Awgichew, Kassahun: Comparative performance evaluation of Horro and Menz sheep of Ethiopia under grazing and intensive feeding conditions



3.1 Experimental location and climate

This study was undertaken at the then ILCA (International Livestock Centre for Africa) now ILRI (International Livestock Research Institute) Debre Birhan Experiment Station. ILCA‘s Debre Birhan Experiment Station was established in 1979 and is located in the central highlands of Ethiopia, about 120 km north-east of Addis Ababa. The altitude is about 2780 m. a. s. l. with an average monthly minimum air temperature (at 0.5m) ranging from 2.4 0C in November to 8.5 0C in August (ILCA, 1992 and ILCA, 1993). The lowest temperature recorded so far at 0.5 m above the ground level is also reported to be -9.0 0C. Average monthly maximum air temperature at the same height ranges from 18.3 0C in September to 23 .3 0C in June (ILCA, 1992 and ILCA, 1993).

The rainfall pattern is characterised as biannual and generally the highlands have long dry seasons and a relatively cool temperate like climate. The average annual rainfall recorded between 1979 and 1992 was 920 mm. More than 70 % of the average annual rainfall is always recorded during the main rainy season (end of June to beginning of September). The small rains occur during February to April or sometimes May. The long dry period is from October to February with occasional night frosts occurring from November to January.

3.2 Soil and vegetation

Most of the station is bottomland with few upland areas. Due to the accumulation of run off soil or silt, the bottomland is comparatively rich in nutrients. However, this area is subjected to annual flooding during the main rainy season and problems of water logging and frost are often encountered.

The pasture at Debre Birhan Experiment Station is dominated by Andropogon grasses (Andropogon longipes) mixed with Trifolium spp. Depending on the site, pasture composition of the station based on dry matter bases is dominated by grass spp. (93 %) followed by broadleaf weeds (6 %).

The legume component is only 1 %. On dry matter bases, pasture yields range from 1.5 t/ha to about 4 t/ha depending on the season and the site. The crude protein yield is also highly variable, (37 - 110 g/kg DM) depending on the season and legume component of the pasture.


3.3 Breeds and management

3.3.1 Description of the local breeds used

Two indigenous Ethiopian fat tailed sheep breeds, Horro and Menz were used in this study. Both breeds are well known in the country and are of great economic importance in their respective areas of origin.

Horro sheep

This local breed belongs to the long fat tailed breed group. The Horro sheep is a hairy breed found widely distributed in the western highlands of Ethiopia. They are mostly uniform in colour having creamy white, dark tan or spotted short smooth hair. It is a large framed (tall and long) local breed. The fat tail is triangular in shape hanging down with sometimes a twisted end. Males develop a mane around the shoulder and brisket. The mature body weight of ewes is estimated to be between 35 and 40 kg (IAR, 1991).

Menz sheep

The Menz breed belong to the few indigenous woollen short fat tailed sheep population. They are predominantly found in Northern Shewa and Western Wollo regions of the country. They have amixed hair-wool fleece with an upper layer of long coarse hair and a woolly undercoat. This local sheep breed has no particular distinguishing colour. However, the dominating colour seems to be black or dark brown with some white spots on the head, neck and legs. Males have mostly long and twisted horns while females are predominantly polled. The Menz sheep are smaller, shorter and compact.

The mature body weight of Menz ewes is estimated to be between 30 and 35 kg.

As described by Galal (1983), the Menz sheep have a typically distinguishing tail characteristic with a fat tail which drops to the hock and has a slight upward twist at the end.

3.3.2 Mating

To assess the possible effect of season of birth, mating was planned in such a way that lambs would be born into the two main seasons (wet and dry). According to the mating practice of the station, breeding ewes were mated at the middle of the dry season (January/February) and at the beginning of the wet season (May/June) so that lambings could occur within the wet and dry seasons respectively.


Sires were randomly selected from rams kept for breeding purposes. The sires chosen were all over 25 kg and had at least two permanent teeth (incisors). Those selected were again physically checked for any scrotal deformity or any history of reproductive problems. Eleven rams per breed were then randomly picked for each mating season. Rams that were used in the May 1992 mating were also used for January 1993 mating. To facilitate linkage across years, some 30 % of the sires used in the previous two matings were used for the third mating period of May/June 1993.

