4. DISCUSSION
Although the food resources of O. niloticus were quite specific and similar (algae, diatoms, zooplankton, and detritus), but algae and diatoms were the most important items in the diet of native C. rendalli species. Both species in the size class <50 and 51-100 mm fed mainly on the similar food resources. The principal difference between the diets of exotic Nile tilapia and the native species was the size class 101-150 mm in which particular food items came to be more important, since the native utilized algae as a broader food category in different size classes. Our results also indicated thatO. niloticus were more generalist with respect to diet composition and the abundance of food items, and relied more heavily on zooplankton and chlorophytes in the 101-150 mm size class than C. rendalli . This more generalist foraging on microphages, macrophages, microfauna, and detritus probably explains the high abundance ofO. niloticus relative to native species in the river.
Contrary to our main hypothesis, the stomach content analysis results give congruent feeding preferences for both in different size classes in the upper Kabompo River. For both species, similarity in feeding item was evident in the size classes <50 mm, except for size class 51-100 mm which showed a partial niche segregation. Oreochromis niloticus evidently had a more diverse and broader food resource niche than native C. rendalli species. This was observed by dominating fish communities of O. niloticus having been observed to have a wide trophic niche feeding zooplankton and detritus (Kırankaya & Ekmekçi, 2013; Werner & Gilliam, 1984; Winemiller, 1989; Mason et al., 2008; Zengeya et al., 2015). This was evident in our study as O. niloticus was seen have wider feeding spectrum than native C. rendalli . However, competitive interactions and dietary overlap with natives such as O. macro chir may force O. niloticus to occupy a wider trophic niche (Kenzo & Mazingaliwa, 2002; Marshall, 2011; Zengeya et al., 2015). Hence, the observed narrow niche use of native species may partly result from increased competition for food resources following the invasion of O. niloticus . In Lake Chivero, exotic O. niloticus and natives presented a low dietary overlap and wide niche areas, while omnivores had a high trophic niche overlap where food sources were limited (Junor, 1969; Marshall, 2011). However, O. niloticus feeds on larger quantities of food than native species in Lake Victoria (Marshall, 2011; Trewavas, 1983; Zengeya et al., 2015). Inclusion of more detritus and the algaeMicrocystis , Gonium , and Phacus in the diet ofO. niloticus may also indicate differences in foraging tactics relative to native species.
Our results from SCA suggested that the wide dietary niche of O. niloticus is largely due to marked size class differences in niche use. Some size classes of O. niloticus had exceptionally wide or narrow trophic niches, indicating low or high levels of interspecific competition in diets, respectively, and high individual variation in feeding habitats (Werner & Gilliam, 1984; Zengeya et al., 2015). The small number but high quantity of food items consumed by O. niloticus in the 101−150 mm size class is probably due to specialized foraging on abundant food sources in certain habitats, whereas the relative importance of food items consumed by most native species and also some O. niloticus in the <101 mm size classes probably indicates a diet dominated by algae chlorophytes and diatoms. The selection of these food items at size class 151–302 mm of the species is likely associated with their high abundance in the upper Kabompo River, as reported for a nearby section of the river outside the sampled area (Bok & Bills, 2012; AES, 2014).
