Livestock Research for Rural Development 16 (7) 2004

Citation of this paper

A monitoring study comparing production of village chickens between communal (Nharira) and small-scale commercial (Lancashire) farming areas in Zimbabwe


T Maphosa, J Kusina*, N T Kusina*, S Makuza and S Sibanda


Department of Animal Science, University of Zimbabwe
PO Box MP167, Mount Pleasant, Harare, Zimbabwe,
*205 Animal Sciences Laboratory, University of Illinois
1207 W Gregory Drive Urbana, IL 61801, USA




A study was conducted in two adjacent smallholder-farming areas, namely Nharira (communal) and Lancashire (small-scale commercial). The objective was to obtain information on village chicken production under smallholder management. A total of 40 farmers were selected on the basis of having at least a hen and a cock (Nharira n=28: Lancashire n=12). Data on feeding practices, flock sizes, flock composition, egg production, hatchability, live weights and mortality were collected for a year.


Flock sizes were higher in Lancashire than Nharira (35±8 vs. 23±4 birds/household, respectively). The birds depended on scavenging and had a mean weight of 250g at eight weeks of age. Average hatchability was similar for Nharira (69%) and Lancashire (74%). Chick mortality by eight weeks of age averaged 60% for both Nharira and Lancashire.


The major findings were that hatchability was high but flock productivity was hindered by high chick mortality and poor growth performance.

Key words: Chickens, flock size, hatchability, mortality, smallholder, village


There is renewed interest in traditional systems of poultry production world-over (Panda and Mohapatra 1993). Governments and non-governmental development organisations consider projects to improve these systems as vehicles for rural development (Kusina and Kusina 1999; Mupetesi 2000). The development of innovative ideas for improving poultry production requires a complete understanding of the system and its key players. In Zimbabwe, there is limited reliable information on traditional systems of poultry production. The few reports on village chickens by Kusina and Kusina (1999) and Chitate and Guta (2001) were generated from cross-sectional surveys. These once-off surveys, while inexpensive and able to generate data for short-term projects, cannot replace long term observational studies based on in situ monitoring systems (Mallia 1999). This study was therefore undertaken to generate factual information on village poultry production systems as well as to identify strengths, weaknesses and opportunities for improving the systems.

Materials and methods


Study sites


The study was carried out in two areas, namely Nharira and Lancashire, located 170 km south of Harare, in the Chikomba District of Mashonaland East Province, Zimbabwe. Nharira and Lancashire lie between 31 oE and 31o15'E and 19oS and 19o15'S, at 1458 metres above sea level (Mutukumira et al 1996). These two areas are in Agro-Ecological Region III, which is characterised by infrequent heavy rainfall, ranging from 570 mm to 900 mm in the rainy season (Chiduza 1994). Ambient temperatures fluctuate between 5oC and 35oC, with minimum temperatures being recorded between May and July and the maximum in October (Chiduza 1994). Nharira is a communal crop-livestock farming area with average arable land holdings of 2.9 hectares per household. It borders Lancashire, a small-scale commercial farming area with larger arable land holdings averaging 53.2 hectares per household (Francis 2001).


Sampling design


Farmers were selected on the criteria that they had at least a hen and a cock (reproductive unit) and were willing to participate in the study. A total of 28 households from Nharira and 12 from Lancashire were selected using these criteria. Data were collected between March 2000 and February 2001. Prior to the commencement of the monitoring study, qualitative data on the production and management system were obtained through participatory rapid appraisal (PRA) techniques that included informal interviews and key informant interviews. During the monitoring study, data recorded included flock sizes, flock composition, productivity (egg production, hatchability, live weights and mortality).


Statistical analysis


The General Linear Models (GLM) procedures were used to detect differences between Nharira and Lancashire on the following quantitative data: monthly flock sizes, clutch sizes and weights of chicks. All data that were presented in percentage terms (hatchability, mortality and eggs set) had to be transformed in order to normalise them before statistical analysis. The transformation used was the arc-sin square root of the percent value (SAS 1997). Data on flock composition and on housing are presented as descriptive statistics.



