SB Brush and E Meng

Department of Human and Community Development, University of California, Davis, California, USA

Although attempts to formally assign a value to crop genetic resources have been a relatively recent phenomenon, the long-term use and conservation of these resources by farmers, as well as by scientists and other interested parties, provide evidence that value was attributed to them long before a formal valuation process began. Until recently, proponents of genetic resource conservation have been able to provide only anecdotal and rather vague estimates of the value of genetic resources. There are abundant examples of the economic contribution of exotic crops or crop varieties to societies around the world, but these examples do not furnish estimates of the value of crop resources that remain in farmers' fields, in forests and pastures, or in gene banks. The burden of being more specific about the value of genetic resources comes from two directions. Crop resource conservationists need measures of value to justify budgets. Farmers' rights activists want measures of the value of crop resources to back up their attempts to obtain compensation for farmers or for less developed countries. These reasons for specifying the value of crop genetic resources are given legitimacy in the 1992 Convention on Biological Diversity Different methods for valuing non-market resources exist, but it is questionable whether a single method is available to value the vast array of genetic resources.

Estimation of value suffers from lack of data and, in some cases, lack of methodology We assert that genetic resources for agriculture present a unique and separate set of problems and characteristics that must be acknowledged in any valuation exercise. Our approach is to examine the use value of landraces to farmers, as a means of assessing the likelihood of genetic resource loss in landraces and to explore the possibility of in situ conservation of genetic resources by farmers in the agricultural systems where landraces are maintained. We view landraces as a proxy for crop genetic resources because land-races have historically been an important source of germplasm in crop improvement programmes. In the collections of the CGIAR centres, 59% of all

© CAB INTERNATIONAL 1998. Agricultural Values of Plant Genetic Resources

(eds R.E. Evenson, D. Gollin and V. Santaniello) 97

accessions are landraces, 14% wild and weedy relatives, and 27% advanced cultivars and breeders' lines, where the last category derives largely from the first (FAO, 1996).

As a source of raw materials for the crop development process, landraces are a public good with social value. However, to the farmers who choose to cultivate them, landraces are also private goods. Farm households benefit directly from their production and consumption, and dedicate valuable resources to make their cultivation possible. Because private value affects the supply of genetic resources, it also affects their social value. Therefore, consideration of the social value of genetic resources must address the value of these resources to farmers. An attempt to examine the value of a landrace to the farm household should be preceded by the identification of what exactly farmers perceive to be advantageous and therefore valuable about the traditional varieties they choose to cultivate. These characteristics presumably will be the same as, or similar to, factors ultimately influencing the farm household in its land allocation decisions.

The objective of this chapter is to propose a practical alternative for valuation by focusing on the private value of landraces in centres of agricultural diversity

The Loss and Conservation of Crop Genetic Resources

The definition of crop genetic resources has grown increasingly broad and now includes germplasm from both cultivated and wild plants, as well as the ecosystems that contain the germplasm. The germplasm resources of agriculture are found in wild relatives of cultivated plants, weedy forms, locally selected crop varieties (landraces), modern cultivars, and the breeding lines used in crop improvement. The first three have been targeted for conservation due to both their threatened status and their contribution to the development of the latter two types. Moreover, these three sources of crop germplasm are the basis for demands for compensation (Mooney, 1979).

Landraces were the initial target for conservation, and hundreds of thousands of individual accessions have been assembled into national and international collections, stored in gene banks, botanical gardens, and breeding programmes around the world (ex situ conservation). Gene bank collections exist in 130 countries, with approximately 6.2 million accessions, including over 2 million cereal accessions (FAO, 1996). Half (50.4%) of the accessions are located in developed countries, more than one-third (38%) are in developing countries, and a smaller fraction (11.6%) are in international centres of the CGIAR (FAO, 1996).

As conservation of crop genetic resources has progressed, conservationists have concluded that ex situ maintenance alone is not sufficient and that in situ conservation is needed to complement gene banks and botanical gardens. In situ conservation serves not only as a back-up, but it also covers types of genetic resource which cannot be protected in gene banks, such as local knowledge and ecosystem interactions (Oldfield and Alcorn, 1987; Brush and Stabinsky, 1996). Ex situ conservation simply cannot capture or conserve all of the diversity in agricultural systems. In situ conservation does not directly provide genes for crop improvement, but it preserves evolutionary processes which will yield new germplasm in the future (Brush, 1995).

