Phenotypic characterization of indigenous sheep breeds in the Jimma Zone, Oromia, Ethiopia

ABSTRACT

The objective of the study was to undertake phenotypic characterization of sheep, in their production environment. A total of 570 sheep were used in this study for body measurements. After the data were collected, using the Proc GLM model, quantitative data were analysed by SAS version 9.3 (2014) and qualitative data were analysed by SPSS. The dominant coat colour pattern observed was 81.1% plain, 16.4% patchy and sheep with spotted pattern (2.5%), respectively. Brown (43.5%), fawn (17.5%) and red (13.0%) were the most frequently observed coat colour types. The majority (91.1%) of sheep were polled whereas (8.9%) of the sheep were horned. The effect of the district on body weight and most of the linear body measurements were significant (p < 0.05) except head length, canon bone length, ear length and tail length. Chest girth explains more variation than any other linear body measurements in both ewes (94%) and rams (93%). The prediction of body weight based on regression equation y = −21.82 + 0.68x for female sheep and y = −49.90 + 1.08x for male sheep, where y and x are body weight and chest girth, respectively. In general, it could help as an input for efficient utilization, conservation and improvement in the future.

KEYWORDS:

• Phenotype
• characterization
• linear measurement
• correlation
• quantitative

1. Introduction

The livestock sector in Ethiopia contributes 12% and 33% of the total and agricultural Gross Domestic Product (GDP), respectively, and provides livelihood for 65% of the population (FAOSTAT 2013). The sector also accounts for 12–15% of the total export earnings (Tewodros 2015). These livestock genetic resources are very important to the development of the economic, social and environmental of one country (ESAP 2004).

Sheep production is among the most important agricultural activities in the highlands of Ethiopia where crop production is unreliable (Kocho 2007). Approximately 75% of the sheep are kept in small-scale mixed farms in the highland regions, which cover regions of over 1500 m.a.s.l. altitude and receive over 700 mm of annual rainfall, while the remaining 25% are found in the lowlands. Thus, Ethiopia is one of the largest African countries for sheep resources which have long genetic diversity for the livelihoods of the rural poor (Abegaz 2007). According to Gizaw et al. (2007), sheep are the second most important livestock species next to cattle with nine diverse breeds in Ethiopia. They have adapted to a range of environments from the cool alpine climate of the mountains to the hot and arid pastoral areas of the lowlands (Mirkana 2010). Currently, the total sheep population in Ethiopia has been estimated to be about 42.9 million, out of which 99.52% of sheep belong to indigenous breeds (CSA 2021).

Sheep are the living banks for their owners and source of immediate cash and insurance against crop failure (Tibbo 2006). Thus, sheep play an important economic role and make a significant contribution to both domestic and export markets through the provision of food (meat and milk) and non-food (manure, skin and wool) products. They also play a major role in the food security and social well-being of rural populations living in extreme poverty (Duguma et al. 2011). They are relatively drought tolerant, small in size, easily manageable and saleable resources.

But, their performance is poor, so there is a need to improve their productivity through selection and breeding (Wollny 2003). Genetic improvement of the local livestock through appropriate techniques or selection and breeding programme is the need of the day (Yakubu 2010). The environmental pressure also maintains a wide range of genotypes adapted to a specific set of circumstances. According to Solomon (2008) there are nine known sheep breeds in Ethiopia; however, there is no clear phenotypic or genetic evidence to show their names rather they were represented by their geographical locations. Most often those locations are believed to be home tracts to those breeds or ecotypes (Mirkena 2010). To make the best use ofsheep-keeping operation, it is important and a prerequisite to have a comprehensive understanding of the whole situation through assessing the production environment, the production and productive and adaptive characteristics of the sheep breeds (Sisay 2009). Because characterizing of the production system is the first step to designing any genetic improvement programme (FAO 2012).

Phenotypic characterization is used to describe and classify breeds of farm animal species (Traore et al. 2008). It includes information on population size, flock size, composition, information on the production environment and husbandry conditions, which are known to play vital roles in trait expression (Mesfin 2015). The first phase of characterization is surveying to identify populations based on morphological, geographical distribution, uses and husbandry and production environments (Traore et al. 2008). The usefulness of breed characterization of indigenous livestock particularly sheep is never doubted, because characterization is used to inventory and monitor animal genetic resources and is essential to their sustainable management and effective planning of how and where they can be used and developed (FAO 2015). Because these different studies have been carried out on the characterization of indigenous sheep breeds (types) in Ethiopia like Abegaz (2002), Solomon (2008), FARM-Africa (1996), Markos (2006), Mengiste (2008); Gizaw et al. (2008), Shigdaf et al. (2009), Tesfaye (2008) and other researchers have been done on phenotypic characterization of indigenous sheep in a different part of Ethiopia. Utilizing animal genetic resources efficiently and optimally is crucial for both food security and the sustainable development of the country (Gebretsadik and Gebreyohanis 2012).

Even though FARM Africa (1996) and other authors were able to characterize the existing sheep in the country including the Horro and Bonga sheep breed in the western part of the country, but after many factors occurred which changed sheep phenotypic characterization and their environment. According to this, phenotypic information is a pre-request to identify potential opportunities and a good understanding of the environment. The finding of available sheep resources is important to making appropriate decisions for necessary improvement intervention programmes. Similarly, the sheep population in the study area is high and suitable for the sheep production system (Jimma Zone Livestock Office 2021). However, phenotypic characterization of indigenous sheep in the study is limited. So, to fill this gap, phenotype characterization of indigenous sheep and their breeding practices was carried out. Therefore, the objective of this study was to characterize indigenous sheep based on qualitative and quantitative traits in the Jimma zone.