Flock size and structure


The household flocks in the Nharira communal area were smaller (P < 0.05) than those in the Lancashire small-scale farming area during the one-year monitoring period. Although they fluctuated during the course of the year, the average flock sizes were 23 ± 4 and 35 ± 8 for Nharira and Lancashire, respectively. The cock to hen ratio was 1:4 and 1:6 for Nharira and Lancashire, respectively. Peak flock sizes for Lancashire were recorded in September. In Nharira, flock sizes increased gradually between March and June, after which they remained at 25 until February when there was a significant decline (P < 0.05).


Housing and feeding management


Most farmers (> 90%) had some structures that were specifically constructed for poultry housing. The structures varied widely within and across farming areas and the following categories were adopted for easy of discussion; A (Extremely ventilated), B (poorly ventilated, very small windows) and C (Adequately ventilated and a fenced outer perimeter). The farmers in Nharira had poor (60%) or no housing (9%), while 85% of the farmers in Lancashire had Category C type of housing. Category C was actually the ideal type of housing that combines a more solid enclosure for night-time housing and a well ventilated but still protected component for day-time housing. All the housing types observed had no special provisions for laying hens as no nests were in place. However, some farmers preferred to accommodate laying hens in their own houses. In some instances, the hens would look for a secure place for themselves.


The chickens were released to scavenge for an average of 11.0 ± 0.5 hours, during which period they would occasionally come back to the homestead for water and kitchen refuse. Most farmers did not offer feed supplements to their chickens regularly, but would sometimes give them either maize or sunflower grain. The amount of supplement given was never measured and varied from a few grams to a kilogram for the whole flock at any feeding time. The chicks competed for the same supplement with older birds. All the 40 farmers monitored claimed that they provided the birds with clean water at least once a day. Besides letting the chickens out of their fowl runs in the morning, giving them water and penning them in the evening, there were no other daily chores specific for chickens.


Productivity and mortality


There were no differences (P > 0.05) observed in the clutch sizes of hens between Nharira and Lancashire within and across seasons (Table 1). However, there was a tendency to larger clutch sizes for hens in Lancashire than those in Nharira. The proportion of eggs set for incubation was similar between areas and among seasons. There was no seasonal effect on hatchability, but the effect of area on hatchability approached significance (P = 0.08). Lancashire had an average hatchability of 75%, while Nharira had an average of 69% across the year.


Table 1.  Productivity parameters (mean ± SE) of village chickens by area and season




