The value of offtake for chicken in Kenya based on national statistical abstract shows that gross marketed production from poultry and eggs was 12,069.2 KSH million, representing an increase of 13.06% compared to the year 2018 (KNBS, 2019a). Furthermore, a report by the Intergovernmental Authority on Development Center for Pastoral Areas and Livestock Development in the year 2013 had estimated annual consumption of chicken meat from both commercial broilers and indigenous chicken at 20,889   million   metric    tonnes    with    a    per    capita consumption of 0.54 kg per year (IGAD, 2013). These economic contribution was obtained from all poultry raised in Kenyan farms, majority of which are free-ranging indigenous chicken (79%), with small numbers of commercial layers (14%) and broilers (6%) (Behnke and Muthami, 2011; FAO, 2007; KNBS, 2019b).

The rising demand for livestock products to meet nutritional needs of the human population can compromise animal welfare standards in a livestock production     system.     The     World     Animal     Health Organisation definition of animal welfare refers to how well an animal can cope with the environment where it lives (OIE, 2016). This definition is based on five fundamental freedoms: freedom from hunger and thirst, freedom from discomfort, freedom from pain, injury, and disease, freedom to express normal behaviour, and freedom from fear and distress.

However, poultry diseases have been reported to be rampant in Kenyan farms which present welfare challenges (Carron et al., 2017; Onono et al., 2018). As a response to disease occurrence, farmers often use drugs, which for the most part are not administered under prescription. This practice not only compromises health of animals but also presents food safety challenges to people who consume poultry products. Indeed antimicrobial resistance (AMR) has recently been prioritized in Kenya, with the development of a national action plan to mitigate its impact (GoK, 2017).

There are few studies within the African context which have reported progress on research in broiler welfare. For example, a study that investigated ammonia gas production in poultry houses and its effect on growth of broiler chickens amongst smallholder producers in Zambia concluded that accumulation of ammonia gas had a significant impact on growth rate. They concluded that blending of feeds with charcoal was an effective way of reducing levels of ammonia gas in poultry houses (Maliselo and Nkonde, 2015). Another study investigated the effects of ventilation and atmospheric ammonia gas on health and performance of broiler chickens. From this paper the conclusion was that broiler chicken kept under conventional atmospheric conditions had lower feed conversion efficiency of 1.98, when compared with those subjected to high level of ammonia gas with a feed conversion efficiency of 2.66 (Alloui et al., 2013).

From a global context, several studies have reported research findings on broiler welfare. A study investigated willingness of Dutch broiler and pig farmers to convert to production systems with improved welfare, and reported varied level of acceptance for adoption of welfare practices (Gocsik et al., 2015). From that study, 40% of broiler farmers were willing to convert to specific systems that allowed reversible changes if they could earn the same income as they did in their current system. The authors concluded that barriers for adopting high welfare systems included perceived uncertainty about price premiums, lack of space on farms for expansion, and poultry disease occurrences and their risks of spread.

Another study reported the effects of broiler housing conditions on performance in large-scale farms in Brazil (Mesa et al., 2017). Based on their study, farms with positive ventilation, metal and clay roof, dirt floors, and owner management of flocks had reduced mortality rates, while farms with concrete floors, negative ventilation, blue curtains in housing, fluorescent lighting, owner management, and use of tap and well water were associated    with   improved   feed   conversion  ratio  for broilers. Furthermore, Olanrewaju et al. (2014)investigated effects of genetic strain and light intensity on growth performance and carcass characteristics of broiler chickens grown to heavy weights. Their study found that chickens raised in low light intensity of 10 lux and 5 lux had better growth performance and meat yield when compared with those raised at a high light intensity of more than 25 lux. This finding is further supported by results from a study by Deep et al. (2013)who investigated minimum light intensity threshold needed to prevent negative effects on broiler production and welfare. The study concluded that a minimum of 5 lux was ideal for optimal growth for broiler chickens. Likewise, studies also show that broiler chickens raised indoor with outdoor access have significant welfare outcomes based on their ability to stand, walk, investigate, dust-bathing, and preen when compared with those raised exclusively indoors (Zhao et al., 2014).

