The Occurrence of Quill Mites (Arachnida: Acariformes: Syringophilidae) on Bee-Eaters (Aves: Coraciiformes: Meropidae: Merops) of Two Sister Clades

Simple Summary Parasitic mites of the family Syringophilidae (quill mites) represent the most diverse prostigmatan family associated with birds. Here, we aim (i) to investigate quill mites of a well-defined monophyletic clade of nine bee-eater species, containing mostly migratory African and Asian species; (ii) to establish ectoparasite geographic ranges; and (iii) to discuss patterns of host-parasite relationships and possible host-switches. We have found that despite quill mites being highly specific on the host species or genus level, host-switches may occur, particularly when the ecology of host species overlaps. Abstract We studied the quill mite fauna of the family Syringophilidae, associated with bee-eaters. We examined 273 bird specimens belonging to nine closely related species of the genus Merops, representing two phylogenetic sister clades of a monophyletic group. Our examination reveals the presence of two species of the genus Peristerophila, as follows: (1) a new species Peristerophila mayri sp. n. from Merops viridis in the Philippines, M. leschenaulti in Nepal and Sri Lanka, and M. orientalis in Sri Lanka; and (2) P. meropis from M. superciliosus in Tanzania and Egypt, M. persicus in Sudan, Tanzania, Liberia, Senegal, Kenya, and D.R. Congo, M. ornatus in Papua New Guinea, M. philippinus in Thailand, Indonesia and Sri Lanka, and M. americanus in the Philippines. The prevalence of host infestations by syringophilid mites varied from 3.1 to 38.2%. The distribution of syringophilid mites corresponds with the sister clade phylogenetic relationships of the hosts, except for P. meropis associated with Merops americanus. Possible hypotheses for the host lineage shift are proposed.


Introduction
The family Syringophilidae (Arachnida: Acariformes: Prostigmata: Cheyletoidea) represents the most diverse prostigmatan family associated with birds. To date, there are about 400 described species, and the range of their hosts comprises 24 orders and 95 families from all zoogeographical regions, except Antarctica [1]. All species of this family are permanent and obligatory ectoparasites of birds, and most of them display a high degree of host specificity, being mono-or oligoxenous parasites [2][3][4][5].

Materials and Methods
The mite material used in this study was collected according to the technique proposed by Skoracki (2011) [3] [13]. All collected feathers were checked under the stereomicroscope and opened using fine forceps. Infested feathers were placed in Eppendorf's tubes with Nesbitt's solution at 40 • C for about 24 h, and then mites were mounted on microscope slides in Hoyer's medium. Slide-mounted mites were examined under a light microscope (ZEISS Axioscope2™; Carl Zeiss AG, Jena, Germany) equipped with DIC optics and camera lucida.  Marks et al. [9]) and associated quill mite species.
In the descriptions, the idiosomal setation follows Grandjean [15], as adapted for Prostigmata by Kethley [16]. The nomenclature of leg chaetotaxy follows that proposed by Grandjean [17]. Measurements (ranges) for paratypes are given in parentheses following data for the holotype. All measurements are given in micrometers. Depositories of mite specimens are given using the following abbreviations: AMU-Adam Mickiewicz University, Department of Animal Morphology, Poznan, Poland; ZSM-Bavarian State Collection of Zoology, Munich, Germany.
Host data ranges were taken from BirdLife International [25]. The map was drawn using QGIS 3.16 software (Open Source Geospatial Foundation, Beaverton, Oregon, United States) with WGS 84 geographical projection.

Species Composition
In total, we investigated 273 host individuals belonging to nine Merops species, of which 56 specimens (21.1%) were infested by quill mites, whose prevalence varied from 3.1% to 38.2% (Table 1). We detected two quill mite species of the genus Peristerophila: P. mayri sp. n. and P. meropis (Skoracki et al., 2017). Their occurrence on the respective hosts corresponds with phylogenetic relationships within the genus Merops, as identified by Marks et al. is infested by P. mayri. Unexpectedly, P. meropis is found on M. americanus belonging to the second subclade ( Figure 1).

Etymology
This species is named in honor of Ernst Walter Mayr (1904Mayr ( -2005, arguably the most preeminent evolutionary biologist of the twentieth century. Mayr's contributions during the course of his remarkable life are manifold; among these, he had a lifelong special interest in ornithology and biogeography (e.g., [26]). When investigating the evolution of populations and species, he also referred to bee-eaters studied in this paper in his works (e.g., [27]).

Differential Diagnosis
Peristerophila mayri sp. n. is morphologically most similar to P. coraciidus Skoracki, Hromada et Sikora, 2020, described from Dollarbird Eurystomus orientalis (Coraciidae) from Papua New Guinea [28]. In females of both species, the hysteronotal shield is absent, and each medial branch of the peritremes has two or three chambers, whereas each lateral branch has four chambers. This new species differs from P. coraciidus by the following features: in heteromorphic females of P. mayri, the medial propodonotal shield is present; the tarsal fan-like setae of legs III and IV have 11-13 tines; genital (g1-g2) and pseudanal (ps1-ps2) setae are subequal in the length; the lengths of setae ag1, ag2, and ag3 are 110-125, 25, and 125-130, respectively; the pygidial shield is apunctate, well developed and with the rounded anterior margin. In heteromorphic females of P. coraciidus, the medial propodonotal shield is absent; the tarsal fan-like setae of legs III and IV have 17-19 tines; genital setae (g1-g2) are 4-5 times longer than pseudanal setae (ps1-ps2); the lengths of setae ag1, ag2, and ag3 are 180-265, 65-85, and 255-265, respectively; the pygidial shield is punctate and reduced to small region bearing bases of setae f1 and f2, with indiscernible anterior margin.

