|Livestock Research for Rural Development 9 (3) 1997||
Citation of this paper
Bangladesh Livestock Research Institute Savar 1341 Dhaka Bangladesh
*Danish Institute of Animal Science, 8830 Tjele, Denmark. E-mail: firstname.lastname@example.org
**Bangladesh Smallholder Livestock Project, Strindbergsvej 104, 2500 Valby, Denmark. E-mail: email@example.com
***Institute of Political Science, University of Aarhus, 8000 Aarhus C, Denmark. E-mail: firstname.lastname@example.org
A model for semi-scavenging poultry rearing has been developed in Bangladesh, which is applied with considerable success to the benefit of rural poor women. 297 landless and mostly illiterate women who participate in the Bangladesh Smallholder Livestock project acted as hosts in a field trial comprising 1272 pullets of eight breed combinations from four hatches (May August November and January) which were tested for egg production in three agro-ecological zones. The objectives of the trial were: to test a breeding system where hybrid parents are used as grand parents for female parents; and to compare the performance of different breed combinations under semi scavenging conditions.
The results of the trial showed that the combination Rhode Island Red x Fayoumi had the highest egg production, lowest mortality and highest profit per hen.. However, none of the combinations produced to their genetic potential as egg production per hen-day was 35 - 40 per cent and age of first egg 32 to 34 weeks. In this experiment the effect of location was higher than that of breed combination and season. It is concluded that management aspects which account for local variation in feed available for scavenging and supplementation and reduction in mortality are important factors to address to improve the technical efficiency of the model.
The rural poultry production system in Bangladesh is typically a smallholder free range scavenging operation. The indigenous (desi) hens in that system have small body size of around 1,140 g (Huque 1993) and produce 45 eggs per hen per year (Ahmed and Hasnath 1983). They constitute about 80% of the country=s chicken population. Their egg laying potential was investigated under intensive conditions by Sazzad (1986) and it was found to be 64 eggs per hen per year with a behaviour of pronounced brooding.
Desi birds are ideal for reproduction by natural brooding. However, in a semi-scavenging model with artificial hatching (Saleque and Mustafa 1997; Jensen 1996 1997), the egg yield is too low. Some experiments have been conducted concerning the introduction of exotic, high yielding poultry breeds and their crosses into the scavenging and semi-scavenging system (Ahmed et al undated; Amin et al 1992; Chandrasiri et al 1994; Quader et al 1989; Hossain et al 1992; Mahbub et al 1992; Paul et al 1995; Sazzad 1992; Roberts and Senarantne 1992; Rahman et al 1995). They showed that exotic breeds have a higher egg yield than desi hens under both scavenging and semi-scavenging conditions, but this was accompanied by a high mortality rate in the scavenging situation.
A model for semi-scavenging poultry rearing has been developed in Bangladesh comprising small units of rice husk hatcheries, breeders, chicken rearers, and producers with a small flock of 10 hens (Saleque and Mustafa 1997; Jensen 1996 1997).
The Directorate of Livestock Services (DLS) maintains breeds of White Leghorn (WHL) Rhode Island Red (RIR) and Fayoumi at the Government Poultry Farms and supplies parent stocks to the breeders in flocks of 25 hens per breeder. Earlier work has shown the most successful parent breeds are RIR male and Fayoumi females and Amber (personal communication) named the cross between these two breeds as Sonali (which means golden in Bengali) and this name provides appropriate connotations of its brown to red colour. The main constraint of using Fayoumi as the parent hen is the low egg yield of 141 eggs (El Salhia 1984; cited by Horst 1989).
Using commercial hybrids as parent hens and an improved breed as parent males (eg: New Hampshire) has been a common system for smallholders in countries with industrial poultry production including Denmark. The main advantage of such a system is the low cost of parent stock and consequently lower cost of day old chickens.
The constraint of using Fayoumi as parent hens in the semi-scavenging model in Bangladesh is the low egg yield. The small breeders keep the hens in confinement and feed them with balanced feed and consequently the production cost of hatching eggs is rather high.
Using commercial hybrids as parent hens will reduce the cost of day old chickens and by using another breed as parent male it is still possible to transmit scavenging traits into the chickens used at village level.