Stocking density in broiler housing has been associated with an impact on body weight gain and feed intake. For example, Abudabos et al. (2013)reported that increasing stocking densities from 28 to 40 kg/m² in poultry house affected the body weight gain, feed intake, and decreased packed cell volume. The study further revealed presence of haemodilution in blood and increased level of aspartate aminotransferase enzyme, which is an indicator of hepatocellular damage. Therefore, raising broilers under high stocking density would affect performance and compromise animal welfare standards. Indeed, Simitzis et al. (2012)reported that raising broiler chickens under high stocking density decreases performance by reducing weight gain and feed intake while increasing physiological and oxidative stress levels, which reduces locomotor activity of chicken.

Few studies have investigated the impact of misuse of antibiotics in poultry farms with regard to development of antimicrobial resistance (AMR) and occurrence of antimicrobial residues in Kenya. For example, a study recently reported detection of high level of residue for sulfamethoxazole and sulfachloropyridazine in tissue samples of broiler meat in Nairobi with a mean residue level of 0.064 µg/g as compared to maximum allowable limits of 0.002 µg/g (Muthuma et al., 2018). Furthermore, studies on poultry value chains in Kenya have reported the presence of diseases in farms as a major constraint, while practices including use of dead birds as a source of family food within informal settlement areas, disposal of dead birds in public damping sites, or feeding of dead birds to dogs are biosecurity and food safety risk to connected farms (Onono et al., 2018).

Furthermore, economic costs associated with poor animal welfare in farms are not well documented in developing countries. The objective of this study was to investigate economic cost of poor animal welfare practices; determine biosecurity and food safety risks and willingness by farmers to adopt broiler welfare practices. The findings will support  policymakers  and  practitioners to understand barriers for uptake of improved animal welfare, the costs of poor animal welfare and their contribution to biosecurity and food safety risks in connected farms.

Study area and design

A cross-sectional study was done in 2018 with both large-scale and medium-scale broiler chicken farms from selected counties: Nairobi, Kajiado, Nakuru, Muranga, Machakos, and Kiambu (Figure 1). This study design was considered suitable to provide data on average position in farms with regard to units of farm inputs and their market prices, level of production, and farmers’ perceptions on broiler chickens welfare and farm-level biosecurity practices (Kothari and Garg, 2014). Additionally, secondary data were obtained from records of an integrated large-scale broiler meat processing plant on types of chicken products, causes and rate of condemnations, and downgrade of broiler chicken products at the point of slaughter. Administrative counties selected for this study comprised regions with a high concentration of large-scale broiler farms in Kenya, with about 51% of the total broiler chicken population of 2,914,840 birds kept in the counties (KNBS, 2019b).

Selection of study units

The study units were large-scale and medium-scale broiler chicken farms. Large-scale farms were described as farms keeping over 10,000 broiler chickens per production cycle, while medium-scale farms kept an average of 1,000 broilers per cycle. The farms were identified by field extension agents of an integrated large-scale company that sells day-old chicks and have an established network of extension  agents  who   offer  training   services   to   farmers  on commercial poultry production. Data were collected from 60 large-scale broiler farms which were part of contract farming agreement with an integrated large-scale company that processes broiler meat, and 60 medium-scale broiler farms. A similar study which was carried out to investigate economics and welfare of broiler chickens and pigs in Netherlands had used a sample size of 15 large-scale farms (Gocsik et al., 2017).

Data collection

Primary data were collected using a semi-structured questionnaire from respondents at farm level. The sections of the questionnaire are highlighted as:

(a) Farm demographics and broiler chicken farming practices

(b) Farmers’ perception on barriers for adoption of broiler welfare practices

(c) Farmers willingness to adopt welfare practices, which were investigated by the following questions: What percentage of farm income they were willing to forgo to implement welfare practices? And, for farms not willing to implement welfare practices, “what percentage rise in income they were willing to accept to implement welfare practices”?

(d) Farm biosecurity risks, food safety and disease risks management which were investigated by the following questions: How were dead broilers disposed from the farm? Which antibiotics were used in farms, and what were the withdrawal periods? How was litter material disposed from the farm and challenges of litter management? Which diseases were frequently seen in farms? Which forms of extension services and materials were used for farmer training? 

