International Journal of Poultry Science 4 (2): xx-xx, 2005 © Asian Network for Scientific Information, 2005 1 Evaluation of the Growth Potential of Local Chickens in Malawi T.N. Gondwe* and C.B.A. Wollny Institute of Animal Breeding and Genetics, Animal Breeding and Production in the Tropics, Georg-August Universität Göttingen, Kellnerweg 6, 37077, Göttingen, Germany E-mail: tgondwe@gwdg.de Abstract: The growth potential of local chickens in Malawi was evaluated by comparing their growth performance under cage-fed and free-range management conditions. Chicks (n = 106) were collected from 39 farmers in 19 villages and individually raised in cages from an average age of 9 weeks to 20 weeks. On- farm made growers mash (17 % CP) was fed and birds were treated against common diseases and parasites. Hatch mates (n = 141) of cage-raised chicks remained on farmer household flocks and were raised by their dam hens under scavenging conditions. These birds were raised in two batches between October and December 2002, and between January and March 2003, which were corresponding with hot-dry and warm-wet seasons, respectively. Sex of chickens, village, management and management x batch interaction significantly (p<0.05) influenced growth traits. The values for birds under cage-managed conditions were significantly (p<0.05) 27, 39, 42, 25 and 41% higher than for birds under scavenging conditions, for weight at 20 weeks, overall daily weight gains, specific growth rate and growth efficiency, respectively. Phenotypic variance for daily weight gains and specific growth rates were 17 and 21%, respectively lower for cage-fed than for free range birds. Correlation coefficients of growth traits measured between cage-fed and scavenging conditions were low (r = 0.21-0.53, p<0.05), indicating possible genotype by environment interaction. Gross margin over feed costs was MK26.00 per bird (SD, MK27.00). This was 35 % rate of return on feed costs (SD, 38 %) or 24 % rate of return on initial bird value plus feed cost (SD, =26). It is concluded that growth potential of local chickens is only partially exploited under scavenging conditions primarily due to feed constraints. Key words: Growth potential, cage fed, scavenging, local chickens Introduction The recognized importance of local chickens in providing meat, cash income, socio-cultural values to rural people and their efficient scavenging system has led to increased research on the species during the past 10 years. However, most of these studies have been baseline surveys and interviews (Pedersen, 2002). These studies have generated information on local chicken production, their functions to rural households and demonstrated that the system is complex with many Materials and Methods constraints such as low productivity of meat and eggs, Experimental site: This study was conducted at Bunda and high mortality (Guèye, 1998). Because of little or College of Agriculture (BCA) located 30 km west of no investment into the system, production of local Lilongwe. Fifty (50) metabolic cages of 34 x 33 x 33 cm chickens is low cost, makes use of by-product size were constructed locally from welding wire and resources and is thus efficient (Aini, 1990). placed inside a building, which had open sides covered Growth is a compound trait influenced by genetic and with wire mesh for ventilation. Each cage was fitted with management, especially nutrition and health. The village water and feed troughs locally made from burnt clay and scavenging condition is variable, without standard curved on top to minimize feed spillage. husbandry system (Kitalyi, 1998). Performance of local The trial on scavenging conditions was conducted in the chickens is thus also variable under traditional surrounding villages on households of farmers who production system. While it is important to know how participated in the village poultry project. These villages chickens perform under scavenging conditions, surround BCA, and are located within the coordinates knowledge of their production potential is also essential 14.10 S, 33.47 E and altitude is approximately 1200 m (Pedersen, 2002). This knowledge can guide sound above see level (Garmin GPSMAP 76CS, Garmin Ltd, formulation of strategies to improve local chickens. www.garmin.com). Over the past four years, annual Trials under controlled environments can help to precipitation for Lilongwe averaged 932 mm (National determine production potential, especially when compared to the scavenging village conditions. The objectives of this study were to determine