Craniometric and Morphometric Characterization of Ouled Djellal Algerian White Arab Sheep

Even though infrequent in Algeria, zooarchaeological excavations have uncovered sheep skull remains, providing valuable information on animals exploited by ancient civilizations for meat, milk and wool (Kherbouche et al., 2016; Merzoug et al., 2016). However, the lack of modern references on local breeds complicates comparisons with current populations. The skull, an essential part of the skeleton, plays a vital role in taxonomy and understanding changes due to selection (Marzban Abbasabadi et al., 2020; Sathapathy et al., 2022; Sathapathy et al., 2021). Its unique structure in each animal enables species and breed distinction and allows for examining individual differences on a finer scale.
       
The Algerian White Arab sheep, known as “Ouled Djellal”, is the typical sheep of the Algerian steppe and high plains. This breed is the most numerous among Algerian sheep, with about 12 million heads, accounting for 63% of the sheep population (Feliachi et al., 2003). Originating from the town of Ouled Djellal, approximately 100 km Southwest of Biskra, from which it derives its name. Phenotypically, it is characterized by its size, slender appearance, thin head, sub-convex or convex profile, long and hanging ears and white skin. It has long limbs suited for walking and white, refined and less coarse wool. The breed is primarily used for meat production, is well-adapted to the nomadic lifestyle and arid regions and is known for its remarkable long-distance walking ability (Ami, 2014; IANOR, 2007).
       
This study is part of a series of osteo-biometric research on Algeria’s native ruminants, including sheep (Ami, 2014; Guintard and Tekkouk-Zemmouchi, 2010), goats (Guintard et al., 2018; Ridouh et al., 2019) and camels (Babelhadj et al., 2016; Dib et al., 2024). Though the White Arab breed is the most widespread in Algeria, its osteometry is little studied. Ami (2014) focused on its craniometry, while Guintard and Tekkouk-Zemmouchi (2010) studied its metapods. However, no studies have yet explored the potential relationship between cranial measurements and those taken on live animals.
       
This study aims to establish correlations between external body measurements of live animals and craniometric measurements to create a reference for zooarchaeology, where measurements of live animals are unknown. Additionally, it aims to position the White Arab breed among other sheep breeds of French origins (Guintard and Fouché, 2008), categorizing them into groups such as rustic, meat, or dwarf, thereby enhancing understanding of the breed characteristics.

This study analyzed 30 White Arab female sheep, including only apparently healthy individuals. All subjects were collected from Ain Fakroun and Teleghma slaughterhouses (North-east of Algeria) between March 2022 and May 2023. Two age groups were defined: young adults (YA) aged 2 to 4 years and adults (A) over 4 years old. The research was conducted at the Gestion Santé et Productions Animales Research Laboratory in the Institute of Veterinary Sciences of El-Khroub. Before slaughter, eight external measurements on the living animals were taken using a tape measure (in centimeters), including: wither height (WH), scapulo-ischial length (SIL), thoracic perimeter (TP), cannon perimeter (CP); head length (hL), head width (hW), ear length (eL) and ear width (eW) (Fig 1). In addition, live weight (LW, in kilograms) was estimated by barymetry method, according to the equations of Rasamisoa (1988): LW= 0.635 CT - 23.026 et LW = 0.7536 SIL - 19.2234. After slaughter, the heads were collected, numbered for identification, linked to the external data of the animals and then prepared by dissecting the soft tissues. Further, the bones were boiled for several hours, cleaned with running water and air-dried. Each skull was labelled and numbered to match its data sheet to secure the consistency of the data. Skull weight (SW) was recorded using a balance in grams. Sixteen craniometric measurements were taken, listed in Table 1 and illustrated in Fig 2A-D, using a ruler (lengths CL1, CL2), a 0.02 mm precision digital caliper and a thickness compass (CH6) in millimeters, according to the methods of Ridouh et al., (2019) and Von den Driesch (1976).
 

 

       
Two indices (RC5 and RC7) were calculated from cranial measurements. RC5 was calculated by dividing the least frontal breadth by the total length (CB8/CL1) × 100. RC7 was calculated by dividing the least height of the occipital by the total length (CH5/CL1) × 100. The results obtained were recorded in Excel 2019^®.
       
All statistical analyses were performed using R (Team, 2023) via the R Studio 4.3.1® interface. Descriptive statistics mean (m), minimum (min) and maximum (max) were calculated for each age class and the total population (TP). Variability was estimated using the standard deviation (s) and coefficient of variation (CV% = (s/m) × 100), freeing the measure from unit dependence.
       