Ewes within breed were assigned to sire groups by a stratified random sampling method taking body weight and parity into consideration. To obtain contemporary groups of lambs at the designated lambing periods, the oestrus of ewes were synchronised using progesterone impregnated intrauterine sponges (Intervet International B. V., Boxmeer, Holland). The sponges were removed after 10-12 days and ewes were randomly grouped for single sire mating for 30 days.

3.4 Experimental design and data collection

Between 240 to 248 ewes from each breed were mated per season, a total of about 845 lambings were recorded over the three lambing seasons. The number of sires used were 33 and 32 for Menz and Horro breeds, respectively. Only lambs of first and second parity dams were included in the study.

At lambing, lambs were identified with numbered plastic ear tags and birth date, weight at birth, dam ID, sire ID, mating group, season of birth and other relevant information were recorded.

The number of sires and ewes used over the three mating seasons are shown in Table 10. The distribution of lambs born in the three lambing periods is also indicated.


Table 10: Distribution of sires, ewes and lambs born by breed and season of birth



Lambs born in:




Dry season/92

Wet season/93

Dry season/93











































































* 31 Menz ewes and 38 Horro ewes have lambed twice in the study period (Dry season'92 and '93)

3.4.1 Grazing management

Animals grazed natural pasture daily for about 7 hours and for shorter period on data measurement days. Ewes and lambs were herded together until weaning of lambs at 90 days of age. After weaning, female and male lambs were separated but grazed the same paddocks in a rotational grazing schedule.

3.4.2 Supplementary feeding and health care:

All groups were supplemented with hay and concentrate in the evening. Hay was offered ad libitum while concentrate was group fed at the rate of about 200g/head/day providing 10-12.5 MJ ME/kg dry matter. The crude protein content of hay was about 3.5 % .The concentrate composition was 66 % wheat bran, 33 % Noug cake (local oil seed cake), and 1 % salt.

Animals in the feeding experiment were individually fed 1500 g /head/day hay providing about 5.12 MJ ME/kg DM and 300 g concentrate/head/day (50 % wheat bran, 30 % maize, 20 % cotton seed cake) at the beginning of the experiment which was raised to 400 g two weeks into the four months fattening period providing about 18.01 % CP and 10-12.5 MJ ME/kg DM. They were also provided with mineral licks and water ad libitum.


All flocks were routinely checked for any health problems and when animals fell sick, the identification of the animal, the date and cause of illness were registered by the health group so that the number of times the animals fell sick and the health category to be calculated. Animals were drenched on regular basis against liver flukes using either Fasinex (Triclabendazole) or Ranide (Rafoxnide) and were also vaccinated for pox, pasteurellosis and clostridial infection. Lambs were also treated with Panacur (Fenbendazole) prior to or after weaning if the faecal egg count is above or equal to 2000 epg.

3.4.3 Measurements Body weight and linear body measurements

Ewes and rams were weighed monthly in the last week of each month using a 0.1 kg precision scale. Lambs were weighed at birth and fortnightly thereafter up to weaning. After weaning at the age of about 90 days they were weighed monthly together with the rest of the flock. Lamb body weights were adjusted for age.

The average daily weight gain (ADG) was calculated using the following formula


ADG g = Average Daily Gain in gram

W1 kg = Birth weight or weight at the preceding age

W2 kg = Weight at a given age

A = Age in days or days between weighing dates

Average daily gain was calculated for the following stages of growth:

  1. ADG1 = Birth to Weight 30
  2. ADG2 = Birth to Weight 90a
  3. ADG3 = Weight 90 to Weight 180
  4. ADG4 = Weight 180 to Weight 270
  5. ADG5 = Weight 270 to Weight 365
  6. ADG6 = Birth to Weight 365

a Weight 90 = Weaning Weight


Average daily weight gain of male lambs in the fattening experiment was also calculated using the following formula


FWT = Final body weight

IWT = Initial body weight

D = number of fattening days

Linear Body measurements were taken together with monthly weight measurements. All body measurements were taken with a measuring tape in centimetre and measured to the nearest 0.5 cm. The following linear body measurements were taken:

  1. Heart girth (the circumference of the chest posterior to the forelegs at right angles to the body axis)
  2. Wither height (the highest point measured as the vertical distance from the top of the shoulder to the ground)
  3. Body length ( the distance between the crown and the Sacrococcygeal joint)
  4. Tail length (from the point of attachment to the tip)
  5. Tail width ( directly behind the Tuber ichiad)
  6. Tail circumference ( directly behind the Tuber ichiad)
  7. Tail volume ( estimated by water displacement)

Tail volume was measured only on lambs in the feeding experiment. A ten litre plastic beaker was filled with water and was put in a basin. The animal was held suspended by two people with its back facing the ground allowing the tail loose. The animal was then lowered slowly until the tail is immersed completely into the water. The amount of water displaced from the beaker is then collected from the basin and measured by a graduated cylinder. The displaced water measured in cubic centimetre (cc) is taken as the tail volume.

40 Feeding experiment

Thirty four Menz and twenty six Horro male lambs born in the first group (Dry'92) and which were 380 - 413 days of age at the start of the experiment were included in a feeding experiment for 123 days. The lambs were randomly assigned to 30 metabolic crates. The randomisation was done within breed. Each metabolic crate held two animals separated by a metal partition. The feeding troughs for hay and concentrate feeding were built inside the crates. Water was provided in buckets which were placed in front all the time. The lambs were adapted for two weeks before taking experimental measurements. Apart from water and the mineral blocks, the amount of hay and concentrate provided and refused was recorded. During the feeding experiment, faeces was collected for each animal and bulked for a weekly dry matter determination. Dry matter calculation of hay and concentrate was also done weekly.

Dry matter intake for each animal was calculated as


HDM = DM % of hay offered

CDM = DM % of concentrate offered

RHDM = DM % of refused hay

RCDM = DM % of refused concentrate

Dry matter digestibility was calculated as the difference between DMI minus faecal DM.

Dry matter (DM) determination of hay and concentrate feed was done after oven drying the sample at 105°C over night. Dry Matter determination of dissected carcass and non-carcass components and faeces samples was done after freeze drying the samples.

The following formula was used to calculate DM %


W1 = weight of sample

W2 = weight of empty crucible

W3 = weight of crucible + oven dried sample

C1 = correction for W2 read from the balance (due to hot weighing)

C2 = correction for W3 (due to hot weighing)


Samples for the determination of ash were put in a furnace at 600°c over night to burn all organic matter. Ash is then the inorganic material which has not volatilised at the temperature indicated above.

The Ash % is then calculated using the following formula


W1 = weight of air dried sample

W2 = weight of empty crucible

W3 = weight of crucible + Ash

C1 = correction for W2 read from the balance (due to hot weighing)

C2 = correction for W3 read from the balance (due to hot weighing)

DM % = dry matter percentage of the sample Carcass evaluation

All lambs of the fattening experiment were transported immediately after the end of the experiment to an abattoir in Debre Zeit city which is about 170km from the experimental site. At the abattoir, lambs were kept in a barn without feed and water for about 16 hours. Pre slaughter body weight was taken shortly before slaughtering. This weight was taken as the slaughter weight on which the calculation of dressing percentage was based. The lambs were slaughtered by severing the jugular vein and the carotid arteries.

After recording weight of fresh carcass and non carcass parameters, carcasses and internal organs were stored in a cold room at 4°C overnight. Each part of the gastro-intestinal tract was weighed with and without the contents and recorded. The content was then determined as the difference.


The following carcass and non-carcass measurements have been taken:

Carcass parameters measured

  1. Undissected: Hot carcass weight, Cold carcass weight, Kidney and kidney fat, Weight of half carcasses
  2. Dissected: lean, bone, fat, sundry trimmings
  3. Composition of whole carcasses were estimated from the composition of the dissected side as follows

Non-carcass parameters measured

  1. Mesenteric and Omental fat
  2. Fore- and hind gut (full and empty)

3.4.4 Laboratory chemical determination

The following samples were retained for laboratory chemical analysis :

  1. From the dissected left half carcasses: lean, subcutaneous fat, rump fat, sundry trimmings
  2. Others: tail fat, gastro-intestinal tract, gut fat, kidney and channel fat, urogenital fat

All samples were separately minced in a meat processing plant and stored in deep freezers. Depending on the availability, all or up to 70 g of the minced samples were freeze dried by forced removal of moisture using a LABCONCO freeze dryer Model 75040-01. The freeze dried samples were milled by a cross mill beater type LAB MILL 8000 rpm with 2mm sieve size and stored for ether extraction.

