Welfare of Broilers Ingesting a Pre-Slaughter Hydric Diet of Lemon Grass

I Graduate Student in Animal Science, College of Agrarian Sciences, Federal University of Grande Dourados, Rodovia Dourados Itahum, km 12, P.O. Box 533, CEP 79804-970, Dourados, MS, Brazil. II Professor, Animal Science Department, College of Agrarian Sciences, Federal University of Grande Dourados, Rodovia Dourados Itahum, km 12, P.O. Box 533, CEP 79804-970, Dourados, MS, Brazil. III Laboratory of Statistics and Geoinformation, Department of Statistics, Federal University of Paraná, P.O. Box 19.081 CEP 81.531990, Curitiba, Paraná, Brazil. IV Pos-Doctorade Studies, College of Agrarian Sciences, Federal University of Grande Dourados, Rodovia Dourados Itahum, km 12, P.O. Box 533, CEP 79804-970, Dourados, MS, Brazil. V Professor, Federal University of Tocantins, Araguaína, TO, Brazil.


INTRODUCTION
During the pre-slaughter period before harvesting, broilers are kept at high densities, and are submitted to stresses, such as feed withdrawal, poor ventilation in the transport crate, and handling and harvesting handling practices ( Barbosa Filho, 2012). This management negatively affected broiler welfare and interferes with their corticosterone balance (Macari & Luqueti, 2002), which is a hormone released due to stress exposure. Corticosterone may also induce cardiovascular diseases, ascites, and modifications in immune functions (Grandin, 1998), in addition of causing lymphoid tissue hypotrophy and hypoplasia (Compton et al., 1990). In addition of blood changes, an increase in the heat exchange in the featherless body parts of broilers under stress is also observed (Dahlke et al., 2005) due to the increase in peripheral vasodilatation (Borges et al., 2003), leading to variation in surface temperature (Nääs, 2011).
Infrared technology is a valuable tool to evaluate heat exchange in broilers as it accurately records surface temperature (Cangar et al., 2008;Yahav et al., 2008), allowing the estimation of metabolic heat exchange, which is associated with heat stress.
Another key aspect in the estimation of broiler welfare is behavioral observations (Pereira et al., 2007), including the analysis of their social behavior. It is possible to infer on broiler welfare status is a flock by determining the quantity and quality of social interactions, (Nääs et al., 2005;Nazareno et al., 2011).
Some plants contain chemical active principles that may reduce anxiety and stress in humans (Rocha et al., 2008). Lemon grass (Cymbopogon citratus Stapf) affects the sympathetic nervous system, reducing stress (Negrelle & Gomes, 2007). Although some researchers investigated the use of phytotherapy in poultry production, most are related to animal nutrition (Wenk, 2003;Hashemi & Davoodi, 2010) and microbiological control (Arshad et al, 2008;Lourenço et al., 2013). This study aimed at evaluating the use of lemon grass (Cymbopogon citratus Stapf) to reduce broiler pre-slaughter stress.

MATERIALS AND METHODS
This study was carried out at the experimental poultry sector of the Federal University of Grande Dourados, MS, Brazil, and it was approved by the Ethics Committee (CEUA; 007/2012). An open-sided poultry house (50-m long; 10-m wide, and 3-m high) was divided into 56 pens measuring 4.5 m 2 each. Each pen was equipped with a bell drinker and a tube feeder. The house environment was controlled by the use of side double curtains and ventilation fans and foggers distributed throughout the house, as well as an evaporative cooling system controlled by a software.
The experiment was carried out in September, 2012. The birds were reared according to the genetic strain manual, at a final flock density of 16 birds/m². House environment (dry bulb temperature and relative humidity) was maintained as recommended by the genetic strain manual. Litter consisted of rice hulls (5cm deep). Broilers were reared until 42 days old and were offered water and feed ad libitum. Commercial balanced feeds, supplying the broilers' requirements for the pre-starter, starter, grower, and finisher phases, were fed.
In total, 2592 42-d-old broilers were distributed according to a completely randomized experimental design in a 3x2x2 factorial arrangement, with four replicates of 54 birds each. Treatments consisted of three different levels of a lemon grass (Cymbopogon citratus Stapf) infusion added to the drinking water (0, 0.1, or 5 g/L of water), two sexes, and two genetic strains (Ross ® 308 e Cobb ® 500).
The oven-dried lemon grass (Cymbopogon citratus Stapf) leaves were acquired from a commercial company (Flores e Ervas Comércio Farmacêutico Ltda., São Paulo, Brazil). The infusions (0, 0.1, or 5 g of leaves/L of water) were prepared by crushing the dried leaves and immersing them in boiling water (100° C). The infusions were left to cool until reaching room temperature, filtered, and then poured into the drinkers (Cruz et al., 2007). The infusions were offered to the broilers during the 6-h pre-harvesting fasting period (Castro et al., 2008) on the 42 nd day of the grow-out. The infusion was made available to the birds starting at 02h:00 min. During the same period, ambient temperature (Ta) was recorded using a digital thermo hygrometer Company,.
Thermal images were recorded three times during the trial using an infrared camera (Testo®, model 880) with an accuracy of ± 0.1 °C and a series of 7.5 µm of the infrared spectrum, placed at 1 m height from the birds. Images were recorded one hour after the infusions were offered (03h:00min), at 05h:00min, and 07h:00min. Two thermal images per treatment, per replicate, and per recording time were analyzed using a thermal-imaging software (IR-Soft, Testo®, 2009), with iron filter and emissivity coefficient of 0.95 for the broiler body surface (Nääs et al., 2010). In order to accurately determine broiler surface temperature (Ts), 30 points were randomly selected on the broiler body, and the average value was calculated. Temperature data were submitted to analysis of variances and means were compared by the test of Tukey at 5% significance levels, using R Statistical Software (2012).
Broiler behavior was evaluated by indirect monitoring using video footage. A total of three video-recordings (5 min long each) were made per treatment. Three broilers were selected for observation during the total video footage, and the video was played and re-played until each bird was observed for 5 min. Broiler behaviors (Table 1) were observed and recorded as suggested by Souto (2003). The activities were computed in percentage within the observational time, with 5 min equivalent to 100%. The results were submitted to analysis of variance and means were compared by the Kruskal-Wallis test at 95% of significance level, c (2012).
Twelve broilers per treatment were selected (three per replicate, as recommended by Rosa et al., 2002). Birds were duly identified and placed in transport crates (10 birds per crate) and transported to the Meat Laboratory of the Federal University of Grande Dourados (UFGD). Before slaughter, 5 mL of blood were collected per

