The survey was conducted in five purposively selected districts, viz., Gondar zuria, Fogera, Dera, North Achefer, and Bahir Dar zuria, from three zones of the Amhara region, Ethiopia ( Figure 1 ). All of the selected districts surround Lake Tana as a border on one side. The sample zones were North Gondar, South Gondar, and East Gojjam. As indicated in CSA (2017), all of the sites follow a crop–livestock mixed farming system in which livestock plays a major role in household income and family nutrition. The livestock population of the sampled zones and the contribution share are presented in Table 1 .

Ethiopia is home to more than 29 cattle breeds (DAGRIS, 2007), including the Fogera cattle breed that populate and produce mainly in the Amhara region (Addisu et al., 2010). The breed is predominant in the belt of Lake Tana to Abay fall ( Figure 2 ). The breed is characterized and phenotypically well known by its pied coat color of black and white or black and grey ( Figure 3 ); short, stumpy, pointed horns; thoracic to cervico-thoracic hump range; folded and moderate to large dewlap; and docile temperament. The breed is used for draft, milk, and meat (Rege and Tawah, 1999; DAGRIS, 2007). It is highly tolerant to heat stress and solar radiation and is also known for its tolerance to high altitudes, parasite and disease infestation, fly burden, wet soils, and swampy areas. The breed is classified as an intermediate zebu-Sanga type (Zenga type) (Alberro and Haile-Mariam, 1982).

The data were collected in two rounds during 2015 and 2016. Study districts were selected based on the available secondary information. With the full involvement of the district experts, study kebeles were selected by stratifying the districts based on the availability of the true type of Fogera cattle: [highly populated (> 80% of the herd resembles Fogera cattle), populated (50%–80%), and less populated (below 50%)]; proximity to Lake Tana (very proximate (up to 1 km radius from the shore of the lake), proximate (1–4 km), and far (above 4 km)); and grazing system (fully grazing and semi-grazing). Based on these strata, a total of 23 kebeles were sampled and used for the survey ( Table 2 ). Similarly, interviewed farmers were stratified based on their wealth status in terms of livestock ownership, viz., rich (having > 10 mature cattle), medium (5–10 mature cattle), and poor (< 5 mature cattle); from the total sampled kebeles, 479 farmers were selected based on equation 1, adopted by Nyangaga et al. (2005).

where

n = The minimum number of sample farmers required in each stratum for a good estimate of the mean number of cattle kept per household

A = The desired level of confidence when estimating the mean number of cattle kept per household. For our surveys, we used 1.96 for the 95% level of confidence.

B = The coefficient of variation

C = The desired level of precision in estimating the mean number of cattle kept per household. For our surveys, we used 20% (i.e., ± 20%) of the population mean.

Primary data directly from farmer interviews and secondary data from published and unpublished documents were used. For the identification of pure Fogera cattle from the mixed and highland zebu breeds, color photographs ( Figures 3 , 4 ) were used to guide the enumerators.

The data were collected by six researchers (considered as enumerators) working on the breeding team of the Andassa Livestock Research Center (ALRC), who were able to identify the pure phenotypic feature of the Fogera breed. Kebele administrators as well as district and zone officials assisted in the operation to identify the true type of Fogera cattle.

Assuming that n kebeles are sampled from N kebeles in a district, the sample mean ( m ) was calculated by adding the number of animals in the sampled kebeles and dividing by n . To obtain the total number of animals in the district, N was multiplied by m . Variance (s²) estimation, standard error, and estimated number of cattle in the district were calculated by equations 2, 3, and 4, respectively, and cluster formation for total number of cattle in a district was estimated by equation 5 (John et al., 2003; Workneh et al., 2004).

To calculate the standard error, calculating variance (s²) from the sample of Peasant Association (PA) is needed and formula (2) was used.

The standard error is the square root of the variance; then

The estimated number of cattle (TC) in the district was then calculated by equation 4.

Clustering of kebele was done for the data collection, and equation 5 was used to estimate the number of cattle in the district (TCD).

The effective population size was determined by equation 6

where Ne is the effective population size, and M and F are the numbers of breeding males and breeding females, respectively.

