Animal- and management-based welfare indicators for a conventional broiler strain in 2 barn types (Louisiana barn and closed barn)

Animal- and management-based welfare indicators for a conventional broiler strain in 2 barn types (Louisiana barn and closed barn) Helen Louton,∗,1 Shana Bergmann,∗ Sven Reese,† Michael Erhard,∗ Josef Bachmeier,‡ Beatrice R¨osler,∗ and Elke Rauch∗ ∗

Department of Veterinary Sciences, Chair of Animal Welfare, Ethology, Animal Hygiene and Animal Husbandry, Faculty of Veterinary Medicine, LMU Munich, 80539 Munich, Germany; † Department of Veterinary Sciences, Chair of Anatomy, Histology and Embryology, Faculty of Veterinary Medicine, LMU Munich, 80539 uterei S¨ ud ZN of the BWE-Br¨ uterei Weser-Ems GmbH & Co. KG, 93128 Regenstauf, Munich, Germany; and ‡ Br¨ Germany ABSTRACT The aim of this observational study was to describe health- and management-related welfare indicators for a conventional broiler strain housed in 2 barn types (Louisiana barn and closed barn) on the same farm and to assess the impact of age and housing conditions on these indicators. Broilers were examined 4 times in each of 7 fattening periods. Their weight gain, gait score, and further animal-based indicators did not differ between the barn types. On average 46% showed no sign, 51% a minor but visible, and 3% a profound lameness at the end of the fattening period (fattening day 39). Soiling of the plumage, skin scratches, foot pad dermatitis, and hock burns worsened with increasing age. Soiling of the plumage was correlated negatively with litter depth (Pearson, r = −0.549, P = 0.042) and positively with litter quality (Pearson, r = 0.641, P = 0.013). Skin scratches occurred in 89% of the broilers (32% deep with penetration of dermis) on fattening day 39 and were correlated posi-

tively with cumulative mortality (Pearson, r = 0.615, P = 0.019), indicating a severe animal welfare impact. Foot pad dermatitis did not correlate with health- or management-related indicators, whereas hock burn correlated positively with the broilers’ weight (Pearson, r = 0.853, P < 0.001) and with ammonia concentrations (Pearson, r = 0.577, P = 0.031). Management-related indicators (antibiotic treatments, dead on arrival) and cumulative mortality did not differ between the barn types. At the end of the fattening period, the litter quality was worse and concentrations of ammonia and peaks of high carbon dioxide concentrations were higher in the Louisiana than in the closed barn. The light intensity was on average 20 times higher in the Louisiana than in the closed barn without any negative impact. Summarizing, the barn type did not seem to influence the investigated welfare indicators, but Louisiana barns might need a more precise management to maintain the required ranges of noxious gases and litter quality.

Key words: welfare indicator, animal-based measure, broiler, barn type, Louisiana barn 2018 Poultry Science 0:1–14 http://dx.doi.org/10.3382/ps/pey111

INTRODUCTION

To assess poultry welfare on-farm, authors proposed the use of animal-based measures (Forkman and Keeling, 2009; Butterworth, 2013). Resource- and management-based measures can supplement the use of animal-based measures (Forkman and Keeling, 2009). Animal-based indicators are measures that can be assessed directly on the animal. They represent characteristic traits with respect to animal health and behavior (Zapf et al., 2015). Using these indicators, it is possible to draw a conclusion on the welfare of poultry on-farm and reveal pain, suffering, or injuries of the animals. According to Arnould et al. (2009), breast blisters, the condition of the feet, and hock burn are valid parameters to evaluate welfare in broilers. These authors state that wounds (scratches) are rare and the measurements are not validated. The cleanliness of the body could also be a good indicator of the quality of

Several factors of housing and management are known to influence the health of broilers. According to the German Order on the Protection of Animals and the Keeping of Production Animals (2006), a maximum stocking density of 39 kg/m2 for the housing of broilers is permitted, whereby in the European Union, the defined maximum stocking density of 39 kg/m2 may be increased by 3 kg/m2 if certain criteria are fulfilled (Council of the European Union, 2007). The maximum stocking density of the “BEST Chicken Standard” (2014) is 35 kg/m2 .  C 2018 Poultry Science Association Inc. Received October 25, 2017. Accepted April 19, 2018. 1 Corresponding author: [email protected]

1 Downloaded from https://academic.oup.com/ps/advance-article-abstract/doi/10.3382/ps/pey111/4995452 by University of California, Santa Barbara user on 15 May 2018

2

LOUTON ET AL.

management, although this indicator has not been evaluated scientifically. Plumage has protective effects on both the skin and the animal’s temperature regulation. If the plumage is soiled, these effects are lost, and consequently the welR fare of the broilers is affected (Welfare Quality , 2009). According to several authors, the soiling of the plumage may be influenced by the body weight of the broiler (Rauch et al., 2017), the health of the foot pads (De Jong et al., 2014; Saraiva et al., 2016), and the quality of the litter (Saraiva et al., 2016; Rauch et al., 2017). Further important animal health indicators are scratches. Cellulitis as a result of scratches is the main reason for condemnation at slaughter (Elfadil et al., 1996). According to the literature, one major aspect influencing the prevalence and severity of scratches is the stocking density (Harris et al., 1978; Dozier et al., 2005; Allain et al., 2009). According to Aviagen (2014a), “scabbing” can be mitigated by preventing excessive bird activity, preventing overcrowding, maintaining correct feeder and drinker space, and providing access to feed at all times. Weise (2007) mentioned high light intensities as a factor that can stimulate the activity of the broilers and the development of skin scratches. Other commonly observed health problems in broilers are foot pad dermatitis (FPD), hock burn lesions, and changes in gait. McKeegan (2010) noted that the mechanism for the development of hock burn lesions differs from that of FPD in that hock burn lesions are related more to sex and weight and less to litter characteristics and quality. However, both types of lesions can be induced by wet litter and high air ammonia concentrations. Furthermore, McKeegan (2010) stated that these lesions are potentially painful and that affected birds walk more slowly than healthy birds. Clinical features of lameness are commonly associated with lesions of the skeleton (McNamee et al., 1998). In addition, authors reported an influence of high body weights (Kestin et al., 1992; Bergmann et al., 2016), stocking density (Sanotra et al., 2001; Knowles et al., 2008), feeding regime, genotype, and age (Kestin et al., 2001) on the deterioration of gait scores. A gait score of 3, which is defined by “visible lameness of one extremity,” is considered painful (Sørensen et al., 2000). Environmental conditions may also affect the wellbeing of broilers (Bergmann et al., 2016). Although 15% of the broilers in Germany were reared in Louisiana barns in the year 2000, no scientific research results have been published about the influence of this housing system on broiler health. Louisiana barns are opensided and use natural ventilation that is controlled automatically by a curtain system (Scientific Committee on Animal Health and Animal Welfare, 2000). This system has the advantage of lower construction and maintenance costs as compared with forced ventilation systems (Damme and Hildebrand, 2002). However, the negative side effect is a possible insufficient fresh air supply, especially at high outside temperatures in combination with high humidity and calm wind,

because this ventilation system depends on the weather conditions (Richter, 2006). Meluzzi and Sirri (2009) described that the housing conditions (litter quality, temperature, and humidity) are important factors influencing the growth rate of broiler. According to the authors, ventilation rate is an important parameter to reduce the negative effect of high stocking densities and wet litter, which in turn directly influences the air quality (dust, ammonia, and humidity). Our main goal with this observational study was to investigate the prevalence of animal- and managementbased welfare indicators in a fast-growing broiler strain. We examined a possible age-related development of these indicators and the potential impact of housing conditions on the animal-based indicators at a target stocking density of 35 kg/m2 . We expected the results to show which factors might have an influence on animal health and to set the ground for future confirmatory experiments. This study was conducted in 2 barn types (Louisiana barn and closed barn) to compare the described animal- and management-based indicators between these barn types.

MATERIALS AND METHODS Animals and Barns This study was performed as part of the “BEST Chicken” (BEST = B¨ undnis f¨ ur Exzellenz, Sicherheit und Transparenz; Alliance for Excellence, Safety and Transparency) research project. Broilers of the conventional strain of slow-feathering Ross 308 were housed in 2 barns on 1 farm and examined over 7 fattening periods. The broilers from each barn hatched on the same day and were housed at the same time. The parental stocks were from the same breeder and housing. However, it was not possible to receive chicks from the same parental stock in both barns in all fattening periods. The production week of the parental stock was documented for each fattening period. The broilers were supplied by Br¨ uterei S¨ ud ZN of the BWE-Br¨ uterei Weser-Ems GmbH & Co. KG, Regenstauf, Germany, transported by Spedition D-LOG GmbH, Burglengenfeld, Germany, and housed from fattening day 0 (FD 0) onward in these 2 barns. The barns differed in the base area and the total number of broilers. The main differences were the ventilation system and the ratio between window area and ground area. The window-to-ground area ratio is calculated as follows: [(window area:ground area) × 100]. The resulting percentage is a reference of the incidence of natural light in the barn and must be at least 3% in Germany, according to legal regulations (German Order on the Protection of Animals and the Keeping of Production Animals, 2006). Barn 1 was a Louisiana barn with unrestricted ventilation, a ground area of 1554 m2 , a capacity for 28,100 broilers, and a windowto-ground area ratio of 19.2%. Barn 2 was a closed barn with forced ventilation, a ground area of 2340 m2 ,