2. Materials and methods

2.1. Description of the study area

The study was carried out in Seka, Dedo and Omo nada districts of the Jimma zone, Oromia, Ethiopia. Jimma area is characterized by cash crop, cereal and livestock-integrated farming systems. The area is predominantly rich in coffee and chat cash crops. It has a diverse agro-ecology classified as highlands (dega), mid-lands (woina-dega) and semi-dry lowlands (Kolla) covering 15%, 67% and 18%, respectively. This zone receives mean annual rainfall ranging from 1200 to 2800 mm and the mean monthly maximum and minimum temperatures of the zone are 11.3 and 26.2°C, respectively.

2.2. Sampling technique and sample size determination

A purposive sampling technique was employed for the selection of districts and peasant associations for the study based on (Workneh and Rowlands 2004). Both districts and peasant associations were selected purposively. This means in the first stage, districts known for sheep populations were identified followed by the identification of potential peasant association. Potentials of sheep production area were used as criteria for selecting the study sites. Thus, three districts were purposively selected. From each district two peasant associations (PA) were selected purposively based on the same criteria. The actual survey was taken to a sampling site during which qualitative and quantitative measurements were made on mature sheep. Both primary and secondary data were used in this part of the study.

For selecting sample sheep from the three kebeles of each district, castrated sheep, pregnant doe, kids, buck kids and doe kids were avoided from the sheep population to enhance accuracy for body weight and linear body measurements (LBMs) and to represent the adult sheep population. Then, sample sheep were taken by using a simple random sampling method. Dentition was used to determine the estimated age class of sheep and sheep which had one or more pairs of permanent incisors (1PPI) were used for body measurements and qualitative trait descriptions. Adult sheep were classified into four age groups one pair of permanent incisors (1PPI), two pairs of permanent incisors (2PPI), three pairs of permanent incisors (3PPI) and four pairs of permanent incisors following the description of African sheep. The sample size of the sheep was determined according to the formula given by Cochran (1977). FAO (2012) recommended for phenotypic characterization of livestock for simple random sampling. $n={\displaystyle \frac{Z^2\ast \left(p\right)\left(q\right)}{e^2}}$where n = sample size for infinite population; Z = standard normal deviation (1.96 for 95% confidence level); e = level of precision (0.05); p = the estimated value for the variability proportion of the population, 15% estimated population variability; and q = 1 − p (0.85). $n={\displaystyle \frac{Z^2\ast \left(p\right)\left(q\right)}{e^2}={\displaystyle \frac{\left(1.96\right)2\ast 0.15\ast 0.85}{0.05\ast 0.05}=190}}$n (190) is for only the district, to get for the three districts, we can multiply n by 3 and get 570. Therefore, based on the formula, a total of 570 indigenous sheep were used for collecting qualitative and quantitative traits data.

2.3. Morphometric data collection

Data (for quantitative and qualitative traits) were recorded based on the breed morphological characteristics descriptor list of FAO (2012) for the phenotypic characterization of sheep. Data for body length, height at withers, heart girth, pelvic width, tail length, tail circumference, canon bone length, canon bone-circumference, ear length and scrotal circumference were collected using a measuring tape while body weight (BW) was measured using a suspended spring balance. Data were generated for qualitative traits, such as coat colour pattern, coat colour type, back profile, head profile, ear orientation, horn present/absent, horn orientation, tail type and hair type, were visually observed. Linear body measurements and body weight were taken in the morning before feeding started.

2.4. Data management and analysis

This indicated that different type of statistical analysis was used depending on the nature of the data. The quantitative data were analysed by SAS version 9.3 (2014). Data generated from observations were described and summarized by using descriptive statistics. Chi-square (χ²) test was carried out to assess the statistical significance among categorical variables using district effect. The General Linear Model (GLM) procedure of SAS was used to analyse the linear body measurements. The sex of the animal, the district and the age group were fitted as fixed effects while the linear body measurements were fitted as dependent variables.

1. The model employed for the analysis of adult (mature) body weight and other liner measurements except scrotum circumference was $Y_{\mathrm{ijkl}}=\mu +A_i+B_j+S_k+(\mathrm{AS}{)}_{\mathrm{ik}}+e_{\mathrm{ijkl}}$

where Y_ijkl= the lth observation in the ith age group, jth location group and kth sex; µ = overall mean; A_i = effect of the ith age group (i = 1, 2, 3, 4); B_j = the effect of jth districts (j = 1, 2, 3); S_k = effect of kth sex (k = 1, 2); (AS) ik = the effect of interaction of i of the age group with k of sex; and E_ijkl = random residual error.

Pearson’s correlation coefficients were estimated among body weight and linear body measurements and between linear body measurements for females and males (SAS 2014). Body weight was regressed on body linear measurements (height at whither (HW), body length (BL), chest girth (CG), pelvic width (PW), head length (HL), cannon bone length (CBL), cannon bone circumference (CBC), ear length (EL), tail length (TL) and tail circumference (TC).

The stepwise multiple regression procedure of (SAS 2014) to determine the best-fitted regression equations for the prediction of body weight from linear body measurements for adult animals. The same body measurements including scrotum circumference were considered. The following models were used for the estimation of body weight from LBM.