Welfare of Broilers Ingesting a Pre-Slaughter Hydric Diet of Lemon Grass
bird by ulnar vein puncture in order to quantify blood cortisol levels (Valle et al., 2008). Blood samples were collected in test tubes containing one drop of heparin (15 μL heparin/1 mL blood)and identified per treatment. The tubes were centrifuged at 4000 rpm for 10 min (Centribio, 80-2B, SP-Brazil) to, and the supernatant was homogenized, resulting in three samples per treatment. Samples were stored at 20°C (Valle et al., 2008) and analyzed using a commercial kit (Cortisol EIA, E52061 Westbrook, ME, USA).
The results were compared using the Kruskal-Wallis test at 95% of significance level, using R Statistical Software (2012). Cortisol levels were converted from nmol/L (titration unit) to µ/dL of blood using a correction factor of 27.59.

RESULTS AND DISCUSSION
Broiler surface temperature was not affected by sex, genetic strain, or lemon grass infusion levels (p > 0.05) during the pre-slaughter fasting period (Table  2). Air ambient temperature and relative humidity (RH) measured each time the birds' surface temperature was recorded were: recording 1= 20.5º C and 58.9 RH; recording 2 = 22.0 ºC and 52.1 RH; recording 3 = 24.0°C and 51.3 RH. No interaction was found between the times infrared images were recorded and broiler surface temperature, differently from results of Santi (2012). This author, using infrared imaging, observed lower broiler surface temperature at the times when environmental temperature were lower (17.9° C; 20.0° C; 25.7° C), confirming the relationship between body and environmental temperatures. According to Borges et al. (2003), broilers become sensitive to heat stress when air relative humidity and ambient temperature exceed the thermal neutral zone values, hindering heat dissipation and increasing their body temperature to hazardous limits. This has a negative effect of live performance and may result in death. The recommended ambient temperature and air relative humidity for broilers are 21-23º C and 65-70 %, respectively (UBA, 2008). Ross® 350 broilers exhibited higher frequency of open beak behavior (p<0.05) than Cobb® 500 broilers. This behavior was previously described by Pereira et al. (2007), who reported that Ross® 350 broilers were more susceptible to stress than Cobb ® 500 broilers, and may present high metabolic rates when faced with stressful situations, such as pre-slaughter handling (Roque-Specht et al., 2009). Broilers usually change their behavior when facing stress, and try to adapt to it (Dahlke et al., 2005). Some behavioral patterns found in the present study ( Figure 1) were also reported by other authors, such as wing flapping and stretching (Alcoock, 2011). However, these are kinetic activities and may not be associated with stress (Morrison,

Welfare of Broilers Ingesting a Pre-Slaughter Hydric Diet of Lemon Grass
2007). Except for pecking other birds, no negative interactions among birds were observed during the observation period (Table 3). Preening is described as the behavior of cleaning and tiding its feathers with the beak (Pereira et al., 2013), and it is an indication of normal behavior. It was observed in several birds of all groups during the observation period; however, no differences were found among treatments (Figure 1). Blood cortisol levels were not different between sexes or genetic strains. In addition, lemon grass levels did not influence blood cortisol levels (p>0.05), which is consistent with the findings of Quinteiro Filho (2008). Several factors may increase blood cortisol levels when birds are exposed to stressors (Macari et al., 1994;Caires et al., 2008). In the present experiment, it was expected that, due to its known phytotherapeutical effect of reducing human anxiety, the lemon grass infusion would be able to reduce broilers blood cortisol levels. However, the results indicate that the active chemical principles of lemon grass were not able to change blood cortisol levels of broilers during the preslaughter period.