LARGE DAMS AND RESERVOIRS



There is an important distinction between the reservoir impounded
by a dam, regardless of the dam's purpose, and the irrigation
system that may accompany the dam.  In the Volta, Kariba, and
Kainji hydrodams, there were no irrigation plans and the health
effects are mostly associated with populations living along the
shores of the reservoirs.  In contrast, the Aswan dams in Egypt
and the Sennar dam in the Sudan are interesting because of the
irrigation plans they helped, not because of the reservoirs
themselves.  Watson (1958) noted that the reservoirs associated
with large dams in the Middle East and North Africa do not seem
to be major sources of schistosomiasis transmission.  The Aswan
and Sennar dams caused problems only because of their irrigation
potential.

Similarly, Malek (1958) noted that the Sudan had two large dams,
the Sennar dam on the Blue Nile and the Jebel Awliya dam on the
White Nile south of Khartoum, which impounds the river for about
300 miles in a long, narrow lake.  Research implicates only the
Sennar dam, associated with the famous Gezira irrigation scheme,
in local schistosomiasis transmission.  Also, the "Biomphalaria"
snails, hosts for the "S. mansoni" parasite, are present in the
Gezira irrigation area, but not in the Sennar reservoir itself. 
Interestingly, the Jebel Awliya dam played an indirect role in
spreading the "S. mansoni" infection to Gezira.  It displaced
part of a population already highly infected by this parasite and
caused them to move to the new Gezira scheme to find work (Greany
1952).

The case of the Aswan High Dam and Lake Nasser is of some
interest because of the alarms raised before the dam's
construction as well as rumors that exist today.  Miller (1978)
challenged widely published reports of early claims or fears of
widespread transmission in the lake ("e.g." Brown and Deom 1973,
Van der Schalie 1974, Farid 1975).  Miller asserted there was no
evidence of transmission of schistosomiasis in Lake Nasser.  The
few fishermen working in the lake continue to receive careful
screening to prevent transmission (WHO 1985).

Similarly, there was much alarm about the fate of the Nubian
population displaced by the Aswan High Dam.  Fortunately,
excellent studies on both the Egyptian and Sudanese Nubian
populations may allay this concern.  In Egypt, the prevalence of
"S. haematobium" infection among Nubians declined from 15.7%,
before displacement in 1964, to an adjusted figure of 7.2% in the
Kom Ombo area after resettlement (Miller "et al." 1979).  The
decline occurred because resettlement communities, located away
from the irrigation courses in the Kom Ombo plain, received piped
water and electricity.  Also, the Nubians hired local workers to
till their fields (Miller 1978).

The Sudanese Nubians were somewhat less fortunate.  Omer (1980)
reports that the Nubians, resettled from Wadi Halfa to Khashm el
Girba in the east of the Sudan, exchanged the urinary
schistosomiasis, prevalent in their own homeland, for the
intestinal form prevalent in their new settlement.  "S. mansoni"
infection rose from zero to 8.3% by 1970-71 and to 13.2% in 1972,
while the prevalence of infection with "S. haematobium" fell from
11.6 to 3.4%.

The influence of the Aswan High Dam on downstream hydrology led
to reports expecting increased transmission of both forms of
schistosomiasis.  This was said to be the result of interrupting
the annual Nile flood in the Nile Delta, the Nile's low
turbidity, abundant aquatic vegetation, irrigation canals, and
marked increases in the densities of both "Bulinus truncatus" and
"Biomphalaria alexandrina" snails (Malek 1975).  Later, El Alamy
and Cline (1977) and Abdel-Wahab "et al." (1979) reported instead
that changes in the Nile's hydrology actually resulted in major
declines in "Bulinus" snail populations.  As a result, the
prevalence of "S. haematobium" infections in the Nile Delta
declined sharply.  In a large-scale 1983 survey, Cline et al.
(1989) reported an average infection of "S. haematobium" of  5%
in selected Delta villages compared to an average of 55% in the
same villages in 1935.  For one locality studied in 1972 and
1978, Abdel-Wahab et al. (1979), found the percentage of urine
samples positive for "S. haematobium" decreased from 30 to 9%. 
Conversely, the "Biomphalaria" snails that transmit "S. mansoni,"
increased and may have benefited from the changes in Nile
hydrology.  This species continued to expand south within the
Delta with some reports of infestations in Upper Egypt.  It is
debatable whether this expansion is a result of the Aswan High
Dam or a continuation of a long-term phenomenon of unknown
origin.  The prevalence of "S. mansoni" infections increased in
the southern Delta but has probably remained stable in the
northern Delta (Cline "et al." 1989).