For nearly as long as the crop breeding industry has existed, scientists have worried that the ultimate source of crop germplasm in centres of crop origins and evolution is threatened by modern conditions - especially the size and growth of the human population, technological change and markets (Frankel, 1970). While threat of genetic erosion came into sharp focus in technical conferences of the FAO in the 1970s, scientific understanding and measurement of this problem is weak. The lack of a firm estimate for genetic erosion and uncertainty about the supply of genetic resources has a potentially large impact on our ability to measure the value of these resources. Diffusion of modern crop varieties (MVs) is the process most widely used to indicate genetic erosion, but this is an indirect measure. The experience in developed countries is that diffusion of MVs can have a devastating effect on the diversity of local crops (Griliches, 1958; Duvick, 1984). While the spread of semi-dwarf wheat and rice in Asia convinced many agricultural scientists that genetic erosion experienced in temperate industrial countries was about to repeat itself in tropical, less developed ones, some evidence suggests that a different pattern may prevail in centres of crop diversity.

Genetic erosion in less developed countries was expected to follow two of the trends observed in developed countries. First, it was believed that the diffusion of MVs would be rapid and thorough. Second, MVs were expected to completely replace landraces. This extrapolation from developed to less developed countries was supported by aggregate data (e.g. Dalrymple, 1986), but more detailed studies suggest that the two trends may not apply broadly in diverse, less developed countries (Brush, 1995). The diffusion of MVs in the latter is complex, with a rate and extent that is context specific. In many situations, farmers decide to plant both MVs and landraces on their farms. Regions thus exist in several countries where landraces either persist as the major type of cultivar or where landraces have been decreased but not eliminated by MVs.

From the observed cultivation of landraces, we can infer that these varieties hold some value for the households that cultivate them, and this household value is one component of the overall value of the genetic resource. Estimating their value to farm households involved in their cultivation can provide a lower-bound estimate of the total value of these resources to society. The change in focus from private, on-farm value for farm households to broader measures of value for society is not necessarily straightforward. In the context of valuing biodiversity for pharmaceutical research, Simpson et al. (1996) point out that previously estimated values have probably been overestimated due to the failure to account for issues of scarcity and redundancy. They emphasize that an accurate assessment will value a marginal species on the basis of its incremental contribution. A similar precaution should be applied to any valuation of land-races.

Additional information about specific varietal characteristics that the farmer finds important will provide insights into household preferences and behaviour. Ideally, the availability of the appropriate information of preferences, both in terms of crop varieties as well as production and consumption characteristics, could facilitate the estimation of household value of landraces. Improved knowledge of preferences and behaviour will also assist in assessing the likelihood that farm households will continue to include traditional varieties in their production decisions. If continued cultivation of landraces appears uncertain and if on-farm maintenance of a specific target level of diversity is a stated policy objective, then the accurate valuation of these genetic resources from the viewpoint of the farm household could provide crucial information for the development of policies to guarantee their existence in the future.

Farmer's Valuations of Crop Resources in Turkish Wheat

Several theories of behaviour have been developed to explain the varietal choice of farm households. The desire to diversify and shield against risk (Just and Zilberman, 1983; Clawson, 1985; Finkelshtain and Chalfant, 1991; Fafchamps, 1992), transactions costs restricting households access to markets (Strauss, 1986; Falchamps, 1992; Goetz, 1992), and the presence of environmental constraints (Richards, 1986; Bellon and Taylor, 1993; Zimmerer, 1996) have all been advanced as important influences in the land-use decisions of farm households. Each of these theories has been empirically tested separately, and alone can explain to some extent the occurrence of the partial adoption of improved crop varieties by farmers in developing countries. Factors such as production stability, joint production, performance in different agroclimatic zones and fit to household schedules may lead to the selection of landraces instead of or in addition to MVs. The presence of consumption risk and the demand for specific quality characteristics can also result in a household decision to cultivate a traditional variety if transportation costs, search costs or other transaction costs prevent household access to markets. Field research on the maintenance of landraces in regions where MVs are grown confirms the multiple objectives of farmers and the importance of heterogeneity in the physical, economic and cultural contexts of local agriculture (Brush, 1992).