Hot-Wet season

Cool-Dry season

Hot-Dry season

Hot-Wet season

Cool-Dry season

Hot-Dry season

Clutch size

13.4 ± 1.02ab

13.3 ± 0.70ab

12.6 ± 0.92a

16.6 ± 1.74b

14.3 ± 1.07ab

12.4 ± 1.62ab

Eggs set, %

96.0± 1.87a

93.0 ± 1.30a

90.0 ± 1.83a

91.2 ± 3.24a

94.9 ± 1.96a

92.1 ± 2.99a

Hatchability, %

68.3 ± 3.01a

70.2 ± 2.10ab

69.7 ± 2.84ab

71.2 ± 4.90ab

73.2 ± 3.18ab

78.9 ± 4.43b

Mortality by 8 weeks, %

61.9 ± 4.26ab

52.5 ± 2.91a

62.4 ± 3.83b

72.5 ± 6.75b

44.2 ± 4.30a

65.8 ± 5.93b

Weight at 8 weeks, g

260 ± 23.9b

221 ± 11.8ab

189 ± 17.6a

290 ± 34.1b

294 ± 19.2b

215 ± 33.9ab

a,b,c Values without superscripts in the same row are different at P < 0.05


Mean weights of eggs (and chicks five days after hatching) were 43.9 ± 3.68 (35.1 ± 1.15) g in Nharira and 49.1 ± 1.35 (40.6 ± 1.65) g in Lancashire. Mean growth rates of chicks to eight weeks correspondingly averaged 3.69 g/d and 4.44 g/d for Nharira and Lancashire. As shown in Table 1, chick weights varied with seasons, but were generally higher in the hot-wet season and lowest in the hot-dry season. The body weights of mature birds (>10 months) were recorded during the rainy season. Mature birds in Nharira and Lancashire were similar in body weight, although those from the latter were slightly heavier. The roosters in both areas were generally heavier than the hens, for example in Nharira (2.4 ± 0.35 kg versus 1.5 ± 0.20 kg) and Lancashire (2. 5 ± 0.44 kg versus 1. 7 ± 0.32 kg), respectively.


Mortality was observed to be the major limitation to village chicken production. Most chickens died during the hot-wet and hot-dry season. There were no differences in chick survival rates between Nharira and Lancashire, with only two out of five chicks making it to the end of brooding (8 weeks). Figure 1 shows causes of chick mortality in the first eight weeks of life. A total of 1 860 losses were recorded from both Nharira and Lancashire households during the one year of monitoring. There were no differences in losses encountered in Nharira or Lancashire. There were interactions detected between cause of loss of chickens and season, three-way interaction between area, causes of loss and season, while the rest of the interactions were non-significant. About 45% of the losses in Nharira and Lancashire were due to unknown causes, especially in the hot-dry season and these ranked highest (P < 0.05), followed by losses through predation. Birds of prey (hawks, eagles and crows), rats, dogs and wildcats, in that order of importance, were found to be the predators of concern in Nharira and Lancashire. Diseases, accidents and cold weather all caused lower losses (P < 0.05) than predation but they were not significantly different from each other. Disease was a greater cause for concern (P < 0.05) in Nharira than in Lancashire. The highest losses due to disease and predation were in the hot-wet season. Accidents, cold weather and ectoparasites did not have any seasonal patterns. Diseases were difficult to identify because farmers could not provide enough details to allow for a conclusive diagnosis. Lack of refrigeration facilities also prevented farmers from keeping fresh samples to allow for post mortems. However, fowl pox and infectious coryza were identified to be the two most important diseases in the area that contributed to the high losses in the hot-wet season.

Figure 1. Causes of chick losses in Nharira and Lancashire



The differences in average flock sizes recorded for Nharira and Lancashire might have been due to the difference in landholding sizes. Lancashire is a small-scale commercial farming area and had larger areas of land for the chickens to scavenge, while Nharira is a congested communal area, where the scavenging feed resource base was limited. Therefore, competition for food was high in Nharira and survival of the fittest was the norm, hence flock growth was stalled. Mwalusanya (1998) in Tanzania also attributed differences in flock sizes in different zones of the country to the capacity of the scavenging feed resource base, among other things.


The Nharira flocks, though smaller than those of Lancashire, were still slightly larger than flocks from other communal areas in Zimbabwe. Kusina and Kusina (1999) reported an average flock size of 15 in Guruve District, while Lambrou (1993) gave a national mean estimate of 20. When making comparisons of flock sizes, it is important to use the same base that is type and timing of survey, as there are likely to be variations over time. In this study, for example, fluctuations were noted in the Lancashire flock sizes across the year. The peak flock sizes observed in September were mainly due to the high number of chicks hatched during this period. Chitate and Guta (2001) similarly observed that hatching peaked around this time. Huchzermeyer (1973) explained this as the effect of increasing day length, which stimulated many hens to lay. The cock to hen ratios for Nharira and Lancashire were close to those reported in other studies (Kusina and Kusina 1999; Chitate and Guta 2001). Lambrou (1993) advised on a mating ratio of 1 male to 15 females and frequent culling of the male lines to introduce new blood and avoid inbreeding. The flock structures for the two areas were similar, but the 55% proportion of chicks observed in the flocks was smaller compared to the 75% reported by other authors in Zimbabwe (Oakley 1998; Chitate and Guta 2001). The low chick to hen ratio was due to the presence of idle hens in the flock and the high chick losses. During the study period a quarter of the hens were observed as unproductive (not laying, incubating or rearing chicks).


