Fertility, Hatchability and Prediction of Egg Weight from Egg Quality Indices of Nigerian Helmeted Guinea Fowls and the French Strain

1. Introduction

Poultry production is an integral component of agriculture. It serves as the most important and quickest sources of animal protein and household income, thereby improving family nutrition, health and livelihood of high- and low-level farmers in the rural and urban communities [1,2,3,4]. In Nigeria, one of the traditional poultry species is the helmeted guinea fowl (Numida meleagris) [5]. Economically, the birds are attractive in the tropics due to less demand with respect to diet; rusticity and better adaptation to the traditional poultry system [6,7]. There is high value for eggs of guinea fowl because of their nutrient contents, thick shells, longer shelf life, and the premium prices they attract compared to chickens [5,8]. The bird is a seasonal layer and its breeding activity peaks in the summer period [9,10].

In spite of the benefits derived from guinea fowl, its characterization in the tropics is poor, thereby limiting the exploitation of its full potential for sustainable economic growth and development [7]. In Nigeria, previous research and development endeavours focused on chicken, while other poultry species like guinea fowl have been neglected. Also, reproductive performance (fertility and hatchability characteristics) which is crucial to guinea fowl profitability [11] and egg quality traits of guinea fowl [12,13,14] in the country have not been widely exploited. Such information is vital to unravel the potential of the underutilized poultry genetic resource in order to map out appropriate selection and breeding strategies for improvement in production and productivity of the birds. It has been reported that egg quality indices are useful for breed assessment and standardization [15]. Also, of interest is the establishment of relationship between traits and the prediction of egg weight from egg quality parameters [16]. Additionally, in the egg industry, egg quality parameters affect consumer preference, quality grading, price, fertility, hatchability, weights of the chicks [17]. However, there is dearth of information on fertility and hatchability characteristics and egg quality traits of guinea fowls in Nigeria [5]. Based on this, the current study aimed to characterise the fertility, hatchability and egg quality parameters of indigenous helmeted and exotic guinea fowls. It also predicted egg weight from egg quality parameters of the birds using multiple linear regression and decision tree.

2. Materials and Methods

2.1. Location of the study

The experiment was conducted at the Guinea Fowl Improvement Unit of the Teaching and Research Farm of Landmark University, Omu-Aran in Kwara State. The Unit is co-managed by the University of Agriculture, Abeokuta, Landmark University and Nasarawa State University, Keffi, Shabu-Lafia Campus to facilitate the implementation of the TETfund National Research Grant for guinea fowl improvement in Nigeria. The farm lies on latitude 8.9ºN and longitude 50.61ºE within the guinea savannah zone of North Central Nigeria.

2.2. Experimental procedure

The base population of the indigenous guinea fowls came from randomly selected birds from three agro-ecological zones of Nigeria, namely the Sudano-Sahelian, Southern Guinea Savanna and Tropical Rainforest as reported earlier [18]. The exotic guinea fowls of the French strain were imported from the neighouring Benin Republic. After acclimatization for two weeks, a total of 300 randomly selected adult guinea fowls (8 months old) comprising 150 indigenous (30 males and 120 females) and 150 French strain (30 males and 120 females) were utilized in the study. Each genetic group was replicated three times in a standard poultry house using a mating ratio of 1:4 as adopted by Atawalna et al. [19]. The birds were fed a diet containing 16.0% CP and 2650 ME (Kcal/kg). Feed and clean drinking water were supplied ad-libitum. Other standard rearing practices and medication were carried out while biosecurity measures were taken. The entire study lasted three months (July-September, 2020).

2.3. Collection of eggs for hatching

Eggs laid in the third week of August were collected which comprised 120 each from the indigenous guinea fowl and the French strain. The eggs were labelled with the code given at the time of collection based on genotype. A total of 240 eggs were set in an automatic incubator. The incubation temperature was set at 37.5°C while the humidity was 45-70%. The essence was to guarantee good embryo development and maximum hatchability [20]. Turning device was adjusted in accordance with the manufacturer’s recommendations. Candling was done on the 7^th, 14^th and 18^th day of incubation. Hatching of eggs was done on the 28^th day of incubation.