For male: $Y={\text{β}}_0+{\text{β}}_1{x}_1+{\text{β}}_2{x}_2+\cdots +{\text{β}}_{13}{x}_{13}+\mathrm{ej}$where y = the response variable (body weight); β₀ = the intercept; x₁ … x₁₃ were body measurement (variables) such as body weight, body length, height at Withers, canon bone length, canon bone circumference, etc. including scrotum circumference; β₁ … β₁₃ were regression coefficients of the variables x₁ … x₁₃; ej = random error

For female $Y={\text{β}}_0+{\text{β}}_1{x}_1+{\text{β}}_2{x}_2+\cdots +{\text{β}}_{12}{x}_{12}+\mathrm{ej}$where Y = the dependent variable body weight; Β_o = the intercept; x₁ … x₁₂ were measurements (variables) such as body weight, body length, height at withers, canon Bone length, canon bone circumference, etc. except scrotum circumference; β1 … β₁₂ were regression coefficients of the variable x₁ … x₁₂; and ej = random error.

3. Result and discussion

3.1. Phenotypic characterization of indigenous sheep population

3.1.1. Qualitative traits of the sample population

Qualitative traits of the indigenous sheep population in the study area are presented in Table 1. There is an increasing interest in the characterization of African small ruminant populations because they play a major role in the maintenance of genetic resources as the basis of future improvement at both the production and the genetic levels (Nsoso et al. 2004). Similarly, Tassew (2012) states that knowing the potential of the local sheep population and trait preferences is useful for making better-informed decisions in developing interventions to improve the contribution of sheep to the livelihoods of their keepers.

Table 1. Qualitative trait of sample sheep population in the study area.

Traits and level	Seka				Dedo				Omonada				Overall
	Female		Male		Female		Male		Female		Male
	N	%	N	%	N	%	N	%	N	%	N	%	N	%
Coat colour pattern
Plain	138	80.7	15	78.9	145	84.8	14	73.7	133	77.2	17	89.5	462	81.1
Patch	30	17.5	3	15	21	12.3	4	21.1	34	19.5	2	10.5	94	16.4
Spotted	3	1.8	1	5.3	5	2.9	1	5.3	4	2.3	0	0	14	2.5
χ^2 value													2.908^ns
Coat colour type
Brown	69	40.4	10	52.6	71	41.5	5	26.3	80	46.8	13	68.4	248	43.5
Black + white	4	2.3	0	0	4	2.3	1	5.3	2	1.2	0	0	11	1.9
Pure Black	1	0.6	0	0	5	2.9	1	5.3	4	2.3	0	0	11	1.9
Pure White	7	4.1	0	0	14	8.2	2	10.5	4	2.3	2	10.5	29	5.1
Red	21	12.3	1	5.3	33	19.3	3	15.8	14	8.2	2	10.5	74	13.0
Red + white	15	8.8	3	15.8	9	5.3	1	5.3	26	15.2	1	5.3	55	9.7
Red + Black	4	2.3	1	5.3	5	2.9	0	0	3	1.8	0	0	13	2.3
Brown + white	10	5.8	0	0	8	4.7	3	15.8	7	4.1	1	5.3	29	5.1
Fawn	40	23.4	4	21.1	22	12.9	3	15.8	31	18.1	0	0	100	17.5
χ^2 value													36.18*
Head profile
Straight	164	95.9	17	89.5	156	91.2	15	78.9	167	97.7	18	94.7	537	94.2
Slightly convex	7	4.1	2	10.5	15	8.8	4	21.1	4	2.3	1	5.3	33	5.8
χ^2 value													10.03*
Back profile
Straight	122	71.3	10	52.6	127	74.3	15	78.9	106	62	18	94.7	398	69.8
Concave	49	28.7	9	47.7	44	25.7	4	21.1	65	38	1	5.3	172	30.2
χ^2 Value													4.063^ns
Horns
Present	10	5.88	9	42.1	14	8.2	10	52.6	4	2.3	3	15.8	50	8.9
Absent (Polled)	160	94.1	11	57.9	157	91.8	9	47.4	167	97.7	16	84.2	520	91.1
χ^2 value													10.041*
Horn shape
Polled	160	93.5	11	57.8	157	91.7	9	47.4	167	97.7	16	84.2	520	91.1
Straight	0	0	0		3	1.8	1	5.3	1	0.6	0	0	5	1.1
Curved	8	4.7	7	36.8	9	5.3	6	31.6	3	1.8	2	10.5	35	6.1
Spiral	3	1.8	1	5.3	2	1.2	3	15.8	0	0	1	5.3	10	1.7
χ^2 value													12.228 ^ns
Horn orientation
Upward	1	9.0	8	42.1	0	0	7	36.8	0	0	3	15.8	1	0.2
Backward	6	54.0	1	5.3	9	5.3	3	15.8	4	2.3	0	0	37	6.5
Forward	3	27.0	0	0	0	0	0	0	0	0	0	0	3	0.5
Lateral	1	9.0	0	0	5	2.9	0	0	0	0	0	0	10	1.7
χ^2 value													22.935*
Ear orientation
Semi-pendulous	95	55.6	6	31.6	101	59.1	9	47.4	75	43.9	8	42.1	294	51.6
Horizontally	76	44.4	13	68.4	70	40.9	10	52.6	96	56.1	11	57.9	276	48.4
χ^2 value													7.96*
Hair type
Smooth/short	155	90.6	16	84.2	144	84.2	15	78.9	158	92.4	18	94.7	506	88.8
Coarse/long	16	9.4	3	15.8	27	15.8	4	21.1	13	7.6	1	5.3	64	11.2
χ^2 –value													8.061*
Toggle
Present	41	23.9	6	31.6	32	18.7	3	15.8	56	32.7	5	26.3	143	25.1
Absent	130	76.1	13	68.4	139	81.3	16	84.2	115	67.3	14	73.7	427	74.9
χ^2 value													9.484*
Tail form
Straight and dawn ward end	145	84.8	15	78.9	140	81.8	12	63.2	127	74.3	13	68.4	452	79.3
Straight and twisted end up	26	15.2	4	21.1	31	18.2	7	36.8	44	25.7	6	31.6	118	20.7
χ^2 value													3.211^ns

