Neotrombicula inopinata (Acari: Trombiculidae) – a possible causative agent of trombiculiasis in Europe

Background For over a decade, the presence of trombiculid mites in some mountain areas of La Rioja (Northern Spain) and their association with seasonal human dermatitis have been recognized. This work aimed to establish the species identity of the agent causing trombiculiasis in the study area. Methods Trombiculid larvae (chigger mites) were collected from vegetation in the Sierra Cebollera Natural Park and in Sierra La Hez during an outbreak of human trombiculiasis in 2010. Three specimens collected from a bird were also examined. Identification was made using morphological and morphometric traits based on the most recent taxonomic sources. A comparison of those mites with specimens of the same species collected throughout Europe was performed by means of cluster analysis with multiscale bootstrap resampling and calculation of approximately unbiased p-values. Results All collected mites were identified as Neotrombicula inopinata (Oudemans, 1909). Therefore, this species is the most likely causative agent of trombiculiasis in Spain, not Neotrombicula autumnalis (Shaw, 1790), as it was generally assumed. No chigger was identified as N. autumnalis in the study area. Neotrombicula inopinata clearly differs from N. autumnalis in the presence of eight or more setae in the 1st and 2nd rows of dorsal idiosomal setae vs. six setae in N. autumnalis. Comparison of N. inopinata samples from different locations shows significant geographic variability in morphometric traits. Samples from Western and Eastern Europe and the Caucasus formed three separate clusters. Conclusion Since the taxonomical basis of many studies concerning N. autumnalis as a causative agent of trombiculiasis is insufficient, it is highly possible that N. inopinata may be hiding behind the common name of “harvest bug” in Europe, together with N. autumnalis.


Background
Nowadays, four chigger mite species are proven to attack humans in Europe causing trombiculiasis -Neotrombicula autumnalis (Shaw, 1790), Kepkatrombicula desaleri (Methlagl, 1928), Blankaartia acuscutellaris (Walch, 1922), and Trombicula toldti Winkler, 1953 [1]. The last one is a mite of unclear taxonomic position, which is still known solely from its type locality in Austria [2]. Blankaartia acuscutellaris is associated with reedy margins of water bodies (rivers, lakes, swamps, etc.). This widespread species principally infests birds inhabiting such biotopes, and humans are occasional hosts [1]. Kepkatrombicula desaleri parasitizes ungulates and has only been recorded on humans by the author who originally described it [3]. Unlike most trombiculids, this species is characterized by the presence of a sucker disk in the apical part of the hypostome that allows it to feed successfully on large hosts [4].
Neotrombicula autumnalis has been reported as the most frequent causative agent of human trombiculiasis in Europe [5][6][7][8][9][10], and its role as "harvest bug" has been attributed without sufficient taxonomic criteria [11,12]. To date, the harvest bug's taxonomy remains unclear. Thus, Richards distinguished several forms of this species when designating the neotype of N. autumnalis [12]. Later, Kepka raised them to the subspecies level and equated two of them with the previously described taxa, Neotrombicula inopinata (Oudemans, 1909) and Neotrombicula vernalis (Willmann, 1942) [13]. Nowadays, both have been assigned the species rank [14][15][16][17]. Therefore, it is highly possible that the common name of "harvest bug" comprises more than one trombiculid species in Europe.
Neotrombicula autumnalis has been found in different locations in northern Spain [18][19][20]. Thus, cases of human and canine trombiculiasis have been recorded annually in La Rioja (North of Spain) from late summer to midautumn [21,22]. Dermatitis associated with chigger bites has repeatedly been a cause of public health concern in La Rioja, and in 2008 a study of trombiculiasis was financially supported by the Regional Government. One of the aims of this project was to correctly identify the chiggers present in this Region, since all cases had been ascribed to N. autumnalis despite no proper taxonomic identification ever having been carried out. Trombiculid larvae were collected in areas where the presence of mites and seasonal dermatitis cases in humans had been previously confirmed [21]. The collection of questing chiggers from vegetation allowed us to suggest the possible causative agent of trombiculiasis in the study area.