growth potential of local chickens under cage-managed system; and to compare their growth with local chickens under free-ranging system in the villages. The hypothesis to be tested was that the village free-ranging system limit expression of growth potential of local chickens. o o Gondwe and Wollny: Growth potential of local chickens 2 Statistical Office, 2003). Two seasons are distinct; a wet These were on free-range (scavenging) with their dam- warm season (November to April) and a dry season hens. Their growth was monitored during the same (May to October). Most of the smallholder farmers belong period as their contemporary counterparts in cages. to the Chewa tribe and practice subsistence agriculture Weighing of all birds was on the same day on weekly in a crop-livestock integrated system. About 82% of intervals. Management followed what farmers practiced these farmers own local chickens using free-range, including the participatory communal Newcastle scavenging low input-low output production system disease vaccination. (Gondwe and Wollny, 2002). Different phenotypes and different age-groups scavenge together. Farmers Data collected: Birds were in the cages until they erratically provide supplement feed to their chickens, reached 20 weeks of age. Altogether there were three mainly using maize bran (10.4% CP). Through batches of chicks introduced in the cages. However, the community participation, farmers were vaccinating their first batch was on pre-trial basis and was not be chickens against Newcastle disease (NCD) using La included in the analysis. With two batches, 100 chicks sota live vaccine (1000 doses cloned, Lohmann Animal were observed. Of these, 70% were female and 30% Health GmbH) at three monthly intervals, between May males. Distribution in terms of colour were 4.76% and December. Farmers shared the cost of the vaccine. Chiphulutsa, 20.95% Kawangi, 12.38% Mawanga, Experimental chickens in cages: Fifty growing chicks a weekly basis, live weights of chicks were taken using collected from farmers in the study area were individually a digital scale (Ohaus CS5000, Ohaus Corp, Pine allocated into the cages at random. From each farmer, Brook, NJ, USA; maximum of 5 kg, graduated to 2 g). up to three chicks from the same hen were randomly Data on free-range birds (n = 147) were combined with chosen. The hatch dates of these chicks and their data from cage birds and subjected to similar mother hens were recorded. Chicks were of an average calculations described for cage-managed chickens age of eight (8.5) weeks when introduced to cages. pertaining to growth traits. Altogether, there were 247 Chicks were tagged, sexed and phenotypically birds included, 40% in cages and 60% under village characterized. Initial weight was taken for each bird. In management. Female birds constituted 71%. total, 39 farmers from 19 villages in the study area Distribution by batch is shown in Table 2. contributed to the study. The periods for the pre-trial and batches 1 and 2 were, respectively, August - September Data calculation and analyses: Growth performance 2002; October - December 2002; January - March 2003. was determined using live weights and growth rate Each batch covered on average 12 weeks. parameters weight gains, specific growth rates and Feeding in cages: During the study period, chicks were offered on-farm formulated growers’ mash (Table 1). (1) Growers’ mash was chosen since it is fed to commercial layer chicks during growing phase. Feed Where, WG is weight gain (daily, weekly or overall) per and water were offered ad lib in the cages. time period in g; LW is live weight at particular week = t ; Health management of chicks in cages: Birds in cages were treated against helminths (Piperazine from CAPS, Zimbabwe), coccidiosis (Amprolium from Netherlands) (2) and other prophylaxis (Triple Sulfa from Antec Health care Africa Ltd, South Africa). Those that had external Where, SGR is the specific growth rate in percent growth parasites were dusted and smeared with tick greeze per day at a particular time; ln(LW ) is natural log of live (Cooper Ltd, Zimbabwe). Other treatments were weight at week = t ; ln(LW ) is natural log of live weight at administered upon noticing a problem on individual previous week = t ; (t -t ) is the period of weighing birds. Health was monitored on daily basis. Treatment converted to days was, however, administered to all birds to prevent possible infection to other birds. (3) If birds died during the study, it was arranged to return to farmers a replacement