Downloaded from https://academic.oup.com/ps/advance-article-abstract/doi/10.3382/ps/pey111/4995452 by University of California, Santa Barbara user on 15 May 2018

ANIMAL- AND MANAGEMENT-BASED INDICATORS FOR BROILERS

and a capacity for 42,100 broilers. The window-toground area ratio was 5.1%. The total ventilation rate was 5.02 m3 /kg live weight per hour in Barn 2. The Louisiana barn was mainly managed without additional ventilation. Additional ventilation systems were used in summer with additional ventilation systems in the gable of the roof and opening of the gates of the barn. The additional artificial lighting was enabled by 19 dimmable fluorescent tubes with an output of 9.22 W/m2 in Barn 1 and 18 tubes with an output of 8.70 W/m2 in Barn 2. Daily photoperiod was performed with 23 h of artificial light and 1 h of darkness on the first 3 d after housing, then 18 h of artificial light and 6 h of darkness during the night from d 4 until 3 d before slaughter. This light regime was in accordance with the German Order on the Protection of Animals and the Keeping of Production Animals (2006), and exceptions occurred only due to veterinary indication. Both barns were equipped with spray cooling systems to keep the temperature in the desired range. Three days before housing the chicks, the barns were heated to a base plate temperature of 30◦ C with an air temperature of 33 to 34◦ C. Both barns were heated via a heat exchange system providing waste heat from a biogas plant. Furthermore, both barns had a sluice for change of clothing. Between the fattening periods and before the housing of chicks, the entire barns were mucked out, washed, and disinfected. Pelletized straw was used as litter material (1 kg litter material per m2 in both barns), and no change of bedding took place during a fattening period, but litter was additionally dispersed if required. The feed (MEGA Tierern¨ ahrung GmbH & Co. KG, Visbek-Rechterfeld, Germany) was provided during the respective age phases of the broilers with “Starter Flour,” “Pelleted Starter,” “Breeding Feed I Slow Growth,” “Breeding Feed II Slow Growth,” and “Finisher Feed SG.” Main components of the feed were corn, wheat, soy bean meal extract, and soy bean oil. Deposits of colza meal extract, wheat bran, colza cake, sunflower meal extract, and maize gluten were possible. Feed was supplied ad libitum in round feeders (LAE Anlagenbau GmbH, Cuxhaven, Germany), which were continuously adapted to the height of the chicks. The chicks had a space of 0.74 cm feeder per kg live weight in Barn 1 and 0.88 cm feeder per kg live weight in Barn 2. Water was provided ad libitum in nipple drinkers (LUBING Maschinenfabrik Ludwig Bening GmbH & Co. KG, Barnstorf, Germany; installed by LAE Anlagenbau GmbH, Cuxhaven, Germany). One nipple drinker was provided for 14 broilers in Barn 1 and for 11 broilers in Barn 2. The chicks were vaccinated according to the BEST Chicken standard with the first vaccination against infectious bronchitis on FD 0 in the hatchery with the vaccine Poulvac IB-Primer (Zoetis Deutschland GmbH, Berlin, Germany). The second vaccination was performed on FD 12 against Newcastle disease and Gumboro (infectious bursitis) orally via the drinking

3

system. On FD 18, the chicks received another vaccination against infectious bronchitis.

Methods of Assessment The day of housing was defined as FD 0. Animal health was assessed 4 times during the fattening period according to the following schedule: first assessment day (AD 1) on FD 5, second assessment day (AD 2) on FD 15, third assessment day (AD 3, FD 32) 2 d before thinning, and fourth assessment day (AD 4, FD 39) 2 d before final slaughter. Thinning is a measure, performed approximately 1 wk before final slaughter, to reduce the stocking density. Approximately one-third of the birds are taken out of the barn for slaughter, whereas the rest stay to grow further. In the presented study, a body weight of 1850 g on FD 31 to FD 33 and of 2500 g on FD 38 to FD 40 was defined as target weights. These target weights were expected to vary slightly between the fattening periods. A stocking density of 35 kg/m2 was defined as target density (according to the BEST Chicken Standard). The target stocking densities were calculated incorporating the expected mortality and weight gain of the broilers. For each assessment, generally 100 animals were examined with respect to several health indicators (exception: first and second fattening periods, only 50 animals per barn due to organizational structure). The broilers were randomly picked from the different areas of the barn. To prevent an unbalanced sampling, the catching was performed during dimmed lights and the same investigators performed the health examination on each assessment day. Based on a detailed assessment sheet, the health status of every broiler was documented and deviations from the physiological status were noted. Respective scores for the assessed health parameters are presented in Table 1. After catching, the broilers were placed into a small ring in the barn and taken from there for the examination. On AD 4, the broilers were first motivated to walk about 1 m from an opening in the ring so the assessor could examine the gait; afterwards, the rest of the animal health parameters were recorded. The weight of the broilers was recorded with a Valor 2000 scale (Ohaus Corporation, Parsippany, New Jersey, USA, and Soehnle [Leifheit AG], Nassau, Germany). The climatic conditions were recorded at every assessment (AD 1 to AD 4). In each barn, 3 locations (front, middle, back) were defined and an imaginary transverse line was drawn across the barn. On this line, 13 (Barn 1) or 15 points (Barn 2) were selected for the measurement of the climatic conditions, resulting in total numbers of 39 (Barn 1) and 45 (Barn 2) measurement points. These points included different functional areas (resting, feeding, and drinking). At these points, the parameters litter depth, litter quality, and a snap shot of ammonia were measured. Additionally, at 3

Downloaded from https://academic.oup.com/ps/advance-article-abstract/doi/10.3382/ps/pey111/4995452 by University of California, Santa Barbara user on 15 May 2018

4

LOUTON ET AL.

Table 1. Assessed animal health parameters and scores corresponding to observed changes. Small area: <5-mm wide; large area: ≥5-mm wide. Soiling of plumage

Score

Skin scratches

Foot pad dermatitis (revised, R according to Welfare Quality , 2009)

Hock burn (revised, according R to Welfare Quality , 2009)

Gait score (revised, according R to Welfare Quality , 2009) Normal gait, dexterous and agile, chicken-typical gait Slight abnormality, but difficult to define Slurred lameness, but no extremity identifiable

0

Plumage clean

None

No lesion

No lesion

1

Mild soiling (only ventral) Moderate soiling (ventral and dorsal) Severe soiling (complete body)

Mild scratches (superficial) Moderate scratches (dermis penetrated, 1-sided) Severe scratches (dermis penetrated, 2-sided) n/a n/a

Superficial lesion, small area

Redness

Superficial lesion, large area

Superficial lesion, small area

Profound lesion, small area

Superficial lesion, large area

Distinct lameness, extremity is identifiable

Profound lesion, large area n/a

Profound lesion, small area Profound lesion, large area

Broiler walks only a few steps Incapable of walking

2 3 4 5

n/a n/a n/a = not applicable.

locations across each transverse line, that means 9 points in each barn, the light intensity was recorded. The litter depth (cm) was measured using a yard stick. For the evaluation of the litter quality, a 5-staged rankR ing system according to the Welfare Quality assessment protocol for poultry was used with the following scores: “Score 0 = completely dry and flaky, moves easily with the foot; Score 1 = dry but not easy to move with foot; Score 2 = leaves imprint of foot and will form a ball if compacted, but ball does not stay together; Score 3 = sticks to boots and sticks readily in a ball if compacted; Score 4 = sticks to boots once the cap or compacted crust is broken” (Welfare R Quality , 2009). A snap shot measurement of gaseous ammonia was done on the height of the chicks’ head at each assessment day with 2 devices of Altair Pro SingleGas Detector-Ammonia, (MSA, Cranberry Township, Pennsylvania, USA). These devices measure with a precision of ±1 ppm from a threshold value of 5 ppm ammonia onwards. The light intensity was measured using a Pocket-Lux 2 (LMT Lichtmesstechnik GmbH, Berlin, Germany) based on a 6-sided measurement. It was recorded on AD 1 and AD 2. On AD 3 and AD 4, the light intensity was not measured because the broilers were examined under dimmed light to prevent stressful reactions. Additionally, a continuous measurement of gaseous ammonia and carbon dioxide at the body height of the broilers was conducted in the middle of Barn 1 (as of the third fattening period) and Barn 2 (all fattening periods) with a specially developed “climate case” (Berndt Messtechnik GmbH, Munich, Germany). These cases have a detector for gaseous ammonia of the type PrimaX (MSA, Cranberry Township, Pennsylvania, USA; measurement range to 100 ppm, precision of measurements ±2 ppm) and a detector for carbon dioxide of the type Guardian Plus (Edinburgh Instruments Ltd., Livingston, UK; measurement range to 1.00%, precision of measurements ±0.02 Vol.%). A control unit of the type 9020 LCD (MSA, Cranberry Township, Pennsylvania, USA) converts the data of the

sensors into a digital signal. The data were recorded at 5-min intervals and stored in an Ecograph T, RSG35 (Endress + Hauser Messtechnik GmbH & Co. KG, Munich, Germany). For the evaluation of gas concentrations, averages of 5 d before the respective assessment day were calculated.

