Reduced host-specificity in a parasite infecting non-littoral Lake Tanganyika cichlids evidenced by intraspecific morphological and genetic diversity

Abstract

Lake Tanganyika is well-known for its high species-richness and rapid radiation processes. Its assemblage of cichlid fishes recently gained momentum as a framework to study parasite ecology and evolution. It offers a rare chance to investigate the influence of a deepwater lifestyle in a freshwater fish-parasite system. Our study represents the first investigation of parasite intraspecific genetic structure related to host specificity in the lake. It focused on the monogenean flatworm Cichlidogyrus casuarinus infecting deepwater cichlids belonging to Bathybates and Hemibates. Morphological examination of C. casuarinus had previously suggested a broad host range, while the lake's other Cichlidogyrus species are usually host specific. However, ongoing speciation or cryptic diversity could not be excluded. To distinguish between these hypotheses, we analysed intraspecific diversity of C. casuarinus. Monogeneans from nearly all representatives of the host genera were examined using morphometrics, geomorphometrics and genetics. We confirmed the low host-specificity of C. casuarinus based on morphology and nuclear DNA. Yet, intraspecific variation of sclerotized structures was observed. Nevertheless, the highly variable mitochondrial DNA indicated recent population expansion, but no ongoing parasite speciation, confirming, for the first time in freshwater, reduced parasite host specificity in the deepwater realm, probably an adaptation to low host availability.

Figure 1. Host species information.

(a) Geographical positions of sampling localities in Lake Tanganyika with indication of host species (pictures by Ad Konings). (b) Schematic phylogenetic tree of the Lake Tanganyika cichlid radiation, showing the phylogenetic position and relative divergence of the tribe Bathybatini and its major lineages. (c) Host species pictures (Ad Konings). The map was created using SimpleMappr software v7.0.0. (available at http://www.simplemappr.net. Accessed February 20, 2016).

Figure 2. A biplot of PCA (first two axes) based on measurements of haptoral sclerotized structures only showing the five best fitting morphological characters selected by CANOCO.

Symbols denote host species and their origin in each of the three subbasins of Lake Tanganyika. DALENGTO – Dorsal anchor, Length to notch, DATotlLn- Dorsal anchor, Total length, DBMaxmSt – Dorsal bar, Maximum straight width, VATotlLn – Ventral anchor, Total length, VBBranLn – Ventral bar, Branch length.

Figure 3. Scatterplots of the first two relative warps showing shape variation of the dorsal and ventral anchor with deformation grids (thin-plate) depicting mean anchor differences among groups.

Symbols denote host species and sampling localities: (a) dorsal anchor, separation according to the host species; (b) ventral anchor, separation according to the host species; (c) dorsal anchor, separation according to the sampling localities; (d) ventral anchor, separation according to the sampling localities. The number of specimens investigated is indicated in brackets.

Figure 4. Haplotype network of C. casuarinus COI sequences (n = 42).

The circles represent different haplotypes with size proportional to the number of individuals represented. Haplotypes are connected with lines, indicating number of mutations. Colours correspond with the host species (pictures by Ad Konings).

Figure 5. Demographic history of Cichlidogyrus casuarinus.

(a) Mismatch distribution. The black bars show the observed frequency of pairwise differences. The grey lines refer to the expected distribution based on parameter estimates (plus 95% confidence limits) under a model of population growth. The sum of squared differences (SSD) and raggedness index (rg) and their respective P-values are given to describe the fit of the observed distribution to the expectations based on growth parameter estimates, as well as τ, the modal value of the mismatch distribution. (b) Bayesian skyline plot (BSP) based on 402 base pairs of COI sequences of Cichlidogyrus casuarinus showing the effective populations size through time, assuming a substitution rate of 10% per site per million years. The thick line represents the median values; the thin lines denote 95% highest posterior density (HPD) intervals.