Chapter 1 General Introduction

Introduction

Poultry farming is emerging as a strong agro-based industry in Bangladesh. The traditional rearing system has been a common feature in rural households for many years. Beside this, a commercial farming system has developed in the country during the last two decades. P oultry farming is gradually taking the shape of a large

industry, and it is now one of the intensive forms of agri-business in our country.

Small-scale commercial poultry farming has proved to be a beneficial occupation for the small and marginal farmers. It is also a potential source of income generation that creates employment opportunity for the educated and unemployed youths, and distressed women. Presently, poultry meat and eggs provides the cheapest quality animal protein to the millions of low -income people. Thus , the industry has shown its great potential in income generation, and poverty alleviation, as well as in improving human nutrition.

In order to achieve the Millennium Development Goal (MDG) , Bangladesh is committed to develop the poultry sector. Raising poultry has become a poverty reduction tool in Bangladesh with about 100 NGOs involved in promoting the poultry sector for reducing poverty as well as for women’s empowerment. The Government has now liberalized loan procedures, lowered interest rates, and introduced tax holidays that have encouraged the growth of poultry sector in the country. There are around 1.5 lakh small, medium and large poultry farms in the country. The total poultry population, both backyard and commercial, accounts to approxima tely 246 million, provid ing 5400 million pieces of eggs annually and nearly 15% of total animal protein. The total investment in the commercial poultry sector is around Taka 13 thousand crore (Taka 130 billion) of which a bout 6 thousand crore (60 billion) comes from landing institutions. This sector employs about 5 million people of the country and has experienced a long-term growth rate of about 4.5%, which is one of the highest in the economy and is believed to have accomplished a silent revolution in Bangladesh. However, this sector is now facing a hazardous situation with recent outbreak of Avian Influenza (AI) posing a great threat to the growing poultry industry.

In mid-December 2003, an epizootic of high pathogenic avian influenza (HPAI) started in the Far-East, quickly spr ead to many countries of South-East Asia,

and subsequently, to other parts of Asia, Europe and Africa. The disease has moved westward and affected more than 63 countries of 5 continents across the globe. Bangladesh has experienced HPAI infection first in early 2007 and after an initial decline the incidence reappeared in the country in late 2007 to continue till early 2008 and the country is still vulnerable to this disease.

Avian influenza (AI) was recognized as a highly pathogenic viral disease of poultry in 1878 and was used to be called “fowl plague”. The virus belongs to the family Orthomyxoviridae. Many species of birds have been shown to be susceptible to infection with influenza A viruses. Low pathogenic avian influenza (LPAI) viruses are frequently isolated from wild birds, especially waterfowl (Alexander, 1982) that acts as a reservoir. One special feature of Avian Influenza Virus (AIV) is its ability to mutate easily from low pathogenic (LP) one to high pathogenic (HP), but this is only observed when LPAIV infects industrial flocks. Mutation from LP to HP virus occurs mainly by antigenic drift or antigenic shift (Saif et al. 2003). Incubation period of this disease is usually 3 to 7 days in individual birds and may increase up to 21 days in a flock. In the case of laying hens, they may lay soft-shelled eggs at first, but will soon stop laying eggs. Combs and wattles become cyanotic, and may have petechial or ecchymotic haemorrhages at their tips. The head is swollen and oedematous. The mortality rate varies but could reach up to 100%. In broilers, signs of the disease are frequently less obvious with severe depression, inappetence and a marked increase in mortality.

Avian influenza is an enveloped virus with single stranded segmented RNA genome. The surface of the virion is covered with two different types of spikes or projections namely haemagglutinin (HA) and neuraminidase (NA) proteins. The influenza viruses are primarily classified into Types A, B and C on the basis of differences of their nucleoprotein and matrix protein antigens. Type A virus is further categorized into subtypes according to the HA and NA antigens it carry on its surface. There are 16 haemagglutinin subtypes (H1 to H16) and 9 neuraminidase subtypes (N1 to N9) of influenza A virus es (Islam and Giasuddin, 2007). H5 and H7 viruses may circulate in the nature as low pathogenic strains for certain period of time before they mutate to highly pathogenic strains. Therefore, irrespective of pathogenicity, all H5 and H7 AI viruses are considered as ‘N otifiable ’ avian influenza (NAI) viruses. All types of birds can be infected with AI viruses that may or may not cause clinical disease.

Usually, “avian influenza virus” refers to influenza A virus found chiefly in birds, but infections with the virus may occur in humans. C onfirmed cases of human infection from several subtypes of avian influenza infection have been reported since 1997. Most cases of avian influenza infection in humans have resulted from contact with infected poultry (e.g., domesticated chicken, ducks, and turkeys) or surfaces contaminated with secretion/excretions from infected birds. Symptoms of avian influenza in humans have ranged from typical human influenza-like symptoms (e.g., fever, cough, sore throat, and muscle aches) to eye infections, pneumonia, severe respiratory diseases (such as acute respiratory distress), and other severe and life-threatening complications. Till 25 June 2008, 15 countries have experienced human casualties comprising 385 cases, of which 243 cases were fatal. Nevertheless, Bangladesh was considered free from human infection up to May 21, 2008. However, one human case was announced by the authorities of Bangladesh on May 22, 2008. The case was confirmed by Center for Disease Control and Prevention (CDC) in Atlanta, USA. With this case, Bangladesh has become the 15^th country to report human influenza H5N1 virus infection. The other 14 countries are:Azerbaijan, Cambodia, China, Djibouti, Egypt. Indonesia, Iraq, Lao, Myanmar, Nigeria, Pakistan, Thailand, Turkey and Viet Nam. All the human cases were associated with the ongoing H5N1 outbreaks in poultry and wild birds in Asia and parts of Europe, the Near East and Africa, most cases have occurred in previously healthy children and young adults and have resulted from direct or close contact with H5N1-infected poultry or H5N1-contaminated surfaces. In general, H5N1 remains a very rare disease in people. The H5N1 virus does not infect humans easily, and even if a person is infected, it is very difficult for the virus to spread to another person.

Bangladesh was considered as a high-risk country for infection with HPAI since the outbreak in early 2006 in the neighboring countries like India and Myanmar with whom Bangladesh shares a long border. Despite this risk factor, Bangladesh remained free from HPAI till February 2007, due to effective prevention and control measures taken by the Government. However, the country first experienced HPAI in early 2007 and the National Reference Laboratory for Avian Influenza (NRL-AI) at BLRI diagnosed and confirmed the presence of H5 sub-type virus on 15 March, 2007. Up to 15^th May 2008, NRL-AI has analyzed 399 samples of which 286 were found to be H5 and three H9 positive. Some of those samples were re-examined and reconfirmed with same results by International Reference Laborator y in U.K. and a regional laboratory in Thailand. In February 2008, NRL-AI recorded 93 incidences of AI, which was the highest in number in a month. In the next month, the number declined to 70 positive cases. In April 2008, there were only 8 incidences of HPAI whereas in May 2008 only one case was reported. If we look back to the last one year’s outbreak frequencies, it becomes clear that usually from the month of November the number of incidence goes up and it begins to decline significantly from April.

After the outbreak of AI in early 2007, around 1.6 million birds were culled and disposed off. Moreover, 2.2 million eggs were destroyed concurrently. The disease was detected from 287 incident points covering 128 Upazillas, 14 Metropolitan Thana of 47 Districts. As a result of death and culling of poultry, sharp price fall of poultry and eggs, price hike of poultry feed, bird flu panic and lack of insurance coverage, as many as 17,990 poultry farms were pushed out of business since July 2007 to February 2008. The amount of loss estimated by the Poultry Farm Owners’ Associations of Bangladesh was worked out to be Tk. 4,165 crore (Taka 41658 million) and around 5 lakh people were claimed to have became jobless due to the outbreak of this disease. A recent study conducted by BLRI showed that broiler prices declined by 16% and egg prices declined by 17% in May 2007 than that of the previous year (Alam et al., 2008). The situation was even worse in February 2008, which however, improved very quickly in March 2008 and the sale of eggs and chicken increased in the city and elsewhere in the country over the next few weeks. Preventive initiatives and awareness building activities of Government and Non Government Organizations regarding Avian Influenza have proved to be useful in this regard. However, a shortage of supply of poultry and eggs is still predicted in the next few months due to effect of the outbreak of this disease.

Influenza viruses have the ability to mutate and scientists are concerned that H5N1 virus might one day be able to infect human and spread easily from one person to another. If H5N1 virus gain the capacity to spread easily from person to person, an influenza pandemic (worldwide outbreak of disease) might begin. None can predict when such a pandemic would occur. However, experts around the world are watching the H5N1 situation in Asia and Europe very closely.

Under the above cercumsentance s, a study was undertaken with three distinct objectives to find out the origin and means of spreading of AIV and to assess the socio-economic impacts  of its outbreak in Bangladesh. The specific objectives of the study were as follows :

Organization of the Report

This report consists of five Chapters. The present chapter is followed by presentation of evidence and discussion on the phylogenetic study regarding the origin of the disease. Chapter 3 presents results of the epidemiological study. The direct and indirect economic losses due to AI and its impact on prices and sales volume of poultry meat and eggs are presented in Chapter 4. Finally, the conclusions and recommendations are given in Chapter 5.

Chapter 2 Phylogenetic Study

Phylogenetic Analysis of Highly Pathogenic Avian Influenza (H5N1) Virus Isolates of Bangladesh

M.R. Islam, M. Giasuddin, M.A. Samad, M.J.F.A Taimur, M.A. Baqi, A.T.M. Mahbub-E-Elahi and M.M. Amin

Summary

The study was undertaken to determine the phylogenic relationship of Bangladeshi isolates of H5N1-HPAI virus with those of other countries across the world.

A total of 47 AI virus samples were sent to FAO/OIE reference laboratory for Avian Influenza, the veterinary laboratories agency (VLA) We ybridge where nucleotide sequence of the HA gene segment of 25 isolates were de termined. In addition, VLA also provided the results of a phylogenic analysis showing relatedness among all the Bangladesh isolates.

It was clearly demonstrated that Bangladesh isolate clustered with those from middle China, South mid Asia, Middle-East, Europe and Africa belonging to the sub-clade 2.2 which is popularly known as Qinghai Lineage or Euro-Asia Africa lineage.

It may be noted that in the phylogenic tree, the closest neighbour of Bangladeshi isolates were these from Kuwait, Mongolia, Russia and Afganistan. Three Indian isolates from outbreaks of 2006 also fall in the same sub-clade but these do not appear to be the closest neighbours (phylogenic tree) of Bangladeshi isolates.

The origin of HPAI in Bangladesh still remains obscure and that the country (Bangladesh) does not have any poultry trade link with the countries mentioned above. The close similarities between the HPAI isolates of these distantly situated countries would suggest that migratory birds might be responsible for initial introduction of the HPAI in Bangladesh.

A regional collaborative molecular epidemiological study is warranted for future.

Introduction

Highly pathogenic avian influenza (HPAI), a devastating viral disease of birds, has affected all continents of the world. The frequency of outbreaks increased since the middle of twentieth century but usually remained confined within a country and were subsequently controlled or eradicated by stamping out and / or vaccination. The situation has changed drastically with the present panzootic of H5N1 HPAI that started in the Far East and South East Asia during December 2003 and spread across most parts of the world. The current H5N1 HPAI virus has been transmitted by poultry to humans and has a potential for triggering a human pandemic.

In Bangladesh, HPAI outbreaks started on 5 February 2007 and as of 20 June 2008 a total of 287 outbreaks have been reported. The outbreaks may broadly be identified into two waves; the first wave peaked in March 2007 and gradually declined with no cases reported in August 2007. The second wave of reported outbreaks started in September 2007, reached its peak in February and March 2008 and then started to decline; Only 12 outbreaks was reported in April and one outbreak in May as compared to 92 and 79 outbreaks in February and March 2008, respectively.

In Southern Asia, India, Pakistan and Myanmar experienced H5N1-HPAI outbreaks before Bangladesh. India experienced a first outbreak in the state of Maharastra during 2006, with subsequent spread to the states of Madhya Pradesh and Gujrat. A total of seven outbreaks were reported during that episode. In 2007, a single outbreak was recorded in Manipur State. Fresh waves of outbreaks started in West Bengal state on 4 January 2008 and until 01 May 2008, a total of 37 HPAI outbreaks from 14 out of 19 districts of West Bengal and 3 outbreaks from Tripura state have been reported to the World Organisation for Animal Health (OIE).

In Pakistan, a total of 12 H5N1 outbreaks have been reported between 23 February and 3July 2006. The disease re-emerged in Pakistan on 1 February 2007 and as of 31 January 2008, a total of 38 outbreaks have been reported to OIE.

Myanmar also experienced outbreaks in 2006 reporting a total of 78 outbreaks between 8 March and 27 April. In 2007, Myanmar reported 12 more outbreaks between 26 February and 28 July. Subsequently, 3 further outbreaks occurred in Myanmar between October and December 2007. Myanmar has now declared freedom from HPAI.