With a sample of 30 individuals, the Wilcoxon-Mann-Whitney test was first applied, with a significance threshold of p<0.05. Pearson’s correlation coefficient (r) was then evaluated for each pair of variables, using thresholds: 0-0.10 for low correlation, 0.10-0.39 for a weak correlation, 0.40-0.59 for a moderate correlation, 0.60-0.79 for a strong correlation and 0.80-1 for a very strong correlation, along with p-values to assess the significance of the correlations across the population. Then, the 2-way ANCOVA evaluated the homogeneity of the correlation between young adults and adults.
       
Finally, the White Arab breed was compared to French breeds analyzed by Guintard and Fouché (2008), using bivariate analysis CB19 = f(CL2) and categorizing it as either meat, rustic, or dwarf breed.

Univariate analysis
 
Table 2 shows the morphometric description for the two age groups and for the total population. The differences in mean values between young adults and adults were non-significant, except for the wither height (p = 0.04).
 

       
The means values recorded for the White Arab breed in this study, including live weight (47.72 kg), wither height (87.13 cm) and thoracic perimeter (110.40 cm), are comparable to those reported by Guintard and Tekkouk-Zemmouchi (2010), who recorded 45.1 kg, 80.7 cm and 112.6 cm, respectively. Whereas, Djaout et al., (2018) and Dekhili and Aggoun (2013) reported lower wither height values of 79.64 cm and 60.9 cm, respectively. The live weight observed in our study falls within the expected standard range of 42 to 81 kg, although the wither height appears slightly above with values varying from 61 to 82 cm (IANOR, 2007).
       
Table 3 summarizes the craniometric measurements and the two calculated indices. The differences in mean values between young adults and adults were non-significant except for CL31, CB14, CB18 and CH5. Young adults had higher CL31, related to the eruption of the third upper molar and CH5, indicating a higher neurocranium. Adults had higher CB14 and CB18, indicating a wider viscerocranium.
 

       
Overall variability for craniometric measurements was low, with CV% ranging from 3.66% to 15.05%. It has to be mentioned that adults showed higher CV%, in skull weight, CB18 and CL7, indicating a higher heterogeneity and a late bone development.
       
For the craniometric parameters such as total length (CL1), condylobasal length (CL2) and oblique length of the muzzle (CL7), the White Arab breed showed values comparable to or higher than Hasmer and Hasak of Turkey (Can et al., 2022), Xisqueta of Spain (Parés et al., 2010), Female Zell of Iran (Marzban Abbasabadi et al., 2020), Sharri of Macedonia (Jashari et al., 2021), French breeds (Guintard and Fouché, 2008), Iranian Native (Monfared, 2013), Bardhoka Autochthonous of Kosovo (Gündemir et al., 2020), Morkaraman and Tuj of Turkey (Özcan et al., 2010) and South Karaman of Turkey (Özüdoğru et al., 2022). The exception was the Konya Merino of Turkey, which had a greater total skull length (Özüdoğru et al., 2023).
       
However, for the median frontal length (CL10), the White Arab breed presented the lowest value compared to the Bardhoka Autochthonous of Kosovo (Gündemir et al., 2020), French breeds (Guintard and Fouché, 2008), Konya Merino of Turkey (Özüdoğru et al., 2023), Xisqueta of Spain (Parés et al., 2010) and Hasmer and Hasak of Turkey (Can et al., 2022).
       
Regarding widths, such as the least breadth between the orbits (CB10) and the greatest palatal breadth (CB14), the White Arab breed exhibited values comparable to or lower than those observed in the other breeds mentioned. Consequently, the skulls of the White Arab breed are characterized by being longer and narrower with a shorter neurocranium.
       
The values of the two indices were slightly higher in young adults due to increased CL1 in adults. Even though young adults’ heads appeared to be more massive, there was a non-significant difference (p = 0.061).
 
Bivariate analysis
 
Correlations in the total population
 
By the whole of the correlations in external body measurements, in craniometric measurements and between external and cranial measurements, 174 out of 378 correlations were significant.
 
Craniometric measurements
 
Most lengths were moderately to strongly correlated with each other, particularly between CL1, CL2 and CL7 (r>0.79) with significant p-values, except for CL31 with CL10 and CL20. Similarly, most widths were correlated with each other, except for CB8 with CB2, CB3 and CB10 and CB14 with CB2. Additionally, most lengths showed moderate to strong and significant correlations with most widths, e.g., CL34 and CB18 (r = 0.67), as shown in Fig 3A.
 

       
Furthermore, concerning the height measurements, CH6 had significant moderate to strong coefficients with lengths, e.g., CH6 and CL7 (r = 0.78) (Fig 3B).
       
The correlation between CL1 and CL10 was found to be higher in the White Arab breed (r = 0.69) than in the Mehraban sheep of Iran (r = 0.36) (Karimi et al., 2011). Yet, it was lower than in the South Karaman sheep of Turkey (r = 0.84) (Özüdoğru et al., 2022). Additionally, the White Arab breed exhibited correlations between CL2 and CL7 and between CL1 and CB10, with coefficients of r = 0.83 and r = 0.60, respectively. These values were similar to those observed in the Konya Merino sheep of Turkey (r = 0.88) (Özüdoğru et al., 2023) and the Mehraban sheep of Iran (r = 0.60) (Karimi et al., 2011), respectively.
 
