Abstract:

The Sub-Saharan Great Lakes Region has a rich ichthyofauna most notably, the cichlids. Among
the cichlids, the most important and commercially exploited are the tilapine species of which O.
niloticus Linn, 1758 (Nile tilapia) is the most common. This vital resource has been subjected to
immense anthropogenic threats primarily through fish translocations, overfishing, habitat
destruction, climate change, alien invasive aquatic weeds etc. Hereby, the broader aim of this
thesis was to investigate the impact of anthropogenic activities on the evolutionary trajectory of the
East African O. niloticus particularly in nonnative habitats following the aforementioned activities
in the region. This was investigated using the hypothesis that anthropogenic activities might have
altered the natural morphology and genetic structure of O. niloticus through admixture and
introgression with the indigenous congeneric species. Geometric morphometries and molecular
genetic markers were used to delineate the cichlid in the region from different East African
localities, covering low and high-altitude environments. Samples were collected from three locality
types; farms, lakes with natural but non-native (introduced) fish and lakes with native (indigenous)
fish. The farms and non-native localities were only limited to Uganda. Additionally, samples at the
continental level were collected, including populations from the West African country of Burkina
Faso and Ethiopia to test and compare the genetic structure with the East African ones. For
morphometric analysis only the East African populations were included. During geometric
morphometries analysis, two thin plate spline (Tps) programs; Tps Utility (TpsUtil) and Tps digitizer
(TpsDig) were utilized. These programs were used to acquire landmarks on fish digital
photographs and henceforth generation of x, y coordinates which were then imported into MorphoJ
program for subsequent multivariate analyses (chapter 2). At molecular level, 64 microsatellites
(SSR) loci were developed and 26 were employed to test the four East African/Uganda
populations. Here, the latest genotyping by sequencing approach using SSR loci (SSR-GBS) was
compared with the traditional fragmented length analysis by assessing the information content
recovered with the two approaches (Chapter 3). SSR-GBS was then applied in Chapter four to
investigate all the sub-Saharan Nile tilapia populations. A total of 40 SSR loci were used for
amplifying all the present studied populations. Geometric morphometric analysis generally
revealed that all the East African populations were somewhat homogeneous in shape, implying
close relation. However, when principle component scores (PC1 and PC2) were plotted against
the axes, it was clear that the non-natives; Victoria and Kyoga, were morphologically divergent
from for instance some native populations (Kazinga Channel and Edward). But the non-native
Victoria and Kyoga appeared undifferentiated from some natives like Albert, George, River Nile
and Turkana. The shape variation among the East African Nile tilapia populations was suspected
I
viii
Keywords: Nile tilapia. Geometric morphometries. Genotyping by sequencing. Admixture.
Geneflow
to admixture and hybridization, partly induced by anthropogenic activities. However, other
environmental cues like water turbulences (for example in River Nile), predation threat effects (e.g.
Nile perch in native lakes; Albert and Turkana), severe and adverse environmental conditions (like
declined water levels aggravated by high temperatures in L. Turkana), were taken into
consideration. Apparently, use of SSR-GBS genotyping approach yielded more information
content than the traditional based amplicon length procedures. For instance, SSR-GBS method
did better to reveal the genetic diversity of the East African O. niloticus than the conventional
amplicon length (AL). These results were consistent with a detailed genetic structure generated by
SSR-GBS which was undetectable with AL. Additionally, based on polymorphic information
content (PIC), it was evident that the SSR-GBS showed more polymorphic loci compared to AL.
More importantly the markers developed in this thesis and the genotyping techniques were
demonstrated how they may contribute useful information for further investigations regarding the
East African Nile tilapia for sustainable exploitation and conservation. By evaluating the genetic
diversity and variability patterns among Nile tilapia populations at sub-Saharan level, the Ethiopian
populations were indicated more divergent than the East African ones. This was particularly
prominent for lake Tana which indicated a high level of distinctiveness. In East Africa, genetic
structure analyses revealed four O. niloticus groups which were consistently supported by Fst
results. Moreover, the East African populations genetically differed from their Ethiopian and West
African counterparts. Also, O. niloticus from Lake Turkana (Kenya) was isolated, while the
Ugandan populations were grouped according geographical locations/water catchments and
anthropogenic activities i. Signals of genetic erosion were detected in Ethiopian and some East
African populations being likely a consequence of either bottlenecks associated with fragmentation
or founder effects. O. niloticus in Lakes Tana, Victoria and the cluster of Lakes Edward, George
as well as Kazinga Channel, may require further exploration for taxonomic recognition. Apart from
Lake Tana, the level of isolation of the cichlid in Lake Victoria and the genetic group in the western
Rift Valley Lakes, have not been reported before. This thesis indicates that anthropogenic activities
have contributed to a significant change in the genotypic composition of O. niloticus which may
compromise the species’ sustainability. Over all, the molecular analyses were congruent to the
geometric morphometries, which shows the robustness and importance of the current approach in
understanding the state the of the East African O. niloticus. In the ongoing work, out of scope of
this thesis, these populations are followed at mitochondrial genome level to precisely investigate
the extent of introgression and phyletic patterns.