Statistical Analysis The statistical analysis was performed using IBM SPSS Statistics 24.0 software (IBM Deutschland GmbH, Ehningen, Germany). To evaluate the effect of assessment day (representing broiler age) on animaland management-based indicators, an analysis of variance was performed and a post-hoc test according to Dunnet was used. To present relationships between factors related to animal condition and factors related to barn management, the animals’ condition at the fourth assessment was correlated with the barn management on this assessment day. Exceptions from this procedure were the factors “assessment day,” “stocking density at all ages,” and “light intensity.” The latter was correlated with the health condition at the first and second assessments because only on these days was light intensity measured. To evaluate correlations, Spearman correlations were used for non-normally distributed data as well as ranked categorical data and Pearson correlations for normally distributed data. The interpretation of the degree of correlation is defined as follows: 0.00 to ≤0.10 = negligible, 0.11 to ≤0.20 = weak, 0.21 to ≤0.40 = moderate, 0.41 to ≤0.60 = relatively strong, 0.61 to ≤0.80 = strong, and 0.81 to ≤1.00 = very strong (Rea and Parker, 2014). To compare the barn types, a t test was used for normally distributed data and a MannWhitney U test for non-normally distributed data. The Shapiro–Wilk test was used to divide metric variables in normally distributed data and non-normally distributed data. A P-value <0.05 was determined as statistically significant.

Downloaded from https://academic.oup.com/ps/advance-article-abstract/doi/10.3382/ps/pey111/4995452 by University of California, Santa Barbara user on 15 May 2018

5

Downloaded from https://academic.oup.com/ps/advance-article-abstract/doi/10.3382/ps/pey111/4995452 by University of California, Santa Barbara user on 15 May 2018

0.529

n/a

n/a

n/a

n/a n/a n/a n/a n/a n/a 0.08 0.13 0.927

0.424

0.557

n/a n/a n/a n/a n/a n/a 1.62 1.58 0.066

0.13 0.20 0.14 0.24 0.21 0.18 0.34 0.50 0.749

0.564

0.248

0.35 0.28 0.27 0.33 0.92 1.05 1.42 1.44 0.073

0.01 0 0.06 0.15 0.07 0.58 0.22 0.64 0.398

0.942

0.388

0.01 0 0.06 0.16 0.25 0.45 0.44 0.72 0.662

0.932

4

3

n/a = not assessed. 1 statistical analysis with a t test for normally distributed data. 2 statistical analysis with a Mann-Whitney U test for non-normally distributed data.

0.632

0.392

0.206 0.816 2

P-value1 SD

Gait

Average P-value1 SD

Hock burn

Average P-value2 SD

Foot pad dermatitis

Average P-value1

n/a

0 0 0.09 0.08 0.28 0.31 0.17 0.27 0 0 0.13 0.09 1.27 1.28 1.19 1.08 0.643

0.22 0.37 0.09 0.12 0.13 0.21 0.20 0.17 0.36 0.44 0.36 0.44 0.43 0.51 0.70 0.75 0.645

20.5 18.1 77.4 83.4 121.8 99.2 131.5 174.0 111 107 511 501 1,912 1,906 2,499 2,536 1 2 1 2 1 2 1 2

With increasing age, an increasing percentage of animals were affected by changes in animal-based indicators such as soiling of the plumage, skin scratches, FPD, and hock burn. The weight gain of the broilers corresponded to the target weight and did not differ between Barns 1 and 2 (Table 3). Throughout the different fattening periods, the weight gain differed but was similar in the 2 barns in each fattening period (Table S1). The broilers reached an average weight of 1912 g (Barn 1) and 1906 g (Barn 2) by AD 3 and 2499 g (Barn 1) and 2536 g (Barn 2) by AD 4. The average weight gain per day, calculated for the 5 d before each assessment day (recorded and provided by the farmer), was on average similar between the barns (Tables 4 and S2). At the beginning of the fattening period (AD 1 = FD 5), the average weight gain was 12 g/d (Barns 1 and 2). Toward the end of the fattening period before thinning (AD 3), the weight gain was 86 g/d (Barn 1) and 90 g/d (Barn 2), and before final slaughter (AD 4), it was 88 g/d (Barn 1) and 85 g/d (Barn 2). The average weight gain within the average of 38 fattening days was 65 g/d (Barn 1) and 66 g/d (Barn 2). The weight gain differed between all 4 assessment days (post-hoc tests, Dunnet, P < 0.001). The weight

1

Animal- and Management-Based Indicators (Age-Related Prevalence and Severity)

SD

The stocking density of 35 kg/m2 was defined as target density (according to the BEST Chicken standard). However, due to the individual weight development and unpredictable mortality, the actual (effective) stocking density could slightly differ during the fattening periods. These variations are presented in Table 2 as “effective stocking densities.” The production weeks of the parental stocks are presented in Table 2 and ranged from production wk 9 to 27. The results are presented with averages of the 7 fattening periods. Detailed results from each of the fattening periods can be found in the Supplementary Tables posted online. These tables will be cited as Tables S1 to S6.

Skin scratches

RESULTS

Average

14 22 15 22 20 25 9

P-value1

15 23 16 23 20 27 9

SD

31.5 33.8 35.8 34.7 35.7 33.7 26.9

Soiling of plumage

32.2 34.4 35.6 35.6 34.0 35.5 31.4

Average

Barn 2

P-value1

Barn 1

SD

Barn 2

Weight (g)

1 2 3 4 5 6 7

Barn 1

Average

Fattening period

Production week of parental stocks

Barn

Effective stocking density (kg/m2 )

AD

Table 2. Deviations from the target stocking density of 35 kg/m2 due to weight increase and mortality resulting in “effective stocking densities” and the production week of the parental stock in the 7 fattening periods.

Table 3. Animal-based indicators (average score) in the 2 barns (1 = Louisiana barn, 2 = closed barn) on assessment days (AD) 1 to 4 (on average: AD 1 = fattening day [FD] 5, AD 2 = FD 15, AD 3 = FD 32, AD 4 = FD 39).

ANIMAL- AND MANAGEMENT-BASED INDICATORS FOR BROILERS

6

LOUTON ET AL.

Table 4. Animal-based indicators in the 2 barns (1 = Louisiana barn, 2 = closed barn) on assessment days (AD) 1 to 4 (on average: AD 1 = fattening day [FD] 5, AD 2 = FD 15, AD 3 = FD 32, AD 4 = FD 39). Data provided by farmer and summarized as the average of the 5 d before each AD. Selection rate due to leg weakness (%) AD 1 2 3 4

Average daily mortality (%)

Average daily weight gain (g)

Barn

Average

SD

P-value1,2

Average

SD

P-value1,2

Average

SD

P-value3

1 2 1 2 1 2 1 2

0 0 0.02 0.03 0.03 0.03 0.01 0.02

0 0 0.01 0.01 0.04 0.04 0.01 0.03

0.530

0.12 0.10 0.08 0.09 0.13 0.11 0.19 0.18

0.06 0.02 0.04 0.02 0.07 0.07 0.02 0.12

0.896

12.18 11.49 40.51 41.09 85.97 89.72 88.27 84.80

2.35 2.04 6.09 6.92 3.36 8.65 9.31 11.49

0.567

0.619 0.462 0.430

0.865 0.441 0.024

0.872 0.316 0.670

1

statistical analysis with a t test for normally distributed data on AD 1, 3, and 4. statistical analysis with a Mann-Whitney U test for non-normally distributed data on AD 2. 3 statistical analysis with a t test for normally distributed data on all AD. 2

of the broilers on AD 4 correlated positively with the stocking density on AD 4 (weight per m2 ) (Pearson, r = 0.760, P = 0.002) and with the production week of the parental stock, i.e., the age of the parents in the respective fattening period of the chicks on AD 4 (Pearson, r = 0.666, P = 0.009). With the ongoing assessment period, soiling of the plumage was observed in an increased percentage of broilers. Most plumage soiling was identified as mild soiling (AD 4: 62.3%), whereas moderate (AD 4: 3.9%) and severe (AD 4: 0.2%) soiling were found less frequently. The average score of plumage soiling on AD 4 was 0.70 in Barn 1 and 0.75 in Barn 2 (Tables 3 and S1). High and low scores of plumage soiling were observed in the examined seasons, e.g., Barn 1 had average scores of 0.90 on AD 4 in the fattening period during the summer (fattening period 2) and 0.47 in the fattening period summer–autumn (fattening period 3) (Table S1). The assessment day influenced the prevalence of plumage soiling, and more broilers with soiled plumage were observed on AD 4 than on AD 1 (post-hoc test, Dunnet, P < 0.001). The prevalence of plumage soiling on AD 4 correlated negatively with the litter depth (Pearson, r = −0.549, P = 0.042) and positively with the litter score (Pearson, r = 0.641, P = 0.013). After an initial increase in the severity of skin scratches up to the third assessment (AD 3), less severe skin scratches were noted at the fourth assessment (average score of skin scratches on AD 3: 1.27 in Barn 1 and 1.28 in Barn 2; on AD 4: 1.19 in Barn 1 and 1.08 in Barn 2; Tables 3 and S1). On AD 1 (FD 5), no skin scratches were detected. On AD 2 (FD 15), broilers with scratches of Score 2 were observed sporadically (0.6%). However, on AD 3 (FD 31 to FD 33, 2 d before thinning), 6.2% of the broilers were observed with severe 2-sided scratches with penetration of the dermis (Score 3) and 32.7% with 1-sided scratches with penetration of the dermis (Score 2). The assessment day influenced the prevalence of skin scratches, with more broilers showing skin scratches on AD 3 and AD 4 than on AD 1 (post-hoc tests, Dunnet, P < 0.001). The prevalence of skin scratches on AD 4 correlated positively with the

