Captive individuals of endangered Philippine raptors maintain native feather mites (Acariformes: Pterolichoidea) species

Endangered species of hosts are coupled with endangered species of parasites, which share the risk of co-extinction. Conservation efforts sometimes include breeding of rare species in captivity. Data on parasites of captive populations of endangered species is scarce and the ability of small numbers of captive host individuals to support the biodiversity of native parasites is limited. Examination of ectosymbionts of the critically endangered Philippine eagles and the endangered Mindanao Hawk-Eagle kept at the Philippine Eagle Center, Philippines, revealed three feather mite species despite regular treatment with insecticide powder. No other ectosymbiont taxa were detected. Studies in morphology and molecular phylogeny of these feather mites based on mitochondrial and nuclear DNA markers indicate that species found were typical for Accipitridae. Three new pterolichoid feather mite species (Acari: Pterolichoidea) were described from two species of eagles (Accipitriformes: Accipitridae) endemic to the Philippines: Hieracolichus philippinensis sp. n. (Gabuciniidae) and Pseudalloptinus pithecophagae sp. n. (Pterolichidae) from the Great Philippine Eagle Pithecophaga jefferyi Ogilvie-Grant, 1896, and Pseudogabucinia nisaeti sp. n. (Kramerellidae) from the Mindanao Hawk-Eagle Nisaetus pinskeri Gould, 1863. The presence of H. philippinensis on P. jefferyi supports the recent finding that the Great Philippine Eagle belongs to the lineage of serpent eagles (Circaetinae) rather than to the Harpy and other eagles.


Introduction
Parasites represent an important component of the ecosystem (Hudson et al., 2006) and support the diversity of the host populations by exerting selective pressure upon their hosts (Dawkins, 1990;Rózsa, 1992). Parasites of endangered species encounter a dual problem. On the one hand, parasites may negatively affect the natural and captive populations of their hosts threatened with extinction (De Castro and Bolker, 2005;McCallum and Dobson, 1995), and on another hand, these parasites often represent endangered species by themselves (Gomez and Nichols, 2013;Rózsa and Vas, 2014). The latter is especially relevant for host-specific parasites (symbionts), such as many ectosymbionts of birds and mammals that often face co-extinction with their host (Buckley et al., 2012).
Host populations of small size harbor reduced diversity of symbiont species due to the parasite loss (Altizer et al., 2007;Lloyd-Smith et al., 2005). The case of the Great Philippine Eagle Pithecophaga jefferyi Ogilvie-Grant, 1896 represents an extreme of minimal population size, both because of being a naturally uncommon apex predator in the islands and of current environmental change and habitat fragmentation, with an estimated 250-750 individuals in total (IUCN, 2017). The extremely low number of Philippine eagles increases the probability of loss for their parasites. Moreover keeping and breeding of rare bird species in captivity for the conservation purposes is also accompanied by the loss of their ectosymbionts mostly due to the antiparasitic treatment (Dunn et al., 2009). Therefore, the survival of the ectosymbionts on the captive group of Philippine eagles was under the question. Besides, its position as apex predator in the ecosystem could facilitate Philippine eagles to adopt alien parasite species from its prey. We tested whether the captive individuals of the Philippine raptors maintained the ectosymbionts and if the ectosymbionts found represented the native fauna of the Philippines eagles studied.
No data on parasites for critically endangered Philippine eagles was available so far; therefore, the study of biodiversity of ectosymbionts in these birds represents an essential need. During ectosymbionts examination of captive Philippine eagles in the Philippine Eagle Center, feather mites were found in spite of the annual antiparasitic treatment (dusting the body, wings and the tail with the powder containing carbamates, Gamma powder, a local producer).
Diurnal birds of prey (Accipitriformes and Falconiformes), a group containing the Philippine eagles, are of the most poorly explored major groups of recent birds in relation to their specific feather mite fauna (Astigmata: Analgoidea and Pterolichoidea). Most collections of feather mites from raptors, especially rare species, have been made from museum skins (Gaud, 1983a;b;Gaud and Atyeo, 1996). Nowadays most species of raptors are endangered and highly protected; therefore, they are not easily accessible for parasitological examinations. All data on parasite-host associations of feather mites and raptors published before the end of 20th century were summarized by Philips (2000). After that, just a few papers on mites from raptors have been published (Dabert and Mironov, 2015;Hernandes, 2017;Mironov et al., 2007;Galloway, 2003, 2014;Pedroso et al., 2015;Proctor et al., 2006).