Between external and cranial measurements
 
A strong correlation (i.e., r≥0.60) was observed between some external measurements and certain cranial lengths and widths. The strongest correlation and the most significant was found between thoracic perimeter (TP) and condylobasal length (CL2) (r = 0.69) with p<0.0001 (Fig 5C). Significant correlations were identified between live weight and the cranial measurements of CL1, CL2 and CB8, with correlation coefficients equal to 0.43, 0.48 and 0.39, respectively. Thoracic perimeter showed correlations with all cranial lengths (r values ranging from 0.38 to 0.69), the width CB10 (r = 0.36) and the height CH6 (r = 0.45). Cannon perimeter was correlated with most cranial lengths and widths, excluding CL10, CL20, CB8, CH6 and SW. Head measurements were correlated with the cranial measurements CL1, CL2, CB3 and CB10. Additionally, head length was also correlated with CL7, CL10, CB14 and RC5, while head width showed a correlation with CB18 and SW. Ear length (eL) was correlated with CL1, CL2, CL31 and CH6, but eW had more correlations, including with various lengths (CL1, CL2, CL31, CL34), some widths (CB2, CB8, CB18) and height CH6. However, scapulo-ischial length (SIL) and wither height (WH) showed non-significant correlations.
 
Correlations by age
 
Among all the correlations tested, strong relationships (i.e., r≥0.60) were more prevalent in adults (N=168) than in young individuals (N=106) (Fig 4). Out of 168 pairs, 60 were between cranial and external body measurements, indicating that the adult’s skulls align more closely with their body measurements.
 

       
The 2-way ANCOVA results, presented in Table 4, reveal notable differences in correlation strength between adults and young adults for certain pairs of external and cranial variables. These findings highlight the importance of considering age when interpreting the relationships between external and cranial variables. Pairs of variables with stable and moderate correlation coefficients were found within the total population and young adults, while elevated correlations were observed within adults, e.g., between live weight and CB8 (Fig 3D) and ear length and CL1 (Fig 3E). Conversely, other pairs exhibited stable and moderate coefficients for the total population and adults but were elevated in young adults, e.g., between the cannon perimeter and CB18. These results indicate that the strength of correlation between cranial and external measurements varied significantly within the two age groups and across all specimens, suggesting that age, along with other zootechnical factors, significantly impacted some of these relationships.
 

       
However, some pairs showed moderate to strong correlations and remained stable across the three categories (young adults, adults and all specimens), highlighting relationships that were not influenced by age. e.g., between the condylobasal length (CL2) and certain external measurements such as thoracic perimeter, head length and ear width. This finding was particularly interesting given that the Algerian White Arab breed is distinguished by its characteristic head and ear shapes and sizes (IANOR, 2007).
       
The scatter plot, shown in Fig 5, positions the White Arab breed in the upper right quadrant between the rustic and meat categories, highlights its dual quality as both a rustic and meat-producing breed, in accordance with the literature references and confirms its characteristics previously reported by Feliachi et al., (2003) and Chekkal et al., (2015).
 

Our results have shown that it is possible to estimate one craniometric measurement from another measurement taken on skull fragments found in zooarcheological excavations. More interestingly, it is also possible to infer a measurements of the standing animals of ancient civilisations based on craniometric measurements. For example, the thoracic perimeter can be estimated from the Condylobasal length (CL2). Also, significant correlations were more pronounced in adults than in young adults, indicating that adult skulls exhibit more coherence with both external and cranial measurements, particularly with external body measurements. Young adults have a higher neurocranium and are more massive, whereas adults have a wider viscerocranium. The variability observed in adults shows that the White Arab breed is heterogeneous and has a late bone development, at least for the females. This breed’s rustic and meat-producing characteristics have been confirmed from an osteometric point of view and their skulls were found to be the longest compared to other breeds, except for the Konya Merino sheep. A more extensive study, including a sample of their mandibles and at later stage, of males, will provide further details and new perspectives on sexual dimorphism in the White Arab breed.

The authors would like to thank Dr. BOUKERROU Abderrahmane for his assistance in collecting the samples, Dr. KOUTCHOUKALI Hafida for her help in preparing the skulls, Prof. BERERHI El Hacene for his valuable guidance, Prof. LAKHDARA Nedjoua for refining the English translation of the manuscript and all the staff of the Ain Fakroun and Teleghma slaughterhouses for their cooperation.

Funding statement

This research received no funding.

The authors declare that there is no conflict of interest.