2.4. Reproductive assessment

Eggs’ fertility was determined using the following equation:

$$\mathrm{Fertility}=\frac{\mathrm{No}.\mathrm{of}\mathrm{fertile}\mathrm{eggs}}{\mathrm{No}\mathrm{of}\mathrm{eggs}\mathrm{set}}\times 100$$

Hatchability of fertile eggs was calculated as indicated below:

$$\mathrm{Hatchability}=\frac{\mathrm{No}.\mathrm{of}\mathrm{hatched}\mathrm{keets}}{\mathrm{No}.\mathrm{of}\mathrm{fertile}\mathrm{eggs}}\times 100$$

2.5. Egg quality assessment

A total of 240 clean eggs with no defects comprising 120 each from the indigenous guinea fowl and the French strain were utilized for the egg quality assessment. The seventeen egg quality parameters assessed included egg weight, egg length, egg width, shell thickness, shell weight, egg shell index, egg shape index, yolk weight, yolk height, yolk diameter, yolk ratio, yolk index, albumen weight, albumen height, albumen diameter, yolk/albumen ratio, and haugh unit. Egg weight was taken using a sensitive balance with ±0.01 g accuracy [17,21]. Egg length and egg width were measured using a digital caliper to the nearest 0.01 mm [22]. Egg shape index was determined as the ratio of egg width to egg length multiplied by 100 [23]. The shell thickness was taken as the average of the thicknesses of blunt, middle and sharp points of the egg and was measured using a micrometer gauge [12]. Shell weight, after drying at room temperature for one day [24], was determined using a sensitive balance with ±0.01 g accuracy [25]. Egg shell index was obtained from the ratio of shell weight to egg weight multiplied by 100 [26].

Albumen height and yolk height were measured by tripod micrometer (accuracy of 0.01 mm) after the albumen of the broken eggs was carefully separated from the yolk [11,22]. Yolk weight was measured by using a sensitive balance [21] while albumen weight was calculated as: Albumen weight (g) = Egg weight – (yolk weight + shell weight) [24]. Albumen and yolk diameters were determined using a digital caliper of 0.01 mm accuracy [21]. Yolk index was taken as the ratio of yolk height to yolk diameter multiplied by 100 [24]. Yolk ratio was calculated as ratio of yolk weight to egg weight [27]. The yolk/albumen ratio was determined as ratio of yolk weight to albumen weight [24]. Haugh unit [21,28] was calculated using the following equation by Raymond Haugh in 1937:

HU= 100* log (H+7.57)- (1.7*W0.37)

Where,

HU= Haugh unit; H= Albumen height (mm); W= Egg weight (g)

2.6. Statistical analysis

Descriptive statistics were computed for fertility and hatchability parameters. In order to find the differences between the indigenous guinea fowls and the French strain, egg parameters were subjected to t-test. Significant differences were declared at P<0.05. The egg quality parameters of local and exotic guinea fowls were subjected to Pearson’s product-moment correlation. Principal component analysis (PCA) was carried out to explore hidden relationship between egg quality traits [29]. This was to permit appropriate grouping of the guinea fowls based on genotype. The criterion used for the extraction of the PCs was Eigenvalues greater than 1 [18]. Egg weight (dependent variable) of the local and exotic guinea fowls was predicted from other egg traits and their PC scores (independent variables) using multiple linear regression model [16,30,31]. The multiple linear regression involved the use of the stepwise method (where the independent variables included the PC scores) as well as the ridge method (the independent variables did not include the PC scores). In the ridge regression, the regularized CATREG procedure was used. The conditions were set at: supplementary object= 17, validation type=crossvalidation (10 folds), convergence =0.0001 and maximum iterations = 100. Egg weight prediction was also done without PC scores using decision tree model [Chi-square automatic interaction detection (CHAID)] as described by Orhan et al. [32] and Portillo-Salgado et al. [16]. The Bonferroni adjustment and ten-fold crossvalidation were applied in CHAID [18]. All analyses were done using IBM SPSS [33].

3. Results

3.1. Reproductive performance

The fertility and hatchability values of the Nigerian indigenous guinea fowls and their exotic counterparts are shown in Table 1. Percentage fertility (90.0 versus 73.3%) and hatchability (66.7 versus 56.5%) were higher in the exotic birds compared to their indigenous counterparts

3.2. Egg quality parameters

The egg quality traits of the Nigerian indigenous and exotic guinea fowls are presented in Table 2. Higher values (P<0.05) were obtained in exotic birds with the exception of egg shell index where the values were higher in their indigenous counterparts (18.88±0.79 versus 16.41±0.69) and haugh unit (92.37±3.13 versus 91.09±3.22).

3.3. Relationships between egg quality traits

The phenotypic correlation coefficients of the egg quality parameters of the Nigerian indigenous and exotic guinea fowls are shown in Table 3. In the indigenous birds, egg weight (EW) was highly (P<0.01) correlated with EWD (0.97), SW (0.95), YW (0.93), AW (0.92), ST (0.90), ESI (0.76) and EL (0.60). With regard to the exotic birds, the correlation (0.96) between EW and EWD was also significantly (P<0.01) highest, followed by AW (0.94), YW (0.92), SW (0.90), EL (0.89) and ST (0.88). The correlations among other egg parameters range from high and low, and positive to negative values in both genetic groups.