As a centre of domestication and diversity for wheat (Zohary and Hopf, 1993), the cultivation of wheat in Turkey for over 8000 years has resulted in a large number of traditionally grown wheat varieties in addition to the existing wild and semi-domesticated wheat relatives. Modern (dwarf) varieties have been available in Turkey since the early part of the century and semi-dwarf varieties were introduced from Mexico in 1966. However, the level of adoption in the country varies greatly from region to region. The spring wheat regions of Thrace and the largely winter wheat areas of the Anatolian plateau are exten-

sively planted in MVs, but CIMMYT (1993) reports that only 31% of Turkey's wheat area was planted in MVs in 1990. The western transitional zone, located between the major wheat-producing Anatolian Plateau region and the western coastal plains, was selected for this research due to its considerable variation from village to village in percentage of area cultivated with MVs. Our surveys show that the adoption of modern wheat varieties varies within as well as between provinces.

Our data were collected in 1992 from a household-level survey covering 287 households, in 24 villages selected from the three provinces of Eskisehir, Kutahya and Usak. Villages also vary with respect to agroclimatic zones, which can be divided into three categories: valley land, hillside land and mountain land. Valley land is more likely to have irrigation and to be connected with urban markets, while mountain land is most distant from markets and situated in or around forested zones. Hillside land shares attributes of both valley and mountain land. The socioeconomic survey covers a broad range of information regarding household characteristics, detailed production data and consumption preferences. Households in the sample include those that cultivate only MVs, those that cultivate both modern and traditional varieties, and those that cultivate only traditional varieties. Differentiation between households also exists with respect to percentage of production output marketed. Table 6.1 summarizes several household characteristics by province and agroclimatic zone for the surveyed sample.

In our Turkish research, we find evidence that environmental heterogeneity, risk and the high household transaction costs of obtaining desired qualities in wheat contributed to the continued cultivation of landraces in specific areas of a nation which has successfully promoted MVs. An econometric analysis of the factors influencing plot-level households varietal selection decisions in the Turkish data confirms the importance of acknowledging multiple household motivations for cultivating traditional varieties (Meng et al., 1995). The probability of cultivation of traditional varieties in a given household plot significantly increases, for example, when the plot is situated in less fertile soil or when distance to market and bad road quality increase a household's cost of accessing markets.

A household's ability to access markets plays a significant role in its decision to cultivate traditional varieties. Table 6.2 shows the relationship between the percentage of household wheat sales and selected household characteristics. Households with little or no wheat sales are often located furthest from markets and experience the least amount of interaction with extension agents. Moreover, these households are likely to be the smallest landowners, both in terms of overall area and irrigated area.

Our survey also reveals that a substantial percentage of households participate in some kind of market, but that market integration and efficiency are not fully represented by road quality and distance to market. Once the market is accessed in Turkey, the sale of any wheat variety is guaranteed since it is the stated policy of the government state purchasing authority, the Toprak Mahsulleri Ofisi (TMO), to accept all varieties. The purchase of a specific variety, o rs)

Table 6.1. Household characteristics.

Distance

Table 6.1. Household characteristics.

Distance

n

Off-farm income (%)

Age

School

Household size

Total land (ha)

Total irrigated land

Total plots

Mean plot size

No. of household wheat plots

from market (km)

All households

285

26.5

49.1

4.1

4.5

12.3

2.6

12.4

13.1

7.0

16.6

Eskisehir

96

21.9

48.2

4.6

4.4

15.5

6.7

13.2

18.0

7.2

20.4

Kutahya

96

22.7

49.7

3.8

4.6

10.4

0.9

20.3

7.1

10.8

12.9

Usak

93

35.1

49.5

3.9

4.3

10.8

0.1

8.5

14.3

2.8

16.6

Valley 100 17.0 50.1 4.4 4.7 15.0 6.7 15.8 16.9 7.3 10.6

Hillside 101 22.5 49.1 4.3 4.3 13.9 0.6 16.2 12.1 8.4 13.5

Mountain 84 42.4 48.0 3.4 4.4 6.9 0.02 9.5 9.9 4.8 24.0

Table 6.2. Household characteristics by percentage production sold.