Ether extract determination:

The number and type of carcass and non-carcass samples analysed for percent of ether extract (EE %) are shown in Table 11.


Table 11: Number and types of carcass and non-carcass samples analysed


Sample type



Lean meat

Subcut. fat


Tail fat

Rump fat


Urogen.3 fat

Gut fat

Sundry trimmings

Sample total


































1 G.I.T = Gastro intestinal tract; 2 KCF = Kidney and channel fat; 3 Urogen. fat = Urogenital fat; * indicates duplicate sample size

Ether was extracted in a conventional method (A.O.A.C., 1990; Harris, Lorin E., 1970 and ILCA, 1994) using Goldfish fat extraction apparatus. Each sample was analysed in duplicates and the average was taken as an EE % value for a particular sample. However, if the difference between the two values was more than 10 % the analysis was repeated to get a third value.

The procedure followed is described below:

  1. 1.5 to 2.0 g freeze dried and milled carcass or non-carcass sample (W1) was transferred into a clean alundum thimble and covered with defatted cotton.
  2. The thimble was then placed in a sample container and fixed under the condenser of the fat extraction apparatus.
  3. Solvent beaker was dried in an oven at 105°c for 30 minutes, cooled in a desicater to room temperature and weighed (W2).
  4. 30 to 40 ml di-ethyl-ether was poured into the solvent beaker and then placed on the condenser with a ring and hand tightened.
  5. The water cooling system was turned on to cool the condenser.
  6. The hot plates were then raised until they were in contact with the beakers, and the heater was turned on.
  7. The apparatus was then left to operate for 8 hrs with occasional checks for ether leaks until extraction was complete.
  8. After extraction was completed, the hot plates were lowered and the thimble was left to drain empty. The extracted samples were removed and ether reclaiming glass tubes were placed under the condenser. The hot plate was again raised to distil the ether into the reclaiming tubes.
  9. The beakers were removed from the hot plate and left in the hood for some time to allow the remaining ether evaporate.


  10. The ether extract was dried in a forced-draft oven at 105°c for 30 minutes, cooled in a desicater to room temperature and weighed (W3).

Ether extract values for all samples were calculated on dry matter (DM) bases using the following formula


W1= weight of ground air dried sample

W2= weight of empty beaker

W3= weight of beaker + EE (Ether extract)

DM %= dry matter % of the extracted sample

Total ether extract of whole carcasses were estimated as follows:



TOTEE = Total ether extract

LEANEE = Ether extract of total lean

SUBCFEE = Ether extract of subcutaneous and intermuscular fat

TAILFEE = Ether extract of tail fat

GITEE = Ether extract of the gastro-intestinal tract

RUMPFEE = Ether extract of total rump fat

GUFEE = Ether extract of gut fat

UROGFEE = Ether extract of uro-genital

RENFEE = Ether extract of renal fat

SUNDTEE = Ether extract of sundry trimming


3.5 Methods of statistical analysis

The PROC GLM procedure (SAS for Windows Release 6.11/12; 1989-1995) was used to analyse the data fitting breed, sex, birth type, season of birth, parity, and sire within breed as main effects in the model.

The independent variables include body weight, average daily weight gain, survival rates, linear body measurements, feed intake (on DM basis), carcass parameters, body composition, etc.

Assuming that the relation between body weight and the body measurements taken is linear and checking for outliers, linear models were found to be most appropriate for the analysis of body weight and linear body measurements. The Stepwise procedure Proc Reg of the SAS System for Windows Release 6.11/12 (1989-1995) was used to see the effects of additional variables in analysing the relationship between body weight and linear body measurements. Regression equations were then developed by using only those independent variables which contributed significantly in the stepwise regression analysis. Graphics were done using Microsoft Excel 97 programme (Microsoft Office 98).

