--
( ~,
*
/ ~
pA
/--0
-low
TABLE OF CONTENTS
Page
A C K N O W LE D G E M E N T S ........................................................... .................................... ................ 2
EX EC U T IV E SU M M A RY ............................................................ .................................................. 3
SUMMARY OF RECOMMENDATIONS.................... .......................................................... 3
IN T R O D U C T IO N ................................................................................................................................... 4
M E T H O D S ............................................................................................................................................... 4
R E S U L T S .................................................................................................................. ..................... 6
POND SITE E VALUATION S ........................................................ ........................................................ 6
G ik o ro ............................................................................................................................................. 6
B iren g a ........................................................................................................... ..................... 6
C y u n g o ............................................................................................................................................ 6
T ab a ................................................................................................................................................. 6
K an am a ................................................................................................................ ....................... 6
M ab an za ........................................................................................................ ...................... 7
C yim b ogo ....................................................................................................... ..................... 7
B u g a ra m a ........................................................................................................................................ 7
B w afu .............................................................................................................................................. 7
K azabe ......................................................... .......................... . . . ............................ 7
STREAM AND CANAL SITE EVALUATIONS......................................................................... ..................... 7
D ISC U SSIO N ................................................... ....................................... . . . ................ 8
FISH CULTURE AND WATERBORNE DISEASE................................................................... ..................... 8
INTEGRATED AQUACULTURE AND WATER POLLUTION................................................. ..................... 8
ENVIRONMENTAL IMPACT ASSESSMENT.......................................................................... ...................... 9
EXOTIC SPECIES INTRODUCTIONS .............................. ......................................................................... 9
R E F E R E N C E S C IT E D ........................................................................................................................ 11
APPENDIX: CHAPTERS 3,7, AND 8 FROM THE REPORT OF MALEK (1983)
CONCERNING THE IMPACT OF FISH PONDS ON
PUBLIC HEALTH IN RWANDA........................................................................................................12
3. SUMMARY AND CONCLUSIONS.................................. ........................................... 12
7. FIELD OBSERVATIONS OF FISH PONDS IN RWANDA REGARDING
W ATER-R ELATED D ISEASES...............................................................................................................13
8. CONTROL OF SNAILS AND MOSQUITO LARVAE IN FISH PONDS ............................ ...................... 14
ACKNOWLEDGMENTS
Mr. Hishamunda Nathanal, Director of the National Fish Culture
Project of Rwanda, provided transportation and background informa-
tion and accompanied me to all fishpond and stream sampling sites.
Mlle. Mukakarera Colette, fifth year License of Biology student at the
National University of Rwanda, assisted with all field collections and
laboratory analyses of samples.
Dri D. Thys van den Audenaerde, Dr. Guy Teugels, and Mr. Jos
Snoeks allowed me access to pertinent scientific literature and valu-
able unpublished information at the Musie Royal de l'Afrique Cen-
trale, Tervuren, Belgium. Ms. Karen Veverica, Research Associate for
Auburn University and Oregon State University, and Dr. Robert Sny-
der, Agricultural Specialist for the Baptist Missions in Rwanda,
offered valuable background information about aquaculture in Rwan-
da, as well as hospitality. Ms. Veverica also provided laboratory space
at the Rwasave Fish Culture Station of the National University of
Rwanda, use of scientific equipment, and a computer.
Drs. Bryan Duncan, John Grover, and E. Cliff Webber, Depart-
ment of Fisheries and Allied Aquacultures, Auburn University,
reviewed the manuscript and made helpful suggestions for its
improvement.
EXECUTIVE SUMMARY
From August 6-17, 1990, an environmental assessment 
was con-
ducted at 10 potential sites (in seven prefectures) for 
cooperatively
managed, integrated aquaculture farms in the marais 
(wetland valleys)
of Rwanda. Sites were evaluated for existing pond 
construction and
management practices, and recommendations were 
made to minimize
the risk of waterborne disease and other negative 
environmental
impacts.
Snail hosts for schistosomiasis and sheep and 
cattle liver fluke
occurred at more than half of the 10 ponds sampled. Mosquito 
larvae
were less common, and anopheline mosquitos (vectors 
of malaria)
were found at only one site. Most snails and mosquitos occurred 
at
poorly built and managed ponds. The presence 
of a well-trained
extension agent at a site was a critical factor to ensure 
well-managed
ponds and thus reduce risk of waterborne disease.
Two supply canals and three discharge canals were sampled 
for
macroinvertebrates to determine possible downstream 
impacts of
fishponds. Benthic communities were characterized by 
low biodiversi-
ty and high tolerance to organic enrichment, probably 
because of fre-
quent dredging of canals, intermittent water flows, 
and organic runoff
from human and livestock activity. No negative environmental 
impact
directly attributable to fish culture activity was detected.
At the scale and management intensity of the proposed aquaculture
centers, fishponds would have relatively low volume 
and infrequent
discharges that should not cause undue environmental 
degradation
downstream.
Four streams with a wide range of organic pollution 
were sampled
for macroinvertebrates to evaluate methods for a biomonitoring 
pro-
gram. Use of a biotic index proved appropriate and 
practical for assess-
ing stream water quality in Rwanda. Protocols with field 
methods, data
interpretation tables, and a cumulative list of macroinvertebrate taxa
in Rwanda are provided.
Risks of indiscriminate introduction of non-native species 
to Rwan-
da are discussed, and a review and decision model for evaluating 
pro-
posed exotic fish introductions is presented.
SUMMARY OF RECOMMENDATIONS
Training and placement of extension agents should remain 
a high
priority of the National Fish Culture Project. To reduce the risk of
waterborne diseases (schistosomiasis, malaria, and liver fluke), 
exten-
sionists should especially help farmers to design ponds and 
eliminate
shoreline vegetation.
To prevent livestock disease and excess organic loading of ponds,
extensionists should help farmers to determine the number 
of animals
that can be kept well-fed and clean in appropriately sized enclosures.
Because cut shoreline vegetation may contain snail vectors of 
disease
(eggs and adults) and parasitic cysts, it should be thoroughly 
dried or
composted away from the pond and from sheep, goats, and cattle.
Quarterly biomonitoring of at least four "trend stations" 
(e.g. small
streams on watersheds with integrated aquaculture projects) 
is recom-
mended to document long-term impact of fish farming, 
pesticide pol-
lution, and other potential disturbances.
Introduction of non-native fishes to Rwanda has altered 
natural
aquatic ecosystems and future introductions should 
only follow a thor-
ough justification and description of the non-native 
species, as well as
detailed and controlled studies on the predicted environmental 
conse-
quences.
Environmental Assessment of Ten Aquaculture
Sites in Rwanda, Africa
William G. 
Deutsch'
INTRODUCTION
Recent research has greatly increased the potential for fish produc-
tion in Rwanda (45). At the completion of a 5-year, USAID-sponsored
technical assistance program with the National Fish Culture Project
(PPN) of Rwanda (44) it was concluded that proper construction and
management of fishponds can result in more than a four-fold increase
in tilapia production, relative to previous culture practices (29).
Because commercial feeds and fertilizers are generally unavailable,
aquaculture in Rwanda is usually limited by nutrient inputs to fish-
ponds. Present research at the Kigembe Station of the PPN has
focused on integration of fish culture with small livestock husbandry.
These integrated systems have successfully used animal manures to
improve pond fertility and further increase fish production.
In an effort to extend improved aquacultural techniques to rural
farmers, the PPN has proposed the establishment of up to seven coop-
eratively managed fish culture centers. These centers would be dis-
persed throughout regions of Rwanda where aquaculture is now
practiced, and would receive technical support from the PPN so that
they function as model farms for training and extension. USAID-
Kigali supported the PPN proposal but requested both an engineering
and environmental assessment of potential sites prior to funding the
project.
The objective of this study was to conduct an environmental assess-
ment of potential sites for PPN-integrated aquaculture centers. This
involved an evaluation of present fish culture practices and recom-
mendations for fishpond construction and management that would
minimize negative environmental impacts. Special attention was given
to fishponds and waterborne disease, and effects of pond discharge on
the downstream environment. Protocols for environmental impact
assessment and evaluation of exotic species introductions also are pro-
vided.
METHODS
From August 6-17, 1990, an environmental assessment was made at
10 pond sites (in seven prefectures) in the marais (wetland valleys) of
Rwanda, figure 1. Each site was evaluated for the construction and
management of existing ponds as well as the potential environmental
impact of an expanded aquacultural facility that might include integra-
tion with small livestock (47).
Construction features that were noted included pond surface
dimensions and depth and characteristics of the dikes and canals that
supplied water and received discharge. Management features noted
included the method of fertilization, water color and clarity, and shore-
line vegetation control. All ponds were stocked with fish, Nile tilapia
(Oreochromis niloticus), and were managed by families, school
groups, or small cooperatives.
'Deutsch is an environmental and training specialist in the International Cen-
ter for Aquaculture and Aquatic Environments.
Several points along the shoreline of at least one pond at each site
were sampled for snails, mosquito larvae, and other macroinverte-
brates using a dip net. The presence and relative abundance of the fol-
lowing human and livestock disease vectors were especially noted:
1. Biomphalaria spp. (B. pfeifferi and B. sudanica) - snail interme-
diate hosts for Schistosoma mansoni (blood flukes causing human
intestinal schistosomiasis).
