Biometric Characterization of the Portuguese Autochthonous Hens Breeds

Simple Summary Autochthonous poultry breeds have been forsaken, forgotten even, since they have always been of less importance in the rural socio-economic context, associated with the domestic economy and, above all, regards from a perspective of self-consumption. The study, protection, improvement and dissemination of breeds, has had an almost inexplicable absence of works on the subject with the first reference to Portuguese poultry breeds in the 30’s of the last century. The biometric study of the breeds is fundamental for the knowledge of the morphological characteristics and the productive potentialities. The aim of this study is to characterize the Portuguese hens breeds (“Pedrês Portuguesa”, “Preta Lusitânica”, “Amarela” and “Branca”), using different biometric measures and live weight and to evaluate, under production conditions specific to the artisanal system, the effect of several factors in each of the studied breeds. The results revealed a high sexual dimorphism and that the “Branca” breed stands out in all the biometric measures. Autochthonous Portuguese hens present morphological traits which would made them more prone to meat production (“Branca”), although the dimensions of certain morphological variables could make them suitable for double-purpose production (“Pedrês Portuguesa” and “Amarela”) and is imperative to consider breeding programs that underline their productive potential. Abstract Promotion of the conservation and preservation of local breed’s biodiversity combined with the concept of sustainable agriculture and development of economically marginal areas are important policies to implement in modern society. The biometric characterization, contributing to maintain phenotypic traits, is a significant tool in breeding programs, which revaluate local breeds, allow the preservation of animal biodiversity and support consumer demands. This paper approaches the biometric characterization of the Portuguese poultry breeds through the study of sexual dimorphism and breed differentiation using six zoometric measures as differentiation criteria. A total of 429 fowl (66 males and 426 females) were studied and the parameters recorded were body weight, body length, chest circumference, shank length, shank diameter and wingspan. A highly sexual dimorphism was evident, in all breeds, with the “Branca” breed being the most zoometrically distant. Concerning Principal Component Analysis, the highly correlations observed between body length, wingspan and shank length, determined the generalized animal form and could be used as selection criteria for improving body size. Breeding programs aiming to preserve these local genetic resources should consider the dual purpose of these breeds: sustainability and cultural legacy, and the offer to urban consumers a source of differentiated high-quality products.


Introduction
Animal biodiversity, the grouping of populations into breeds and domestication have evolved over the centuries, from prehistoric times to the present agricultural sedentary animals and type of production (meat or eggs) [13,15,16]. The equipment is rudimentary, without much technical or technological evolution. The same rusticity is present in the nutritional aspect-farm fodder and the use of surplus or by-products of human or animal feeding complement the reduced needs of these animals, as well as the search for animal protein (insects, worms), in the very typical and leisurely act of etching the soil [13,14,16].
The sustainability of production is becoming an increasingly strong consumer argument for the choice of products and producers that are part of their diet [17][18][19]. The breeding system for indigenous species is balanced by the use of natural resources, land and water, making it environmentally "friendlier", and particularly adjusted in less favored regions where these resources are scarce [16,[20][21][22][23][24].
Traditionally, the production of local chicken breeds (breeding, fattening and then slaughtering of males) has been for gastronomic purposes, in a generally slow-growing and late-maturing period, especially when grown in systems with reduced inputs. The husbandry practices, in this artisanal system, are characterized by the use of rustic animals in free-range conditions with a low capital investment, in a very efficient productive management to develop purposes of high-biological-value protein such as meat and eggs [25][26][27]. In fact, recent studies confirmed the high quality of the eggs of Portuguese native breeds, matching or superseding the quality of commercial breeds product, an interesting opportunity for the recent specialized market niches [28].
This incalculable genetic value, comparable to any other patrimony, even monuments, needs to be characterized, preserved, promoted, disseminated and valued so that future generations can study and know it for their advantage [2,4,11,13,14]. After a long period lacking actions related to the conservation of local genetic resources, with the Genealogical Register in an early stage, breed characterization studies began to be carried out, under a genetic conservation program for its morphological measures, egg production, growth and reproductive performance, and genetic makeup [11].
The aim of this study was to perform the zoometric characterization of the Portuguese indigenous breeds, evaluating the large existing phenotypic variability in these populations and the different productive factors that may contribute to this variability. Zoometric traits play an important role in the live weight prediction and subsequently in the performance of animal carcasses [29,30], constituting a high potential economic selection criterion, with a significant impact in the paternal line of autochthonous populations.
The knowledge of the zoometric and productive traits will support the implementation of conservation strategies aimed to ensure the survival of low-efficiency local breeds. In addition to being unprecedented, this data is expected to contribute to the creation of an ICAR working group and, a basis to the implementation and validation of poultry breeding strategies.