bird of similar age taken from the Where, GE is growth efficiency per time period = t ; WG is weight gain at time = t ; LW is live weight at time = t Experimental chickens under free-range: The clutch normality using proc univariate, normal and plot mates of chicks (offspring from same hatch) brought to procedure of SAS (SAS, 1999). Normality was cages remained at the farmers’ home in the villages. considered at over 90 % using Shapiro - Wilk (W) test for 1 26.67% Yakuda, 15.24% Yofira and 20.00% Yoyera. On growth efficiency. These were calculated as follows ti i LW is live weight for the previous period = tto o ti i to o i o i ti College stock. i to o All measured and calculated parameters were tested for Gondwe and Wollny: Growth potential of local chickens 3 Table 1: Ingredients and nutritive contents of growers’ (xn), taken as a covariate; h is the random effect of hen; mash ration fed to chicks (on as fed basis) > is the residual error assumed NID (0, F >). Ingredient Amount (kg/100kg) Maize 67.61 Soybeans, full fat, roasted 26.12 Fish meal 3.67 Vitamin and Mineral premix 0.30 Iodised salt 0.30 DL-Methionine 0.01 Lime 2.00 Total 100.00 Nutrient contents1 Dry matter, % (analyzed) 95.34 interaction was done by correlating performance of local Crude Protein % (analyzed) 17.95 Calcium % (calculated) 1.00 ME, kcal per kg (calculated) 3233 Phosphorus % (calculated) 0.83 Crude Fibre, % (calculated) 2.43 Lysine % (calculated) 0.93 Methionine % (calculated) 0.32 Analsyzed means nutrients were analyzed in the lab; calculated1 means nutrients were calculated based on nutrient values of ingredients. Table 2: Distribution of chicks allocated into cages and on-farm with number of their hens Batch Number Number of chicks of hens ------------------------------------ Cage On-farm 1 22 50 79 2 24 50 68 100 147 normality (n < 2000). All parameters were normallyobs distributed and data analyses proceeded without transformations. Model of analyses: The model of analysis included the management effect to compare growth variation among birds between cage-managed and free-range managed systems. Effects of colour, two-way (except batch x management) and three-way possible interactions were not significant (p>0.05) and were hence dropped from the model. Initial weight was more important as a covariate than initial age and hatch weights. All analyses were based on overall production in this case. The final model fit to data was y = µ + b + m + s + v + h + $(x - xn) + > (4)ijklmn i j k l m ijklmn ijklmn Where, y is the observed measure for bird n; µ is theijklmn overall mean to all birds; b is the fixed effect of batch ofi production (I = 1,2); m is the fixed effect of managementj (j = cage, scavenging); s is the fixed effect of sex of birdk (k = 1,2); v is the fixed effect of village (l= 1,2 …, 19); $ isl the linear regression coefficient of the measure on initial weight of bird; (x - xn) is the observed initial weight ofijklmn the n-th bird adjusted from the overall mean initial weight m ijklmn 2 Batches of production were confounded with season. Batch one was during hot-dry season and batch two in hot-wet season. Batch / seasonal effects will be used interchangeably. Proc mixed REML procedure of SAS (SAS, 1999) was used during the analysis of variance for the various parameters. Least square means were separated into significant differences by the least significant difference procedure (LSD). An estimate of genotype x environment (G X E) chickens between two management systems based on dam-hen REML BLUP values obtained using proc mixed procedure of SAS (SAS, 1999). Economic evaluation of feed costs: Feed costs (FC) during cages were calculated by multiplying total feed intake (TFI) by price per kg feed. Revenue (RV) was calculated by multiplying weight at 20 weeks per bird by farm-gate price of MK142.79 per kg live weight of chicken (price determined from participating farmers’ selling prices for chickens). Initial value of chicken was calculated by multiplying initial weight per bird by farm- gate price per kg live weight. Gross margin over feed cost (GMOFC) was calculated by subtracting FC from RV. Return on FC was calculated as GMOFC as a percentage of FC, while return on bird and FC was calculated as GMOFC as a percentage of initial bird value plus FC. These parameters were normally distributed and were analyzed for their means and standard deviations. Analysis of variance was performed to test effect of batch and sex using general linear model procedure of SAS (SAS, 1999). Results Fixed effects of