It is believed that the viruses of the present H5N1 strain originated in the province of Chinese Guangdong in 1996 and arrived in the Far East and South East Asia in 2003 (Sims et al. 2005; Wang et al., 2008). As the virus continues to circulate, genetic drift has mutated the virus into many clades and sub-clades. The HA (haemagglutinin) sequences of the majority of H5N1 circulating in avian species now cluster into 10 phylogenetic clades. A standard clade nomenclature system based upon the evolution of the HA has been jointly developed in 2007 by WHO, OIE and FAO (Annon. 2007). Phylogenetic analysis of gene sequences of HPAI viruses have been found to be very useful for studying evolution of the viruses in a region or across the globe (Ducatez et al., 2007; Salzberg et el., 2007). There is little information available on the origin and spread of HPAI in Southern Asia (Kamal et al., 2007; Islam et al., 2008).

The present study was undertaken to determine the phylogenetic relationship of Bangladeshi isolates of H5N1-HPAI viruses with those of other countries across the world.

Materials and Methods

Virus sequences

A total of 47 avian influenza virus samples were sent to the FAO/OIE Reference Laboratory for Avian Influenza, the Veterinary Laboratories Agency (VLA), Weybridge, U.K. where nucleotide sequence of the HA gene segment of 25 isolates was determined. These included 15 isolates from 2007 and 10 isolates from 2008. Out of these, the sequence data of 9 isolates from 2007 were made available to the Study Team. In addition, VLA also provided the results of a phylogenetic analysis showing relatedness among all the 25 Bangladeshi isolates.

In the beginning of HPAI outbreaks of Bangladesh 5 samples were also sent to the National Institute for Animal Health (NIAH), Bangkok, Tha iland from where the sequence data of all the 8 genome segments of one isolate was also provided.

Haemagglutinin gene sequence information of 32 H5N1 HPAI viruses belonging to different clades and sub-clades from 20 different countries were downloaded from the GenBank Database of the National Center for Biotechnological Information (NCBI), USA.

The sequences are listed in Table 1.

Table 1. Virus sequences used in phylogenetic analysis

Phylogenetic analysis

EditSeq and MegAlign modules of Lasergene DNAStar software were used for sequence editing and phylogenetic analysis, respectively. Multiple alignment was performed using Clustal V algorithm and neighbour joining method. A total of 35 sequences were subjected to phylogenetic analysis, which included the sequences of three isolates from Bangladesh and 32 isolates from other countries. A 1210 base pair gene segment from the 5′ end of the coding region of HA gene was used in alignment. The alignment generated a phylogenetic tree. The 1959 Scotish isolate served as outgroup in the phylogenetic tree. Different clades and sub-clades in the tree were identified according to the OIE-FAO-WHO suggested standard clade nomenclature system.

 Results and Discussion

Phylogenetic relationship among the 25 Bangladeshi isolates are shown in Fig. 1. Percent identity between Bangladeshi isolates is shown in Table 2. As in Fig. 1, it may be observed that all the isolates from Bangladesh are very close to each other. Table 2 demonstrates that Bangladeshi isolates from 2007 and 2008 are 99.1 to 100% identical indicating a single introduction of the virus in Bangladesh. (The Fig. 1 and Table 2 were provided by Dr. Ian Brown of VLA, Weybridge, U.K.).

Fig. 1. Phylogenetic relationship based on HA gene sequence among 25 Bangladeshi H5N1 HPAI isolates (provided by Dr. Ian Brown of VLA, Weybridge, U.K).

Table 2. Sequence identity / distances based on HI gene sequences among 25 H5N 1 HPAI isolates from Bangladesh.

Provided by Dr. Ian Brown, VLA, Weybridge, U.K.

Out of these 25 almost identical virus sequences of Bangladesh origin, three were subjected to a wider phylogenetic analysis along with 32 sequences of 20 different countries. The results are presented in Fig. 2. which clearly demonstrates that Bangladeshi isolates clustered with those from middle China, south and mid Asia, Middle -East, Europe and Africa belonging to the sub-clade 2.2, which is popularly known as Qinghai lineage or Euro-Asian-African lineage. On the other hand, all the South-East and Far-East Asian isolates belong to other clades or sub-clades.

Fig. 2. Phylogenetic relationship based on HA gene of Bangladeshi H5N1 HPAI isolates with those from other countries

It may be noted that in the phylogenetic tree the closest neighbours of Bangladeshi isolates were those from Kuwait, Mongolia, Russia and Afganistan. Three Indian isolates from outbreaks in 2006 also fall in the same sub-clade but these do not appear to be the closest neighbours of Bangladeshi isolates.

It is interesting to note that one Indian isolate (India/m777/2007) has clustered in Clade 1, which is rather unusual. Because, Clade 1 viruses are now quite rare even in South-East Asia. No detail is available about the origin of this isolate. The information provided in the GenBank indicates that it is an ‘escape mutant’, presumably a laboratory strain.

The origin of HPAI in Bangladesh still remains obscure. Although HPAI isolates from Bangladesh are closely similar to those from Kuwait, Mongolia, Russia and Afganistan, Bangladesh does not have any poultry trade link with these countries. Therefore, from the findings of the present study poultry trade cannot be blamed for introduction of HPAI in Bangladesh unless a missing common poultry trade link between these countries could be identified. The close similarities between the HPAI isolates of these distant countries would suggest that migratory birds might be responsible for initial introduction of HPAI in Bangladesh.

One limitation of the present study was that the sequence information of HPAI isolates of Pakistan and Myanmar is not available in the GenBank. Although India submitted sequence data of 2006 isolates in the GenBank but those from 2008 isolates are not yet available in the GenBank. It is very important to know the molecular epidemiology of HPAI viruses in the region, particularly if the same or similar viruses are circulating in the region or viruses of different genetic make up have already entered in the region. Moreover, it is also important to know if during the period 2006-08, the viruses have undergone any significant evolution, mutation or reassortment. These pieces of information are vital to establish a baseline for future virological surveillance in the region. This would also be of immense value should the countries in the region decide to go for vaccination in future. Therefore , a regional collaborative molecular epidemiological study should be undertaken.

Chapter 3:Epidemiological Study

Summary

A case-control study was undertaken to identify possible risk factors of the multiple outbreaks of highly pathogenic avian influenza registered in commercial poultry farms in Bangladesh during 2007-2008. Structured information on farm placement, profiles including input sources, area, size, type, dwellings, farm management, human traffic, movements and bio-security measures were collected from 50 case and 100 control chicken farms (2:1 ratio) by using a pre-tested questionnaire and personal interviews administered by 15 trained vet. graduates. A total of 13,340 pair wise correlation coefficient among the predictor variables were examined of which 19 variables showed moderate to high correlation (Spearman’s rho >

0.5 and Pearson’s Correlation Coefficient >0.5). Measures of association with being a HPAI case (0=control 1 =case) farm and categorical potential predictors were determined using Pearsons chi square test/F test and a screening selection pool for inclusion in bivariate and subsequent multivariate analysis was constructed using a cut off P<0.1.

For univariate and bivariate, each risk factor (RF) available for analysis, the Mantel– Haenszel odds ratio was calculated and checked for their possible conf ounding and effect modification role. The risk for influenza A (H5N1) infection was assessed by using adjusted odds ratios based on bivariate and multivariate logistic regression analysis. Finally multivariate models were developed and tested for goodness of fit using Pearson’s goodness of fit for logistic model (P<0.01-0.05).

The result of descriptive study indicate d layer farms to have mostly experienced (90%) with the outbreak. The average age of affected birds was 212 days. Location of farm being away from city (peri-urban/rural) (OR=2.52) , farms distance from live bird market (>1 km) (OR=1.6), farms having pullet as input source (OR=2.13), presence of backyard farms around (OR=1.68), farms having more than 4 staffs (OR=1.61), farms have their vets visit scheduled but infrequently (OR=1.61), farms have scheduled visit of feed seller/representatives ( OR=1.75), farms used hired vaccinator(OR=1.84), farms used both ready mix and own formulated feed (OR=3.69), farms allowed loading of egg/live bird vehicle inside gate (OR=2.74), farms practices irregular cleaning and disinfection practice before each restocking (OR=2.85), farms practices disposal of dead bird thrown outside/open places (OR=3.27) and farms practices disposal of litter drained in pond (OR=1.67) recorded to have higher probability of being a HPAI positive farms.

Although not statistically significant, sickness/mortality was reported in backyard poultry preceding the HPAI outbreak (68%). Therefore, intensive surveillance, early detection of HPAI infection in backyard poultry and an effective response could possibly reduce the risk of HPAI introduction to the neighboring commercial farms. Since layer type farms were assumed to be at higher risk for HPAI virus introduction, more specific control measures need to be applied to prevent future outbreaks. Human traffic, input output movement to and from farms and disposal practices apart from other bio-security steps need particular attention to reduce the spread of influenza related to live poultry markets, which are common place across Asia.

Introduction

Highly Pathogenic Avian Influenza (HPAI) has emerged as a big global concern for its socio-economic and public health impact due to devastating infection in poultry and concomitant human cases since 2003. The disease has been posing a serious threat to the poultry sector in many Asian countries including Bangladesh.

Epidemiology of the disease

Avian influenza (AI) viruses form a group of virus strains of diverse virulence under the family Orthomyxoviridae. HPAI viruses typically produce a severe, systemic disease with high mortality in chickens and other gallinaceous birds. However, these same viruses usually produce no clinical signs of infection or only mild disease in domestic ducks and wild birds (Spackman et al., 2002, Boender et al., 2007, Swayne, 2007).

The virus infects almost all avian species but some of the subtypes may infect mammals including humans. The disease is transmitted by direct contact between infected birds, contaminated fomites, transport and humans. The virus can survive longer in comparatively cold temperature and in organic materials. Ducks and other water fowl may carry the virus without showing any clinical sign.

The current hypothesis about virus introduction and transmission comes under the interest of epidemiological research. It has been postulated that virus is introduced into poultry by wild migrating birds (Hinshaw et al., 1979, 1980; Pomeroy, 1982; Alexander and Gough, 1986; Webster et al., 1992; Alexander, 1995; Alexander and Brown, 2000). HPAI virus may subsequently spread to other poultry farms via several routes, either via direct or indirect contact between poultry. In this respect, transport of live poultry, persons (Wells, 1963; Homme et al., 1970; Alexander and Spackman, 1979; Halvorson et al., 1980; Glass et al., 1981), and mechanical transfer of faeces of infected birds in which very high virus concentrations are present (Utterback, 1984) are considered to be the most important transmission routes. Consequently, these routes mainly consist of movements of people (e.g., farm owners and their staff), materials (e.g., egg trays) and vehicles (e.g., lorries that transport egg trays and eggs).

Other routes of transmission are via ‘contiguous spread’, a comprehensive term that includes transmission from infected farms over short distances by unknown or poorly understood introduction routes (Henzler et al., 2003).

Status of Poultry Sector in Bangladesh

The poultry production practices of Bangladesh can broadly be classified into family poultry or backyard poultry and commercial poultry. The number of chickens and ducks are 210 millions and 38 millions, respectively (Source: DLS). Family chickens are distributed almost equally across the rural Bangladesh and about 70% of the rural households rear backyard chickens (UNICEF, 2007). Backyard chickens are also raised in urban and sub­urban area but in a significantly low number.

The share of commercial and family poultry is almost equal though commercial system (layers and broilers) has emerged recently. Commercial poultry is distributed across the country but highly concentrated in the central part (Fig. 1 and 2). All the four poultry sectors classified by FAO are present in the country. Small scale commercial poultry flocks range between 300-1000, which account for majority of the production in the commercial poultry sector. These farms fall under sector -3 of FAO designated category. It is generally presumed that bigger operations follow higher bio-security standard but in many instances that is not true. This is revealed during an epidemiological investigation carried out after an outbreak in a big commercial farm (Kalam et al. 2007). There are about 100 parent stock breeder operations in the country. Some of them have environmentally controlled house and some operate in open-sided house. There are 7 broiler grand parent stock farms in the country. Duck population in Bangladesh is one of the highest in the world. Ducks are generally kept along with backyard chickens and in cases where they are kept separately they intermingle with chickens during forage. The concentration of duck population is higher in the areas of large water bodies. The highest concentrations of ducks are seen in north eastern region of the country (Fig. 3).

Likewise, pigeons and quails are reared by many villagers in the country. Cock fighting is not common in Bangladesh and fighting cocks are limited in small pockets of Bangladesh. Pet birds are seldom kept in the rural areas and in most cases they are caged and are not allowed to move outside freely.

Poultry market chain in Bangladesh

Backyard chickens and eggs are either consumed by owner or sold to local market or to beparies (egg or chicken collector or trader or middleman). Beparies collect eggs or chicken from households or from local chicken markets. Collected chicken and eggs are transported to the big city centers on the roof of buses. Some time chickens are transported to a distance over approximately 300 kilometers.

Commercia l poultry are grown across the country. Commercial poultry farmers depend on the parent breeder farms for supply of day old chicks. Day old chicks (DOC) are delivered to commercial farmers by the breeders through dealers. Breeder companies supply to dealers using their own transport or by hired transport. The farmers either collect chicks from the dealers’ shop or the dealers supply chicks to farms.

