LARGE-SCALE IRRIGATION SYSTEMS



The two clearest examples of irrigation systems spreading
schistosomiasis come from the Nile Valley.  First, in the Nile
Valley south of Cairo in Upper Egypt, researchers have known for
a long time that displacing basin irrigation (by the flood waters
of the Nile) with perennial irrigation results in a dramatic
increase in schistosomiasis.  Reports verified (Khalil and Azim
1938, Khalil 1949) that four locations changing irrigation
methods in the mid-1930s, following Aswan Low Dam construction,
experienced a surge in "S. haematobium" infections.  Infection
rates went up from 2 to 11% in 1934 to 44 to 75% in 1937. 
Similarly, cross-sectional data collected by Scott (1937) shows
urinary schistosomiasis prevalence at 60% in areas of perennial
irrigation in Upper Egypt and only 5% in areas with basin
irrigation.  Later, schistosomiasis increased in Sohag
governorate from 3% in 1937 to 42% in 1955 with the shift from
basin to perennial irrigation.  In three districts with perennial
irrigation, prevalence ranged from 56 to 68%, while in two
districts with basin irrigation, prevalence was 9 and 13% (Wright
1973).  Basin irrigation allows the land to dry out seasonally,
killing almost all snails.  Under perennial irrigation, the land
is wet year round.

The major risk of the Aswan High Dam for increasing
schistosomiasis transmission was in areas that switched from
basin to perennial irrigation or areas being cultivated for the
first time because irrigation water from the dam made it
possible.  Farooq (1973) estimated that these changes would
affect 5 million people. Seventy percent of these people would
become infected, adding 2.7 million new cases to 14 million
existing cases in Egypt.  However, Miller claims that by 1968,
almost all of Upper Egypt was already under perennial irrigation.

Therefore, the Aswan High Dam could not have a major additional
adverse effect on schistosomiasis (Miller "et al." 1979). 
However, Mobarak (1982) reports that after 1970, prevalence of
urinary schistosomiasis in Upper Egypt reversed a previous
downward trend and began to increase again because of the shift
from basin to perennial irrigation.  In either case, increased
use of chemotherapy is bringing schistosomiasis under control in
Egypt.  In the northern part of Upper Egypt, prevalence dropped
from 29.4% in 1977 to 11.5% in 1983, while further south,
prevalence reportedly fell from 26.4% in 1980 to 16% in 1983 (WHO
1985).

The other Nile valley large-scale irrigation system implicated in
schistosomiasis transmission comes from the Gezira irrigation
scheme in the Sudan.  In this area irrigated by the Sennar Dam
(constructed in 1924 but extended after 1950), "S. haematobium"
prevalence increased from under 1% before 1940 to 5% in adults
and 15% in children in 1950 (Greany 1952).  In the late 1950's,
schistosomiasis seemed temporarily under control in the Gezira.
Wright (1973) cites an unpublished 1961 report by Farooq stating
that among 7-year-old entrants to 19 Gezira schools, prevalence
of "S. haematobium" fell from 28.3% in 1957 to 3.3% in 1960.  "S.
mansoni" infections reportedly held steady or declined during the
1950s.

The major increase in prevalence of schistosomiasis in the Gezira
came after 1960.  One factor, instituted by the Gezira scheme,
was the cropping rotation change to include "winter" wheat. 
Farmers kept the canals filled with water from March to May, when
they were previously dry (Fenwick 1989). Another factor was the
creation of the adjoining Managil extension which left tenants
without adequate water supplies or sanitation facilities.  Also,
there was an influx of migrant laborers in the original Gezira
scheme, who lived near irrigation canals, under very poor
sanitary conditions.  Tameim "et al." (1985) reported low
prevalence of "S. haematobium" but 50 to 60% prevalence of "S.
mansoni" infection during 1981-82, in six sentinel villages in
the Gezira and Managil schemes.  Fenwick states (1989) that
prevalence of "S. mansoni" was much higher in Managil villages
than in Gezira villages.  He found 65% in a representative
Managil village versus 31% in a typical Gezira village. 

Most experts agree that applying the proper combination of
sanitary engineering, water control management, snail control,
infection surveillance, and treatment drugs, can avert
irrigation's effect on schistosomiasis.  Even without water
control measures and environmental sanitation, chemotherapy with
or without treating water to kill snails may adequately control
schistosomiasis transmission.  Fenwick (1989) notes that in the
newer Rahad irrigation scheme, east of the Gezira plan in the
Sudan, because of the use of drugs and snail control,
schistosomiasis remained very low, despite very poor sanitary
conditions.  In 1985, prevalence of "S. mansoni" in 12,801 Rahad
school children was 9.0% and "S. haematobium" was only 2.3%. 
Relaxing control measures, however, would cause a predicted surge
in schistosomiasis, presumably reaching levels similar to those
in Managil.

