Hydrobiologia 452: 79–88, 2001. © 2001 Kluwer Academic Publishers. Printed in the Netherlands.

79

The impact of water hyacinth, Eichhornia crassipes (Mart) Solms on the abundance and diversity of aquatic macroinvertebrates along the shores of northern Lake Victoria, Uganda Wanda Fred Masifwa1 , Timothy Twongo1 & Patrick Denny2,∗ 1 Fisheries

Research Institute, P.O. Box 343, Jinja, Uganda Fax: 256-43-130123/121322. E-mail: [email protected] 2 IHE, Delft, P.O. Box 3015, 2601 DA Delft, The Netherlands Fax: +31-15-212-2921. E-mail: [email protected] (∗ Author for correspondence) Received 5 July 2000; in revised form 8 February 2001; accepted 19 February 2001

Key words: water hyacinth, impact, abundance, diversity, aquatic macroinvertebrates, Lake Victoria

Abstract This study examined the impacts of the alien waterweed, water hyacinth, on the abundance and diversity of aquatic macroinvertebrates in the littoral areas of northern Lake Victoria in Uganda. The weed had undergone explosive growth on the lake causing serious disruption to people, the economy and the ecosystem. This study was confined to impact of the weed in the littoral zone, not to the large floating mats of vegetation which float across the lake and clog large areas of shoreline. The littoral area studied comprised of fringing mats of Eichhornia crassipes (Mart) Solms (water hyacinth) to the lakeward of Cyperus papyrus; water hyacinth mats undergoing colonisation by Vossia cuspidata (Roxb.) Griff.; and a typical Cyperus papyrus L shore with no outer floating mat of water hyacinth. Numerical abundance (Nos. m−2 ) and diversity (No. of taxa) of macroinvertebrates recovered from pure Eichhornia crassipes and the Eichhornia-Vossia succession increased from the fringe of the Cyperus papyrus towards the open water. In the typical Cyperus papyrus fringe, in the absence of water hyacinth, abundance was highest at the papyrus/open water interface and dropped off sharply towards open water. The Shannon–Weaver diversity index (H ) of macroinvertebrates decreased progressively from pure Eichhornia crassipes stands, to Vossia/Eichhornia beds and Cyperus papyrus stands (H =0.56, 0.54 and 0.34, respectively) but were not significantly different. Dissolved oxygen decreased from open water into vegetation where it approached anoxia. Water hyacinth appeared to enhance the abundance and diversity of aquatic macroinvertebrates at the interface with the open water. The impoverished abundance and diversity of the macroinvertebrates deeper into the vegetation mats suggested negative environmental impacts of the water hyacinth when the fringe is too wide. Further research is recommended to establish the optimum width of the fringe of stationery water hyacinth that promotes maximum abundance and diversity of aquatic macroinvertebrates and, possibly, of other aquatic life. Since this study in 1997, there has been a dramatic decrease in Eichhornia infestations and by June 2000 it appeared largely to exist only as fringing vegetation in bays and inlets.

Introduction Eichhornia crassipes (Mart) Solms (water hyacinth) is an obnoxious, surface-floating waterweed of the tropics. A large body of literature addresses its growth

and methods for control (Mitchell, 1985; Gopal, 1987; Harley et al., 1997) whilst relatively little investigates its ecological interactions. This is hardly surprising as the exceptionally high growth rate of water hyacinth (Harley, 1990) causes physical obstruction of water-

