The fish community of Lake Victoria, East Africa, has continued to exhibit an intriguing degree of resiliency despite stress from anthropogenic activities. Frequent environmental perturbations include wild fluctuations in fishing pressure, limnological conditions, and lake levels. Surprisingly, many of the endemic and other indigenous fishes have survived, and some have even increased in numbers in recent years. While a positive development, this resiliency is a red flag to scientists because we do not understand it, or know whether it can be expected to continue. Furthermore, besides issues of immediate human well-being, there remain grave concerns about long term resource sustainability. This paper explores possible reasons for ecological resilience in the Lake Victoria fish community, with a focus on omnivory and functional redundancy as possible explanations. The analysis used published data based on both traditional gut content analysis (GCA) and stable isotope analysis (SIA). Trophic plasticity is ubiquitous among the surviving fishes of Lake Victoria. This, combined with an overall simplification of the food web, contributed to the lake community's current resilience.
Lake Victoria (69,000 km2), East Africa, is the largest freshwater lake in the tropics. Its watershed encompasses varied habitats including deep and open lake, shallow and sheltered bays, swamps, satellite lakes, dams and rivers. Together these contained one of the world's most speciose fish communities. The lion's share of the fishes were members of the celebrated diversification of haplochromine cichlids, which owed much to the lake's spatial layout and temporal dynamics (Mayr, 1963, Meyer, 1993; Kaufman et al., 1997). Chronic and cumulative human stressors, beginning in the 1920s with the advent of modern and efficient gear, wrong fishing practices, introduction of alien species in the 1950s (Cadwalladr, 1965; Ogari, 1992; Verschuren et al., 2001) and rapid eutrophication, were associated with a collapse of the endemic haplochromine species flock and dominance of the fish community by only three species. Two are introduced species (Nile tilapia, Oreochromis niloticus and predatory Nile perch, Lates niloticus) and one is a native cyprinid, Rastrineobola argentea.
Recent surveys revealed dense but localized residual populations of two indigenous migratory riverine fishes (Labeo victorianus and Barbus altianalis) in the adjoining rivers (Ojwang et al., 2007). These and other indigenous species like Schilbe intermedius and Bagrus docmak are now considered economically extinct in the lake but occasionally appear in landings. A slight resurgence of a few haplochromines has been noted in the pelagic and rocky shoreline habitats, likely in response to fishing pressure on Nile perch (reducing predation pressure) and phenotypic plasticity of morphological features which enables adaptive response to environmental changes (Witte et al., 2008). However, haplochromines are far from reassuming the diversity and dominance of the ecosystem and system processes that they exhibited prior to the mid 1980's (Seehausen, 1999; Balirwa et al., 2003).
The resiliency of the post-haplochromine Lake Victoria fish community presents something of a puzzle. Many studies have alluded to omnivory and functional redundancy as possible explanations for resiliency and stability in ecosystems. Here we review evidence that the surviving fishes of Lake Victoria exhibit these attributes, using published data/information based on both traditional gut contents analysis (GCA) and stable isotope analysis (SIA). SIA is an important adjunct to GCA in that it distinguishes actual trophic links from those potentially possible and therefore clarifies trophic relationships.
Omnivory and Functional Redundancy
Omnivory and functional redundancies are two concepts that have continued to dominate most discussions among ecologists (McKaye, 1995; Polis and Strong, 1996; Persson, 1999; Tanabe and Namba, 2005). Omnivory is defined as feeding on more than one trophic level (Pimm, 1982). It has been proposed that omnivory will work as a stabilizing agent and may be the glue for binding natural communities (McCann et al., 1998). Many studies have looked at different aspects of omnivory including behavioral traits of omnivorous species, the phylogenetic origin of omnivory, and the benefits and costs of omnivory (see Coll and Guershon, 2002). Omnivory was previously considered as rare in nature because of what most modeling studies considered as its destabilizing effects (Pimm and Lawton, 1978). However bony fishes are mostly known to be omnivorous, and traditional GCA studies have shown this to be true of today's Lake Victoria fishes as well (Table 1). Furthermore, Lake Victoria fishes exhibit a tendency toward greater omnivory over the last half century. Some that were hitherto piscivorous like B. docmak and S. intermedius now show a greater propensity to feed at more than one trophic level (Goudswaard and Witte, 1997). The three species currently of greatest economic importance, L. niloticus (introduced), O. niloticus (introduced) and R. argentea (native) are also now all omnivorous (Wanink, 1998; Mkumbo, 2002; Njiru et al., 2004). Indigenous tilapiines were and remain more stenotrophic: e.g. O. esculentus (Ngege) maintained a narrow diet range of mostly phytoplankton (mostly diatoms) and Tilapia zillii stuck with a preponderance of vascular plant material, even after being confined to small water bodies within the lake watershed (Opiyo and Dadzie, 1994; Mbabazi et al., 2004). One may argue that using GCA can be misleading because ingestion does not guarantee assimilation (Vander Zanden et al., 1999). Recent studies in Lake Victoria (see Campbell et al., 2003; Mbabazi, 2004; Ojwang et al., 2004; 2007) have used SIA to circumvent this shortcoming. It is evident from this review that omnivory is prevalent amongst the surviving fishes of Lake Victoria irrespective of the approach or method used (Figures 1, 2 and 3; Table 1). Even surviving morphological specialists show increased trophic breadth as revealed by SIA. For example, Ojwang (2006) observed that although the subterminal mouth of the cyprinid L. victorianus is well suited to its typical habit of grazing aufwuchs from rocks in clearwater streams, recently it fed more on insects in perturbed and turbid waters. The most fundamental question therefore is whether the ubiquity of omnivory in Lake Victoria- a recent development- could contribute to the fish community resiliency amidst continuing environmental perturbations.