The parent stock poultry farms get their Parent Stock DOCs either from Grand Parent farms of the country or import from AI-free countries. Import of day old commercial chicks and hatching eggs are not allowed in the country but non formal trade of day old poultry can not be ruled out.

Farmers use manufactured finished feed from millers or prepare feed from purchased ingredients. The feed ingredients are either imported from outside or collected from rural markets. Vaccines and drugs are either collected from local chemists or pharmaceutical company representatives who supply at farm gate.

Migratory birds, wild birds and pet birds

According to Government sources, about 244 species of migratory birds visit Bangladesh each year or during a migratory season. Moreover, there are many species of local wild birds including aquatic species in the country. Stray dogs and cats are seen almost every where.

Animal health service and surveillance

The official veterinary service is extended up to upazila. The private veterinary service is yet to develop. However, with the growing commercial farming practices the demand of veterinary service has been under transformation and recently farm based consultancy along with scheduled visit of the farm by both vet and other representatives are not uncommon. In absence of diagnostic facilities in place mostly post-mortem and symptomatic curative and preventive practices are common.

Epidemiological investigation for any disease outbreak is rarely attempted as it is constrained by the absence of trained resources and of an effective epidemiological unit in DLS. Recently, FAO has helped establish an epidemiology unit at DLS headquarters and has provided primary training to the unit members. The field offices rely on farmers’ report and have limited capacity to carry out active surveillance. Recently, FAO has fielded 150 animal health workers to conduct active surveillance in the country.

Overview of the disease (HPAI) in the world

Between 2003 and early 2008, a total of 61 countries/territories reported the occurrence of highly pathogenic avian influenza (HPAI) virus subtype H5N1. In late 2003 and in 2004 the disease was restricted to South-East Asia and in 2005 spread to Central Asia, Russia and Eastern Europe. In 2006, it reached the African continent and the Middle East for the first time and spread to Western Europe, where mainly wild birds were infected. In 2008 (up to 31 March), eleven countries/territories have notified the reoccurrence of HPAI (H5N1) following its previous eradication, thus indicating that the virus is continuing to circulate. In 2007, 29 countries (4 in Africa, 14 in Asia, 3 in the Middle East and 8 in Europe) reported HPAI subtype H5N1 in poultry, in wild birds, or in both poultry and wild birds. In six of these countries, Bangladesh, Benin, Ghana, Kuwait and Saudi Arabia, there were the first ever reports of the disease.

In 2007, the number of countries that reported outbreaks was smaller (29 countries) than in 2006, when 47 countries were affected. Starting in 2007, the trend has been for a decline in both the number of affected countries and the number of notified outbreaks.

Overview of the disease scenario in Bangladesh

Bangladesh was free from Highly Pathogenic Avian Influenza till March, 2007. The presence of the disease was declared on 22nd March of 2007 after detection of the disease in Biman Poultr y complex. However, according to the report submitted to OIE the index case was Naz Poultry Farm of Jamalpur district where outbreak started back in 05 February, 2007 (OIE Report, 2007). Since then 287 outbreaks have been detected in 47 districts of Bangladesh (Fig. 4). First wave of outbreak was observed during the first half of 2007. The highest number of outbreaks was seen in March 2007 during the first wave. The disease started to decline after March and not a single case was reported during the month of August, 2007.

The disease reappeared in Bogra in the month of September, 2007. Number of cases was gradually climbing up and reached its peak in Feburary, 2008 with 93 cases. A

regression started again from the month of March, 2008 and only a single case was detected in May, 2008, which was on 18 May 2008 in Tangail.

Human health situation

Bangladesh had a single reported case of H5N1 till 30 June, 2008. Government of Bangladesh declared a human case from findings of sample collected back in January, 2008. However, the boy who got infection recovered (WHO, 2008). Till 28 May, globally there have been 383 confirmed human cases with 241 fatalities.

Control measures and options

The Government of Bangladesh has been following stamping out policy without vaccination to control the disease. Until recently, Government used to depopulate all birds in 1 km region, impose movement restriction of poultry and poultry products and disinfect the premise. Despite the measures undertaken, the disease continued to spread. Now Government has shifted to focal culling that involves culling of infected farm and its contacts when a commercial farm is infected and in case of infection in rural poultry, culling around an area of 500 meter is recommended.

Rationale of undertaking this research:

Exploring the putative risk factors (RF) for spread of HPAI into commercial poultry farms is critical for effective intervention. The identification of important risk factors, and understanding their channel of operation might prove very useful for the improvement of existing control measures.

Recently Ministry of Fisheries and Livestock has emphasized on the need to identify the major risk factors of HPAI outbreak in Bangladesh to formulate necessary actions for effective control of spread, prevention of future outbreak and minimizing the public health risk of the disease. Government directed a group of scientists headed by the Director General, BLRI to conduct an epidemiological investigation to identify the key causes of trans mission of the disease. The present study was, therefore, undertaken with the following goal and objectives.

Objective:

The specific objectives of the study were to:

1. 1.         To develop a  profile and portray of risk population/farm through descriptive epidemiology
2. 2.         To determine potential farm level risk factors of HPAI outbreak in commercial poultry in Bangladesh
3. 3.         To develop idea on possible causes of spread of HPAI in commercial poultry in Bangladesh

Overall Aim/Goal:

.                      • To enhance the national policy by formulating an applicable strategy to minimize the risk of introduction and spread of HPAI outbreaks in commercial poultry farm.

.                      • To identify risk areas for targeted disease surveillance.

.                      • To contain the disease in poultry and prevent large economical losses in the commercial poultry industry.

.                      • To minimize the threat of human infection with HPAI.

Fig. 1. Distribution of layer chickens in Bangladesh Fig.2. Distribution of broiler chickens in Bangladesh

Fig. 3. Distribution of ducks in Bangladesh

Fig. 4. Distribution of HPAI outbreaks in Bangladesh. (Points are not based on exact coordinates, but indicate number of outbreaks in the respective upazila)

Methodology Population and Sampling

Study Population

Reference/Target Population: All commercial poultry farm in Bangladesh

Source Population:

Govt. registered c ommercial poultry farms under each district and upazilla in Bangladesh brought under the surveillance database of the Department of Livestock Services, Bangladesh

Sampling Frame

List of commercial poultry farms of each upazilla in Bangladesh brought unde r the disease reporting surveillance database of the Department of Livestock Services (DLS)

Study Design: Case Control Study Case: Control Ratio: 1: 2

Case Definition

Case Farm

Farms within a case of upazila (registered at least one outbreak/positive samples in one farms) registered at least one sample diagnosed positive (H5) at National Reference Laboratory for Avian Influenza (BLRI).

Control Farm

Farms (neighbourhood) within a case of upazila have not experienced with a single clinical case (till the date of interview) or samples diagnosed negative to H5 at NRL-AI (BLRI) and located outside the 1 km radius of the case farm.

All farms within a control of upazila (no confirmed HPAI outbreak) were considered as control farm for the purpose of this study.

Selection of Case and Control farm:

The DLS disease reporting database was used here for selection of case farms. Total 287 case farms were available for the selection of case farms. The selection of both case and control farm was randomized as per following strategy

Sampling Strategy

Selection of Case farm

Total 47 districts registered with at least one outbreak of HPAI in the database. Initially 20 districts were selected from the list on the basis of density of outbreaks (Fig. 5). Case farm were then selected from a customized sampling frame (pooled with matched criteria) following simple random sampling within the selected districts (n=20) with a ratio of 1 case farm in every 5 outbreaks under same districts until the total of 50 case farm being chosen. Commercial farms were deliberately preferred and selected as having more organized records and backyard case farm were not considered to include in this study due to anticipated difficulties in tracing back, unwillingness to support and likely not to have sufficient record back up.

Selection of Control farm

For each case farm 2 control farms were selected and recruited in the study. Farms close to the border of the (1 km radius) case farm even might have experienced culling/depopulation have included. Each unaffected/negative to HPAI farm within the defined zone adjacent to case farm was assigned with a unique identification number, and 100 farms were selected meeting criteria by using numbers generated with a random number generator in Survey Toolbox (Cameron, 1988). In case of absence of control farm within set radius the closest neighbouring farm even in adjacent upazila meeting criteria was selected to meet the pool of control farm for this study.

Data Collection

A preset questionnaire was used to collect data. A structured questionnaire was developed with the contribution of expert national and international epidemiologists. The questionnaire contained mostly closed (mixed) and few open-ended questions. Of the most questions, majority meant for multiple value answers and some offered binary or single answers.

The intended questions gathered to collect farm level detailed information on exposure or potential variants to indicate and establish differentiable differences between case and control poultry farms. The questionnaire was piloted before main administration in selected farms in 2 upazilas to modify and refine the questions and expected responses, if any.

A total of 15 veterinarians (fresh graduates) were employed for data collection. Two days training programme was arranged to develop and evaluate the efficiency, techniques and performance of directed interviewing for the employed veterinarians. The questionnaire set contained 116 variables (40 Questions). Data on geographic location, farm characteristics, stock information, flock health history, farm bio -security, farm management, and marketing practices were collected by the above mentioned trained interviewers during farm visits. Additional information such as farm area, number of sheds, and incoming day-old chick numbers were obtained from observation, farm records and requested exploration.

Risk factor (RF) or Predictors/Exposures Definition

Based on the epidemiological nature as described in literature elsewhere and also hypothesizing some major putative factors that might influence the risk of  HPAI in countries including Bangladesh the following variants have been considered as the potential customer to collect details information on (Standardized questionnaire- Appendix).

In the need of the current study it has been hypothesised that introduction of HPAI into a flock could be facilitated through or caused by certain factors, in order of following relative importance:

1) Poultry transports 2) Neighbourhood infection : This route of virus spread is generally used. This route is defined as transmission over a short distance via various, but unknown routes.

3)   Personnel who, as a profession, deal with poultry in particular and as such arelikely to enter the houses (e.g. farm staff, vet/representatives and egg/live bird tradwes/transporters)

4) Biological Career/Transfer : Infected wild birds/other birds/carrier.

5)   Mechanical transfer of faeces of infected commercial poultry by wild birds or vermin. It was presumed that wild birds or vermin c ould transfer the virus to commercial poultry via fomites, voiding or into the courtyard. Consequently, farmers/visitors could transfer the virus via their boots into ground floor systems.

With the requirement to collect information and identify the hypothesized possible risk factor/routes the following specific group or areas of questions were intended and recruited in the questionnaire:

A. Farm Location and Surroundings

�.a. Farm Location (GPS)

�.b. Farm distance to live bird market

�.c. Density of farms around

�.d. Presence of layer as neighbor farm

�.e. Presence of backyard neighbor farm

B. Farm Area, Size and Dwellings

�.a. Total farm area

�.b. Farm size (total birds/farm)

�.c. Dwellings type in farm premises (owners/staff/other animals/mixed farm)*

* to also be considered for human/other animal traffic/carrier

C. Farm Management

�.a. Housing Type

�.b. Input type and sources

�.c. Feed type and sources

 D. Human Traffic/Carrier/Source of transmission

�.a. Number of total staff

�.b. Type of employment

�.c. Frequency and type of Vet visit at farm

�.d. Frequency of visit of feed seller/input supplier/other representatives in farm

�.e. Type of vaccinator

E. Movement of Input and Output to and from Farm (including mechanical transfer)

�.a. Types of movement of output (egg/live bird) off the farm

�.b. Loading place of vehicle to transport farm output

�.c. Place of sale of farm output (egg/live bird)

�.d. Types of vehicle/transport means to and from farme) Types of egg tray used

F. Bio-security

�.a. Vegetation (access of wild bird/animals)

�.b. Use of foot bath at entry

�.c. Use of footwear at entry

�.d. Change of cloth at entry

�.e. Cleaning and disinfecting practice of house/floor before restocking

�.f. Disposal practice of dead bird and litter

Some descriptive questions only meant to be collected from the case farms have also been assigned. Questions like age of affected birds, type of farm affected, clinical signs, whether any new change in management (new introduction of bird, new staff, new feed), any report of outbreak in neighboring farms, list of prevalent diseases and vaccination practices information were intended to hypothesize profile of a case farm and their likely spreading route.

Data Entry, Management and Analysis:

The each coded responses from the questionnaire were entered into a spreadsheet database (Microsoft Excel XP (R) ) and then transferred into a statistical package for analysis (STATA ® 9.2 College Station Texas, USA and SPSS for Windows version 11.0, SPSS Inc., Chicago, IL, USA). Data were duly sorted and checked for inconsistencies and missing value and or duplications. Descriptive statistics was used to calculate distributions of all variables by case and control status.

Correlation between all the variables was examined. The Spearman Rank Correlation was used in preference to the Pearson’s Correlation Coefficient because of the ordinal data type used for the binary questionnaire answers (Hollander 1973). A Spearman’s rho of 0.50 to 0.75 indicated moderate to high correlation and a Spearman’s rho greater than 0.75 indicated very high correlation. For the normally distributed data; “Total farm area”, “Number and species of birds” and “Number of staff” the Pearson’s Correlation Coefficient was used to determine correlation. A Pearson’s Correlation Coefficient of 0.50 to 0.75 indicated moderate to high correlation and a Pearson’s Correlation Coefficient greater than

0.75 indicated very high correlation.