growth performance: Table 3 shows fixed factors and their effects on overall values for growth traits. Batch of production significantly (p<0.05) influenced SGR while sex of birds, village, management and management x batch interaction significantly (p<0.05) influenced more growth performance traits. Initial live weight was not significant (p>0.05) for overall and daily weight gain. Growth performance between cage-fed and scavenging chicken mates: Birds under cage-fed condition were superior to clutch mates under free- range village conditions (Table 4). Based on overall values, birds under cage-fed management were significantly (p<0.05) 27.04, 39.01, 41.54, 25.25 and 41.18% superior than birds on free-range for live weights, overall weight gains, daily weight gains, SGR and GE, respectively. Gondwe and Wollny: Growth potential of local chickens 4 Table 3: Effects of fixed factors on overall growth characteristics of local chickens under cage and free-range management Effect Weight at 20 weeks Overall weight gain Daily weight gain SGR GEF Batch of production ns ns ns *** ns Sex of birds *** *** ** ns * Village * * *** ns ns Management *** *** *** *** *** Batch x management * * * * ns Initial weight *** ns ns *** *** SGR = specific growth rate (% growth per day); GE = growth efficiency (g final weight gain / g initial weight); Significant levels (F – Test), * = p<0.05; ** = p<0.01; *** = p<0.001; ns = not significant (p>0.05) Table 4: Overall productivity of local chickens under cage-fed and free – range village conditions (lsmeans and standard errors) Trait Batch Management ------------------------------------------------------------------------------------- Cage-fed Free- range (villages) ----------------------------------- ------------------------------------ Mean SE Mean SE Live weight (g) at 20 weeks 1 1131.85 47.36 835.23 49.37a b 2 1022.04 40.33 860.17 46.98a b Overall 1076.94 32.12 847.70 34.55a b Weight gain (g/period) 1 871.75 47.36 575.13 49.37a b 2 761.94 40.33 600.07 46.98a b Overall 816.84 32.12 587.60 34.55a b Daily weight gain (g/day) 1 10.68 0.63 6.61 0.66a b 2 10.58 0.53 8.41 0.63a b Overall 10.63 0.42 7.51 0.46a b SGR (% / day) 1 1.665 0.119 1.167 0.123a b 2 2.045 0.105 1.796 0.116a b Overall 1.855 0.087 1.481 0.091a b GE (g/initial weight) 1 3.471 0.201 2.417 0.209a b 2 3.297 0.169 2.377 0.202a b Overall 3.384 0.133 2.397 0.146a b SGR = specific growth rate (% growth per day); GE = growth efficiency (g final weight gain / g initial weight) Variance components and G X E interaction: Effect of evaluation of feed costs (Table 7) showed a positive management was compared on behaviour of variance mean gross margin with a wide variation. Similarly, both components as shown in Table 5. Management did not rate of return on feed costs and on feed and bird costs affect phenotypic variance for final weight and weight showed wide variation. Batch and sex effects were not gains. Phenotypic variance for daily weight gains and significant (p>0.05, Table 8) SGR were 16.9 and 21.3%, respectively, lower for cage- fed than for free-range birds. On the other hand, phenotypic variance for GE was 7.8% higher in cage-fed than in free-range birds. The between dam-hen variance was larger in free-range birds than in cage-fed birds. The between dam-hen variance was also larger than the within bird variance comparing within free-range birds. Correlation coefficients between cage-fed and free- range local chickens (Table 6) were significantly (p<0.05) different from zero but lower than 50% for all traits except for daily weight gains. SGR had lowest coefficient. Economic evaluation of feed costs in cages: Economic Discussion Growth potential: Demeke (2003) observed higher live weights (1300g, intensive; 985 g, free-range) in Ethiopian local chickens at 20 weeks of age than the results from this study. The difference could in part be due to different initial live weights, which were not adjusted for in Demeke’s study. In this study, feeding started from 9 weeks while in Demeke’s study, feeding started from one day old. However, daily weight gains are comparable. On the other hand, the trend for live weights and daily weight gains agree with those in this study, whereby their birds under intensive system were 24% heavier than birds under scavenging conditions. Gondwe and Wollny: Growth potential of local chickens 5 Table 5: Variance components by management of local chickens for growth traits Component Weight at Weight gain Daily weight SGR GE 20 weeks gain ------------------------ ---------------------- -------------------------- --------------------- --------------------------- ó SE ó SE ó SE ó SE ó SE2 2 2 2 2 Between dam hen Cage 12109 7114 12109 7114 0.8388 0.846 0.0366 0.022 0.2764 0.150 (ó b) Free-range 25107 9404 25107 9404 4.2914 1.891 0.0835 0.050 0.5320 0.2032 Within birds Cage 24416 4426 24416 4426 4.9720 0.841 0.0956 0.017 0.5844 0.102 (ó ) Free - range 11832 2206 11832 2206 2.7045 0.522 0.0845 0.017 0.2666 0.0502 residual Total (ó ) Cage 36525 36525 5.8109 0.1322 0.86082 p Free - range 36939 36939 6.9959 0.1680 0.7986 SGR = specific growth rate (% growth per day); GE = growth efficiency (g final weight gain / g initial weight) Table 6: Correlation coefficients between BLUP values of dam hens for cage fed and free - range managed local chickens Trait Weight at 20 weeks Weight gain Daily weight gain SGR GE Coefficient 0.384 0.384 0.529 0.219 0.376 Significance level ** ** *** * * SGR = specific growth rate (% growth per day); GE = growth efficiency (g final weight gain / g initial weight); Significant levels , * = p<0.05; ** = p<0.01; ***=p<0.001 The status and limitation of scavenging feed resource variance for free-range than for cage-managed birds is base is the main area of research in many countries in expected since free-range birds continued with their Africa (Roberts, 1999; Olukosi and Sonaiya, 2003). In dam-hens, and hence had more common influence their prediction, Olukosi and Sonaiya (2003) estimated from their dam-hens. Nature of data in this comparison a daily feed intake of 20 g per bird per day under did not allow for determining the between bird variance. scavenging condition, which was lower than feed intake Hence, the dam-hen variance includes all genetic, observed in the study (45-59 g per day) under cage maternal and common environment effects. Since management. The findings support views that village effect was taken care of, the common scavenging feed resource base is often inadequate environment in this case was flock of the birds. Birds quantitatively and qualitatively (Huque, 1999; Ndegwa et under cage management were separated from their al., 2001; Dana and Ogle, 2002), depending on flock dam-hens, and hence the low between dam-hen size, environment and season (Gunaratne, 1999; variance that was in this case, a carry over effect. Roberts, 1999). Kitalyi (1998) reported that Hu et al. (1999) reported that maternal effects are scavenging feed is a constraint to local chicken growth and reproductive potential, and emphasized on the need to provide supplement feeds to birds. The major input to birds under cage management was feeding of balanced ration, grower’s mash. Disease intervention may have differed in intensity (between cage-fed and village chickens) but was also taking place in the villages. The difference in growth performance observed in the study was therefore due to feeding management. The significant superiority of birds under cage managed conditions over village free-range birds shows that feed constraint limits expression of growth potential in local chickens. The significant effect of village on final weight, and weight gains is probably due to differences in flock structure, management, scavenging biomass and nutritional pressure under free-range, and disease challenges that vary from village to village. Village effect was more pronounced in birds under free-range than those fed in cages. This is obvious since birds in cages were exposed to village conditions only before they were brought to cages. Variance components: The larger between dam-hen moderate in poultry, only contributing less than 10% of total variance and depend on traits. Falconer (1989) reasons that maternal affects are more important in mammals. Pinchasov (1991) reported that maternal effects in chickens disappear within the first three weeks of chicks life. Prado - Gonzalez et al. (2003) observed significant effect of maternal effect for weights at hatch and fourth week of Creole chickens in Mexico and not thereafter. The variances in this study show potential maternal effects that are high and persist for long time in chickens. Maternal effects from literature arose from the dam-hen influence on the egg (size, weight, shell quality and yolk composition) that is described as the only vehicle for maternal effects in poultry (Sewalent and Johansson, 2000). This is true when birds are raised under intensive system and are separated from their dam-hens at hatch. In free-ranging chickens, a dam-hen takes care of chicks till weaning, hence post hatch maternal effects