80 ways, hydropower generation, water abstraction units, etc. which has serious socio-economic consequences often requiring immediate action. Water hyacinth was introduced into Africa from South America in the early 1900s (Mitchell, 1985; Gopal, 1987), but since the 1950s it has caused problematical weed infestations in Southern Africa, the Congo basin and in the Upper Nile (Rzoska, 1974; Denny, 1984). An aerial survey to establish the water hyacinth infestation on Lake Victoria in Uganda was undertaken in 1989 (Taylor, 1993). Results of this survey created subsequently a major politico-socio-economic challenge for its management and control (Twongo & Balirwa, 1995, Twongo et al., 1995). Its proliferation in the East African waterways, which are a primary source of fish for food protein, income and employment amongst the riparian communities, has highlighted the need for more ecological studies (Twongo & Balirwa, 1996). The weed forms a permanent floating fringe (often replacing the obligate acropleustophyte, Pistia stratiotes L.) at the highly productive wetland/open water interface zone (Denny, 1991), alters the food web (Balirwa, 1995, 1998) and affects biological diversity. In some places, the mats have become consolidated into the plant succession by the fringing grass Vossia cuspidata (Roxb.) Griff. (hippograss) which penetrates and stabilizes Eichhornia floating vegetation (Howell et al., 1988; Balirwa, 1995). The abundance and diversity of aquatic macroinvertebrates in the interface zone are crucial links in the food chains of littoral fish (Balirwa, 1995). Green et al. (1976) reported a range of microand macro-invertebrates associated with the roots of Eichhornia on a crater-lake, Rana Lamongan, in Indonesia. In East Africa, Bailey & Litterick (1993) described the invertebrate communities in the fringing vegetation of the Sudd swamps of the Upper Nile following colonization by water hyacinth in the 1950s. The study of the invertebrate fauna on Pistia roots in Lake Volta, Ghana, (Petr, 1968) makes an interesting comparison. Our research investigates the impact of water hyacinth on the macroinvertebrate communities of the interface zone making comparisons between those associated with pure stands of Eichhornia, with the Vossia cuspidata/Eichhornia plant succession and with the Cyperus papyrus/open water area beyond the vegetated zone. Since this study, there has been a dramatic decline in the water hyacinth biomass following the introduction of two species of weevils, Neochetina eichhorniae and N. bruchi. In June 2000 in Uganda, it

was only seen as a thin fringe in sheltered bays around the lake (Denny, pers. ob.).

Materials and methods Study area and sampling sites Field investigations were undertaken in the dry season from October 1996 to March 1997 along the northern shores of Napoleon Gulf, Lake Victoria, Uganda, near Jinja (Lat. 0◦ 22 N; 0◦ 30 N & Long. 33◦ 10 E; 33◦ 26 E) (see Fig. 1). Sampling sites were selected to compare three shore environments namely one under the cover of stationary, floating fringes of pure Eichhornia crassipes, the second was covered by floating mats of the waterweed interlaced with the strongly rhizomatous grass, Vossia cuspidata and the third, a floating-weed-free zone. At all three locations, emergent vegetation was dominated by Cyperus papyrus at the back. The pure stands of Eichhornia were characterized by shelter from offshore, but exposed to onshore, winds that encouraged accumulation of the mats. The second location was similar with respect to the wind regimes but the floating mats of Eichhornia crassipes were undergoing colonisation by Vossia cuspidata that stabilised the vegetation mat. The third location was swept by along-the-shore winds that precluded establishment of Eichhornia crassipes. At each of the three locations, sampling was at 10 m intervals along three parallel transects set at roughly right angles to the shoreline, approximately 5 m apart. Access into the floating beds of Eichhornia crassipes and Vossia-Eichhornia was facilitated by wooden planks at least 25 cm wide buoyed up by the vegetation. Triplicate samples of macrophytes and bottom sediment were collected at each sampling site. Measurements of dissolved oxygen were made under the water surface and close to the lake bottom, using a portable oxygen meter (Microprocessor Oximeter, WTW OX196 SN. 42079061). Sampling and sorting out macro-invertebrates A square quadrat of 0.25 m2 was laid over the floating vegetation. A large, strong nylon net (0.04 mm pore size) mounted on a rectangular metal frame connected to a long handle was placed carefully underneath the floating vegetation and the rhizomatous root mass was cut at the water surface around the edges of the quadrat using a pair of shears. The cut block of vegetation with the intact root mass was lifted carefully out

81

Figure 1. Map of northern Lake Victoria, Uganda, showing sampling sites.

