The conservation of the genetic resources of crops and their wild relatives has become acrucial task for mankind as soon as it has been realized that the gene pools of thoseresources drastically erode. This problem is particularly exacerbated for the gourd familyCucurbitaceae, a worldwide economically important plant group, which contains manyoverlooked and barely known useful species. Different classification approaches in the familyresulted in controversies about phylogenetic relationships of species, genera, and tribes.Moreover, knowledge of genetic patterns and ecological demands of species has beenscarce, particularly in West Africa. This dissertation aims at contributing to halt loss ofcucurbit resources by shedding light onto relationships among cucurbit species occurring inthe phytoregions of West Africa, and to investigate the structure of the genetic variation oftwo particularly important species of this region, Lagenaria siceraria and Momordicacharantia.I use phylogenetic, cytological and population genetic approaches to resolve longstandingambiguities regarding West African cucurbits. My results indicate that all species ofBenincaseae (mostly from Africa) fall within their traditional tribal group. However, thesubdivision of the tribe into subtribes Cucumerinae and Benincasinea is not supported, asmost species of Cucumerinae are dispersed throughout the Benincaseae. In additionCucumis melo subsp. melo (watermelon) shows pronounced differences to subsp. agrestis,what is congruent with morphological data.Detailed analyses on cytological and genetic variation in Lagenaria siceraria(calabash) revealed a clear correlation of genome size with two different seed or fruit usagetypes and with growing elevations in West Africa. The genome size differences in seed typesof L. siceraria might indicate differences in their evolutionary history. AFLP polymorphism inthe species was in general low, but principal co-ordinate analysis, Bayesian assignmenttests, and phylogenetic analyses revealed three significantly differentiated groups. In additionto the already known subsp. siceraria and asiatica, a third taxon, described as subsp. egusiwas inferred based on these analyses.Morphometric analyses clarified also the long-standing ambiguity for the identificationof Momordica charantia observed in Benin by clearly discriminating two groups characterizedby leaf shape differences: the common subsp. charantia and the newly described subsp.macroloba, endemic to the Dahomey gap and the Sudano-Guinean phytoregion of Benin andTogo. Phylogenetic analyses using AFLP indicated the existence of a paraphyleticmacroloba group, which clusters within subsp. charantia, and indicated some gene flowbetween both types. Within subsp. charantia all populations collected in the dry Sudanianzone were monophyletic. Although not formally described I refer to this group as sudanicaclade. Ecoclimatic niche modelling revealed different preferential distribution areas for thethree entities, and showed a good fit between potential and realised ecoclimatic niches.Apart from the clear structuring of the species, our data show that genetic diversity withinperipheral and centre populations of M. charantia are similar. Adequate conservationstrategies could therefore be suggested for the promotion and the sustainable use of thesevaluable resources.