are expected. The observed variance and the difference show possible post hatch maternal influence on the birds under free-range system. The magnitude of such maternal effect is however, trait and management specific. Gondwe and Wollny: Growth potential of local chickens 6 Table 7: Descriptive statistics for economic parameters of feed costs for chickens under cage management Parameter n Mean SD Minimum Maximum Total feed intake, g 100 3473.5 884.7 1048.0 5536.0 Feed cost per bird, MK 100 83.36 21.23 25.15 132.86 Income per bird, MK 97 109.44 33.47 26.13 194.19 GMOFC per bird. MK 96 25.69 26.62 (48.42) 90.08 Return on feed cost, % 96 34.80 38.10 (42.33) 150.63 Return on bird & feed cost, % 96 24.13 25.90 (29.92) 106.15 n = number of birds; SD = standard deviation; GMOFC = Gross margin over feed cost; MK = Malawi Kwacha (1US$ = MK85.00); Values in brackets are negative. Table 8: Least square means and SE for gross margin over feed cost and rates of returns by batch and sex of birds Factor GMOFC, MK Return on feed cost, % Return on bird and feed cost, % Lsmean SE Lsmean SE Lsmean SE Batch 1 30.68 4.09 43.15 5.85 29.73 3.98 2 24.63 3.98 30.76 5.70 21.59 3.88 Sex Female 22.67 3.24 31.79 4.63 21.97 3.15 Male 32.65 5.01 42.11 7.17 29.35 4.88 GMOFC = Gross margin over feed cost; MK = Malawi Kwacha (1US$ = 85.00 MK) G x E interaction: In general, genetic parameters could chickens, as observed by Demeke (2003). Pedersen be better estimated under these conditions of improved feeding. However, there is need to check if local chickens express some genotype by environment (G x E) interaction, if the results determined under improved feeding could be applicable to local free-ranging environment. In genetic analysis, genetic correlations between 0.9 and 1.0 suggest that two traits are the same (Kerr et al., 2001). Lower correlation coefficients imply small covariance between the two observations. This indicates presence of a stronger interaction (Lin and Togashi, 2002). The dam-hen effect and ranking was expected to be similar in the two management systems and hence show high correlation coefficients. The results obtained showed correlation coefficients ranging from 0.22 - 0.53. Common covariance in these correlations was due to genetic and maternal effects. The low correlation coefficients show possibility of individual G x E interaction being expressed in growth traits for local chickens. This means local chickens are sensitive to environmental changes. Use of parameters determined under improved feeding in this case is only to show potential performance of local chickens under free- ranging environment (Prado - Gonzalez et al., 2003). Economic evaluation of feeding local chickens: Gross margins and rates of returns on feed costs were positive. The observed rates of return are a possible contribution of high costs of feed and poor feed conversion efficiency of these local chickens. This return declined when initial value of birds was included. When all costs, such as labour involved with intensification and treatment are taken into consideration, it would not be economically justifiable to improve feeding of local (2002) found negative gross margins of US$28 for intensively managed local chickens from day old in Zimbabwe. Local chickens are appropriate for the low input scavenging system. Lwesya et al. (2004) reported positive response of supplementing local chickens a locally formulated ration on early chick growth rate and return to lay of hens. However, the ration was not economically viable. This poses a challenge to utilize the advantages of the low-input system while at the same time attempt to achieve their genetic potential. On assumption that the low-input system will prevail in rural communities and that currently there is no suitable breed to substitute local chickens, investigations into optimal feeding strategies ranging from supplement feeding to whole feeding are encouraged within the framework of the farming system. Conclusion: The growth potential of local chickens is not fully exploited under free-range (scavenging) conditions due to inadequate feeds. Feeding management contributes to about 30% of their growth potential. Growth of local chickens can be enhanced through improved management under free-ranging conditions. The option of improved feeding of local chickens under confined conditions is, however, economically not attractive enough to warrant farmers adopting it. References Aini, I., 1990. Indigenous chicken production in South – East Asia. World’s Poult. Sci. J., 46: 51 - 