The Lake Victoria Species Survival Plan (LV-SSP) of the American Zoo and Aquarium Association (AZA) sent me to participate in field work in Uganda in April 1996. My assignment involved participation in ongoing surveys of haplochromines in Lake Victoria; development of exhibit aquaria at the Ugandan Fisheries Research Institute; development of in situ aquaculture of endangered cichlids; and recovery of founder stock of important haplochromine taxa from Lake Nawampasa (Kaufman, 1996).

The Lake Victoria Species Survival Plan represents an organized effort by North American zoos and public aquariums to conserve as much of the diversity of the endemic cichlids therein as possible. Captive breeding of endangered species plays an important role in these efforts. Unfortunately much of the cichlid species diversity (representing many entire trophic groups) had already gone extinct in Lake Victoria before founder populations could be secured for the program. Therefore, only a handful of species representing relatively few trophic groups and specializations are now being protected from extinction in breeding programs. Periodically, species in the program are reviewed as to the priority they should be given. To do so, the status of a species in the wild must be considered, using updated information collected during recent field efforts. The major limiting factor in the program is the number of participating institutions and the aquarium space each can make available.

Much of the existing debate concerning the future direction of LV-SSP activities centers around the best ways to maintain genetic integrity of the fishes in the program. Species kept in aquaria or ponds will over time become genetically distinct from wild founder stocks, a situation that is not particularly surprising given the very different selection pressures operating on captive fishes. One option that would maintain these fishes under natural selection pressures and eliminate the need for costly, labor-intensive efforts in North America and Europe is to locate and protect satellite lakes that still support important taxa. Satellite lakes are small, isolated bodies of water left behind when water levels in the Lake Victoria basin dropped to present levels. Provided these lakes are granted protection from environmental degradation and overexploitation, they could serve as sanctuaries for native cichlids as well as other endemic flora and fauna of the basin. Ugandan and American researchers have recently discovered that one of these satellites, Lake Nawampasa, is home to haplochromine taxa representing trophic specializations that have been extirpated from Lake Victoria. This provides an incredible opportunity, at the very least, to secure founder stocks for these taxa, and at best, to preserve a natural "living museum" that grants a second chance to fishes already thought lost forever. Because of its importance to the LV-SSP, and the fact that Lake Nawampasa is already under exploitation pressures from ornamental fish exporters, a reconnaissance of this lake was a priority of my trip to Uganda.

This year's field work (1995) began at the British Museum of Natural History in London during a lengthy stopover on my way to Entebbe (Uganda). Here I met up with Dr. Mark Chandler of the New England Aquarium in Boston to study preserved specimens of Haplochromis species that we were unfamiliar with but expected to encounter in Lake Victoria and Lake Nawampasa. Admittedly, the viewing of preserved specimens became more meaningful only once we were able to compare them with live fishes from the lake.

After arriving in Entebbe, we drove eastward for three hours along the north shore of Lake Victoria to the Fisheries Research Institute (FIRI) in Jinja. Once we were settled in a comfortable guest house (only a 10-minute walk from FIRI), work plans were established for the next fortnight. The first week would be spent sampling in the Napoleon Gulf of Lake Victoria near Jinja; the second would be consumed in building aquaria at FIRI and collecting fishes in Lake Nawampasa.

The first week afforded us an opportunity to observe the many changes that have taken place in the lake ecosystem over the last three decades. Hillsides surrounding the Napoleon Gulf are being deforested to produce charcoal fuel. This increases run-off of silt into the lake, which in turn causes increased turbidity. More importantly, the increased inflow of nutrients has resulted in significant eutrophication. The introduced water hyacinth (Eichhornia crassipes) is now common in almost all areas of the gulf, a clear indication of the increased nutrient levels in the lake. Aside from being an indicator of nutrient levels, water hyacinth acts as a barrier to the exchange of oxygen at the air-water interface in areas where it collects as a result of wind. Unfortunately, this often happens to be in smaller bays and along shorelines, typically the most productive habitats for haplochromine cichlids.

In a 6-m wooden canoe fitted with an outboard motor, we visited Mark's five study sites in the Napoleon Gulf, reflecting the major habitats represented: papyrus shorelines; intermediate zones with mixed rock and sand substrates; and fringing swamps. Dr. Chandler's research requires regular sampling of these sites with gill nets. At each site, one net is set along the shoreline in shallow water; a second is set parallel to the first about 20 meters offshore; and a third is set ca. 200 meters offshore parallel to the first two. Nets are composed of five panels, each with a different mesh size, for sampling fishes of varying sizes. At each site, temperature, dissolved oxygen, and Secchi disk (water transparency) measurements were taken. The nets are left in position for twelve hours but are checked twice during this period, once after about eight hours and then a second time just before they are moved to the next site. As the nets are checked, any haplochromines caught are identified, measured, and recorded. For less commonly encountered species or fishes that we could not positively identify, tissue samples and photographs were also taken. The relatively-abundant characoid Brycinus sadleri, Nile perch (Lates niloticus), and introduced tilapias, particularly the Nile tilapia (Oreochromis niloticus), were measured and counted, then set aside by the Ugandan crew for further studies of a culinary nature!