Previous studies suggest that dietary niche overlap between species can be interpreted as evidence of multiple shared food sources (Helfman et al., 1997; Pilger et al., 2010) and therefore as an indicator of potential competition (Eloranta et al., 2015). The exotic speciesO. niloticus and the native species C. rendalli exhibited high trophic niche overlap in <50 mm size class. However, their wider trophic niche area suggests that the O. niloticus are opportunistic and can change to overcome possible competition in trophic position or diet composition, and may explain the coexistence of these 2 species (Lowe-McConnell, 1987; Winemiller, 1991, Winemiller & Kelso-Winemiller, 2003; Agostinho et al., 2015). In contrast, all native species in our study had narrower dietary niches. For example, although cichlids might include many species, individual species demonstrate a narrow dietary niche area, suggesting that these species are specialists. Orochromis macrochir and Coptodon rendalliwhere caught in relatively similar abundant of the aforementioned size class during the study. The species are of similar size classes, feeding mainly on filamentous algae, benthic invertebrates, and diatoms (Kenzo & Mazingaliwa 2002). Thus, the dietary niche of these species may overlap with that of O. niloticus at the base of the trophic web. Nowadays this native species is among the less abundant species in the upper Kabompo River (DoF 2018). In contrast, the O. niloticuspopulation has greatly increased in the last decade, and highly abundant as recorded during this study. Oreochromis niloticus is a native pelagic herbivore which feeds mainly on detritus, zooplanktons, small invertebrates (in size classes 51-100 and 101-150 mm), and algae (in size class <50 mm), suggested a potential dietary niche overlap with C. rendalli (Kenzo & Mazingaliwa 2002, Marshall 2011, Zengeya et al., 2015).
Niche differentiation can explain how species co-exist in an ecosystem, but empirical support has been sparse (Mason et al., 2008; Zengeya et al., 2015). However, niche specialization has been regarded as a common phenomenon in fish community assemblages (Pilger et al., 2010; Eloranta et al., 2015). In the upper Kabompo River, there seems to be dietary niche specialization among native species, supported by their relatively low trophic niche overlap. The only substantial overlaps among native species were observed in <50 mm size class for C. rendalli. Oreochromis niloticus and Coptodon rendalli in the 51−100 mm size class showed partial segregation in their diets (Minshull, 1969; Junor, 1969; Kenzo & Mazingaliwa, 2002; Zengeya et al., 2015). In contrast, the high dietary niche overlap was observed for both species in <50 mm size class suggests potential competition for food resources (Brendonck et al., 2003; Marshall, 2011; Zengeya et al., 2015). The presence of O. niloticus can also affect populations of native species indirectly. In many studies, invasion by O. niloticus has been related to cascading effects in the numbers of aquatic plants and phytoplankton biomass, and a reduction in macroinvertebrate and macrophyte populations (Moriarty & Moriarty, 1973; Marshall, 2011). For example, in the Olifants River in the Western Cape Province of South Africa, Nile tilapia became the dominant species within a few years of introduction and eliminated macrophytes within a few years of establishment, coinciding with significant declines in previously abundant native species within the community. As a result, natives are now confined to smaller tributaries and headwaters (Zengeya et al., 2015). Furthermore, the presence of O. niloticus can lead to further trophic cascading effect on ecosystem functioning (Marshall, 2005; AES, 2014; Zengeya et al., 2020).
The narrow dietary niche of native species observed in 51-100 mm size class suggests that a small change in food resource abundance could generate a big shift in dietary intraspecific competition of native species in the population sizes. This narrow dietary niche could contribute to population decreases of native species, particularly, C. rendalli , which showed a relatively high degree of dietary overlap with O. niloticus . This may be because the system has been changing rapidly over the last few years due to anthropogenic activities in the surrounding area, as predicted by a 2010 biological monitoring study (Bok & Bills, 2012; DoF, 2018). Mining activities may affect fish habitat condition and affect populations of native fishes. Another factor is dietary niche overlap among natives, suggesting shared resource use and potential competition. However, as indicated by wide dietary niches in <50 mm class size,O. niloticus species is able to switch food sources depending on the abundance of these sources, while native species that feed only on a small range of food sources are more vulnerable to changes in resource abundance. Such vulnerably contributed by presence of exotic species may create intraspecific competition within size classes of the species and interspecific competition of the species, leading to habitat displacement and possible extinction of native species.
In conclusion, partial dietary segregation observed in size class ofO. niloticus and C. rendalli confirm the coexistence and interspecific competition between them in the upper Kabombo River. Further studies are required to investigate food web structure that may help to explain population decreases of some native species due to invasion in the upper Kabompo River and other rivers across African where the species coexist.