cumulative mortality rate throughout the fattening period (Pearson, r = 0.615, P = 0.019). In fattening period 4 (autumn), the highest average scores of scratches were observed in both barns on AD 3 and AD 4 (Table S1). The development of FPD over time as an average score for each assessment day is presented in Tables 3 and S1. The foot pads showed increasing impairment over time in both barns, but the development of FPD was more severe in Barn 2 (the closed barn) than in Barn 1 (average score increase from 0.0 on AD 1 to 0.4 on AD 4 in Barn 1 and from 0.0 on AD 1 to 0.7 on AD 4 in Barn 2; Table 3). The broilers in Barn 1 were affected by FPD during more fattening periods than those in Barn 2. However, when FPD was recorded in the broilers of Barn 2, the percentage of affected birds was always larger than that in Barn 1. The assessment day influenced the severity and prevalence of FPD, and more broilers were observed with FPD on AD 3 and AD 4 than on AD 1 (post-hoc tests, Dunnet, P = 0.019 for AD 1 vs. AD 3; P < 0.001 for AD 1 vs. AD 4). The prevalence of FPD on AD 4 did not correlate with any of the animal- or management-based indicators on AD 4. As with the other health parameters, the hock burn changes showed an increased severity with progressing age and differences in different fattening periods, but the assessed scores were similar in both barns (average score increase from 0.4 on AD 1 to 1.4 on AD 4 in Barn 1 and from 0.3 on AD 1 to 1.4 on AD 4 in Barn 2; Tables 3 and S1). Hock burns of Score 2 (superficial, small area) were already observed at the first assessment. In several fattening periods (Barn 1: fifth, sixth, seventh; Barn 2: fourth, fifth, sixth; Table S1), broilers with hock burn of Score 3 (superficial, large area) were observed at the beginning of the fattening period (on FD 5). Hock burns of Score 5 were rare and seen only at the fourth assessment (0.2%). Thus, the assessment day affected the prevalence and severity of hock burns, and more broilers showed hock burn on AD 3 and AD 4 than on AD 1 (post-hoc tests, Dunnet, P < 0.001). The prevalence of hock burn on AD 4 correlated positively with the weight of the broilers (Pearson, r = 0.853,

Downloaded from https://academic.oup.com/ps/advance-article-abstract/doi/10.3382/ps/pey111/4995452 by University of California, Santa Barbara user on 15 May 2018

7

ANIMAL- AND MANAGEMENT-BASED INDICATORS FOR BROILERS

Table 5. Climatic, management-based indicators in the 2 barns (1 = Louisiana barn, 2 = closed barn) on assessment days (AD) 1 to 4 (on average: AD 1 = fattening day [FD] 5, AD 2 = FD 15, AD 3 = FD 32, AD 4 = FD 39). Due to technical reasons, light intensity was only measured on AD 1 and AD 2. Litter score, according to Welfare R Quality (2009)

Litter depth (cm) AD 1 2 3 4

Light intensity (lx)

Barn

Average

SD

P-value1

Average

SD

P-value1

Average

SD

P-value2

1 2 1 2 1 2 1 2

0.63 0.52 1.26 1.13 2.97 3.09 3.48 3.56

0.06 0.03 0.09 0.09 0.26 0.26 0.23 0.30

0.003

0.68 0.94 1.69 1.99 2.38 2.47 2.40 2.32

0.17 0.23 0.35 0.40 0.30 0.19 0.30 0.27

0.034

372.95 24.73 576.27 21.69 n/a n/a n/a n/a

457.82 5.15 824.23 6.24 n/a n/a n/a n/a

0.003

0.024 0.385 0.582

0.165 0.498 0.605

0.025

1

statistical analysis with a t test for normally distributed data. statistical analysis with a Mann-Whitney U test for non-normally distributed data. n/a = not assessed. 2

P < 0.001), the stocking density (weight per m2 ) (Pearson, r = 0.711, P = 0.004), and the average gaseous ammonia concentration on AD 4 (Pearson, r = 0.577, P = 0.031), and with the production week of the parental stock on AD 4 of the respective fattening period (Pearson, r = 0.617, P = 0.019). The gait scores, which were only assessed on AD 4, were similar in the 2 barns and constant within the fattening periods (Barn 1: average = 1.62, SD = 0.08; Barn 2: average = 1.58, SD = 0.13; Tables 3 and S1). In all fattening periods, broilers with Score 0 (normal gait, dexterous and agile, and chicken-typical gait) were noted in both barns, with an average of 11.6%. The majority (34.3%) of broilers were observed with a “slight abnormality, but difficult to define” (Score 1). Lameness, either as a slurred lameness but no limb identifiable (Score 2) or as a distinct lameness with limb identifiable (Score 3) was observed in 41.1 or 10.0% of the broilers, respectively. Differences between the barns were observed in terms of severe gait changes and the inability to move. Broilers were assessed with a gait score of 5 (incapable of walking) in the second, third, sixth, and seventh fattening periods in Barn 1 and in the third, fourth, and seventh fattening periods in Barn 2. On average 1.0% of the broilers were assessed with Score 5. Gait score 4 (broiler walks only a few steps) was found in broilers during almost every fattening period in each barn, with an average of 2.0% of the broilers being affected. The only correlation of the gait score was found with the litter score on AD 4 (Pearson, r = 0.545, P = 0.044). Management-based indicators such as litter depth, litter score, gaseous ammonia and carbon dioxide concentrations, temperature, and humidity differed between the assessment days (Tables 5, 6, S3, and S4). The litter depth and litter score differed between AD 1 and all other assessment days (post-hoc tests, Dunnet, P < 0.001). The continuous measurements of gaseous ammonia and carbon dioxide revealed that the concentration of ammonia was higher on AD 4 than on AD 1 (post-hoc test, Dunnet, P = 0.002). The concentration of carbon dioxide was higher on AD 2, AD 3,

and AD 4 than on AD 1 (post-hoc tests, Dunnet, P ≤ 0.05). Gaseous ammonia and carbon dioxide concentrations differed between the fattening periods, with higher concentrations during the autumn and winter time (Table S4). The continuous measurements of gaseous ammonia and carbon dioxide on AD 4 did not correlate with each other; however, the snap shot and the continuous measurements of ammonia correlated positively (Pearson, r = 0.782, P = 0.003). Furthermore, the average daily mortality and the average daily weight gain increased over time and differed within the fattening periods (Table 4). In fattening period 4, a higher average selection rate due to leg weakness and higher average daily mortality were observed compared with the other fattening periods (Table S2). Accordingly, the assessment day influenced the average daily mortality (post-hoc test, Dunnet, P = 0.007 for AD 1 vs. AD 4) and the average daily weight gain (post-hoc test, Dunnet, P < 0.001 for AD 1 vs. AD 4). Due to the increasing weight of the broilers, the stocking density (weight per m2 ) was higher on AD 2, AD 3, and AD 4 than on AD 1 (post-hoc tests, Dunnet, P < 0.001).

Barn Comparison The barn type did not significantly affect any of the examined animal-based indicators such as weight, plumage soiling, skin scratches, FPD, hock burn, and gait score on any of the 4 assessment days (Table 3). Although FPD was on average less severe in the Louisiana barn than in the closed barn, this difference was not significant, and in several fattening periods, FPD was more severe in the Louisiana barn (Table S1). Of the indicators in Tables 4 and 7, only the average daily mortality (%) at the fourth assessment was higher in the Louisiana than in the closed barn (MannWhitney U test, P = 0.024). The selection rate due to leg weakness, average daily weight gain, cumulative mortality, dead on arrival, number of antibiotic treatments, and number of antibiotics used or days of antibiotic treatment within one fattening period did not

Downloaded from https://academic.oup.com/ps/advance-article-abstract/doi/10.3382/ps/pey111/4995452 by University of California, Santa Barbara user on 15 May 2018

8 0.747

0.594

0.714

0.115

0.03 0.06 0.06 0.09 0.04 0.05 0.05 0.06 0.086

0.332

0.000

0.11 0.17 0.25 0.24 0.22 0.20 0.21 0.20 0.000

0 0 0 0 2.69 2.79 5.72 3.35 0.271

0.238

0.317

0.20 0.40 0.20 0.40 3.52 1.88 7.44 2.48 1.000

0.006

0.057

1

4

3

2

1

2

0.027

0.207

0.959 0.909

29.80 28.37 25.10 25.18 19.72 22.10 18.74 21.46 1 2 1 2 1 2 1 2

statistical analysis with a t test for normally distributed data. statistical analysis with a Mann-Whitney U test for non-normally distributed data on AD 2. 3 statistical analysis with a t test for normally distributed data on AD 1, 3, and 4.

0 0 0 0.04 4.60 1.83 4.11 1.81 0.00 0.00 0.00 0.01 4.12 1.80 3.86 1.90 0.410

10.99 6.47 12.89 5.75 10.54 11.89 2.76 6.87 0.041

0.97 1.33 1.06 1.48 1.37 2.64 1.06 1.86

54.22 58.69 61.48 61.21 65.37 57.36 68.92 61.85

P-value1 SD Average P-value1 SD Average P-value2,3 P-value1 AD

Barn

Average

SD

Average

SD

P-value1

Average

SD

CO2 continuous measurement (%) Ammonia continuous measurement (ppm) Ammonia snap shot (ppm) Humidity (%) Temperature (◦ C)

Table 6. Climatic, management-based indicators in the 2 barns (1 = Louisiana barn, 2 = closed barn) on assessment days (AD) 1 to 4 (on average: AD 1 = fattening day [FD] 5, AD 2 = FD 15, AD 3 = FD 32, AD 4 = FD 39).