Omnivory has also been associated with the success of invasive species e.g. the establishment of O. niloticus has been attributed to its ability to feed on different trophic levels (McKaye et al., 1995; Batjakas et al., 1997). Ontogenetic omnivory exhibited by large fishes like L. niloticus in Lake Victoria also allows for coexistence at different life stages between the predators and prey; e.g. schooling of juvenile Nile perch and adults of the native zooplanktivorous cichlids Yssichromis spp. (Ojwang et al., 2004). Prey switching in Nile perch onto the most abundant prey first (Mkumbo, 2002), relieves predation pressure and provides refuge for a number of fish species at low densities. This could help to explain the survival of relict populations of indigenous species in the lake despite the presence of Nile perch.
The original Lake Victoria fish assemblage has obviously been destabilized, and the current fish assemblage is not consonant with the value of preserving biological diversity, but this does not mean that it is definitionally unstable or dysfunctional. The “Diversity-Stability” hypothesis, while perhaps operational in the fish community's prior incarnation, is clearly not the sole driver of system stability today, and whether the system is stable or not depends upon the definition of stability.
Ecosystem stability may be defined in terms of the maintenance of a particular species assemblage or instead, in terms of functional groups’ capacity for adaptive, differential response to perturbation (McCann, 2000). The two are not necessarily mutually exclusive, but are in Lake Victoria on its current trajectory. The high level of omnivory observed amongst surviving fishes in the Lake Victoria may thus be a contributor to ecosystem stability (McCann and Hastings, 1997). Omnivory can enable fish species and the systems to respond differentially to variable effects of environmental stress even at much lower species diversity than existed in Lake Victoria up to the 1980's (Diehl, 1993; Fagan, 1997). Omnivory can effectively shorten food chains, thereby promoting their persistence (Sprules and Bowerman, 1988). Functionally short chains are believed to be more stable than longer ones (Vander Zanden et al., 1999; Pimm, 1982). This particular view of circumstances in Lake Victoria will depress conservationists, but will give fisheries officers the needed impetus to continue with the on going implementation of the management plan for the lake fisheries.
In most aquatic systems, omnivores near the top of a food chain derive much of their subsistence from organisms on the lower, highly productive trophic levels where energy is presumably unlimited, rather than at intermediate trophic levels (Vander Zanden et al., 1999; Coll and Guershon, 2002). This is what we observe in Lake Victoria and nearby lakes modified by fish introductions and eutrophication: SI data indicate that even large Nile perch feed heavily at surprisingly low trophic levels in Lake Victoria, Lake Kyoga, and nearby satellite lakes where they have been introduced (Mbabazi, 2004).
The other ecological attribute which is the flip side of omnivory in Lake Victoria, and perhaps also important to the system's surprising endurance, is the ability of seemingly very different species to perform similar ecological functions i.e. the biogeochemical activities of an ecosystem or the flow of materials (nutrients, water, atmospheric gases) and processing of energy (Naeem, 1998).
Until the early 1980's Lake Victoria had 80% of its ichthyobiomass made up of haplochromine species, 20% of which were zooplanktivorous. The loss of zooplanktivores and subsequent increase in native zooplanktivorous minnow, R. argentea raises the question whether the native minnow functionally replaced the dwindling zooplanktivorous haplochromines. Goldschmidt et al. (1993) reported that the abundant freshwater shrimp, C. nilotica had functionally replaced the nearly extinct detritivores. Studies by Ojwang et al. (2004) on the three fishes (the two Yssichromis species and R. argentea) in the pelagic component of Lake Victoria further depicted functional redundancy (Table 2).
