The origin and significance of host specificity are intriguing questions in parasitology. In the case of single-host versus multiple-host parasites, this topic integrates with the concept of the specialist/generalist trade-off. We use the model of sucking lice Polyplax serrata and rodent hosts Apodemus, to address these concepts. Polyplax serrata was shown to form a complex genetic structure, with a strictly specific S lineage living on Apodemus flavicollis, and a less specific N lineage on A. flavicollis and Apodemus sylvaticus. Moreover, the S lineage formed two mitochondrial clades with geographically exclusive distributions and a narrow hybrid zone, providing an opportunity to test the hypothesis that hybrids suffer a decrease in fitness. We sampled 451 individual lice from two host species at 103 localities. We used prevalences and intensities as proxies of fitness, which the parasites realize on their host. The S lineage, strictly specific to Apodemus flavicollis, reached significantly higher prevalences and intensities on its host compared with the N lineage. Conversely, the N lineage occurred with high prevalence and intensity on A. sylvaticus but tended to use also A. flavicollis when the louse populations became too dense. We discuss possible mechanisms behind this difference (particularly interspecific competition as a typical phenomenon in the specialist/generalist systems). We conclude that a parasite's "choice", not accessibility of the host or interspecific competition, is the main factor affecting the louse prevalences. We suggest that historical differences in geographic distribution of both lice and mice may provide a possible explanation for the observed life strategy differences. In contrast to the convincing picture in S and N lineage prevalences, we did not detect an expected drop in fitness in hybrids. We consider instability of the hybrid zone, or decline in abundance of the respective hosts, as possible explanations for this result. (c) 2024 Australian Society for Parasitology. Published by Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