LOUTON ET AL.

differ significantly between the barn types. Regarding the cumulative mortality throughout the fattening periods, fattening period 4 stands out with very high mortality rates of 8.5% (Barn 1) and 7.6% (Barn 2). Within the fattening periods, the barns were quite similar regarding cumulative mortality, number of antibiotic treatments, number of antibiotics used or days of antibiotic treatment, and dead on arrival (Table S5). In fattening period 3, the average percentages of dead on arrival were higher in both barns than in the other fattening periods. Common diagnoses before antibiotic treatments were omphalitis, aerosacculitis, yolk sac inflammation, polyserositis, pericarditis, necrosis of the femur head, ascites, and arthritis (Table S5). Of the climatic indicators, especially the light intensity differed between the 2 barns (Tables 5 and 6). On average, the light intensity was 20 times higher in the Louisiana barn than in the closed barn, without a negative effect on the animal health measures. Especially in the Louisiana barn, differences in the light intensity throughout the different fattening periods were observed (Table S3). The temperature was on average 1.5◦ C higher at the beginning of the fattening period (AD 1) in the Louisiana barn, and 2.7◦ C higher at the end of the fattening period (AD 4) in the closed barn (Table 6). The humidity was different in the barns throughout the fattening periods. In several fattening periods (fattening periods 2, 6, and 7), we observed a lower humidity on AD 1 and a higher humidity on AD 4 in the Louisiana than in the closed barn (also presented by the average values). Very low humidity values were observed in the Louisiana barn on AD 1 in the fattening periods in the winter–spring (38.6%) and spring (45.7%) seasons. The results of the measurement of the climate outside of the barn are presented in Table S6. The temperature development corresponded to the expected temperature of the climate in Germany for the respective seasons. The highest light intensities outside of the barn were observed in the summer periods. The litter depth was significantly higher and the litter score lower in the Louisiana barn than in the closed barn at the beginning of the fattening period (Tables 7 and S4). Higher gaseous ammonia concentration and an average increase in moisture of the litter in the Louisiana barn than in the closed barn were observed at the end of the fattening period (Tables 5, 6, S3, and S4).

DISCUSSION Animal- and Management-Based Indicators The development of the broiler weight corresponded to the previously defined target weights, and no difference was observed between the 2 barn types, Louisiana barn (Barn 1) and closed barn (Barn 2). Assuming an initial body weight of 42 g, an average weight gain of 65 g/d in Barn 1 and 66 g/d in Barn 2 within 38 to 40 fattening days seems reasonable for the conventional

Downloaded from https://academic.oup.com/ps/advance-article-abstract/doi/10.3382/ps/pey111/4995452 by University of California, Santa Barbara user on 15 May 2018

0.811 0.51 0.44 0.37 0.41 0.895 1.70 2.71 5.29 5.00 0.424

2

1

statistical analysis with a t test for normally distributed data. statistical analysis with a Mann-Whitney U test for non-normally distributed data.

0.53 0.69 1.43 1.14 0.884 0.49 0.76 1.29 1.29 0.735 1.71 1.41 4.83 4.54 1 2

P-value2 SD Average P-value2 SD Average P-value2 SD Average P-value2 SD Average P-value1 SD Average Barn

Days of antibiotic treatment during FP Number of antibiotics used during FP Number of antibiotic treatments during FP Cumulative mortality (%)

Table 7. Animal- and management-based indicators in the 2 barns (1 = Louisiana barn, 2 = closed barn) throughout the fattening period (FP).

Dead on arrival (%)

ANIMAL- AND MANAGEMENT-BASED INDICATORS FOR BROILERS

9

broiler strain Ross 308. In the Ross 308 Broiler Performance Objectives (Aviagen, 2014b), the average weight of the broilers on fattening day 38 is 2429 g. Assuming the initial body weight of 42 g, the average daily weight gain would be 63 g. The positive correlation we found between broiler weights and stocking density was expected because the stocking density is defined as weight per m2 . Furthermore, the production week of the parental stock correlated positively with the weight development assessed in their offspring before slaughter during 7 fattening periods. Lambio et al. (1998) also observed an influence of the age of the parental stock, with higher body weights in broilers of parental stocks with an age of 58 wk compared with 29 wk, but lower weights in broilers of 70-week-old parental stocks than in those of 65-wk-old parents. Similar to findings by Rauch et al. (2017) and Greene et al. (1985), a majority of the assessed broilers in our study showed soiling of the plumage. We found a correlation of the prevalence of plumage soiling with litter depth (negative correlation) and litter quality (positive correlation). Furthermore, the assessment day (representing the age of the broilers) influenced the prevalence of soiled plumage. Rauch et al. (2017) examined broiler strains growing more slowly than the animals in this study and observed a positive correlation of plumage soiling with the weight of the broilers. This was not the case in this study. Similar to our findings, strains kept on dryer litter quality had cleaner plumage than did strains on moist litter in the study by Rauch et al. (2017). In addition, the examined strain with the highest gait score and highest average weight had the most soiled plumage in the study by Rauch et al. (2017). Saraiva et al. (2016) observed a positive correlation between the severity of FPD and soiling of the plumage and assumed a relationship between these 2 factors via the litter quality, which could influence both. De Jong et al. (2014) also stated that broilers that showed FPD due to moist litter would have a more soiled plumage than would broilers with healthy foot pads. We did not observe a relationship of the soiling of the plumage with FPD. Greene et al. (1985) observed that broilers would be cleaner if housed on dry instead of moist litter. Similarly, in our study, a positive correlation of plumage soiling with the litter moisture (quality score) was observed. Broilers keep their feathers clean by preening, which protects them from being moist or dirty. Soiled plumage will lose its protective characteristic for the skin and its function in microclimatic regulation and thus affects the welfare of the broilers (Welfare R Quality , 2009). To prevent soiled plumage, the broilers should be housed on dry litter. Cellulitis as a result of skin scratches is one of the main condemnation reasons at slaughter, and other authors observed skin scratches in broilers beginning at the age of 4 wk (Elfadil et al., 1996). We observed mild and moderate skin scratches in the broilers of this study from the age of FD 15 onward. At the third (FD

Downloaded from https://academic.oup.com/ps/advance-article-abstract/doi/10.3382/ps/pey111/4995452 by University of California, Santa Barbara user on 15 May 2018

10

LOUTON ET AL.

31 to FD 33) and fourth (FD 38 to FD 40) assessments, we also documented severe scratches, and the assessment day significantly affected the prevalence of skin scratches in our study. Elfadil et al. (1996) observed amongst other factors an influence of average body weight of broilers on the occurrence of cellulitis and suggested an association with a high growth rate. However, on each of the 4 assessment days, the weight of the broilers did not correlate with the prevalence of skin scratches in our examination. Furthermore, the prevalence of skin scratches did not correlate with other animal-based or management-related indicators except for the cumulative mortality. These mortality factors are possibly the direct consequences of cellulitis and skin scratches. Norton et al. (1999) considered skin scratches as an essential element for the development of cellulitis. As described by Allain et al. (2009) and Harris et al. (1978), the development of skin scratches may be related to the effective stocking density. Dozier et al. (2005) observed more severe scratches in broiler carcasses at slaughter with increase in stocking densities. As a climatic factor, high light intensity can stimulate the activity of the broilers and the development of skin scratches and subsequently necrotic dermatitis (Weise, 2007). However, Newberry et al. (1988) and Kristensen et al. (2006) did not find a negative effect of high light intensities on broiler performance. In our study, neither light intensity nor stocking density correlated with the prevalence of skin scratches. Scratches were observed in both examined barns irrespective of these factors. It seems that the development of scratches is complex and possibly not related to only 1 factor. Because skin scratches can reduce the animals’ welfare and raise the cumulative mortality, more research on the cause and possible preventive measures is needed. The best assessment time to evaluate skin scratches in the animals seems to be before thinning and final slaughter, because at this time, if existent, severe scratches would be detected. In contrast to other authors (Haslam et al., 2007; Allain et al., 2009; McKeegan, 2010; De Jong et al., 2014), we did not observe a correlation between moist litter and the prevalence of FPD. However, increasing age (assessment days) correlated positively with the prevalence of FPD. This finding is consistent with the studies by De Jong et al. (2012), Bergmann et al. (2016), and McKeegan (2010), who observed that the age of broilers had an effect on the severity of FPD. Sørensen et al. (2000), Haslam et al. (2007), and Dozier et al. (2005) found an influence of the stocking density on the development of FPD and concluded that stocking densities beyond 30 kg/m2 had negative effects on the performance of heavy broilers. Even though it seems reasonable to conclude that high stocking densities could lead to moist litter and thus to FPD, we did not find a correlation of either of these factors with the prevalence of FPD. Note that the differences between the recorded prevalence of FPD were small, and FPD

