Study areas

The study was carried out in the Gamo and the Gofa zones. Three districts were selected from each zone: Kucha, Daramalo and Dita districts from Gamo zone, as well as Zala, Denbagofa and Oyda districts from Gofa zone (Figure 1). Weather and agroecology-related information of the selected districts are presented in Table 1.

Site and animal selection

Representative samples of Gamo and Gofa cattle breeds were selected from their respective breeding areas. Information on their breeding regions and distribution zones was gathered using secondary sources. The Gamo cattle breed is reported to be native to the Gamo zone, with its distribution extending into the neighbouring Gofa zone (Chebo et al., 2013; Rege & Tawa, 1999). Thus, three districts – Kucha, Daramalo and Dita – were randomly chosen from Gamo zone to represent the indigenous Gamo cattle. Similarly, Gofa cattle are reported to be primarily found in Gofa zone, with their distribution reaching into the neighbouring Gamo zone (Kebede et al., 2017; Rege & Tawa, 1999). Accordingly, three districts – Zala, Denbagofa, and Oyda – were randomly selected from Gofa zone to represent the indigenous Gofa cattle. From each district, two sampling sites (known as ‘Kebeles,’ the smallest administrative units) were randomly chosen. Twenty-five households-raising cattle were then randomly selected from each sampling site. From each household, two unrelated adult cattle, aged four years and older, were randomly chosen. One male cattle was sampled every two households within each kebele. These animals were carefully monitored by their owners and trained labourers. Aggressive cattle that were unable to stand properly on flat ground were excluded from measurements.

Data collection

Morphometric and morphological data were collected following theFAO (2012) guidelines. Data collection was conducted in the morning to minimize the effects of feeding and watering on the measurements. Three researchers were involved in the data collection process: two handled the morphometric data, while the third recorded the morphological data. To minimize bias, the same researchers performed the data collection in all sites throughout the study. The animals were measured using a textile measuring tape in centimetres. A total of 600 cattle (486 females and 114 males) were subjected to nine morphometric measurements (Table 2) and 11 qualitative (morphological) traits (Figure 2, Table 3). For data analysis, the cattle were grouped into three age categories based on the classification by Tatum (2011): group one (3–5 years), group two (6–7 years) and group three (8 years and older).

Qualitative characteristics

The coat colour of the two studied breeds and sexes is presented in Figure 2. The coat colour was not significantly affected by breed, but it was affected by their sex. The majority of the studied cattle possessed red and light red coat colour. Sex-wise results revealed a high proportion of red-coloured females, while males possessed a light-red coat colour.

The effect of breed and sex on the qualitative characteristics of Gamo and Gofa cattle breeds is presented in Table 3 along with the respective chi-square values and levels of significance. Relatively, breed affected more morphological traits than sex: three out of the ten morphological traits were affected significantly by the breed of the animals, while sex affected only one trait. Accordingly, the majority of the studied cattle populations possessed uniformly-patterned coat colour, upward-oriented straight-shaped horns with black colour, laterally-oriented round edge ears, straight face profile, small hump size, short coat hair size, and medium tail length.

Morphometric traits

Table 5; Table 4 present the least square means, standard errors and pairwise comparisons showing the effect of breed, sex and age on the morphometric traits of the studied cattle populations. Breed affected two out of the nine morphometric traits, with Gofa cattle exhibiting greater pelvic width and horn length measurements compared to Gamo cattle. However, the effect of breed on the morphometric measurements was found to be sex-dependent. Breed significantly affected pelvic width and horn length measurements of the females, with values higher in Gofa females. However, these measurements of the male cattle populations were not significantly affected by breed differences. On the other hand, the heart girth measurement of the males was significantly affected by breed, with Gamo males having higher measurements than their counterparts from Gofa. However, the heart girth measurement of the females was not significantly affected by breed.

Similarly, sex affected seven out of the nine morphometric traits, with males exhibiting larger measurements than females in most of the traits except for ear and horn length. Age affected five of the nine morphometric traits, with most measurements increasing as the animal grew older. Body length and heart girth measurements were found to be stable after the age of six years. However, muzzle circumference and horn length measurements of middle-aged animals were found to be the highest among the compared age groups.