Univariate analyses was used to test for associations between disease status and each explanatory variable The Pearson chi-square (?2) statistic was used for categorical variables (Hope, 1968; Patefield, 1981). For the continuous data variables, “Total farm area”, “Number and species of birds” and “Number of staff”, means were compared using the Student t-test (O’Mahony, 1986). Where appropriate, responses were categorized before analysis, with categories selected on the basis of the distribution of responses for that variable.

Only variables with a significant unconditional association with the dependent variables were retained for the subsequent bi-variate and multivariate statistical model. A level of significance of 0.10 was chosen as the cut-off. For the data subsets, after removing non-significant association variables, a binomial logistic regression model was built that determined the most simple (limiting number of variables) model which adequately described the data (best fitting). Multivariate logistic regression analysis was then used to assess associations between independent variables and the outcome of interest (case or control status), while controlling for other possible risk factors. The potential risk factors were first evaluated by calculating the crude odds ratio (OR) and 95%confidence intervals (CI). Secondly, the OR was adjusted for confounding by other suspected variables, using the Mantel–Haenszel procedure to produce an adjusted OR (Mantel and Haenszel, 1959). For adjusted OR, farms were grouped into strata of expected confounders. The stratum-specific estimates were considered similar enough for pooling if the Breslow–Day statistic was not significant (p > 0.05) (Breslow and Day, 1980). An association was considered statistically significant if the probability that it arose from chance was <0.05. Final model was constructed by using both forward and backward stepwise procedures. “Best subsets” were adapted with approach, by which variables of particular interest were forced into the initial equation and the influence of key variables was tested by using the fit of various possible equations. Model fit were assessed by using the Hosmer-Lemeshow goodness-of-fit test and the Pearson Chi square fit for logistic model (Collett, 1991; Hosmer, 2000). An overall measure of discriminatory accuracy for the model was estimated by calculating the area under a non-parametric receiver operating characteristic (ROC) curve, which is a plot of sensitivity versus (1- specificity) over the range of all possible cut-off values (Greiner, 2000; Eng, 2006).

Result and Discussion

A total of 150 case (n=50) and control farms (n=100) were recruited in this study which altogether was selected from 20 districts [having relatively higher cluster (.=5 outbreak) of outbreak] and from 36 upazilla. The above two bar char t showing the selected Upazilla and districts from where the case and control farms were sampled with following a case control ratio of 1:2.

Fig. 6. Number of case and control farms selected from different districts

The descriptive study (carried out only in case or HPAI positive farms) explored that most of farms experienced the outbreak were of layer type (90%) (Fig. 7). Layers/layer finisher type has also been recorded to be significantly associated with the presence of HPAI virus elsewhere (Thomas et al., 2005). It has been explained that the reason why layers become more susceptible to HPAI introduction than broilers lies with the nature of management practices which allows different potential factors like input supply, movement of output off the farms, human traffic or biological carrier and poor practices of bio-security to get this nature of farming exposed for relatively longer period of time than a short cycle business like broiler (Alexander, 1995). More precise information on affected layer farms was unavailable, because all the farms were depopulated, and farmers were not revisited.

Descriptive study also recorded that the mortality in backyard birds preceded about 68% of the neighbouring layer farms to experience with the HPAI outbreak (Fig. 8). On the other hand, it was also reported that the new birds (backyard) were introduced within 21 days before the outbreaks (18%; Fig. 9) in or around HPAI affected farms. However, in the absence of a robust surveillance system in place and without having a trace back done it seems difficult to connect these with the reason of introduction in a more plausible way.

Fig. 8. Mortality in backyard poultry before the outbreak (N=47) Fig.9. Introduction of new bird/batch in or around HPAI affected farms 21 days before outbreak

The age distribution of HPAI affected flocks has been presented in Fig. 11. The mean age of the affected flock at the time of outbreak recorded to be 212 – 114.49 (7 months) .This descriptive finding albeit limits would indicate that possibly day old chicks (private hatcheries) as input source may not be the major source of introduction, rather other

mechanical transmission through human/biological contact, poultry transport and bio-security

Figure11:Other disease(s) occurredon your farm in the year before the HPAI outbreak? (N=47)

Newcastle disease (NCD) and Gumboro disease recorded to be prevalent as the existing disease in the HPAI affected farms (Fig. 11). The frequencies of concurrent diseases in HPAI affected farms, however, reflected endemic nature and possibly do not have any role of exposing the farms to HPAI outbreaks.

Correlation analysis and PCA

A total of 13,340 pair-wise correlation coefficients among the predictor variables were examined. There was moderate to high correlation between 19 variable pairs (Spearman’s rho > 0.5 and Pearson’s Correlation Coefficient > 0.5). The highest correlation

(0.65) was found between HPAI incursion and practice of using vehicle for feed and egg transportation (Table 1). All 19 results with a Spearman’s rho > 0.5 or Pearson’s Correlation Coefficient > 0.5 (medium to high correlation) were significant (P<0.01).

Measures of Association

Distribution of potential on-farm, environmental and other putative risk factors in both case and control farms are shown in Table 1. The risk factors were grouped and categorized to examine their stratum specific association with the HPAI. Uni-variate measures of association with HPAI and different categorical exposure variables were calculated using Pearsons Chi square test/F test where appropriate (Table 1). Farms being away from the city (P<0.06), loading of vehicle inside the farm to move off the farm output (egg/live bird) (P<0.001), use of foot bath and change of footwear at entry to farm more frequently (P<0.001, P<0.01), cleaning and disinfecting practices before each restocking (P<0.02) and disposal practices of dead bird (P<0.06) significantly associated with the HPAI outbreak.

The likely association above explained that the possibility of spread could be attributed with the level of bio -security practices which might have exposed the human/biological contact/carrier and movement of farm output/input to and from farm to invite and spread the HPAI between farms having shared facilities and closer in vicinity in terms of location and distance to live market.

Table 1: Distribution of farm and environmental risk factors of HPAI grouped into different categories (farm location, characteristics,

management, movement and bio-security practices :

The results of the simple unconditional association were shown in the rows in bold, indicate the variables with significant association and odd ratios. The following significant risk and protective factors could be identified (for odd ratios and 95% confidence intervals)

.                      � Farms being away from the city (peri-urban/rural) were with 2.5 times risk of being a HPAI positive farm than farms within city areas

.                      •     Farms >1 km distant from live bird market were at increased risk (OR-1.61) of HPAI outbreak

.                      � Having backyard farm around (1 km radius) increased the likelihood of being a HPAI farm 1.68 times

.                      � The probability of farms to become a HPAI positive increased 2.13 times if the input source was pullet in comparison to farm whose input source was DOC

.                      •     Total number of staff >4 increased of risk of HPAI  1.61 times

.                      •  Short term employment (compared to long term regular staff) increased the
chance of an HPAI outbreak on a poultry farm.

.                      •     Frequency and practice of vet visit scheduled but infrequent increased the chance of HPAI 1.61 times

.                      •  No visits of company representatives of poultry and feed suppliers increased the chance of HPAI.

.                      •  Farms used hired/outside staff for vaccinating their bird had 1.84 times higher risk of HPAI than farms used their own staff for vaccinating their bird.

.                      •  Farms offered ready feed (OR=2.46) and a combination of ready (OR=3.69)and own formulated feed had increased risk of HPAI than farms offered only their own formulated feed.

.                      •  Farms used shared/company vehicle to trade/move of f output had higher risk (OR=1.6)

.                      Loading vehicles outside the farm gate reduced the risk of HPAI and loading
vehicles inside the farm gate increased the risk of HPAI (OR=2.74).

.                       Selling eggs at distant markets (and not at local markets) reduced the risk of HPAI  on the farm

.                       Sharing egg trays with other poultry farmers increased the risk of introducing HPAI on the farm.

.                       Density of farm >10 increases the risk of the HPAI.

.                     The consistent use of foot baths seemed to increase the risk of HPAI on a farm,
while irregular use of foot-baths reduced the risk of HPAI.

.                      Changing shoes also seem to increase the risk of HPAI, while not changing
footwear seemed protective.

.                       Cleaning and disinfecting between restocking reduced the chance of HPAI

.                      Disposal practice of dead birds (thrown outside/open places) increased the risk of
HPAI 3 times than disposing dead birds in fixed pit

.                      Disposal practice of litter (drained to pond) increased the risk of HPAI (OR=1.67)

Table -2 : Univariate Analysis : Identification of potential risk factors for HPAI outbreak (Measures of strength of association

hpai | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval]

One thing becomes clear with this study, and was previous ly identified in other countries; biosecurity and the movement of people, animals and vehicles are crucial factors in the spread and control of HPAI in Bangladesh. Simultaneously backyard poultry also plays an important role for spreading of disease.

For future outbreaks of HPAI, the findings of this study suggest that the contacts between layer type farms, live markets, human traffic, and restricted vehicle entrance to load farm output should be limited as much as practicable, and might be a first step in a more sophisticated approach to eradication of the AI virus from the poultry population. In practice, if the current depopulation policy is upheld, either prioritisation of high-risk farms and their contacts at depopulation in a certain radius around infected farms or prohibition of the collection of eggs might decrease mechanical transmission.

1.0

Sensitivity

Limitations of the Study

As like the inherent this case control studies also have limitation of selection bias, recall bias (interviewer bias) and misclassification (exposure).

From the analysis it is found that some variants are insignificant due to confounding. Most the case farms are not existing or restocking. That is why it was difficult to enquire the plausibility of information.

Confounding, recall bias and poor interview techniques could possibly account for apparent inconsistencies and contradictory outcomes. Footbaths are described in the literature as highly effective to prevent poultry diseases on a farm. Leading questions or social embarrassment may cause incorrect answering of questions.

Since the study is carried out even after 112 mean days following outbreak (range 4­478 days), the recall bias and selection bias is very likely.

Due to time constraints, the study design became limited. The backyard poultry was not included in this study which played an important role to spread the disease. The duck was not incorporated in this study which would be responsible to harbor the disease.

The farmers sometimes gave in correct information. But it was very difficult to proof the evidence due to absence of the bird (depopulation) in the case farms.

Conclusion

The findings indicate that­

.                      •          Introduction of birds from market causes spreading of disease into the backyard level which extends to the farm level later on.

.                      •          Majority of risk factors of HPAI outbreak in commercial poultry farm are mostly related to market chain.

.                      •          Inadequate bio-security recorded is also an important factor to spread the disease.

.                      •          Movement of human, vehicles, farm equipments evidenced to be responsible to spread HPAI between farms.

Recommendations

.                      •          Improvement of the surveillance in backyard poultry and wet market by following the OIE guideline.

.                      •          Improvement in bio-security and movement control.

.                      •          Policy regulation and enforcement in trade and live bird market.

.                      •          Further prospective study is needed to follow up and confirm finding of this study and to formulate effective control policy for HPAI in Bangladesh.

Appendices

Standardized questionnaire:

Category

Question

Farm location Area and distance Distance to main or inter-district road Distance to live bird market Number of poultry farms in a 1 km radius Neighbouring farms

Farm area size and dwellings

Total farm area Dwellings Farm type

Grandparent stock Parent flock Layers Broilers Cockerels

Management Housing Ventilation

Movements

Replacement stock Source of pullets and DoCs Number of staff Staff employment

Veterinary visits Visits of chick/feed supplier representative Vaccinators Feed transportation

Feed composition Movement off the farm of eggs or birds Loading of vehicle Sales of eggs and birds Egg trays Cleaning and disinf ection of egg trays

Biosecurity Fence against birds and animals Vegetation density around sheds and around farm Number of entrances Foot baths

Changing foot wear on entry Change of clothes Cleaning and disinfecting before restocking Name of disinfectant Disposal of dead birds Litter Do farm staff have poultry of their own

Are staff involved in poultry trading, vaccinating or farming Drinking

Chapter 4 Socio-economic Study

Impact of Outbreak of Avian Influenza in 2008 on Production,
Marketing and Consumption of Poultry Meat and Eggs in Bangladesh

Jahanghir Alam, S. M . Ataur Rahman, M. Aktaruzzaman and  Salahuddin Polash

Summary

The poultry industry of Bangladesh faced a terrific financial loss in 2008 which was estimated at Taka 3858.31 crore. The prices of broiler declined by about 28 per cent while the price of eggs decreased by 26.5 percent. Many farm owners were compelled to abandon poultry raising as they lost their capital. The affected farmers lost their business and many of them were reluctant to go back to the production system. This study was undertaken to assess the impact of AI on poultry sector of Bangladesh. In this regard, 68 affected and non-affected poultry farms, 110 market intermediaries and 550 consumers were surveyed during the period from January to May 2008.

It is evident that the average loss per farm was Taka 4, 31, 543 for broiler and Taka 59, 28,971 for la yer farms. The total loss of 547 culled farms was estimated at about Taka 258.31 crore (2583.10 millions), and the total national loss (direct and indirect) was Taka 3858.31 crore (38583.10 millions). There was a high level of backward and forward linkages in the commercial poultry farms, which were also affected.