Note: N = the number of sheep, % = percentage, * = significant, ns = non-significant, χ²=  chi-square.

Qualitative traits differed significantly (p < 0.05) between indigenous sheep populations in coat colour type, head profile, horn presence, horn orientation, toggle presence, hair type and ear orientation. Whereas non-significant difference is observed in coat colour pattern, tail form, back profile and horn shape of the sample sheep population.

The most frequent coat colour patterns 81.1% plain, 16.4% patchy and sheep with spotted pattern (2.5%) were observed, respectively. The dominant coat colour types were brown (43.5%), fawn (17.5%), red (13.0%), red and white (9.7%) and other types of colour types that contributed small proportion were pure white, brown with white, black with white and pure black also observed in varied proportion. This finding is in agreement with (Zewdu 2008) who reported that brown coat colour type was the dominant colours of Bonga sheep. The proportion of blacks is very small in the current study. According to FGD the study area, strongly supported coat colour types depended on the preference of farmers such as brown, white and red colours but they are against the black colour for which the farmers are exercising some kind of selection for the preferred ones.

The most dominant hair type of the sampled sheep populations was the short smooth hair type that accounted for (88.8%) this type of trait help to fatten easily as it makes the sheep free from external parasite and the feed required for hair production could be used for meat production. The remaining (11.2%) coarse long hair types were also rarely observed but it will be caused by external parasites that will affect the health and productivity of sheep.

The majority (91.1%) of sheep in the study area were polled, whereas (8.9%) of the sheep were horned. These findings are contrary to the results of Gizaw, who reported that above 50% of Arsi Bale female sheep were horned. However, out of the horned sampled sheep population, 6.1% had curved horn shapes, the remaining 1.7% had spiral and 1.1% had straight horn shapes. In terms of horn orientation in the current study from the total sampled sheep population 6.5% of the sheep had backward horn oriented, 1.8 lateral, 0.5% forward and 0.2% upward horn orientation.

In the study area, about 51.6% of the sampled sheep had semi-pendulous ear orientation out of the total sampled sheep and the remaining (48.4%) of the population had carried horizontal ear orientation. This result slightly agrees with the report of Zewdu (2008) who reported that the majority of Bonga sheep had semi-pendulous ear orientation. The majority of the sampled sheep population had straight (94.2%) head profiles, whereas the remaining (5.8%) had slightly convex profiles. Concerning the back profile about (69.8%) of the sampled sheep population had a straight back profile and (30.2%) had a concave profile.

The majority of the sampled sheep population had straight and downward end (79.3%) tail types while (20.7%) had straight and twisted end tail types. In the study area, the majority (74.9%) of sampled sheep had no toggle and the remaining (25.1%) had toggle. In terms of qualitative traits, the dimensions of the identified sheep populations from the three districts share some common characteristics. The possible reason might be due to the geographic proximity of the three districts.

3.2. Body weight and linear body measurements

Information on body weight and physical linear measurements of specific sheep populations at constant age has paramount importance in the selection of genetically superior animals for production and reproduction purposes. The least-square means and standard errors of body weight and other body measurements by sex, age, location and sex-by-age interaction are presented in Table 2. The body measurements are considered as qualitative growth indicators which reflect the conformational changes occurring during the life span of animals. Although live body weight is an important growth and economic trait, it is not always possible to measure it due to the lack of weighing scales, particularly in rural areas. Body measurement can also be used routinely in weight estimation and selection programmes based on its utility in determining breed evolution trends (Getahun 2008).

Table 2. live body weight and linear body measurements of sheep in the study area.