Materials and Methods
The trial was carried out in accordance with EU Directive 2010/63/EU; it complied with the Portuguese legislation on animal care (DL n. 113, 7 August 2013), and adhered to the internal rules of the Polytechnic Institute of Viana do Castelo.
All animals, over the age of 6 months, were listed in the Genealogical Register of the respective breed and are originated from 19 explorations in the region considered to be the breeding area (districts of Viana do Castelo, Braga and Porto). These farms are characterized by a small number of animals (less than 50 F) divided into several flocks and usually from different breeds. Each flock has, on average, 1 male for every 10 females. Traditionally, the production of autochthonous chickens has been undertaken for double Animals 2021, 11, 498 4 of 13 purposes: egg production (hens), and breeding, fattening and slaughtering (roosters), with the ideal slaughter weight being achieved in about 9 to 12 months.

Zoometric Measures
Biometric variables were measured and following procedure, according to FAO (2012) guidelines for adult animals (older than 6 months) [31], is shown in Table 1. Quantitative data was obtained using a digital scale, a gauge with 0.02 mm accuracy, and a measuring tape. Body weight (BW) was estimated using the multifunction scale-KERN HDB with a maximum weight of 5 kg and an interval of 5 g.

Data Analysis
Descriptive statistics [mean, standard deviation (SD), minimum/maximum values] were generated for all the variables in the dataset. The animals were grouped in 3 productive cycles-between 180-360 days old (group 1), from 361 to 720 days old (group 2) and more than 720 days old (group 3). The two-way ANOVA test was used to determine the effects of sex, breed and age group for distinct data categories and differences between means were determined by Tukey's test using the general linear model analysis of IBM SPSS Statistics 23.0 [32]. All statements of significance were based on testing at the p. 0.05 level. The Pearson phenotypic correlation matrix was estimated for BW and zoometric measures (ZM) and principal components analysis (PCA), a method of transforming the original ZM in a new set of orthogonal variables (uncorrelated), called principal components (PC), a linear combination of the original variables, was carried out [33]. The PCA has been used as a tool in the assessment of the body conformation which can be conducted to understand of the complex growth process in the bodily dimensions of an animal during the growth period. Results of principal component analysis not only impact the management of animals but also help in conservation and selection of multiple traits by breeders [34].

Sexual Dimorphism and Breed Effect
The morphometric analysis indicated highly significantly (p ≤ 0.05) sexual dimorphism, as shown in Table 2, with the superiority of the roosters' weight and zoometric measures. Concerning breed analysis, as for males, the "Branca" breed roosters were significantly (p ≤ 0.05) the heaviest, largest and biggest in shank diameter, comparing to the other breeds and presented, as well, a tendency (p > 0.05) to a larger chest circumference, a greater shank length and wingspan. Conversely, the "Preta Lusitânica" roosters breed were the significantly (p ≤ 0.05) less robust (lighter, shorter, with a smaller chest circumference) males, and presented (p > 0.05) the shorter shank length, the smallest shank diameter and wingspan. a,b,c,d Different letters in the superindex are indicative of the existence of significant differences among sex or among sex*breed interaction for the zoometric measures and body weight (p ≤ 0.05). If the same letter is present in different sex or sex*breed interaction, within each analyzed measure, then, no significant difference is found. F: female; M: male; AM: "Amarela"; BR: "Branca"; PL: "Preta Lusitânica"; PP: "Pedrês Portuguesa"; N: number; SD: standard deviation; BW: body weight; BL: body length; CC: chest circumference; SL: shank length; SD: shank diameter; WS: wingspan.
No notorious significant differences were observed between hens' breeds ( Table 2), although the "Branca" showed the significantly (p ≤ 0.05) highest diameter and shortest shank. The "Amarela" was the shortest (p ≤ 0.05) hen and with smallest shank diameter, the "Preta Lusitânica" revealed the smallest chest circumference and the "Pedrês Portuguesa" the largest wingspan.