3.4. Variables’ contributions to variation and loadings on the principal components

The eigenvalues, percentages of the total variance and the communalities of the egg quality traits of the Nigerian indigenous and exotic guinea fowls are presented in Table 4. The communalities ranged from 0.574-0.978 and 0.590-0.987, respectively in the indigenous and exotic birds. Three principal components (PCs) were extracted from the indigenous guinea fowls with eigenvalues of 5.949 (PC1), 2.148 (PC2) and 1.825 (PC3), accounting for 82.7% of the total variance. In their exotic counterparts, three PCs were also extracted with eigenvalues of 5.739 (PC1), 2.164 (PC2) and 1.799 (PC3), explaining 80.8% of the generalised variance. The reliability of the PCA was confirmed using the Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy (KMO = 0.66 and 0.64, respectively for indigenous and exotic birds) and Bartlett's Test of Sphericity (chi-square= 2993.944 and 2944.584; P<0.01, respectively for indigenous and exotic birds).

3.5. Principal component factor score coefficients

The principal component factor score coefficients of the Nigerian indigenous and exotic guinea fowls are presented in Table 5. These factor scores could be used instead of the original interdependent egg quality traits in estimating the egg weight of guinea fowls.

3.6. Egg weight prediction using stepwise multiple linear models

The prediction models of egg weight from original egg quality traits and on their principal component factor scores in the Nigerian indigenous and exotic guinea fowls are presented in Table 6. Four models were obtained in the indigenous birds from the stepwise regression of egg weight using the original egg indices as predictors. Egg width was the sole predictor in the first model; egg width and shell weight were the explanatory variables in the second model; egg width, shell weight and egg shell index were the independent variable in the third model while egg width, shell weight and egg shell index yolk/albumen ratio were the predictors in the fourth model. The respective coefficients of determination (R²) to estimate reliability of the models were 0.936, 0.958, 0.999 and 0.999. Two principal component models (PC1; PC1 and PC2) were obtained to predict egg weight with R² values of 0.908 and 0.953, respectively. In exotic birds, however, three models were obtained for the estimation of egg weight from original egg indices. The first model solely contained egg width as the predictor; egg width and shell weight were the independent variables extracted to predict egg weight in the second model while the third model contained egg width and shell weight and yolk/albumen ratio. The R² values were 0.927, 0.948 and 0.959, respectively. Here, three principal component models (PC1; PC1 and PC2; PC1, PC2 and PC3) were obtained to predict egg weight with R² values of 0.913, 0.950 and 0.952.

3.7. Egg weight prediction using ridge regression

The estimation of egg weight using ridge regression model in Nigerian indigenous guinea fowls and the French strain is presented in Table 7. In the indigenous birds, nine variables (egg length, egg width, shell thickness, shell weight, egg shell index, egg shape index, yolk weight, albumen weight and yolk index) were found to be significant (P<0.05; P<0.01) in predicting egg weight. The coefficient of determination (R²) was 0.959. However, in the French strain, the significant (P<0.05; P<0.01) explanatory variables were egg length, egg width, shell thickness, shell weight, egg shape index, yolk weight, albumen weight, yolk ratio and yolk/albumen ratio.

3.8. Egg weight prediction using CHAID model

The CHAID model for the prediction of egg weight is shown in Figure 1 and Figure 2, respectively. In the indigenous guinea fowl, egg width (>33.50 mm) was found as the most important variable to estimate egg weight, and together with albumen weight (>17.36 g) in terminal node 8 made a better prediction (optimal egg weight was 36.23 g). The resubstitution estimate (0.20) and standard error (0.03) were low. 2.866 while the prediction accuracy was high (R² =0.930). With regard to the French strain, egg width (>38.63 mm) was also found as the most important variable to estimate egg weight, and together with albumen weight (>21.75 g) in terminal node 8 made a better prediction (optimal egg weight was 43.87 g). The resubstitution rate (0.26) and standard error (0.04) were also low while the prediction accuracy was high (R² = 0.917).