In the present work, we studied the fauna of feather mites found on two eagles endemic to Philippines based on both morphology and molecular phylogenetic analysis (genes COI, EF-1α, 18S, 28S). We provided descriptions of three species of pterolichoid feather mites and investigated whether these feather mite species likely represent native fauna of Philippine eagles as opposed to species recently acquired through prey-to-host transmission.

Material and methods
The mite material used in the present study was collected in the Philippine Eagle Center (Davao City, Malagos, The Philippines, 7°11′6.29″N, 125°24′55.17″E) from two species of endemic raptors, the Great Philippine Eagle Pithecophaga jefferyi Ogilvie-Grant, 1896 and the Philippine Hawk-Eagle Nisaetus pinskeri Gould, 1863, during annual medical examination of birds by OOT in 2016. Parts of the feathers bearing mites were removed using forceps and a magnifying glass, placed in the tube with 96% ethanol and kept at 4°C for subsequent studies.

Taxonomic study
Some of the collected mites were mounted on microslides in Hoyer's medium according to the standard techniques used for many groups of small acariform mites (Krantz et al.,. 2009). Investigation of mite specimens and drawings were made by SM using a Leica DM 2500 light microscope with differential interference contrast (DIC) and equipped with a camera lucida. Descriptions of new species and measurement methods follow the formats elaborated for corresponding taxonomic groups of mites (Hernandes, 2017;Hernandes and Mironov 2015;Mironov et al., 2007Mironov et al., , 2015Pedroso et al., 2015). General morphological terms and leg chaetotaxy follow Gaud and Atyeo (1996); idiosomal chaetotaxy also follows these authors with corrections for coxal setation by Norton (1998). Descriptions provide the measurements for a male holotype with a range for paratype males in parentheses, and a range for female paratypes. All measurements are in micrometres (μm). Collection data indicate the places of origin and dates of taking of bird individual from nature.

Molecular study
DNA was isolated from specimens fixed in 96% ethanol using Holterman's method (Holterman et al., 2006) with addition of proteinase K and mercaptoethanol in the lysing solution. Sequences of cytochrome oxidase subunit I (COI), elongation factor 1 alpha gene (EF1), partial sequences of 18S and 28S ribosomal DNA subunits 18S and 28S molecular markers were amplified using an EncycloPlus PCR Kit (Evrogen, Russia) with the parameters recommended by the producer on a Biorad T100 amplifier (United States). The sequences of primers used are given in Table 1. Polymerase chain reaction (PCR) products were visualized in gel, cut out, and cleaned using the SV Gel and PCR Clean-Up System kit (Evrogen, Russia). They were then precipitated by ethanol in the presence of ammonium acetate to increase the efficiency of DNA precipitation. DNA sequencing was performed at the Genome Center for Collective Using (Genome, Russia). Molecular markers used and GenBank accession numbers for the sequences of the species studied are presented in Table 2. The sequences were combined and aligned using the ClustalX program after the addition of sequences from the GenBank (Thompson et al., 1997). Subsequently, the sequences were edited using the Genedoc 2.7 program (Nicholas et al., 1997). The phylogenetic trees were reconstructed in the Mr. Bayes 3.2.3 program (Huelsenbeck and Ronquist, 2001) and RaxML (Stamatakis, 2014) in the CIPRES server (Miller et. al., 2010) with the evolutionary model which was selected based on the results of the analysis in jModelTest2 program (Darriba et al., 2015). Sequences of the Amerodectes turdinus (GenBank accession number KU203310) and Amerodectes sp. (GenBank accession numbers KU202819 and KU202968) were used as outgroups for phylogenetic reconstructions. The genus Amerodectes (Analgoidea: Proctophyllodidae) was selected as an outgroup for the Pterolichoidea feather mites studied because this genus is well defined morphologically and represents another superfamily, Analgoidea, a sister lineage to all pterolichoidean mites used in our analysis. Taxa of feather mites used for phylogenetic analysis, their systematics and hosts are summarized in Table 3.