Effect and levels	N	BL	WH	CG	RH	PW	CBL	CBC
		LSM ± SE	LSM ± SE	LSM ± SE	LSM ± SE	LSM ± SE	LSM ± SE	LSM ± SE
Overall	570	61.12 ± 0.16	63.28 ± 0.17	71.38 ± 0.17	64.20 ± 0.16	11.37 ± 0.03	11.36 ± 0.04	8.21 ± 0.03
CV%		2.77	2.98	2.05	2.82	5.43	8.65	10.17
R^2		0.83	0.79	0.87	0.80	0.49	0.07	0.17
District		*	*	*	*	*	ns
Seka	190	63.2 ± 0.21a	65.10 ± 0.23a	72.88 ± 0.18ab	66.00 ± 0.22a	11.34 ± 0.08b	11.73 ± 0.12	8.15 ± 0.10b
Dedo	190	63.19 ± 0.22a	65.19 ± 0.24a	73.49 ± 0.19a	66.04 ± 0.23a	11.58 ± 0.07a	11.52 ± 0.12	8.69 ± 0.01a
Omo/nada	190	62.05 ± 0.21b	64.36 ± 0.24b	72.78 ± 0.18b	65.00 ± 0.23b	11.52 ± 0.08ab	11.55 ± 0.12	7.98 ± 0.01b
Sex		*	*	*	*	ns	*	ns
Male	57	64.40 ± 0.23a	66.56 ± 0.26a	74.77 ± 0.20a	67.16 ± 0.24a	11.52 ± 0.08	11.87 ± 0.14a	8.33 ± 0.12
Female	513	61.04 ± 0.07b	63.21 ± 0.08b	71.34 ± 0.24b	64.15 ± 0.08b	11.42 ± 0.02	11.33 ± 0.04b	8.21 ± 0.04
Age		*	*	*	*	*		ns
1PPI	166	58.58 ± 0.22d	60.64 ± 0.24d	68.200 ± 0.18d	61.61 ± 0.23d	10.57 ± 0.07c	11.13 ± 0.13c	8.02 ± 0.10
2PPI	134	61.37 ± 0.22c	63.40 ± 0.25c	71.52 ± 0.19c	64.37 ± 0.23c	11.16 ± 0.08bc	11.43 ± 0.13bc	8.30 ± 0.10
3PPI	136	64.41 ± 0.25b	66.36 ± 0.29b	74.76 ± 0.22b	67.26 ± 0.28b	11.88 ± 0.09ab	11.74 ± 0.15ab	8.33 ± 0.13
4PPI	134	66.53 ± 0.29a	69.15 ± 0.33a	77.74 ± 0.25a	69.37 ± 0.31a	12.26 ± 0.10a	12.09 ± 0.17a	8.44 ± 0.15
Sex by age interaction		*	*	*	*	*	*	ns
Male,1PPI	18	61.09 ± 0.40d	63.04 ± 0.45de	70.37 ± 0.35d	63.98 ± 0.44cd	10.58 ± 0.15cd	11.08 ± 0.24bcb	8.07 ± 0.20
Female,1PPI	148	56.06 ± 0.14e	58.23 ± 0.16f	66.03 ± 0.12e	59.24 ± 0.15e	10.57 ± 0.05d	11.18 ± 0.18bc	7.96 ± 0.07
Male,2PPI	18	62.90 ± 0.41c	64.87 ± 0.46c	72.52 ± 0.36c	65.82 ± 0.44b	11.07 ± 0.15bc	11.61 ± 0.24abc	8.46 ± 0.20
Female,2PPI	116	59.83 ± 0.16d	61.94 ± 0.18e	70.52 ± 0.14d	62.93 ± 0.17d	11.24 ± 0.06b	11.26 ± 0.09c	8.15 ± 0.08
Male, 3PPI	12	66.10 ± 0.49ab	67.84 ± 0.55b	76.43 ± 0.43b	68.69 ± 0.53a	11.87 ± 0.18a	12.12 ± 0.29ab	8.40 ± 0.24
Female,3PPI	124	62.72 ± 0.15c	64.87 ± 0.17cd	73.10 ± 0.13c	65.84 ± 0.16bc	11.89 ± 0.06a	11.36 ± 0.08bc	8.26 ± 0.07
Male, 4PPI	9	67.52 ± 0.56a	70.50 ± 0.63a	79.77 ± 0.49a	70.14 ± 0.60a	12.55 ± 0.20a	12.68 ± 0.33a	8.41 ± 0.28
Female, 4PPI	125	65.53 ± 0.15b	67.79 ± 16b	75.72 ± 0.13b	68.60 ± 0.16a	11.96 ± 05a	11.51 ± 0.08bc	8.48 ± .070
Effect and levels	N	HL	EL	TL	TC	SC	BW
	570	LSM ± SE	LSM ± SE	LSM ± SE	LSM ± SE	LSM ± SE	LSM ± SE
Overall		14.66 ± 0.05	10.58 ± 0.03	31.63 ± 0.14	15.59 ± 0.12	23.57 ± 0.12	27.36 ± 0.13
CV%		8.26	8.86	10.82	16.07	3.72	4.95
R^2		0.04	0.05	0.05	0.25	0.60	0.83
Districts		ns	Ns	*	*	*	*
Seka	190	14.89 ± 0.15a	10.86 ± 0.11	32.33 ± 0.42ab	15.55 ± 0.31b	23.72 ± 0.20b	29.28 ± 0.16a
Dedo	190	14.61 ± 0.15	10.70 ± 0.12	33.05 ± 0.43a	19.18 ± 0.32a	24.23 ± 0.21a	29.62 ± 0.17a
Omonada	190	14.94 ± 0.15	10.66 ± 0.11	31.18 ± 0.43b	16.29 ± 0.31b	23.65 ± 0.20b	28.69 ± 0.16b
Sex		ns	*	*	*		*
Male	57	14.98 ± 0.17	10.58 ± 0.04a	32.82 ± 0.47a	18.75 ± 0.35a	23.86 ± 0.12	31.19 ± 0.19a
Female	514	14.65 ± 0.05	10.90 ± 0.13b	31.54 ± 0.15b	15.26 ± 0.11b		27.18 ± 0.06b
Age		ns	*	ns	ns	*	*
1PPI	166	14.55 ± 0.15	10.44 ± 0.11b	31.88 ± 0.43	16.76 ± 0.32	22.56 ± 0.21d	25.03 ± 0.17d
2PPI	134	14.95 ± 0.16	10.56 ± 0.12b	31.69 ± 0.44	16.96 ± 0.33	23.18 ± 0.21c	27.48 ± 0.18c
3PPI	136	14.69 ± 0.18	10.79 ± 0.14ab	32.33 ± 0.52	17.09 ± 0.38	24.37 ± 0.26b	30.53 ± 0.20b
4PPI	134	15.06 ± 0.21	11.18 ± 0.16a	32.83 ± 0.59	17.23 ± 0.43	25.34 ± 0.29a	33.72 ± 0.23a
Sex by age interaction		ns	*	ns	ns	*	*
Male,1PPI	18	14.73 ± 0.29	10.37 ± 0.23b	32.92 ± 0.82	18.56 ± 0.60	22.56 ± 0.21d	26.39 ± 0.33e
Female,1PPI	148	14.38 ± 0.09	10.50 ± 0.07b	30.82 ± 0.28	14.95 ± 0.20		23.66 ± 0.11f
Male,2PPI	18	15.18 ± 0.29	10.47 ± 0.23b	32.22 ± 0.82	18.50 ± 0.50	23.18 ± 0.21c	28.79 ± 0.33d
Female,2PPI	116	14.71 ± 0.11	10.64 ± 0.08b	31.16 ± 0.32	15.40 ± 0.23		26.16 ± 0,13e
Male, 3PPI	12	14.79 ± 0.35	10.97 ± 0.27ab	32.64 ± 1.00	19.07 ± 0.73	24.37 ± 0.26b	32.80 ± 0.39b
Female, 3PPI	124	14.58 ± 0.10	10.60 ± 0.08b	32.02 ± 0.30	15.12 ± 0.12		28.25 ± 0.12d
Male, 4PPI	9	15.22 ± 0.40	11.80 ± 0.31a	33.49 ± 1.15	18.87 ± 0.84	25.34 ± 0.29a	36.75 ± 0.45a
Female, 4PPI	125	14.90 ± 0.10	10.55 ± 0.08b	32.17 ± 0.30	15.58 ± 0.22		30.69 ± 0.12c