Total

Total

Percentage of land in:

Total

wheat

irrigated

Home bread

Market

Extension

Off-farm

Household

No. of

land

land

land

Modern

Traditional

(%

distance

visit

income

sales

households

(ha)

(ha)

(ha)

varieties

varieties

households)

(km)

(%)

(%)

0

91

7.4

2.7

2.1

9.1

90.9

98.9

18.1

19.8

36.3

(10.1)*

(2.1)

(1.6)

(7.3)

0<S<25

34

10.8

5.4

4.3

13.2

86.8

97.1

18.4

23.5

35.3

(10.7)

(4.3)

(4.4)

(7.5)

26<S<50

49

13.4

8.3

5.8

34.6

65.4

87.0

15.9

36.7

12.2

(13.3)

(7.9)

(6.6)

(7.8)

51<S<75

51

13.9

10.4

5.5

43.8

56.2

74.0

15.7

19.6

31.4

(10.0)

(13.0)

(5.8)

(8.4)

76<S<99

60

17.4

11.9

4.2

63.0

37.0

80.7

14.8

43.3

11.7

(16.2)

(15.5)

(5.6)

(8.2)

100

4

14.7

3.5

0.8

75.0

25.0

75.0

12.8

50.0

25.0

(10.9)

(4.4)

(1.2)

(9.9)

*Standard deviation.

Table 6.3. Location of sales by variety type.

No. of

Total

Percent of

Percent of

households

No. of

quantity

sales of

sales of

Location sold (%)

cultivating

households

sold

total wheat

total variety

Variety type

variety type

with sales

(kg)

production

production

Local

Merchant

Not local

Modern

136

108

2,899,478

76.3

74.3

1.5

24.4

73.3

Traditional

221

127

901,955

23.7

46.7

5.8

65.3

28.0

Table 6.4. Seed acquisition.

Activity

Yes (%)

No (%)

Purchase seed

35.4

64.6

Borrow seed

11.2

88.8

Purchase government seed

7.7

92.3

Purchase government seed

42.6

57.4

(previous years)

Sell seed

15.8

84.2

Sell seed (previous years)

31.9

68.1

Location purchased

Location purchased

(previous years)

No.

Percent

No.

Percent

Village 45

31

20

13.8

Merchant 21

14.5

10

6.9

Neighbour 16

11

8

5.5

Cooperative 12

8.3

6

4.1

Government 5

3.4

5

3.4

No purchase 46

31.7

64

44.1

Total responses 145

100

113

100

however, is much more difficult. Private and public wheat buyers categorize the grain by colour and hardness, a policy that results in a mixture of varieties in each market class. Quality aspects that might be attributed to specific varieties are not included in the market classes, so that search costs and other transaction costs for obtaining a specific local variety are likely to be high. Table 6.3 gives a breakdown of sales of modern and traditional varieties by location sold and shows that sales of traditional varieties make up only a small percentage of wheat sales. Half of the production in traditional varieties is sold in contrast with the high percentage of sales in modern wheat varieties. Local sales at the village level amount to only 9% of traditional variety sales, indicating that these varieties are primarily used to meet consumption needs. A slightly higher percentage (12.4%) of wheat sales to the TMO take place outside the village. While merchant sales take place in or near the village, the percentage of this wheat remaining in the local community is unknown. In conclusion, it appears that if the household wishes to consume a specific variety, on-farm cultivation appears to be the most certain method of guaranteeing availability.

Similarly, a secure supply of seed of traditional varieties for a particular household can best be obtained through on-farm cultivation. A market for traditional variety seed is hard to detect in our survey. Table 6.4 shows that just over a third (35.4%) of the households purchased any wheat seed and even fewer (11.2%) borrowed seed. The bulk of seed that is sold and bought locally is seed of MVs.