In this study, unlike in many reports on village chickens (Kusina and Kusina 1999; Bamhare 2001; Mavale 2001; Moreki and Masupa 2001) most farmers housed their chickens. Chitate and Guta (2001), likewise, reported that more than 90% of the households they surveyed in rural Zimbabwe provided housing to their chickens. There was a greater proportion of farmers in Lancashire (85%) with a chicken house with a fenced chicken run than in Nharira (40%). The lack of proper housing might have affected flock sizes. However, Cumming (1992) concluded that village chicken production in developing countries was too unreliable to warrant farmers committing any large investment of their limited resources. A common feature in all houses observed was the lack of nest boxes for hens. The alternative housing that was often used for laying and incubating hens were kitchens and sleeping rooms. Accommodating these birds this way has several weaknesses. The major weakness is when the hen is denied access to either the nest or scavenging if the farmer happens to lock up the house. When hens are subjected to this stress and lack of privacy in the fowl run, the tendency is to lay away (Huchzemeyer 1973). The option of laying away exposes eggs to the rain and predation. For this reproductive wastage to be reduced, farmers should be advised to construct nest boxes.


The ad hoc feed supplements that were not based on any assessed needs were common in both areas. This unplanned system of feed supplementation is obviously not the best use of limited resources. Chicks had to compete with mature birds for the supplement. Cumming (1992) and Chitate and Guta (2001) agreed that most chicken losses were a result of nutritional stress. The solution to this lies in the suggestion that chicks should be given preferential access to feed supplementation (Gunaratne et al 1992). Giving supplements indiscriminately, as was the case in Nharira and Lancashire can result in overfed adult birds and starved chicks.


The birds were out scavenging for an average of 11.0 ± 0.5 hours during the day. Gunaratne et al (1992) made a similar observation in Sri Lanka. The farmers in these areas were not in the habit of confining birds during the cropping season as was observed in Tanzania (Mwalusanya 1998). The daily routine for a village chicken farmer was to release birds from the fowl run in the morning, give them water and pen them in the evening.


The clutch sizes decreased with decreasing feed availability. They were smallest in the hot-dry season and largest in the hot-wet season, the former being a period of low feed availability. Kusina and Kusina (1999) and Chitate and Guta (2001) reported mean clutch sizes of 14 eggs in some communal areas of Zimbabwe. These values are similar to those observed in this study. In some parts of Nigeria and Mali, clutch sizes that are half those observed in Zimbabwe have been reported (Wilson et al 1987; Dipelou et al 1996). This variation might be due to differences in the genetics of the birds or the type of management practised in Zimbabwe, Nigeria and Mali.


More than 90% of the eggs laid were incubated. The reasons for incubating greater than 90% of the eggs laid were, firstly, that the clutch sizes were small and, secondly, for restocking because mortality was high. Huchzermeyer (1973) also attributed this to the farmers' preference for poultry meat over eggs, which led them to leave more eggs to incubate so as to increase the flock size. Hatchability in both Nharira and Lancashire averaged 70%. This was relatively high, but hatchabilities of 82% have been observed elsewhere in Zimbabwe, indicating that there is still potential for improvement (Kusina and Kusina 1999). Etches (1996) identified males as being mostly responsible for infertility since the presence of the egg assures the contribution of the hen. In most cases of infertility in poultry it has been observed that the males are to blame (Etches 1996). Contrary to observations by Wilson et al (1987) in Mali that hatching was lower in the hot-dry season, this study did not detect any significant seasonal differences.


