Introduction

Poultry and egg production systems are economically efficient because they increase bird population per unit area and decrease the time to commercialization. These leads to abnormal behavioral patterns characterized by collective panic attacks or avian hysteria and cannibalism (^{Garcia-Belenguer and Mormede, 1993}), due to the physical stress caused by shelters with inappropriate environmental temperatures, hunger, thirst, damage or noise and psychological damage by mismanagement. However, there is a new market for poultry products, which is differentiate in terms of animal welfare, and grows as information, awareness and society’s perception of animal production increase (^{Raineri et al., 2012}).

The bird management impact, in regard animal welfare, is evaluated by assessing the development stability (DE) of an organism, because it reflects the individual´s ability to ideally produce under certain conditions and shows their intrinsic abilities to resist accidents and external disturbances during their growth and development (^{Clarke, 1998}, ^{Benítez and Parra, 2011}). A tool to estimate DE is to evaluate the bilateral morphological structures by a fluctuating asymmetry index, expressed as the absolute value of the difference between the left and right (L-R) sides of the animals, whose value close to zero would indicate perfect symmetry, which would ideally shows that the sides are identical (^{Klingenberg et al., 2003}; ^{Van Dongen, 2006}). However, there is no absolute similarity because the bilateral anatomical structures are not similar in size and shape; they always differ from each other, giving rise to a FA evaluated through their differences by a standard normal distribution (^{Cocilovo et al., 2006}). Other tools, such as the directional asymmetry (DA) and antisymmetry (AS) are evaluated by character distribution. These because, when one side is larger than the other is, the DA has a normal distribution with a mean other than zero and the SA as a non-normal distribution with a mean of zero (^{Palmer, 1994}; ^{Auffray et al., 2003}; ^{Knierim et al., 2007}). Genetic factors related to loss of genetic variability, homozygosity, directional selection, mutations, and hybridization, as well as environmental factors related to adverse temperatures, nutritional stress, and chemical factors or population density, generate variations that the fluctuating asymmetry reflects (^{Parsons, 1990}; ^{Campo et al., 2006}).

Another animal welfare estimator is the relationship between heterophiles and lymphocytes (H:L). This is an indicator of acute stress in birds due to the diphasic response when faced to stressful situations, leading to changes in the blood cellular components, such as heteropenia and lymphopenia due to the increase in cholesterol levels. Still, the H:L ratio does not indicate a greater or lesser degree of disease susceptibility (^{Tejeda et al., 1997}) because heterophiles are responsible for the defenses of the organism against bacteria, while lymphocytes recognize and destroy a large variety of pathogens (^{Campbell, 1995}; ^{Davis et al., 2008}). According to ^{Gross and Siegel, (1983)} there are three characteristic values for the H:L ratio: 0.2 for low stress, 0.5 for optimal stress and 0.8 for high stress. Tonic or muscular immobility (MI), also known as animal hypnosis, is a good indicator of the psychological well-being associated to fear in birds (^{Gallup, 1979}), which is a normal response when faced with their predators, called fake death. This low level of reaction to a stimulus propitiates the animals to change posture, which is considered a behavioral problem. ^{Campo et al. (2002}) mentioned that the factors that cause MI are related to management deficiencies of the birds due to inadequate transport, forced movements of the flocks, tethered animals management, pain and persecution in enclosed areas. The main features of MI are stress, absence of physical pathologies, head, eyes, ears and legs movements, and electrical sensitivity.

The objective in this study was to determine the influence of the management (cage and freedom) and sex (females and males) on animal welfare indicators estimated by fluctuating asymmetry, muscle immobility and heterophile/lymphocyte ratio on a population of autochthonous domestic turkeys in the municipality of Villaflores, Chiapas.

Materials and Methods

We carried out the study at the town of Jesús María Garza in the Villaflores municipality, from August to October 2015 in order to assess the females’ posture seasonality and the sexual maturity of the males during the summer and early fall season. We performed the study in two phases, one for adaptation, 7 d, and another 60 d of experimentation. The turkeys were housed in individual cages (1 m²), identified with metal rings and on each one, four measurements were made. The first at the beginning of the experiment and others at 20 d intervals. The turkeys were 23 males from 7 to 10 months age and 14 females at laying stage. These were at family production units within the Villaflores municipality. Females at the beginning of the posture stage and similar age were chosen, for which rectal palpation was performed by inserting an index finger into the cloaca sewer and checking for egg formation.

