BackgroundHost-parasite relationships are expected to be strongly shaped by host specificity, a crucial factor in parasite adaptability and diversification. Because whole host communities have to be considered to assess host specificity, oceanic islands are ideal study systems given their simplified biotic assemblages. Previous studies on insular parasites suggest host range broadening during colonization. Here, we investigate the association between one parasite group (haemogregarines) and multiple sympatric hosts (of three lizard genera: Gallotia, Chalcides and Tarentola) in the Canary Islands. Given haemogregarine characteristics and insular conditions, we hypothesized low host specificity and/or occurrence of host-switching events.MethodsA total of 825 samples were collected from the three host taxa inhabiting the seven main islands of the Canarian Archipelago, including locations where the different lizards occurred in sympatry. Blood slides were screened to assess prevalence and parasitaemia, while parasite genetic diversity and phylogenetic relationships were inferred from 18S rRNA gene sequences.ResultsInfection levels and diversity of haplotypes varied geographically and across host groups. Infections were found in all species of Gallotia across the seven islands, in Tarentola from Tenerife, La Gomera and La Palma, and in Chalcides from Tenerife, La Gomera and El Hierro. Gallotia lizards presented the highest parasite prevalence, parasitaemia and diversity (seven haplotypes), while the other two host groups (Chalcides and Tarentola) harbored one haplotype each, with low prevalence and parasitaemia levels, and very restricted geographical ranges. Host-sharing of the same haemogregarine haplotype was only detected twice, but these rare instances likely represent occasional cross-infections.ConclusionsOur results suggest that: (i) Canarian haemogregarine haplotypes are highly host-specific, which might have restricted parasite host expansion; (ii) haemogregarines most probably reached the Canary Islands in three colonization events with each host genus; and (iii) the high number of parasite haplotypes infecting Gallotia hosts and their restricted geographical distribution suggest co-diversification. These findings contrast with our expectations derived from results on other insular parasites, highlighting how host specificity depends on parasite characteristics and evolutionary history.Electronic supplementary materialThe online version of this article (10.1186/s13071-018-2760-5) contains supplementary material, which is available to authorized users.

No abstract

Hepatozoon spp. (Apicomplexa: Haemogregarinidae) are the most commonly reported hemoparasites from snakes. Of over 300 Hepatozoon species identified, more than 120 were described from snakes. However, recent genetic assessments have found Hepatozoon lineages recovered from both prey and predators, indicating that diet may play an important role in the infection of final vertebrate hosts. Here 4 different snake genera with different diets were assessed. Hepatozoon spp. prevalence varied greatly between the genera, but only lineages already identified from potential prey, i.e., gecko and lacertid lizards, were recovered from the snakes. Interestingly, the Hepatozoon spp. lineage known from geckos was the most common in the snakes, but this does not reflect their diet. Higher parasitemia levels, reported for some geckos relative to lacertid lizards, may play a role. Alternatively, this lineage may be more effective at parasitizing snakes or may occur, despite being unrecorded, in other vertebrate groups consumed by snakes.

:   Hemogregarines are the most-commonly reported hemoparasites in reptiles. In this work we analyzed samples from 572 individuals of 6 species of the wall gecko genus Tarentola from European and African countries adjacent to the Mediterranean Sea as well as from the Macaronesian islands. Screening was done using hemogregarine-specific primers for the 18S rRNA gene. Positive amplifications were sequenced so that the diversity of the hemogregarines from these hosts could be assessed within a phylogenetic framework. The results from the phylogenetic analysis showed that within Tarentola, the detected parasites are comprised of at least 4 distinct main lineages of Hepatozoon spp. In clades A and B, the new sequences clustered closely together with the ones previously known from individuals of the genus Tarentola and other species of geckos but also with those from other vertebrate host groups including skinks, snakes, iguanids, and rodents. Clade C included a sample from Tarentola angustimentalis of the Canary Islands. This sequence is the first molecular characterization of these hemogregarines in this archipelago. Until now, this lineage had only been found in lacertids, skinks, and snakes, so this infection extends the host range for this clade. Lastly, in the newly detected clade D, the retrieved parasite sequences form a group currently identified as exclusive of geckos. Our results show that geckos of Tarentola spp. harbor a great diversity of hemogregarines but also that further sampling and other tools, including a multi-locus approach using faster-evolving genetic markers, and identification of definitive hosts are needed to better understand the biology, diversity, and distribution of these parasites.

Species of Hepatozoon Miller, 1908 are blood parasites most commonly found in snakes but some have been described from all tetrapod groups and a wide variety of hematophagous invertebrates. Previous studies have suggested possible associations between Hepatozoon spp. found in predators and prey. Particularly, some saurophagous snakes from North Africa and the Mediterranean region have been found to be infected with Hepatozoon spp. similar to those of various sympatric lizard hosts. In this study, we have screened tissue samples of 111 North African and Mediterranean snakes, using specific primers for the 18S rRNA gene. In the phylogenetic analysis, the newly-generated Hepatozoon spp. sequences grouped separately into five main clusters. Three of these clusters were composed by Hepatozoon spp. also found in snakes and other reptiles from the Mediterranean Basin and North Africa. In the other two clusters, the new sequences were not closely related to geographically proximate known sequences. The phylogeny of Hepatozoon spp. inferred here was not associated with intermediate host taxonomy or geographical distribution. From the other factors that could explain these evolutionary patterns, the most likely seems series of intermediate hosts providing similar ribotypes of Hepatozoon and a high prevalence of host shifts for Hepatozoon spp. This is indicated by ribotypes of high similarity found in different reptile families, as well as by divergent ribotypes found in the same host species. This potentially low host specificity has profound implications for the systematics of Hepatozoon spp.