2. Bulinus spp. (B. globosus and B. strigosus) - snail intermediate
hosts for Schistosoma haematobium (blood flukes causing human uri-
nary schistosomiasis).
3. Lymnaea natalensis - snail intermediate host for Fasciola gigan-
tica (sheep and cattle liver fluke)
4. Anopheles spp. - mosquito intermediate host for Plasmodium
falciparum, P malariae, and P vivax (human malaria).
If the supply or discharge canal of a pond site had a current velocity
of at least 5-10 cm/s, and an area relatively free of vegetation, it was
sampled for macroinvertebrates to determine a family-level biotic
index (28). The index is a relative measure (on a scale of 0 to 10) of
organic pollution in flowing water, and is based on the tolerance level
of each family of macroinvertebrates in the sample, table 1.
TABLE 1. EVALUATION OF STREAM WATER QUALITY USING THE
FAMILY-LEVEL BIOTIC INDEX'
Family Biotic Index
2  
Water quality Degree of organic pollution
0.00-3.75 Excellent Organic ollution unitkel
3.76-4.25 Very good Possible slight organic polution
4.26-5.00 Good Some organic pollution probable
5.01-5.75 Fair Fairly substantial pollution likely
5.76-6.50 Fairly poor Substantial pollution likely
6.51-7.25 Poor Very substantial pollution likely
7.26-10.00 
Very poor 
Severe organic pollution 
ikely
'Field methods used: (A) sample as often as necessary to obtain 100-200 in-
vertebrate specimens sufficienty mature to identify to family, using a di net in
a riffle or shallow run (ideally where current is greater than 30 cm/s); (B) place
net contents in a shallow white pan with water for observation (specimens
clinging to the net should be included); (C) sort specimens by family into
containers of 70% ethanol, excluding Hemiptera and Coleoptera (other than
Dryopoidea) and individuals too sm91l to identify; and (D) record number of
specimens per family.
2
CalculatehFBI by multiplying the number in each family 
(n,) by the tolerance
value for that family (t, s umming the products and dividing by the total
number of arthropods (N) in the sample (i.e. FBI = I (n, t) / N).
A biotic index was also determined for macroinvertebrate commu-
nities at four stream sites, figure 1, that had a wide range of organic
pollution. One station was relatively pristine and in an area of low
human population density (Karamba, in the Nyungwe Forest). Two
sites were in areas of moderate human and livestock activity (Rushashi
and a tributary of the Nyabarongo River), and one site received heavy
organic loading from an upstream butchery area (Kamabuye). Results
from stream samples were used to evaluate the applicability of
bioassessment techniques (developed in the United States) for Rwan-
RWANDA
Pond Sites
1 Kanama
2 Bwafu
3 Kazabe
4 Cyungo
5 Mabanza
6
7
8
9
10
Taba
Gikoro
Birenga
Cyimbogo
Bugarama
Stream Sites
A Rushashi
B Tributary of the
Nyabarongo River
C Kamabuye
D Karamba
SProject Headquarters-
Kigembe
A Regional Fish Station
FIG. 1. Map of Rwanda with pond, canal, and stream sampling sites.
da, as well as to relate indices from fishpond canals to a spectrum of
organic pollution levels.
Organisms collected at ponds, canals, and streams were hand sort-
ed from shallow trays of water and preserved in 70% ethanol. Samples
were returned to the laboratory of the Rwasave Fish Culture Center
and organisms were enumerated and identified using a dissecting
microscope, hand lens, and taxonomic keys of Durand and L6veque
(20,21).
RESULTS
POND SITE EVALUATIONS
Gikoro (Kigali Pr6fecture)
The Gikoro site had one pond (150 mn
2
) stocked with tilapia, though
other ponds were being constructed. Ponds were managed by a group
of local women. An extension agent has not been assigned to Gikoro,
but one is presently intraining at the Kigembe facility. Soils had a high
clay content and were suitable for pond construction. The stocked
pond had a compost enclosure filled with vegetation, household
scraps, and manure. A large amount of plant fragments was floating in
open water, outside of the enclosure. The shoreline was irregular with
numerous indentations that were filled with emergent vegetation. The
pond was constructed without levelling instruments and had an exces-
sive bottom slope. About 25-30% of the pond was too shallow, as evi-
denced by emergent vegetation 3-5 m from shore in the shallow end.
Depth near the discharge was 1-1.2 m.
Dip net samples in shoreline emergent vegetation revealed that
there were hundreds of snails (especially Biomphalaria pfeifferi with
some Lymnaea natalensis) per linear meter of pond bank. These and
other macroinvertebrates incidentally collected with snails are listed in
table 2. No mosquito larvae were observed.
Birenga (Kibungo Pr6fecture)
The Birenga site had one pond (500 m
2
) managed by about 17
young adults. They intend to build more ponds on either side of the
existing pond and raise pigs. An extensionist for this area is in training
at Kigembe. This area of the marais was relatively undeveloped and
was being cleared of trees. Dikes were straight and steep-sided, but
had tree stumps imbedded in them that would probably result in
water leakage. The small amount of compost in a corner enclosure had
not been turned recently, as evidenced by 50-cm-high grasses growing
in the floating vegetation. Some vegetation was floating on the pond
surface, outside of the enclosure.
The pond was wider than the available harvesting net, and had been
stocked a second time after a partial harvest. Present stocking density
was, therefore, unknown. Construction of the discharge canal for this
pond was not completed, so an adequate method of complete harvest
was not available.
Some floating vegetation (Nymphaea sp.) was observed in the pond
2-3 m from shore. Emergent grasses at the shoreline contained the
snail, L. natalensis. No mosquito larvae were observed, table 2.
Cyungo (Byumba Pr6fecture)
The Cyungo site had a large (2,500-3,000 m
2
) single-family pond
with about 15 compost enclosures around its perimeter. Farmers have
chosen to raise rabbits and chickens at this site, and pigs will be raised
in the area at other ponds. The dike is broken to drain the pond and is
rebuilt after harvest.
This pond was well managed. There was a healthy phytoplankton
bloom and tilapia actively fed on Colocasia leaf fragments thrown on
the water surface. The pond was well built with little shoreline vegeta-
tion. Only one snail, Bulinus sp., and a single mosquito larva was found
after sampling several meters of shoreline, table 2.
Taba (Gitarama Pr6fecture)
The Taba site had two ponds (200 m
2
each), with six additional
ponds planned that would be 1,500-2,000 m
2
each (3 ha total surface
area). Ponds were managed by students of a CERAI school, and rab-
bits and cows are already raised at the site. Being close to Kigali, there
is a good market for meat and fish produced here. Ponds had relative-
ly straight, steep-sided dikes and no emergent vegetation in open
water. Shoreline vegetation was abundant, however, and the emergent
grasses and smartweed (Polygonum spp.) contained three genera of
snails, Bulinus sp., L. natalensis, and B. pfeifferi, table 2. No mosquito
larvae were found.
Kanama (Gisenyi Pr6fecture)..
The Kanama site had about 40 ponds that have been built over the
last 3 to 5 years. Ponds were of various sizes and shapes, and were dis-
organized in their layout. Soils have a high organic matter content and
dikes of some ponds were poorly compacted and unstable.
In general, ponds were poorly built and managed. Water levels var-
ied and some ponds had large areas that were less than 20 cm deep.
Some ponds were completely covered with duckweed and water meal
(Lemna sp. and Woffia sp., respectively); others had large amounts of
floating vegetation (Potamogetop sp.) in open water. Most ponds were
infertile, with little to no compost. Pond water was generally clear and
some ponds were darkly stained with humic compounds. Shoreline
grasses and smartweed were abundant and contained the snail, L.
natalensis and numerous mosquito larvae, table 2.
TABLE 2. CUMULATIVE LIST OF SNAILS, MOSQUITOS, AND OTHER
MACROINVERTEBRATES COLLECTED FROM SHORELINES OF
PONDS AT TEN SITES IN RWANDA AUGUST 6-17, 1990
Taxon 
Pond Site'
1 2 3 4 5 6 7 8 9 10
Snails (Gastropoda)
Bulinidae
Bulinus sp. P P P P
Lymnaeidae
Lymnaea natalensis P P P P P P
Planorbidae
Biomphalaria pfeifferi P P P P P P
Thiaridae
Melanoides tuberculata P P
Mosquitos (Diptera)
Culieidae
Anophelinae P
Culicinae P P
Other Macroinvertebrates
Ephemeroptera
Baetidae P P P P P P P
Caenidae P
Odonata
Aeshnidae P P P
Coenagrionidae P P P P P P P
Libellulidae P P P
Macromiidae P
Hemiptera
Belostomatidae P P P P P P
Corixidae P P P
Gerridae P P
Hydrometridae P
Mesoveliidae P P P P
Naucoridae P P P
Nepidae P P P P P
Notoneetidae P P P P P
Pleidae P P
Coleoptera
Dytiscidae P P P P
Gyrinidae P P
Hydrophilidae P
Diptera
Ceratopogonidae P
Chironomidae P P P P
Ephydridae P
Sites of study: 1 = Gikoro; 2
6 = Mabanza; = Cyimbogo; 8
present
= Birenga; 3 = Cyungo; 4 = Taba; 5 = Kanama;
Bugarama; 9 = Bwafu; 10 = Kazabe; P =
Mabanza (Kibuye Pr6fecture)
The Mabanza site had four large ponds (3,000-4,000 m
2 
each) that
were constructed about 15 years ago. This was previously a govern-
mental station that cultured O. macrochir and common carp (Cypri-
nus carpio) with minimal feed or compost inputs. It was managed 
by a
cooperative of local men who use compost enclosures to fertilize
ponds stocked with O. niloticus. A woman extensionist, in training at
Kigembe, will be stationed there.