Note: Means with different superscripts (a, b, c, d, e, f) within the same column and class are statistically different.

BW = body weight, BL = body length, CG = chest girth, WH = wither height, RH = rump height, PW = pelvic width, EL = ear length, CBL = canon bone length, CBC = canon bone circumference SC = scrotal circumference, TL = tail length, TC = tail circumference, HL = head length, N = number of sheep; ns = Non-significant; * = significant

In the study area, the overall mean of body length, body weight, height at wither, chest girth, ramp height, pelvic width, canon bone length, canon bone circumference, tail length, tail circumference, head length and scrotal circumference were 61.12 cm, 27.36 kg, 63.28, 71.38, 64.20, 11.37, 11.36, 8.21, 31.63, 15.59, 14.66 cm and23.57 cm, respectively. The values obtained for body weight (27.36 kg) in this study were lower than those obtained by Zewdu (2008) 30.75 kg for Bonga and for Horro 29.66 kg and also by Solomon (2007) for the Bonga breed which was 35 kg. This much lower value in body weight in the present study (27.36 kg) might be due to the difference in nutritional status of the animals or because the change of environmental variation or maybe the result of breed dilution through the mixing of flocks leading to increase in inbreeding. Additional possible reason may be this particular study of body weight and linear body measurements was taken during the dry season which is a period of critical feed shortage and this might be the consequence of lower body weight in the current study.

3.2.1. Sex effect

Sex had a significant (p < 0.05) effect on body weight (BW), body length (BL), chest girth (CG), wither height (WH), rump height (RH), tail length (TL), ear length (EL), canon bone length (CBL), canon bone circumference (CBC) and tail circumference (TC) whereas pelvic width (PW) and head length (HL) (p < 0.05) were not affected (p > 0.05) by sex. The sex differences in live weight and most of the LBMs observed in this study showed that these parameters are sex influenced. Male sheep had consistently higher measurement values than females across all the significantly affected variables except some that were not significant (p < 0.05). This finding could be in agreement with Sowande and Sobola (2007) who reported that ewes have a slower rate of growth and reach maturity at the smaller size compared to males due to the effect of oestrogen which restricts the growth of the long bones of the body weight.

3.2.2. Age effect

The linear body measurements and body weight were significantly (p < 0.05) affected by age except tail length and cannon bone circumference. The size of sheep increased as the age increased from youngest (1PPI) to the oldest (4PPI) which is in agreement with Getachew (2008) the size and shape of the animal increases until the animal reaches its maturity and the effect of age on body weight and other body measurements was also observed in different sheep breeds of Ethiopia. The average value of body weight for the age group was 25.03, 27.48, 30.53 and 33.72 kg for 1ppi, 2ppi, 3ppi and 4ppi, respectively. This implies that the growth patterns of the animal might be explained well by body measurements as age advances. The current finding was similar to Jemal et al. (2018) who reported that age had a highly significant (P < 0.001) effect on body weight and all other linear body measurements except head length which was not significant (P > 0.05).

In general, the body weight of indigenous sheep increases with an increase in the age of the animal. Thus, the body weight of the indigenous female sheep population increased by 2.50, 2.09 and 2.41 kg as the animal grew from 1PPI to 2PPI dentition class, from 2PPI to 3PPI dentition class and from 3PPI to 4PPI dentition class whereas for the male sheep were 2.40 kg, 4.01 and 3.95 kg, respectively (Table 2). The change in body weight was higher for males between the age class 2PPI and 3PPI, which was approximately 4.01 kg which is relatively better but not significantly higher while in the other age groups in both sexes almost constant growth.

