The Planning of Emergency Seed Supply for Afghanistan in 2002 and Beyond
Composite Version (345k MS Word file)
Contents Findings Part I Part II Part III References Abbreviations/Glossary Appendix 1 2 3 4 5 Maps
Part I : Context of Providing Seed Aid to Afghanistan « previous page | next page »
1.0 - Introduction
1.1 - Food Production
1.2 - Recent Wheat Improvement Efforts
1.3 - Are different strategies needed for the irrigated and non-irrigated areas?

1.3 - Are different strategies needed for the irrigated and non-irrigated areas?

1.31 The nature of robustness: traditional use of genetic diversity

Afghan agriculture has been robust. The resilience and robustness of Afghan agriculture is reflected in the fact that from 1978 to 1999 rural production systems in Afghanistan continued to support the rural population under conditions of extreme difficulty. Although malnutrition and hunger have been constantly observed, the situation did not degenerate into a situation as catastrophic as in some other less fortunate countries of other continents.

Crop robustness is a condition that to a large degree is conferred by the use of relevant genetic diversity. In highly diverse and stressed agroecosystems genetic diversity has been shown to be an important risk-reducing mechanism for farmers. In addition, it has been shown again and again that relevant diversity will in stressed environments generally increase yield over genetically-homogenous crops (Wolfe,1985). Genetic diversity is an ancient time-tested mechanism that has been used in agriculture. Its value has in the last decades been discovered again scientifically and is increasingly being used again in modern agriculture. Local landraces and mixtures contain a diversity of characteristics, including resistance to many diseases. They are the source of resistance for virtually all resistance used in modern breeding programs. The landraces and mixtures have been little studied and hence little is known of how farmers manipulate resistance genes in genetically diverse systems. However, it is known that farmer varietal mixtures of Phaseolus beans contain a substantial number of component varieties that are resistant to disease and that the proportion that these varieties occupy in the mixture increases with increased disease pressure (Trutmann et al., 1993). It is likely that a similar mechanism of gene deployment is used in landraces of other crops. The value of using genetic diversity in the form of multilines and mixtures has been demonstrated also for wheat(Browning, 1967, 1974). Most recently it was shown that that significant disease reductions as well as substantial yield increases, equal or better to the best yields in single HY varieties, could be obtained over large areas by enhancing the spatial distribution of genetic diversity by intercropping a hybrid and a local variety of rice (Zhu et al.,2000). Under the intercropping of varieties approach, farmers no longer needed to use pesticides.

Genetic diversity, especially in the case of landraces and varietal mixtures, works in a unique manner to mitigate the effects of especially biological stresses and to increase yield. In genetically variable crops that vary in the level of resistance to diseases, the resistant plants in the crop will act to protect susceptible plants. The mechanism is multiple and includes resistant plants acting as barriers, increased space between susceptible plants, induced resistance of susceptible plants to virulent strains by non-virulent races of the pathogens, and compensation by resistant plants growing into photosynthetic space lost from killed plants. The overall phenomenon of disease reduction and increased yield over the expected is called "the mixture effect" (Wolfe, 1985,Trutmann and Mukishi,1994).

Genetic diversity also acts simply as a means of reducing risk. For instance, if 45% of a landrace or mixture is more tolerant to drought, then, if there is a drought that kills susceptible plants, the farmer will still retain 45% of a normal crop.

1.32 Rain-fed systems

In the traditional northern "bread basket" of Afghanistan rain-fed communities depend on lalmi wheat. These have been especially severely hit by the drought (Fitzherbert, 2000).

In the rain-fed areas landraces are still grown widely. Landraces are made up of similar but highly diverse germplasm adapted over centuries by selection to the local conditions. Under broad local conditions where no external inputs are used, these landraces perform better than most if not all "improved" varieties. Although no studies are known to have been conducted in Afghanistan, it is likely that the reasons for the very limited, slow diffusion and acceptance of improved varieties by farmers in rainfed and some irrigated regions have a sound basis that should be respected. In the absence of a viable economic infrastructure to support a highly developed market system that would provide on-going seed supplies and ensure profitable returns and markets for farmers.

