AFGHANISTAN NATIONAL LIVESTOCK CENSUS 2003 INTERIM REPORT OSRO/AFG/212/ITA AFGHANISTAN NATIONAL LIVESTOCK CENSUS 2003 INTERIM REPORT FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS Contents Acknowledgements iii Executive summary iv Background v ACTIVITIES 1 ■ Organization and planning 1 ■ Recruitment and training 2 ■ Data collection 2 ■ Data entry and checking 3 ■ Data analysis 3 ■ Results 4 ■ Districts 4 VILLAGES, COMMUNITIES AND FAMILIES 6 LIVESTOCK NUMBERS 8 HERD STRUCTURES 14 ■ Female cattle older than two years 14 ■ Young stock 16 ■ Cattle 16 ■ Smallstock 16 ■ Draft animals 18 CHANGES IN NUMBER OF FAMILIES OWNING LIVESTOCK 20 DISCUSSION 22 APPENDICES 1. Explanation of district comparisons between the official list (Afghanistan Ministry of the Interior), 2003, AIMS 2001 and the FAO Livestock Census field data, 2003 2. Total livestock on a district basis 3. Adult cows on a district basis 4. Young stock on a district basis Acknowledgements This work was only possible thanks to the help of many diverse groups including the Government of Italy through the Italian Cooperation Fund, the Afghanistan Ministry of Agriculture and Animal Husbandry (MAAH), the Afghanistan Ministry of Interior (MoI), the Central Statistical Office (CSO) in Kabul, the Dutch Committee for Afghanistan, Oxfam (UK) and the University of Kabul. Assistance and cooperation was gratefully received from many staff in the MAAH including Mr Jawat, Deputy Minister of MAAH, Dr Hanif, General President of Animal Husbandry, and Dr Raufi, General President of Veterinary Services. Supervision of fieldwork would not have been possible without the enthusiastic support of many staff from the MAAH, the Emergency Operations and Rehabilitation Division (TCE) of FAO, the University of Kabul, the Dutch Committee for Afghanistan and Oxfam UK. Fieldwork was dependant upon the conscientious attention of almost 900 staff, mainly from Veterinary Field Units who visited villages and collected the data. A concentrated period of data entry and checking was undertaken by a small group of hard working young staff. Clarification of district names and locations was undertaken with assistance from the MoI, the CSO and Dr Pinney from the Afghanistan Ministry of Reconstruction and Rural Development. Assistance from UN personnel across diverse projects in Afghanistan and Rome is acknowledged, including particularly Dr Thieme, Dr Majok, Dr Favre, Dr Crowley, Mr Mack and Mr Miagostovich. The project in Afghanistan was managed and the final report written by Dr Len Reynolds. Dr Habib Narwos was Assistant Manager in Afghanistan. Data analysis was undertaken by Dr Wolfgang Pittroff of the University of California, Davis, and Prof Fred Dham of Texas A and M University. This report is produced thanks to the dedication of all the above. The authors believe that it justifies all the hard work of the participants and will provide a baseline upon which future Afghan livestock development can be built. iv Executive summary • After the end of a four-year drought and a period of insecurity in Afghanistan, the numbers of livestock were reported to have fallen dramatically since previous estimates were made. In early 2002, the Ministry of Agriculture and Animal Husbandry [MAAH] of the Transitorial Administration of Afghanistan requested FAO to organize a national livestock census. • The work started in Afghanistan in October 2002 with funding from the Italian Government, and fieldwork was completed by April 2003. Every village in Afghanistan was visited, with the exception of Barmal District in Paktika Province due to insecurity, and parts of Ghor Province due to inaccessibility over the winter period. • A team of 28 supervisors, 24 women surveyors and 821 enumerators were employed for data collection, and 14 data entry personnel and 10 data checkers undertook data entry and checking. This was all overseen by a Census Manager and an Assistant Census Manager. • Basic data (Level I) covering livestock numbers, herd structures and changes in ownership as a result of the drought were collected at the community level from every village in Afghanistan (with the above exceptions). Production system data (Level II) was collected by supervisors from selected households in randomly selected villages. A women’s survey covering production systems and women’s attitudes to livestock was collected from women in selected households, in randomly selected villages. • The collected data covered 3 044 670 families in 53 214 communities across 36 724 villages. • There were a total of 3.72 million cattle, 8.77 million sheep, 7.28 million goats, 1.59 million donkeys, 0.18 million camels, 0.14 million horses and 12.16 million poultry. • The number of cows kept per family was low, with only Khost, Kunar, Laghman