The objectives of the experiment were:
The results of the experiment will be used to develop further the semi-scavenging system in the Bangladesh Smallholder Livestock Development Project and - it is expected - in other ongoing and future poultry projects.
The study covered a period of three years with establishment of the parent population in the autumn of 1993 and completion of last field trials in August 1996.
Breeding for production of the chicken for the experiment was conducted at the DLS Central Poultry Farm Mirpur Dhaka with 1 259 females and 177 males of Lohmann Brown (AB) RIR WLH Fayoumi AxRIR and RIRxFayoumi as parents. Artificial insemination was performed to produce the experimental chicken of the eight breed combinations including one commercial hybrid as shown in table 1.
Eggs were hatched in four batches with intervals of eleven weeks. The chicken were reared by chick rearers (see the model in Saleq and Mustafa 1997; Jensen 1996 1997) living in the experimental sites and grown in confinement up to 8 weeks of age in flocks of 250 chicken per rearer.
|Table 1. Distribution of experimental pullets / key rearer by districts and seasons|
Numbers of pullets / key rearers
Numbers of pullets / key rearers
|Breed combinat-ions||Jessore||Manikgonj||Rajshahi||Hatch - 1 (May)||Hatch - 2 (August)||Hatch - 3 (Novemb-er)||Hatch - 4 (January)|
|1. Lohmann brown (AB)||29 / 10||66 / 19||58 / 10||26 / 10||58 / 10||40 / 9||29 / 10|
|2. (A x RIR) x Fayoumi||52 / 10||31 / 7||82 / 18||43 / 9||52 / 10||39 / 9||31 / 7|
|3. Fayoumi x AB||60 / 18||43 / 10||48 / 10||24 / 8||43 / 10||36 / 10||48 / 10|
|4. RIR x AB||32 / 10||33 / 7||88 / 17||56 / 10||32 / 10||32 / 7||33 / 7|
|5. RIR x Fayoumi||43 / 10||63 / 18||54 / 10||28 / 9||54 / 10||35 / 9||43 / 10|
|6. RIR x WLH||88 / 20||40 / 9||45 / 10||34 / 10||40 / 9||54 / 10||45 / 10|
|7. (RIR x Fayoumi) x AB||52 / 10||54 / 18||36 / 8||21 / 9||36 / 8||52 / 10||33 / 9|
|8. WLH x AB||52 / 10||45 / 10||78 / 18||40 / 8||52 / 10||45 / 10||38 / 10|
|Total||408 / 98||375 / 98||489 / 101||272 / 73||367 / 77||333 / 74||300 / 73|
1272 / 297
1272 / 297
Selected key rearers (Saleq and Mustafa 1997) received about 4 - 6 pullets at eight weeks of age from one particular breed combination and a small amount of feed to ensure gradual adaptation of the birds to the semi scavenging system.
For the present paper records were used of a total of 1 272 pullets from 8 breed combinations placed with 297 landless and mostly illiterate rural women participating in the Bangladesh Smallholder Livestock Development Project (see accompanying paper by Alam 1997 for details on socio-economic status of the women). Primary contact with the women was by the NGO Bangladesh Rural Advancement Committee (BRAC). The birds were distributed to women in three districts who kept the birds until they were about 1.5 years old or had produced eggs for almost one year. At start of lay the number of pullets per key rearer ranged from 2 to 6. The distribution of pullets / key rearer is shown in table 1. The birds were allowed to scavenge during the day in and around the homestead and in the neighborhood after the crop was harvested and they were kept in a shelter at night and during unfavorable weather conditions. Farm by-products (wheat bran poor quality paddy broken rice rice polishings bran of pulse crops) or household waste (leftover cooked rice and other dining and kitchen wastes) or small amounts of grain or balanced diets for laying hens were offered as a supplement either as a single ingredient or in a mixture of 2 to 3 ingredients.
The birds were housed in a bamboo shelter when offered supplementary feed. There was a continuous supply of drinking water in the shelter and there was easy access during the day when they were scavenging
De-worming was done every two months. Vaccination against Newcastle and Fowl Pox diseases was according to a set program. Data regarding egg production mortality and supplementary feed were recorded twice a week by six field assistants. The protein and energy contents of the supplementary feed were calculated on the basis of book values.
Supplementary feed cost (see in the appendix for prices used) and gross margin per hen were estimated on the basis of free market prices.