In some cases when farmers move into a new area, or if seed is lost and farmers have to start from scratch, as much as possible farmers will collect local seed, but will sow seed from as many sources as possible and harvest the surviving seed. In this manner a selection is made for a new locally-adapted landrace.

This is not to suggest that farmers are averse to trying out new germplasm. On the contrary, farmers, even the resource-poorest, in most countries have been found to be eager to try out new technologies including new germplasm. New germplasm is tested separately (Voss, 1992), or if already tested locally used quite readily. However, it typically is not used to replace the local landrace outright, but rather to augment the available diversity.

Table 1.1. Exotic Improved Rainfed Rice Cultivars Recently Extended in Afghanistan
# Year of release Name of cultivar Introduced from Area for which recommended
1 1996 Ghori- 96 Mexico /Syria Western rainfed area
2 1996 Diama-96 CIMMYT Northern Rainfed areas
3 2000 Lalmi-1 (Fow-1) CIMMYT and ICARDA rainfed areas
4 2000 Lalmi-2 (Bobwhite1//Mn..) CIMMYT and ICARDA from Syria Rainfed areas
5 2000 Lalmi-3 (Florkwa-3) CIMMYT and ICARDA Rainfed areas

Table 1.1 above is a list of exotic improved rainfed varieties that FAO-Afghanistan has released to rainfed farmers since 1996. To our knowledge, none of these cultivars contains any genetic material originating from Afghanistan. No local wheat variety, rainfed or irrigated, is named in the two recent FAO reports we have accessed (Tunwar, 1998; FAO,2001) which gives the impression that wheat farmers are "on their own" when it comes to replicating and reselecting local variety seed. Although rainfed wheat farmers make up approximately 44% (the 1998 areal percentage) of all households growing wheat in Afghanistan, only five of 23 modern varieties released in recent years, or 22%, is a rainfed variety. This sort of bias towards irrigated cropping systems can have adverse effects on social equity.

It is unclear what areas these five rainfed varieties - all of them released in the last 2-6 years--occupy on farm and thus what proportion of rainfed wheat is grown with improved varieties. Such information would seem critical to making objective assessments of need and for implementing the "replicating what has been there" principal of emergency relief.

1.33 Irrigated systems

The irrigated systems, are the principal agricultural production areas in Afghanistan. It is here that input-responsive high yielding dwarf wheat appears to have been adopted on a larger scale. However, again we have been unable to obtain reliable figures on areas covered by various HYVs. Table 1.2 below lists the newer HY wheat varieties that have been tested and released by FAO-Afghanistan. . Again, for varieties appropriate for irrigated areas, we have so far not been able to find a single local wheat variety named in either FAO report we have accessed (Tunwar, 1998; FAO,2001), and, again, we find this troubling. It appears that local wheat varieties have just not been a part of FAO-Afghanistan's strategy of support to Afghani farming communities.

We are concerned that apparently no maps yet exist showing the distribution of HY wheat varieties, let alone the distribution of particular varieties. We suggest that the Russian 1:50,000 map series, possibly using a digitized version, could be overlayered with survey information showing areas of cereal cropping taken from RS imagery‹perhaps 2-3 sets from a base year in order to distinguish winter wheat from Spring wheat‹and another layer indicating SCA survey data.

The map, "Cereal I," shows the location and size of yield of wheat and barley-growing areas as of the early 1980s. The map, "Cereal II," shows the location and size of yield of the maize, rice, and the "millet and sorghum" crops for the same time period. And the "General Economic Map" shows the location of irrigated farming in the early 1980s and distinguishes between double cropping for rice, wheat, maize and barley and single-cropping for wheat, maize, and barley. The same map displays rainfed areas (wheat and livestock) in yellow.

Due to the need to keep this version of the report readily e-mailable, these and other maps are viewable at and are not displayed here in the paper. We think it likely that the extent of modern wheat varieties in Afghanistan today is limited to some subset of the olive green (yearly irrigated double cropping) and orange (irrigated yearly single cropping) and much less frequent in the beige areas (irrigated wheat, maize, and barley, one crop every two or three years).