The index method was adopted to analyze the data collected by the rank method to obtain the average weights of ordinal types of data. The method was based on the weight of each value and multiplying that by the total ranks indicated by equation 7 (Musa et al., 2006).

where

Rn = Value given for the least-ranked level (for example, if the least rank is 5th, then Rn-5, Rn-1 = 4, and … R1 = 1)

Cn = Counts of the least-ranked level (in the above example, the count of the 5th rank = Cn, and the counts of the 1st rank = C1)

In addition, descriptive statistics and weighted averages for multiple responses were analyzed by SPSS for windows (2009). MS-Excel (2010) was also used to calculate the estimated number of Fogera cattle population based on equation 2 to equation 5.

Based on the current findings, which depend on cattle owners as a source of information, cattle production is dominant throughout the survey districts. The overall cattle holding per household was 7.03 ± 0.2 heads. Poultry holding is the next main livestock sector reared in the study districts, which might be because chickens can be reared by both men and women in households with the lowest initial capital ( Table 3 ). As indicated in the table, the cattle holding of Bahir Dar zuria was higher than the rest of the working districts, mainly due to the availability of grazing land, the production of supplementary forage through irrigation, and also because the selected kebeles of the district had experienced a nomadic nature of cattle production.

The multiple-response analysis indicated that the main breeding objectives of Fogera cattle in the study area were draft power, milk production, and manure, in order of importance ( Table 4 ). Hide production and ceremonial and cultural use were the least important breeding objectives indicated by the respondents. A primary breeding objective for draft power and milk production was also reported by Azage et al. (2009); Donald et al. (2011); Mekonnen et al. (2012); Bayou et al. (2014); Destalem et al. (2015), and Fisahaye (2016). Endashaw et al. (2012) also reported that there were multiple reasons for keeping the Mursi cattle.

Compared to the mixed and highland zebu cattle kept in the same production system, the Fogera cattle had superior quality traits, including greater milk production, drought tolerance, and a better growth rate. Of the overall comparable 13 phenotypic traits, the respondents categorized the Fogera breed as good on eight (61.5%) of them ( Figure 5 ). Milk yield (100%), color (100%), traction power (100%), body size (99.2%), and growth rate (96.8%) were also reported as superior traits of the Fogera cattle (Girma et al., 2016). As reported by Endashaw et al. (2012), farmers of the Bodi area prefer their cattle breed for its better milk production (0.37), adaptability to the local environment (0.17), ability to utilize available feeds (0.15), coat color (0.09), and large body size (0.08). Similarly, reports indicated that the Horro cattle were reared for milk fat content and the Kereyu cattle for milk yield performance (Debir, 2016).

Culling in the production system of cattle is done mainly for maintaining the desired quality of the animals. The main methods of culling are selling, castration, and slaughtering (Debir, 2016). According to the respondents’ responses, they (64.65%) practice culling in their herds. Male animals were culled due to old age (51.76%), individual performance (32.81%), and smaller body size (27.93%) ( Table 5 ), while female animals were culled due to old age (45.7%), poor fertility (41.79%), and individual performance (23.83%) ( Table 6 ). Similar results on culling of animals from the herd by old age, reproductive failure, reduction of production performance, health problems, the need for cash for household use, and slaughter were also reported (Donald et al., 2011; Mulugeta, 2015; Fisahaye, 2016; Girma et al., 2016). There is no significant (p< 0.05) difference between study districts in culling criteria, which might be due to the presence of similar criteria to cull unproductive and inferior animals based on the farmers’ indigenous knowledge.

Of the total respondents, 68.76% keep breeding bulls for breeding purposes (18.75%), draft power (28.91%), and for both breeding and draft power (21.1%). Bull keeping for mating purposes (91.3%) and for both mating and draft purposes (8.7%) was reported (Girma et al., 2016). Multiple-response analysis indicated that respondents use the coat color of the bull (56.84%), body conformation (55.08%), and body size (47.46%) as the selection criteria for breeding bulls ( Table 7 ). The selection criteria for male cattle such as coat color, fertility, body size, horn shape, and its dam’s milk yield were reported by Endashaw et al. (2012). Similar morphological bull selection criteria were also reported by Debir (2016) and Mulugeta (2015). Pedigree performance for bull selection, including milk performance of the dam, mating ability, and fast growth, was reported by Bayou et al. (2014) and Donald et al. (2011).

