth rate of broiler chickens with a low energy diet does not improve cortical bone quality. British Poultry Science, 39, 24 - 30. 45 50 LEWIS, N. J. & HURNIK, J. F. (1990) Locomotion of broiler chickens in floor pens. Poultry Science, 69, 1087-1093. MCGEOWN, D., DANBURY, T. C., WATERMAN-PEARSON, A. E. & KESTIN, S. C. (1999) Effect of carprofen on lameness in broiler chickens. Veterinary Record, 144, 668-671. E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 18 of 26 British Poultry Science PRAYITNO, D. S., PHILLIPS, C. J. & STOKES, D. K. (1997) The effects of color and intensity of light on behavior and leg disorders in broiler chickens. Poultry Science, 76, 1674-1681. 5 RATH, N. C., HUFF, G. R., HUFF, W. E. & BALOG, J. M. (2000) Factors regulating bone maturity and strength in poultry. Poultry Science, 79, 1024-1032. 10 rR 25 SANOTRA, G. S., LUND, J. D., ERSBOLL, A. K., PETERSEN, J. S. & VESTERGAARD, K. S. (2001) Monitoring leg problems in broilers: A survey of commercial broiler production in Denmark. World's Poultry Science Journal, 57, 55-69. ee 20 REITER, K. & BESSEI, W. (1996) Effect of the distance between feeder and drinker on behaviour and leg disorders of broilers. Proceedings of the 30th International Congress of the International Society for Applied Ethology. Guelph, Canada. rP 15 REITER, K. & BESSEI, W. (1995) Influence of running on leg weakness of slow and fast growing broilers. Proceedings of the 29th International Congress of the International Society of Applied Ethology. Exeter, U.K. Fo SORENSEN, P., SU, G. & KESTIN, S. C. (1999) The effect of photoperiod:scotoperiod on leg weakness in broiler chickens. Poultry Science, 78, 336-342. ev 30 SORENSEN, P., SU, G. & KESTIN, S. C. (2000) Effects of age and stocking density on leg weakness in broiler chickens. Poultry Science, 79, 864870. iew SU, G., SORENSEN, P. & KESTIN, S. C. (1999) Meal feeding is more effective than early feed restriction at reducing the prevalence of leg weakness in broiler chickens. Poultry Science, 78, 949-955. 35 On THORP, B. H. & DUFF, S. R. I. (1988) Effect of exercise on the vascular pattern in the bone extremities of broiler fowl. Research in Veterinary Science, 45, 72-77. 40 45 50 ly 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 THORP, B. H. & WADDINGTON, D. (1997) Relationships between the bone pathologies, ash and mineral content of long bones in 35-day-old broiler chickens. Research in Veterinary Science, 62, 67-73. VESTERGAARD, K. S. & SANOTRA, G. S. (1999) Relationships between leg disorders and changes in the behaviour of broiler chickens. Veterinary Record, 144, 205-209. WEEKS, C. A., DANBURY, T. C., DAVIES, H. C., HUNT, P. & KESTIN, S. C. (2000) The behaviour of broiler chickens and its modification by lameness. Applied Animal Behaviour Science, 67, 111-125. E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps British Poultry Science Page 19 of 26 WILLIAMS, B., SOLOMON, S., WADDINGTON, D., THORP, B. & FARQUHARSON, C. (2000) Skeletal development in the meat-type chicken. British Poultry Science, 41, 141-149. 5 WILLIAMS, B., WADDINGTON, D., MURRAY, D. H. & FARQUHARSON, C. (2004) Bone strength during growth: Influence of growth rate on cortical porosity and mineralization. Calcified Tissue International, 74, 236-245. 10 ZHANG, B. & COON, C. N. (1997) The relationship of various tibia bone measurements in hens. Poultry Science, 76, 1698-1701. Fo iew ev rR ee rP ly On 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps British Poultry Science Table 1. Prevalence of two measures of leg health of focal birds from batches 1-3 (n=70) at 6 weeks of age against light treatments Light treatment Control Step change 30 32 (43%) (46%) 6 2 (9%) (3%) 12 10 (17%) (14%) 9 17 (13%) (24%) 15 7 (21%) (10%) No Rotation of tibiotarsus P = 0.149 Hock burns (Increasing severity) P = 0.064 Fo 5 Yes 0 (no discolouration or lesions) 1 (reddening) 2 (scabbing) iew ev rR ee rP ly On 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 20 of 26 E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 21 of 26 Table 2. Mean hourly activity across focal birds at each age (n = 24) 5 Age of birds (weeks) 2 4 6 Mean hourly activity (grids crossed) (+/- 95% CI) 163.5 ± 16.77 77.0 ± 10.57 53.6 ± 7.55 Std. deviation 39.72 25.04 17.89 Fo iew ev rR ee rP ly On 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 British Poultry Science E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps British Poultry Science Table 3. Descriptive data from right (RL) and left (LL) legs of Batch 3 focal birds (n=23) Minimum Maximum 1017.90 973.47 22.30 RL horizontal plane moment of inertia (mm4) 97.74 294.24 168.99 45.89 126.17 288.51 196.55 46.70 LL cortical density (mg/cc) 895.52 1072.08 968.98 38.03 LL horizontal plane moment of inertia (mm4) 68.47 295.56 172.91 46.95 314.81 196.44 47.94 1980 317.0 1055 2595 iew ly On 15 101.28 ev Weight (g) rR ee rP LL vertical plane moment of inertia (mm4) 10 Std. Deviation 938.92 RL vertical plane moment of inertia (mm4) 5 Mean RL cortical density (mg/cc) Fo 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 22 of 26 20 25 E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 23 of 26 Figure 1. Step changes employed for illuminance. 200 10 Fo 15 3h 24 10 20 0 Time of day 21.00 20.30 03.30 iew ev rR ee rP Illuminance (lux) 5 03.00 ly On 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 British Poultry Science E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps British Poultry Science Figure 2. Relationship between mean weight and gait score at all ages (weeks 2 to 6). 2.5 n = 20 n = 20 n = 165 1.5 n = 145 1 0.5 0 ee rP Mean weight (kg) 2 Fo 0 1 rR 2 3 Gait Score (0 = sound, 3 = poor) Error Bars: 95% CI iew ev ly On 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 24 of 26 E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps Page 25 of 26 Figure 3. Decline in median activity of focal birds in Batch 3 (n=24) as age increases for each light treatment. 5 Light treatment ● control 10 ○ step change Fo 15 30 iew ev rR 25 ee 20 rP ly On 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 British Poultry Science E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps British Poultry Science Figure 4. The interaction between age and time in relation to median activity of focal birds in Batch 3 (n=24). 5 Age (weeks) 400 10 15 4 300 6 200 100 20 04.00-05.00 09.00-10.00 14.00-15.00 Time (Error Bars: 95% CI) iew ev rR 25 ee 0 rP Median activity (grids crossed per hour) 2 Fo ly On 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 26 of 26 E-mail: br.poultsci@bbsrc.ac.uk URL: http://mc.manuscriptcentral.com/cbps