Ponds were generally well constructed but needed repair. Dikes
were high (1-1.5 m above water surface) and undercut in several
places. Shoreline vegetation was abundant and contained the snails, L.
natalensis and B. pfeifferi, table 2; no mosquito larvae were observed.
A cow grazed near the ponds and a boy swam in one of them at the
time of sampling. This indicated that there may be regular human and
animal contact with pond water that contained snail hosts for schisto-
somiasis and cattle liver fluke.
Cyimbogo (Cyangugu Pr6fecture)
The Cyimbogo site had 300-400 m
2 
ponds managed by families
that own the land of the marais. Cultured Tilapia rendalli and
O. macrochir were replaced by O. niloticus within the past 5 years.
The people are willing to convert the surrounding sweet potato and
cassava fields into fishponds because of the high demand for animal
protein within this Pr6fecture and from nearby Zaire. They wish to
integrate fish culture with chickens and goats, and they also may raise
rabbits. An extension agent was active at Cyimbogo and has previously
improved pond design. Farmers have kept good fish production
records.
Ponds were well built and managed. Dikes were straight with little
shoreline vegetation. Compost enclosures were properly maintained,
and ponds had healthy phytoplankton blooms. Water from the supply
canal was filtered with a screen before it entered the ponds. No
mosquito larvae or snail hosts for schistosomiasis or liver fluke were
observed after sampling several meters of shoreline, table 2. A single
specimen of the predaceous snail Melanoides tuberculata was found.
Bugarama (Cyangugu Prdfecture)
The Bugarama site had four ponds (200-300 m
2
) managed by a
group of Boy Scouts. Ponds were at the lowest elevation in Rwanda
(900 min). A rice/fish culture project has been proposed 
for this area,
and farmers are willing to replace existing rice fields with fishponds.
An extensionist is presently being trained at Kigembe, and a neighbor-
ing extensionist assists here 
occasionally.
Ponds were fertilized with cut sedges (Cyperus sp.) in compost
enclosures. Dikes were higher than necessary and shoreline grasses
were abundant. One pond seemed to have the water level recently
lowered and contained emergent vegetation in open water. Two snail
species, B. pfeifferi and M. tuberculata were found there. table 2. This
was the only site at which anopheline mosquito larvae were collected
from fishponds.
Bwafu (Gisenyi Prefecture)
The Bwafu site had five government-owned ponds (about 2,000-
3,000 m
2 
each), adjacent to a Dutch agricultural project. An extension-
ist from the PPN had not been assigned here. Local farmers were
interested in producing more fish and beginning livestock husbandry.
Ponds were terraced with about a 3-m difference in the water level of
the first and fifth ponds. Dikes were extremely high (up to 2 m above
the water surface) and large portions of them were subsiding into the
ponds. Some ponds were greater than 2 m deep at the water level con-
trol monks, and the outer slope of all ponds was up to 3-m high and
leaking.
In general, ponds were too large and deep to effectively manage
under current conditions. There was little or no compost in ponds and
existing compost enclosures were grossly undersized. Water was infer-
tile and brown except in the smallest pond which was composted and
used to produce tilapia fingerlings for stocking in other ponds. Stock-
ing and harvest records were unavailable, but only about 5 kg of
O. niloticus, stocked from Kigembe 1 year ago, have been harvested.
There was little emergent vegetation, and dip net samples revealed
that shoreline macroinvertebrates were generally scarce. A few Buli-
nus sp. and B. pfeifferi snails were collected along the shoreline, but
no mosquito larvae were observed, table 2.
Kazabe (Gisenyi Pr6fecture)
The Kazabe site had eight government-owned, terraced ponds
(200-300 m
2 
each) and eight additional ponds that were drained.
Ponds were designed and managed much like those at nearby Bwafu
and an extensionist from the PPN had not been assigned here. A
group of local women was interested in improving these ponds and
integrating them with small livestock. Ponds were not fed or compost-
ed and water was infertile and brown. Shoreline vegetation contained
three snail species, Bulinus sp., L. natalensis, and B. pfeifferi. No
mosquito larvae were found, table 2.
STREAM AND CANAL SITE EVALUATIONS
A biotic index (BI) was determined for four streams and five canal
sites, table 3 and figure 2. Macroinvertebrate data which formed the
basis for the BI for each site are presented in Deutsch (15). The BI of
streams reflected known conditions of organic pollution and, there-
fore, seemed applicable for bioassessment in Rwanda. The most pris-
tine stream, at Karamba, had the greatest biodiversity and the lowest
BI, figure 2. Water quality for this stream was rated "very good, near
excellent", table 1. The Kamabuye stream, with known organic pollu-
tion, had only one-third the biodiversity of that at Karamba, and the
highest BI, figure 2. One genus of Chironomidae (Chironomus spp.)
composed more than 90% of the total number of organisms present.
Water quality for this stream was rated "very poor, severe organic pol-
lution likely", table 1. Streams at Rushashi and near the Nyabarongo
River had intermediate biodiversity and water quality ratings of "good,
some organic pollution probable", table 1. Results suggested that
macroinvertebrate families in Rwanda were "ecologically equivalent"
to their counterparts in North America in terms of pollution tolerance,
and that criteria of the BI were appropriate for this study.
Most canals at pond sites were not evaluated with a BI because they
did not have flowing water or were overgrown with vegetation. Canal
sites that were sampled had macroinvertebrate communities with low
biodiversity (few taxa) and high tolerance to organic enrichment.
Cyimbogo, where fishponds were located downstream from a large
agricultural area, had the fewest number of taxa and highest BI of the
canals sampled, figure 2. Gikoro, in a less developed marais, had ben-
thic communities in canals that were slightly more diverse and with a
lower BI. At the two sites where both the supply and discharge canals
were sampled (Cyimbogo and Gikoro), macroinvertebrate communi-
ties were relatively similar upstream and downstream of fishponds,
figure 2. This indicated that negative environmental impacts on ben-
thos were not attributable to fish culture activities.
Biotic index
9 -2 
Canal
2 3
^ 8.01 7,92 4 [777 o+ ...
1 2 3 4 5 6 7 8 9
Sampling site
FIG. 2. Number of macroinvertebrate families and family-level biotic
index for nine sampling sites in Rwanda, 6-17 August 1990
TABLE 3. AN EXAMPLE FIELD SHEET FOR CALCULATION OF THE FAMILY-LEVEL
BIOTIC INDEX AT A STREAM AT KARAMBA, RWANDA AUGUST 13, 19901
n, ti (natj) n, ti (njtl)
EPHEMEROPTERA ISOPODA
Baetidae 119 4 476 Asellidae 8
Caenidae 7
Ephemerellidae 1
Ephemeridae 4
Heptageniidae 4 TRICHOPTERA
Leptophlebiidae 3 2 6 Hydropsychidae 11 4 44
Oligoneuridae 2 Hydropfilidae 4
Tricorythidae 4 Lepidostomatidae 1 1 1
Leptoceridae 1 4 4
Limnephilidae 4
Philopotamidae 3
Polycentropodidae 2 6 12
Psycomyiidae 2
Phyacophilidae 0
Sericostomatidae 1
ODONATA
Aeshnidae 3
Calopterygidae 5
Coenagrionidae 9
Cordullidae 5
Gomphidae 1
Lestiae 9 COLEOPTERA
Libellulidae 9 Elmidae 4
Macromiidae 3
DIPTERA
Blephariceridae 0
Ceratopogonidae 1 6 6
Chironomidae (red) 8
PLECOPTERA Chironomidae (pink) 6
Perlidae 1 Ephydridae 6
Psychodidae 10
Simuliidae 1 6 6
Tabanidae 6
AMPHIPODA Tipuidae 3 3 9
Gammaridae 4
Talitridae 8
Family Level Biotic Index (FBI) = I (nit1) / N = 564 / 142 = 3.97
'The list of macroinvertebrate families in Rwanda is from Durand and
Liveque (20,21).
Sampling Site: Stream at Karamba (Cyangugu Prefecture), Rwanda, Africa
Date: August 13, 1990 Collector(s): WGD, MC Identifier(s): WGD, MC
Notes: Biotic Index of 3.97; Number of Families = 9; 6 Hydrophilidae and 3
Gyrinidae (Coleoptera) were also colected.