The price of poultry meat and eggs began to fall after the outbreak in comparison with pre-avian flu period and it was lowest during January-February 2008. The price began to increase since March 2008 and became normal in May 2008. It was found that the sales volume of eggs decreased, but at a lower rate than that of the broiler. People were more afraid of processing and cooking poultry meat than eggs.

The net marketing margins of different traders decreased due to a decline in the volume of business. However, the net marketing margin of retailers was higher after the occurrence of avian influenza than that of other intermediaries. The cross-price elasticity indicated that the consumers shifted from poultry to either beef or fish resulting in social welfare loss by the consumers.

Majority (about 70 percent) of the consumers received information on avian flu from TV followed by their neighbors (about 63 per cent). About forty per cent of the respondents felt that avian flu is a contagious and dangerous disease. Thirty six per cent of consumption of poultry meat reduced during avian flu and the consumers opined that it is harmful for health. After the tension of severe outbreak of avian influenza was over, all consumers increased their consumption of poultry products but that was less than the pre avian influenza situation.

Problems relating to avian influenza were identified and a set of recommendations are suggested.

Introduction

Background of the Study

The situation of avian influenza is rapidly evolving in Asia, including countries in the South and South-East Asia region. There is growing concern regarding the potential and the imminent threat of an influenza pandemic which could have the most devastating consequences. Since January 2004, events affecting both human and animal health have brought the world closer to an influenza pandemic than at any time since 1959 (Table 1.1). Experts estimate that a modest pandemic lasting over one year might cause a loss as high as 3% of Asian GDP and 0.5% of world GDP. This is presently equivalent to about $ 150-200 billion in GDP. Therefore, in addition to high morbidity and mortality, the next pandemic may cause massive social, political and economic disruption. The global commitment to prevent an influenza pandemic is reflected in the World Health Assembly Resolution, endorsed in 2005, that articulates the need for global action and strengthening of national capacity to respond to the threat of a pandemic. In 2005, the WHO Regional Committee for South-East Asia endorsed the Asia Pacific Strategy on Emerging Diseases and reiterated the call of the World Health Assembly. There is also a growing realization that a pandemic can be be tter fought collectively. Regional organizations such as SAARC and ASEAN, established with the central objective of economic cooperation between countries, are now being utilized to extend collaboration in public health as well. Recognizing that the formulation of plans should be a country-led process, WHO stands ready to assist Member States in the preparation or finalization of national plans either through country missions from the Regional Office or support through WHO country offices. These plans will provide not only a policy and strategy framework for a multisectoral response to avian influenza at the country level but also serve as an instrument for resource mobilization. The multi-bilateral agencies such as the World Bank, USAID and the European Commission are also willing to support in this area.

The poultry sector of Bangladesh was not infected by avian influenza until 2006, but the people of Bangladesh were in suspicion during the rumors of avian influenza in Southeast Asia in 2004. The death of 20 people at that time by unknown disease in different places of Bangladesh made this suspicion stronger. Later it was proved that the unknown disease did not occur for avian influenza but it occurred for the virus of NIPAH and HENDRA (Ahmed, 2004). However, the poultry industry of Bangladesh faced a terrific financial loss for the rumors of avian influenza. The price of broiler decreased by about 50 per cent in the wholesale and retail market from Taka 56-60 to Taka 25-30 (Ahmed, 2004). The price of day old chicks decreased from Taka 20-25 to Taka 5-10. The poultry farm owners of Bangladesh, especially broiler farm owners faced a huge amount of financial loss. Many farm owners were compelled to stop poultry raising as they lost their capital.

Table 1.1 lists the primary Highly Pathogenic Avian Influenza (HPAI) outbreaks in poultry since 1959. From the Table, it can be seen that half of the 20 outbreaks occurred between 1959 and 1991 but the ten most recent ones have all occurred in the last decade.

Source: Lister (2004).

** China, South Korea, Japan, Laos, Vietnam, Taiwan, Cambodia, Thailand, Indonesia and Pakistan.

At the end, the avian influenza occurred severely in Bangladesh in March 2007 and again in January 2008. Till July 04, 2007, a total of 2, 38,781 chickens were culled (MoFL Web page) from 73 poultry farms located in 16 districts in Bangladesh (Table 1.2). There was a great fear of avian influenza to the farm owners. Many farm owners lost their farm business and government made a moderate effort to provide them with incentives. In January 2008, the second wave of avian influenza was observed with intensity in Bangladesh. By 23 June 2008, a total of 47 districts and 187 farms were infected and 16, 37,266 chickens were culled.  The

incidence of avian influenza in Bangladesh by division is shown in Table 1.3. The highest incidence of avian influenza was observed in Dhaka division followed by Rajshahi division.

Source: MoFL Web Page (www.mofl..gov.bd/)
Table 1.3: Division wise incidence of avian influenza during 22-03-2007 to 23-06-2008 in Bangladesh

Source: DLS

The incidence of avian influenza by month is shown in Table 1.4. It can be observed that the incidence of avian influenza was high mainly in March-June 2007. The incidence of the disease repeated in December 2007 and continued with severity until March 2008. The highest incidence was found during January-March 2008, which collapsed the poultry sector of the country. The poultry association of Bangladesh estimates that the economic loss due to avian flu was Taka 4165 crore (41650 millions), which is about 5 percent of the national budget in 2007-08. The losses were incurred from the cost of culling birds, wage loss, productivity loss etc., although their calculation might have some margin of upward bias. Moreover, the market intermediaries were disappointed from their business; as a result, thousands of people become jobless. In addition, the consumer welfare losses were observed because they decreased consumption of poultry meat and eggs and shifted to beef and fish consumption by spending more amount of their limited income.

Preparation of Survey Schedule

Keeping in view the objectives of the study, three different types of interview schedules were prepared to collect the expected information from broiler farm owners, intermediaries and consumers. Before preparing the final interview schedule, draft schedules were pre-tested to verify the relevance of the questions and the nature of response from farm owners, intermediaries and consumers. After pre-testing, necessary corrections , modifications and adjustments were made in the survey schedule.

Map 2.1: Map of Bangladesh showing the study areas 2.4 Method of Data Collection

The primary data were collected by direct interviews with selected respondents. The data were collected during the period from February 2008 to June 2008 by the appointed research assistants of this project. The researchers monitored the data collection through several personal visits for ensuring data quality. All the data were coded, tabulated, summarized and processed for analysis after collection from the field.

Focus Group Discussion

After generation of information from the field, they were verified. Some of the results were presented in field workshops. The errors and inconsistencies were corrected through focus group discussion (FGD). A total of 33 sessions were conducted taking 3 in each area of the consumers. Moreover, inconsistencies regarding farm owners and intermediaries were also discussed in separate FGD sessions. Thus, a total of 55 FGD sessions were carried out.

In addition to generation of field data, secondary data were collected from the Directorate of Livestock Services (DLS) under the Ministry of Fisheries and Livestock (MoFL), FAO bulletins and other unpublished reports. Also desk reviews were done at BLRI, BAU and CDIL offices and libraries.

 Data Management and Analysis

Profit function

The data were analyzed on the basis of the objectives of the study. In the present study, mostly tabular and descriptive techniques were used for analysis of data. Tabular techniques were applied with the help of average, percentage etc. to find out the meaningful results. In this study, costs and returns analysis were done on both variable and fixed cost basis. The following profit function was used to assess the profitability of poultry farming:

µ = PbQb + PLQL -? (Pxi.Xi) – TFC Where,

µ = Profit or loss per poultry farms per batch of 1000 birds

Pb = per unit price of live broiler (Tk/kg)

PL = per unit price of used litter and excreta (Tk/sack)

Qb = quantity of live broiler (Tk/kg)

QL = quantity of waste litter (sack/1000 birds)

Pxi = per unit price of ith (variable) inputs

Xi = Number/quantity of ith input

TFC = Total fixed costs.

Cross-price elasticity

In economics, the cross elasticity of demand and cross price elasticity of demand measures the responsiveness of the quantity demand of a good to a change in the price of another good. It is measured as the percentage change in quantity demanded for the first good that occurs in response to a percentage change in price of the second good.

The formula used to calculate the coefficient of cross elasticity of demand is

D D

A

_{Q Q}A _PB

Cross elasticity (E _{A, B}) =

B

P

–

D

=

D

A

Q

B

QP

B

P

A

.

–

Where

DQA= the change in quantity demanded of product A

DP_B = the change in price of product B.

Generally, cross elasticity is positive whe n the products are substitute goods and

negative when the products are complementary goods. Cross elasticity became one when products are perfectly substituted and became zero when there is no substitute product.

Regression technique was used to determine cross elasticities of demand. In this case Double Log Linear Regression Model has been employed using the following consumption

function.

LnY = Lna + b₁LnX₁+b₂LnX ₂+b₃LnX₃+b₄LnX₄

Where,

           Y = Monthly per capita quantity consumption (kg) of broiler
X₁ = Monthly price of big fish (Tk/Kg)
X₂ = Monthly price of small fish (Tk/Kg)
X₃ = Monthly price of beef (Tk/Kg)
X₄ = Monthly price of goat meat (Tk/Kg)
a = Intercept
b₁, b ₂, b₃, b₄ = Coefficient

However, these measures of elastic ity must be used with caution because of the statistical problems involved in finding them and because of the same product elasticity tend

to vary from place to place and also from short run to the long run.

3. Economic Losses Due to Infection of AI in 2007 and 2008

In this study, 38 avian influenza affected and 30 non-affected farms were surveyed of which 33 were broiler and 35 were layer farms (Table 3.1). The layer farms were largely affected while the broiler farms were of short duration and could not be affected at large. On average, the total chicks affected both in broiler and layer farms were 2058 and 2894, respectively and the number of adult birds per farm was 1733 and 2520 respectively , which indicated that the layer farms were mostly damaged. The number of chicks and adult birds died in the broiler farms were 632 and 750, respectively. These figures in the cases of layer farms were 752 and 885, respectively (Table 3.2).

Estimation of Direct Losses

FAO estimated the economic costs for the avian influenza affected countries in 2004 (AGAP/FAO, 2004). These costs include d lost birds, production down time, lost trade, losses in associate industries and direct cost of control measures. In this study, we have considered some other items, which becomes a component of losses either directly or indirectly to the specific poultry farms.

The direct costs include d the cost of death and culled chicks and birds in the farms. The damages of eggs, feeds and medicines were also considered. About 80% of the total direct losses occurred due to the losses of chicks and birds that dominated in total direct losses (Table 3.3).

 Estimation of Indirect Losses

The indirect losses were observed due to wage losses of associate industries both having backward and forward linkages , steriliziation (disinfection) cost, productivity loss of the birds, and reduction of meat and egg consumption by the owners and other specified consumers and finally the profit losses (Table 3.4) which were forgone due to infection of AI. It was found that many of the wage laborers were deprived of due to outbreak of avian influenza in broiler farms while a significant amount of profit losses were found in layer farms (Table 3.5

Source: Recalculated from Field Survey (2008)

Estimation of Total Losses at Farm and National Level

The total losses of poultry farms were estimated and it was found that the direct losses were higher (71.33%) in broiler farming while that was only 12% in the case of layer farming (Table 3.6). It means that the different stakeholders were involved in layer farming and loss of productivity of the parent stocks was many-folds by which the indirect loss beca me much more higher than the broiler (Table 3.5).  It was observed that a lot of market intermediaries were involved in layer farming and they were affected due to AI, while in the case of broiler farming the farm owners were directly losing. In all cases, the production down time loss was considered 30% over 2007 and 2008 due to the affect of avian influenza, which continued severely for about 6 months. Yalcin (2006) calculated 12% production down time loss of total value of poultry production in Turkey where the duration of avian influenza was only one month. In the case of Bangladesh, the duration of AI in 2007 and 2008 was about 6 months and the magnitude of incidence was considerable.

Rahman (2004) estimated 4 years back that the total value of poultry products was about Taka 5,000 crore (50,000 millions) every year. This has increased significantly in recent years and the present value of poultry products and by-products is estimated at about Taka 12,000 crore (120,000 millions). Thus, the imputed value of production down time of poultry and poultry products in 2007 and 2008 was worked out at Taka 3,600 crore (36,000 millions) and the total value of loss for the nation was Taka 3,858.31 crore or 38,583.10 millions(Table 3.7).  This estimate seems to be at a lower side than that of the Poultry Association of Bangladesh. One can imagine that the later estimate (Taka 4,165 crore or 41,650 millions) might have some margin of upward bias.

Table 3.6: Total direct and indirect losses of farm owners per farm per batch during bird flu situation

4. Backward and Forward Linkages of Commercial Poultry Farms

An underdeveloped economy is characterized by very weak backward and forward linkages in agriculture. In this chapter, the impact of 2008 avian flu is discussed on the backward and forward linkages of poultry industry in Bangladesh.

The Inter-linkages of Different Actors in Poultry Industry

According to Ghatak and Ingersent (1984), backward linkage measure the ratio of intermediaries inputs purchased from other indus tries to a particular industries’ total value of production. The forward linkage measures the ratio of intermediate inputs sales to other industries to a particular industries total sale.

The backward and forward linkage between commercial broiler/layer industry and their production scale is shown in the following diagram.

