Institutional reforms for getting an agricultural knowledge system to play its role in economic growth

Pakistan Development Review, Winter, 1999 by Jock R. Anderson

The study by Oldeman, Hakkeling and Sombroeck (1990), conveniently summarised in Oldeman (1992), found that, of the LDC total of 6164 million ha of land in annual and permanent crops, permanent pasture, and forest and woodland, 1548 million ha (25 percent) was degraded to some extent. Water and wind erosion were responsible for 83 percent of the degraded land (i.e., 1292 million ha), The other 17 percent was degraded by chemical processes (e.g., soil nutrient loss or salinisation) and physical processes (e.g., soil compaction caused by animal traffic). Oldeman, Hakkeling and Sombroeck (1990), as reported in Oldeman (1992), show the LDC degraded land by severity of degradation only for the land degraded by water and wind erosion. Crosson used the data of Oldeman (1992) to calculate that 4475 million, 582 million and 235 million ha are lightly, moderately and strongly eroded, respectively.

Oldeman, Hakkeling and Sombroeck (1990) did not estimate the percentage losses of soil productivity for each of their categories of degradation. Crosson estimated these losses, using the percentages of loss by degradation category from Dregne and Chou (1992), to give an average productivity loss on the 6164 million ha of land of 4.2 percent. This is the cumulative loss over 45 years from the end of World War II to about 1990 [Oldeman, Hakkeling and Sombroeck (1990)], so the average annual rate of loss is only about 0.1 percent, an estimate of productivity loss dominated by the assumption that 75 percent of the 6164 million ha of LDC land in crops, permanent pasture and forest/woodland has suffered no productivity loss because of soil degradation.

Water: Degradation of water resources considered here is confined to water used for irrigation. Irrigation water is degraded primarily by the uptake of salts from the soil as the water moves downstream and is used and reused by one farmer (or other user) after another. Waterlogging, while less important generally than saline irrigation return flows, is the other main form of degradation of water for use in irrigation. Both of these processes lead to the build-up of salts in the soil that, unless checked, will in time reduce crop yields.

Totally reliable information about current rates, extent and productivity consequences of degraded irrigation water is scarce. The problems of defining and measuring the magnitude of irrigation-related salinisation and waterlogging are considerable. Dregne and Chou (1992) concluded that 30 percent of the 145 million ha of irrigated land in dry areas (globally, not just in the LDCs) is moderately to very severely degraded by salinisation and waterlogging seemingly on the high side of reality. Alexandratos (1995, p. 138) gives an estimate close to that of Postel (1992, pp. 35-54), some 25 million ha of irrigated land according to her, saying that about 10-15 percent of the irrigated land in the LDCs is "to some extent" degraded by waterlogging and salinisation.

Using the Dregne and Chou (1992) magnitudes because they give quantitative estimates of the effects of salinisation and waterlogging on yields of irrigated land, a weighted average cumulative yield loss on this land, in dry areas of the LDCs, of 11 percent. Most of this land was brought under irrigation over the past 30 to 40 years. Taking 40 years as the period over which the loss occurred, the average annual rate of loss would be about 0.3 percent, a conservatively high estimate for the present purpose.