In general, the benthos of canals upstream and downstream of
ponds seemed to be altered from such factors as frequent disturbance
of the substrate (dredging of canals), intermittent water flow, and
organic enrichment from nearby human and livestock activities. It is
important to note that the BI of canals is not directly comparable to
the BI of streams, and canal water quality should not be interpreted
according to table 1. The canal aquatic environment is unstable, and
usually had flows less than those required for appropriate application
of the BI . Canal water quality would be better evaluated using chemi-
cal or bacteriological (fecal coliform and streptococcus) methods.
DISCUSSION
FISH CULTURE AND WATERBORNE DISEASE
Fishponds in tropical countries have frequently been implicated in
the spread of human and livestock disease (30,36,39). Desowitz (14)
cited examples of fish culture development projects that resulted in
expanded mosquito habitat and a large increase in the incidence of
human malaria. Others have noted that fishponds may be favorable
habitats for snail vectors of schistosomiasis (5), and numerous efforts
have focused on the control of these snails (12,41,42,43,56).
Recent documentation of schistosomiasis in Rwanda (25) has
resulted in concern over the role of fishponds in the spread of this dis-
ease. Malek (40) conducted a survey of several fishponds in Rwanda
and concluded that the snail hosts for schistosomiasis and sheep and
cattle liver fluke were widespread. Ponds have the potential for facili-
tating the spread of these diseases if infected humans or livestock con-
taminate them, however, the trematode parasites do not yet seem to
be prevalent in Rwanda. Recommendations for controlling mosquito
and snail hosts for malaria, schistosomiasis, and liver fluke were relat-
ed, and included maintenance of fish in ponds at all times, control of
shoreline vegetation, removal of floating debris, and screening of sup-
ply canal water. Unedited excerpts from the report of Malek (40),
which give background information on fish culture and waterborne
disease and are pertinent to the present study, are presented in the
Appendix.
In the present study, the findings of Malek (40) were substantiated
in that snail hosts for schistosomiasis and liver fluke were found in
more than half of the ponds sampled, table 2. Mosquito larvae were
less common and anophelines were observed at only one site.
Control of shoreline vegetation is perhaps the single most impor-
tant factor in eliminating mosquito and snail habitats in stocked fish-
ponds. Without emergent plants, mosquito larvae and trematode
cercariae are vulnerable to fish predation, and snails are deprived of
their preferred, firm substrates (plant stems and leaves) for grazing.
Because of the prevalence of L. natalensis in ponds, shoreline vegeta-
tion may contain metacercarial cysts of the liver fluke as well as snails
and their eggs. It is, therefore, important to not compost cut shoreline
plants in fishponds nor feed them to cattle, sheep, or goats. Shoreline
vegetation should be thoroughly dried or composted away from water
and livestock to break the life cycle of the parasite.
There was a clear relationship between design and management
features of ponds and the occurrence of disease vectors. For example,
the poorly built pond at Gikoro had the most snails, and ponds with
abundant shoreline and open-water vegetation at Kanama and
Bugarama harbored mosquitos. Conversely, well-built and well-man-
aged ponds at Cyungo and Cyimbogo had almost no disease hosts.
The role of an active, well-trained extensionist cannot be underesti-
mated in minimizing risks of waterborne disease. At the sites visited,
they seemed to be the critical factor in this regard. The fact that exten-
sionists have effectively helped farmers to keep pond dikes straight
and relatively free of shoreline plants is a credit to the training at
Kigembe, and continued training and placement of extensionists
should remain a high priority of the PPN.
INTEGRATED AQUACULTURE AND
WATER POLLUTION
Because most fishponds in Rwanda are nutrient limited, integration
of aquaculture with livestock husbandry has the potential for increas-
ing fish production while supplying farmers with additional meat and
animal by-products. Integration is one of the most efficient ways to
convert manures into edible protein (38,49). Although integrated
aquaculture has been successfully practiced in Asia for centuries,
there is concern that these culture methods may spread disease or pol-
lute the downstream environment. For example, some virologists
warn that integrated aquaculture with pigs, poultry, and fish, as prac-
ticed in Thailand, creates an environment in which harmful viruses
may be transmitted to people (19,48). Pigs are particularly implicated
as a source of mutated viruses which infect human populations.
To minimize the spread of disease in integrated systems, livestock
should be kept well-fed and clean in adequately sized enclosures. Fish
culture with livestock in properly built pens at Kigembe has demon-
strated that animal manures may be safely metabolized in ponds to
increase tilapia production in Rwanda. Use of manures may even
reduce disease risks by enhancing pond fertility and phytoplankton
blooms that, in turn, shade out emergent vegetation and eliminate
snail and mosquito habitat.
Proper matching of pond size with manure inputs (e.g. kg of animal
per unit surface area of water) is important to avoid excess organic
loading and unhealthy, septic conditions that may pollute the down-
stream environment when water is discharged. Fortunately, the fish
serve as one of the most important checks against over-manuring
because the resulting low dissolved oxygen or build-up of toxic
metabolites (e.g. ammonia) would preclude their culture. Extension-
ists at Kigembe should be trained to assist farmers in housing the
appropriate number of animals at their ponds.
Although much of the organic inputs from manures would be con-
verted to fish flesh, some will be released downstream when ponds are
drained. This study found that the aquatic environment immediately
below fishponds is already disturbed from other factors, and that the
relatively few taxa that occur there are tolerant of high levels of organ-
ic enrichment, figure 2. In large aquacultural systems that receive
intensive feed and chemical inputs, fishpond discharges may be of low
quality and adversely affect downstream use (4,6,22,55). At the scale
and management intensity of the proposed aquaculture centers, how-
ever, fishponds would have relatively low volume and infrequent dis-
charges that should not cause undue environmental degradation
downstream.
As freshwater wetlands, marais are among the most productive
ecosystems, and much of the excess organics discharged from fish-
ponds will quickly become incorporated into plant biomass (cultivated
and natural vegetation). Some organics will probably reach surface
streams or ground water and become a source of non-point pollution.
For this reason, environmental monitoring of streams is recommended.
ENVIRONMENTAL IMPACT ASSESSMENT
Basic ecological information is needed to form a baseline from
which sound environmental management decisions can be made.
Although there has been some attempt to evaluate and control aquatic
pollution in Africa (1), baseline information on water quality trends in
Rwanda is sorely lacking. This study made a preliminary evaluation of
the usefulness of the biotic index for assessing organic pollution in
streams and canals of Rwanda, table 1 and figure 2. Several macroin-
vertebrate families were collected and used in this effort, table 4.
Aquatic macroinvertebrates are typically abundant, relatively
immobile, and have a wide range of pollution tolerances. For these
reasons, they have been used in pollution studies for a number of years
(26). Recent research has developed rapid bioassessment methods
that require a minimal amount of equipment (27,28,37). Fish commu-
nities have also been used in impact assessment (23,34), and a concur-
rent study of water chemistry, macroinvertebrates, and fish provides a
thorough description of non-point sources of organic pollution in
streams (16).
The biotic index would be useful in stream studies in Rwanda, and
it is one of the most practical ways to determine the long-term impact
of fish culture or other human and livestock activities on water quality.
Although taxonomic and ecological information on macroinverte-
brates in Rwanda is scarce (7,9,20,21,32), family-level tolerance values
determined for benthos in the United States seemed applicable for
this study. The BI could be refined for local conditions and made more
accurate by determining tolerance values for organisms at lower taxo-
nomic levels (genus or species). For example, a biotic index that used
genera and species of macroinvertebrates was applied to stream
assessments in South Africa by Chutter (8). The BI is best applied to
first and second order streams and is not recommended for use in
canals, table 1.
Stream macroinvertebrates also could be used to detect other
forms of disturbances such as pesticide pollution. Because inverte-
brates such as aquatic insects and worms form a critical link in the food
webs of Rwandan fishponds, and they are particularly intolerant of
most pesticides used in agriculture (33), it may be important to moni-
tor upstream use of chemicals that may threaten the feasibility of
aquaculture. At a USAID-sponsored aquaculture project in El Sal-
TABLE 4. CUMULATIVE LIST OF MACROINVERTEBRATES COLLECTED FROM
PONDS, CANALS, AND STREAMS IN RWANDA, AUGUST 6-17, 1990
Arthropoda
Insecta
Ephemeroptera
Baetidae
Caenidae
Odonata
Aeshnidae
Coenagrionidae
Libellulidae
Macromiidae
Hemiptera
Belostomatidae
Corixidae
Gerridae
Hydrometridae
Hydronwtra sp.
Mesoveliidae
Naucoridae
Nepidae
Nepa sp.
Ranatra sp.
Notonectidae
Pleidae
Plea sp.
Tricoptera
Hydroysychidae
B jropsyche spp.
Aiacronema sp.
Lepidostomatidae
Goerodes sp.
Polycentropodidae
Arthropoda (cont.)
Insecta (cont.)
Coleoptera
Dytiscidae
Gyrinidae
Hydrophilidae
Diptera
Teratopogonidae
Chironomidae
Culicidae
Anophelinae
Culicinae
Ephydridae
Mollusca
Gastropoda
Bulinidae
Bulinus 
spp.
Lymnaeidae
Lymnaea natalinsis
Planorbidae
Biomphalaria pfeifferi
Thiaridae
Melanoides tuberculata
vador, for example, it was impossible to maintain fish in ponds until
upstream use of pesticides was controlled (3).