Table 6.5. Variety attributes by household type.

Attribute*

Table 6.5. Variety attributes by household type.

Attribute*

Drought

Disease

Baking

Household type

n

Yield

resistance

resistance

quality

Taste

All

352

2.13

2.24

1.86

1.93

1.86

(0.87)

(1.07)

(0.87)

(0.87)

(0.78)

Modern varieties

133

1.73

2.72

1.97

2.22

2.15

(0.67)

(1.02)

(0.81)

(0.92)

(0.81)

Traditional varieties

219

2.37

1.95

1.79

1.75

1.69

(0.89)

(1.00)

(0.90)

(0.79)

(0.72)

Traditional varieties are often attributed with more attractive characteristics than MVs. An examination of household rankings of specific attributes on a scale from 1 (best quality) to 5 (worst quality) from the Turkish survey data sheds light on the association of these characteristics with individual varieties. Households ranked each variety cultivated with respect to taste, bread quality, milling quality, yield, disease resistance and drought resistance. These characteristics reflect important considerations for the household on both the production and consumption side. Table 6.5 presents the average scores among all households of five characteristics for both traditional and modern varieties. In general, yield attributes are ranked higher for MVs while traditional varieties are ranked higher in terms of taste and baking quality. Traditional varieties also appear to be associated with better drought resistance.

Unfortunately, information is not available from individual households ranking the characteristics of the available varieties against each other. However, particularly for households that cultivate only one variety, it is reasonable to assume that the qualities most characteristic of the chosen variety are the ones that the household considers the most important relative to other varietal traits. Data on varietal rankings from households that did not grow the variety are likewise unavailable. Nevertheless, we are able to examine the rankings assigned to modern and traditional varieties by the subsample of 71 households cultivating both types of varieties. As shown in Table 6.6, households that grow both modern and traditional wheat varieties exhibit no marked differences relative to other households in the sample with respect to variables such as age and education of household head, number of household members or total land in wheat. The number of household plots, distance to market and percentage of households baking bread at home for these households, however, fall outside the range found for other household types. The attribute rankings for the households cultivating both modern and traditional varieties that are presented in Table 6.7 exhibit a pattern of rankings similar to that of the other households in the sample.

Additional information on the relative importance of varietal attributes comes from the responses given by households regarding their reasons for discarding a previously cultivated variety. Table 6.8 gives similar information for

Table 6.6. Selected household characteristics by household types.

Household type

Total

Total

Total

No. of

Home bread

Market

(by type of variety

land

wheat

irrigated

household

(%

Livestock

distance

No. in

No. of

Educatioi

cultivated)

n

(ha)

land

land

plots

households)

(no.)

(km)

household

varieties

Age

(years)

Both varieties

71

15.7

9.6

5.9

23.8

75.7

3.8

14.6

5.9

2.3

48.6

4.6

(16.2)

(11.3)

(6.7)

(21.3)

(43.2)

(2.8)

(7.7)

(2.2)

(0.9)

(11.2)

(1.7)

All households

285

12.1

7.3

4.0

16.5

88.5

3.7

16.6

5.9

1.6

49.1

4.1

(12.7)

(10.4)

(5.0)

(16.6)

(32.0)

(3.7)

(7.9)

(2.6)

(0.8)

(11.9)

(2.3)

Only modern

63

15.6

12.4

3.6

16.0

84.7

4.7

17.7

5.8

1.7

47.1

3.9

varieties

(14.6)

(15.9)

(5.5)

(17.5)

(36.3)

(6.3)

(7.7)

(2.8)

(0.9)

(11.9)

(2.2)

Only traditional 147 8.9 4.0 3.1 13.2 95.9 3.3 16.9 6.0 1.2 50.1 3.6

varieties (8.6) (3.8) (3.1) (12.3) (20.0) (2.2) (7.7) (2.7) (0.4) (12.2) (2.4)

Table 6.7. Attribute rankings for households growing both variety types.

Attribute

Table 6.7. Attribute rankings for households growing both variety types.