was more severe in the closed than in the Louisiana barn. However, on average an increased moisture of the litter (litter quality score) was observed in the Louisiana compared with the closed barn. Deep et al. (2010) reported that the development of FPD was influenced by the intensity of light. Even though we also observed slightly fewer broilers with FPD in the 20 times brighter barn, the difference was not significant. As described by McKeegan (2010), we found that the severity of hock burn lesions increased with increasing age of the broilers. Greene et al. (1985) observed hock burns from d 22, whereas the broilers in our study already showed superficial small changes at the first assessment on FD 5. However, a hock burn score of 5 was assessed only rarely and only at the fourth assessment on FD 38 to FD 40. In accordance with the studies by Sørensen et al. (2000), Broom and Reefmann (2005), Haslam et al. (2007), Bergmann et al. (2016), and Saraiva et al. (2016), the likelihood for the development of hock burn correlated positively with a heavy weight of the broilers, with heavier broilers developing hock burn more severely or more frequently. Furthermore, the stocking density correlated positively with the prevalence of hock burn in our study, likely because the stocking density depends on the weight of the broilers. Sørensen et al. (2000) also observed this relationship. As also described by Haslam et al. (2007), we found no correlation of the development of hock burn with the litter quality. However, the ammonia concentration of the snap shot measurement correlated positively with the prevalence of hock burn on AD 4. McKeegan (2010) noted that even though hock burn lesions are little related to litter characteristics and quality, they can be induced by wet litter and high ammonia concentration. The positive correlation we found between hock burn changes and the age of the parental stock could be due to the previously explained relationship between body weight and the age of the parental stock. The optimal assessment time to detect hock burn changes, based on our results, seems to be at the end of the fattening period, before thinning or final slaughter, as the assessment day influenced the prevalence of hock burn. However, small hock burn lesions can already be detected in broilers at the age of FD 5. The hypothesis by McKeegan (2010) that the mechanisms of development differ between hock burn lesions and FPD can be supported by our study. Hock burn lesions were related more to weight and less to litter characteristics. Haslam et al. (2007) described a weak positive correlation between the prevalence of hock burn, but not of FPD, and the gait score in broilers, but our results did not indicate this relationship. In our study, the gait score correlated only with the litter score (positive correlation). No other animal- or management-based indicator related to the gait score. In contrast to studies by Bergmann et al. (2016) and Kristensen et al. (2006), in which the body weight had a significant effect on the

Downloaded from https://academic.oup.com/ps/advance-article-abstract/doi/10.3382/ps/pey111/4995452 by University of California, Santa Barbara user on 15 May 2018

ANIMAL- AND MANAGEMENT-BASED INDICATORS FOR BROILERS

gait score of broilers, we found no correlation between these 2 variables. Kestin et al. (1992) found that the gait score increased to 2 with increasing body weight, whereas birds with gait scores worse than 2 had lower body weights. Similarly, the broilers in our study had increasing weights up to gait score 3 and lower weights at gait scores 4 and 5. The reason could be that broilers with an inability to walk or suffering from pain might not be able to intake the necessary amount of food to gain the expected weight. Sanotra et al. (2003) stated that the differences in the prevalence of a gait score above 2 assessed in their study were due to different body weights and management factors such as housing conditions. Other authors observed large differences in the walking ability of birds between different genotypes, ages, and feeding regimes: slow-growing vs. fast-growing genotypes, younger birds of the age of 54 d vs. 81 d, and birds on non-limited diet vs. Label Rouge diet were less lame than their counterparts (Kestin et al., 2001). However, we did not examine the possible effects of genotypes, age, and feeding regimes on the gait score. Authors (Sanotra et al., 2001; Knowles et al., 2008) observed an influence of the stocking density on the gait score. Sanotra et al. (2001) concluded that 2 effects could be responsible: the reduced locomotion and the deterioration of the litter causing contact dermatitis. The authors assumed that the rapid growth itself can reduce the locomotion. The stocking density in our study did not correlate with the gait score, as also reported by Dozier et al. (2005). This could be due to the relatively small differences in stocking densities combined with a relatively equal distribution of gait scores between the fattening periods. Sørensen et al. (2000) found an influence of sex and age on the gait score, with male and older broilers showing less ability to walk than female and younger broilers. Furthermore, higher stocking densities led to higher gait scores, and the gait score was correlated positively with hock burn and FPD (Sørensen et al., 2000). However, the authors stated that it was unclear whether hock burn and FPD caused the leg weakness or if they were the result of a lame bird being in longer contact with the litter due to longer duration of sitting and less activity. In our study, hock burn as clinical parameter was not correlated with the gait score, whereas the litter quality was. Sørensen et al. (2000) concluded that birds with gait scores of 3 or worse suffer from pain when walking. McGeown et al. (1999) observed that lame birds that received carprofen could transverse a course faster than if they did not receive the non-steroidal antiinflammatory drug. The authors concluded that birds with a moderate lameness suffer pain when they walk, because the application of an analgesic increased the speed of walking in lame birds. The observation that 13.0% of the birds in our study showed a lameness of Score 3 or worse and thus must have suffered from pain should be regarded as critical. Kristensen et al. (2006)

11

observed an “abnormal” gait in 25% of the broilers (Score 2 or higher). According to their categorization, 54% of the broilers were affected by an abnormal gait in our study. It should be considered that broilers with severe leg weakness were selected and culled throughout the fattening periods because they would have suffered had they stayed alive. This measure could have caused lower rates of high gait scores. These animals are included in the “selection rate due to leg weakness.” With increasing age of the broilers, the climatic condition in both barns changed and gaseous ammonia and carbon dioxide concentrations increased. In addition, the litter quality (score) deteriorated in both barns. The concentrations of gaseous ammonia and carbon dioxide did not correlate with each other. The origins of high concentrations of the gases are different. Ammonia is generated through bacterial decomposition of protein and urea (Kristensen and Wathes, 2000). Because high concentrations of ammonia have negative effects on the welfare of the housed broilers (Costa et al., 2012), a reduction of ammonia should be implemented. High carbon dioxide concentrations are the result of insufficient ventilation in the barn (Richter, 2006). The origin of carbon dioxide is the respiration air of the broilers and fermentation processes of organic material (Hoy et al., 2006). According to Hoy et al. (2006), carbon dioxide can be considered as an indicator gas because it reveals barns that are insufficiently supplied with fresh air. They concluded that if high concentrations of carbon dioxide are measured, higher concentrations of other gases are possible. This assumption was not confirmed in our study, probably because high concentrations of carbon dioxide were also observed at the beginning of the fattening period, whereas high concentrations of gaseous ammonia were found at the end of the fattening period. Meluzzi and Sirri (2009) stated that in the winter, ventilation is commonly reduced due to economic reasons. This measure should not be performed and ventilation should be adjusted to the climatic needs of the broilers. Because the snap shot measurement and continuous measurement (5 d before assessment) of gaseous ammonia correlated positively, it should be discussed whether the snap shot measurement would be sufficient to evaluate the ammonia concentrations 5 d before examination.

Barn Comparisons Not much is known about the effect of the barn type (Louisiana vs. closed barn) on animal health indicators such as FPD, hock burn, or gait scores or managementbased indicators. In our study, the health status of the broilers and the cumulative mortality in the examined barn types were similar. Body weights, soiling of the plumage, skin scratches, hock burn, and gait score were the same in both barns. Although FPD was on average less severe in the Louisiana than in the closed barn type, this difference was not significant.

Downloaded from https://academic.oup.com/ps/advance-article-abstract/doi/10.3382/ps/pey111/4995452 by University of California, Santa Barbara user on 15 May 2018

12

LOUTON ET AL.