Fig. 1: Backward and forward linkages of poultry industry in Bangladesh

The above diagram distinctly depicts the status of backward and forward linkages of poultry farms in Bangladesh. An account of backward and forward linkages industries are placed in Table 4.1. The table indicates that both public and private sectors in the poultry industry have made a large amount of investment. The country has 0.15 million of commercial poultry farms where 0.2 million tons of broiler meat and 3500 million pieces of poultry eggs are produced in every year which is amounted to about Taka 50,000 million. The total amount of investment is about Taka 100,000 million by the private and public sectors where about 6 million people are directly and indirectly engaged. The hatcheries as well as the feed producing mills have been established and the demand for poultry based feed is high.

Source: Rahman (2004).

 Impact of Avian Influenza on Feed Industry

A total of 30 feed mills are available at the moment.  A significant number of people as well as transport chains are engaged both in backward and forward linkages activities of feed mills. There are thousands of feed sellers through out the country and the feed sellers were at loss of their business by at least Taka10,000 per broiler poultry farm while it was about Taka 20,000 in the case of layer farm.

Impact of Avian Flu on Veterinary Service Sector

The veterinary services, which are considered here, include direct cost of sterilization (disinfection) of the infected farms. The cost ranged from Taka 1,500 to Taka 7,500 in each poultry farm. The cost of sterilizing of the layer farms was five times higher than that of the broiler farms because the parent stock of the layer farms were more expensive and farmers were more interested in treatment of layer birds.

 Impact of Avian Flu on Market Intermediaries

There are different types of market intermediaries involved in poultry industry. Everybody was more or loss affected due to the avian flu in 2008. It was found that the sales volume as well as the profit of the market intermediaries significantly decreased and the producers were deprived of reasonable prices of poultry meat and eggs. Thus a huge amount of money was lost by market intermediaries of the poultry sector.

 Impact of Avian Flu on Consumer

A detailed discussion of consumers (either person or restaurant owners) is elaborated in Chapter 7. It was found that many consumers have decreased their consumption of poultry meat and eggs due to the outbreak of avian flu. As a result, there was a reduced sale of poultry meat and eggs.

1. 5. Prices and Sales Volume of Poultry Meat and Eggs
2. 5.1 Marketing Chain of Poultry and Egg

According to American Marketing Association, “Marketing chain is the structure of the intra-company organization units and extra-company agents and dealers, wholesalers and retailers through which a commodity, product or service is marketed”.

Marketing chain plays an important role in achieving the marketing objective of any organization. With the expansion in transportation and communication network, changes in structure of demand and the deve lopment of markets, marketing chain for farm products in Bangladesh have undergone a considerable change, both in terms of length and quality.

The marketing chains of poultry and eggs as found in Bangladesh are shown in Figure 2.

Fig. 2: Marketing chains of poultry and eggs as found in Bangladesh

The first participant in the chain of poultry and egg marketing was producers who raised broilers and layers throughout the year. They used to sell their poultry and eggs to the intermediaries of different markets.  Major portion of the produce was sold to wholesaler and wholesaler cum retailer of poultry and eggs. The wholesalers were other market participants in poultry and egg marketing chain of Bangladesh. They purchased a large amount of poultry and eggs from producers and sold to the retailers, hotels, mega shops and institutions. The wholesaler-cum-retailers were one of the important intermediaries in the chains of poultry and egg marketing chain. They were professional traders who had a fixed establishment in the market places with adequate storage facilities and performed both wholesaling and retailing activities of broiler. They purchased a large number of poultry and eggs mainly from producers and wholesalers and sold these birds to retailers of different markets at wholesale rate and to consumers at retail rate. The retailers were the professional traders who sold their purchased poultry and eggs to consumers directly. Among all types of intermediaries involved in broiler marketing, they were largest in number. Hotels, mega shops and institutions like university hostel, government and private office canteen purchased broilers from producer, wholesaler, wholesaler-cum-retailers or retailers. As a regular customer they got some discount from traders.

Effect of Avian Influenza on Price of Poultry Meat and Eggs

The prices of poultry and eggs declined significantly during the period of avian influenza in Bangladesh in 2008. But after few weeks both the price of broiler and eggs started to increase. The initial reduction of prices of both the commodities during the Avian Influenza in January 2008 was very sharp. However, the prices began to recover from March. The movement of prices of poultry meat from July 2007 to June 2008 is shown in Table 5.1 and Figure 3.

Table 5.1: Comparison of broiler prices between same months of previous and current years

An attempt was made to compare between the prices of same months of previous (2006-07) and current year (2007-08) to find out the change in prices both during avian influenza and normal situation. The price of the current year from the month of January to March was higher than that of the previous year. In avian flu situation, a sharp decrease in prices of broiler was observed. Compared to December 2007 (base price), the highest decline of 27.78 percent was recorded in the month of February 2008. It appears that the price change in 2008 was more severe than that of 2007. Alam et al. (2007) observed that during the avian flu in 2007, the price decreased by about 16 per cent compared with the base price. However, the prices of broiler in the month of April, May and June 2008 went up very sharply in Bangladesh due to an acute shortage of broiler supply and an effort on massive campaign by BLI, DLS, relevant ministries and civil society organizations (CSOs).

Figure 3 shows that the retail prices of poultry were in the limit between Tk. 65 per kg. to Tk. 120 per kg. in 2007-08. During avian influenza situation, the price began to fall down initially in comparison with pre-avian flu situation. The price of poultry was almost stagnant during November–December 2007 (Taka 90 per kg) and decreased slightly in January 2008 and was minimal in February 2008. It appears that the market collapsed during January to March 2008 due to a sharp decline in broiler consumption. The price began to increase from March 2008 and reached the peak in June 2008. The changes were found the highest (33.33 percent) in June 2008 in comparison with the base price of December 2007 (Fig. 4). The higher prices of poultry meat were observed not necessarily for increase in consumption but for decrease in supply of poultry meat. Field evidence showed that most of the affected farms were not in operation even after 3 months of its infection. Even some of the owners opined that they were thinking of shifting from poultry to other alternative business.

An attempt was made to compare the prices of eggs between 2007 and 2008 to find out the change in prices both during avian influenza and normal situation. The prices of eggs was lower in 2006-07 compared to prices in 2007-08, although the prices in January and February of 2008 was lower than that of 2007. The retail price of eggs was in the limit between Tk. 305 – 510 per 100 pieces in 2007-08 (Fig. 5). The egg price began to decrease in the month of January 2008 in comparison with the previous month and this situation existed up to March 2008. In February, the price declined by 26.50%. After the month of March, the price of eggs increased sharply although the avian influenza situation existed. The positive temporal price effect was the highest 22.89% compared to the base price of December 2008 (Table 5.2, Fig. 6.). The higher prices of eggs were observed not necessarily for increase in consumption but for decrease in supply of poultry eggs. Field evidence showed that most of the affected layer farms were not in operation even after 3 months of its infection. Moreover, the poultry hatcheries lost a lot and they were afraid of further investment in poultry hatchery. That might have reduced day-old chick production and supply at home.

Marketing Margin of Poultry Meat and Eggs

Market functionaries move the commodities from producers to consumers. Every function or service involves cost. The intermediaries or middlemen make some profit to remain in the trade after meeting the cost of the function performed. Hence the marketing margin is the sum of these costs and profit. According to Tomek and Robinson (1972, p.110) marketing margin may be defined alternatively as the price of a collection of marketing services, which is the outcome of the demand for and supply of such services. Table 5.5 and Figure 7 shows that the net marketing margins of wholesaler (Tk. 501.19) and wholesaler cum retailer (Tk. 441.05) were the highest for 100kg live poultry. During avian influenza situation, wholesaler cum retailer’s margin was the highest (Tk. 213.82).

• Shadow indicates the first and second attack of Avian Influenza in Bangladesh

 Effect of Avian Influenza on Sales Volume of Broiler and Egg

There was a declining trend in selling of both broiler and eggs during AI in 2008. It was found that the sales volume of broiler decreased more significantly than that of eggs (Table 5.3 and Table 5.4). The fear of broiler meat processing was found more severe in consuming households than that of eggs. The avian flu germ was believed to have destroyed after being fried which was reflected on comparative higher consumption of eggs than broiler.

[

 Marketing Margin of Poultry Meat and Eggs

Market functionaries move the commodities from producers to consumers. Every function or service involves cost. The intermediaries or middlemen make some profit to remain in the trade after meeting the cost of the function performed. Hence the marketing margin is the sum of these costs and profit. According to Tomek and Robinson (1972, p.110) marketing margin may be defined alternatively as the price of a collection of marketing services, which is the outcome of the demand for and supply of such services. Table 5.5 and Figure 7 shows that the net marketing margins of wholesaler (Tk. 501.19) and wholesaler cum retailer (Tk. 441.05) were the highest for 100kg live poultry. During avian influenza situation, wholesaler cum retailer’s margin was the highest (Tk. 213.82).

Table 5.5 and Figure 8 show that the net marketing margins of wholesaler (Tk. 24.95) and wholesaler cum retailer (Tk. 22.00) were the highest for 100 pieces of eggs marketed. But during avian influenza situation, wholesaler’s margin was the highest (Tk. 23.78) followed by retailer (Tk. 19.75). The volume of business reduced during avian influenza situation but the degree of reducing the business was lower for wholesaler than the business of other intermediaries.

Table 5.6 indicates the impact of avian influenza on net marketing margins of all intermediaries of poultry and egg in Bangladesh. The impact of avian influenza was less on egg traders than on broiler traders. The highest decline in net margin was observed (-61.65 per cent) for wholesaler-cum-retailers followed by wholesalers and retailers (Table 5.6). In the case of egg marketing, only retailers’ net margin increased (24.61%) during avian influenza situation in comparison with the normal situation. But other interme diaries ’ net margin decreased because of decline in the volume of business.

1. 6. Estimation of Cross-price Elasticity
2. 6.1 The Elasticity Function

Cross-price elasticity of demand for broiler for different types of substitute products of households included in this survey have been presented in Table 6.1 and are discussed below.

The overall cross elasticity function,

LnY = 1.44 + 0.32*LnX1+0.41*LnX2+0.55*LnX3+1.29LnX4

Table 6 .1: Cross-price elasticity of broiler according to the occupation groups for different substitute products

 Estimated Results of Cross-price Elasticity

Small fish

The cross-price elasticity of small fishes was found 0.32, which means that if the incidence of AI increases by 1 percent the consumption of small fishes increases by 0.32 percent. The cross-price elasticity of small fishes for the day-laborers and farmers were respectively 1.26 and 0.72 which were significant at 5 percent and 10 percent level. Small fis h had no significant elasticities for businessme n, service holders and religious leaders. It had significant cross effect on farmer s and day laborers.

Big fish

The cross-price elasticity of big fishes was found significant to the farmers. Big fish was found to have statistically significant cross elasticity for farmers and all households. This implies that farmers shifted to big fish consumption due to a decline in the consumption of broiler. Day-labor ers had no impact on consumption of big fish due to a decrease in the consumption of broiler because they had limited capability of buying big fish.

Beef

Except for day labor ers, the cross elasticity values of beef for broiler consumption were found to be statistically significant. For a day laborer, it was difficult to buy beef for consumption due to recent price booming. Goat meat

Among the meat items of Bangladesh, goat meat poses the highest unit prices and beyond the limit of maximum consumers of our country except higher income groups. The cross elasticity of goat meat was positive for businessmen and service holders but negative for farmers, day laborers and religious leaders.

Discussion on the Estimated Results

It was found that the consumption of small fish significantly increased for the farme rs, day laborers and religious leaders during the avian influenza, while the consumption of beef increased for the service holders and businessmen. The consumption of big fish as well as goat meat also increased positively for them due to a shift from poultry meat and eggs after the outbreak of avian influenza. Thus, it was observed that the consumption of substitutes of poultry meat and eggs increased significantly with the increase in the incidence of avian influenza and the prices of substitutes increased by 23 to 30 per cent (Table 6.2) after the incursion of avian influenza in Bangladesh.

Table 7.1 indicates the sources of getting information about avian flu on the basis of occupational status of the consumer. It can be seen from the table that most of the consumers heard about avian flu through television, neighbors and newspapers. These are regarded as the major sources of information. Majority (70 percent) of the consumers received information from TV followed by neighbors (66.66 per cent).

 Consumers’ Knowledge, Belief and Attitude about Avian Influenza

Table 7.2 indicates consumers’ knowle dge, belief and attitude about avian influenza. It can be seen from the table that a lot of the consumers (38 percent) have medium knowledge about avian flu. A significant percentage of the respondents felt that avian flu is a contagious and dangerous disease. More than half of the consumers having high and medium knowledge of avian flu reduced their consumption during the outbreak and 15 percent opined that it is harmful for health.

 Impact of Avian Influenza on Consumption of Poultry Meat

Table 7.3 indicates the impact of avia n influenza on consumption of poultry meat in Bangladesh. The consumption pattern is different for consumers having different status of occupation. Table 7.3 demonstrates that the family members of businessmen consumed the highest quantity of broiler and it was 1.21 kg per month per person. But the wage laborers consumed the lowest amount of broiler and it was

0.16 kg per month per person. Most of theoccupational groups reduced consumption of broiler while wage laborers increased their consumption during the outbreak of avian influenza . However, as the crisis was over , all group members increased their consumption but that was less than the pre-avian influenza situation. The consumption  of wage laborers was reduced at the recovery stage because of increase in broiler price.