Attribute

Drought

Disease

Baking

Variety type

n

Yield

resistance

resistance

quality

Taste

Modern varieties

71

1.65

2.78

2.17

2.35

2.29

(0.61)

(0.89)

(0.88)

(0.96)

(0.88)

Traditional varieties

71

2.71

1.71

1.74

1.96

1.76

(0.85)

(0.82)

(0.93)

(0.89)

(0.78)

Table 6.8. Reasons given for ending cultivation of traditional varieties.

All households responding Households cultivating MVs (n = 163/285) (n = 88/118)

Reason No. Percent No. Percent

Yield

124

38.8

78

38.8

Cold susceptibility

38

11.9

20

10.0

Production-related

27

8.4

16

8.0

Quality problems

26

8.1

6

3.0

Drought susceptibility

24

7.5

20

10.0

Lodging

23

7.2

21

10.4

Other

18

5.6

14

7.0

Seed availability

16

5.0

10

5.0

Marketing problems

12

3.8

9

4.5

Disease susceptibility

5

1.6

1

0.5

Price

4

1.3

4

2.0

Climate adaptability

2

0.6

1

0.5

Soil adaptability

1

0.3

1

0.5

Total responses

320

100

201

100

all households that have given up the cultivation of traditional varieties and also for those households that continue to grow MVs and have recently discarded a traditional variety. Yield is the most frequent reason given for their decisions. Quality-related reasons decrease sharply between the sample of all responding households and the subsample of households that currently cultivate MVs. These findings suggest that quality issues are no longer of great importance for those households that have given up traditional varieties. Table 6.9 presents the reasons provided for opting against the cultivation of MVs. Two sets of responses are presented: firstly, for all households responding to the question, and secondly, for those households that continue to cultivate traditional varieties and have given up the cultivation of a MV. Again, yield remains the most important reason in both cases, but different reasons may pertain here than for those households that listed yield as the primary reason for giving up a traditional

Table 6.9. Reasons given for ending cultivation of modern varieties.

Households cultivating All households responding traditional varieties

Table 6.9. Reasons given for ending cultivation of modern varieties.

Households cultivating All households responding traditional varieties

Reason

No.

Percent

No.

Percent

Yield

43

25.4

22

25.9

Drought susceptibility

26

15.4

19

22.4

Production-related

20

11.8

7

8.2

Quality problems

14

8.3

10

11.8

Disease susceptibility

13

7.7

1

1.2

Seed availability

12

7.1

3

3.5

Cold susceptibility

10

5.9

10

11.8

Price

10

5.9

3

3.5

Other

9

5.3

4

4.7

Lodging

4

2.4

1

1.2

Climate adaptability

4

2.4

4

4.7

Soil adaptability

4

2.4

1

1.2

Marketing problems

0

0.0

0

0.0

Total responses

169

100

85

100

variety. For MVs, it is more likely that the yield response reflects a failure to attain expected yield. In addition. the percentage of households listing quality and drought resistance as reasons for giving up MVs increases for the subsample of households that continue to grow traditional varieties. These results further suggest that influence of both consumption demand and environmental constraints in variety choice and the advantages of consumption quality and adaptability to climate hardships associated with landraces.

An understanding of farm households' perspectives about landraces is relevant for the effective design, implementation and cost estimation of in situ conservation. Information about varietal selection is useful in identifying households that value landraces the most, those likely to carry on de facto conservation, and the ones least costly to incorporate into conservation programmes. Socio-economic surveys can link the probability of cultivating certain varieties or types (e.g. traditional or modern) with the household costs of production and the profitability of different varieties. In combination with genetic assessment, these surveys provide a way of determining the most cost-effective way to conserve diversity in an agro-ecosystem. Given the difficulty of estimating the total benefit of landrace conservation to society, this cost-side approach provides a method for estimating the cost of in situ conservation of landraces. The costs of such a conservation programme depend largely on the scope of the programme - the number of landraces targeted, the area needed to conserve these, and the number of households deemed necessary to maintain these in cultivation. While cost estimation does not necessarily require specific knowledge of how households value landraces, this knowledge can be helpful in reducing costs.

Was this article helpful?

0 0

Post a comment