of the water with the help of the net, which trapped any macroinvertebrates dislodged by the operation. The 0.25 m2 blocks of vegetation were collected in triplicates and placed in separate polythene bags for sorting in the laboratory. The macroinvertebrates in the bottom sediments were sampled in triplicate using a Ponar grab sampler with a jaw area of 0.02 m2 . The bulk of the macroinvertebrates in the samples were dislodged from the root masses by washing the material with tap water through sieves of 0.2 mm and 0.04 mm. The root material was then flushed with 95% alcohol to excite and dislodge any remaining organisms. Larger macroinvertebrates were picked up with forceps and preserved in 10% formalin in vials. Sediments were washed with tap water through identical sieves. The fine sediment residues containing the smaller organisms were spread into large white

plastic trays where drops of 95% ethanol were added to excite them. They could then be seen with the aid of a hand lens and removed with forceps for preservation. The contents of the vials were flushed with tap water through a 0.04 mm sieve to wash off the preservative before sorting into taxonomic groups and enumeration. Species diversity With the limited resources for identification of macroinvertebrates, identifications were taken to the family level (Peckarsky et al., 1990; Thorp & Covich, 1991). To determine the diversity of macroinvertebrates, the Shannon–Weaver Diversity Index (H ) was applied on the families using the formula: H  = −i pi lnpi ,

82

Figure 2. Abundance of macroinvertebrates from the edge of the papyrus zone, through the floating vegetation where it exists, towards the vegetation/water interface and open water. (a–c) Abundance of invertebrates in the bottom sediments: (a) below the Vossia/Eichhornia vegetation, (b) beyond the papyrus zone deplete of floating vegetation towards open water and (c) below Eichhornia crassipes. (d–f) Abundance of invertebrates amongst the root mass of (d) Vossia/Eichhornia (e) papyrus and (f) Eichhornia crassipes. Note the differences in the vertical scales between the upper and lower histograms.

83 where pi is the proportion of all the macroinvertebrates which belong to the ith species.

Results Abundance of macroinvertebrates The results reported here are based on the mean abundance (Nos. m−2 ) of macroinvertebrates found both in the sediment below and amongst the macrophyte root mats. Figure 2a–c and 2d–f show the mean abundances of macroinvertebrates from sediments and root mass, respectively (from triplicate samples at each site) along the three transects in each of the three environments. The outer (lakeward) edge of Cyperus papyrus is taken as the starting point for each transect and the distances of the sample sites are measured from off these points, through the floating fringe of vegetation, towards open water. In general, there was an increasing trend in the abundance of macroinvertebrates from the root mats at the edge of the Cyperus papyrus zones through the floating, stationary weed stands towards the open water interface (see Fig. 2d,f). Where the papyrus fringe was not capped by floating mats of Eichhornia (the weed-free environment) Cyperus papyrus had a similar abundance of macroinvertebrates as the outer (lakeward) zones of weed stands (see Fig. 2e). No significant differences (t-test, p=0.05) were found in the abundance of macroinvertebrates at the open water interface with the three vegetation stands. Generally, the stationary root masses of water hyacinth contained a higher and more evenly distributed abundance of macroinvertebrates (see Fig. 2f) than stands of Vossia/Eichhornia (see Fig. 2d). All the sediment samples contained lower numbers of macroinvertebrates than the root mats (see Fig. 2a–c) and no clear pattern of distribution was discernable. Figure 3 a–c show percentage composition of the macroinvertebrate taxa at the interfaces of the three vegetation stands with the open water. Outer fringes of floating weed stands were dominated by Chironomidae (50% or more), whilst Ephemeroptera constituted the highest percentage (50%) at the Cyperus papyrus/open water interface. The two gastropod genera, Biomphalaria and Bulinus, vectors of bilharzia, were encountered in the three environments (see Fig. 4a,b). There was a general tendency for the two genera to increase in abundance from the Cyperus papyrus fringe at the back of the stationary, floating weed stands towards the open water interface especially in the Vossia/Eichhornia

Figure 3. Percentage composition of major macroinvertebrate taxa in the root mass at the vegetation/open water interface of (a) Vossia/Eichhornia vegetation (b) papyrus and (c) Eichhornia crassipes.

succession. At the Cyperus papyrus/open water interface in the Eichhornia-free environment, Bulinus dominated; Biomphalaria favoured especially the pure Eichhornia root mats. Shannon–Weaver diversity of macroinvertebrates The Shannon–Weaver diversity indices, based on the families, indicated that a more diverse macroinvertebrate fauna occurred in stands of pure water hyacinth and Vossia/Eichhornia than at the open water interface with Cyperus papyrus. The overall diversity index was slightly higher in the water hyacinth stands than in the Vossia/Eichhornia beds and lowest at the Cyperus papyrus interface (see Fig. 5) but no significant differences (p=0.05) were realised among the diversity indices of the three habitats.