Collections
Chigger mites were collected during the scheduled samplings of a three-year study (2008-2010) by members of the Infectious Diseases Department, Center for Biomedical Research of La Rioja. A proportion of the chigger specimens collected on randomly-selected sample days were included in this study. According to the seasonality of human trombiculiasis in La Rioja, field collections were carried out from September to November 2010. Unfed trombiculid larvae ready to seek hosts were collected as follows: 1. The black-plate method, which is commonly used when collecting free trombiculid larvae from the ground, was combined with the flag technique, used to collect ticks on vegetation. Thus, a 20×40 cm black plastic plate with two strings at opposite corners was dragged over the ground. Chiggers were transferred into tubes containing distilled water with the aid of a brush. 2. By direct capture: chiggers on the upper surface of the leaves or on small dry branches were directly collected into tubes.
Additionally, three chiggers were removed with a swab from an Erithacus rubecula (L.) specimen captured during a bird ringing campaign carried out in the same area by Aranzadi Sciences Society and Abies Environment Resources Inc.

Chigger collection sites
Three sampling areas in La Rioja (North-West of the Central Iberian System, North of Spain) were chosen according to the proven existence of human and canine trombiculiasis:

Preparation and examination
Specimens from La Rioja selected for this work were fixed in 70% ethanol; 87 specimens were then mounted on microscopic slides in Faure-Berlese medium under uniform conditions. Clarified preparations were examined under a compound microscope MBI-3 (LOMO plc, St. Petersburg, Russia) with phase contrast optics. Measurements were made with an ocular micrometer and converted to micrometers. Digital photographs of mites were acquired with a Leica DM5000B microscope equipped with a DFC320 digital camera (Leica Microsystems Wetzlar GmbH), using differential interference contrast microscopy. All examined specimens are archived at ZIN (Nos 8250-8336).

