Patterns of cryptic hybridization revealed using an integrative approach: a case study on genets (Carnivora, Viverridae, Genetta spp.) from the southern African subregion

Abstract

Recent years have seen the development of molecular-based methodologies to investigate hybridization and its impact on the evolutionary process. However, morphological characterization of hybrid zones has only scantily been considered, especially in zootaxa. Thus, the level of congruence between molecular and morphological characters when attempting to detect hybrids remains a poorly tackled area. The genets (genus Genetta) provide an ideal case study for further investigation of the respective contribution of morphology and DNA in hybrid zone characterization because (1) their morphology has recently been exhaustively explored and (2) the existence of hybrid zones in southern Africa was proposed in the literature. We assessed levels of hybridization among the southern African genets, and questioned the role of ecological factors on the hybridization patterns detected. We used an integrative approach involving nine discrete morphological characters and a diagnostic discriminant function, geometric morphometrics and sequences of cytochrome b including collection specimens. The combination of independent materials allowed us to accurately reassess the level of hybridization in southern African genets, and revealed cryptic, interspecific gene flows. Morphology unambiguously detected a low number of G. maculata × G. tigrina hybrids and rejected the hypothesis of a large intergradation zone in KwaZulu-Natal, thus supporting the species status of the two genets. Cytochrome b analyses revealed: (1) cryptic, massive hybridization between G. tigrina and the sympatric G. felina, and (2) a trace of reticulation (one sequence) between G. tigrina and the allopatric G. genetta. The type specimen of G. mossambica Matschie, 1902 is considered to be a morphological hybrid between G. maculata and G. angolensis. Remarkably, the morphological approaches (discrete characters and morphometrics) proved complementary to conclusions derived from cytochrome b sequences. Whilst morphometrics was generally unable to accurately identify all putative hybrids, this approach revealed diagnostic cranial shape differences between recognized species as well as the cryptic G. ‘letabae’ (included in the super-species G. maculata). Morphometrics also confirmed the diagnostic value and age dependency of discrete characters. Our integrative approach appeared necessary to the detection of cryptic hybridizations and to the comprehensive characterization of hybrid zones. The recurrent detection of hybrids exhibiting tigrina-like coat patterns may suggest (1) asymmetric hybridization of G. tigrina males to females of other species and (2) positive selection for tigrina-like phenotype in South African habitats, but these hypotheses will have to be further tested using other sources of evidence. Despite the precise mosaic of hybrid zones identified in southern African genets, the environmental factors that shape patterns of distribution of hybrids remain unclear. Nevertheless, in the light of our range reassessment, it appears that seasonality of precipitation and periods of annual frost may play stringent roles in the distribution of genets. The complementarity of our results based on morphology and molecules is regarded as encouraging for the further development of integrative approaches in order to better understand the complex phenomena that underlie hybridization processes.