Experimental methodology

To test the effect of the management system and sex on stress indicators, turkeys were randomly assigned to four treatments, resulting from the factorial combination of two types of management (a₁: free range, a₂: confinement) and two sexes (b₁: males, b₂: females), with different number of repetitions. The distribution of the turkeys to the treatments was: T₁: 11 males in cage, T₂: 12 males in free range, T₃: 7 females in cages and, T₄: 7 females in free range. We used, for the data analysis the following fixed effects model with two classification and interaction criteria:

where: y _ij =response variable (AF, IM, H:L), µ =general mean, t _i =management effect (type of housing) (i=1,2), s _j =sex effect (j=1, 2), ts _ij =managementẊsex effect interaction, ε _ijk =experimental error.

For this analysis the GLM procedure of the SAS software (^{SAS Institute Inc., Cary, NC, 2008}) was used.

Results and Discussion

The interaction managementẊsex was not significant (p˃0.05) on the bilateral wing length character, in the asymmetry indicators (Table 1). The between sexes comparison was different (p≤0.01) for the bilateral indexes: (R+L)/2, |R-L| and Relative Fluctuating Asymmetry (RFA). In addition, the average RFA=2.84 for caged turkeys was higher (p≤0.01) than those managed in free range (RFA=1.10), that is, females had a higher RFA than males (p≤0.01). These results indicate that stress levels are caused by cage management. This causes a change in their natural behavior, due to the reduction of the housing area, blocking the possibility of turkey to flap, move, dig, and take sand baths and pecking. All these cause turkeys instability in its development regard this trait (^{Parsons, 1992}; ^{Watson and Thornhill, 1994}).

Table 1 Mean and standard errors (EEM) of the asymmetry^† of five bilateral characters in autochthonous domestic turkeys (guajolote) at Villaflores, Chiapas. 

^{a, b} Mean in a row with different superscript indicate significant differences (p≤0.05). ^†Relative fluctuating asymmetry expressed: |R-L|/ (R+L). ^¶R: Right side; L: Left side. ^§FA: fluctuating asymmetry; AS: antisymmetry; AD: directional asymmetry.

The analysis of the tarsus length showed interaction between the management and sex (p≤0.01) for the index of asymmetry (R+L)/2, the mean of this character was different (p≤0.01) between genders. The caged males showed superior values to those of the females, which were modified when changing from a cage management to free range management. Also, the females did not present differences when changing the management. The interaction was significant (p≤0.05) between the management type and the sex for the index (R-L). These differences in the length and width of the tarsus could be due to genetic differences in the turkeys, and to environmental differences because of greater exercise in the males in free range, due to more walking to look for their food. Besides, the stress that causes an aggressive behavior among them due to the competition by females during the mating season, which manifests itself in fights, pickings and sexual aggressiveness that leads to greater tarsus development. This does not occur with breeding females, which during chicks incubation restrict their movements, thus have lower energy expenditure and a smaller development of their tarsus compared to males in both management types (^{Millman et al., 2000}).

Differences in the mentioned characters do not indicate problems in the functioning of their extremities, but the development of the bilateral characters of the turkeys affected by the activity level, mainly by locomotion, favoring an asymmetric growth of the extremities size (^{Alados et al., 1993}; ^{Palmer, 1996}). In the development of the turkeys, small disturbances in their bilateral characteristics caused by environmental differences, as a result of the competition of the animals by the females, by the food, lodgings or defense from predators, are common, which can cause instability in their development. These disturbances occur in a small part of the organism, so the effects are expected to accumulate on the left or right side (^{Polak and Starmer 2001}; ^{Klingenberg, 2003}). According to ^{Yang et al. (1997)}, ^{Yang and Siegel (1998)} and ^{Campo et al. (2000)}, not all bilateral asymmetries in domestic poultry can be related to FA. Therefore, some indicators of asymmetry evaluated in this study, such as R-L were not consistent when related to the FA, as directional asymmetry indicators.