(e) Types of farm inputs, quantities and corresponding average market prices

(i) Vaccines: Newcastle, infectious bronchitis and infectious bursal disease

(ii) Energy sources (charcoal, electricity, gas and paraffin)

 (iii) Broiler feeds (starter, grower and finisher)

(iv) Rodent blocks, vitamins, water, and water treatment, cost of transportation

(f) Secondary data obtained from records of a large-scale integrated company

(i) Number of day-old chicks purchased by farms

(ii) Number of broiler chickens sold at point of maturity

(iii) Number of dead broilers per production cycle

(iv) Number of condemnation due to emaciation, ascites, scalding, and cadavers

(v) Numbers and types of products downgraded

(vi) Market prices for broiler products

Welfare and rights of respondents who participated in the study were protected. Before the questionnaire was administered, the objective of the study was explained, and respondent’s right to withdraw from the study explained. Furthermore, ethical clearance was obtained from University of Nairobi, Faculty of Veterinary Medicine institutional ethical review committee (Reference number: FVM/BAUEC/2018/138).

Data management and analysis

Descriptive statistical measures were computed from quantitative data, while qualitative data were summarized using thematic analysis (Gale et al., 2013). Feed conversion ratio for broiler chickens were calculated by dividing quantity of feed (kg) fed to a broiler per cycle by cold dressed weight (kg) per bird to determine quantity of feed required to produce a kilogram of broiler meat (Jeremiah et al., 2015).

Contingent valuation method was used to evaluate farmer’s willingness for monetary trade-off for adoption of welfare practices. The farmers were asked their willingness to adopt the welfare criteria with consequences on family income (Gocsik et al., 2017). For each welfare criteria, a dichotomous choice question was asked and respondents had to decide whether they would adopt or not a given broiler welfare practice, given that this would have an effect on family income. If their answer was yes, they had to indicate on a pre-defined scale ranging from 0 to 50%; what percentage of their income they were willing to give up to implement welfare practice. If they answered no, they were asked to indicate percentage increase in income they would require to adopt the given broiler welfare criteria (on a scale ranging between 0 and 50%). And, welfare practice which farmers were willing to forgo with high percentage reduction on income was considered acceptable, whereas welfare practices for which farmers were willing to forgo with the least percentage reduction on income was not acceptable.

The economic analysis was done through calculation of gross margin per production cycle; this analysis considered gross income from sale of mature broiler chicken. The farm-level variable costs included energy sources (charcoal, paraffin, electricity), feed (starter, grower, and finisher feeds), purchase of day-old chicks, vaccines, and veterinary medicines, transportation, wood shavings used as litter material, water, wages, consultation fees, rodenticide, and disinfectants. Furthermore, cost of broiler mortality, downgrade loss, and cost of condemnation at processing plant were factored in economic analysis. The cost of mortality was estimated using Equation 1 (Gocsiket al., 2017). It was assumed that broilers died at almost half the growing period, and therefore total economic weight of broiler chickens delivered was halved to take care of this assumption.

Cost of mortality = (PD + ((PP × Wt) – PD / 2) - CD) × MR            (1)

where PD is purchase price for day-old chicks; PP is farm-gate price for mature broiler chicken; Wt is average cold dressed weight for broiler chicken; CD is cost  of delivery  of  mature  broiler  to  the processing plant, and MR is the mortality rate of broilers in farms (total number of dead broilers presented as a proportion of total number of day-old chicks purchased by farms).

The cost of condemnation of broiler chicken at slaughter was estimated using Equation 2 (Gocsik et al., 2017). The cost of condemnation was determined as a product of farm-gate price for mature broiler chicken multiplied by cold dressed weight for broiler chicken and condemnation rate.

Cost of condemnation = (PP × Wt) × CR                                       (2)

where PP is farm-gate price paid to broiler chicken farmers; Wt is average cold dressed weight for broiler chicken; and CR is the condemnation rate, which is number of broiler chicken delivered to processing plant as cadavers and are rejected, presented as a proportion of the total number of mature broilers delivered to processing plant.