84

Figure 4. Abundance of bilharzia vectors at the different habitat types: (a) Biomphalaria and (b) Bulinus.

Dissolved oxygen Sub-surface dissolved oxygen concentrations of the water under the weed mats were higher than those close to the lake bottom and there was a general trend towards decreasing oxygen concentrations from the outer to the inner zones of floating vegetation (see Fig. 6a,b). Off-shore, sub-surface oxygen concentrations were at approximate saturation (∼8 mg l−l ); at

the outer fringe they varied depending upon wind and current but were generally between 4 and 5 mg l−1 and in the back zones of floating vegetation, they were near to zero. Anoxic conditions occurred close to the bottom sediment. No significant differences (p=0.05) in sub-surface dissolved oxygen were found at the open water interfaces between Vossia/Eichhornia and the Eichhornia stands.

85

Figure 5. The Shannon–Weaver diversity indices at the outer vegetation fringes of the three habitat types.

Figure 6. Dissolved oxygen concentrations recorded from the edge of the papyrus zone, below the floating vegetation towards the vegetation/open interface, to the open water, (a) sub-surface oxygen and (b) concentrations close to the bottom sediment.

Discussion The establishment and proliferation of Eichhornia crassipes on Lake Victoria has had devastating repercussions on the economy of the region and, more particularly, on the riparian people: some two million families. Thankfully, now, the explosive growth of the weed is abated following the introduction of

weevils as a biological control strategy. This study was undertaken at a time of exponential growth of the weed. It had formed freely floating, wind-blown and current-driven rafts of vegetation some two kilometres in diameter and had become established around the shoreline and in the numerous bays and inlets as semi-permanent, stationary, floating fringe vegetation.

86 We confined our study to the macroinvertebrate populations of the fringes of the swamps; not to the free-floating, mobile rafts of vegetation. The convolutions of fringe vegetation increase naturally the contact surface area of the interface with the lake water and is vital for the lake’s littoral ecology (Balirwa, 1998). The oxygen-rich wetland/open water interface is a species-rich and highly productive environment (Denny, 1991) but the Eichhornia must have altered the ecology of the zone. Do the root mats of Eichhornia provide additional habitats for the macroinvertebrates thereby increasing biodiversity and abundance, or do they stifle existing populations? Does the root mat of Eichhornia, an alien species, host a higher or lower diversity of macroinvertebrates compared to indigenous flora? Eichhornia will not be eradicated: what will be the long-term impact on the littoral macro-invertebrate fauna? Inevitably, sampling in such unstable (and hazardous) conditions is liable to considerable experimental errors: our relatively high error bars from nine replicates at each sample site bear witness to this and statistically significant differences are few, but trends are identifiable and patterns discernible. Abundance and diversity of macroinvertebrates Mean abundance of macroinvertebrates varied widely within each vegetation bed and between the different habitat types. At the open water/macrophyte interface, the beds of water hyacinth supported higher populations of macroinvertebrates than the consolidated Vossia/Eichhornia areas or the Cyperus papyrus/open water fringes. This suggests no particular aversion of the indigenous macroinvertebrates to the alien plant species. Petr (1968) showed that in root mats of Pistia stratiotes on Lake Volta, Ghana, numbers of invertebrates increased as the number of plants decreased through a season from a minimum of less than 3000 m−2 to a maximum of 15 000 m−2 . His maxima were some threefold higher than those we found on Eichhornia in Lake Victoria. This may suggest a preference for Pistia but we have no direct evidence for this as Pistia had been practically eradicated by Eichhornia from the Napoleon Gulf. Care must be taken with comparisons of data from different surveys in other regions as methodologies vary and season clearly affect abundance (Petr, 1968; Bailey & Litterick, 1993). In Lake Victoria, the surface-floating fringe vegetation at the open-water interface supported a richer abundance and diversity of macroinvertebrates than