3.2.3. District effect

In the current study, the effect of the district on body weight and most of the linear body measurements was also significant (p < 0.05) except head length, canon bone length, ear length and tail length. This is similar to Michael (2013) who reported that the district had a significant (p < 0.05) effect on live body weight and most of the linear body measurements across the studied districts. The present result also was in agreement with the earlier study result that showed the district had a significant effect on the body measurement of indigenous sheep in west Shewa (Yadeta 2016). There was a variation of body weight in study districts especially in the Omo-nada district the body weight and other linear measurements were lower as compared to the other districts. This difference could be the result of the Management; environmental differences or maybe the mixing of inbreeding of breeds across the study area.

3.2.4. Age-by-sex interaction effect

Age-by-sex interaction had a significant effect (p < 0.05) on body weight and other linear body measurements except for head length, tail length and cannon bone circumference. For all age groups, males had higher body weight and other linear measurements (p < 0.05) than females but pelvic width and some of the linear measurements were similar for the two sexes at all age groups. These differences might be due to the function of sex-related differences between sex-differential hormonal effects on the growth of sheep (Semakula et al. 2010).

3.3. Correlation between body weight and linear body measurements

The association between body weight and linear body measurements of sheep in the study area is presented in Table 3. Almost all of the linear body measurements had a positive and significant (p < 0.05) correlation with body weight except ear length in females and tail length (TL), canon bone circumference (CBC), tail circumference (TC) and head length (HL) for males. The positive and highly significant (P < 0.0001) correlations between body weight and most of the body measurements imply that these measurements can be used as indirect selection criteria to improve live weight (Tesfaye 2008) or could be used to predict live body weight of the sheep. Correlations between the quantitative traits in the sampled sheep population showed low to strong positive significant and non-significant values. In males, chest girth (r = 0.97), body length (r = 0.92), wither height (r = 0.94) and rump height (r = 0.89) had strong and significant (P < 0.05) positive associations with body weight. Scrotal circumference (r = 0.83) and pelvic width (r = 0.75) had moderate correlation with body weight, cannon bone length (r = 0.37) and head length (r = 0.46) had a low weak relationship with body weight; however; ear length (r = 0.17), tail length (r = 0.14), canon bone circumference (r = 0.08), tail circumference (r = 0.11) were non-significant values. The high, positive and significant correlation between body weight and chest girth suggests that this variable could provide a good estimate for predicting the live weight of sheep. In addition, the high correlation coefficients between body weight and body measurements for the sampled sheep populations showed that either of these variables or their combination could provide a good approximation for predicting live weight. In females, chest girth (r = 0.92), body length (r = 0.90), wither height (r = 0.89) and rump height (r = 0.88) had strong and significant (P < 0.05) positive associations with body weight, pelvic width (r = 0.59) had a moderate correlation with body weight, whereas ear length, tail length, canon bone circumference, tail circumference, cannon bone length and head length had low weak relationship with body weight. A strong correlation was observed between Rump height and Height at wither (r = 0.88), body length r = 0.90 in females and in males Height at wither (r = 0.94), Rump height (r = 0.89) but Chest girth (r = 0. 97) and (r = 0.92) for male and female, respectively correlated with body weight that indicated Chest girth haas an appropriate variable for predicting live weight for both sexes in this study than other measurements. Similarly, Michael (2013), chest girth was the recommended variable to estimate the body weight of the sheep at the farmer’s level when there are no other instruments like spring balances to measure the exact live body weight of sheep.

Table 3. Coefficient of correlations between body weight and linear body measurements of sampled sheep population (above the diagonal for females and below the diagonal for males; female = 513 and male = 57).

	BL	HW	CG	RH	PW	CBL	CBC	HL	EL	TL	TC	BW
BL	1	0.98*	0.94*	0.98*	0.59*	0.13*	0.22*	0.09*	−0.002ns	0.13*	0.12*	0.90*
HW	0.96*	1	0.93*	0.98*	0.58*	0.11*	0.22*	0.08*	0.008ns	0.13*	0.10*	0.89*
CG	0.93*	0.93*	1	0.93*	0.62*	0.12*	0.22*	0.12*	0.007ns	0.15	0.08*	0.92*
RH	0.95*	0.95*	0.87*	11	0.59*	0.11*	0.20*	0.08*	0.005ns	0.15ns	0.12*	0.88*
PW	0.68*	0.69*	0.77*	0.64*	1	0.09*	0.19*	0.13*	0.05*ns	0.01ns	0.08ns	0.59*
CBL	0.30*	0.31*	0.38*	0.27*	0.38*	1	0.08	−0.05ns	0.09*	−0.08ns	−0.04ns	0.14*
CBC	0.17ns	0.16ns	0.20ns	0.20ns	0.39*	−0.05ns	1	0.06ns	−0.01ns	0.29*	0.12*	0.26*
HL	0.05ns	0.06ns	0.09ns	0.08ns	0.07ns	0.10ns	−0.15ns	1	0.02ns	0.10ns	0.07ns	0.14*
EL	0.48*	0.44*	0.48*	0.40*	0.47*	0.26ns	0.28*	−0.02ns	1	−0.10*	0.01ns	0.003ns
TL	0.14ns	0.16ns	0.14ns	0.19ns	0.01ns	−0.08ns	−0.29*	0.10ns	−0.20ns	1	0.10*	0.14*
TC	0.10ns	0.09ns	0.10ns	0.14ns	0.19ns	−0.21ns	0.49*	0.002ns	−0.03ns	0.25ns	1	0.09*
BW	0.92*	0.94*	0.97*	0.89*	0.75*	0.37*	0.08ns	0.46*	0.17ns	0.14ns	0.11ns	1
SC	0.75*	0.75*	0.82*	0.73*	0.69*	0.27*	0.28*	0.09ns	0.36*	0.10ns	0.23ns	0.83*

