This study showed that Chorioptes sp. was frequently detected in the outer ear canals of moose (Alces alces) in Sweden. In wildlife, localization of Chorioptes mites in the outer ear canals has also been described in reindeer, Rangifer tarandus , red-flanked duiker, Cephalophus rufilatus  and giant panda, Ailuropoda melanoleuca . In this study the skin of other parts of the body of the moose was not examined for mites and therefore it is not known if the Chorioptes infestation was restricted to the ear canals. Moreover, Chorioptes sp. have been found on alopecic areas of the skin in moose necropsied at SVA, and in one case Chorioptes were also demonstrated in the outer ear canals (pers. comm. C. Bröjer). However, these mites have not yet been closely compared to the Chorioptes sp. isolated from the external ear canals of the moose of this study.
The evidence presented here indicates that the Chorioptes sp. found is a previously undescribed species, which will be described in detail in a later publication. The Chorioptes sp. was closest to C. texanus, based on both morphology and genetics including cloning and sequencing of multiple independent PCR products from skin scrapings, in order to recognize different sequence types that might originate from different individual mites present in the skin, or from different copies of the rDNA which represents a multi-copy gene .
Morphological differentiation between C. texanus and C. bovis is mainly based on the setae at the opisthosomal lobes in male mites (Figure 2). The outer opisthosomal seta, called seta 1  or seta ae , is much longer in C. bovis than in C. texanus, with a length >100 μ m and <100 μ m, respectively [5, 7, 31]. The mites from moose possessed shorter setae than those of C. bovis. Furthermore, they had longer opisthosomal setae 1, 2 and 3 compared to C. texanus mites from cattle. Whether the ear mites of reindeer, affiliated to C. texanus , might actually belong to the same species as the ear mites from moose of this study, can not be decided, since mites were not available for direct morphological and molecular comparison.
When the genotype sequences from the Chorioptes mites from moose were compared with sequences from C. texanus and C. bovis obtained using the same protocols , pairwise identities from the different Chorioptes moose genotypes were 89–91% (C. texanus) and 86–89% (C. bovis). These inter-species identities were thus of the same order as the identity of 89–93% found between C. texanus and C. bovis , for which separate species status is generally accepted. Furthermore, phylogenetic analyses (which were robust to the form of analysis used) indicated that the moose mites formed a monophyletic group, with a sister-group relationship to C. texanus. This separation is interpreted here as an indication of taxonomic separation and not as a reflection of the geographic origin of the mites, because the corresponding sequences of C. texanus and C. bovis grouped according to taxonomic entities and not the geographical origin of the mites, and the mites from Swedish moose did not group together with Chorioptes mites of their nearest geographical origin. Thus, there is no molecular support, either from genetic similarity or sister-group relationships, for placing the Chorioptes mites from Swedish moose in either C. texanus or C. bovis.
In the past, moose have been described as being parasitized by C. texanus in Poland [11, 12], although neither pictures nor descriptions of the mites were given. These mites might be different from those that we found in moose in Sweden, because the mites in Poland were not found in the auricles of the animals but only at several other body sites . We did not investigate non-auricular body sites in our study, leaving open the question as to whether other species of Chorioptes, such as C. texanus, are also found in moose in Sweden or if the Chorioptes sp. identified in this study affects other parts of the skin besides the ear canal.
The species status of the other three Chorioptes species seems doubtful . C. crewei was based on morphological features of only four females and two males taken from the ears of a red-flanked duiker (Cephalophus rufilatus) in Cameroon [30, 32], and detection of this species has never been documented afterwards. Mites of C. mydaus were isolated only once from a stink badger (Mydaus lucifer) in Borneo , and those of C. panda were found in the ears of giant pandas (Ailuropoda melangolenca) in zoos in France  and China . Affiliation of the mites from Swedish moose with C. mydaus or C. panda was difficult to determine, since mites of these questionable species were not available for direct comparison. Morphological comparison was thus limited to published descriptions and drawings [18, 19, 33], which allowed only limited conclusions. The mites from the moose differed from C. mydaus by a shorter seta ae, which was 85 μ m on average compared to 100 μ m in C. mydaus . Additionally, setae l4 and d5 were of a distinct flattened, blade-like shape in the mites from moose (Figure 2), while they were described and drawn to be small or only slightly widened in C. mydaus . The moose mites were regarded as morphologically closer to C. panda based on drawings . The most evident differences were that in C. panda seta ae was less separated by setae l4, l5 and d5, and all of these 4 setae arose at an oblique line from the opisthosomal lobe , while there was a distinct gap between seta ae and the other three seta mentioned in the mites from the moose, and the seta ae arose proximal to a line of the points of origin of setae l4, l5 and d5 (Figure 2). These differences, however, could not be detected in the drawings of C. panda by other authors .