the same vegetation rooted into the bottom mud indicating the value of the floating root mats as habitats for aquatic invertebrate fauna. The floating, vegetated interface had more macroinvertebrates than the adjacent weed-free zones, which emphasizes the importance of the plants for the provision of niches. Other studies have come to similar conclusions (Petr, 1968; McLachlan, 1969; Green et al., 1976; Whiteside et al., 1978; Bailey et al., 1978; Cyr & Downing, 1988; Kornijow, 1989; Hargeby, 1990; Brown & Lodge, 1993). It has been suggested that the greater niche diversity provided by plant biomass, periphyton, detritus and silt collected by root mats, especially floating root mats, altogether support a more complex and richer food-web structure than the bare bottom sediment (Hargeby, 1990; Brown & Lodge, 1993). However, Kornijow (1989) believes that the associations of macroinvertebrates with macrophyte roots are not due to trophic relationships (since the plant tissues are a direct food to a very small fraction of the invertebrate fauna) but merely to the provision of a substratum. There was a marked reduction of numbers of macroinvertebrates from the outer edges of vegetation into the floating mats. Bailey & Litterick (1993), from their studies of the newly established water hyacinth fringe vegetation of the Sudd, Upper Nile, recorded lower mean counts of invertebrates but trends were similar with higher numbers (735 m−2 ) at the outer edge of the fringe than 6 m shorewards (182 m−2 ). The reduction in abundance and diversity of macroinvertebrates along our transects was correlated with a changing environment, especially with a reduction in dissolved oxygen concentrations. Like Bailey & Litterick (1993), we believe that the oxygen concentrations regulated the invertebrate distributions. In Lake Victoria, the macroinvertebrate taxa, Chironomidae, Oligochaeta, Hirudinea and Coleoptera, tolerant of oxygen-poor conditions, were commonly found in the inner zones of the vegetation beds. Willoughby et al. (1993) felt that water hyacinth was probably a threat to biodiversity and abundance of fish and invertebrates in Lake Victoria. Our findings generally provide a note of optimism for the littoral ecosystem. Where the alien weed occurs as a narrow fringe of vegetation (in contrast to a blanket over the water surface) it provides a rich habitat for a diversity of macroinvertebrate communities. There is evidence that it also provides a refuge for fish fry (Balirwa, 1998; Wanink, 1998). Its contribution to the diversity of littoral habitats as a refuge for many

87 aquatic species, and hence to the food webs thereof, may be of benefit to species diversity and fish production in the lake. Indeed, at present (June 2000), it is no longer a troublesome weed and the latest reports from local fishermen suggest that fish catches (especially Oreochromis niloticus) in this zone are improving. Some cichlid fish species that were assumed to have been predated to extinction by Lates niloticus (Oguta-Ohwayo, 1990, 1995; Witte et al., 1992) are also being found in this zone. The future of water Hyacinth in lake Victoria is uncertain: it is too soon to know whether it will go through another ‘explosive growth’ phase or whether it will continue to be controlled at the present density by the beetles. Clearly, the ecological dynamism of the littoral zone in relation to the presence and abundance of Eichhornia crassipes warrants close investigation. Therefrom, community structure and food web models would be valuable tools for predicting possible impacts on species diversity and fish production. The potential for bilharzia A special mention needs to be made concerning bilharzia (schistosomiasis). Bilharzia is a debilitating and often lethal disease from waterborne parasites transmitted by freshwater snails in tropical countries (Obeng, 1969). The wetland/open water interface zone is an ideal environment for the snails and the potential for infection and transmission of the parasite cercaria to human hosts is high. The two gastropod vectors of the disease, Bulinus and Biomphalaria, were abundant in the floating vegetation root mats and although infections of the snails were not ascertained, earlier studies (e.g. Okedi, 1990) in Murchison Bay of Lake Victoria reported prevalence of the disease. Also, reports have indicated an increase in the incidence of bilharzia patients at medical centres around the lakeshores in recent years. Clearly, increased abundance of these two gastropods poses a potential health hazard to the lakeside communities, particularly to fisherfolk and those who make regular contact with the lake water.