It is recommended that several "trend stations" be established on
small watersheds in Rwanda for periodic bioassessment studies. A
quarterly monitoring of four or five key sites would require approxi-
mately four person-weeks per year for a qualified technician with
expertise in macroinvertebrate taxonomy. Such a collection of ecologi-
cal information could be invaluable for determining conditions and
appropriate future uses of water in the country.
EXOTIC SPECIES INTRODUCTIONS
There has been relatively little control over the introduction of
exotic fishes to Africa, and some non-native species have irreversibly
altered natural aquatic environments. In Rwanda, there have been
introductions of the grass carp, Ctenopharyngodon idella (1979), the
common carp, Cyprinus carpio (1960s), the silver carp, Hypoph-
thalmichthys molitrix (1979), the clupeid, Limnothrissa miodon
(1960s), and the tilapias, O. macrochir (1950s), and T rendalli (1956),
among others (54). Ogutu-Ohwayo and Hecky (46) noted some of the
"species extinctions, introgressive 
hybridizations, and ecosystem 
alter-
ations" that occurred following fish introductions in African waters,
and a recent study by DeVos et al. (18) documented the complete
change in a cyprinid-based fishery in Lake Ruhondo, Rwanda, follow-
ing the introduction of Tilapia.
New species of fish are continually being described from Rwanda
(17,51,52) and there may be hundreds of invertebrates, plants, and
other organisms which are unknown to science and potentially useful
to humans. Clearly, the indiscriminate introduction of non-native
organisms, which may threaten these species, should be strongly dis-
couraged. Introductions should only follow a thorough justification
and description of the potentially introduced species, as well as
detailed and controlled studies on the predicted environmental conse-
quences. Even the best evaluations may fail to detect negative impacts
that become impossible to correct after introduction. A decision
model for exotic introductions, figure 3, was developed by Kohler and
Stanley (35) and has been applied to potential fish transfers in Africa
by Slootweg (50).
Proposal for exotic species 
introduction
f
LEVEL OF REVIEW 1
1. Determine validity for introduction.
2. Determine population abundance in native
range and current levels of exploitation.
3. Determine potential for inadvertent
introduction of 
disease and 
parasites.
4. Characterize site of proposed introduction.
NO OR
S UNCLEAR
LEVEL OF REVIEW 2
1. Determine the acclimatization potential.
YES OR
UNCLEAR
LEVEL OF REVIEW 3
1. Predict ecological benefits and risks.
2. Predict benefits and risks to humans.
UNCLEAR
tYES
APPROVE
DECISION BOX 2
1. Would the exotic species be able to survive and
reproduce in the range of habitats that would be
available?
AtNO
APPROVE
IL4iNO
APPROVE
LEVEL OF REVIEW 4
1. Conduct detailed literature review to
develop a FAO species synopsis.
$r
I UNCLEAR
DECISO10N BOX 4
1. Was data base adequate to develop a
complete species synopsis?
NO
2. Does the data base indicate desirability for
introduction? NO
tYES
APPROVE
DECISION BOX 5
1. Based on all available information, do the benefits
of the exotic fish introduction out-weigh the risks?
YES NO
APPROVE REJECT
FIG. 3. Review and decision model for evaluating proposed exotic introductions.
10
REJECT
1. Are reasons for introduction valid?
YES NO
2. Is the exotic species endangered,
threatened, or rare in its native range?
NO YES'-I
3. Would adequate safeguards be taken
to guard against introduction of disease
and parasites? 
YES 
NOYESNO
4. Would the exotic species be maintained
in a closed system with little chance for
escapement?
DECISION BOX 3
1. Would the exotic species have major adverse
ecological impacts?
NO YES-
2. Would the exotic species potentially be
hazardous to man? YES-*-
REJECT
REJECT
LEVEL OF REVIEW 5
1. Conduct research necessary to complete
species synopsis.
2. Conduct research to assess potential im-
pact on indigenous species and habitats.
ff
I
1 2MI.-
I
a
a a
son
BX~fREJECT
i nrI
f UNCLEAR I
1 00-
REFERENCES CITED
(1) ALABASTER, J. S. 1981. Review of the State of Aquatic Pollution of
East African Inland Waters. CIFA Occasional Paper No. 9. Food
and Agriculture Organization of the United Nations. Rome. 36
PP.
(2) BARD, J. AND L. MVOGO. 1963. Note d'Information sur l'Astatore-
ochromis alluaudi Poisson Molluscophage Utilisable Dans La
Prophylaxie De La Bilharziose. Bulletin de la Societe de Patholo-
gic Exotique 56: 119-124.
(3) BAYNE, D. R. 1974. Progress Report on Fisheries Development
in El Salvador. Research and Development Series No. 7. Project
A.I.D./ la.-688. International Center for Aquaculture, Agricul-
tural Experiment Station, Auburn University. 11 pp.
(4) BERGHEIM, A. AND A. SIVERTSEN. 1981. Oxygen Consuming
Properties of Effluents from Fish Farms. Aquaculture 22:185-
187.
(5) BERRIE, A. D. 1966. Fish Ponds in Relation to the Transmission
of Bilharziasis in East Africa. East African Agricultural and
Forestry Journal 21:276-282.
(6) BOYD, C. E. 1978. Effluents from Catfish Ponds During Fish
Harvest. Journal of Environmental Quality 7:59-62.
(7) CHRISPEELS, A. 1959. Larves de Chironomidae (Diptera: Nema-
tocera). pp. 137-188 In: Exploration Hydrobiologique des Lacs
Kivu, Edouard et Albert (1952-1954), Vol. III, fascicule 3. Insti-
tut Royal des Sciences Naturelles de Belgique. Bruxelles.
(8) CHUTTER, F M. 1972. An Empirical Biotic Index of the Quality
of Water in South African Streams and Rivers. Water Research
6:19-30.
(9) DAMAS, H. 1956. Etude Limnologique de Quelques Lacs
Ruandais. IV. Les Sediments et Leur Faune. Conclusions
G6nrales. Academie Royale des Sciences Coloniales. Classes
des Sciences Naturelles et M6dicales. Memoires in-8o Nouvelle
Serie. Tome III, Fascicule 4.
(10) DEBONT, A. F. 1956a. Controle Biologique des Mollusques d'eau
Douce et des Maladies Qu'ils Transmettent. Annales de la Soci-
ete Belge de Medecine Tropicale 36:667.
(11) . 1956b. Lutte Contre les Mollusques Dans les
eaux Africaines. Bulletin Agricole du Congo Belge 47: 337-380.
(12) . AND M. J. DEBONT HERS. 1952. Mollusc Control
and Fish-Farming in Central Africa. Correspondence 23 August.
Nature 170 (4321): 323-324.
(13) . AND M. J. DEBONT HERS. 1956. Haplochromis
mellandi Blgr. Poisson Malacophage (Fam. Cichlidae). Revue de
Zoologie et de Botanique Africaines 53:370-376.
(14) DESOWITZ, R. S. 1976. How the Wise Men Brought Malaria to
Africa and Other Cautionary Tales of Human Dreams and
Opportunistic Mosquitos. Natural History 85 (8):36-47.
(15) DEUTSCH, W G. 1990. Environmental Assessment of Ten Aqua-
culture Sites in the Marais of Rwanda. Auburn University, Inter-
national Center for Aquaculture and Department of Fisheries
and Allied Aquacultures. 
4 7 
pp.
(16) W. C. SEESOCK, E. C. WEBBER, AND D. R.
BAYNE. 1990. The Impact of Poultry Rearing Operations on
Water Quality and Biological Communities of Second Order
Streams in Cullman and Winston Counties, Alabama, 1988-89.
Auburn University, Department of Fisheries and Allied Aquacul-
tures. 
6 2 
pp.
(17) DEVOS, L. AND T. VAN DEN AUDENAERDE. 1990. Description de
Barbus cludinae sp. n. (Cyprinidae) Avec Synopsis des Grandes
Especes de Barbus du Rwanda. Cybium 14 (1):3-25.
(18) , J. SNOEKS, AND D. T. VAN DEN AUDENAERDE. 1990.
The Effects of Tilapia Introductions in Lake Ruhondo, Rwanda.
Environmental Biology of Fishes 27:303-308.
(19) Discover Magazine (Up Front). 1988. As If the Flu Weren't Bad
Enough. Vol. 9 (7):14.
(20) DURAND, J.-R. AND C. LtVIQUE. 1980. Flore et Faune Aquatiqes
de l'Afrique Sahelo-Soudanienne. Tome I. Paris. 
389
pp.
(21) . 1981. Flore et Faune Aqua-
tiqes de l'Afrique Sahelo-Soudanienne. Tome II. Paris. 
4
8
4 
pp.
(22) ELLIS, J. E., D. L. TACKETT, AND R. R. CARTER. 1978. Discharge
of Solids from Fish Ponds. Progressive Fish-Culturist 40: 165-
166.
(23) FAUSCH, K. D., J. R. KARR, AND P. R. YANT. 1984. Regional Appli-
cation of an Index of Biotic Integrity Based on Stream Fish Com-
munities. Transactions of the American Fisheries Society
113:39-55.