The angular finger in turkeys is their main support for balance, locomotion or flight and a poor development of this finger modifies their activity, puts them at disadvantage compared to other turkeys and causes stress situations. This study showed interactions (p≤0.01) managementẊsex with RFA, for the length of the coronal finger, but not for the width (p˃0.05) in any of the indicators. There were differences (p≤0.01) with respect to sex for the mean length and width of the middle finger, in RFA for the length (p≤0.03) and width (p≤0.06) of this character, but also in the R-L indicator (p≤0.02). Values were higher for males, compared to females. There was a slight difference (p≤0.03) for the width of this same character in the handling, favoring the caged turkeys.

Regarding types of asymmetry to estimate variation within the individual as a measurement of developmental instability, a correction is required for the average degree of directional asymmetry or antisymmetry (^{Van Dongen, 2006}). There are indications that the three different forms of asymmetry are interrelated and would be transitions from one to other (^{Graham et al., 1993}). However, there is a lack of research on turkeys and it is necessary to conduct studies that relate the different forms of asymmetry with practical applications (^{Lens and Van Dongen, 2000}). Our study showed fluctuating asymmetry (FA), estimated by the R-L difference, in the wing length character and angular finger conformation, in the both management types. In free range handled turkeys, there was a directional asymmetry (DA) for the tarsus length and finger width, showing that the left side was smaller than the right. These differences had a normal distribution with a mean different from zero. In caged turkeys, antisymmetry (AS) was found in the tarsus width and length of the toe in males, and directional asymmetry in the wing length in males and the width of the toe in females.

When analyzing turkey’s behavior respect to their stress levels, we observed that the high levels relate to an increase of the FA with a reduction of their growth rate and reproduction in the turkey populations (^{Koehn and Bayne, 1989}; ^{Sommer, 1996}). FA could provide advantages over other stress bioindicators, because of its easy, inexpensiveness and rapid application (^{Clarke, 1998}). In addition, FA is related to the biological fitness of the females in sexual selection (^{Moller, 1990}), therefore, a change in FA must be biologically relevant (^{Sommer, 1996}).

To obtain greater precision in the character measurement, it is necessary to estimate the repeatability rate or the FA constancy, indicated by ^{Palmer and Strobeck (1986)} as a stress indicator in birds. Table 2 shows the rates of intraclass correlation or repeatability for domestic turkeys in this study. The wing length and the tarsus conformation in the males had superior repeatability values compared to females, regardless of the management system.

Table 2 Repeatability values of five bilateral characters in females and males of autochthonous domestic turkey. 

^{Van Dongen (2006)} mentioned that it is necessary to perform character repeatability analysis so that the observed differences between R-L reflect differences between individuals and not attributed to a possible sampling error. The highest repeatability value was obtained in males, for each of its sides. However, when analyzed, their side’s difference indicator decreased. According to ^{Van Dongen (1998)}, the repeatability estimation is a statistical measure of the consistency between repeated measures of the same character in the same individual, allowing biases correction in such a way that the observed patterns for the FA allow to make inferences on the alleged instability in the turkey development. The FA interactions found in all characters in this study were significant (p≤0.05). ^{Balmford et al. (1993)} in an avian asymmetry study obtained repeatability values of 0.77 (p≤0.001), evidence of the natural selection of tails and symmetrical wings in birds.

There were no significant correlation (p˃0.05) between the absolute value of the asymmetry and the size of the bilateral characters (Table 3). There was a high correlation in turkeys managed in free range regard the tarsus length in females and tarsus width in male’s traits. ^{Leung and Forbes (2000)} mention that if there is a developmental stability component that affects the whole organism, a low correlation between the FA values of the characters would be expected, but usually not significant.

Table 3 Correlation coefficient of the fluctuating asymmetry^† with the absolute value^¶ of the difference of the sides of five bilateral characters, with two types of autochthonous domestic turkey management. 

^†[(D+R)/2]; ^¶ [|D-I|]; * p≤0.05; *** p≤0.001.

Heterophiles relationship: lymphocytes and muscle immobility

There were significant differences (p≤0.05) between the management types. Turkeys in individual cages had a higher H: L ratio (0.55) than those maintained at free range (0.46), but there were no differences (p˃0.05) between sexes (Table 4).

