Project scope and farmer selection

The Ghana component of the TPGS project, managed by the Animal Science Department of the University of Ghana with support from the Ministry of Food and Agriculture (MoFA), supplied Kuroiler and Hubbard chickens to selected poultry farmers across two management systems – intensive and semi-intensive – in the forest and coastal savannah agroecological zones of Ghana. The Kuroiler is a tropically adapted dual-purpose chicken developed in India by Kegg Farms Private Limited. It is widely reported that this cross-bred chicken originated from crossing Rhode Island Red, White Leghorn, Barred Plymouth Rock, and two Indian local chicken breeds with some broiler blood infusion to obtain specific broiler characteristics (Dessie & Getachew, 2016). Hens attain 2.5kg within 12 months, begin laying eggs at five to six months, and then lay 150–200 eggs during their 12–16-month egg-laying period, initially more than 20 eggs per month (Dessie & Getachew, 2016). Males reach 4kg in 12 months and weigh at least 1kg at around three months. The breed thrives well on household waste, scraps and vegetation and, thus, does not compete with human food for grain or require any special feeding (Dessie & Getachew, 2016). Hubbard has been a worldwide reference for broiler breeding stock, with the company supplying day-old grandparent and parent stock chicks all over the world. Hubbard chickens are characterized by good growth, feed conversion ratio, excellent viability and good meat yield. The Hubbard JA57 reach between 2.3 and 2.4kg on average at 20 weeks (https://www.hubbardbreeders.com/).

The present project was undertaken in the Greater Accra, Central, Eastern, Bono and Ashanti Regions of Ghana (Figure 1), the main areas of chicken production in the country. Stakeholder engagement, particularly with the Poultry Farmers Association, Regional and District Livestock Officers of the MoFA and the Women in Poultry Value Chain, enabled us to identify and agree with the targeted beneficiaries their needs and challenges in adopting improved technologies to enhance their production efficiency. Two service providers located in the Eastern and Ashanti Regions of Ghana were contracted to hatch fertilized eggs and brood the chickens for a period of five weeks before they were distributed to the participating poultry farmers. The study sampled poultry farmers from five regions across two agroecological zones in Ghana. These farmers were provided with Kuroiler and Hubbard chicken genotypes, which had been selected for improved growth and egg production performance, to raise under semi-intensive or intensive systems. In these systems, housing, feeding and veterinary care were provided for the chickens, with birds allowed to freely roam around the farm during the day in the semi-intensive system. Six communities (villages/towns) were sampled from each of the selected regions. Within each community, nine households (smallholder farmers) were selected as project beneficiaries. Thus, 54 households per region and 270 farmers nationwide participated in the programme. Three enumerators with minimum qualifications that met the standard selection criteria by ILRI were chosen in each region to support the project in terms of data collection. The Regional Livestock Officers of MoFA in the participating regions were assigned as supervisors to oversee the activities of the enumerators. Enumerators and their supervisors were trained on the data collection protocols, which were captured in real time using tablets installed with ODK software (ICTD, 2010). Throughout the project, regular communication was established between enumerators and the project team to ensure speedy resolution of any challenges encountered in the field.

Egg incubation and mothering of chickens

In total, 10,080 fertile Kuroiler eggs were imported from Uganda on 6 December 2021 and hatched by Akro Farms, Akropong, Eastern Region, one of the selected private hatcheries whilst a second batch of 5,040 fertile eggs each of the Hubbard JA57-I and JA57-II strains were received from Morocco on 9 April 2022 and hatched by Topman Farms, Kumasi, Ashanti Region. The day-old chicks were raised intensively in mother units for up to five weeks. The feed provided was a commercially prepared chick mash with 3,080kcal/kg of energy and 21% crude protein. All prescribed prophylactic vaccines, including the 1^st and 2^nd Newcastle and Gumboro vaccinations, were provided. Two hundred and seventy (270) beneficiary farmers received 40 birds of each of the three chicken strains at random, with no special preference for sex and size.