The cost of downgrade was estimated using Equation 3, where the difference between the producer price and downgrade price was considered as lost income to farmers per kilogram of weight delivered to processing plant. This difference in price was multiplied with weight of broiler chickens delivered and downgrade rate.

Cost of downgrade = Wt × DR (PP-DP)                                         (3)

where Wt is cold dressed weight; PP is farm-gate price paid to broiler farmers; DR is the downgrade rate which is the number of mature dressed broilers which are downgraded due to defects presented as a proportion of total number of mature broilers; while (PP – PD) is the difference between farm-gate price paid to broiler farmers (PP) and downgrade price for broiler products (PD). The economic analysis was done using a bio-economic model developed in Microsoft excel software (Currency exchange rate 1 USD = 100 KSH).

Socio-economic characteristics of selected broiler chicken farms

Respondents from medium-scale farms were drawn from individually owned farms, and about 59% of the respondents were female. Most respondents had obtained some basic level of education, with only 14% of respondents not having attained any form of education. Similarly, about 44% of the respondents had no formal training in agricultural sciences. The farms had employed on average one person to manage broiler chicken, and about 67% of the respondents were willing to implement welfare practices in farms. Of the sixty large-scale farms that were included in the study, 28 (47%) were owned by companies, while the remainder of 32 (53%) were owned by individual farmers. For large-scale contract farms, handling of broiler chickens was mostly done by hired workers (63.5%), but men (28.5%) and women (9%) were also involved in catching broilers. Of these broiler handlers, 88.5% had been trained on proper ways of catching broiler to minimize injuries. Approximately 61% of the farm personnel were trained by field extension agents from a large-scale integrated company selling day-old chicks, while other broiler handlers were trained by  government  extension  agents  through seminars and representatives of companies selling animal feeds. Large-scale broiler farms could not provide data on willingness for adoption of welfare practices since it was at a requirement that broiler chickens raised in these farms should have a growth rate of over 50 g per day.

Description of inputs and outputs for broiler chicken farms

The average number of day-old chicks purchased by large-scale farms was 22,595 per production cycle, and the number sold at point of harvesting was 20,959, with an average cold dressed weight of 1.75 kg per broiler (Table 1). For medium-scale farms, the average number of day-old chicks purchased was 1,654 per cycle, and 1,482 sold at time of maturity with a cold dressed weight of 1.64 kg. The feed types varied across farms, with medium-scale farms only feeding starter (1 kg) and finisher (3 kg) per broiler in a cycle, while contracted large-scale farms fed starter (0.42 kg), grower (1.26 kg) and finisher (1.6 kg) per broiler per cycle. In addition, water intake was estimated at 290 L per 1,000 broilers kept per cycle. Based on amount of feed to broilers in farms and average cold dressed weight of broiler meat, the feed conversion ratio for medium-scale farms was 2.44 and 1.87 for large-scale farms. The market price for cold-dressed weight for broilers for large-scale farms was 187 KSH per kilogram, while second-grade products were sold at 165 KSH per kilogram. Based on secondary data, mortality rate in contracted large-scale farms was 6.4 and 3.1%, in medium-scale farms, while culling rates were 1 and 0.3% for both large-scale and medium-scale broiler farms, respectively. Sources of energy for lighting broiler  houses  included   liquefied   gas,  electricity,  and charcoal used during brooding, with cost of electricity estimated at 11 KSH per broiler chicken (Table 2)  

The cost of vaccination was based on estimates from a large-scale integrated company that sells broiler chicken with estimates of 0.7 KSH per dose for Gumboro (infectious bursal disease) vaccine and 0.3 KSH per dose for a cocktail of Newcastle and infectious bronchitis vaccine. Veterinary fee included payment of consultation charges per farm visit, purchase price for vitamin supplements, rodenticide blocks, and disinfectants. Other costs included wood shavings, wages for casual labour, day-old chicks, transportation, and repairs and maintenance. Most farm inputs were reportedly available in local agro-vet shops (outlets selling agricultural inputs), but feeds and energy sources were considered expensive by respondents, and charcoal was not readily available during brooding stage. Some farmers reported that use of liquefied gas was most suitable due to its low costs and reliability. Of the inputs, day-old chicks were reported to be available, but wood shavings used for litter were difficult to find, while feeds and vitamins were very expensive. Some fixed inputs (feeders and drinkers) were of poor quality and would often break-up while in use, especially equipment’s made from plastics. Similarly, timber used for construction of poultry houses was expensive, and metallic materials were prone to rust.