Data were subjected to analysis of variance by the least square principle (SAS 1990). The models that were used were as follows:
is the observed value of the mte farm mean having its start in the lte month belonging to the kte breed combination situated in the jte location from the ite hatch .
is the fixed effect of the ite hatch . i = 1 4.
is the fixed effect of the jte location j = Jessore or Manikgonj or Rajshahi
is the fixed effect of the kte breed combination k = 1 8.
is the fixed effect of the lte month on start of lay l = 1 .14
is the fixed effect of duration of the production in months
is the fixed effect of the age at start of lay in months
is the regression of the duration (D) on the observed production
is the regression of age at start of lay on the observed production
is the unexplainable residual
The effects were simultaneously analyzed by use of an analysis of variance using the GLM procedure of SAS (1990) and then tested by the F-test. Multiple comparisons among the various locations breeds and hatches were tested by their least square means corrected for other effects in the model.
Mortality was calculated according to: (3)
mortdt = numbers of chickens dead due to diseases during the tth month/(numbers of chicks at beginning of the month - number of chickens sold in the month)
Cmortdt = Mortality due to diseases in the tth month adjusted for mortality in earlier months
Cmortpt = Mortality due to predators in the tth month adjusted for mortality in earlier months
A similar equation was used for mortality due to predators.
In order to calculate mortality to the same length of production period the following formula was used:
Survt = 1 - Cmortdt - Cmortpt or the survivability in the actual production period in a farm measured at "t" month
= is the geometric mean of mortality per month.
The experiment was designed to test eight different breed combinations in three locations over four hatches within one year.
The hens used in the semi-scavenging model (Saleq and Mustafa 1997; Jensen 1996,1997) are breeds and breed combinations based mainly on White Leghorn, Rhode Island Red and Fayoumi. The breeding stocks are kept in small flocks of 25 hens in confinement and fed with balanced feed. The low egg yield, combined with high feed expenditures, make cost of producing eggs for hatching rather high. A way to decrease the production cost of hatching eggs is to use commercial hybrids as parent females and another breed as parent males. This will give the breeders the full advantages of the high genetic potential in the hybrids and the key rearers (Saleq and Mustafa 1997; Jensen 1996 1997) will receive hens with a higher genetic potential for egg production traits while the scavenging traits are provided through the male line. To increase further the genetic potential at key rearer level a combination (AxRIR)xFayoumi was established where A is the male line in Lohmann parent stock.
The genetic potential for egg production traits is increased in breed combinations where hybrids are one of the parental breeds. However, the question is: will this higher potential express itself under semi-scavenging conditions?
The combination RIR x Fayoumi (Table 2) had the best performance with the highest egg production, lowest mortality and highest profit per hen. However, in statistical terms the egg production differences are only significantly different to four of the other combinations, while for mortality it is superior to only to one other breed combination. It is interesting to note that the combination (AB) with the highest genetic potential for egg production did not express this genetic potential under the semi-scavenging conditions. Their parent lines averaged 63% hen-day egg production at Mirpur Central Poultry Farm (Rahman 1995) in comparison to 35 - 40% in the present experiment and started laying at the age of 20-21 weeks in contrast to 32 - 34 weeks recorded for their offspring in the field.
|Table 2: Summary of performance of the different breed combinations.|
( Least square mean values)
|Parameters||AB (Lohmann brown)||
(Ax RIR) x Fayoumi
|Fayoumi x AB||RIR x AB||RIR x Fayoumi||RIR x WLH||(RIR x Fayoumi) x AB||WLH x AB|
|Age of first egg (week)||34.5||32||32||34||33||32||32.5||34|
|Eggs/hen/year *(hen day).
Actual no of eggs per hen. (hen day)
|Mortality % (excluding predator loss)||22.1ab||35.0b||27.6ab||32.6b||15.9a||25.2ab||21.2ab||22.8ab|
|Mortality % due to predator||1.38||0.87||1.40||2.71||2.09||0||5.31||2.61|
|Supplementary energy, kcal./bird / day||146b||122a||136ab||144b||130a||134ab||146b||135ab|
|Suppl. feed cost, Taka/bird/laying/yr||202b||177a||193ab||203b||184a||190ab||205b||192ab|
|Gross margin, Taka/hen up to one laying year (Income - Cost)||134b||169ab||113c||133b||205a||145b||132b||154b|
|Egg production period (Months)||8.54||9.84||9.17||9.37||9.84||9.13||8.74||8.97|
Figures with same or no superscript in a row are not significantly
* Corrected for 12 months egg production period.