4. Discussion

The exotic birds performed better than their indigenous counterparts in terms of hatchability and fertility. This may be as a result of differential genetic constitution. According to Zeleke et al. [13], the differences among genetic groups of guinea fowls in terms of fertility and hatchability could be a reflection of previous interventions on selection and breeding. However, the values obtained in the indigenous stock are comparable to the hatchability range of 58.9-76.7% and fertility rate of 49.2-70.0% reported by Obike et al. [34]. In the current study, the fertility value of the Nigerian indigenous stock was higher than the value of 55.97% [35] and 57% reported by Zeleke et al. [13]. Also, the indigenous stock had higher hatchability value compared to the 50.4% reported by Atawalna et al. [36]. However, a lower hatchability percentage was obtained in comparison with the hatching rate of 72.92 and 82.2% reported by Yakubu et al. [35] and Adu-Aboagye [37], respectively. In other studies, fertility of guinea fowl eggs has been reported to range from 70 to 85% [38,39], while hatchability ranged from 72 to 80% [40,41]. The observed variations in the indigenous guinea fowls of Africa could possibly be due to environmental factors (as they were reared in different agro-ecologies), varying management systems, health status, sex structure, age and weight of guinea hens, egg size and incubator conditions. Also, the guinea fowls of the current study were naturally mated; there is every possibility that if artificial insemination is adopted, better results would be obtained. This is congruous to the submission of Hudson et al. [42] that artificial insemination is a veritable tool in the improvement of guinea fowl’s fertility and hatchability.

Guinea fowls are highly valued for their eggs because of their nutritional and health benefits. On the average, eggs from the exotic birds seemed to be better than those from their indigenous counterparts in most parameters investigated. The superior edge of the egg quality characteristics of the exotic birds, especially egg weight is not surprising considering the fact that the birds have been subjected to artificial selection. According to Krunt et al. [15], higher egg weight is a product of the selective process involving birds of superior advantage in terms of performance. Therefore, there is every possibility that if the indigenous birds are upgraded using the exotic gene pool, there may be an improvement in the egg quality parameters with emphasis on egg weight, which has been reported as a crucial indicator for breed standardization, quality grading and consumer evaluation [43,44].The mean egg weight value of 41.50 g in exotic birds is comparable to the value of 40.37 g reported for French Broiler Guinea Fowl guinea fowl strain in Kastina State, Nigeria [45], but less than 43.44 g and 51.68 g reported by Krunt et al. [15] and Kouame et al. [46], respectively. The egg weight value of 34.1 g in indigenous birds is lower than the 37.3 and 37. 5 g reported for guinea fowls in Sarki and Birnin Kebbi, Nigeria by Idowu et al. [14]. However, apart from genetics, other non-genetic factors such as age, nutrition and system of management can influence egg quality parameters [47,48].

The estimates of correlation are comparable to those reported in similar studies [16,24,49,50]. The strong and positive association of egg weight with egg width, egg length, shell, yolk and albumen weights, and shell thickness are in consonance with the submissions of earlier workers [24,27]. The negative correlation between egg weight and haugh unit is consistent with the findings of Bernacki et al. (27) and Khaleel (28). The strong relationship existing between egg weight and some egg parameters may be useful as selection criterion, as it is possible that they under the same gene action. This, therefore, provides a basis for the genetic manipulation and improvement of the indigenous guinea stock for better egg quality traits. High correlation coefficients among the variables also make it possible to predict egg weight from egg quality parameters.

Three principal components were extracted (out of a total of sixteen original independent variables), which were able to account for a good percentage of the generalized variance in the egg quality parameters investigated. These, according to Malfatti et al. [51], can be used to assess the relationship between the different egg quality indices as well as assignment into various groups. The stepwise regression revealed the importance, especially of egg width in the prediction of egg weight. However, due to the problem of multicollinearity, the ordinary least squares method (stepwise regression) estimates may be biased compared to estimates from the principal component factor scores and ridge regression [30,52]. Also, in the CHAID model, egg width singly and in combination with albumen weight was the best in the prediction of egg weight in indigenous guinea fowl and the French strain. The CHAID model, which is non-parametric, does not impose assumptions on the independent variables compared to the multiple linear regression. In this study, CHAID model was more consistent in the estimation of egg weight in both genetic groups. Therefore, it might be an indispensable tool in the poultry industry as regard egg quality classification. The is in accordance with earlier reports [32,53]. Egg width can easily be measured. This becomes imperative as this information could be exploited in estimating egg weight where resources are limited. The prediction of egg weight from other egg quality parameters has been reported [16]. Albumen weight as observed in the present study has also been reported to be a good indicator of egg weight [27].

5. Conclusions

The exotic birds performed better than their indigenous counterparts in terms of hatchability and fertility. Percentage fertility and hatchability were higher in the exotic birds compared to their indigenous counterparts. Three principal components sufficiently accounted for the variations in the egg quality traits of both genetic groups. The multiple linear regression and the CHAID models revealed the importance, especially of egg width in the prediction of egg weight. The indigenous stock performance can be improved in terms of egg quality parameters, especially egg weight by upgrading using the exotic guinea fowls. The obtained CHAID model with egg width as the sole predictor can be used to estimate egg weight where there is no availability of digital weighing scale.