Farms were involved with expansion activities such as construction of new housing facilities for dairy and pigs, while some had introduced layer chickens and other slow-growing improved indigenous chicken breeds. An integrated large-scale meat processing company with signed contractual agreements with farms was the buyer of mature broilers from large-scale farms. The company supplied farm inputs including day-old chicks and feeds to the  farm  and farmers only pay for inputs when mature broilers are slaughtered and marketed. However, for medium-scale farms, mature broilers were sold through several outlets: brokers, butcheries, fast food restaurants, bars, and hotels.

Barriers for uptake and willingness to adopt welfare practices by farmers

Table 3 presents perceptions of respondents’ on barriers for implementation of broiler welfare practices. The result shows a state of  uncertainty  regarding  price  premiums, increase in profits and availability of land and ease of acquiring land for farm expansion to support welfare practices. The median ranks for barriers for adoption of welfare practices averaged 3, but for ease of acquiring land, and availability of land for expansion, the average scores were 2.52 and 2.83, respectively. This reveals that these two factors can hinder adoption of welfare practices in farms.

Respondents who were willing to implement broiler welfare  criteria  in  their  farms  indicated  that they would forgo up to 30% of their family income to implement two welfare criteria: maintaining a stocking density of 30 kg/m² and provision of quality litter on the floors (Table 4). However, they only indicated to forgo a paltry 1.5% of their family income for support of keeping slower growing breeds.

Respondents further indicated they would forgo up to 15% of their income to support introduction of bales of hay and other enrichments to broiler houses. For respondents who indicated they would not implement welfare criteria, farmers would require up to 15% rise in their family incomes to support maintaining a stocking density of 30 kg/m² per square meter and spreading litter material in broiler housing (Table 5). However, they indicated low support for keeping slower growing breeds, introduction of bales of hay and other enrichments, and introduction of 6 h of darkness at night to control feed intake, even if incomes were increased. The medium-scale farmers were practicing other income-generating activities to supplement their family incomes, including vegetable farming, dairy cattle and goat farming, building of rental houses and operating green groceries.

Biosecurity risks, poultry diseases, food safety challenges and extension services

Poultry diseases that were frequently reported were similar in large-scale and medium-scale farms: Coccidiosis, gumboro, Newcastle disease, ascites, and respiratory  infections.   In  most  chicken  farms,  farmers reported that they used antibiotics to manage diseases whenever outbreaks occurred in flocks to minimize losses or through regular vaccinations. In large-scale farms, antibiotics such as doxycycline, fosbac, trimovet, tylosin 75% were often used, with an average withdrawal periods of 7 days (Table 6). Large-scale farmers were more knowledgeable on names of drugs used for treatment of diseases and withdrawal periods. However, most medium-scale farmers did not provide specific names for antibiotics used for treatment of sick broilers apart from general description such as drugs used to treat respiratory system, while other farmers reported providing supplements and vitamins. Furthermore, they did not know the withdrawal periods for drugs. Large-scale farms used condemnation pits for disposal of dead broilers, but some would burn or buried them in farms, while for medium-scale farms, these would also be fed to dogs, cats and pigs.

Most large-scale farms reported the use of litter materials in crop farms to increase soil fertility, while medium-scale farms, would also feed them to dairy cows. Wet litter occurred in broiler houses during rainy seasons, and their accumulation presented a health hazard to poultry workers, through presence of high concentration of ammonia gas, and itchiness of hands for personnel who clean litter from broiler house. In addition there were shortages of buyers of used litter during dry seasons. Training and extension services were provided to large-scale farmers by field agents of an integrated company selling day-old chicks. This company was running a large-scale  contract  farming  business model with broiler farmers where mature birds were sold back for slaughter and they would market processed meat products. The extension materials covered topics on disease control, poultry vaccinations, brooding management, and biosecurity management in farms: litter management, handling and disposal of dead birds, ventilation, and lighting in poultry houses, feeding and watering of birds.