The combination RIRxFayoumi has previously proven to have superior performance to other presently prevailing breeds in Bangladesh (Amber, personal communication). The results of the present experiment support the earlier finding that the combination RIRxFayoumi is superior to other breeds or breed combinations under the conditions of this study. Apart from that, the combinations with hybrids have the same or better performance as the combination RIRxWL, one of the traditional combinations.
It is known that the seasons have influence on hen performance. Increasing or decreasing day-length (table 3) causes lower or higher egg production and harvesting season influences availability of feed to scavenge or supplement. However, very little is known with respect to the influence of these factors on exotic breeds in the semi-scavenging system and in the geographical position of a country like Bangladesh where the difference in day-length between shortest and longest day is less than three hours (table 3).
|Table 3. Average day length (natural) and harvesting seasons for major grain crops in the experimental sites by months.|
|Average day length (Hours.)||11.0||10.7||10.8||11.3||11.9||12.6||13.1||13.5||13.4||12.9||12.3||11.6|
Sites: j = Jessore m =Manikgonj and r = Rajshahi
The four batches used in the experiment were hatched in May, August, November and December. Thus a comparison of the performance of the hatches automatically introduces a seasonal comparison (table 4). Except for the age of first egg the differences seem insignificant. The decrease in day-length is expected to delay the start of laying for the batch hatched in May and slightly for the batch hatched in August. When day-length begins to increase by end of December this causes an earlier start to laying for the batch hatched in November and to some degree for the batch hatched in January. However, the results do not fully comply with these expectations and it is very likely that other factors, such as availability of feed for scavenging, influence the results.
|Table 4. Comparison of performance by hatch (and season).|
Least square means
|Hatch - 1 (May)||Hatch - 2 (August)||Hatch - 3 (Nov.)||Hatch - 4 (Jan.)|
|Age of first egg (weeks)||36c||30b||28a||37c|
|Eggs / hen / year (hen day)||154a||121b||139ab||138ab|
|Mortality % (excluding predator loss)||13.3||19.7||30.1||38.2|
|Mortality % due to predator||0||1.67||3.12||6.07|
|Supplementary energy, kcal /bird/day||154c||137b||131ab||125a|
|Supplementary protein, g/bird/day||7.6c||6.8b||6.5ab||6.1a|
|Supplementary feed cost,(Taka/bird/laying year||215c||195b||188ab||176.a|
|Gross margin,Taka/hen up to one laying year (Income - Cost)||158ab||128b||144ab||163a|
|Egg production period,months||8.75b||10.5a||9.90a||7.66c|
Figures with same or no superscript in a row are not significantly different (P<0.05).
The main conclusion from the experiment is that there are seasonal effects. However, while the causal direction of increasing and decreasing daylength on egg production is well understood in a semi-scavenging system this may be confounded by other factors such as availability of feed and weather.
The prevailing scavenging condition is naturally expected not only to influence the performance of hens in the semi-scavenging system but also the amount of feed available to supplement. The experiment was not designed to measure the differences in the local scavenging conditions. On the other hand three different ecological zones were deliberately chosen which had different cropping patterns as specified in Appendix table 1. The results are presented in table 5.
The egg laying performance of the hens in the sugarcane/grain area was by far the lowest. But there was also a difference in the mortality between the grain/fiber area and the grain dominant area. This may have been caused by flooding in Manikgonj (the grain dominant area) and the protection a sugarcane field provides to predators (Rajshahi). The present study was not designed to study such effects. However, the socio-economic study conducted under the same project and reported by Alam (1997) also found location to have a significant influence on production.
|Table 5. Performance of experimental birds reared under semi scavenging condition at three locations with different cropping patterns.|
Least square mean
Grain/f iber (Jessore)
Grain dominant (Manikgonj)
|Age of first egg, week||34b||31a||34b|
|Eggs/hen/year, hen day||154a||157a||103b|
|Mortality, % (excluding predator loss)||12.3a||43.9b||19.8a|
|Mortality % due to predator||0.30a||0a||6.27b|
|Supplementary feed cost, Taka /bird/laying year||243c||159a||178b|
|Gross margin, Taka/hen up to one laying year (Income - Cost)||140b||224a||80.3c|
|Egg production period, months||10.4a||8.62b||8.62b|
Figures with same superscript in a row are not significantly different (P<0.05).