By week's end it was Mark's impression that there were fewer haplochromines at the sites we sampled than in previous years. The lake's rising turbidity increases the efficiency of Nile perch predation on the visually-oriented cichlids better adapted for life in clear water. However, the effects of water hyacinth on dissolved oxygen levels in inshore habitats may be beneficial to cichlids in that haplochromines can take advantage of the cover in these habitats despite the low oxygen, whereas Nile perch are intolerant of such oxygen levels. This would explain the significantly higher numbers of cichlids caught in the shoreline gill nets as opposed to the nets offshore. While there is little doubt that the lake is still responding to environmental changes and fishing pressures, it is certain that the diversity of haplochromine taxa that once flourished in the lake is now greatly reduced.

Part of FIRI's commitment to conserving the cichlid biodiversity of the region is the development of aquaria and pond aquaculture for education and captive breeding purposes. I suggested using the more secure indoor aquaria to display and breed Lake Nawampasa cichlids and the outdoor ponds to rear offspring. By this arrangement, valuable brood stock can be protected; reproduction can be carefully controlled; and juveniles would have the requisite space for proper growth.

At the start of our second week, we made a trip to Kampala, the capital, to order and pick up glass needed for the planned exhibit aquaria to be built and housed at FIRI. A room (6 m x 4 m) with a sink, drainage, electrical power, and windows (to control ambient light) accomodated a metal stand constructed to hold twelve 150-l aquaria. The arrangement of these tanks permits further expansion as resources become available and husbandry skills develop. A primary aquarist or technician and a back-up person will staff the aquarium room for weekends, holidays, and other times when the primary person is not around. To make the facility complete, a supply of such tools as buckets, nets, siphon hoses, fish foods, medications, and resource books are needed. While we were setting up this facility, we also made plans and took the initial steps to construct concrete lined rearing ponds on the "back lawn" at FIRI. They were designed to operate on a gravity flow system and would be built above ground to facilitate husbandry work.

Lake Nawampasa is small (ca. l square kilometer) and very shallow with a maximum depth of only three meters. The lake is accessible by dirt road although a 4-wheel drive truck is required to negotiate the route given its poor condition. There is a village alongside the lake whose residents are primarily involved in ranching. The area around the lake has little relief so the lake (which is surrounded by tall fringing grasses) is difficult to locate from a distance. One almost needs to get wet feet to know where the lake actually starts!

Perhaps the dominant (certainly the most impressive) physical feature of Lake Nawampasa is the extensive fringe of submerged and emergent macrophytes in the littoral zone, which provides habitat and food for an abundance of resident cichlids. The dominant submerged macrophyte appeared to be Ceratophyllum sp. (hornwort), although Nymphaea sp. and Potamogeton sp. were also readily observed in the clear, shallow waters. In places the submerged macrophytes were so dense that it made travel in the canoe difficult, in contrast to the inshore zones of the Napoleon Gulf of Lake Victoria where I observed no submerged macrophytes. The latter condition is due at least in part to the presence of the introduced Tilapia zilli, which has been described as "a goat with fins" based on its habits of foraging on vegetation.

The water of the littoral zone appeared clear and was not stained despite abundant plant debris on the bottom. The substrate in this zone was predominantly a combination of mud and plant debris. The littoral zone conditions contrasted sharply with the turbid, seemingly macrophyte free, open waters of offshore areas.

Geological conditions around Lake Nawampasa are dramatically different from the lateritic soils found in the Jinja area of Lake Victoria. Lake Nawampasa is situated on an ancient shield, responsible for the grey, sandy soils, which also makes its presence known by the occasional massive rock outcrop. It was not surprising to find dissimilar water chemistry from the values observed at Jinja. When tested with a Tetra Laborett kit, the pH of Lake Nawampasa was 7.3 (range: 7.07.5) with a KH of 11 DH. Dissolved oxygen levels were near saturation at most sites where measurements were taken, but this could depend on the amount of exposure to wind and the density of submerged macrophytes.

During our second week, we made two trips to Lake Nawampasa with our Ugandan colleague, S. B. Wandera. Our objectives were to survey the habitat and collect founder stocks for specific Haplochromis taxa not represented in the LV-SSP captive breeding program. On the first trip we arrived late in the afternoon; while some of the FIRI crew set up camp, we had a local take us onto the lake. The canoe we used seemed to have been designed to float at the expense of any ability to keep out water and lacked all but the absolute minimum of stability. However we did manage to complete our work on (rather than in) the lake! Several inshore and offshore sites were identified and sampled using unbaited minnow traps. Also at these sites, temperature and dissolved oxygen measurements were taken. The following morning the minnow traps were checked and removed. The only fish found in the traps were juvenile cichlids less than 3 cm TL. The first trip ended with an unfortunate accident that resulted in the demise of the fish captured from the lake. The plug in the transport cooler came out during the 3.5 hr drive back to Jinja along less than perfectly level roads. Luckily, our second trip later in the week was far more successful.