Prof. Azam Gul states that "about 90% of the irrigated wheat grown in Kabul province is modern variety wheat" (conversation with J.Dennis, 16 Jan. 2002).

However, a survey of four district in three provinces lying between Kabul and the Pakistani border in year 2000 gives us considerable concern about the percentage of wheat and maize farmers who would chose to grow modern varieties. We have so far only seen Appendix L (MS Word) of the study report and that does not indicate whether any of the 350 villages surveyed grew rainfed wheat and maize, but were think most of the survey involved irrigated agriculture. Tables 1.3 and 1.4 below summarize the survey findings in the report's Appendix L.

Of 350 villages surveyed, some farming households in 292 of them, or about 83% of survey villages, reported using at least one kind of chemical fertilizer. By contrast, there was no use of modern wheat varieties reported by any surveyed household in 68% of the villages surveyed and no modern maize varieties reported by any household surveyed in 91% of the 350 villages surveyed. As these three provinces bracket the main transport route between Peshawar, Pakistan and the capital, Kabul, and as chemical fertilizer is used in the majority of villages, neither expense nor difficulty of access would seem to plausibly explain the gap between chemical fertilizer use and the much lower use of modern variety seeds. M. Omar Anwarzay, director of the Afghan Survey Unit that carried out the survey in 2000 for UNOPS has explained that farmers in this area do not care for the eating quality of the two modern maize varieties available.

Table 1.2. Improved Wheat Cultivars Currently Grown in Irrigated Areas of Afghanistan near Seed Production Stations
No. Variety Characteristics
01 Ataya-85 Winter Wheat. Susceptible to Rust. Being gradually phased out.
02 Bakhtawar92 (Kauz) Facultative
03 Gul-96 Facultative and cold tolerant.
04 HD 2232 (Balkh 66) Facultative
05 HD 2285 Facultative
06 HD 2329 Facultative
07 Inqilab-91 Facultative
8 Pamir-94 Facultative and cold tolerant.
9 PBW 154 Facultative
10 PS-85 Facultative. Susceptible to rust. Being gradually phased out.
11 Rana-96 Facultative and cold tolerant.
12 Roshan-96 Facultative
13 Sonali (HP 1633) Spring Wheat. Rust Prone. Being phased out
14 Takhar-96 Facultative
15 WH-542 Facultative
16 Mazar-99 Facultative
17 Amu-99 Facultative
18 heart-99 Facultative
Source: adapted from FAO, 2001.

Table 1.3 Summary of HY Wheat and Maize and Fertilizer Use in Four Districts in 2000
District, Province No. of Villages Sampled Villages with Chemical Fertilizer Used Villages withs HY Wheat Used Villages with HY Maize Used
Hesarak, Nangahar Prov. 97 89 30 8
Kahk-e-Jabar, Kabul Prov. 37 34 12 1
Gardez Dist.,Paktia Prov. 107 82 35 6
Sayed Karam, Paktia Prov. 109 87 35 16
Total: 350 292 112 31
% of villages using input: --- 83.4% 32.0% 8.9%
Source: data derived from Appendix L of Year 2000 Survey for UNOPS Carried out by Afghan Survey Unit.

Table 1.4 Modern Wheat and Maize Varieties Reported in Three Provinces in 2000
Province Paktia Paktia Nangarhar Kabul
District Sayed Karam Gardez Hesarak Kak-e-jabar
Villages surveyed: 109 107 97 37
%Villages using HY vWheat varieties 32% 33% 31% 32%
Atay85 X X   X
Bakhtawar X   X  
Inqila91   X X  
Pamir94 X X   X
Pirsabak85     X  
Unknown X X X X
%Villages using HY Maize varieties 15% 6% 8% 3%
Sarhad X X    
Shaheen X X X  
Unknown X X X X
Source: data derived from Appendix L of Year 2000 Survey for UNOPS Carried out by Afghan Survey Unit.