The effect of location in this experiment appears to have been higher than that due to breed or to season. Consequently, management aspects due to local conditions, which influence supplementary feed available, opportunity to scavenge and reduction of mortality including protection against predators, are essential elements in development of the semi-scavenging poultry model. We feel that while the present work has established the levels of production that can be expected at present with this model in Bangladesh it has raised many more questions which should be addressed in future research.
The authors gratefully acknowledge the support and cooperation of the women who hosted the experiment in the field. We acknowledge the support of Dr Nazir Ahmed former Director General Directorate of Livestock Services Dr. Kazi Abdul Fattah Director General Directorate of Livestock Services Dr. A. Jalil Ambar Project Director and Dr. Ziauddin Ahmed Liaison Officer of the Bangladesh Smallholder Livestock Project and the funding from IFAD and DANIDA. The DANIDA Fellowship Centre funded a stay at the Danish National Animal Science Institute by the senior author. In the field, staff of Bangladesh Rural Advancement Committee (BRAC) and of the Directorate of Livestock Services provided valuable support.
Ahmed A and Hasnath M A 1983 A study on the heritability estimates of body weights of indigenous chickens. Bangladesh Vet. J: 17: 19-24.
Amber Jalil (no date) Personal communications
Ahmed N U R , Amin N P, Magor N I, Miah N, Islam A, Soha A, Khan H and Ali H No date. Livestock related FSR activities of the Bangladesh Rice Research Institute Rice Farming Research Division Bangladesh Rice Research Institute Gazipur PP.171.
Alam J 1997 Impact of Smallholder Livestock Development Project in some Selected Areas of Bangladesh. Livestock Research for Rural Development. Volume 9, Number 4: 17-:
Amin R M, Hoque M M, Islam Q M S and Khan M M R 1992 The performance of cross bred and indigenous chicken under scavenging system. Bangladesh Journal of Animal Science 21(1-2):77-81.
Chandrasiri A D N, Gunaratne S P, Wickramaratne S H G and Roberts J A 1994 The egg and meat production potential of village chickens under scavenging system. In: Sustainable animal production and the environment. Proc. of the 7th. AAAP Animal Science Congress Bali Indonesia pp 73-75.
Horst P 1989 Native fowl as reservoir for genomes and major genes with direct and indirect effects on the adaptability and their potential for tropically oriented breeding plans. Arch. Geflugelk. 53(3):93-101.
Hossain M B, Islam M B and Hossain M A 1992 Introduction of chick rearing unit among the rural house hold of Bangladesh. Bangladesh Journal of Animal Science 21(1-2): 15-19.
Huque Q M E 1993 Indigenous animal species conservation versus intensive production in Bangladesh. Asian Livestock xviii (10):132-137.
Jensen Hans Askov 1996 Semi-Scavenging Model for Rural Poultry Holding. Invited paper presented at the Symposium on: Rural Poultry Development with focus on Employment Income and role of Women. Proceedings of XX World Poultry Congress New Delhi India September 2-8 1996 Volume 1.
Jensen Hans Askov 1997 Semi-scavenging Poultry Flock. In: Frands Dolberg and Poul Henning Petersen (eds) Integrated Farming in Human Development. Proceedings of a Workshop. March 25-29 1996. Tune Landboskole Denmark: 56-63.
Mahbub A S M, Uddin M K, Hoque A K M Z and Jahangir 1992 Chick rearing unit as affected by different levels of rationing with partial scavenging at rural condition of Bangladesh. Bangladesh Journal of Animal Science 21(1-2):119-121.
Mukharjee T K 1990 Breeding and selection programs in developing countries. In: Poultry Breeding and Genetics. Ed. R. D. Crawford (Elsevier). pp. 1049-1060.
Paul D C , Huque Q M E and Haque M F 1995 Growth pattern of exotic pullets under farmers management. Bangladesh Journal of Scientific and Industrial Research 30(4): 15-20.