Age Group Effect
When considering the age group effect (Table 3) in males, significant differences (p ≤ 0.05) were observed in the body weight, between all groups, and in the shank diameter, between the first and third groups. In relation to body weight, these results are particularly due to the "Branca" and "Pedrês Portuguesa" breeds' effect contribution. The same observation was found in the female body weight (Table 4), significantly different between age groups (p ≤ 0.05), in particular between the first and the third, in all breeds. A high variability of the body weight was verified, particularly in "Amarela" and "Branca" roosters and "Branca" hens, that could be explained to the recent breeding programs and different practices of production. Additionally, in hens, zoometric modifications were observed (p ≤ 0.05), in the measurements of chest circumference and wingspan, with the enlargement of the chest over age, in the" Branca" and "Pedrês Portuguesa" breeds, and, in relation to wingspan, between the first and third groups of the "Pedrês Portuguesa" hen.   SD  306  389  337  283  354  345  373  505  283  261  349  321  272  388  479  Lower Limit  940  1070  1530  940  1360  1530  1390  1070  1710  1290  1435  1675  1375  1220  1690  Upper Limit  2960  3590  3050  2210  2940  2910  2960  3590  2870  2650  2795  2830  2395  2880

Phenotypic Correlations
The phenotypic correlations between linear body measurements and body weight are given in Table 5, for all birds. All the phenotypic correlations, for all the population. between body weight and the body measurements were positive and highly significant (p ≤ 0.01) ranging from 0.549 to 0.687. High significant (p ≤ 0.01) positive correlations were recorded, for all the animals, between the body length and wingspan (0.76), body weight and chest circumference (0.69), body length (0.66), shank diameter (0.66) and wingspan (0.63).  Similarly, in males (Table 6), high significant (p ≤ 0.01) positive relationships were observed for body weight and shank diameter (0.80), wingspan (0.57), chest circumference (0.56) and body length (0.55), and body length and wingspan (0.58). In females (Table 6), significant (p ≤ 0.01) high positive relationships were obtained for body weight and chest circumference (0.64), and body length and wingspan (0.59).  Table 7 presents the eigenvalues, percentage of the total variance along with the rotated component matrix and communalities of the body measurements. The communalities rep-    Table 7 presents the eigenvalues, percentage of the total variance along with the rotated component matrix and communalities of the body measurements. The communalities represent estimates of the variance in each variable accounted for by the components. It ranged 0.588-0.867, 0.234-0.835 and 0.440-0.855 in total, M and F respectively. The eigenvalues showed the amount of variance out of the total variance explained by each of the factors.   In males, two principal components were extracted with eigenvalues of 3.321 and 0.878 for PC1 and PC2, respectively and accounted for 69.9% of the total variance present in the original variables. PC1 had high positive loadings on body length (0.835), wingspan (0.809) and body weight (0.783). PC2 most highly correlated with shank length (0.844) (Figure 2). In females, the two principal components extracted accounted for 62.8% of the total variance in the original variables with eigen values of 2.809 and 0.809 for PC1 and PC2, respectively. PC1 was most highly correlated with body weight (0.855), chest circumference (0.817) and body length (0.631) and PC2 had high positive loadings on shank length (0.834) and wingspan (0.692) (Figure 3).

Discussion
Several studies of morphometric index are performed worldwide, the vast majority directly related to the breed characterization and conformation [10,20,24,35,36], but the use of zoometric measures as a strategy to facilitate the implementation of conservation policies aimed to ensure local resources survival, is still beginning in avian populations [37], particularly in Portugal [11,13,38].
The morphometric measurements show highly significant sexual dimorphism, due to hormonal growth effects [39], in accordance with several authors [17][18][19][20]22,24,[40][41][42][43][44]. Breed had significant effect (p ≤ 0.05) in males, with the superiority of the "Branca" roosters, being the heaviest, largest and with the highest shank diameter and the "Preta Lusitânica" roosters, the lightest, shortest and with the smallest chest circumference. In females, the two principal components extracted accounted for 62.8% of the total variance in the original variables with eigen values of 2.809 and 0.809 for PC1 and PC2, respectively. PC1 was most highly correlated with body weight (0.855), chest circumference (0.817) and body length (0.631) and PC2 had high positive loadings on shank length (0.834) and wingspan (0.692) (Figure 3).

Discussion
Several studies of morphometric index are performed worldwide, the vast majority directly related to the breed characterization and conformation [10,20,24,35,36], but the use of zoometric measures as a strategy to facilitate the implementation of conservation policies aimed to ensure local resources survival, is still beginning in avian populations [37], particularly in Portugal [11,13,38].

