Malaria parasites (Apicomplexa, Haematozoea) and their relationships with their hosts: is there an evolutionary cost for the specialization?

Parasite–host specialization is frequently considered to be a derived state such that it represents an ‘evolutionary dead end’ that strongly limits further evolution. In this study, it was tested whether this theory is applicable to the relationship of malaria parasites and their vertebrate hosts. For this, we revisited Perkins and Schall (2002) analysis of the phylogenetic relationships of the malaria parasites (belonging to the genera Plasmodium, Haemoproteus and Hepatocystis) based on the mitochondrial Cytochrome b gene sequence, and inferred, using a maximum likelihood (ML) approach, the putative ancestral vertebrate hosts. As the topology in this study presents several unresolved branches and is slightly different from that of Perkins and Schall, a Shimodaira and Hasegawa (SH; 1999) test has been performed in order to properly consider several alternative topologies. The results of this study suggest that the common ancestor of all these malaria parasites was a reptile (more specific of the order Squamata), and that the host switches from Squamata to Aves and vice versa were quite frequent along the evolution of these parasites. On the contrary, a strong evidence that the host shift from Squamata to Mammalia had occurred only once during the evolution of these organisms was found. This evidence (added to the current knowledge about the association of the malaria parasites with their vertebrate hosts) allows us to suggest, at least considering the species included in this study, that the adaptation in mammals had required a high level of specialization. Hence, the acquisition of this host class had culminated in an evolutionary dead end for the mammalian malaria parasites. Zusammenfassung Parasit‐Wirt‐Spezialisierungen werden oft als stark abgeleitete Merkmale aufgefasst, die in eine evolutive Sackgasse führen und eine weitere Evolution erheblich einschränken. In dieser Arbeit haben wir getestet, ob sich diese Theorie auf die Beziehung zwischen Malariaparasiten und ihren Wirbeltierwirten anwenden lässt. Zu diesem Zweck haben wir die Untersuchungen von Perkins und Schall (2002)über stammesgeschichtliche Beziehungen der Malariaparasiten (Plasmodium, Haemoproteus und Hepatocystis) aufgegriffen, die auf Sequenzvergleichen des mitochondrialen Cytochrom b basieren und haben mit einem maximum‐likelihood‐Verfahren auf die mutmaßlichen Wirbeltierwirtvorfahren geschlossen. Da unsere Topologie mehrere ungeklärte Zweige enthält und sich leicht von der Topologie vo