The eggs in Lancashire were heavier than those in Nharira, perhaps because the hens in Lancashire were heavier than those in Nharira. Leeson and Summers (1997) reported that body weight influences egg size. In turn, a larger egg produces a heavier chick. Therefore, chicks from Lancashire were hatched with a weight advantage, improving their chances of survival. The Nharira chicks were still lighter at eight weeks, which was the time they were separated from the hen. In addition to the influence of initial weights, the quantity of scavenged food might have had a significant role in the superiority displayed by the Lancashire chicks as the feed base was greater in Lancashire than in Nharira. Sexual dimorphism was clearly defined in the body weights of mature birds recorded from both sites. The age at which sexual dimorphism manifested itself was not determined. Wilson et al (1987) and Ramlah (1996) in Mali and Malaysia, respectively, reported that sexual dimorphism occurs at fifteen weeks.


Mortality during rearing of chicks was very high. In this study it was observed that many losses could not be accounted for. Losses due to unknown causes could be interpreted as a sign of negligence by farmers. They could also be a result of silent killers such as endoparasites, which are never observed because no post mortems are carried out. However, Huchzermeyer (1973) reported that internal parasites are of little importance when birds are kept on free range unless birds drink from dirty and unprotected water sources.


Unlike in other reports where Newcastle Disease was reported as the major cause of poultry losses (Alemu and Tadelle 1998; Gue'ye 1998; Kusina and Kusina 1999), no single Newcastle Disease outbreak was observed during this study. Fowl pox and infectious coryza were the two diseases that were observed to significantly contribute to the observed chicken losses in the hot-wet season. In the case of fowl pox, this was expected as it is spread by mosquitoes which breed in the hot-wet season (Lambrou 1988). Infectious coryza is a bacterial disease and bacteria usually proliferate in the humid environment of the hot-wet season. Infectious coryza is also worsened by poor ventilation in some of the housing structures. Losses in the hot-dry season were attributed to the following factors: predation, disease, cold, ectoparasites, accidents and a number of unknown causes. Chitate and Guta (2001) felt the primary cause of most of these losses was nutritional stress.




We are grateful for the financial support of the UZ/RVAU/DANIDA Project and to the 'Key Stakeholders' the farmers of Nharira and Lancashire may the good Lord abundantly bless them.



Alemu Y and Tadelle D 1998 The Status of Poultry Research and Development in Ethiopia. In the Proceedings of the 5th Conference, ESAP. pp. 40-60.


Bamhare C 2001 Country Report: Namibia. Alders R G and Spradbrow P B (Editors) Proceedings of the SADC Planning Workshop on Newcastle Disease Control in Village Chicken 6-9 March 2000, Maputo, Mozambique pp 26-31.


Chiduza C 1994 Farm Data Handbook: Crop and Livestock budgets, Communal Areas, Agro-ecological zones I-V. Agricultural services Division and FAO, Zimbabwe.


Chitate F and Guta M 2001 Country Report: Zimbabwe. . Alders R G and Spradbrow P B (Editors) Proceedings of the SADC Planning Workshop on Newcastle Disease Control in Village Chicken 6-9 March 2000, Maputo, Mozambique pp 47-52.


Cummming R B 1992 Village chicken production: Problems and potential. Spradbrow P B (Editor) Proceedings of an International Workshop on Newcastle disease in village chicken, control with Thermostable Oral Vaccines 6-10 October 1991, Kula Lumpur, Malaysia, pp 21-24.


Dipelou MA, Eruvebetine D and Williams T J 1996 Indigenous Chicken Rearing Under Village Conditions. International Journal of Animal Science, Volume 11: 63-67.


Etches R J 1996 Reproduction in Poultry. CAB International. University Press Cambridge. Singapore. UK.


Francis J 2001 Analysis of Integrated Crop- livestock Farming Systems in Nharira-Lancashire Smallholder Dairy Farms in Zimbabwe Through Farmer Participation. D Phil Thesis, University of Zimbabwe, Harare, Zimbabwe.