The SAS CATMOD procedure (SAS for Windows Release 6.11/12 (1989-1995) was used to analyse survival rate of lambs. The CATMOD procedure analyses survival rate as a binomial variable ( 0 = dead, 1 = alive) with effects described in Model 2 and generate a frequency table showing the number of lambs survived from birth to the various stages of growth. The package also generates the Maximum-Likelihood Analysis of a chi-square Variance table. The estimated values from the analysis of Maximum-Likelihood will then be multiplied with the Covariance Matrix of the Maximum-Likelihood estimates using LOGMLVAR (Rege, 1992) to calculate the survival rates of lambs from birth to 15, 30, 90, 180, 270 and 365 days of age.


Models fitted for lamb growth and survival

Lamb growth rate was calculated from the fortnightly and monthly weight records. Mortality of lambs was also recorded. The following statistical models were used to analyse the data.

Model 1. Growth and Average Daily Gain (ADG) of all lambs (birth - 180 days of age):

Yijklmno = µ + Bi + Sj +Btk + Pl +Bsm + Sire(B)in + (B x Bs)ij + eijklmno

Where :

Yijklmno = Body weight/ADG at different stages of growth (birth-12 months of age)

µ = Overall Mean

Bi = fixed effect of the ith breed (i = Menz, Horro)

Sj = fixed effect of the jth sex of lamb (j = male, female)

Btk = fixed effect of the kth lamb birth type (k = single, twin)

Pl = fixed effect of the lth dam parity (l = 1, 2)

Bsm = fixed effect of the mth birth season (m = Dry'92, Wet'93, Dry'93)

Sire(B)in = Nested effect of the nth sire within breed

(B x Bs)ij = breed by birth season interaction effect

eijklmno = Effect of the oth random error

Model 2. Growth and Average Daily Gain (ADG) of male lambs (birth - 365 days of age):

Yijklmno = µ + Bi + Btj + Pk + Bsl + Sire(B)mi + (B x Bs)ij + eijklmn

Where :

Yijklmn = Body weight/ADG at different stages of growth (birth-12 months of age)

µ = Overall Mean

Bi = fixed effect of the ith breed (i = Menz, Horro)

Btj = fixed effect of the jth lamb birth type (k = single, twin)

Pk = fixed effect of the kth dam parity (l = 1, 2)

Bsl = fixed effect of the lth birth season (m = Dry'92, Wet'93, Dry'93)

Sire(B)mi = Nested effect of the mth sire within breed

(B x Bs)ij = breed by birth season interaction effect

eijklmn = Effect of the nth random error


Model 3. Weight and linear body measurements of male lambs (180 - 365 days of age):

Yijklmno = µ + Bi + Btj + Pk + Bsl + (B x Bs)ij + eijklm

Where :

Yijklmno = Body weight and linear body measurements at 180, 270 and 365 days of age

µ = Overall Mean

Bi = fixed effect of the ith breed (i = Menz, Horro)

Btj = fixed effect of the jth birth type (j = single, twins)

Pk = fixed effect of the kth parity (k = 1, 2)

Bsl = fixed effect of the lth birth season (l = Dry'92, Wet'93, Dry'93)

Sire(B)in = Nested effect of the nth sire within breed

(B x Bs)ij = breed by birth season interaction effect

eijklmno = Effect of the oth random error

Model 4. Survival rate:

Yijklmno = µ + Bi + Sl + Btk + Pl + Bsm + eijklmn

Where :

Yijklmno = Survival rate at various ages

µ = Mean

Bi = Effect of the ith breed

Sj = Effect of the jth sex of lamb

Btk = Effect of the kth lamb birth type

Pl = Effect of the lth dam parity

Bsm = Effect of the mth birth season

eijklmn = Effect of the oth random error


Models fitted for lamb fattening performance

Model 5. Body weight gain, DM intake, carcass and non-carcass parameters:

Yijklmno = µ + Bi + Sire(B)ij+ eijk

Where :

Yijk = Body weight gain, DM intake, carcass and non-carcass parameter

µ = Mean

Bi = Effect of the ith breed

Sire(B)ij = Nested effect of the jth sire within breed

eijk = Effect of the kth random error

© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.

DiML DTD Version 2.0
Zertifizierter Dokumentenserver
der Humboldt-Universität zu Berlin
HTML - Version erstellt am:
Fri Aug 3 13:07:13 2001