(24) GAMET, A., H. BROTTES, AND L. MVOGO. 1964. Premiers Essais
de Lutte Contre Les Vecteurs des Bilharzioses Dans les Etangs
d'une Station de Pisciculture au Cameroun. Bulletin de la Soci-
ete de Pathologic Exotique 67:118-120.
(25) HANOTIER, J. AND P. L. GIGASE. 1981. Note on a New Focus of
Schistosomiasis (S. mansoni) in Rwanda. Annual of the Belgian
Society of Tropical Medicine 61:93-98.
(26) HILSENHOFF, W L. 1977. Use of Arthropods to Evaluate Water
Quality of Streams. Wisconsin Department of Natural Resources
Technical Bulletin 100. 15 pp.
(27) . 1987. An Improved Biotic Index of Organic
Stream Pollution. The Great Lakes Entomologist 20:31-39.
(28) . 1988. Rapid Field Assessment of Organic
Pollution with a Family-level Biotic Index. Journal of the North
American Benthological Society 7 (1):65-68.
(29) HISHAMUNDA, N., AND J. MOEHL. 1989. Rwanda National Fish
Culture Project. International Center for Aquaculture Research
and Development Series No. 34. Auburn, Alabama: Department
of Fisheries and Allied Aquacultures, Auburn University. 19 pp.
(30) HUGHES, C. AND J. M. HUNTER. 1970. Disease and Development
in Africa. Social Science in Medicine 3:443-493.
(31) IRVINE, F. R. 1947. Fish and Fisheries of the Gold Coast. Crown
Agents for the Colonies. London. 352 pp.
(32) JACQUEMART, S. 1959. Trichoptere du Lac Albert. pp. 107-136 In:
Exploration Hydrobiologique des Lacs Kivu, Edouard et Albert
(1952-1954), Vol. III, fascicule 3. Institut Royal des Sciences
Naturelles de Belgique. Bruxelles.
(33) JOHNSON, W W AND M. T. FINLEY. 1980. Handbook of Acute
Toxicity of Chemicals to Fish and Aquatic Invertebrates. United
States Department of the Interior Fish and Wildlife Service
Resource Publication 137. Washington D. C. 98 pp.
(34) KARR, J. R. 1981. Assessment of Biotic Integrity Using Fish Com-
munities. Fisheries (Bethesda) 6:21-27.
(35) KOHLER, CH.C. AND J. G. STANLEY. 1984. A Suggested Protocol
for Evaluating Proposed Fish Introductions in the United States.
pp. 387-407 In: Courtenay and Stauffer (eds.). Distribution, Biol-
ogy and Management of Exotic Fishes. Johns Hopkins University
Press, Baltimore.
(36) LAIRD, M. 1988. The Natural History of Larval Mosquito Habi-
tats. Academic Press. London. 555 pp.
(37) LENAT, D. R. 1988. Water Quality Assessment of Streams Using a
Qualitative Collection Method for Benthic Macroinvertebrates.
Journal of the North American Benthological Society 7(3):222-
233.
(38) LITTLE, D. AND J. MUIR. 1987. A Guide to Integrated Warm
Water Aquaculture. Institute of Aquaculture Publications. Uni-
versity of Stirling. 
23 8 
pp.
(39) LOCKHART, J. D. F, R. B. HIGHTON AND J. P. MCMAHON. 1969.
Public Health Problems Arising Out of Man-made Fish Ponds in
the Western Province of Kenya. East African Medical Journal
46:471-480.
(40) MALEK, E. A. 1983. Impact of Fish Ponds on Public Health in
Rwanda with Special Reference to Schistosomaisis. Department
of Fisheries and Allied Aquacultures. Auburn University, AL
(prepared for United States of America Agency for International
Development). 
4
0 pp.
(41) MCCULLOUGH, F. 1981a. Appraisal of the Potential Use of Fish
for Control of Disease Vectors Other than Mosquitoes. Unpub-
lished World Health Organization document. TDR/BCV/IC.81.2
/WET22.
(42) _______ 1981b. Plan of Action and Support for Imple-
mentation of a National Schistosomiasis Control and Surveillance
Programme in Rwanda. Report on a Mission to Rwanda, May 11-
26, 1981. Unpublished World Health Organization document.
(43) MCMAHON, J. P., R. B. HIGHTON, AND T E DE. MARSHALL. 1977.
Studies on Biological Control of Intermediate Hosts of Schistoso-
miasis in Western Kenya. Environmental Conservation 4:285-
289.
(44) MOEHL, J. F. JR. (ed.). 1988. ICA technical Assistance to Rwanda
National Fish Culture Project Successfully Completed. ICA
Communicae 11 (1-2). Auburn University.
(45) ________ K. L. VEVERICA, B. K. HANSON, AND N.
HISHAMUNDA. 1988. Development of Appropriate Pond Man-
agement Techniques for Use by Rural Rwanda farmers. pp. 561-
568 In: R. S. V. Pullin, T. Bhukaswan, K. Tonguthai, and J. L.
Maclean (eds.), The Second International Symposium on Tilapia
in Aquaculture. ICLARM conference proceedings. Manila,
Philippines: International Center for Living Aquatic Resources
Management.
(46) OGUTU-OHWAYO, R. AND R. HECKY. 1990. The State and Known
Impacts of Fish Species Transferred to or Within Continental
Africa. Abstract from the Ecological and Genetic Implications of
Fish Introductions (FIN) Conference, held at the Great Lakes
Institute, University of Windsor, May 17-19. In: Li. H. W (ed.)
Newsletter of the Introduced Fish Section. American Fisheries
Society. 10 (2):7-8.
(47) PHELPS, R. P. 1981. Environmental Considerations in Aquacul-
ture. International Center for Aquaculture. Auburn University.
73 pp.
(48) SCHOLTISSEK, C. AND E. NAYLOR. 1988. Fish Farming and
Influenza Pandemics. Nature 331:215.
(49) SCHROEDER, G. L. 1980. Fish Farming in Manure-loaded Ponds.
pp. 
7 3
-86 In: R. S. V. Pullin and Z. H. Shehadah (eds.), integrated
Agriculture-Aquaculture Farming Systems. Proceedings of the
ICLARM-SEARCA Conference on Integrated Agriculture-
Aquaculture Farming Systems, Manila, Philippines, 6-9 August
1979. 258 pp.
(50) SLOOTWEG, R. 1989. Proposed Introduction of Astatoreochromis
alhsaudi, An East African Mollusc Crushing Ciclilid, As a Means
of Snail Control. pp. 61-64 In: M.-D. Crapon-de.Caproma and B.
Fritzsch (eds.). Proc. of the Workshop in Biology, Ecology and
Conservation of Cichlids. Ann. Kon. Mus. Mid. Aft., Zool.
Wetensch., Annales Museie Royal de l'Afrique Centrale, Sc.
Zoologie, Vol. 257. (Bielefeld, West Germany, 24-27 February
1988).
(51) SNOEKS, J., L. DEVOS, AND D. F E. THYS VAN DEN AUDENAERDE.
1987. Description de Haplochromis scheffersi sp. nov. (Pisces,
Cichlidae) du lac Kivu. Revue de Zoologie Africaine 100: 443-
452.
(52) SNOEKS, J., L. DEVOS, E. COENEN, AND D. THYS VAN DEN AUDE-
NAERDE. 1990. Description de Deux Nouvelles Especes d'Hap-
lochromis (Teleostei, Cichlidae) du Lac Kivu, Rwanda. Cybium
14 (1):63-76.
(53) VAN DEN BERCHE, L. 1939. Les Schistosomes et Les Schistoso-
moses an Congo Belge et Dans Les Territoires du Ruanda-Urun-
di. Memoires Institut Royal Colonial Belge (Section des Sciences
naturelles et medicales). Collection 8, Vol. VIII, 1
5
2 pp.
(54) WELCOMME, R. L. 1981. Register of International Transfers of
Inland Fish Species. FAO Fisheries Technical Paper 213. 120 pp.
(55) WIESMANN, D., H. SCHEID, AND E. PFEFFER. 1988. Water Pollu-
tion with Phosphorus of Dietary Origin by Intensively Fed Rain-
bow Trout (Salmo gairdneri Rich.). Aquaculture 69:263-270.
(56) World Health Organization. 1982. Biological Control of Vectors
of Disease. Technical Report Series No. 679. Geneva.
APPENDIX
CHAPTERS 3, 7, AND 8 FROM THE REPORT OF MALEK (1983) CONCERNING THE IMPACT
OF FISH PONDS ON PUBLIC HEALTH IN RWANDA
3. SUMMARY AND CONCLUSIONS
In order to assess the impact of fish ponds in Rwanda on the preva-
lence of scbistosomiasis and malaria, examination of ponds was carried
out in five prefectures, namely, Butare, Gikongoro, Gitarama, Gisenyi,
and Kigali. In the examination, emphasis was placed on the layout of
the ponds, the water supply, the fish species being cultured, presence
or absence of aquatic weeds and of bank vegetation hanging in the
water, species of snails and their infection with schistosomes and other
trematodes, presence or absence of mosquito larvae and use of the
ponds by humans. Chemical analysis of pond water was carried out in
several ponds.