The majority of respondents (48.63%) indicated that they use selected Fogera bulls for mating their heifers and cows ( Table 8 ). In the Fogera district, 24.81% of the respondents reported that they use selected bulls distributed by the ALRC for community-based conservation and improvement programs. Girma et al. (2016) reported that farmers who kept Fogera cattle practiced pure breeding of Fogera, and interbred with Fogera and non-Fogera types of breeds.

Respondents (40.04%) indicated that their bull source was from their herd ( Table 8 ). Breeding bulls from their source (43.6%) and from neighbors (41.6%) were reported for the Mursi cattle (Endashaw et al., 2012). Keeping their own bulls for mating was also reported by Destalem et al. (2015); Mulugeta (2015), and Bayou et al. (2014).

Respondents followed group mating (38.87%), hand mating (28.52%), and uncontrolled mating (22.27%) in their herds ( Table 9 ). Group mating is conducted when the farmer selects a breeding bull and lets the herd stay with the selected bull for mating; hand mating is conducted when the farmer selects a bull and lets the cows mate individually; and uncontrolled mating is done on the grazing land. Similar results of hand mating and uncontrolled mating with 55.4% were reported by Endashaw et al. (2012). Unlike the current findings, Bayou et al. (2014) reported uncontrolled mating as a major mating system.

The population count of Fogera cattle was performed using the three categories of pure Fogera, mixed Fogera, and highland zebu breeds. From the total population estimated in the study districts, households own 41% relatively pure Fogera, 35% Fogera–highland zebu mix, and 24% highland zebu based on the breeds’ phenotypic features. In total, 76% of the estimated population was reported to have Fogera blood. This report is much lower than what was reported by Addisu et al. (2007), who stated that 98.1% of the respondents in northwest Ethiopia include Fogera cattle in their herds. This might be because some additional sites were considered in this study compared to the sites covered by these authors. It might also be due to the increased transhumance of the Fogera breed to the highland areas and in turn, crossing with highland zebu cattle during the rainy season to escape inundation of the plain by Lake Tana. In addition, the preference of highland zebus by the farmers over Fogera (due to high feed requirement) might also imply the decline in the ownership trend of Fogera cattle from the general perspective.

In this study, the total estimated number of the breeding Fogera cattle was 55,646 ± 16,579 heads, and the effective population size (Ne) of the breed was 51,770 heads ( Table 11 ). The literature reported that the population size of the Fogera breed decreased at an alarming rate from 800,000 in 1981 (Alberro and Haile-Mariam, 1982), 636,000 in 1998 (ILRI, 2007), and 86,800 in 1999 (Rege and Tawah, 1999) to 15,000 in 2004 (Gebeyehu et al., 2004). However, the current population size shows an increase compared to the last report. The increase might be due to the adoption of a better estimation methodology and wider sampling size and area coverage. The higher number reported by Alberro and Haile-Mariam (1982) considers the entire cattle population count in one district (Fogera district) where the breed is produced, and the study by Gebeyehu et al. (2004) reported the population of selected kebeles of two districts, viz., Fogera and Dera. However, due to the increased distribution of the breed ( Figure 2 ) and the transhumance by herders, the current paper covered five districts and followed a systematic breed population count methodology, which increased the current population size of the breed. Despite higher figures reported in this study than Gebeyehu et al. (2004), the population trend of the breed continues to decline.

According to the current estimation, the Fogera–highland zebu cross of different blood levels also covered 35% of the total cattle population in the study districts, creating a pressing need to back-cross to 100% Fogera blood levels through the introduction of pure Fogera breeding bulls. In line with this, important activities are being undertaken by the ALRC to increase the number of Fogera cattle through community-based bull service, which needs further scale-up.