and Nuristan Province reporting more than 1.5 cows per family. The data on calves suggests that in many areas, restocking of depleted cattle herds would not be possible at the current reproduction rates. • Similarly reported numbers of young sheep and goats appear low in many areas, and rebuilding of herds will be slow to achieve. • The numbers of families without livestock have increased as a result of the drought from 4 families to 14.4 families per community. However, at the same time, numbers of family per community has increased. • There was no pre-drought livestock census data against which direct comparisons could be made. However, earlier survey results combined with the information from the present Livestock Census indicate that stock holding per family has decreased sharply over the period. • When analysis of Level II and the Women’s Surveys is completed, further interpretation of the Level I data will be undertaken. v Background ivestock plays a fundamental role in Afghan agriculture, but existing information on the numbers of animals in the country and their distribution predates a four-year drought and L nationwide disruption. It is said that animal numbers have fallen sharply since earlier surveys in 1995 and 1998, but there are no statistics to substantiate the claim. Farmers are aware of the need to vaccinate their animals against diseases, but animal numbers are required for planning veterinary campaigns. Reliable statistics are also required to provide the basic information to prepare livestock development activities. However, to formulate plans we need to know the starting point as well as our end target position. A livestock census provides the basic data required for livestock development activities. In 2002 the Italian Government agreed to contribute funds for the agricultural subsectors of the Intermediate and Transitional Assistance for the Afghan people (ITAP), including the provision of funds for the National Livestock Census. vi 1 Activities ■ ORGANIZATION AND PLANNING The National Livestock Census started in Afghanistan in September 2002 with the arrival of an external consultant and the Livestock Census Manager. The consultant was responsible for the design of census survey forms in consultation with the Census Manager. A basic form (Level I Livestock Census) was drawn up covering ruminants and monogastric animals including poultry, to determine livestock numbers and herd breakdown by age and sex (Appendix 1). Level I was designed to be conducted by enumerators on every village in Afghanistan, or on component communities within large villages. A second census form (Level II Survey), to be administered by the Supervisors, covered details of production and marketing systems (Appendix II). The Level II form was more detailed and designed to be administered to individual households in randomly selected districts, villages and households. A further survey, the Women’s Survey, was administered by female enumerators to women livestock keepers, selected randomly in selected districts (Appendix III). Work plans were drawn up by early October 2002, with the aim of completing data collection before livestock moved out of their winter quarters with the onset on spring weather. The start of fieldwork was divided so that provinces with predominantly mountainous areas, where winter conditions would hinder the movement of the data collectors, would start earlier than lower altitude provinces. Two levels of management were established. The Census Manager and the Assistant Manager based in Kabul were responsible for the nationwide management of activities. Supervisors covering two to three provinces provided the second management level, with Data Collectors (field staff) based in each district who were responsible for visiting all villages in their district and collecting information on animal numbers. A national database kept by the Afghanistan Information Management System (AIMS) administered under the United Nations Development Programme (UNDP) provided a list of provinces, districts and villages. Afghanistan comprised 32 provinces, 329 districts and 30 172 villages. Provinces varied from 4 districts (Saripul) to 22 (Nangarhar). Rural districts varied from 4 villages (Andkhoy in Faryab Province) to 620 (Daykundi in Uruzgan Province). Based on population estimates from AIMS, and assuming 7 persons per family, village sizes varied from 6 families per village in the Wormamay District (Paktika Province) to 3 016 families in the Andkhoy District (Faryab Province). However, much of the village data was derived from information compiled in the 1970s. The AIMS material was taken as the planning base for the Livestock Census. Complications arose from variations in the number of districts from the pre-2000