Multivariate analysis

Canonical discriminant analysis

Canonical correlations and eigenvalues for both male and female cattle populations are shown in Table 8. In line with the low partial R-Square and F-value outputs in Table 6, the eigenvalues were also small enough to discriminate between the two cattle breeds in both sexes. However, relatively higher Eigenvalues were observed for males than females. Similarly, the canonical correlation, which was used to build the canonical variate 1 (Can 1) from the used traits, was also low. However, a relatively higher canonical correlation was observed for males than females.

Table 8: Multivariate statistics outputs of the canonical structures.

Multivariate Statistics	Females	Males
Canonical correlation	0.5140	0.7087
Eigen value	0.3591	0.9060

Pairwise squared Mahalanobis distances between the breeds were calculated as 1.43 for females and 3.68 for male cattle, indicating that males were more distantly related than females. However, both distances were small and insufficient to indicate a significant distance between the breeds. The overall multivariate analysis results showed low and non-significant distances between Gamo and Gofa cattle breeds.

Plots of the first two canonical variables to discriminate the cattle breeds are presented in Figure 3. In line with the result of the Mahalanobis distances, the studied Gamo and Gofa breeds were found to be inseparable and categorized in the same group, while a relative separation was observed between the males.

Qualitative characteristics

Qualitative traits, due to their easily observable nature, are valuable for distinguishing between cattle breeds. Coat colour and coat colour patterns are among the most easily observed traits used to differentiate breeds. However, in this study, these traits did not distinguish the Gamo and Gofa cattle breeds, as most animals from both breeds exhibited uniformly patterned red and light red coat colours. The similarities in coat colour and pattern, along with other morphological and morphometric similarities between the breeds, may suggest genetic relatedness ( (Mustefa et al., 2021) for Raya cattle; (Getachew, Abegaz, Misganaw, & Fekansa, 2014) and (Mustefa et al., 2023) for Ogaden cattle). Therefore, the observed similarities in coat colour and pattern between Gamo and Gofa cattle imply phenotypic resemblance, which may reflect underlying genetic similarities. These findings, however, should be validated through genetic analysis.

The red-dominant coat colour observed in this study aligns with the findings of Rege and Tawa (1999), who identified red as the primary coat colour in Gofa cattle. Similarly,Kebede et al. (2017) also reported that red and white were the most common coat colours in Gofa cattle. In agreement with these findings,Chebo et al. (2013) reported that dark and light red were the predominant coat colours in Gamo cattle. The farmers’ preference for red coat colour in both breeds may be linked to farmers' selection criteria, as red is a preferred colour in the studied areas. According to Kebede et al. (2017), coat colour was a significant selection criterion for farmers, following milk yield.

Similarities in other qualitative traits, such as horn, ear, hump, face, hair, and tail lengths, were also observed between the Gamo and Gofa cattle breeds, which challenges their classification as distinct breeds. The minor differences observed could be attributed to variations within the breeds. Such intra-breed variations across different sampling locations were also reported byTerefe, Dessie, Haile, Mulatu, and Mwai (2015) in Mursi cattle,Mustefa et al. (2021) in Raya cattle, andMustefa (2023) in Harar cattle.

Morphometric traits

Morphometric measurements, in conjunction with qualitative traits, provide reliable information for assessing the degree of relationship between breeds. Most of the morphometric measurements were similar between the two cattle breeds, which supports the observed qualitative similarities. Consequently, no significant differences were found between the Gamo and Gofa cattle breeds that could serve to differentiate them. As mentioned in the qualitative section, the slight differences observed may reflect within-breed variations. These variations are crucial for designing conservation and genetic improvement programmes.Terefe et al. (2015) in Mursi cattle,Mustefa et al. (2021) in Raya cattle, andMustefa (2023) in Harar cattle also reported morphometric variations within the same breed across different locations.

In comparison with the previous study on Gofa cattle by Kebede et al. (2017), the Gofa females in this study showed similar body length, heart girth and wither height. However, they had larger muzzle and hock circumferences, and lower ear and horn lengths. In contrast, the Gofa males in this study exhibited comparable body length but lower values for other morphometric measurements. These findings suggest a reduction in body size of Gofa males over the past seven years, possibly due to negative selection practices by farmers. Similarly, most of the morphometric measurements for the cattle breeds in this study were lower than those reported byChebo et al. (2013) for Gamo cattle, but comparable to those reported by Zeleke et al. (2017).

