Small Dams


Throughout the semi-arid areas of Africa, people have constructed
many small earthen dams to provide irrigation water for dry
season cultivation.  While reports often blame these dams for
spreading schistosomiasis ("e.g." Masaba 1983), there is little
evidence to substantiate such claims.  Two important exceptions
exist in Ghana and Mali.

In East Africa, Webbe (1962) noted that in northern Tanzania
researchers found snail vectors for schistosomiasis in seasonal
pools of water.  They did not find them in the year-round
reservoirs created by small dams that provided water for
livestock during the dry season.  Taylor and Makura (1985) report
that 10,000 small dams recently constructed in Zimbabwe did not
increase schistosomiasis prevalence.  They compared prevalence
estimates from each region of the country to estimates from the
1960s.  The 1981-82 national survey could not substantiate the
claim by Clarke (1977) and Shiff "et al." (1973) that connected
schistosomiasis increases to these and other water resources
development projects.

Small dams in the semiarid zone of West Africa get the greatest
amount of attention.  In northern Nigeria, studies show that the
dominant snail species change when people build dams (Tayo and
Jewsbury 1978, Olofin 1982, Betterton 1984).  However, this does
not necessarily change transmission patterns because transmission
can occur mainly through a vector species found only in other
water sources. Betterton "et al." (1988) report that two
different villages near Kano experienced transmission of
schistosomiasis not from newly-constructed dams, but from rain-
fed pools used by most children to bathe. Research suggests that
"Bulinus senegalensis," which inhabits seasonal pools, may be the
primary host for "S. haematobium" in this area. Therefore,
although the two dams built for dry-season irrigation extended
the range of "Bulinus rohlfsi" snails, a common vector for "S.
haematobium" in West Africa, this did not cause any noticeable
increases in prevalence of urinary schistosomiasis.

A well-publicized case in international scientific literature is
the small earthen dam at Ruwan Sanyi in Malumfashi, Kano State,
northern Nigeria.  Pugh and Gilles (1978) initially predicted
that this dam, built in 1977, would increase transmission of
schistosomiasis.  They said this would spread the disease in the
region since a World Bank-funded agricultural development project
would build 80 similar dams. However, Pugh "et al." (1980)
reported the 55.8% infection rate in school boys in 1979 was no
higher than it was in 1971 (59.3%) before the Sahelian drought of
1972-73.  But the rates were higher than in 1975 (33.8%) and 1976
(41.2%), before dam construction.  They did not speculate whether
it was the dam or a return to more normal rainfall that caused
the change.

Less cautiously, Gilles "et al." (1983) report in their project
summary, "Although a rise in prevalence was  already occurring
two years after the drought (presumably because the dry season
pools of water were now once again available to the community) a
further and much sharper rise occurred in 1979."  Actually, the
annual increment between 1976 and 1979 was only 4.9%, lower than
the 7.4% increase between 1975 and 1976.  Despite the
inconclusive nature of this evidence, others use the articles by
Gilles and colleagues to support claims that small earthen dams
generally spread schistosomiasis.

Several studies in the same Sudano-Sahel ecological zone as
northern Nigeria noted that evidence linking small earthen dams
to schistosomiasis was lacking.  In Burkina Faso, a comparison of
three villages near Kaya showed schistosomiasis infection was
lowest in a village with rice irrigation since 1967-68.  The rate
was almost as low in a village with access to natural lakes. 
Schistosomiasis was highest in a village with a small stream. 
For the three villages, urinary schistosomiasis in males was 12,
18, and 40%.  In females it was 8, 9, and 25%.  Apparently, the
"Bulinus globosus" snails in this area prefer small streams over
lakes or reservoirs (Le Bras "et al." 1982).  Similarly, Doumenge
(1984) reports that, in Burkina Faso, natural seasonal ponds
infested by "Bulinus truncatus" snails are more common and
dangerous locations of infections than are artificial reservoirs
created by dams or impoundments.