Data collection and analysis

Data on body weight, age at first egg (AFE), egg number and weight were collected from participating farmers using ODK tools. The first phase of TPGS Ghana took place between December 2021 and March 2023. During this period, on average, poultry farms were visited monthly for sensitization and body weight measurements, starting when the birds were five weeks of age for approximately 8 visits in total. Data entries were double-checked by supervisors before the final authorization to upload onto the server at ILRI. A total of 180 eggs made up of 60 eggs each from Kuroiler, local forest and local coastal savannah chicken ecotypes aged 34–36 weeks, were randomly sampled from farmers and kept at room temperature for three days before measuring egg quality parameters including egg weight, Haugh unit, albumin height and yolk colour using an electronic egg analyzer (ORKA Food Technology LLC, 2006) at the Molecular Genetics Laboratory, Department of Animal Science, University of Ghana, Legon. Shell thickness was measured with an electronic Vernier calliper by taking the average thicknesses of the large, central and narrow ends of individual eggs (Melesse, Maak, & Lengerken, 2010).

The one-way analysis of variance (ANOVA) procedure of GenStat 12th Edition (2009) was used for the data analysis. The effect of age and genotype of chicken were estimated based on the following model:

$y_{ijk}$ = $\mu + a_i$ + $g_j+{{ag}_{\left(ij\right)}+e}_{ijk}$

Where $y_{ijk}$ = body weight of chicken; $\mu$ = the population mean; $a_i$ = effect of age of chicken; $g_j = \mathrm{effect} \mathrm{of} \mathrm{genotype}; a{g}_{\left(ij\right)}= \mathrm{age} \mathrm{and} \mathrm{genotype} \mathrm{interaction};e_{ijk}$ = residual or the random error term.

The effect of management system and sex on growth performance of Kuroiler chickens at week 17 was analyzed based on the following model:

             $y_{ijk}$ = $\mu + m_i$ + $s_j+{{ms}_{\left(ij\right)}+e}_{ijk}$

Where $y_{ijk}$ = body weight of chicken; $\mu$ = the population mean; $m_i$ = effect of management system; $s_j = \mathrm{effect} \mathrm{of} \mathrm{sex}; m{s}_{\left(ij\right)}=\mathrm{management} \mathrm{and} \mathrm{sex} \mathrm{interaction};e_{ijk}$ = residual or the random error term.

The effect of genotype on egg quality traits of sampled eggs of Kuroiler and two local chicken ecotypes from the forest and coastal savannah zones of Ghana was analyzed using the following model:

$y_{ij}$ = $\mu$ + $g_i{+e}_{ij}$

Where $y_{ij}$ = egg quality parameter; $\mu$ = the population mean; $g_i$ = effect of chicken genotype $;e_{ij}$ = residual or the random error term.

Where the ANOVA indicated a significant genotype effect, the means were separated using Tukey’s test at the 5% margin of error. Tukey’s test was chosen for mean separation on account of its robustness, requiring a single value judging the significance of all mean differences and is thus quick and easy to use (Steel & Torrie, 1981). We also surveyed farmer trait preferences in the introduced chicken genotypes. Farmers were asked to rank ten traits (fast growth, good survivability, good disease resistance, low feed intake, laying more eggs, laying big eggs, good mothering, plumage colour, tough meat and soft meat) indicating their preference for the introduced chicken genotypes compared to local chickens. In terms of hatchability, we used the two-proportion Z-Test to calculate the true difference in the hatching performance percentages between Kuroiler and Hubbard chicken eggs.

Body weight and growth performance

The growth rates of both the Kuroiler and Hubbard chicken genotypes are shown in Figure 2. The growth rates of both the Kuroiler and Hubbard chickens were superior to those of the local chickens of Ghana kept under semi-intensive management reported in an earlier study (Osei-Amponsah, Kayang, & Naazie, 2012). This finding of superior growth performance and productivity of the introduced chicken genotypes compared to the local chickens aligns with previous studies that have reported on the benefits of using adopted breeds selected for improved production in smallholder systems (Bamidele et al., 2023). The significant reduction in the number of sampled chickens over time was largely due to the disposal (sale) of particularly male birds after 17 weeks to reduce the feeding costs of the farmer.