We tested the congruence of operational taxonomic units (OTUs) by the application of two analytical methods: Generalized Mixed Yule Coalescent (GMYC) (Pons et al., 2006) and Automatic Barcode Gap Discovery (ABGD) (Puillandre et al., 2012). GMYC represents a modelbased approach, aiming to discover the maximum likelihood solution for the threshold between the branching rates of speciation, while ABGD detects the statistically inferred barcode gap -difference between the greatest intraspecific distance and the smallest interspecific distance -and uses it to partition the data.
Depositories of type material and voucher specimens used for molecular study are as follows: UMICHZ -Museum of Zoology of the University of Michigan, Ann Arbor, USA; ZISP -Zoological Institute of the Russian Academy of Sciences, Saint Petersburg, Russia.  (Cho et al., 1995;Regier, 2008), (Klimov and OConnor, 2008)
Differential diagnosis. Among previously described species, Hieracolichus philippinensis sp. n. is more similar to H. dobyi Mouchet, 1959 described from Stephanoaetus coronatus (Linnaeus, 1766) in Africa (Gaud and Mouchet, 1959;Gaud, 1983b) in having, in males, setae e2 extending to the level of setae h2 and f2, and relatively short and narrowly lanceolate setae h1. Hieracolichus philippinensis differs from this species by the following features: in both sexes, setae c3 are long, filiform and exceed 100 μm in length, and genual solenidion σ is situated at the base of genu III; in males, setae g are situated almost at the level of anterior genital papillae; setae h1 are short (22-25 μm), and the inner margins of opisthosomal lobes have a pair of noticeably convex membranes in the anterior part of the terminal cleft; in females, the hysteronotal shield is shaped as an inverted trapezium and the posterior one third of the opisthosoma is devoid of sclerotization except the posterior margin, and tarsus IV completely extends beyond the posterior margin of the opisthosoma. In both sexes of H. dobyi, setae c3 are narrowly lanceolate at base with filiform apex (80-90 μm long), and genual solenidion σ is situated at the midlength of genu III; in males, setae g are situated anterior to the level of genital papillae; setae h1 are narrowly lanceolate, curved and 30-35 μm long, and the inner margins of opisthosomal lobes are almost straight; in females, the hysteronotal shield is shaped as an inverted trapezium and the posterior one third of the opisthosoma is devoid of sclerotization except for the posterior margin, and tarsus IV slightly (by ¼ the length) extends beyond the posterior margin of the opisthosoma.
Etymology. The specific epithet is derived from the country, where The genus Pseudalloptinus originally included pterolichine mites associated with birds from the orders Accipitriformes, Falconiformes, Gruiformes, Ciconiiformes and Psittaciformes (Dubinin, 1956;Gaud and Mouchet, 1959). After a revision (Gaud, 1988), the content of this genus was reduced to five species associated exclusively with birds of the order Accipitriformes. The genus Pseudalloptinus is readily distinguishable from other pterolichine genera in having, in most species, a unique structure in males: the postgenital sclerite [ = fossette post-genitale of Gaud (1988)]. This sclerite, being apparently a derivative of adanal apodemes, is situated between the genital apparatus and anal field and usually is stirrup-shaped or roughly ovate.
Differential diagnosis. The new species, Pseudalloptinus pithecophagae sp. n. is most similar to P. africanus Gaud, 1988 andP. milvulinus (Trouessart,1884) in having the following features: in both sexes, setae c3 are lanceolate; in males, opisthosomal lobes are well developed, with semi-ovate terminal membranes; and in females, the striated sejugal area is large and constitutes about 1/5th of the total length of the idiosoma. Pseudalloptinus pithecophagae sp. n. differs from these species by the following features: in males, the genital apparatus is situated at the level of the anterior margin of trochanters IV, epimerites IVa are long and almost extending to the genital arch, and setae e2 are filiform, situated at the level of the anterior end of supranal concavity and not do not extend to lobar apices; in females, the hysteronotal shield is entire, the epigynum is semicircular and extends to the level of setae 4b, setae c1 is situated on the hysteronotal shield, external copulatory tube is minute (only 2-3 μm long), and setae g are situated at the level of setae 3a. In males of P. africanus and P. milvulinus, the genital apparatus is situated at the level of the posterior margin of trochanters III, epimerites IVa are poorly developed, and setae e2 are spiculiform, situated posterior to the supranal concavity and extend beyond the lobar apices; in females, the hysteronotal shield is spit into a large anterior piece and a small pygidial fragment covering the very posterior end of the opisthosoma, the epigynum is bow-shaped and does not extend to the level of setae 4b, setae c1 are situated on striated tegument near the anterior margin of the hysteronotal shield, the external copulatory tube is about 15 μm long and curved ventrally, and setae g are situated posterior to the level of setae 3a.