Discussion
Several studies of morphometric index are performed worldwide, the vast majority directly related to the breed characterization and conformation [10,20,24,35,36], but the use of zoometric measures as a strategy to facilitate the implementation of conservation policies aimed to ensure local resources survival, is still beginning in avian populations [37], particularly in Portugal [11,13,38].
Concerning the females, no evident differentiations between breeds were observed, although the "Branca" hen presented the largest diameter and shortest shank. The "Amarela" was the shortest hen, indicating the lower potential for egg production, as body and dorsal lengths along with head length are relevant indicators to measure productivity [45,46].
The results observed also reflect the breeds high rusticity and the recent implementation of the selection and breeding programs. Complementary in agricultural production, breed in traditional systems well adapted to the environment, with low nutritional requirements and productivity [20,37,47], these populations, naturally or due to the absence of human intervention, maintained their particular ancestral characteristics [48,49].
The "Branca" revealed to be a strongly built breed, mainly the roosters, and with robust legs. The dimensions of the leg have been related with the type of production, with animals presenting higher dimensions (both in width and length), being more appropriate for meat production and carnic breeds characteristic [44]. On the other hand, "Branca" breeding program began later, only in 2014, and, for decades, its morphological and productive characteristics and color plumage, led to crossbreeding and higher genetic proximity with commercial lines, heavier and with thick hips.
The "Preta Lusitânica" breed was the smallest hen (shorter, lighter, with a smaller chest circumference) hen, reflecting the population with the least productive characteristics, meat and eggs, more used to cultural and religious practices. Without intervention of selection programs and less human action, this breed is genetically closer to the avian populations' ancestor [12].
Zoometric measures are, generally, stabilized during the first year of age, allowing for the biometric characterization and to contribute to the selection so that breeding programs could be carried out during this period of life. In roosters, the increase of the body weight is accomplished with the strengthening of the shank, and concerning females, beyond the body weight increase, more evident differences were verified in the stronger and robust breeds, "Branca" and "Pedrês Portuguesa", with chest and wing enlargement. The chest circumference variable proved to be a good indicator of meatiness in most poultry species [45,50].
Morphological traits are essential to implement breeding programs and dimensions of certain morphological variables could make them suitable, from a productive point of view, for meat (breast measurement) or egg productions (body and dorsal lengths, head length) [19,22,24,27,37,51]. Autochthonous Portuguese hens present morphological traits which would made them more prone to meat production ("Branca"), although the dimensions of certain morphological variables could make them suitable for doublepurpose production ("Pedrês Portuguesa" and "Amarela") and is important to consider a breeding programs adjustment to underline their productive potential [36,44,46].
The positive and significant correlations among the body measurements observed in all the groups (total, male and female) indicate high predictability among the variables [34,52,53].
The positive relationship between body weight and the body measurements showed that body weight can be predicted from body measurements in fowls [21,33]. The values of communalities computed for all groups confirm that PCA was appropriate for the data sets and the range of communalities (0.643-0.904; 0.546-0.847; 0.360-0.743) were similar (fowl), slightly inferior (males) or inferior to those reported for body measurements of broilers [39,53,54]. The lower communality observed for shank diameter (0.360) and body length (0.625) in hens interpret the body parameters' weakness in the body measurements total variation explanation.
PC1 showed that the morphological traits' variables explain the largest share of total variance, mainly in total and male groups, correlated highly with body length and wingspan, and could be described as "form factor" [39]. In a principal component analysis of body measurements of hens, with PC1 less accounted for the largest variance in the body measurements, high positive loadings on body weight, chest circumference and body length, according to Mendes [54] that reported PC1 high correlation with breast circumference and body weight of Ross 308 broilers. The presence of wide ranges of variation within hens could be explained to the different conditions of human intervention, whether for cultural reasons, in the productive system, or to the agroecological resources adaptation.

Conclusions
The results revealed the high positive correlations between morphometric measurements and its advantages to define conformation, providing a simple practical methodological framework suited for management, characterization and conservation, to be used in breeding and selection programs.
Portuguese autochthonous poultry breeds are of incalculable ancestral value. These recent selection strategies aim to conciliate the indigenous breeding resistance and adaptation to productive potential. The small size of the population and a strong directional selection may greatly affect the genetic diversity, reinforcing the need for definition and characterization.
Policies that support rural livelihoods, promote local genetic resources and value sustainable products, are a contemporary society requirement. Technical data, such as zoometric measures as a tool in biometric characterization supporting either national or international breeding programs, and scientific strategies is mandatory in the implementation of autochthonous breeds selection programs.

Conflicts of Interest:
The authors declare no conflict of interest.