Within each fattening period, the indicators selecCONCLUSIONS tion rate due to leg weakness, average daily weight For the assessment of animal welfare, several aspects gain, antibiotic treatments, and birds dead on arrival of the broilers should be considered, such as perforwere not different between the 2 barns. Possibly, the mance, animal behavior, physiological and immunologgeneral management of the farm and the fattening peical parameters, mortality, injuries, illness, and medical riod influenced these indicators more than the barn treatments (Broom, 1991). In our study, all of the astype did. A detailed look at the fattening periods insessed animal-based indicators (soiling of the plumage, dicates a problem (high average score of scratches, high skin scratches, FPD, and hock burn) worsened with cumulative mortality, and high average daily mortalprogressing age of the broilers. Soiling of the plumage ity) in fattening period 4 in both barn types, signaling also increased with moist litter. Skin scratches were not the problem being based on general barn management correlated with animalor management-based indicaor fattening period and not linked to the barn type. tors except for the cumulative mortality. The effects The concentration of gaseous ammonia and the litter of housing conditions on skin scratches need more requality differed between the barn types, indicating a search because they influence mortality and the cause is management-related difference between the barn types. not clear. Hock burn development was correlated posiCompared with the closed barn, the Louisiana barn was tively with high body weights, high stocking density, associated with higher gaseous ammonia concentrations high gaseous ammonia concentration, and increasing and increased litter moisture at the end of the fattenproduction week of the parental stocks. ing period, and the light intensity was 20 times higher. The quality of the litter and the concentrations of Even though Weise (2007) reported that high light innoxious gases worsened with ongoing fattening period. tensities could possibly have a negative effect on the However, the concentrations of gaseous ammonia and development of scratches, we did not observe any signs carbon dioxide did not correlate, so these gases should of negative effects. Differences of light intensities in the always be measured individually. Louisiana barn were observed throughout different fatThe barn type (Louisiana vs. closed) did not correlate tening periods and assessment days. High light intensiwith the animal-based health indicators but did corties were especially observed in the summer season in relate with several management-based indicators. Furthe Louisiana barn. thermore, several parameters of management and housCompared with the temperature in the closed barn, ing correlated with each other, indicating that the barn the temperature in the Louisiana barn was on average ◦ ◦ type should be well chosen and has an effect on the de1.5 C higher shortly after housing (AD 1) and 2.7 C velopment of gaseous ammonia concentration and the lower at the end of the fattening period (AD 4). The light intensity. Comparing the 2 housing systems, we average humidity in the Louisiana barn was lower on could not prove that the Louisiana barn has a positive AD 1 and higher at the end of the fattening period. The or negative effect on the health of the broilers, and it broilers should be housed in their comfort zone, and huseems that this barn requires a more accurate and spemidity of the air should be kept above 50% (Aviagen, cific management than does the closed barn type. The 2014b). The humidity on AD 1 was below 50% in the development of gaseous ammonia, the exposure to natLouisiana barn in 2 fattening periods. These low huural light intensities, and the litter quality need a more midity values on AD 1 should be avoided. Yoder et al. precise management in the Louisiana barn. However, no (1977) observed more incidents of airsacculitis in broilnegative effect of higher light intensities was observed ers raised at low humidity compared with high humidon skin scratches or any other animal health parameter. ity. Weaver and Meijerhof (1991) observed an increase In conclusion, the observed positive correlations bein moisture of litter and in ammonia levels with intween the animal- and management-based indicators creases in relative humidity. These observations are in suggest that management targets should include low line with our observations in the Louisiana barn at the concentrations of gaseous ammonia and carbon dioxide, end of the fattening period, in which on average a higher dry litter, and prevention of animals scratching each humidity, higher litter score, and higher ammonia level other. High weights of the broilers are considered critiwere observed. cal as they were associated with a higher prevalence of However, we assessed the effect of management-based hock burn. Management and barn type have an effect indicators and barn types with the results from 2 barns on the development of gaseous ammonia concentration on 1 farm. It is possible that the farm management itand the light intensity and should be considered indiself has an influence on the number of antibiotic treatvidually. ments and the litter characteristics. Nonetheless, the Louisiana barn, with a higher concentration of gaseous ammonia and potentially more severe deterioration of SUPPLEMENTARY DATA litter quality, may need a more detailed management than the closed barn type does. Furthermore, the exSupplementary data are available at Poultry Science posure of broilers to natural light was less controllable online. in the Louisiana barn because of the open sides of the barn. Especially in the summer season, high light inTable S1. Animal-based indicators (average score) in tensities above 1000 lx were possible inside the barn. the 2 barns (1 = Louisiana barn, 2 = closed barn) on Downloaded from https://academic.oup.com/ps/advance-article-abstract/doi/10.3382/ps/pey111/4995452 by University of California, Santa Barbara user on 15 May 2018

ANIMAL- AND MANAGEMENT-BASED INDICATORS FOR BROILERS

assessment days (AD) 1 to 4 (on average: AD 1 = fattening day [FD] 5, AD 2 = FD 15, AD 3 = FD 32, AD 4 = FD 39). Allocation of the 7 examined fattening periods in Barn 1 and Barn 2 to the respective seasonal periods in which the examinations took place: spring = March 1 to May 31, summer = June 1 to August 31, autumn = September 1 to November 30, winter = December 1 to February 28/29. Table S2. Management-based indicators in the 2 barns (1 = Louisiana barn, 2 = closed barn) on assessment days (AD) 1 to 4 (on average: AD 1 = fattening day [FD] 5, AD 2 = FD 15, AD 3 = FD 32, AD 4 = FD 39). Data provided by farmer and summarized as the average of the 5 d before each AD. Allocation of the 7 examined fattening periods in Barn 1 and Barn 2 to the respective seasonal periods in which the examinations took place: spring = March 1 to May 31, summer = June 1 to August 31, autumn = September 1 to November 30, winter = December 1 to February 28/29. Table S3. Climatic management-based indicators (average) in the 2 barns (1 = Louisiana barn, 2 = closed barn) on assessment days (AD) 1 to 4 (on average: AD 1 = fattening day [FD] 5, AD 2 = FD 15, AD 3 = FD 32, AD 4 = FD 39). Allocation of the 7 examined fattening periods in Barn 1 and Barn 2 to the respective seasonal periods in which the examinations took place: spring = March 1 to May 31, summer = June 1 to August 31, autumn = September 1 to November 30, winter = December 1 to February 28/29. Table S4. Climatic management-based indicators (average) in the 2 barns (1 = Louisiana barn, 2 = closed barn) on assessment days (AD) 1 to 4 (on average: AD 1 = fattening day [FD] 5, AD 2 = FD 15, AD 3 = FD 32, AD 4 = FD 39). Allocation of the 7 examined fattening periods in Barn 1 and Barn 2 to the respective seasonal periods in which the examinations took place: spring = March 1 to May 31, summer = June 1 to August 31, autumn = September 1 to November 30, winter = December 1 to February 28/29. Table S5. Animal- and management-based indicators in the 2 barns (1 = Louisiana barn, 2 = closed barn) throughout the fattening period (FP). Allocation of the 7 examined fattening periods in Barn 1 and Barn 2 to the respective seasonal periods in which the examinations took place: spring = March 1 to May 31, summer = June 1 to August 31, autumn = September 1 to November 30, winter = December 1 to February 28/29. Arth = arthritis, Asc = ascites, Hep = hepatitis, NFH = necrosis of femur head, Omph = omphalitis, PC = pericarditis, PS = polyserositis, SP = subcutaneous pyosis, YS-Infla = yolk sac inflammation. Table S6. Outdoor climatic conditions measured on the assessment days (AD) in front of both barns (AD 1 = fattening day [FD] 5, AD 2 = FD 15, AD 3 = FD 32, AD 4 = FD 39). Allocation of the 7 examined fattening periods to the respective seasonal periods in which the measurements took place: spring = March 1 to May 31, summer = June 1 to August 31, autumn = September

13

1 to November 30, winter = December 1 to February 28/29.

ACKNOWLEDGMENTS We thank the farmers and their family for their cooperation and Katinka Wirsch for her dedicated support during data collection.

FUNDING The “BEST Chicken” (BEST = B¨ undnis f¨ ur Exzellenz, Sicherheit und Transparenz; Alliance for Excellence, Safety and Transparency) study was funded by McDonald’s Germany LLC, Branch Office Munich, Munich, Germany.

Ethical statement The work described in this article with research on live animals met the guidelines approved by the institutional animal care and use committee (IACUC).

REFERENCES Allain, V., L. Mirabito, C. Arnould, M. Colas, S. Le Bouquin, C. Lupo, and V. Michel. 2009. Skin lesions in broiler chickens measured at the slaughterhouse: relationships between lesions and between their prevalence and rearing factors. Br. Poult. Sci. 50:407– 417. Arnould, C., A. Butterworth, and U. Knierim. 2009. Standardisation of clinical scoring in poultry. Pages 7–30 in Assessment of Animal Welfare Measures for Layers and Broilers. B. Forkmann, R Reports Series 9. Cardiff and L. Keeling, eds. Welfare Quality University Press, Cardiff, UK. ISBN: 1-902647-79-3. Aviagen. 2014a. Arbor Acres. Broiler Carcass Condemnation and Downgrade Management Pocket Guide. Accessed Feb. 12, 2018. http://cn.aviagen.com/assets/Tech Center/BB Resources Tools/ Pocket Guides/AA-Carcass-Condemned-Pocket-Guide-0714EN.pdf. Aviagen. 2014b. Broiler Ross 308 Performance Objectives. Accessed Sep. 2015. http://en.aviagen.com/assets/Tech Center/ Ross Broiler/Ross-308-Broiler-PO-2014-EN.pdf. Bergmann, S., H. Louton, C. Westermaier, K. Wilutzky, A. Bender, J. Bachmeier, M. Erhard, and E. Rauch. 2016. Field trial on animal-based measures for animal welfare in slow growing broilers reared under an alternative concept suitable for the German market. Berl. M¨ unch. Tier¨ arztl. Wochenschr. 129:453–461. BEST Chicken Standard. 2014. Basiert auf verbindlichen Richtlinien f¨ ur die Aufzucht und Haltung in H¨ ahnchenmast Pilotfarmen. Stand. 03.04.2014. Broom, D. M. 1991. Assessing welfare and suffering. Behav. Processes. 25:117–123. Broom, D. M., and N. Reefmann. 2005. Chicken welfare as indicated by lesions on carcases in supermarkets. Br. Poult. Sci. 46:407–414. Butterworth, A. 2013. On-farm broiler welfare assessment and associated training. Rev. Bras. Cienc. Avic. 15:71–77. Costa, A., S. Ferrari, and M. Guarino. 2012. Yearly emission factors of ammonia and particulate matter from three laying-hen housing systems. Anim. Prod. Sci. 52:1089–1098. Council of the European Union. 2007. Council Directive 2007/43/EC of 28 June 2007, laying down minimum rules for the protection of chickens kept for meat production. Damme, K., and R.-A. Hildebrand. 2002. Gefl¨ ugelhaltung. Verlag Eugen Ulmer, Stuttgart, Germany. ISBN-13: 9783800139293.

Downloaded from https://academic.oup.com/ps/advance-article-abstract/doi/10.3382/ps/pey111/4995452 by University of California, Santa Barbara user on 15 May 2018

14

LOUTON ET AL.