It was revealed that feeder streams often have snails and aquatic
vegetation and no doubt they are the source of infestation of the ponds
with these organisms. Ponds with aquatic weeds, with bank vegetation
hanging in the water and with flotsam, vegetable matter, and debris
have more than one species of snails. On the other hand, well man-
aged ponds without aquatic weeds, where the bank vegetation is
trimmed, without debris or vegetable matter do not harbor snails or
mosquito larvae.
Several ponds harbor the snail intermediate hosts of Schistosoma
mansoni (Bioniphalaria pfeifferi and B. sudanica) and the snail hosts
of Schistosoma haematobium, (Bulinus (Physopsis) globosus and Buli-
nus strigosus). This is the first report of these bulinid snails in Rwanda.
Although no infection with S. mansard or S. haematobium was found
in these snails, their presence in the ponds creates the possibility of
transmission of schistosomiasis. Moreover, it is possible that snails may
be found infected at a different season of the year. Chemical analysis
of the water of some ponds indicated that fish ponds in several prefec-
tures can harbor snails and other molluscs at any time if they are intro-
duced in the ponds.
Fortunately, the majority of ponds examined in this study are not
used by the human population, at least at the time of the visit to these
ponds. People tend to use nearby streams, if they exist, because the
stream water is cleaner than the stagnant and fertilized ponds. Howev-
er, when the pond is at the edge of a town, such as the one at Nyabisin-
du, people tend to use it for various purposes, resulting in
human-water contact and possible infection with schistosomiasis.
Recommendations made, if followed, will minimize or possibly
eliminate the chances of infection with schistosomiasis and malaria.
Environmental means of snail and mosquito larvae control are strong-
ly encouraged. Removal of aquatic and bank vegetation 
and of all flot-
sam and debris should be done regularly to discourage snail and
mosquito larvae breeding. Ponds should be drained occasionially and
vegetation, mud on the bottom and the banks removed, and indenta-
tions along the sides of the pond should be eliminated. Because most
feeder streams were found to harbor snails and weeds, a small metal
screen with 16 meshes to the linear inch, should be installed across the
12
inlet of each pond to detain snails, weeds, and snail eggs adhering to
the weeds. These weeds when found should be removed so that they
do not interfere with passage of water. Introduction of malacophagous
fish and mosquito larvae-eating fish is encouraged. However, well-
controlled observations have to be carried out to determine the effica-
cy of these fish under Rwandan conditions. Ponds should always be
stocked with fish because such ponds stocked with even the species
which are presently being cultured, do not contain mosquito larvae.
All persons who have anything to do with fish ponds should have urine
and fecal examinations every six months to determine if they are
infected with schistosomes. If found infected they should be treated in
a nearby hospital with the available effective chemotherapeutic
agents. This examination should be done as soon as possible because
the snail intermediate hosts are already present in a large number of
ponds. Arrangements have been made with the Laboratory of Para-
sitology at the Medical School in Butare to carry out the examination.
A monitoring or surveillance and control program for snails and
mosquito larvae should begin as soon as possible, and should be con-
sidered an important activity of the Fish Culture Project. A part of the
training course for monitors, which will start at Kigembe in Septem-
ber 1983, should be devoted to the impact of fish ponds on the preva-
lence of schistosomiasis and malaria. Supervisors and agronomes also
should receive such training. If monitors find any snails or mosquito
larvae in their areas they should alert the supervisors who subsequent-
ly will contact the team leader and the fish culture extension specialist.
Control measures, mainly environmental, should be taken immediate-
ly, after which monitoring should continue every two months. During
the first year (1983-1984) emphasis will be placed on monitoring fish
ponds for snails and mosquito larvae in the prefectures of Butare,
Gikongoro, and Gitarama. In the second and third years, the monitor-
ing will be carried out in other prefectures. The fourth year will be
devoted to evaluation of the program.
This study indicates that fish ponds in Rwanda prove satisfactory in
terms of high fish productivity, without or with a minimal risk of trans-
mission of schistosomiasis and malaria, if they are well managed and if
a few simple and affordable control measures are undertaken. Howev-
er, a more detailed study will be required to verify that the disease vec-
tors can be controlled over time.
7. FIELD OBSERVATIONS OF FISH PONDS IN
RWANDA REGARDING WATER-RELATED DISEASES
Fish ponds in Rwanda, if not well managed, could be a health haz-
ard in transmitting schistosomiasis and malaria. Some viral diseases
transmitted by mosquitoes also could be contracted. However, the
ponds are not suitable habitats for the breeding of the larvae of black
flies, Simulium spp. which transmit onchocerciasis.
In Rwanda, fish ponds have been in existence for a long time and
thus are not a new environmental impact. It should be noted that in
any country the area used by fish ponds is very small compared to the
area containing natural streams, rivers, small and large impoundments
and swamps with which the human population is in contact. With
regard to the impact of fish ponds on public health, several factors
have to be taken into consideration, for example, their management,
their size and number, their attractiveness to the human population,
the vectors they harbor, the infection rates among these vectors, and
the prevalence of diseases among the human population.
Fortunately, the majority of ponds examined in this study are not
used by the human population, at least at the time of the visit to these
ponds. People tend to use the nearby streams if they exist, because the
stream water is cleaner than the stagnant and fertilized ponds. Howev-
er, when the pond is at the edge of a town, such as the one at Nyabisin-
do, people tend to use it for various purposes resulting in
human-water contact and possible infection with schistosomiasis. This
is especially true when there are only a few other alternatives in the
form of streams or rivers or piped water. Moreover, when fish ponds
are in the proximity of towns, the stream feeding the pond is usually
surrounded by dwellings which pollute the stream as well as the
ponds. Such is the case in the prefectural fish ponds at Kigali.
The present study of fish ponds in Rwanda showed that the majori-
ty harbor the snail intermediate hosts of the human and animal schis-
tosomes. Moreover, chemical analyses of water in several fish ponds
showed that all are suitable habitats for existence and breeding of mol-
luscs. Thus it is only a matter of time until all become infested with
certain snail species which they do not harbor at present. Although no
schistosome intermediate hosts collected were infected with the schis-
tosomes, certain studies on fish ponds in other African countries
revealed the presence of snails infected with the human schistosomes.
The snails collected during this study in Rwanda were all browsing
on plant surfaces rather than mud surfaces although they occasionally
occur on mud. If macrophytic vegetation is allowed to grow in a pond
or marginal vegetation overhangs into the water, snails are likely to be
found in large numbers. Another favorite substrate is provided by
decaying vegetable matter such as leaves or grass cuttings which fall
into the pond or may be deliberately thrown in as food for the fish,
often in excessive quantities.
Fortunately, the situation can be remedied with regard to both snail
and mosquito vectors of schistosomiasis and malaria respectively. Dis-
ease control should be a primary objective in pond management. We
can have fish ponds with minimal health hazards. Well managed ponds
in which aquatic vegetation and all debris are removed and marginal
vegetation are trimmed are not favorable habitats for snails and
mosquito larvae. Larvae of Anopheles gambiae and A. funestus are
usually found on mats of floating debris or in the semi-submerged veg-
etation at the edges of the water. This is a similar habitat to that for
snails. Snails and aquatic weeds also should be detained at the inlet of
each pond from the feeder stream, so that they will not be introduced
into the ponds.
It was observed during this study that when the pond is stocked
with fish, any species which are presently cultivated in Rwanda, then
there are no mosquito larvae. Evidently the larvae were being con-
sumed by the fish. Therefore the ponds always should be stocked with
fish until they are drained and dried out.
Important Snails Found In Rwandan
Fish Ponds
The important snail species which were encountered in the fish
ponds which were examined are the following:
Lymnaea natalensis
This is a common snail in many African countries and is the snail
intermediate host of the cattle and sheep liver fluke Fasciola gigantica.
None from fish ponds was found infected with this fluke, however,
some specimens collected from an irrigation ditch near Butare were
infected, and when the snails were isolated in water with grass blades,
metacercarial cysts of F gigantica were observed on the grass blades.
However, there were other trematode cercariae emerging from the
specimens collected from fish ponds.
In Zaire, this lymnaeid species in fish ponds was found infected
with a strigeid trematode of the family Diplostomatidae. The cercariae
after leaving the snails encyst in fish to form metacercariae. When the
fish is eaten by birds the metacercariae develop to the adult worm
Diplostomum sp. Infection with the metacercariae in the fish may
cause a serious effect on the fish population and results in low fish pro-
ductivity.
Thus snail control in fish ponds is not only for health reasons but
also to ensure high fish productivity by eliminating metacercarial
infections in fish.
Biomphalaria pfeifferi and B. sudanica
These planorbid snails, especially B. pfeifferi, are common in many
African countries. They are recognized and effective intermediate
hosts of Schistosoma mansoni, which causes human intestinal schisto-
13
somiasis. None were found infected with S. nmansoni in the fish ponds
that were examined. They might, however, be found infected at anoth-
er season of the year. The snails were found infected with other trema-
todes, not of medical importance.
B. pfeifferi and B. sudanica are intermediate hosts of another schis-
tosome Schistosoma rodhaini, a parasite of small rodents and dogs in
Zaire, Uganda, and Kenya. It occasionally infects humans in Zaire.