The production system was found to influence the growth of chickens. On average, Kuroiler chickens raised intensively weighed significantly (p < 0.05) more (2.0kg) than those kept under semi-intensive (1.8kg) management (Table 1). The significant (p < 0.05) effect of the production management system on growth was expected because birds in confinement spend less energy moving around than those in the open. Furthermore, husbandry practices, such as feeding, biosecurity and veterinary care, are expected to be better and easier in confinement. In a previous study on growth performance of chickens, Osei-Amponsah et al. (2012) reported the average body weight (g) of female and male local Ghanaian chickens at hatch, 20 and 28 weeks to be 25.8, 1,033.5 and 1,273.5g, respectively. In the male population, the average body weights at these ages were 26.6, 1,358.5 and 1,593.5g respectively indicating a clear sexual dimorphism. This growth performance in local chickens was however significantly (p ≤ 0.05) lower than the exotic Sasso T44 raised under the same management condition which gave average body weights in females at the same ages as 38.2, 2355.0 and 2,635.0g with corresponding values of 37.6, 2,777.0 and 3,192.0g in males (Osei-Amponsah et al., 2012).

The body weights recorded for Kuroiler chickens in the current study were lower than those reported for the same breed and age on station in Ethiopia (Alemu et al., 2021; Bedru, 2021). This emphasizes the importance of improved management practices for the growth and development of chickens. Local chicken improvement programmes should incorporate supplementary feeding, veterinary care and housing to enhance productivity (Osei-Amponsah et al., 2012). Under improved management, local Ghanaian Savannah and Forest chicken ecotypes weighed between 1.2–1.7kg at seven months of age (Osei-Amponsah et al., 2012). This was lower than that recorded at four months for both Kuroiler and Hubbard chickens in the present study. The superior growth performance of Kuroiler over local chickens is due to the genetic potential of the breed as a commercial strain (Dessie & Getachew, 2016), as it has been bred for superior growth through selection and strain crossing. This finding of considerable differences in live body weight among chicken strains is in line with those of previous studies (Alemu et al., 2021; Guni, Mbaga, Katule, & Goromela, 2021; Kassa, Tadesse, Esatu, & Dessie, 2021; Mulugeta, Goshu, & Esatu, 2020). Sasso chickens have also been reported to be superior in growth compared to Koekoek and Sasso-RIR crosses in Ethiopia and were recommended for semi-scavenging production systems (Fekede, Tadesse, Esatu, & Dessie, 2021). There is thus a need for the selection and improvement of the growth potential of local chickens to take full advantage of their adaptive potential and genetic variability with respect to various environmental challenges (Osei-Amponsah et al., 2013). Our findings indicated a significant (p ≤ 0.05) effect of management system on age at first egg in Kuroiler chickens. Kuroiler chickens raised under intensive management laid significantly earlier (22.3 weeks old) than those raised under semi-intensive management (25 weeks old). Additionally, Kuroiler chickens raised under intensive management grew significantly (p ≤ 0.05) faster than those under semi-intensive production systems (Table 1).

Egg production performance

The egg production performance of the Kuroiler and the two Hubbard lines was also superior to that of local Ghanaian chickens. The age at first egg (AFE) for the Hubbard birds was 17.5 weeks compared to 28 weeks for the local chickens, and the expected egg number and egg weight (Figure 3) of the Hubbard chickens increased with age.

We further compared the hatching performance of fertilized Kuroiler and Hubbard eggs under Ghanaian conditions, based on the data received from the hatchery operator and shown in Table 2. Both genotypes recorded a relatively high percentage hatched on fertile (≥ 83%). Egg hatchability is affected by egg physiology, egg size, environment (bird nutrition and handling of eggs), bird genotype, genetic diversity and their interactions. Hence, the current results could not be solely associated with the genetic makeup of birds. The hatchability percentages of 83 and 85% of incubated Kuroiler and Hubbard eggs, respectively, and can be considered satisfactory, although a higher hatchability of 87.8% has been reported for indigenous chickens in the forest and savannah zones of Ghana (Osei-Amponsah, Kayang, Manu, & Naazie, 2014). Comparison of hatching performance based on the proportion Z-Test indicated significantly (p < 0.001) higher fertility of Hubbard chicken eggs compared to Kuroiler eggs (Table 2), which can be attributed to variations in egg storage and handling by the two hatchery operators.