The Gamo and Gofa cattle breeds were found to be smaller than many other Ethiopian indigenous cattle breeds. This observation is consistent with (Rege & Tawa, 1999) description of Gofa cattle as the smallest strain of Ethiopian cattle. Their morphometric measurements were smaller than those of breeds such as Afar (Tadesse, Ayalew, & Hegde, 2008), Begait (Ftiwi, 2015), Begaria (Getachew et al., 2020), Fogera (Girma, Alemayehu, Abegaze, & Kebede, 2016), Gojjam Highland (Getachew & Ayalew, 2014), Harar (Mustefa, 2023), Kereyu (Nigatu & Tadesse, 2020), Mursi (Terefe et al., 2015), Nuer (Minuye, Abebe, & Dessie, 2018), Ogaden (Mustefa, 2023) and Raya cattle (Mustefa et al., 2021). However, their measurements were larger than those of Abergelle and Irob cattle breeds (Zegeye, Belay, & Hanotte, 2021). Similar morphometric traits were also observed in Arado (Genzebu et al., 2012) and Horro cattle (Bekele, 2015).

Multivariate analysis

High partial R-Square and F-values are necessary to demonstrate significant discrimination between populations. Additionally, low error rates are essential to show the distinctiveness of separate breeds. However, the results from the stepwise analysis (Table 6) exhibited low partial R-Square and F-values, indicating weak discriminatory potential of the morphometric traits. Furthermore, high error rates (Table 7) were observed, indicating greater similarities between the cattle populations, which lowers the likelihood of classifying the breeds into separate clusters. In contrast, lower error rates would indicate distinct differences between the breeds. For example, an error rate of 1% was reported in the classification of phenotypically unrelated Harar and Ogaden cattle breeds (Mustefa, 2023).

An eigenvalue greater than 1 is required for breed discrimination. If the value falls below 1, discrimination between the studied animals is not significant. The lowest eigenvalues observed in both sexes (Table 8) failed to distinguish the cattle populations into separate clusters, suggesting the absence of two distinct breeds. Similarly, high Mahalanobis distances between breeds are needed for clear cluster separation, but the Mahalanobis distance results in this study were low, with males showing slightly higher distances. This could be attributed to the smaller sample size of oxen. The accuracy of the analysis improves with larger sample sizes. Due to the low eigenvalue (< 1) and short Mahalanobis distances in the multivariate analysis, the studied cattle breeds were found to be phenotypically inseparable. However, phenotypic similarities do not necessarily imply genetic similarities between breeds (Zechner et al., 2001).

Effect of sex and age

Sex and age had minimal effects on the qualitative traits of the studied cattle populations since qualitative traits are typically controlled by fewer genes (Falconer, 1989). Therefore, successive planned selection activities are required to bring changes to the qualitative traits. On the other hand, the morphometric traits were influenced by both sex and age, because quantitative traits are influenced by a greater number of genes (Falconer, 1989). This means a few natural or artificial selection activities can produce significant changes to the quantitative traits.

Accordingly, males were generally larger than females in most morphometric traits of both breeds, aligning with Rensch’s rule (Rensch, 1950), which suggests that females of a species are usually smaller than males. These differences could be attributed to testosterone, which promotes the development of skeletal and muscle mass in males (Baneh & Hafezian, 2009). The effect of the endocrine system on growth was also significant, with estrogen's influence on growth being more limited in females (Baneh & Hafezian, 2009; Chriha & Ghadri, 2001). Similar findings, showing male dominance in size, were reported byMustefa (2023) for Harar and Ogaden cattle breeds,Mustefa et al. (2021) for Raya cattle, andTerefe et al. (2015) for Mursi cattle.

Age significantly affected five morphometric traits. Body length and heart girth measurements showed stable body development from the middle-aged group, while muzzle circumference and horn length measurements indicated the middle-aged group as optimal for these traits.

Significantly different breeds need to be registered separately, while the same breed should be registered only once. This is because our next step as a country is to design breeding programmes that include both conservation and genetic improvement activities for each breed individually. Therefore, conducting these programmes separately for breeds without significant differences would be inappropriate. The currently observed differences among these breeds can be considered as within-breed variation; however, this needs to be supported by further genetic characterization studies.