Pending further information, possible explanations include:

  • Villages reporting no use of modern variety seed perhaps don't realize that they are using modern variety seed introduced many years ago and now thought to be local;
  • Modern variety seed has been tested from time to time in the past, but has not been found to be well-adapted to the village areas where it is not used. The higher incidence of modern wheat varieties compared to modern maize varieties could be explained by:
    • Some sample villages not growing any maize;
    • Wheat varieties remaining true to type over many crops due to self-pollination; and
    • HY wheat varieties being better adapted to this area than HY maize.

1.34 Genetic Diversity and Stability.

HYVs have been developed for irrigated areas where the environmental factors important to crop production, especially nutrients and water, are best controlled. Crops like wheat were grown in genetically homogenous uni-varietal monoculture. Still, despite the introduction of progressively better resistant high yielding varieties, the crop loss due to biotic stresses in these systems are staggering.

In Hazarajat, a predominantly irrigated region, a 2000 base line study showed that there was severe crop loss due to biological constraints. It states, "throughout Hazarajat problems of rust and smut are common. ACF in 1998 reported 20-35% of the fields on the Miridineh-Shinideh road being seriously attacked by rust, whilst in the same year Oxfam reports repeatedly underlined the seriousness of rust infestation in the irrigated wheat crop in the districts in which it worked and estimated crop losses at around 40% . The problem is not new - back in 1988 the Swedish Committee for Afghanistan reported that 82% of farmers in Bamiyan identified crop disease as a consistent problem"(SCA,1988).

The reasons for the serous damage caused by rust and smut can be multiple. However, clearly an insufficient level of resistance exists in the varieties grown. Although the main source of seed for farmers appears to be local, it appears that some if not most varieties grown in the irrigated areas are HYVs often derived from breeding programs of CGIAR centers like CIMMYT and ICARDA and evaluated, selected and extended by the national research and extension programs. We have no figures at this time on the extent to which this new germplasm has been adopted and is grown. We are assuming that FAO has good information and is correct in it's strategy to multiply the dominant wheat varieties.

We are assuming it likely that under a top-down, fairly centralized seed program that releases mostly HYV seed, a relatively few HY wheat varieties could come to dominant the total area of irrigated wheat in Afghanistan. In such cases if a variety is, or becomes, susceptible to a pathogen, then the pathogen will find a large expanse of an available food source and under favourable conditions will multiply rapidly. These are the conditions for epidemics and, - "Appendix L" (see below) not withstanding - the direction in which Afghanistan's irrigated wheat systems seem to be moving. The phenomenon is explained well by Wolfe (Wolfe, 2000). Commonly the option has been to provide farmers with new varieties resistant to the new pathogen or races of the pathogen, preferably before the problem reaches epidemic proportions. As a result much of the budgets of national programs, companies and CGIAR centers has been devoted to activities to replace varieties, or maintenance research. For example in 1989, CIMMYT spent 2/3s of its total research efforts on wheat on maintenance research (McKenzie, 1989). For many countries, this is a big expense for which resources are not available.

Analysing varietal diffusion and replacement patterns in Pakistan in the 1980s, Byerlee and Heisey (1990) noted that Pakistan's wheat breeding program had been unable to produce a sufficient number of popular new varieties to replace older wheat varieties, some of which had diminished resistance to rust. While we accept this point, we suggest that a phytosanitary approach that actually bans the use of wheat varieties that have "lost resistance" to rust is in the end self-defeating. By restricting the number of varieties on the landscape in any given crop season, it generates the intense selection pressure on local pathosystems and insect pests and the "homogenous substrate" that results in the emergence of virulence and the breakdown of resistance.

A second option to farmers (but unlikely to be common during the present emergency in Afghanistan due to the cost) is to spray with pesticides. Unless used judiciously, pesticide use increases costs, is often dangerous to farmer and consumer health and has negative impacts on the environment. More over, pesticide use usually encourages the development of resistance to the compounds by the pest (in this case, fungal pathogens). This leads do less effectiveness of the pesticide and often to increased pesticide use by farmers, unless other chemicals are available. If so, the same cycle starts again.

top of page
©2002 | Site by machinelab | Hosting donated by Clarity Connect