Egg production performance of Kuroiler and the two Hubbard lines studied was superior to that of local chickens kept under similar management conditions in Ghana (Osei-Amponsah, Kayang, Naazie, Tiexier-Boichard, & Rognon, 2015). Kuroiler birds raised intensively under on-station management in Tanzania had an age at first egg of 22.7 weeks whilst 24.8 weeks was achieved for on-farm semi-scavenging management. In other studies, egg production traits of chickens selected for fast growth (Sasso, Bovans and Koekock) were generally higher than those of local chickens, with variations observed across the genotypes (Guni et al., 2021). The fact that chickens in the present study began laying from week 23 may imply an adaptation strategy for survival under semi-intensive production conditions. Egg production is very demanding of nutrients; therefore, under free range, it is likely that the bird must first build its reserves adequately before engaging in this nutrient-demanding activity and hence adapt to start later rather than earlier. The mean egg weight, albumin height and Haugh unit of Kuroiler chickens were all significantly (p ≤ 0.05) heavier than those of the indigenous birds; however, the shell thickness and yolk colour were not significantly different (Table 3).

The relatively higher egg weights obtained for Kuroiler chickens compared to local chickens can be attributed to genetic differences. Both albumin height (AH) and Haugh unit were positively and significantly related to egg weight (Bekele, Melesse, Esatu, & Dessie, 2022; Moula, Ait-Kaki, Leroy, & Antoine-Moussiaux, 2013; Osei-Amponsah et al., 2014). Kuroiler eggs in this study were superior to the eggs from local chickens in terms of average albumin height and Haugh unit, similar to their mean egg weight. Previous findings indicate that local chicken eggs have relatively stronger shell and yolk ratios than Sasso T44. However, the egg quality of local chickens is lower than that of Sasso T44 chickens because of the negative correlation between yolk ratio and AH, the main determinant of egg quality (Osei-Amponsah et al., 2014). Haugh unit of eggs from the local chickens was lower than that of exotic chickens (Kejela, Banerjee, & Taye, 2019; Osei-Amponsah et al., 2014). The mean Haugh unit of local chickens in Ghana ranged from 64.97 to 67.34, lower than that of naked-neck and normal-feathered chickens in Nigeria, which were 73.22 and 71.40, respectively (Osei-Amponsah et al., 2014; Yakubu, Ogah, & Barde, 2008).

Farmer perception and acceptance of introduced chicken genotypes

In terms of trait preferences, farmers were asked to rank ten traits (fast growth, good survivability, good disease resistance, low feed intake, laying more eggs, laying big eggs, good mothering, plumage colour, tough meat and soft meat). Fast growth, good survivability, good disease resistance, low feed intake and increased egg number are the most preferred traits in this study (Figure 4). The result of ‘fast growth’ and ‘increased egg number’ was expected. This is because, smallholder poultry farmers in sub-Saharan Africa mainly keep chickens for their meat and eggs hence these traits are of much importance to them (Melesse, 2014; Ochieng, Owuor, & Bebe, 2013; Sonaiya & Swan, 2004). The ‘good survivability’ and ‘good disease resistance’ trait preferences are consistent with previous studies (Faustin et al., 2010; Terfa, Garikipati, Kassie, Dessie, & Christley, 2019). Diseases in all poultry enterprises can lead to huge economic loss hence farmers’ choice of ‘good survivability’ and ‘good disease resistance’ is not surprising (Dar et al., 2018).

Beneficiary farmers provided positive accounts of experimental birds. They attested to the high performance of both Kuroiler and Hubbard in terms of their body weight, egg size and egg number compared to the local chicken. The socioeconomic gains and prestige these birds have brought to them in their communities were emphasized. Almost all beneficiary farmers mentioned income generation through the sale of larger and more eggs as well as cockerels. The adaptation abilities of the introduced birds were evidenced by the relatively low mortality and their resilience to environmental and farmer management conditions. This led to non-beneficiary poultry farmers expressing interest in the TPGS project and requesting inclusion in any future initiatives. The farmers showed a strong preference for the introduced breeds, as they were found to be more market oriented (Bamidele et al., 2023). Furthermore, the use of improved tropically adapted chicken breeds in smallholder flocks in sub-Saharan Africa led to increased production and productivity, generating more income while contributing to food security and social and ecological resilience (Birhanu et al., 2023). As highlighted in the literature, livelihood enhancement programmes often provide beneficiaries with poultry birds and feed to contribute towards food security and income generation (Singh et al., 2022).