Hidden biodiversity in microarthropods (Acari, Oribatida, Eremaeoidea, Caleremaeus)

A challenge for taxonomists all over the world and across all taxonomic groups is recognizing and delimiting species, and cryptic species are even more challenging. However, an accurate identification is fundamental for all biological studies from ecology to conversation biology. We used a multidisciplinary approach including genetics as well as morphological and ecological data to assess if an easily recognizable, widely distributed and euryoecious mite taxon represents one and the same species. According to phylogenetic (based on mitochondrial and nuclear genes) and species delimitation analyses, five distinct putative species were detected and supported by high genetic distances. These genetic lineages correlate well with ecological data, and each species could be associated to its own (micro)habitat. Subsequently, slight morphological differences were found and provide additional evidence that five different species occur in Central and Southern Europe. The minuteness and the characteristic habitus of Caleremaeus monilipes tempted to neglect potential higher species diversity. This problem might concern several other “well-known” euryoecious microarthropods. Five new species of the genus Caleremaeus are described, namely Caleremaeus mentobellus sp. nov., C. lignophilus sp. nov., C. alpinus sp. nov., C. elevatus sp. nov., and C. hispanicus sp. nov. Additionally, a morphological evaluation of C. monilipes is presented.


Molecular phylogenetic analyses (part 1)
Phylogenetic analyses such as Neighbor joining (NJ) and maximum parsimony (MP) were performed using PAUP* version 4.02a [1]. These topologies were statistically supported by bootstrap values (BT; 5,000 replicates for NJ, 1,500 for MP and 1,000 for ML). Using the program MODELTEST version 3.06 [2] the TrN+I+G model was selected for NJ and MP analyses. Maximum Parsimony was performed using a heuristic tree search, random addition of sequences (1,000 replicates) and the TBR branch swapping algorithm (2,000 replicates). Uninformative characters were excluded from all analyses and all characters were considered with equal weights. Maximum likelihood (ML) trees were conducted in RAxML-7.0.3-WIN [3]. The constructed trees were combined in a 50 % majority rule consensus tree by means of the program PAUP*. To assess whether the topologies obtained by four different tree building algorithms differ significantly, a SH-test [4] was performed as implemented in PAUP*.
(1) GMYC was conducted by means of the splits package as implemented in R version 3.3.2 [5], http://rforge.r-project.org/). The steps for GMYC analyses were conducted as recommended by Talavera et al. [6]. Cadence v 1.0.1 [7] was used to evaluate substitution rate differences for COI and EF-1α dataset.
Distribution of branch length was plotted for every individual for each tree. There were no considerable variations in the relative rates among ingroup taxa, therefore a strict molecular clock was applied for both gene fragments. For the GMYC analysis ultrametric input trees are required, therefore a combined ultrametric tree was created (MCMC simulation with 200 million generations, sampled every 1000 th generation, Yule tree model, a clock rate of 2.15 for COI, and an estimated rate for EF-1a was entered) by means of the program BEAUti as implemented in BEAST v1.8.3 [8]. TRACER v.1.7.1 was again used to verify the chains had reached stationarity. The remaining post burn-in trees were combined with TreeAnnotator 1.8.3 (also implemented in the BEAST package, settings after Casquet et al. [9], posterior probability limit of 0.5). A single-threshold was first employed due to its better performance in delimitation [10], however, the multiple-threshold analysis showed same results. (2) The bPTP method is an updated version of the original maximum likelihood PTP and adds Bayesian support values to delimited species on the input tree. bPTP analyses require a bifurcated phylogenetic tree (not an ultrametric tree like the GMYC model). Therefore, the BI of the concatenated dataset (tree settings mentioned above) were uploaded to the web server (https://species.h-its.org/ptp/) applying 500,000 MCMC generations. The removal of the outgroup had no impact on species delimitation results of the ingroup. (3) The ABGD Barcoding gap analysis was conducted with the COI dataset and default settings (simple distance and K2P gave the same results) via the ABGD web server (http://wwwabi.snv.jussieu.fr/public/abgd/abgdweb.html). ABGD is an automatic procedure that sorts the sequences into putative species based on distance. The ABGD is a non-tree-based method and only requires an alignment file. (4) For mPTP, a BI tree in Newick format (same as for bPTP, from the BI analysis with branch length) was submitted to the mPTP web server (https://mptp.hits.org/#/tree). The multi rate poisson tree processes method was selected. (5) BPP, which is a multilocus species delimitation analysis was conducted by means of the program bppX (version 1.2.2), which is the graphical interface of the bpp program (http://abacus.gene.ucl.ac.uk/software.html). To confirm the consistency between runs, we performed five independent runs. We also used different combinations of priors for ancestral population size (θ) and root age (τ0) as recommended by Leaché & Fujita [11] and Yang & Rannala [12], because of the fact that these parameters can affect the posterior probabilities for models. The obtained posterior probability values showed consistency between runs and among different prior settings. Prior five clades (and two outgroup clades) were defined and confirmed as delimited species. Except of the ABGD method (COI dataset) the concatenated dataset was analysed for species delimitation analyses. The use of multiple genes (which are unlinked) by concatenation of data present a more adequate approach in species delimitation, although nuclear genes evolve more slowly and therefore the statistical support or the resolution of revealed clades may be lower [13].
Supplementary references -part 1  Biol. Evol. 29, 1969-1973(2012.        Description of species and remarks (part 2) All light microscopical observations were made with mites mounted in permanent slides. During the hardening of the embedding medium the specimens roll easily on one side because of the shape of the notogaster (in Michael's slides too), therefore most drawings show a body slightly tipped over.
The abbreviations for morphological structures used in text and figure legends follow Norton & Behan-Pelletier [1].