In the semi-arid Sudano-Sahel zone of West Africa,
schistosomiasis transmission typically occurs through seasonal
pools of water, rather than permanent bodies of water.  This
contradicts what we might expect, based on data from other
regions.  The "Bulinus senegalensis" snail that causes this
surprising phenomenon (Wright 1959, Greer "et al." 1990) is
dormant during the summer (Betterton "et al." 1983).  Numerous
rain-filled pools in hard laterite surfaces favor the breeding
habits of this snail species.  Therefore, in the east central
part of The Gambia, schistosomiasis in children was six times
higher in villages located on a laterite plateau than in river
villages cultivating swamp rice (Duke and McCullough 1954, also
cited in Wright 1973).  Similarly, in the middle Senegal valley,
a 1977-78 survey found the infection rate in 10 flood plain
villages was only 0.7%.  Five villages located on higher ground,
where "Bulinus senegalensis" snails commonly live in seasonal
laterite pools, had 10.4% average infection rate (Malek and
Chaine 1989).

Until recently, people in Cameroon assumed that small dams and
irrigation caused a higher rate of schistosomiasis.  However,
Ratard "et al." (1990) reported that a nation-wide survey, at
more than 500 locations in Cameroon, failed to find an
association between the rate of schistosomiasis and small dams. 
Instead, temporary ponds and snail hosts adapted to low seasonal
rainfall permits intense transmission of "S. haematobium" and 
"S. mansoni" throughout northern Cameroon, regardless of dams. 
Previous studies, comparing just two or three locations, did not
have enough observations to reach valid conclusions.

Contrary to the experience in Cameroon, Nigeria, and Burkina
Faso, one study in northern Ghana showed small dams to be linked
to schistosomiasis (Hunter 1981, also cited in WHO 1992).  Data
collected during 1960-61 showed much higher prevalence of
schistosomiasis in the eastern, more densely settled, part of
what was the Upper region of Ghana.  USAID funding had paid to
construct over 120 small earthen dams in this region during the
late 1950s.  The mean prevalence of urinary schistosomiasis in
the region's eastern part was 19.8% in 15 districts without dams,
42.3% in 16 districts with dams 1-2 years old, and 52.0% for 6
districts with dams 3 years old.  Areas in the western part of
the region with few or no dams had infection rates under 10% in 6
districts, and 10 to 29% in 10 others.  In contrast, prevalence
was over 70% in 2 of the 3 western districts containing dams.

The evidence offered from northern Ghana by Hunter (1981) is
strong, even if cross-sectional in nature. However, Lyons (1974:
623) reports that in the western part of the same region of
northern Ghana studied by Hunter, the prevalence of
schistosomiasis in 1969 to 1971 was 17.6% among individuals using
only river, stream, or pond water.  This compared to 8.1% for
people using water from small dams present in 11 of the 43
locations studied.  Most of the reservoirs examined did not have
infected snails and,

"The great majority of infections with "S. haematobium" are
contracted through contact with rivers and streams at points
other than where they have been impounded."

Studies from similar zones in neighboring Burkina Faso also did
not connect small dams and schistosomiasis transmission.

The only place in West Africa where unequivocal evidence exists
on the effect of small dams on snail vectors is the Dogon country
of Mali (Bandiagara district).  In this barren plateau, workers
built 20 small earthen dams in the mid-1970s.  In one year,
between 1976 and 1977, urinary schistosomiasis rose from 79.4 to
93.4%.  Researchers assumed this number was even lower before the
dam construction.  Citing this case and that of northern Ghana,
Hunter "et al." (1982) generalize about the adverse health
effects of small earthen dams to impound water.

Brinkmann "et al." (1988) compare the Bandiagara villages to
selected savanna villages elsewhere in Mali that do not have any
water resource projects.  The average prevalence of urinary
schistosomiasis was 67.2% in the Bandiagara villages and 13.4% 
in the savanna villages.  The mean prevalence of intestinal
schistosomiasis was 12.0 and 1.6% respectively.  Assuming that
the two areas are comparable, this would imply that prevalence of
urinary schistosomiasis increased 6-fold because of the dams. 
Recent evidence confirms that schistosomiasis is much higher in
Bandiagara villages with impounded water than in those without
[note 7].  However, in the dry savanna of western Mali without
dams, "S. haematobium" was as high as 85% in the Koulikoro area
(Wright 1973).  This suggests that the situation in Bandiagara
may not be generalizable even to rural Mali.

[Note 7. Long "et al." (1992) report that the average prevalence
rate of urinary schistosomiasis in 174 children examined for
blood in their urine was 63%.  They report that the rates were
highest, at or close to 100%, in four Dogon villages with dams
and much lower, as low as 14%, in two villages without dams or
access to dams.  Two villages with partial access to dams had
intermediate rates of 31% and 44%.]


Converted from gopher on 8/6/1999