The existence of species with similar ecological roles has received considerable attention with respect to functional redundancy (Walker, 1992, 1995; Naeem, 1998). As noted by Walker (1995) redundancy is an insurance against loss of function when species are lost, for example when the environment changes beyond the tolerance of some species. Proponents of functional redundancy argue that species that undertake similar functions are substitutable and the loss of species that perform the same role will not affect ecosystem function (Walker, 1992; Frost et al., 1995). On the other hand, species redundancy also enables compensatory functional replacement in the face of perturbation due to differential response of species to variable conditions (Naeem, 1998; Lawton and Brown, 1993).
The current Lake Victoria is at least an order of magnitude less diverse than it was but a few decades ago but the ecosystem is still hanging on through the ability of a few species to respond differentially to environmental perturbations. The near-crash in the populations of zooplanktivorous haplochromines must have had a huge impact on the prey available to important fishery species, but may have had little or limited impact on ecosystem function as the numbers of R. argentea, the minnow zooplanktivore, increased fourfold. The minnows must have certainly filled at least some of the functional void left by haplochromines. Recent upsurge in the numbers of piscivorous Prognathochromis spp. in Lake Victoria (Mbabazi, 2004; Ojwang, 2006) in the face of intensive fishing pressure on Nile perch is another good example of trophic redundancy making possible a differential species response, and maintenance of function through species diversity. Although there is paucity of data on the diet of remnant piscivorous haplochromines during the upsurge in numbers of predatory Nile perch, it is reasonable to postulate that competition from the introduced Nile perch caused surviving “piscivorous” haplochromines to resort to taking a larger proportion of insects in their diets than during earlier days when they could roam relatively fearlessly in open waters (Mbabazi et al., 2004). Concurrent studies also showed that other historically piscivorous species such as B. docmak and S. intermedius have switched to feeding mostly on insects (Olowo and Chapman, 1999).
The rebounding of species is a clear sign of resiliency consistent with compensatory functional replacement. Initially, the lake's future was depicted as being rather bleak within the last decade, with expectations of complete ecosystem collapse (Barel et al., 1985; Kaufman, 1992). Indeed, given the levels of environmental stress or pressure delivered upon Lake Victoria in recent years, it is surprising that there are still fish from the lake being exported, with European Union as the dominant market (Odongkara et al., 2009). Harris (1993) proposed that ecosystems are buffered from extinction by redundancy. The level of redundancy in Lake Victoria is now at its lowest level in historic times, and biased with species distributed unevenly among functional groups. Even under high anthropogenic pressure Lake Victoria has continued to provide the services human need. This does not mean that allowing continued adversity is a wise strategy for maximizing human wellbeing around Lake Victoria, now or in the future.
In spite of high levels of anthropogenic pressure, the Lake Victoria fish community in its new incarnation exhibits omnivory and functional redundancy, along with a surprising level of resilience to fishing pressure and environmental perturbations such as lake level fluctuation. Even though both omnivory and functional redundancy are fundamental attributes in the resiliency observed there are other subtle contributory factors such as the phenotypic plasticity of morphological features common in cichlids. But just because Lake Victoria has managed to persist as long as it has does not mean that it can tolerate even more. It is thus critical that the resiliency observed thus far is nurtured through prudent management strategies that would guarantee the future of the fishery for the livelihood of riparian communities. To realize a sustainable fishery, the management plan for Lake Victoria fisheries should shift from the current single-species management to a form of ecosystem-based management that encompasses rivers, satellite lakes and dams, as well as the lake proper, and that includes all key elements of the system, and not just the target fishery species. The Lake Victoria Region still harbors a diverse aquascape including numerous indigenous fish species refugia that represent seeds for the eventual restoration of a bit of the lake's original diversity. It is clear that if persistent elements of the indigenous fish community were nurtured and human activity regulated under a unifying systems approach, a future trophically diverse and functionally efficient Lake Victoria ecosystem could well be achieved. Thus the current ongoing initiatives including Implementation of Lake Victoria Management Plan (IFMP) and Lake Victoria Environmental Management Program (LVEMP II) are timely and apt in ensuring sustainable exploitation of the Lake fisheries and providing alternative livelihoods to alleviate current intense fishing pressure.
We wish to thank the management of Kenya Marine and Fisheries Research Institute (KMFRI) for their support. We also acknowledge the contribution and support from our colleagues at KMFRI-Kisumu Research Center, Robert Hecky and two anonymous readers.