Discussion
Syringophilid mites are tightly associated with their hosts, where most of them represent mono-or oligoxenous species limited to a particular species or group of phylogenetically closely related hosts [3,4]. Because the biology of these obligate and permanent parasites is heavily linked to their avian hosts, their diversification pattern often reflects those of their hosts according to the old, but still actual rules of Fahrenholz [29] and Eichler [30].
In our study, birds belonging to two sister clades of bee-eaters are infested by two species of syringophilid mites (Figures 1 and 4), which shows that the occurrence of Peristerophila mites indeed tightly corresponds with the phylogenies of their avian hosts, except for the presence of Peristerophila meropis on Merops americanus.
We are describing an interesting situation, when two closely related host species, considered until recently to be subspecies of one species, are parasitized by different ectoparasitic quill mites. We suggest the following hypothesis why M. americanus is parasitized by P. meropis, a quill mite species typical for the sister bee-eater subclade: M. americanus occurs sympatrically with M. philippinus on the Philippines archipelago (Figure 4), where they often dwell in the same habitats. Since M. philippinus is a highly colonial species with tens or hundreds of pairs breeding together, they sometimes form mixed colonies with other local bee-eater species. Both M. philippinus and M. americanus breed in ground nest-burrows [38,39] that might be re-used over the years; it is also probable that different species use the same hole. Thus, the infestation of M. americanus by P. meropis could be explained by a host-switch between the two sympatric bee-eater species via their colonial nesting sites.
Due to the host phylogenetic affinities, it seems obvious to expect that M. americanus should be parasitized by P. mayri; however, it is not possible to speculate at the moment, whether P. mayri, maybe once infesting M. americanus and/or its common ancestor with M. viridis, was outcompeted when occurring sympatrically with M. meropis in the same habitat and host (Figures 1 and 4). More relevant data on the biology and ecology of both ectoparasites and their hosts are needed.
Interestingly, there is another, even more distantly related bird species parasitized by Peristerophila meropis living sympatrically on the Philippines-the Collared Kingfisher Todiramphus chloris (Coraciiformes: Alcedinidae) [26]-which, besides nest-holes excavated in arboreal termitaria, rotten tree trunks, etc., also breeds in earth banks and in the ground, sometimes in nest-burrows excavated by another bird species [40]. Until now, it is the only known non-meropid host of this widespread quill mite taxon. However, for the time being, it is not possible to conclude if this kingfisher is another example of an isolated host-switching event, or P. meropis can be established also on other kingfishers, without detailed investigations of quill mites on other related hosts.
Mapping the ranges of both investigated quill mite species enabled us also to denote their zoogeographic demarcations. Easternmost boundaries of P. mayri range are within the classical Wallace line (after Huxley), delineated with the extent of land at the time of the last glacial maximum. Hosts of P. mayri are residents [6,8,9]. On the other hand, the range of P. meropis is crossing these natural boundaries as far as to Australia; all hosts of M. meropis but M. americanus (which is the only species from different clade) are migratory, some of them long-distance migrants [6,8,9].
Another interesting fact in our parasite-host study system is not only the relatively high prevalence of infestation, compared to other wild avian taxa [21,[41][42][43][44][45][46][47], but a situation when actually all taxa in the whole host clade are parasitized by quill mites. We never recorded a similar situation during our studies of quill mites, it is rather usual that some hosts, despite being sufficiently sampled, are not parasitized [21,[42][43][44]. We expect that such high prevalence and presence on all host species are related to bee-eaters high sociality, when up to hundreds of pairs are breeding in the same colony; cooperative breeding, when, besides parents, several helpers are participating in the raising of the offspring in the nest-burrow; and finally, bee-eaters tendency to perch often touching each other [6,8,38,39].  Until recently, M. americanus was treated as a subspecies of the Blue-throated Beeeater M. viridis, i.e., M. viridis americanus [9,[31][32][33][34]. However, recently, this species was elevated to a full species, the Rufous-crowned Bee-eater [7,8,[35][36][37]. The geographic range of M. viridis stretches from S China, Thailand, and Indochina to Sumatra, Borneo, and Java, whereas M. americanus is confined to the Philippines, thus, they are vicariants [7,36].
We are describing an interesting situation, when two closely related host species, considered until recently to be subspecies of one species, are parasitized by different ectoparasitic quill mites. We suggest the following hypothesis why M. americanus is parasitized by P. meropis, a quill mite species typical for the sister bee-eater subclade: M. americanus occurs sympatrically with M. philippinus on the Philippines archipelago (Figure 4),  Institutional Review Board Statement: Ethical review and approval were waived for this study, due to the use of only dead animals (specimens deposited in the ornithological collection).

Data Availability Statement:
All necessary data (such as localities) are available in the text of this article.