The unique feature of P. pithecophagae males, easily discriminating this species from all previously known Pseudalloptinus species, is the absence of the entire postgenital sclerite well separated from the adanal apodemes. In this species, L-shaped tips of adanal apodemes turned anteriorly and flank small median area with setae ps3, apparently corresponding to the lateral pieces of the postgenital sclerite of other species of this genus.
Etymology. The specific epithet is derived from the generic name of the type host and is a noun in the genitive case.
Differential diagnosis. The new species, Pseudogabucinia nisaeti sp. n. is close to P. intermedia (Mégnin et Trouessart, 1884) known from falcons by in having, in both sexes, ambulacral discs of tarsi IV extending to or slightly beyond the posterior margin of the body, and
setae c2 exceeding the distance between internal scapular setae si, and, in females, setae f2 and ps2 being equal to or exceeding the distance between their bases. Pseudogabucinia nisaeti sp. n. differs from that species by the following features: in both sexes, subhumeral setae c3 are long filiform and approximately half as long and humeral setae cp, solenidion ω1 of tarsus II does not extend to the apex of this segment; in males, the supranal concavity does not extend beyond the level of setae e1, setae 4a are situated posterior to the base of the genital arch; in females, the genital papillae are situated distinctly anterior to the level of setae g. In both sexes of P. intermedia, subhumeral setae c3 are about 1/3 the length of setae cp, solenidion ω1 of tarsus II extends to the apex of this segment; in males, the supranal concavity extend far beyond the level of setae e1, setae 4a are situated posterior at the of the level of genital arch base; in females, the genital papillae are situated at the level of setae g. Etymology. The specific epithet is derived from the generic name of the type host and is a noun in the genitive case.

Molecular phylogenetics
We obtained sequences for the genes COI, EF-1α, 18S, 28S from four specimens of P. pithecophagae, two specimens of H. philippinensis and one specimen of P. nisaeti (Table 2). Data on different molecular markers studied for feather mites of superfamily Pterolichidae in GenBank are both sparse and variable in coverage (Klimov and OConnor, 2008). Therefore, we did not provide the resulting phylogenetic tree for COI because there were very few sequences for pterolichoid feather mites available in GenBank. Phylogenetic trees for EF-1α, 18S and 28S molecular markers placed the sequences of the Philippine raptor feather mites studied among the other pterolichoid feather mites (Fig. 10, Figure S1, Figure S2). Although only the phylogenetic tree for elongation factor 1 alpha sequences showed congruent topologies between Bayesian and maximum likelihood analyses (Fig. 10). Operational taxonal unit testing analysis by both GMYC and ABGD algorithms supported delimitation of OTU hypothesized by morphological studies for feather mites P. pithecophagae and H. philippinensis from Pithecophaga jefferyi while for P. nisaeti only the ABGD delimitation was significant, which can be explained by the presence of single specimen of the latter species available for analysis.

Discussion
Most of the birds species host several groups of ectosymbionts, including obligatory feather mites and chewing lice species (Mironov, 2016;Price et al., 2003). However, our examination of captive Philippines Eagles revealed feather mites species and no chewing lice were detected. Although we sampled only three individuals of the Great Philippine Eagles, the fact that we found no chewing lice suggests that these insects are much more susceptible to the antiparasite treatment, and endemic lice will likely not survive on captive birds in the Philippine Eagle Center. Loss of chewing lice is not unusual for small populations of endangered species of birds conserved and bred in captivity (Dunn et al., 2009). The feather mites according to results of our examination are capable of surviving annual antiparasitic treatments for a long time. For example, one of the examined birds, named Thor, was captured in the wild in 1974 and at the day of examination in 2016 hosted a viable population of H. philippinensis. This fact, assuming this population of mites is endemic, suggests that these ectosymbionts have been able to survive 43 years in captivity.