Losses from condemnations and downgrade of broiler products at processing plant

Based on secondary data, approximately 7 million broilers were delivered to a large-scale meat processing plant in the year 2018, and 1.2% were condemned due to ascites (water belly), emaciations, and cadavers. These broilers were delivered dead to  the  processing  plant. Of the total condemned broilers, ascites comprised 81% (68,000 broilers), while emaciation comprised about 14% and cadavers comprised 5% (Figure 2). The other source of loss was downgrade of processed broiler products. The causes of downgrade included freaked breasts, superficial wounds, bruised breasts, bruised legs, bruised wings, broken wings, broken legs, and fractured breasts (Figure 3). For the most part, important causes of product downgrades from farm level were due to poor catching and handling which resulted in bruised wings, legs and breasts. Bruised wings comprised 56% of the total downgraded cases, while bruised legs and breast both comprised 44% of downgraded products. For the product downgrades at the meat processing plant, 34% were bruised wings, 28% bruised leg and 34% fractured breasts. The remainders were from freaked breast, fractured legs and broken wings.

Estimation of returns to broiler chicken producers for medium and large-scale farms

The gross income for large-scale farms was 6,858,833 and 578,454 KSH for medium-scale farms. In large-scale farms, feed and water (starter, grower, and finisher feed) accounted  for  63%  of  total  variable  costs  and day-old chicks’ accounted for 28% (Table 7). For medium-scale farms, feed accounted for 65% of the variable costs, and 24% for day-old chicks. The cost of producing a kilogram of broiler meat in large-scale farms was 172 and 222 KSH for the medium-scale farms. The estimated gross margins per farm were 548,226 and 37,889 KSH for both large-scale and medium-scale farms, respectively.

Additional costs from product downgrade for large scale farms were estimated at 11,279 KSH per cycle at farm level, and 37,297 KSH from factory associated downgrades. Furthermore, mortality of broilers accounted for an additional cost of 24,000 KSH per cycle in large-scale farms. Condemnation of broilers at the processing factory from ascites, emaciation, and cadavers accounted for additional cost of 80,934 KSH per cycle. Therefore, poor broiler welfare practices resulted in additional costs estimated at 202,088 KSH per cycle, accounting for 2.45% of variable costs of production. For medium-scale farms, mortality of broilers in farms resulted in additional costs of 3,500 KSH per cycle.

Animal welfare has increasingly become a driver of change in animal production around the world (OIE, 2016). This has influence on choices that consumers make with regard to animal source food products. Yet, findings from this study suggest  existence  of  resistance by broiler farmers to adopt welfare practice of keeping breeds that have slower growth rate, with medium-scale farmers only willing to forgo 1.5% of their income to support this welfare practice. Indeed, in the African context where livestock producers are faced with several challenges, adoption of breeds that are compatible with environmental stresses is important. Some authors have argued that breeding objectives for a desirable breed for farming systems within the continent should mitigate challenges associated with poor state of infrastructure and supply of inputs, small-scale farming systems, and rural development (Olori, 2008).

Furthermore, farmers expressed uncertainty on obtaining price premiums from implementing the welfare criteria. This finding presents a barrier for uptake of welfare practices in farms, and it agrees with previous report which revealed uncertainty on whether adoption of welfare practices in both broiler chicken and pig farms in Netherlands would result to increased farm profits or price premium (Gocsik et al., 2017). However, medium-scale broiler farmers who were interviewed showed little resistance to adoption  of  appropriate stocking density of 30 kg/m² in poultry houses and provision of quality litter in broiler houses. This could be a result of extensive training and outreach services that broiler farmers have obtained from government extension agents and private businesses on appropriate management practices. These two welfare criteria were listed as topics that poultry farmers got training and extension services on. Indeed, improvement on floor conditions through spreading quality litter materials have been reported to support improved welfare by providing an enabling environment that improves leg health and dust bathing behavior (Shields and Greger, 2013).