Statistical evaluation
The bootstrapping cluster analysis was used to estimate relative morphometric closeness between N. inopinata collected in La Rioja and samples from other regions. The   [26]. Specimens with data missing for any variable were excluded from the analysis. The single specimen from Vitosha Mts was found to be an outlier and therefore was excluded. Lastly, the number of OTUs (Operational Taxonomic Units) was 12, and their sizes were as follows: La Pineda 10, Valdecalvo 5, Ocón 9, Torgny 2, Nový Jičin 10, Kvasivsky Menchul 6, Rakhiv Raion 10, Pirin 8, Nalchik 10, Alagir 7, Mazada 5, and Khnov 1. Mean values were calculated for each of them. Then the UPGMA clustering algorithm (Unweighted Pair Group Method with Arithmetic Mean) applied to the matrix of Euclidean distances between 12 OTUs by the above 17 variables was used. The number of bootstrap replications in the analysis was 1000.
The computations were performed using pvclust module [27] in the "R" software environment ver. 2.14.1 [28]. The package provides AU (approximately unbiased) p-values [29,30] and BP (bootstrap probability) p-values. For a cluster with AU p-value > 0.95, the hypothesis that "the cluster does not exist" is rejected with significance level 0.05 and such clusters can be considered as strongly supported by data [31]. We can also conclude, from our experience [26], that AU p-values slightly lesser than the threshold may also characterize well-delineated clusters.
The comparison of the sample from La Rioja with specimens from the other regions (Table 1) showed that the former was distinguished first of all by shorter dmt and, consequently, smaller m-t (ratio dmt/TaIII where TaIII is a length of leg III tarsus), by a lower number of ventral idiosomal (VS) (Figure 7) and therefore all idiosomal setae (NDV = DS + VS), and by shorter anteromedian seta of scutum (AM). That these differences represent intraspecific variation is evident from their comparison with the ranges of variation in other widely distributed Neotrombicula species [38][39][40]. Moreover, a concordance between morphometric and geographic distances among OTUs can be clearly seen from Figure 8, where three clusters with significant statistical support (AU p-values > 0.95) correspond to groups of geographically closest samples (three OTUs from Spain, two from Caucasus, and three from Ukraine plus Bulgaria). It is noteworthy that upper-level grouping of all 12 OTUs, although without the same support, follows the geographical (meridian) sequence of sample sites (Figure 4), creating three large clusters consisting of Western European, Eastern European, and Caucasian samples, respectively.
The fauna of trombiculids in Spain, based on collection of chiggers from small mammals, reptiles and birds included 18 species, according to Kepka [18,41], Varma [42], Pereira-Lorenzo [19], and Imaz et al. [20]. In particular, 10 trombiculid species were recorded on Apodemus sylvaticus (L.) from Navarra, a region bordering La Rioja [20]. The description of N. inopinata collected from vegetation in La Rioja has raised to 19 the number of trombiculid species in Spain: Blankaartia acuscutellaris (Walch, 1922), Figure 7 Neotrombicula inopinata, ventral aspect of idiosoma. To collect chiggers from vegetation is of primary importance in order to find the agent causing human trombiculiasis. Chigger mites located on grass, shrubs or low tree branches are unfed larvae ready to attack their hosts. Chigger infestation in humans is very probable due to direct contact with these substrates during excursions, hiking or working. All human cases of trombiculiasis recorded in La Rioja have been diagnosed in people that had been in contact with the vegetation from Sierra Cebollera Natural Park during the season when the peak of abundance of N. autumnalis and related species is usually recorded in Europe [6,14,43]. Affected people do not detect the presence of trombiculid mites. In fact, chigger bites are noticed later on, due to the associated intense itching. By then, the mites are detached and examination of the patients skin does not reveal any "harvest bug". The association between chigger mites and human trombiculiasis in the study area is only based on epidemiological evidence. In accordance with this, collections of chiggers from vegetation have been made previously in Spain [21,44]. All collected chiggers have been identified as N. autumnalis but with insufficient taxonomic evidence. Since N. autumnalis is closely related to N. inopinata, the characteristics that distinguish these species should be examined in each case in order to confirm the identification. Neotrombicula inopinata clearly differs from N. autumnalis by the presence of eight or more setae in 1st and 2nd rows of dorsal idiosomal setae ( Figures 5 and 6), rarely six or seven setae in one of these rows, while N. autumnalis always has six setae in both rows. Accordingly, the number of dorsal idiosomal setae is 36-45 in N. inopinata and 26-33 in N. autumnalis. Futhermore, N. inopinata is slightly larger than N. autumnalis (larger scutum, longer setae, and longer legs), but intervals for individual morphometric traits overlapped when our entire sample was compared to specimens of N. autumnalis collected throughout its geographic range [38].
It should be noted that taxonomical basis of many studies concerning trombiculiasis is inadequate or nonexistent. A notable case was the record of N. autumnalis on a little bittern from Turkey [45], where no taxonomic evidence was provided. In fact, the existence of this species in Turkey is not confirmed [17]. Occasionally, when authors try to support their identifications by morphological data, these are surprisingly inappropriate. For example, Guarneri et al. [46] provided some descriptions in order to justify their record of N. autumnalis in Italy. Their method included traits common to larvae of several mite families, such as three pairs of legs with three claws, paired eyes, hooked chelicerae, segmented palps, and some of the typical traits shared by many chigger mite genera such as pentagonal scutum with five setae and two sensilla, nude galeal seta, and tricuspid palpal claw. Kampen [9] treated N. inopinata and N. vernalis as "subspecies or ecotypes" of N. autumnalis following old papers of Kepka [13,47], ignoring all subsequent development of chigger mite taxonomy. Schöler et al. [10] stated that the only species involved in their investigation is N. autumnalis. Nevertheless, only outdated references [13] and no information about the identification process are included. In the light of these facts, most published data concerning N. autumnalis as the causal agent of trombiculiasis should be considered as doubtful. A confirmation of the role of this species in human trombiculiasis, based on new collections and appropriate examination, is required. In Spain, human trombiculiasis has been only described in the study area [21], and no cases have been reported from the remainder of the country. According to our data, N. inopinata is the only chigger species on vegetation during the seasonal outbreak of trombiculiasis. Therefore, it is the only candidate for the causative agent of trombiculiasis in this territory. However, we cannot conclude that N. inopinata is the only mite of medical and veterinary importance in northern Spain. To date, no other trombiculid species has been found in La Rioja, although additional chigger collections from birds and small mammals may increase the trombiculid fauna of the region, especially with the species previously recorded in Navarra [20]. Thus, it is highly probable that N. autumnalis parasitizes small mammals in the study area, although its role as causative agent of trombiculiasis is unclear. The different height of locations where unfed chiggers wait for their hosts (e.g. vegetation, low tree branches or forest floor), could result in the different host spectra of diverse trombiculid species [39,48]. According to our results, since N. autumnalis has not been collected from vegetation, it would be unlikely to attack humans or large animals, whereas N. inopinata would have a privileged position to infest humans. The identification of engorged trombiculid mites that have developed their stylostomes, directly removed from humans or domestic animals is essential to define the etiological agent of trombiculiasis.

Conclusion
Neotrombicula inopinata is the most likely causative agent of trombiculiasis in northern Spain, rather than N. autumnalis, the species usually regarded as the main human-attacking chigger mite in Europe. Taking into consideration that N. inopinata is widely distributed in Europe, this species could be hidden behind the common name of "the European harvest bug" together with N. autumnalis, or even exceed the latter in medical and veterinary importance.