3.4. Prediction of body weight from other body measurements

Multiple linear regression models for predicting the body weight of sheep from linear body measurements are presented in Table 4. The accuracy of functions used to predict live weight or growth characteristics from live animal measurements of immense financial contribution to livestock production enterprises (Tesfaye 2008). Regression analysis is commonly used in animal research to describe quantitative relationships between a dependent variable and one or more independent variables such as body weight and body measurements (chest girth, body length and height at wither) especially when there is no access to weighing equipment (Cankaya 2008).

Table 4. Multiple regression analysis of live body weight on different body measurements of female (n = 513) and male sheep (n = 57) in the study area.

Model	Intercept	B1	B2	B3	B4	B5	B6	B7	R^2	Adj. R^2	C(P)	MSE
For female sheep
CG	−21.82	0.68							0.84	0.8400	55.40	1.40
CG + BL	−20.46	0.21	0.48						0.85	0.0101	20.76	1.31
CG + BL + CBC	−21.37	0.20	0.48	0.19					0.85	0.0031	11.40	1.29
CG + BL + CBC + HL	−22.47	0.21	0.47	0.20	0.09				0.85	0.0014	8.49	1.28
CG + BL + CBC + HL + CBL	−23.33	0.21	0.47	0.09	0.19	0.09			0.85	0.0008	7.44	1.27
CG + BL + CBC + HL + CBL + HW	−23.38	0.13	0.09	0.46	0.09	0.19	0.09		0.86	0.0006	7.08	1.27
CG + BL + CBC + HL + CBL + HW + RH	−22.59	0.23	0.19	0.46	−0.20	0.08	0.18	0.09	0.86	0.0010	5.55	1.26
For male sheep
CG	−49.90	1.08							0.94	0.94	22.80	0.89
CG + RH	−53.95	0.89	0.28						0.95	0.009	11.69	0.76
CG + RH + BL	−52.98	−0.39	1.02	0.48					0.95	0.004	7.67	0.69
CG + RH + BL + SC	−53.34	−0.35	0.93	0.46	0.27				0.96	0.002	5.99	0.66
CG + RH + BL + SC + HW	−52.63	−0.44	0.27	0.87	0.34	0.29			0.96	0.002	5.03	0.65
CG + BL + RH + SC + HW + CBC	−53.38	−0.40	0.30	0.85	0.29	0.20	0.24		0.96	0.002	4.32	0.62
CG + BL + RH + SC + HW + CBC + TC	−53.53	−0.40	0.29	0.84	0.30	0.29	−0.04	0.28	0.96	0.002	4.26	0.60

In this study, Chest girth, body length, canon bone circumference, Head length, wither height and Rump height were the best-fitted model for female sheep, whereas Chest girth, Rump height, body length, Scrotal circumference, wither height, canon bone circumference and tail circumference were the best-fitted model for male sheep. The fitted prediction model was selected with smaller C (p), MSE and higher R² values. Chest girth was selected first, which explains more variation than any other linear body measurements in both ewes (84%) and rams (94%).

However, the change in the R-square of chest girth does not have a strong preceding argument due to the inclusion of additional variables in the model that could serve as the best predictor of body weight under field conditions. Moreover, thus, body weight prediction from chest girth alone would be a practical option under field conditions. This is in agreement with the results of Hizkel (2017) who reported that chest girth was selected first for the prediction of the live body weight of sheep however, the change in R-square was notstrong due to the inclusion of additional variables in the model.

Regression of body weight over independent variables, which have a higher correlation with body weight, was done to set an adequate model for the prediction of body weight separately for each sex. y = −21.82 + 0.68x for female sample sheep population and y = −49.90 + 1.08x for male sample sheep population where y and x are body weight and chest girth, respectively.

4. Conclusion

4.1. Conclusion

The objective of this study was to phenotypically characterize indigenous sheep within their production environment. Characterization of indigenous sheep in this study was vital to farmers and researchers in preserving the genetic resources of sheep breeds as well as to the country. The present study indicated that there is high variation within and between the studied indigenous sheep in most qualitative and quantitative traits, which infers considerable genetic variability. In this study, body weight and linear body measurements were influenced by sex and age. Generally, positive and significant (P < 0.05) correlations were observed between body weight and most of the body measurements. The existing higher variability within and between indigenous sheep breeds would be useful for future genetic improvement. The present study could aid future decisions on the management, conservation and improvement of the indigenous sheep genetic resources. Therefore, to increase the validity of this study, it is important to undertake well-planned genetic characterization of sheep type and then to improve their genetic potential to exploit their best use in the future.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The datasets used to support the finding of the present study is available from the corresponding author upon reasonable request.

Additional information

Funding

The author greatly acknowledges Salale University’s financial support.