The enhanced extension services amongst large-scale farmers had impact on level of knowledge on antibiotics used in the farm. However, the indiscriminate use of antibiotics require monitoring because it increases risks for the development of antimicrobial resistance and antibiotic residues in broiler chicken products which poses significant one-health risk to communities. Indeed, a study has reported detection of high level of residue for sulfamethoxazole and sulfachloropyridazine in tissue samples of broiler meat with a mean residue level of 0.064 µg/g as compared to maximum allowable limits of 0.002 µg/g (Muthuma et al., 2018). Notwithstanding, the common farm practices where farmers reported feeding dead broilers to dogs, cats and pigs which can contribute to maintenance and spread of antimicrobial resistance genes and infectious diseases. These farm level practices can result in antimicrobial resistance through several mechanisms (Fletcher, 2015). These mechanisms include process of horizontal gene transfer through soil-borne resistance from antibiotic resistant genes integrated in successful gene-transmission elements like plasmids, or through animal husbandry and waste management challenges where antibiotic residues are concentration in farm waste leading to increase in number of antibiotic-resistant bacteria in environment.

Feeds and day-old chicks were the main contributors to production costs in broiler farms accounting for over 80% of total production costs, with feeds alone accounted for 65% of total costs for medium-scale farms and 63% in large-scale farms. These estimates agree with findings from a review conducted for South African farms, where cost of feeds alone accounted for up to 70% of total production costs in broiler farms (Tracy et al., n.d). Similarly, from a study done in Sweden, it was revealed that cost of feeds (SEK 5.03-5.44) and day-old chicks (SEK 1.97-2.28) were the main cost centers for broiler farms (Gocsik et al., 2017). However, in Kenyan farms the feed conversion ratio was high, with medium-scale farms having a feed conversion ratio of 2.44 and large-scale farms 1.87, which shows that Kenyan broiler systems are inefficient on feeding utilization. These feed conversion ratios were relatively high when compared with average feed conversion ratio of 1.6 for broiler raised in the United State (Skinner-Noble et al., 2003). But this level of feed conversion ratio is comparable to reports in other African countries;  for  example, Alloui  et  al. (2013)reported a feed conversion ratio ranging between 1.98 and 2.66 in an experiment where they investigated effect of ventilation and atmospheric ammonia on health and performance of broiler in Algeria.

Poor animal welfare practices in farms created additional costs to farmers, accounting for 2.45% to costs of production in large scale farms. These costs accrued from mortality, condemnation of broilers, and product downgrade due to poor catching and handling both at farm and processing plant. The causes of broiler chicken condemnation at slaughter were reported as ascites (water belly), with an estimated condemnation rate of 0.98% of all broilers delivered to meat processing plant. Indeed, cases of ascites have been reported as cause of product downgrade in developed countries (Deep et al., 2013, 2010; Part et al., 2016). From a similar study in Sweden, cost of mortality and condemnation of broilers were estimated at 0.11 and 0.14 Swedish kroner’s (SEK) per kilogram of dressed weight (Gocsik et al., 2017).

The gross margins were too tight, with medium-scale farms having an estimated gross margin of 23 KSH per broiler chicken while large-scale farms had 24 KSH, however, inherent risks associated with market prices and levels of farm inputs used may significantly alter profitability. Nevertheless, the findings are comparable with estimates from large-scale broiler farmers in Sweden, where farmers keeping broilers under restricted feeding were reported to have gross margins of SEK 0.12 while those feeding whole wheat had higher gross margins of SEK 0.68 (Gocsik et al., 2017).

In conclusion, farmers were uncertain on obtaining price premiums from adoption of welfare practices. Furthermore, profits from broiler farming were marginal, with main costs of production associated with feeds and day-old chicks. But additional costs were incurred from mortality, condemnation and product downgrades at processing factory. A reduction of these welfare losses may enhance farm profitability. But disposal of dead broilers in farm environments present biosecurity risks to connected farms. Interventions measures should also mitigate welfare losses along the value chains.

The author has not declared any conflict of interests.