Gue'ye H F 1998 Village egg and Fowl meat production in Africa. World Poultry Science Journal 54: 73-86.


Gunaratne S P, Chandrasiri A D N, Mangalika Hemalatha W A P and Roberts J A 1992 The Productivity and Nutrition of Village Chicken in Sri Lanka. Spradbrow, P B (Editor) Proceedings of an International Workshop on Newcastle Disease in Village Chicken, Control with Thermostable Oral Vaccines 6-10 October 1991, Kula Lumpur, Malaysia, pp 144-148.


Huchermeyer F W 1973 Free ranging "Hybrid" chicken under African tribal conditions. Rhodesian Agricultural Journal, Volume 70: 73-75.


Kusina J F and Kusina N T 1999 Feasibility study of Agricultural and household activities as they relate to livestock production in Guruve district of Mashonaland Central province with emphasis on poultry production, HASP, Harare, Zimbabwe. pgs 129.


Lambrou L C 1988 Poultry diseases in Zimbabwe. The Farmer, January 1988, Modern Farming Publications Trust, Harare, Zimbabwe. pp 12-17.


Lambrou L C 1993 Indigenous poultry in Zimbabwe. Farming World 19 (3): 11-15.


Leeson S and Summers J D 1997 Commercial Poultry Nutrition (Second edition). University Books, Guelph Ontario, Canada.


Mallia J G 1999 In-situ conservation of the Black Maltese and performance evaluation under small-scale intensive system. First INFPD / FAO electric conference on family poultry.


Mavale A D 2001 Country Report: Mozambique. Alders R. G. and Spradbrow P.B (Editors) Proceedings of the SADC Planning Workshop on Newcastle Disease Control in Village Chicken 6-9 March 2000, Maputo, Mozambique pp 20-25.


Moreki J C and Masupa K V 2001 Country Report: Botswana. Alders R. G. and Spradbrow P.B (Editors) Proceedings of the SADC Planning Workshop on Newcastle Disease Control in Village Chicken 6-9 March 2000, Maputo, Mozambique pp 5-10.


Mupetesi T 2000 Smallholder poultry projects within the household Agriculture Support Programme in Zimbabwe. Pedersen G., Pemin A. and Minga U.M. (Editors). Proceedings of a Workshop on Possibilities for Smallholder Poultry Projects in Eastern and Southern Africa, 22-25 May 2000, Morogoro, Tanzania. pp 111-118.


Mutukumira A N, Dube D M J, Mupunga E G and Feresu S B 1996 Smallholder milk production, milk handling and utilization: A case study from the Nharira/ Lancashire farming area, Zimbabwe. Livestock Research for Rural Development, Volume 8 (1): 1-8


Mwalusanya N A 1998 Productivity and Nutritional Status of Local Chicken Under Village Management Conditions. Msc. Thesis, The Royal Veterinary and Agricultural University, Copenhagen, Denmark.


Oakley R D 1998 Emergency assistance for the control of Newcastle disease, consultancy report on Rural poultry production-socioeconomy, May 25th- June 9th 1998, Project TCP/ ZIM/ 8821 (A), FAO, Rome.


Panda B and Mohapatra S C 1993 Poultry Development in India. World's Poultry Science Journal, Volume 49: 127-133.


Ramlah H 1996 Performance of village fowl in Malaysia. World's Poultry Science Journal, Volume 52: 75-79.


SAS 1997 SAS Guide for Personal Computers. Statistical Analysis Systems Institute, Cary, North Carolina, USA.


Wilson R T, Traore A, Traore A, Kuit H G and Slingerland M 1987 Livestock production in Central Mali: reproduction, growth and mortality of domestic fowl under traditional management. Tropical Animal Health and Production, Volume 19: 229-236.

Received 3 February 2004; Accepted 27 May 2004


Go to top