No evidence was found to affiliate the Chorioptes sp. with any of these taxa. The sequence from mites sampled from panda in China [EMBL:EF053123], although not affiliated to a particular mite species, appears to exclude C. panda as a possible identification of the moose mite.
A high percentage of the moose (~81%) was shown to be Chorioptes -infested. Additionally the true prevalence of infestation might actually have been higher since in many cases the moose had been dead for several days, and the mites may had left the carcase. The direct ocular method employed, i.e. raising the temperature of the sampled pieces of ear canals, stimulated the live mites to migrate and thus facilitated their detection and collection. This method is able to detect live mites that are present in relatively high numbers. Detection of mites by skin scrapings, also performed in this study, is more frequently used , and demonstrates only a proportion of the mites actually present. Also, it does not distinguish live from dead mites. In addition, non-burrowing mites may occasionally be found in histological sections, however, most of the mites on the skin surface may be lost during the histological processing.
The majority of the moose were in a poor nutritional state, ranging from subnormal to cachectic. Mite-infested animals were more often recorded at necropsy as cachectic than were uninfested moose. However, possible associations between Chorioptes infestation and the nutritional state and other factors, including age, sex, season and concomitant diseases, were not analyzed because the sample was obtained from animals found dead in the field at different times of the year and sent for diagnostic post-mortem examination. This resulted in a great variability in host related factors and possible biases in sampling.
Since the samples were from wild animals, no observations on the duration of the infestations could be performed. The histopathological changes, however, suggested that most cases were subacute or chronic. Studies describing the pathology of chorioptic mange in various species, including ruminants, are few, and only concern mange caused by C. bovis. The epidermal changes seen in the moose, characterized by hyperplasia with ortho- and parakeratotic hyperkeratosis, sometimes with serocellular exudates and crusts, were similar to those described in sheep and a Japanese serow (Capricornis crispus) [35, 36].
Eosinophils are common in allergic reactions, including those of ectoparasitic origin [37–39]. The inconsistent and sparse dermal tissue eosinophilia in the moose is in contrast to findings in C. bovis -infested cattle and horses, in which eosinophils were numerous [37, 40], but are in accordance with observations in sheep with this mite .
In dermal allergic hypersensitivity reactions in domestic animals, a superficial perivascular distribution of inflammatory cells is commonly predominant [37, 38]. Despite the type of hypersensitivity reaction, T-lymphocytes dominate, while B-lymphocytes and plasma cells are present in lesser amounts [38, 39]. The dermal inflammation in the moose was characterized by perivascular to interstitial infiltrates of T-lymphocytes, plasma cells and B-lymphocytes, suggesting that also in the moose the reactivity may involve a hypersensitivity reaction. In C. bovis -infested sheep, macrophages, lymphocytes and plasma cells were found perivascularly in the superficial dermis, and the reaction was proposed to represent an allergic contact dermatitis .
Dermal infiltrates of plasma cells were prominent in the moose. Plasma cells are reported to be common in interstitial dermatitis in large domestic animals, and may be considered to be of low significance . However, in dogs and cats, plasma cells are described as predominant in late stage pyoderma, but are less frequent in allergic hypersensitivity reactions . A bacterial cause of the prominent plasma cell infiltration in the moose should be considered, because staphylococci, mainly Staphylococcus aureus, were cultured from the moose's ear skin in 6/10 infested ears, but in none of the non-infested moose. Nonetheless, in many species S. aureus belongs to the normal skin flora , and this could also be the case in the skin of the outer ear canals of moose. Irrespective if the ears were cultured positive for staphylococci or not, dermal infiltrates of plasma cells were a feature. Hence, an association between presence of plasma cells and positive culture could not be established. This might indicate another cause than bacterial infection to the plasma cell prominence.