Common characters of all species described below
A very detailed generic description was published recently by Norton & Behan-Pelletier [1]. But it seems necessary for the reader to give an overview and to summarize characters common to all the species described below and to avoid repetitions in the text of the species descriptions. Adaxial of bothridia one pair of inwards curved slightly barbate interlamellar setae (in) of medium length situated on subrectangular sclerotized and foveate areas, which bear an enantiophysis (dt) directed towards the anterior border of notogaster. Between these areas cuticle without foveae.
Notogaster: Clearly separated by a deep dorsosejugal furrow from prodorsum. Contour more or less elliptical in dorsal aspect. The bulges of the notogaster were described correctly by Michael [2], (p. 17) as follows: "Immediately behind the anterior margin there is a broad, rounded, transverse elevation, not reaching the lateral margin. Behind this is a deep, linear depression, and then the centre of the abdomen, until within a quarter of its length from the hind margin, is occupied by a domed lump, followed by a smaller one, which touches the hind margin." (S2Fig. 6). Cuticle smooth with transverse irregular row of foveae in the area posterior of the transverse anterior bulge; at both Gnathosoma: Subcapitulum half-oval; diarthric (S2Fig. 7). Cuticle of genae and rutella smooth.
Fixed digit with two terminal teeth and one subterminal tooth; moveable digit with two terminal teeth and three subterminal teeth. Träghård's organ (T) distinct, elongate and tapered. Setae cha and chb attenuate, cha barbed. Nearby seta cha one to three tiny cuticular spines, their number seems to be species-specific.
Several enantiophyses and cuticular tubercles present across epimeral borders in varying numbers.
The following enatiophyses are well developed in all studied specimens: enantiophyses eSa and eSp on lateral border of deep sejugal furrow as well as enantiophyses e4a and prominent cuspidate e4p; the latter pair spans deep epimeral groove IV laterally. Discidium (dis) directed laterally. Posterior border of epimeron IV with medial fossa; the contour of the fossa varies from trapezoid to parabolic.
Legs: All legs monodactylous, with big falcate claw consisting of two layers: an inner smooth one and an outer one with spines and striae (see S2Fig. 11). Femoral porose areas present as well as trochanteral porose areas on trochanters III and IV. On tarsus I famulus (ε) long with rounded tip.

Description of species
Only species-specific characters are mentioned in the following descriptions which are completed by line drawings and SEM-micrographs.
Caleremaeus monilipes (Michael, 1882) Apart from Michael's description there is a generic diagnosis in Grandjean [3], and several short morphological notes for "Caleremaeus monilipes" are in determination keys; incomplete descriptions with figures are to be found in Subías & Arillo [4] and in Ayyildiz et al. [5]. Some figures are given by Miko & Travé [6]. Mentum without foveae, its cuticular surface finely granulate. On adaxial side of chelicera two small cuticular spines below insertion of seta cha.
Six pairs of genital setae. Rostrum slightly contoured because of irregular cuticular ledges and pits. Rostral setae (ro) similar in size as in.
Ventral region (S2Fig. 18): Several enantiophyses present across epimeral borders except epimeral border I. Genital plates each with six smooth setae; distance between g5 and g6 longer than between other setae.
Derivatio nominis: "mentobellus" means a beautiful mentum, ornamented with many foveae in the anterior half of mentum.