When we arrived at the lake for the second time, we were greeted by young men eager to fish for us. The local fishermen were very productive. In no time at all, they brought us shallow, plastic tubs containing not more than a few liters of water and up to thirty adult cichlids. The tubs were at times covered with a waterlily leaf to maintain water temperature. Despite this, temperatures in the tubs rose very quickly with an obvious effect on dissolved oxygen content, evidenced by the fish "gasping" at the surface.

We were interested in only a few of the taxa caught by the fishermen due to our limited transport and shipping space and the priorities of the LV-SSP. Species of most interest were: Allochromis welcommei, Lake Victoria's only known lepidophage (scale-eater); any Lipochromis species; Haplochromis barbarae; any Prognathochromis, including species of the subgenus Tridontochromis; and any Harpagochromis species. Sorted from the fish tubs were a few specimens of Lipochromis, a paedophagous (embryo-eating) species; several piscivores of the subgenera Prognathochromis and Tridontochromis; and an epiphytic algae-grazing Haplochromis species. Other cichlid species caught included: Pseudocrenilabrus victoriae, Astatoreochromis alluaudi, Oreochromis variabilis and O. esculentus.

We made every effort to sort the fishes as quickly as possible, determining which could be kept and bought and which we would try to convince the fishermen to release back into the lake. At first the concept of releasing unwanted fish made little sense to them. They said if we did not want the fish the first time, releasing them only meant that they could catch the unwanted fish a second time, wasting their effort. The local solution to this problem was to discard the unwanted fish, ensuring that only new, potentially saleable fish would be caught on subsequent trips to the lake. It took some very creative buying practices on the part of S. B. Wandera of FIRI to finally get the fishermen to release unwanted fishes back into the lake.

We were not able to witness the collecting methods used by the locals. Also, we were not present at the lake when the Kampala exporters were, and therefore cannot be certain whether they utilize the locals or implement more destructive collecting methods, such as beach seines. Regardless, based on the activities we observed lakeside, there is a need for education programs aimed at locals concerning fish conservation and proper handling methods, as they are involved in the collection of ornamental fishes. It is critically important that the ornamental trade in these fishes not jeopardize the undisturbed natural habitat that this satellite lake represents.

We packed the fish in oxygen and double-bagged them for the trip back to FIRI. We also loaded several large plastic containers with Lake Nawampasa water for changes and repacking before the intended shipment to New York. The next morning (the last day of the trip) we checked the fish. Although they had produced a lot of waste in the bags, almost all had survived. The fish were given a change of water, reoxygenated and rebagged in preparation for shipment. Unfortunately, upon checking in at the airport in Entebbe, we discovered that the fish could not be shipped as baggage, and it was too late to check them as freight! Therefore the fish were taken back to FIRI to be shipped at a later date. This (the only disappointment of the trip) proves that one cannot always go on past experiences when dealing with an airline. Especially in Africa!

All things considered, the trip was productive, enjoyable, and a valuable learning opportunity for me. The staff at FIRI were excellent co-workers and hosts. The flora and fauna of even this small part of the Lake Victoria basin were amazing with an almost overwhelming diversity of birds, frogs, spiders, and flowering plants and trees. In Lake Nawampasa this wealth of diversity extends to the endemic cichlids present and will hopefully continue to do so.

Unfortunately the proximity of Lake Nawampasa to Lake Kyoga does not bode well for the status quo. Connected to Lake Victoria by the Nile River, Lake Kyoga has also been affected by water hyacinth and introduced fish species. There is only a short (50 meter), relatively low (l meter) land bridge presently separating Lake Kyoga and Lake Nawampasa. Locals drag their boats regularly from one lake to the other during the course of fishing and travelling in the area. It has been speculated that Nile perch (Lates niloticus) could not survive in Lake Nawampasa for various reasons (e.g., diurnal temperature fluctuations; dissolved oxygen profiles; lake size). However, Tilapia zilli and water hyacinth would no doubt find Lake Nawampasa a perfectly suitable habitat. The voracious T. zilli could literally eat the haplochromines of the lake "out of house and home," while the hyacinth could in a short time cover the entire surface of the lake. Certainly a knowledge of what has happened to Lake Victoria burdens all of us with the responsibility to prevent Lake Nawampasa from suffering a similar fate.

References Cited
Kaufman, L.; 1996; Lake Nawampasa, Uganda; where the rubber hits the road for aquarist participation in cichlid conservation. Cichlid News 5(4):17-21.