Deep, A., K. Schwean-Lardner, T. G. Crowe, B. I. Fancher, and H. L. Classen. 2010. Effect of light intensity on broiler production, processing characteristics, and welfare. Poult. Sci. 89:2326– 2333. De Jong, I. C., J. van Harn, H. Gunnink, V. A. Hindle, and A. Lourens. 2012. Footpad dermatitis in Dutch broiler flocks: prevalence and factors of influence. Poult. Sci. 91:1569–1574. De Jong, I. C., H. Gunnink, and J. van Harn. 2014. Wet litter not only induces footpad dermatitis but also reduces overall welfare, technical performance, and carcass yield in broiler chickens. J. Appl. Poult. Res. 23:51–58. Dozier, W. A., III, J. P. Thaxton, S. L. Branton, G. W. Morgan, D. M. Miles, W. B. Roush, B. D. Lott, and Y. Vizzier-Thaxton 2005. Stocking density effects on growth performance and processing yields of heavy broilers. Poult. Sci. 84:1332–1338. Elfadil, A. A., J. P. Vaillancourt, A. H. Meek, R. J. Julian, and C. L. Gyles. 1996. Description of cellulitis lesions and associations between cellulitis and other categories of condemnation. Avian Dis. 40:690–698. Forkman, B., and L. Keeling. 2009. Assessment of Animal Welfare R Reports SeMeasures for Layers and Broilers. Welfare Quality ries 9. Cardiff University Press, Cardiff, UK. ISBN: 1-902647-793. German Order on the Protection of Animals and the Keeping of Production Animals. 2006. Amended and promulgated on August 22nd, 2006, last changed on June 30th, 2017. Verordnung zum Schutz landwirtschaftlicher Nutztiere und anderer zur Erzeugung tierischer Produkte gehaltener Tiere bei ihrer Haltung; Tierschutz–Nutztierhaltungsverordnung in der Fassung der Bekanntmachung vom 22. August 2006 (BGBI. I S. 2043), die zuletzt durch Artikel 3 Absatz 2 des Gesetzes (BGBI. I S. 2147) ge¨andert worden ist. Greene, J. A., R. M. McCracken, and R. T. Evans. 1985. A contact dermatitis of broilers- clinical and pathological findings. Avian Pathol. 14:23–38. Haslam, S. M., T. G. Knowles, S. N. Brown, L. J. Wilkins, S. C. Kestin, P. D. Warriss, and C. J. Nicol. 2007. Factors affecting the prevalence of foot pad dermatitis, hock burn and breast burn in broiler chicken. Br. Poult. Sci. 48:264–275. Harris, G. C., Jr., M. Musbah, J. N. Beasley, and G. S. Nelson. 1978. The development of dermatitis (scabby-hip) on the hip and thigh of broiler chickens. Avian Dis. 22:122–130. Hoy, S., M. Gauly, and J. Krieter. 2006. Nutztierhaltung und hygiene. Grundwissen Bachelor. Verlag Eugen Ulmer, Stuttgart, Germany. ISBN-13: 978-3-8252-2801-9. Kestin, S. C., T. G. Knowles, A. E. Tinch, and N. G. Gegory. 1992. Prevalence of leg weakness in broiler chickens and its relationship with genotype. Vet. Rec. 131:190–194. Kestin, S. C., S. Gordon, G. Su, and P. Sørensen. 2001. Relationships in broiler chickens between lameness, liveweight, growth rate and age. Vet. Rec. 148:195–197. Knowles, T. G., S. C. Kestin, S. M. Haslam, S. N. Brown, L. E. Green, A. Butterworth, S. J. Pope, D. Pfeiffer, and C. J. Nicol. 2008. Leg disorders in broiler chickens: prevalence, risk factors and prevention. PLoS One 3:e1545. doi:10.1371/journal.pone.0001545 Kristensen, H. H., and C. M. Wathes. 2000. Ammonia and poultry welfare: a review. Worlds Poult. Sci. J. 56:235–245. Kristensen, H. H., G. C. Perry, N. B. Prescott, J. Ladewig, A. K. Ersbøll, and C. M. Wathes. 2006. Leg health and performance of broiler chickens reared in different light environments. Br. Poult. Sci. 47:257–263. Lambio, A. L., A. D. Balbarez, and A. S. Doctor. 1998. Performance of broilers from parent stocks of different age groups. Philipp. J. Vet. Anim. Sci. 13:5–10. McGeown, D., T. C. Danbury, A. E. Waterman-Pearson, and S. C. Kestin. 1999. Effect of carprofen on lameness in broiler chickens. Vet. Rec. 144:668–671. McKeegan, D. 2010. Foot pad dermatitis and hock burn in broilers: risk factors, aetiology and welfare consequences.

Research Project Final Report, Faculty of Veterinary Medicine, University of Glasgow, Scotland, UK. Accessed Dec. 2016. http://randd.defra.gov.uk/Default.aspx?Menu=Menu&Module= More&Location=None&Completed=0&ProjectID=14546. McNamee, P. T., J. J. McCullagh, B. H. Thorp, H. J. Ball, D. Graham, S. J. McCullough, D. McConaghy, and J. A. Smyth. 1998. Study of leg weakness in two commercial broiler flocks. Vet. Rec. 143:131–135. Meluzzi, A., and F. Sirri. 2009. Welfare of broiler chickens. Ital. J. Anim. Sci. 8:161–173. Newberry, R. C., J. R. Hunt, and E. E. Gardiner. 1988. Influence of light intensity on behavior and performance of broiler chickens. Poult. Sci. 67:1020–1025. Norton, R. A., K. S. Macklin, and B. L. McMurtrey. 1999. Evaluation of scratches as an essential element in the development of avian cellulitis in broiler chickens. Avian Dis. 43: 320–325. Rauch, E., C. Keppler, K. Damme, M. Hausleitner, J. Bachmeier, J. Hartmann, and H. Louton. 2017. Untersuchungen unterschiedlicher langsam wachsender Premiumh¨ uhner unter Tierschutz-Label-Konditionen. Teil II: Tiergesundheit. Europ. Poult. Sci. 81:1–15. Rea, L. M., and R. A. Parker. 2014. Designing and Conducting Survey Research: A Comprehensive Guide, 4th Edition. Jossey-Bass, San Francisco, CA, ISBN 978–1118767030. Richter, T. 2006. Krankheitsursache Haltung: Beurteilung von Nutztierst¨ allen - Ein tier¨ arztlicher Leitfaden. Enke Verlag, Stuttgart, Germany. ISBN 978-3-8304-1043-0. Sanotra, G. S., L. G. Lawson, K. S. Vestergaard, and M. G. Thomsen. 2001. Influence of stocking density on tonic immobility, lameness, and tibial dyschondroplasia in broilers. J. Appl. Anim. Welf. Sci. 4:71–87. Sanotra, G. S., C. Berg, and J. D. Lund. 2003. A comparison between leg problems in Danish and Swedish broiler production. Anim. Welf. 12:677–683. Saraiva, S., C. Saraiva, and G. Stilwell. 2016. Feather conditions and clinical scores as indicators of broilers welfare at the slaughterhouse. Res. Vet. Sci. 107:75–79. Scientific Committee on Animal Health and Animal Welfare. 2000. The welfare of chickens kept for meat production (broilers). Report of the Scientific Committee on Animal Health and Animal Welfare. European Commission Health & Consumer Protection Directorate General, Directorate B – Scientific Health Opinions. Unit B3 – Management of Scientific Committees II. SANCO.B.3/AH/R15/2000. Sørensen, P., G. Su, and S. C. Kestin. 2000. Effects of age and stocking density on leg weakness in broiler chickens. Poult. Sci. 79:864– 870. Weaver, W. D., Jr., and R. Meijerhof. 1991. The effect of different levels of relative humidity and air movement on litter conditions, ammonia levels, growth, and carcass quality for broiler chickens. Poult. Sci. 70:746–755. Weise, H. 2007. Lichtbedingte Einfl¨ usse auf Verhalten und Leistung in der H¨ ahnchenmast – Eine Feldstudie unter Ber¨ ucksichtigung tierschutzrechtlicher und wirtschaftlicher Aspekte. Diss., Dr. med. vet., Ludwig-Maximilians-Universit¨ at M¨ unchen, Munich, Germany. R R . 2009. Welfare Quality assessment protocol for Welfare Quality R Consortium, poultry (broilers, laying hens). Welfare Quality Lelystad, Netherlands. Yoder, H. W., Jr, L. N. Drury, and S. R. Hopkins. 1977. Influence of environment on airsacculitis: effects of relative humidity and air temperature on broilers infected with Mycoplasma synoviae and infectious bronchitis. Avian Dis. 21:195–208. Zapf, R., U. Schultheiß, W. Archilles, L. Schrader, U. Knierim, H-J. Herrmann, J. Brinkmann, and C. Winckler 2015. Tierschutzindikatoren. Vorschl¨ age f¨ ur die betriebliche Eigenkontrolle. KTBL Schrift 507, Kuratorium f¨ ur Technik und Bauwesen in der Landwirtschaft e.V. (KTBL), Darmstadt, Germany. ISBN 978-3945088-06-7.

Downloaded from https://academic.oup.com/ps/advance-article-abstract/doi/10.3382/ps/pey111/4995452 by University of California, Santa Barbara user on 15 May 2018