Fig 7.1: Impact of avian influenza on broiler consumption

Impact of Avian Influenza on Egg Consumption

Table 7.4 indicates the impact of avian influenza in 2008 on consumption of egg in Bangladesh. It is evident from the table that the family members of businessmen consumed the highest number of eggs and it was 8 pieces per month per person. But the wage laborers and farmers consumed the lowest number of eggs and it was 5 pieces per month per person. All of the occupational groups reduced consumption of egg during avian influenza. After avian influenza panic was over , all group members increased their consumption compared with pre avian influenza situation.

Fig 7.2: Impact of Avian Influenza on egg consumption

Problems Related To PoultryIndustry

Although the poultry industry of Bangladesh is rapidly expanding, the scenario in the poultry farming is that almost every day there are new farms coming up and some of the old ones are closing down. It appears that the business environment is quite hazy and full of risk and uncertainty. Poultry industry of the country is now passing through a crisis period as it has been suffering from various problems including avian influenza. Moreover, Bangladesh’s poultry industry is facing competition regionally and globally. With the limited local production of corn and thus the need to rely on imports, feed costs are high. Similarly, the industry is suffering from lack of adequate and authentic market information. No reliable statistics on supply-demand situation of poultry products are available in Bangladesh. In addition to that absence of appropriate technology coupled with absence of adequate poultry experts are also major problems in the industry.

Production problems faced by the farm owners

In the study areas, poultry farm owners faced various problems related to production. Cash capital is a necessary item for the establishment and operation of broiler farms. Hundred per cent of the farm owners mentioned that lack of sufficient capital was a problem to run their business in both pre and post AI situations. Again, as institutional credit was not available and it required complicated procedure, farm owners had to solve this problem by receiving loan from individuals with high interest rate. Sometimes they purchased inputs on credit. High price of day-old-chick was another problem of poultry farming. Table 7.5 shows that around 87 percent farm owners reported this problem. High price of feed in recent weeks was one of the most important problems of poultry farming. Most of the poultry farm owners reported this problem in both situations. Price fluctuation is another problem reported by majority of the farm owners in the study area. Poultry market is not stable; price is changed within a month, even within a week. Not only the poultry sector, but all other sectors of business in Bangladesh have been facing the problem of uncertainty of electric supply. In both situations , farm owners reported this problem. Outbreak of diseases is a dangerous problem for the development of the poultry industry in the country. Akbar (1991) estimated that around 30 percent of poultry had to die every year due to various diseases. As a result, the loss was estimated at Tk. 800 crore (8000 millions). Ranikhe t, Fowl pox, Fowl cholera, Coccidiosis and Gumboro were common diseases in the study areas. Recently , poultry farm owners had to bear tremendous loss due to avian influenza. Table 7. 6 shows that most of the farm owners reported this problem. High price of medicine is another major problem. The price of different vaccines has been increasing day by day. Table 7.5: Problems faced by farm owners in normal and avian flu situations

 Measures suggested by farm owners

In or der to overcome the problems of poultry farming and to make it more profitable, the farm owners of the study area were asked to suggest solutions to the problems. Their replies are summarized in Table 7. 6. As the tables depict, the farm owners made the following suggestions:

Problem faced by the intermediaries of poultry and eggs

The intermediaries involved in poultry marketing also faced various problems which are summarized in Table 7.7. Around eighty per cent intermediaries of both poultry and eggs faced the problem of lack of capital to run their business in both situations. The intermediaries had to feed the broiler birds for maintaining live weight. But the cost of feed was very high for those intermediaries. Table 7.7 indicates that about 95 percent of all intermediaries faced this problem. Price instability was one of the major problems of intermediaries. Because of price instability, they earned lower profit and sometimes, they incurred loss. Tables 7.7 and 7.8 show that cent percent interme diaries of poultry and around 92 per cent intermediaries of eggs faced this problem. Another vital problem is high transportation cost. Around 70 per cent of poultry intermediaries and 60 per cent of egg intermediaries reported this problem in both normal and bird flue situations. All the respondents opined that hartal, strike, flood and natural calamities caused problems for the smooth transportation system and the delay in transportation caused deterioration in egg quality.

Table 7.8 : Problems faced by intermediaries of eggs in normal and bird flu situations

Measure suggested by the poultry and egg intermediaries

Intermediaries suggested that government should play a significant role so that they can purchase feed at reasonable price (Tables 7.9). They suggested further that the government should take proper steps to ensure the stable and reasonable price for poultry and eggs (Tables 7.9 and 7.10). Moreover, in order to maintain price stability, government should fix -up the price and control it. Some intermediaries demanded to increase the existing marketing facilities. They also demanded specific and permanent market place for running their business. Besides, rehabilitation credit for the intermediaries and improved communication system were suggested.

Table 7.10: Measure suggested by the intermediaries of egg in normal and bird flu situation

8. Conclusions

The duration of peak hours for AI outbreaks lasted for about 3 months in Bangladesh in 2008 resulting in immediate fall of prices and consumer’s panic. This analysis depicts that the extent of losses by the poultry sector and its related industry was indeed very high. On the ba sis of the results obtained from this study, the following conclusion and recommendations are made:

.                      •          The poultry industry in Bangladesh faced a terrific financial loss due to avian influenza in 2007 and 2008 when 16,37,266 poultry were culled up to 23 June 2008.

.                      •          Many farm owners were compelled to stop poultry raising as they lost their capital. A large number of farm owners lost their farm business and Government made a moderate effort to provide them incentives.

.                      •          The economic loss of layer farms was 12 times higher than that of broiler farms and the average estimated loss per farm was Taka 431,543 for broiler farms and Taka 59,28,971 for layer farms. The total loss was about Taka 3858.31 core (38583.10 millions) in 2007 and 2008 due to avian flu. Considering this loss, the compensation and other incentive packages for the poultry sector should be re-determined.

.                      •          There was a high level of backward and forward linkage with the commercial poultry farms, specifically with layer farms.  Broiler farming has a strong association with forward linkages. The study recommends for addressing more specifically the backward linkage industries for supporting the layer farming.

.                      •          The price of poultry meat and eggs decreased by 26 to 28% in all cases. Thus, the price support policy during avian influenza could have influence on producers’ profit.

.                      •          The volume of business was reduced during Avian Influenza but the degree of reducing the business was lower for poultry eggs compared to poultry meat. Thus, the study suggests for campaign on safe consumption of poultry meat by the relevant authority.

                        •          The price of poultry began to fall down and it decreased strongly in January and February 2008 that continued till March 2008. The price began to increase from April 2008 and became higher in the following months. The price of eggs began to fall down from the month of January 2008. It decreased sharply up to the month of February 2008 and then increased sharply since April 2008, although the avian influenza situation was almost over by that time. There was a declining trend of both broiler (about 40%) and egg (about 30%)

                        consumption during the incursion period. . It was found that the sales volume of eggs decreased and was lower than that of broiler. The fear of broiler meat processing was found more severe than that of eggs. Bio-security measures were not properly followed in most of the farms. Strict bio-safety measures along with awareness campaign are suggested in the poultry farms throughout the year.

.                      •          Majority (about 70 percent) of the consumers’ received information on avian flu from TV followed by neighbor (about 63 per cent). About forty per cent of the respondents felt that avian flu is a contagious and dangerous disease. Thirty six per cent poultry meat consumption was reduced during avian flu and consumers opined that it is harmful for health. Thus, the study suggests for conducting training to create awareness, improve bio-safety and increase preparedness of the farmers and consumers on avian flu in the country.

Chapter 5:Conclusions and Recommendations

Conclusions

This study was conducted by BLRI in 2008 to find out the origin and methods of spreading of highly pathogenic avian influenza and to assess the socio-economic impact of its outbreak in Bangladesh. For this purpose, three distinct analyses, namely phylogenetic, epidemiological and socio -economic were carried out.

Results of the phylogenetic study shows that Bangladeshi isolates belong to the sub-clade 2.2 which are similar to those from Kuwait, Mongolia, Russia and Afganistan. Since Bangladesh does not have any poultry trade link with these countries, the investigation suggest that migratory birds might be responsible for initial introduction of HPAI in Bangladesh.

From the epidemiological study, it may be predicted that after introduction of HPAI in Bangladesh the virus spreaded to waterfowl and initially native chicken and ducks might have been infected after having contact with the migratory birds. Then, HAPI virus would have spreaded to commercial farms through movable poultry workers, feed and medicine suppliers, feed and egg carriers, egg tray, wet market, weak bio-security etc. There after the infection spreaded all over the country.

Socio-economic analysis shows that the poultry industry of Banglade sh faced a terrific financial loss in 2008 due to the second wave of avian influenza, which was estimated to be at Taka 3858.31 crore (38583.10 millions). The price of broiler declined by about 28 per cent while the price of eggs decreased by 26.5 percent. More than one third of consumers refrained from consumption of broiler meat and eggs. As a result of market collapse , many farm owners were compelled to abandon poultry raising. Nevertheless, after the tension of severe outbreak of avian influenza was ove r, the consumers increased their consumption and the poultry industry went back to the business, except in some cases like those of layer farms and also small broiler farms.

The study reported here might have its limitations and constraints on area coverage, sample selection and recall bias and not to mention of time constraint and financial hard-up. The short-comings might be improved in a prospective study proposed as a regional one w ith the help of FAO/OIE/WHO provided it is agreed upon by the countries concerned.

Recommendations

The Poultry sector has created an income and a self-employment opportunity in the country. Recent threat of AI has put an obstacle to achieve these desired goals. Also food security will be at stake and the rate of protein intake will be reduced due to outbreak of this disease. As remedial measures, immediate attention on the following issues is recommended:

Specific areas of Bangladesh be earmarked as sanctuarie s for migratory birds. It is recommended further that measures be taken so that native birds (Indigenous birds) do not intermingle with migratory ones.

Early detection of HPAI virus and prompt stamping out be ensured.

Surveillance and monitoring activities be strengthened in the affected and control areas by setting up coordination among Government, private organizations and NGOs.

Bio-security measures be strictly adhered to.

Timely sharing of information and effective coordination between animal and public health personnel be ensured.

The owners of the affected poultry farms be rehabilitated promptly and provided with soft bank loans and other financial incentives.

Insurance coverage to save the industry be introduced with compulsory registration of poultry farms.

Bird flu panic be removed from the consumers by training and sensitizing people through media.

Efforts be made to share information and undertake studies on HPAI bi-laterally and regionally to minimize the risk of sustained endemic nature of avian flu in poultry in the region.

References

Acharya,S.S and N.L. Agarwal (1999): Agricultural Marketing in India Third Edition, Oxford & IBH Publishing Co. PVT. Ltd, New Delhi.

Akbar, M.A., Islam, M.S. and Bhuiya, M.S.U. 1995. Socioeconomic survey of small-scale rural mixed farming (crop/livestock) of Bangladesh with a view to introducing fodder legumes in their cropping system. Research report number 1. Extra Mural Contract A0447. Natural Resources International (UK) and Bangladesh Agricultural University Research System, Bangladesh.

Akey, B.L., 2003. Low-pathogenicity H7N2 avian influenza outbreak in Virginia during. Avian Dis. 47, 1099–1103 Akteruzzaman M. (1993) An economic evaluation of cattle distribution program of BRAC in some selected areas of Mymensingh district. Research Report, Submitted to Winrock International, BARC, Dhaka-Bangladesh. Akteruzzaman, M. A. M. Miah, M. M. Hossain and A. H. M. S. Islam (2006b). Impact of MFTS project for dissemination of livestock technology and improving livelihood of the project beneficiaries in Bangladesh. Progressive Agriculture, 17(1): 339-347 Alam J., Akteruzzaman M., Rahman, S.M.A. and Sayeed, M. A. (1993) The Cost of Semen Production and the Rate of Conception for Artificial Insemination in Cattle. Asian-Australian Journal of Animal Science, Vol. 6 (No. 1) 13-17 Alam J., Kabir M., Ahmed Z., Rahman S.M.A., Akteruzzaman M. and Malek, M. A.  (1992) Status of Adoption of Artificial Insemination Technique in Bangladesh, Jahahgirnagar Review, Vol. 13-16 (Part A) 131-143 Alam, J. 1995. Livestock Resources in Bangladesh: Present Status and Future Potential. University Press Limited, Dhaka. Alam, J. and Akteruzzaman, M. (2003) Current and Potential Market Supply and Demand, Marketing Opportunities and Consum er Preferences for Indigenous Breed Animals/Products, Technical Report , ILRI-BAU, Mymensingh. Alam, J., Aktaruzzaman, M., Rahman, S.M.A. and Palash, M.S. 2008. Impact of recent outbreak of avian influenza on Production, Marketing and Consumption of poultry Meat and Eggs in Bangladesh. Bangladesh Livestock Research Institute, Savar, Dhaka. Alam, J., Akteruzzaman, M., Rahman S.M.A and Ahmed, Z. (1994) Comparative performance of Local and Cross-bred Cows in Bangladesh, Indian Journal of Dairy Science, Vol. 47 (No. 2) 112­117 Alam, J.2008. Incidence of avian influenza in Bangladesh. The Bangladesh Observar, April 12, 2008. Alexander, D. J. 1982. Avian influenza: Recent developments. Vet. Bull. 52:341-359. Alexander, D.J., 1995. The epidemiology and cont rol of avian influenza and Newcastle disease. J. Comp. Pathol.112, 105–126. Alexander, D.J., Brown, I.H., 2000. Recent zoonoses caused by influenza Aviruses. Rev. Sci. Tech. 19, 197–225.