Quader M A, Rahman M M, Moula M G and Khan M R 1989 A comparative study on adaptability of different exotic and local breeds of poultry under two management systems of rearing. Research Report (Rabi 1988-89). On-farm Research Division Region II Bangladesh Agricultural Research Institute Jamalpur.
Rahman M M 1995 HYV Crossbred under semi-scavenging conditions. Paper presented at a seminar organised by the Bangladesh Smallholder Livestock Project at the Officers Training Institute Savar July 12 and 13 1995.
Rahman M M, Islam M N and Sarker N R 1995 An adaptability study of three improved types of layer chicken under rural conditions at Naikhongchari FSR site. Bangladesh Journal of Livestock Research . paper accepted for publication.
Saleque Md A and Shams Mustafa 1997 Landless women and poultry. In: . Integrated Farming in Human Development (Editors: Frands Dolberg and Poul Henning Petersen).. Proceedings of a Workshop. March 25-29 1996. Tune Landboskole Denmark: 37-55.
Sazzad M H 1986 Reproductive performance of desi (indigenous) hens under scavenging and intensive systems of rearing. Proceedings of the First Annual Livestock Research Workshop Bangladesh Livestock Research Institute Savar Dhaka Bangladesh pp. 63-67.
Sazzad M H 1992 Comparative study on egg production and feed efficiency of different breeds of poultry under intensive and rural conditions in Bangladesh . Livestock Research for Rural Development. Volume 4, Number 3: 65-69.
SAS 1990 SAS procedure=s guide Version 6 3rd Edition. SAS Inst. Inc. Cary. NC USA
Received 10 May 1997
|Table1. Cropping patterns of the three locations during the experiment.|
|Grain dominant (Manikgonj)||Grain / fiber
|Sugarcane / grain (Rajshahi)|
|Agro-ecological zone||Low Ganges river flood plain.||Medium high||High Ganges flood plain|
|Major cropping patterns||Rice (B. Aman +B.Aus) -
Rice (B. Aman +B.Aus) - Mustard.
Mustard - Rice (Boro).
Rice (Boro) - Rice (Local) - B.Aman (transplanted).
Rice (B.Aus+B. Aman) / Jute - Khesari.
Rice (B.Aus+B. Aman) - Wheat / Potato / Mustard.
Rice (B.Aman) - Ground Nut.
Rice (B.Aman) - Sesame.
|Rice (B.Aus) / Jute - Rice
Rice (B.Aus) / Jute - Rice (T.Aman) - Wheat / Pulse / Oil seed.
Rice (B.Aus) - Rice (T.Aman) - Rice (Boro).
Rice (B.Aus) / Jute - Wheat / Pulse - Oil seed / Vegetables.
|Sugarcane - Lentil.
Sugarcane - Onion / Garlic.
Rice (B.Aus)- Wheat.
Rice (B.Aus) - Lentil / Mustard.
Rice (B.Aus) - Potato/Maize.
Rice (T.Aman) - Wheat.
Rice (T.Aman) - Vegetables/Maize.
|Table 2. Basis of calculation of nutritive values and cost (Taka 43.8 = 1 US $) of supplemented feed ingredients.(dry matter basis)|
|Supplemented feed ingredients||CP, %||ME,kcal / kg||Cost, Taka / kg|
|F1||Wheat Flour||12||3000||9 10|
|F2||Wheat bran Paddy or small amount of Fish waste / Paddy / Snail / Cockroach / Earth worm||14 - 15||1500 - 1600||5 - 6|
|F3||Rice polish (40 % rice husk)||7 - 8||1000 - 1200||3 - 4|
|F4||Broken rice||8||2300||6 - 7|
|F5||Cooked rice (70 % moisture) left over after meal||8||2400||4|
|F7||Balanced feed for layer (BRAC)||17||2550||9 - 10|
|Table 3. Basis of calculation of gross benefit|
|1 spent hen (each)||70|
|Expenditure||1 pullet (each)||35|
|Supplementary feed / bird - Laying||calculated on the basis of field notes|
|Supplementary feed / bird -
Growing (9th. week to age of first egg).
Growing period supplementary feed /bird /day = 0.5 * Laying period supplementary feed /bird /day.
|Gross margin = (Income - Expenditure)|
Go to top