We describe for the first time feather mites of two endangered Philippine eagles, which, if they prove to be species-specific, are endangered species too. Based on the phylogenetic position of the species described herein and known reference data on associations of their genera and families, it is possible to drawn out very preliminary hypotheses on the origin of the examined feather mite species from eagles of the Philippines. Of 16 genera of the family Gabuciniidae, eight genera, including the genus Hieracolichus, are restricted to birds of the order Accipitriformes (Gaud, 1983b;Gaud and Atyeo, 1974;Mironov et al., 2007). Most representatives of the genus Aetacarus, with exception of a few species, are associated with raptors. Although the primary origin of the family Gabuciniidae as developing on Accipitriformes is not completely proven, gabuciniids have a maximum of diversity in genera and species on Accipitriformes compared to its other host orders, like Coraciiformes, Caprimulgiformes, and Otidiformes. In any case, it is possible to state that the core of the family Gabuciniidae likely arose on the ancestors of the order Accipitriformes and extensively evolved on these birds. In this light, it is possible to suggest that Hieracolichus philippinensis represents the original feather mite fauna on the Great Philippine Eagle rather than a recently acquired feather mite species. Currently the suprageneric system of the family Pterolichidae is not fully developed (Mironov, 2016). Our attempts to study the molecular phylogeny of the family showed a lack of available sequences in Gen-Bank for many molecular markers, which make it difficult to build a reasonable concatenated tree. Nevertheless, based on the distribution of the genus Pseudalloptinus exclusively inhabiting Accipitriformes (Dubinin, 1956;Gaud, 1988), we could conclude that this genus was probably formed on the ancestors of this order and successfully evolved on these birds. In this case, like H. philippinensis, Pseudalloptinus pithecophagae also represents rather ancient and most likely the primary fauna on the Great Philippine Eagle.
Wide and mosaic distribution of the kramerellid genus Pseudogabucinia among birds orders and within Accipitriformes and Falconiformes (Table 4) strongly contrasts with other genera of the family Kramerellidae that are each restricted to a particular host order (Gaud and Atyeo, 1996). Distribution of Pseudogabucinia representatives on phylogenetically distant genera of raptors of two orders allows us to hypothesize that species associated with accipitriforms could represents some remnants of formerly rich fauna of Pseudogabucinia on these birds. On the other hand, mites of this genus could represent invading fauna transferred from other unknown host groups or rather, a transferrable mite grouping between accipitriform and even falconiform hosts.
The Great Philippine Eagles were historically placed in the subfamily Harpiinae related to other eagles but were recently moved to the family Circaetinae based on molecular studies (Lerner and Mindell, 2005;Ong et al., 2011). Although the host distribution of the genus Hieracolichus is not yet well explored, its preferential occurrence on rather basal lineages (see Lerner and Mindell, 2005) of accipitriforms, Fig. 9. Pseudogabucinia nisaeti sp. n. details. Aopisthosoma of male, ventral view, B-Dgenu, tibia and tarsus I-III of male, respectively, dorsal view, Etibia and tarsus IV of male, F, Gtibia and tarsus III and IV of female, respectively, Gtibia and tarsus IV of female, Hspermatheca and spermaducts.
such as Aegypiinae, Circaetinae, Polyboroidinae (Gaud, 1983b;Philips, 2000), can be considered as additional evidence that P. jefferyi indeed belongs to the lineage of serpent eagles Circaetinae, rather than derived lineages of typical eagles as Aquilinae and Harpiinae.

Conclusions
We showed that a small captive group of endangered birds could maintain viable populations of native feather mites, demonstrating the utility of ectosymbiont examination for host individuals even after decades in captivity. We provided the first record of feather mites from endemic raptors or diurnal birds-of-prey in the Philippines, with three new feather mite species described, and revealed the native origin of the feather mites studied. Our work facilitated an understanding of biodiversity in the understudied family of feather mites Pterolichidae, although many more species should be sequenced before the relations in the family can be resolved clearly by molecular phylogeny.
Mohagan and Anna Gonchar for the help in organization of samples collection and Prof. Dr. Sven Klimpel for providing facilities to work with samples. We also thank Amber Longo for English correction and anonymous reviewers for critical comments on the manuscript.