The two Koka dams in the Awash Valley in Ethiopia present an
analogous case of the different effects a dam has on the two
forms of schistosomiasis through the reservoir, downstream
hydrology, and irrigation.  Kloos "et al." (1988) reported that
these two dams, built in 1958 and 1964, appeared to have no
effect on transmission through their reservoirs as assessed in
1968.  The dams did, however, enable "S. mansoni" transmission to
occur in the upper Awash Valley by changing its hydrology and
introducing irrigation.  However, the dams apparently decreased
the risk of "S. haematobium" transmission in the middle Awash
Valley by reducing the supply of water to the swamps where the
"Bulinus abyssinicus" snail hosts live.  In the upper Awash
valley, as a result of the Koka dams, the water is now clear,
with abundant aquatic vegetation and a weak current flow.  These
conditions favor breeding of "Biomphalaria pfeifferi" snails that
were previously absent from the Awash valley.  Below 800 m
altitude, higher water temperatures deter this snail (Kloos
1985), so that intestinal schistosomiasis does not affect
irrigation schemes in the middle Awash valley.

The Akosombo dam in Ghana may have favored schistosomiasis
transmission through both its reservoir (Volta Lake) and through
downstream hydrology in the Volta River.  An early study by
Hilton and Tsri (1970) reported that the Akosombo dam's effect on
downstream hydrology reduced schistosomiasis transmission by
eliminating backwaters created by seasonal flooding.  However, a
subsequent study (Wen and Chu 1984) reported dramatic increases
in urinary schistosomiasis downstream.  At Bator, infection
increased from 17% in 1963 to 27.1%  in 1972 and to 74.6% in
1981.  At Mepe, the corresponding figures were 26.5, 36.4, and
88.0%.  They attribute these changes primarily to the downstream
hydrology effects of the Akosombo dam but don't specify the
mechanisms involved.

Filling Volta Lake, beginning in 1964, undoubtedly caused
schistosomiasis to worsen in central Ghana.  MacDonald predicted
this in 1955, at least 10 years earlier.  Paperna (1970) reports
that a schistosomiasis outbreak occurred on the southwestern
shore of Volta Lake in 1967.  Ewe fisher folk, who migrated to
the new lake shore from the coast, brought the infection from the
Volta delta area where it had already infected most children. 
The explosive growth of aquatic weeds in Volta Lake favors fast
snail reproduction.  Before water filled the dam, prevalence of
"S. haematobium" in the central area was 1-5%;  by 1968,
infection was up to 40%.  Later, a careful study by Scott "et
al." (1982) found prevalence averaged 72-73% at the lake's
southwestern shore during 1973-74.  Transmission was local,
affecting the indigenous Krobos almost as much as the newly-
arrived Ewes.  Prevalence of schistosomiasis was virtually absent
in communities more than 7 km from the lake that did not use lake
water (WHO 1985).

At Kainji Lake in Nigeria, evidence strongly supports the lake's
role in increasing transmission of schistosomiasis.  The
evidence, however, only applies to the period directly after
workers finished the dam.  A WHO team collected data on
prevalence of schistosomiasis from a group of people at lakeside
villages on the southwestern shore of Lake Kainji between 1970
and 1972.  Deom (1975) cites them as reporting that prevalence
increased from 64.4 to 89.1% in those age 10-14, and among 15-19-
year-olds, prevalence rose from 37.5 to 94.7%.  In comparison, at
New Bussa, settled by river people who moved away from the lake,
infection remained stable.  Among those aged 10 to 14, prevalence
declined from 54.5 to 40%, and among those aged 15 to 19, it
remained at 25.0%.  Most significantly, in children under age 5,
at the lakeside villages, the infection rate increased from 15.8%
in 1970 to 64.3% in 1972.  At the same time in New Bussa, away
from the lake, transmission decreased from 4.0 to 0.0% since no
new transmission occurred.  However, Adekolu-John and Abolarini
(1986) note that the same WHO team found much less change in
prevalence between 1970 and 1972 at two towns on the northern end
of Lake Kainji.  Unfortunately, nobody has conducted systematic
follow-up research.