Conclusions The floating root mat of Eichhornia provides new habitats for colonization by aquatic macroinvertebrates and, in effect, becomes consolidated into the natural plant succession. The outer fringes of the vegetation at the interface with the open water are

well oxygenated and support diverse and abundant invertebrate communities. Deeper into the bed of vegetation, the oxygen concentration in and below the root mat declines. This is mirrored by a decline in abundance and diversity of invertebrates typified by low oxygen-tolerant species. It may be assumed that Eichhornia is a permanent addition to the Lake Victoria macroflora and first indications are that in terms of aquatic macroinvertebrate communities it serves a similar function to the surface-floating Pistia that it has replaced.

Acknowledgements We appreciate assistance by the technical staff of FIRRI particularly Mr Hannington Ochieng and Miss Egulance Ganda for their immense role in sorting samples. This research was funded by the Government of Uganda through the Water Hyacinth Research Project based at the Fisheries Resources Research Institute (FIRRI) of the National Agricultural Research Organisation (NARO) in Jinja, Uganda, and the Netherlands Fellowship Programme through IHE in Delft.

References Bailey, R. G., S. Churchfield & R. Pimm, 1978. Observations on the zooplankton and littoral macro-invertebrates of Nyumba ya Mungu reservoir, Tanzania. Biol. J. linn. Soc. 10: 93–107. Bailey, R. G. & M. R. Litterick, 1993. The macro-invertebrate fauna of water hyacinth fringes in the Sudd swamps (River Nile, Sudan). Hydrobiologia 250: 97–103. Balirwa, J. S., 1995. The Lake Victoria environment: its fisheries and wetlands – a review. Wetlands Ecology & Management 3: 209–224. Balirwa, J. S., 1998. Lake Victoria Wetlands and the Ecology of the Nile tilapia, Oreochromis niloticus Linne. PhD thesis. A.A. Balkema, Rotterdam: 247 pp. Brown, K. M. & D. M. Lodge, 1993. Gastropod abundance in vegetated habitats: the importance of specifying null models. Limnol. Oceanogr. 38: 217–225. Cyr, H. & J. A. Downing, 1988. The abundance of phytophilous invertebrates on different species of submerged macrophytes. Freshwat. Biol. 20: 365–374. Denny, P., 1984. Permanent swamp vegetation of the Upper Nile. Hydrobiologia 110: 79–90. Denny, P., 1991. African Wetlands. In Finlayson, M. & M. Moser (eds), Wetlands. Facts on File. Oxford: 115–148. Gopal, B., 1987. Water Hyacinth. Elsevier, Amsterdam: 471 pp. Green, J., S. A. Corbet, E. Watts & Oey Biauw Lan, 1976. Ecological studies on Indonesian lakes. Overturn and re-stratification of Ranu Lamongan. J. Zool., London 180: 315–354. Hargeby, A., 1990. Macrophyte associated invertebrates and effect of habitat permanence. Oikos 57: 338–346.