In the Middle East and North Africa, research has associated even
small-scale irrigation plans, decentralized at the village level,
with increases in schistosomiasis transmission.  Malek (1962)
reports that, in Sudanese villages along the Nile north of
Khartoum, prevalence of urinary schistosomiasis in children was
11-12%, compared to an average of less than 2% in other areas. 
Wright (1973) reports that in the rural area around Baghdad,
Iraq, prevalence of schistosomiasis increased from 10 to 25%
because of the installation of lift pumps.

However, in Iraq, new large-scale irrigation projects never
fulfilled the prophecies of dramatic increases in transmission of
schistosomiasis.  Abdel Azim and Gismann (1956) suggested that
Iraq in the mid-1950s was at the same point as Egypt earlier in
the century.  They reported that large-scale irrigation projects
would trigger a massive spread of schistosomiasis as happened in
Egypt.  Later, Wright (1973: 157) also referred to "vast new
irrigation systems which will inevitably lead to spread of
infection by "S. haematobium" unless exceptionally vigorous
measures of prevention are undertaken."  However, the only
evidence offered is the case of the Musayeb irrigation project. 
Settlers moved there in 1956, and, by the end of 1958, 28% of the
13,000 settlers screened had the disease.

McMullen "et al." (1962), who also cite this case, note that one
could not rule out the possibility that settlers imported the
infection since prevalence was the same or higher throughout
southern Iraq. Yacoub and Southgate (1987) reported that, in
Southern Iraq, mass chemotherapy reduced the prevalence of
urinary schistosomiasis to very low levels, where the highest
endemic areas now average less than 10% prevalence.  These
authors did not address the question of the role irrigation
systems might have played in the changing prevalence of
schistosomiasis in Iraq.

In West Africa, the French colonialists established the "Office
du Niger" system in Mali between 1934 and 1948 to irrigate cotton
and rice.  Deschiens (1970) reported that the irrigation dams at
Sansanding and Markala, near Segou, resulted in a marked increase
in urinary schistosomiasis, even during the colonial period. 
Later, both urinary and intestinal forms of schistosomiasis
became highly endemic in the Office du Niger area.  Brinkmann "et
al." (1988) compared areas in the Office du Niger plan to fishing
and farming communities along the Niger river and the internal
delta.  The irrigation scheme had a mean prevalence of "S.
haematobium" of 64.4% compared to the river communities' 19.9%. 
The scheme had a mean prevalence of "S. mansoni" of 53.9%
compared to the river communities' 1.9%.

However, Wright (1973) reports that upstream, along the Niger
river in western Mali, the prevalence of "S. haematobium" was
about 75% in the area of Kayes, and downstream, in the Niger
Republic, prevalence averaged 45% in river communities. 
Therefore, it is not clear how much the Office du Niger project
increased transmission of urinary schistosomiasis, although the
worsening of intestinal schistosomiasis is unmistakable.

In southern Africa, Pitchford (1958b) warned that large-scale
irrigation plans in the eastern lowveld part of Swaziland would
increase transmission of schistosomiasis.  Wright (1973) cites
unpublished information reporting that this prediction came true.

Specifically, in the Mhlume irrigation scheme, prevalence of "S.
haematobium" increased from 25% in 1959 to 56% in 1962.  In the
Ngonini scheme, the mean prevalence rose from 34% in 1958-59 to
73% in 1963.  In the Big Bend scheme, prevalence rose from 15% in
1958 to 50% in 1962.

Elsewhere in the region, evidence on the effect of large-scale
irrigation systems on schistosomiasis is lacking.  In
southeastern Zimbabwe, it is claimed that a 30,000-hectare
irrigation scheme, developed between 1963 and 1971, worsened
schistosomiasis, but nobody collected data until control measures
took effect (Shiff "et al." 1973).  This irrigation plan was much
larger than contemporary projects in Kenya, Tanzania, or the
Eastern and Southern Africa regions.  Only the Awash Valley
schemes in Ethiopia, totaling 60,000 hectares as of 1986 (Kloos
"et al." 1988), appear to match it.  Nobody has collectively
evaluated the Awash Valley plans in any systematic fashion.  Data
are only available for the Wonji sugar estate (see below).

Andreano argued (1976) that perennial irrigation and double-
cropping can intensify exposure to schistosomiasis through much
larger irrigation distribution networks and year-round
transmission.  After 1958, large-scale gravity-fed irrigation
projects expanded dramatically in China.  He suggests that the
prevalence of schistosomiasis probably increased as a result. 
The number of cases declined by more than 90% between the mid-
1950s and mid-1980s (Chen 1989), contradicting this prediction. 
Wiemer (1984) offers evidence consistent with a temporary
increase in prevalence.  In Kunshan county, Jiangsu province,
prevalence was 55% in 1957.  With a national thrust at control,
the prevalence rate fell to 22-25% during 1959-61.  But
prevalence rebounded to 35% in 1963 and exploded to 62% in 1964,
and 80% in 1965.  A second national control program reduced
prevalence to 3% by 1977.