It is difficult to make generalizations from the experiences of
the dams just discussed.  The lesson drawn by Waddy (1975) is
that constructing other dams will similarly spread
schistosomiasis.  In particular, he expressed concern about the
Cabora Basso dam in Mozambique, the Kossou dam in Cote d'Ivoire,
and the dams on the Nam Pong river in northeastern Thailand.  But
in later years, no reports have linked schistosomiasis to any of
these dams.  Elsewhere, in northern Thailand, a reservoir, filled
in 1963, remains free of the snail vector for human
schistosomiasis (Temcharoen 1992).  Third World countries
constructed many large dams in the last two decades. 
Interestingly, very few reports link them to schistosomiasis
transmission.

Among the few reports on recent dam projects, two from West
Africa implicated large dams in increasing transmission of
schistosomiasis.  When the Diama dam at the mouth of the Senegal
River became operational in 1986, it prevented salt water from
penetrating upriver, allowing irrigation development in the lower
Senegal River valley.  Intestinal schistosomiasis, reported by
Talla "et al." (1990), surged.  Determined from clinical data in
the town of Richard-Toll, the infection rate rose from 2% in 1988
to 72% in 1989.  Diaw "et al." (1991) indicated that population-
based estimates of the prevalence of "S. mansoni" infection rose
from 12% in 1988 to 43% in 1990.  However, research linked the
geographically narrow focus of the outbreak to a massive
irrigated sugar cane cultivation project at Richard-Toll, rather
than to any general effects of the dam.

Brinkmann "et al." (1988) reported prevalence data before and
after construction of the Selingue dam in western Mali.  The
research covered 10 villages displaced by the dam.  In four
villages that previously had no "S. haematobium", infection rates
ranged from 6 to 22%, three to four years after resettlement.  In
two other villages, the prevalence of urinary schistosomiasis
increased from roughly 2% before the dam to 22 to 42% afterward. 
The other four villages had much smaller increases.  The data for
all 10 villages suggest an 8-fold increase in "S. haematobium"
infection from an average prevalence of roughly 4% before the dam
to nearly 32% afterward.  The average prevalence of "S. mansoni,"
on the other hand, decreased slightly, from 2.2 to 1.6%.

There is concern that the new Itaipu hydroelectric dam, on the
border of Brazil and Paraguay, could result in schistosomiasis
spreading to areas in which it is still unknown (Bousfield,
1979).  Some fear the host snails will colonize the lake
impounded by the dam, and infected migrants from elsewhere in
Brazil will result in local transmission and a severe public
health problem.  Little documentary evidence exists on water
resources development projects in Brazil and the spread of
schistosomiasis.  Water development projects in the San Fernando
valley possibly resulted in new endemic foci of intestinal
schistosomiasis (Malek and Chaine 1989).  No studies, however,
tested this hypothesis.  Reviews agree that the spread of
schistosomiasis during this century in Brazil resulted from
infected migrants from the northeastern part to the country
transmitting the infection in areas where the host snails already
exist (Buttner 1958, Almeida Machado 1979, Kvale 1981). 
Therefore, it is not clear whether environmental change,
including large dams or irrigation systems, spread
schistosomiasis in Brazil.

The Amazon jungle, shared by Brazil and its northern neighbors,
of which parts of Venezuela and Suriname are endemic for
schistosomiasis, appears to remain free of schistosomiasis.  In
Suriname, the Brokopondo dam, completed in 1964, did not cause
snail-breeding because of the acidity of the river water (Brown
and Deom 1973, Leentvaar 1975). " Biomphalaria glabrata" snails
exist only in soil with lime.  In most of the Amazon, the soil
and water are acidic and therefore, do not support the snails. 
Deforestation in the Amazon, however, could change this. 
Similarly, in Africa, schistosomiasis is also mostly absent in
dense rain forests (Wright 1973).  Kloos suggests (1988) that
deforestation in western Ethiopia, combined with resettling
infected migrants, could result in a new wave of schistosomiasis.