88 Harley, K. L. S., 1990. The role of biological control in the management of water hyacinth, Eichhornia crassipes. Biological News and Information 11: 11–22. Harley, K. L. S., M. H. Julien & A. D. Wright, 1997. Water Hyacinth: a tropical worldwide problem and methods for its control. Appendix 7. In: Proceedings of the International Water Hyacinth Consortium. World Bank, Washington, D.C., March 1997. Howell, P., M. Lock & S. Cobb, 1988. The Jonglei Canal: Impact and Opportunity. Cambridge University Press: 537 pp. Kornijow, R., 1989. Macrofauna of Elodeids of two lakes of different trophy: relationships between plants and structure of fauna colonising them. Ekol. Pol. 37: 31–48. McLachlan, A. J., 1969. The effect of aquatic macrophytes on the variety and abundance of benthic fauna in a newly created lake in the tropics (Lake Kariba). Hydrobiologia 66: 212–231. Mitchell, D. S., 1985. African aquatic weeds and their management. In Denny, P. (ed.), The Ecology and Management of African Wetland Vegetation. Dr W. Junk Publishers, Dordrecht, Geobotany 6: 177–199. Obeng, L. E., 1969. The invertebrate fauna of aquatic plants of the Volta Lake in relation to the spread of helminth parasites. In Obeng, L. E. (ed.). Man-Made Lakes: The Accra Symposium, Ghana University Press: 398 pp. Ogutu-Ohwayo, R., 1990. The decline of the native fishes of lakes Victoria and Kyoga (East Africa) and the impact of introduced species, especially the Nile perch, Lates niloticus, and the Nile tilapia, Oreochromis niloticus. Envir. Biol. Fish. 27: 81–96. Ogutu-Ohwayo, R., 1995. Diversity and stability of fish stocks in Lakes Victoria, Kyoga and Nabugabo after establishment of introduced species. In Pitcher, T. J. & J. B. Hart (eds), The Impact of Species Changes in African Great Lakes. Chapman and Hall, New York: 59–81. Okedi, J. Y., 1990. Observations on the benthos of Murchison Bay, Lake Victoria, East Africa. Afr. J. Ecol. 28: 111–122. Peckarsky, B. L., P. R. Fraissinet, M. A. Penton & D. J. Conklin, 1990. Freshwater macroinvertebrates of Northeastern North America. Camstock Publ. Associates, Ithaca & London. Petr, T., 1968. Population changes in aquatic invertebrates living in two water plants in a tropical man-made lake. Hydrobiologia 32: 449–484.

Rzoska, J., 1974. The Upper Nile swamps, a tropical wetland study. Freshwat. Biol. 4: 1–30. Tailor, A. R. D., 1993. Floating water weeds in East Africa, with a case study in northern Lake Victoria. In Greathead, A. & P. De Groot (eds), Control of Africa’s Floating Waterweeds. Proceedings of a Workshop, Zimbabwe, June 1991; Commonwealth Sc. Council, CAB International, Biomass Users Network. Thorp, J. H. & A. P. Covich (eds), 1991. Ecology and classification of North American Freshwater Invertebrates. Academic Press Inc. San Diego, California. Twongo, T. & J. S. Balirwa, 1995. The water hyacinth problem and the biological control option in the highland region of the Upper Nile basin – Uganda’s experience. Paper presented at The Nile 2002 conference – Comprehensive water resources development of the Nile Basin: ‘Taking Off’, Feb. 13th–17th, 1995; Arusha, Tanzania. Twongo, T. K. & J. S. Balirwa, 1996. Influence of water hyacinth on the ecological character of shoreline wetlands of Lake Victoria (East Africa). At: 6th meeting, Ramsar Convention, Brisbane, Australia, 19th–27th March, 1996. Twongo, T., F. W. B. Bugenyi & F. M. Wanda, 1995. The potential for further proliferation of water hyacinth on Lakes Victoria, Kyoga and Kwania and some urgent aspects for research. Afr. J. Trop. Hydrobiol. Fish. 6: 1–10. Wanink, J. H., 1998. The pelagic cyprinid Rastrineobola argentea as a crucial link in the disrupted ecosystem of Lake Victoria. PhD thesis, University of Wageningen: 288 pp. Whiteside, M. C., J. B. Williams & C. P. White, 1978. Seasonal abundance and patterns of Chydorid Cladocera in mud and vegetation habitats. Ecol. Soc. am. 59: 1177–1188. Willoughby, N., I. G. Watson, S. Lauer & I. F. Grant, 1993. An investigation into the effect of water hyacinth on the biodiversity and abundance of fish and invertebrates in Lake Victoria, Uganda. NRI Project Report 10066 A032g, Natural Resources Institute, Chatham, U.K. Witte, F., T. Goldschmidt et al., 1992. The destruction of an endemic species flock: Quantitative data on the decline of the haplochromine cichlids of Lake Victoria. Envir. Biol. Fish. 34: 1–28.