Whether it occurs in Rwanda is not known.
Bulinus (Physopsis) globosus and Bulinus (Bulinus) sp.
probably B. (B.) strigosus
These are intermediate hosts, in other countries, of Schistosoma
haematobium, the causative agent for human urinary schistosomiasis.
This is the first report of these snails in Rwanda, and they are prob-
ably the intermediate hosts of the few cases of S. haematobium found
recently in Rwanda. Early reports indicated the absence of these
bulinid snails and the absence of urinary schistosomiasis in Rwanda.
The only bulinid snail which was reported in Rwanda is Bulinus (Buli-
nus) forskalii, a different bulinid snail which is not an intermediate
host of urinary schistosomiasis in any African country.
B. globosus, and B. forskalii are intermediate hosts of another schis-
tosome, Schistosoma intercalatum, although related to S. haematobi-
urn, it produces human intestinal schistosomiasis in Cameroon, Zaire,
Gabon, and the Central African Republic. Whether it is found in
Rwanda is not known. It is possible that the case near Butare, which
was diagnosed as S. haenmatobium, where the eggs were found in the
feces, is caused by S. intercalatum.
The bulinid snails B. globosus and B. strigosus also serve as hosts for
Schistosoma bovis, a schistosome of cattle and sheep. It was reported
in Rwanda by van den Berghe (53), where it also infects antelopes.
Thiara tuberculata = Melanoides tuberculata
This species was found in fish ponds in Kigali and in Lake Ruhundo
near Ruhengeri. T tuberculata and another species T granifera have
been advocated as biological control agents of snails of medical or vet-
erinary importance. The observations in Rwanda (also in Senegal and
the Sudan) indicate that it coexists with biomphalarid, bulinid and
lymnaeid 
snails.
8. CONTROL OF SNAILS AND MOSQUITO LARVAE
IN FISH PONDS
Snail Control
In general, snails can be controlled by chemical, environmental and
biological measures.
Chemical Control
Most if not all the molluscicides available at present are toxic to fish
and are, therefore, unsuitable for use in fish ponds unless completely
harvested prior to the treatment period; a requirement not likely to be
met in most rural situations. Moreover, there is the problem of the
necessary training for the technical helpers who will apply the mol-
luscicide. There are some herbicides which also have molluscicidal
value and they are reputedly not toxic to fish and which may be of
potential use in fish ponds. However, they have not been tested under
field conditions; moreover, attention should be given to their effect on
the micro and macroflora of the ponds which constitute the food sup-
ply of the fish.
Certain parts and extracts of plants which grow in many African,
Asian, and South American countries have molluscicidal value, for
example, Sapindus saponaria and Phytolacca spp., but these also kill
fish. Most of them have generally become known through their use by
the local populations as fish poisons and thus their use should be pro-
hibited in fish ponds.
The conclusion is that chemical control by use of molluscicides is
not considered to be really satisfactory or feasible in fish ponds.
Environmental Control
Several malacologists working in Africa have observed that careful
maintenance of ponds will often prevent the buildup of large popula-
tions of snails. Weeds and overhanging vegetation should be regularly
removed. Occasional drying of ponds may have some effect on snails,
but it seems that this is only likely to be very temporary.
Biological Control
Predators and competitors of snails include insects, fish and other
snails. The most promising biological control measure in fish ponds is
stocking these ponds with mollusc-eating (malacophagous) fish.
DeBont (10,11) and DeBont and DeBont Hers (12,13) reported great
success with Haplochromis mellandi Blgr. (identified in a previous
publication as Serranochromis macrocephala Blgr.) in fish ponds in
Zaire. This fish is reported to be good eating. Another fish, Astatore-
ochromis alluaudi (Pellegrin) has been taken from Uganda to
Cameroon where pond trials have been carried out which showed that
snails were effectively controlled, and that A. alluaudi could be suc-
cessfully cultured together with T nilotica (2,24).
McMahon et al. (43) conducted an experiment in water impounded
by earth dams, for local water supply, in Nyanza Province, Western
Kenya. Control of snails was attempted by introduction of the mala-
cophagous fish Astatoreochromis alluaudi; other species (Tilapia zillii
and T leucosticta) also were introduced. One reservoir was left as a
control without any introduction of fish. Assessment of snail control
was made by scooping for snails, catches being expressed per man-
hour of effort. This was carried out both before and after introduction
of fish, over a total period of 15 years. The data indicated that A. allu-
audi did reduce the numbers of some species of snails, particularly
Biomphalaria pfeifferi (intermediate host of Schistosoma mansoni)
and, to a lesser extent and with less certainty, Bulinus spp. (some are
intermediate hosts of S. haematobium). The other two introduced fish
species, Tilapia ziUlii and T leucosticta, did not appear to be associated
with reduction in snail numbers. There can be no doubt, however, that
in this study Biomphalaria pfeifferi formed the principal diet of A.
alluaudi.
In Brazil, the fish Astronotus ocellatus ocellatus was very effective
in considerably reducing populations of the snail Biomphalaria
glabrata, intermediate host of Schistosoma mansoni. The capacity of
A. ocellatus and Tilapia rendalli in eliminating snails was also demon-
strated in Brazil. They are medium-size fish which are adapted to
ponds and reservoirs. Smaller predatory fish also exist which, although
unable to swallow adult snails of such species as Biomphalaria glabra-
ta and B. tenagophila, kill these snails by wounding their exposed soft
parts. These smaller species of fish include Hemichromis bimaculatus,
the jewel fish, Macropodus opercularis, the paradise fish, and Hap-
lochromis mellandi. The jewel fish, Hemichromis bimaculatus, when
introduced into a 75 m
2 
pond in Brazil infested with Biomphalaria
tenagophila, and which had a substantial area of shallow water, elimi-
nated the snails in six months. At the same time, itself increasing from
20 introduced specimens, to more than 500. A lake nearby populated
with a fish of the same family, Geophagus brasiliensis, showed no
infestation with snails (41). Astronotus ocellatus and Geophagus
brasiliensis are Brazilian in origin; Tilapia rendalli, Haplochromis
bimaculatus are exotic species but are found in certain bodies of water
in Brazil. Astronotus ocellatus swallows the snail whole, while Tilapia
spp. reduces or eliminates the vegetation which shields the snails and
their eggs.
Some species of voraciously herbivorous freshwater fish, for exam-
ple, the Chinese grass carp, Ctenopharyngodon idella, may play a use-
ful role in suppressing, but rarely eliminating, snail host population
densities, particularly in such lentic habitats as drainage and small
feeder canals in irrigation schemes, ponds, etc. In Egypt and the
Sudan trials using herbivorous fish, not only incidentally to suppress
snail host population, but more to control aquatic weed growth, are in
progress. On the other hand there are reports that other herbivorous
14
fish such as Tilapia zillii and T melanopleura were not effective in
eliminating snails by preventing the growth of macrophytic vegetation
in ponds and rendering them less suitable for the snails. McMahon et
al. (43) found that T zillii had little or no effect on snail populations in
the reservoirs which they studied in Kenya. Also at Mahiwa, Tanzania,
Berrie (5) found that fish ponds containing T melanopleura were gen-
erally fairly free from weeds whereas those containing T nilotica were
not, but the snails were still able to thrive on decaying leaves and other
rubbish which the fish did not seem to eat.
Mosquito Larvae Control
As is the case with snail control certain environmental and biologi-
cal measures have been recommended for the control of mosquito lar-
vae.
Lockhart et al. (39) after studying fish ponds in the North Nyanza
District of Kenya, recommended that the sides of the ponds should be
regular as indentations in the banks reduce wave action and provide
ideal habitats for mosquito larvae breeding. Vegetation on the banks
should be kept short because overhanging vegetation provides ideal
cover, especially for Anophelesfunestus larvae.
Biological control of mosquito larvae by the use of certain species of
fish has been advocated by many workers. Irvine (31) in Ghana, states
that Tilapia are effective predators of mosquito larvae. Others have
noted that ponds stocked with fish had few or no mosquito larvae as
compared with other ponds which contain no fish.
In a recent report on Biological Control of Vectors of Disease by
the World Health Organization (56) the use of fish in the control of
mosquito larvae has been advocated. In general, the use of indigenous
rather than introduced fish is encouraged. Oreochromis spilurus is to
be used operationally against Anopheles spp. Other examples of fish
with considerable promise include members of the families Cyprin-
odontidae (e.g. Aphanius spp., Aplocheilus spp. and Oryzias spp.),
Hemirhamphidae, Anabantidae, and Cichlidae (including Tilapia
spp.).
Several species of Gambusia spp. are operational in some areas to
control mosquito breeding, but may destroy local fish and therefore
should not be introduced into new areas without careful study of the
ecology and the fish fauna. Similarly, the use of other exotic fish is to
be avoided, with the notable exception of annual fish such as Notho-
branchius spp.
Several species of bacteria have been recommended for biological
control of mosquito larvae and blackfly larvae (vectors of onchocercia-
sis), but their use is still at an experimental stage. Among these bacte-
ria are Bacillus thuringiensis H-14 and Bacillus sphaericus. Also
certain species of fungi are lethal to mosquito larvae, but their use is
still at an experimental stage.
15