Alexander, D.J., Brown, I.H., 2000. Recent zoonoses caused by influenza Aviruses. Rev. Sci. Tech. 19, 197–225.

Anon. 2007. Towards a unified nomenclature system for the highly pathogenic H5N1 avian influenza viruses .www.who.int/csr/disease/avian_influenza/guidelines/nomenclature/en/index.html

Anthony W. Mounts, Heston Kwong,Hector S. Izurieta, Yuk-yin Ho, Tak-wong Au,Miranda Lee, Carolyn Buxton Bridges,3Seymour W. Williams, Kwok Hang Mak,Jacqueline M. Katz, William W. Thompson,Nancy J. Cox, and Keiji Fukuda(1997): “Case- Control Study of Risk Factors for Avian Influenza A( H5N1) Disease, Hong Kong.”

BBS (2002) Statistical Year Book of Bangladesh. Bureau of Statistics, Statistical Division, Ministry of Planning, Government of the People’s Repablic of Bangladesh, Dhaka. Becker, R. A., Chambers, J. M., Wilks, A. R (1988). The New S Language. Cole, Wadsworth & Brooks. Best, D. J., Roberts, D.E. (1975). “Algorithm AS 89: The Upper Tail Probabilities of Spearman’s rho.”

Applied Statistics 24: 377–379.

BLS (2007). Baseline Survey Report, PKSF, Dhaka.

Brown, I.H., Ludwig, S., Olsen, C.W., Hannoun, C., Scholtissek, C., Hinshaw, V.S., Harris, P.A., McCauley, J.W.,Strong, I., Alexander, D.J., 1997. Antigenic and genetic analyses of H1N1 influenza Aviruses from Europeanpigs. J. Gen. Virol. 78, 553–562.

Collett, D. (1991). Modeling binary data. London, Chapman and Hall.

Delgado C., Rosegrant M., Steinfeld H., Ehui S. and Courbois C. (1999) Livestock to 2020 – the next food revolution. Food, Agriculture and Environment Discussion paper no. 28, IFPRI, Washington, D. C, USA.

Delong, E. R., Delong, D.M., Clark-Pearson, D.L. (1988). “Comparing the areas under two or more correlated reciever operating characteristic curves: a nonparametric approach.” Biometrics 44: 837-845.

Department of Agricultural Marketing (DAM, 2002). Khamarbari, Farmgate Dhaka.

Dirk U. Pfeiffer, Phan Q. Minh, Vincent Martin, Michael Epprecht, Martin J. Otte(2007): An analysis of the spatial and temporal patterns of highly pathogenic avian influenza occurrence in Vietnam using national surveillance data.

Dohoo, I. R., Martin, W., Stryhn, H. (2003). Veterinary Epidemiologic Research. Charlottetown, AVC Inc.

Ducatez MF, Olinger CM, Owoade AA, Tarnagda Z, Tahita MC, Sow A, De Landtsheer S, Ammerlaan W, Ouedraogo JB, Osterhaus AD, Fouchier RA, Muller CP. (2007). Molecular and antigenic evolution and geographical spread of H5N1 highly pathogenic avian influenza viruses in western Africa.. J Gen Virol. 88:2297-306.

Eng, J. (200 6). “ROC analysis: web-based calculator for ROC curves. Baltimore: Johns Hopkins University.” Retrieved 23 June 2007, from http://www.jrocfit.org.

Gert Jan Boender, Thomas J. Hagenaars, Annemarie Bouma, Armin R.W Elbers, Mart C. M. de Jong, Michiel van Boven(2007):Risk Maps for the Spread of Highly Pathogenic Avian Influenza in Poultry.3:e71.

Giasuddin, M, Alam, J., Samad,M.A., Al-Mamun.M.,and Taimur,M.J.F.A. 2006. Highly pathogenic avian influenza: Bangladesh situation.Bang.J.Lives.Res.Vol.13,(1&2),P84-95. Greiner, M., Pfeiffer, D., Smith, R.D. (2000). “Principles and practical application of the receiver-operating characteristic analysis for diagnostic tests.” Prev. Vet. Med. 45: 23-41.

Guan, Y., Peiris, M., Kong, K.F., Dyrting, K.C., Ellis, T.M., Sit, T ., Zhang, L.J., Shortridge, K.F., 2002. H5N1influenza viruses isolated from geese in Southeastern China: evidence for genetic reassortment and interspeciestransmission to ducks. Virology 292, 16–23.

Henzler, D.J., Kradel, D.C., Davison, S., Ziegler, A.F., Singletary, D., DeBok, P., Castro, A.E., Lu, H., Eckroade,R., Swayne, D., Lagoda, W., Schmucker, B., Nesselrodt, A.,( 2003). Epidemiology, production losses, and ontrol measures associated with an outbreak of avian influenza subtype H7N2 in Pennsylvania. Avian Dis. 47,1022–1036.

Hollander, M., Wolfe, D.A (1973). Nonparametric statistical inference. New York, John Wiley & Sons.

Hosmer, D. W., Lemeshow, S. (2000). Applied Logistic Regression. New York, John Wiley.

Hossain, M. I., M. M. Hossain, M. A. Hasim and M. Ahmed (2002) Uses of Slaughterhouse by ­Products in Bangladesh, Bang. J. An. Sc,31 (1-2):37-43

Huque, K.S. 2005. Personal communication. BLRI, Savar, Dhaka.

Islam M. S. (1998) A study on some productive and reproductive parameters of Local x Friesian upgraded dairy cows at Government Dairy Farm, Faridpur. M. S. Thesis , Department of Dairy Science. BAU, Mymensingh.

Islam MR, Baqi MA, Giasuddin M and Samad MA (2008). Molecular characterization and phylogenetic analysis of highly pathogenic H5N1 avian influenza virus of chickens of Bangladesh. Proceedings of the Bangkok International Conference on Avian Influenza 2008: Integration from Knowledge to Control. 23 – 25 January 2008, The Dusit Thani Bangkok, Thailand.

Islam, M. R. and Giasuddin, M. 2007. Highly Pathogenic Avian Influenza Global situation and Bangladesh. 5^th International Poultry show and Seminar, WPSA-BB, 1-3 March 2007, Dhaka, Bangladesh.

Jabbar, M. A. (2003) Livestock in food security and poverty alleviation in Bangladesh, First Graduation, Chittagong Government Veterinary College, University of Chittagong, Bangladesh.

Jabbar,M A and E.Tambi and G.Mullins (1997): A Methodology Characterizing Dairy Marketing System; Market Oriented Small Holder Dairying Research, International Livestock Research Institute, Working Document,No.3,p.11

Jolliffe, I. T. (2002). Principal Component Analysis. New York, Springer.

Kamal RP, Tosh C, Pattnaik B, Behera P, Nagarajan S, Gounalan S, Shrivastava N, Shankar BP, Pradhan HK (2007). Analysis of the PB2 gene reveals that Indian H5N1 influenza virus belongs to a mixed-migratory bird sub-lineage possessing the amino acid lysine at position 627 of the PB2 protein. Arch Virol. 152:1637-44.

Karasin, A.I., Brown, I.H., Carman, S., Olsen, C.W., 2000. Isolation and characterization of H4N6 avian influenzaviruses from pigs with pneumonia in Canada. J. Virol. 74, 9322 –9327.

Khan, M.A.S., Akbar, M.A. and Siddiki, M.S.R. 2006. Studies on improving the low quality roughage feed resources of ruminants for milk and meat in developing countries. In: Challenge of Animal Industry for the Wellbeing of Mankind. Proc. XIIth AAAP Animal Science Congress-2006, held in Bexco Busan, Korea, September, 2006. pp. 18-22.

Kohls, R. L. and Uhl, J. N. (1980) Marketing of Agricultural products . 5^th edition. Macmillan Publishing co. Inc. New York.

M.E. Thomas et al. / Preventive Veterinary Medicine 69 (2005) 1–11 Metz, C. E. (2006). “Rockit 1.1B2, Beta version.” Retrieved 26 June 2007, from

http://xray.bsd.uchicago.edu/krl/KRL_ROC/software_index6.htm#Programs%20used%20to% 20perform%20ROC%20calculations.

Miah, M. A. M. (2002) Production and Marketing of Livestock and Livestock Products in Selected Peri-Urban Areas of Bangladesh. Un-publ ished Ph. D. thesis, Department of Agricultural Economics, Bangladesh Agricultural University, Mymensingh.

National Reference Laboratory for Avian Influenza at Bangladesh Livestock Research Institute, Savar, Dhaka. Ninomiya, A., Takada, A., Okazaki, K., Shortridge, K.F., Kida, H., 2002. Seroepidemiological evidence of avianH4, H5, and H9 influenza A virus transmission to pigs in southeastern China. Vet. Microbiol. 88, 107–114.

OIE Disease Information, 2003. http://www.oie.int .

Patefield, W. M. (1981). “Algorithm AS159. An efficient method of generating r x c tables with given row and column totals. .” Applied Statistics 30: 91-97.

Peiris, J.S., Guan, Y., Markwell, D., Ghose, P.,Webster, R.G., Shortridge, K.F., 2001. Cocirculation of avian H9N2and contemporary ‘‘human’’ H3N2 influenza A viruses in pigs in southeastern China: potential for geneticreassortment? J. Virol. 75, 9679–9686.

Raha, S. K. (2003) Poultry farming under Participatory Livestock Development Project (PLDP): An agribusiness study, Research Report, Submitted to Bureau of Socioeconomic Research and Training (BSERT), Bangladesh Agricultural University, Mymensingh, Mymensingh, Bangladesh.

Rahman M., Mia M. I. A. and Haque S. A. M. A. (2000) A Policy Study on the Dairy Development in Bangladesh BARC, Dhaka, Bangladesh.

Rahman, K. T. (2004) Marketing System and Consumer Preference for Broiler with Special Emphasis on Impact of Bird Flu Rumour in an Area of Bangladesh, MS Thesis, Dept. of Cooperation and Marketing, Bangladesh Agricultural University, Mymensingh.

Saadullah, M. 1995. Integrated Farming for Rural Poor. Rural Poor Programme Task Force of BRDB/SIDA, Dhaka.

Saif, Y.M., Barnes, H. J., Glisson, J. R., Fady, A. M., McDougald, L. R. and Swayne. (2003). Disease of Poultry (11^th Edition). Pp.135-160.

Salzberg SL, Kingsford C, Cattoli G, Spiro DJ, Janies DA, Aly MM, Brown IH, Couacy-Hymann E, De Mia GM, Dung do H, Guercio A, Joannis T, Maken Ali AS, Osmani A, Padalino I, Saad MD, Savic V, Sengamalay NA, Yingst S, Zaborsky J, Zorman-Rojs O, Ghedin E, Capua I.(2007). Genome analysis linking recent European and African influenza (H5N1) viruses. Emerg Infect Dis. 13:713-8.

Scholtissek, C., Burger, H., Bachmann, P.A., Hannoun, C., 1983. Genetic relatedness of hemagglutinins of the H1subtype of influenza A viruses isolated from swine and birds. Virology 129, 521 –523.

Sims LD, Domenech J, Benigno C, Kahn S, Kamata A, Lubroth J, Martin V, Roeder P.(2005). Origin and evolution of highly pathogenic H5N1 avian influenza in Asia [Review]. Vet Rec. 157:159-

64.

SPSS for Windows, 1999. Standard Version, SPSS Inc., 1989–1999. Statistics Netherlands, 2002.

http://statline.cbs.nl/Stat Web/Start.asp?lp=Search/Search&LA=EN&DM=SLEN

Tareque, A.M.M. and Saadullah, M. 1988. Feed availability, requirements for animals and current pattern of utilization in Bangladesh. In Non-Conventional Feed Resources and Fibrous Agricultural Residues, IDRC, Canada.

Wang J, Vijaykrishna D, Duan L, Bahl J, Zhang JX, Webster RG, Peiris JS, Chen H, Smith GJ, Guan Y.(2008). Identification of the progenitors of Indonesian and Vietnamese avian influenza A (H5N1) viruses from southern China. J Virol. 82:3405-14.

WHO (2006): Weekly epidemiological record.26:249-260.
WHO. (2007). http://who.int/csr/disease/avian_influenza/country/cases_table_2007.
Yalcin, Cengiz. 2006. Market Impacts of HAPI Outbreaks: A Rapid Appraisal Process Turkey, FAO

Publication, Italy, Rome.