57. Dana, N. and B. Ogle, 2002. Effects of scavenging on diet selection and performance of Rhode Island Red and Fayoumi breeds of chicken offered a choice of energy or protein feeds. Trop. Anim. Health Prod., 34: 417-429. Gondwe and Wollny: Growth potential of local chickens Local names reported by Gondwe et al. (1999).1 7 Demeke, S., 2003. Growth performance and survival of Lin, C.Y. and K. Togashi, 2002. Genetic improvement in local and White Leghorn chickens under the presence of genotype by environment scavenging and intensive systems of management interaction. Anim. Sci. J., 73: 3-11. in Ethiopia. Research for Rural Development, 15, Lwesya, H., R.K.D. Phoya, A.C.L. Safalaoh and T.N.P. 11: 2003. http://www.cipav.org.co/lrrd/lrrd 15/11/ Gondwe, 2004. Rearing chicks in enclosure under deme1511.htm village conditions: Effect on chick growth and Falconer, D.S., 1989. Introduction to quantitative reproductive performance of mother hens. Livestock genetics. ELBS Third Edition. Longman Scientific Research for Rural Development. Vol. 16, Art. #89. and Technical, England. Retrieved December 7, 2004, from Gondwe, T.N.P., A.J.D. Ambali, F.C. Chilera, H. Lwesya, http://www.cipav.org.co/lrrd/lrrd16/11/ wesr16089. and C. Wollny, 1999. Rural poultry biodiversity in htm Lilongwe and Mzuzu Agricultural Development National Statistical Office, 2003. Malawi in Figures 2003. Divisions (ADD), Malawi. Malawi J. Sci. Tec., 5: 17- www.nso.malawi.net. 25. Ndegwa, J.M., R. Mead, P. Norrish, C.W. Kimani and A.M. Gondwe, T.N.P. and C.B.A. Wollny, 2002. Traditional Wachira, 2001. The growth performance of breeding systems in smallholder rural poultry in indigenous Kenyan chickens fed diets containing Malawi. In: Session 25, Developing Sustainable different levels of protein during rearing. Trop. Anim. Breeding Strategies in Medium to Low input Health Prod., 33: 441-448. Systems. 7 World Congress on Genetics Applied Olukosi, O.A. and E.B. Sonaiya, 2003. Determination ofth to Livestock Production, August 19-23. Montpellier, the quantity of scavengeable feed for family poultry France. on free range. Livestock Research for Rural Guèye, H.F., 1998. Village egg and fowl meat production Development, 15, 5:2003. http://www.cipav.org.co/ in Africa. World’s Poult. Sci. J., 54: 73-86. lrrd/lrrd15 /5/0luk 155.htm Gunaratne, S.P., 1999. Feeding and nutrition of Pedersen, C.V., 2002. Production of semi-scavenging scavenging village chickens. First INFPD / FAO chickens in Zimbabwe. PhD Thesis. Royal Electronic Conference of Family Poultry. Veterinary and Agricultural University, Copenhagen, http://www.fao.org/ag/aga/agap/lpa/fampo1/fampo. Denmark. htm Pinchasov, Y., 1991. Relationship between weight of Hu, Y.H., J.P. Poivey, R. Rouvier, C.T. Wang and C. Tai, hatching eggs and subsequent early performance 1999. Heritabilities and genetic correlations of body of broiler chicks. Br. Poult. Sci., 32: 109-115. weights and feather length in growing muscovy Prado - Gonzalez, E.A., L. Ramirez - Avila, and J.C. selected in Taiwan. Br. Poult. Sci., 40: 605-612. Segura - Correa, 2003. Genetic parameters for body Huque, Q.M.E., 1999. Nutritional status of family poultry weights of Creole chickens from Southern Mexico in Bangladesh. First INFPD / FAO Electronic using an animal model. Livest. Res. Rural Dev., 15, Conference of Family Poultry. http://www.fao. org/ag/ 1:2003. http://www.cipav.org.co/lrrd/lrrd 15/1/ prad aga/agap/lpa/fampo1 /fampo. htm 151.htm Kerr, C.L., R.H. Hammerstedt and G.F. Barbato, 2001. Roberts, J.A., 1999. Utilization of poultry feed resources Effects of selection for exponential growth rate at by smallholders in the villages of developing different ages on reproduction in chickens. Avian countries. In. Poultry as a tool in poverty eradication and Poult. Bio. Rev.,12: 127-136. and promotion of gender equity. Kitalyi, A.J., 1998. Village chicken production in rural http://www.husdyr.kvl.dk/htm/php/tune 99/28- Africa, household food security and gender issues. Roberts.htm Animal Production and Health Paper No. 142. Food SAS., 1999. SAS /STAT Users Guide, Version 8.1, Cary, and Agricultural Organization of the United Nations, NC, USA. Rome. Sewalent, A. and K. Johansson, 2000. Egg weight and Kitalyi, A.J., 1999. Family poultry management systems reproduction traits in laying hens: estimation of in Africa. First INFPD / FAO Electronic Conference of direct and maternal genetic effects using Bayesian Family Poultry. http://www.fao.org/ag/aga/agap/lpa/ approach via Gibbs Sampling. J. Anim. Sci., 70: 9- fampo1/fampo.htm 16.