Table 4 Mean and standard error (MEE) of two animal welfare indicators, coefficient of heterophiles: lymphocytes (H:L) and muscle immobility (MI) in autochthonous domestic turkey at Chiapas, Mexico. 

^{a, b} Averages from columns with different superscript indicate significant differences (p≤0.05).

To judge the observed values we can consider the ^{Gross and Siegel proposal (1983)}, that the H/L ratio will be higher as stress intensity increases, and the H/L values of 0.2, 0.5 and 0.8 can be considered as low, optimum and high stress, each. The relationship obtained in our study was moderate or optimal, and there is heteropenia. However, this indicator of leukocyte stress can be used for parameter prediction, since a high ratio of H:L in birds is associated with infections susceptibility (^{Al-Murrani et al., 2002}), slow growth rate (^{Moreno et al., 2002}) and bird’s survival to the next breeding season (^{Lobato et al., 2005}; ^{Kilgas et al., 2006}).

No differences were found between treatments in muscle immobility (p˃0.05), but high test mean times were obtained, indicating a tendency of native turkeys for a notable death feigned reflex (^{Fraser, 1960}). Mean values for muscle immobility duration and heterophils to lymphocytes agree with those reported by ^{Campo and Redondo (1996)} and ^{Campo et al. (2002)}. ^{Campo et al. (2000)} reported a significantly lower H:L ratio in hens caught in a sand bath compared to a control group, however, muscle immobility was similar in both groups. In addition, ^{Campo et al. (2001)} report that poorly feathered hens had a shorter duration of immobility and a higher leucocyte quotient than well-feathered hens.

In most bilateral characters (wing, tarsus and finger) no correlation was found in the relative asymmetry [2|L-R|/(L+R)*100] between muscle immobility and the heterophiles: lymphocyte ratio (Table 5).

Table 5 Relative asymmetric correlation coefficients^† between five bilateral characters (wing, tarsus and finger) and muscle immobility (MI) duration and the relationship between Heterophiles:Lymphocytes (H: L) and the combination of gender by management in autochthonous domestic turkeys in Chiapas, México. 

^†[(D+R)/2]; ^¶ [|D-I|]; * p≤0.05; *** p≤0.001.

The correlation for muscle immobility with the middle finger length in caged males was positive (r=21 %), which can be attributed to male turkeys being stressed, manifesting great fear when induced during the test. ^{Nestor et al. (1996)} found a negative association between muscle immobility and growth and between animal welfare and quality indicators in turkeys selected by weight, where the heaviest were more fearful in the test. Also, in a study with roosters from the Villafranquina breed, ^{Campo et al. (2000)} reported a negative correlation (p≤0.05) between the wing length and the duration of muscle immobility. Besides, there are records of a highly significant relationship (r=52 %) for finger width with the heterophile/lymphocyte ratio of the females that were released.

Most of the bilateral characters with two animal welfare indicators (Table 6) showed no significant correlations, but the wing length in the females in free range had a positive correlation (r=30 %) with the muscular immobility. This indicates a reflex of fake death in these females. For this same character, there was a negative correlation (r = -0.35) with the relation of heterophiles and lymphocytes, of caged females. ^{Moller et al. (1999)} and ^{Campo et al. (2001)} found a positive and significant relationship in the relative asymmetry of different characters and muscular immobility in hens of the Villafrancine breed. A similar value was found in roosters of the Basque breed between relative asymmetry for leg length and muscle immobility duration (^{Campo et al., 2000}).

Table 6 Correlation coefficients between five bilateral characters^†, with the duration of muscle immobility and the ratio of heterophiles:lymphocytes for the combination of sex by management in autochthonous domestic turkey in Chiapas. 

^† [(I+D)/2]; **p≤0.01

Conclusions

The breeding of autochthonous domestic “guajolote” turkeys in family production units, with a traditional management in free range, generates a well-being environment for this species, compared to management in confinement, since every time there is a reduction in its “comfort” space by handling they exhibit states of stress in their behavior. This relates to the reproduction, at the time of the sexual selection. Besides, turkeys present heteropenia and a tendency to display a remarkable feigned death by fear reflex.