First description of the larva of Dinaraea Thomson, 1858, with comments on chaetotaxy, pupa, and life history based on two saproxylic species from Europe (Staphylinidae, Aleocharinae, Athetini)

Abstract The paper describes the morphological ultrastructure of the previously unknown early (L1) and late larval instars (L2–3) of Dinaraea, including chaetotaxy, pupal cocoon, prepupa, and pupa, based on the saproxylic species D. aequata Erichson and D. linearis Gravenhorst. Diagnostic larval characters for the genus Dinaraea are given for the first time. Morphological differences between mature larvae of these two species relate to the colouration and degree of flattening of the body, details of antennal structure, anterior margin of the labrum, mandibles, and mala. The differences are relatively small, probably because of the similar ecological preferences of both species. As in the case of other aleocharine larvae, L1 in Dinaraea differs from L2–3 in the lack of some setae on the dorsal surface of the head and thorax, and on the abdominal tergites and sternites; the presence of a subapical seta on the urogomphi; egg bursters on some thoracic and abdominal tergites; a darker antennal segment III; and the relatively longer urogomphi and their apical setae. The differences established in the features of the chaetotaxy of L1 and L2–3 between Athetini (Dinaraea), Oxypodini (Thiasophila) and Homalotini (Gyrophaena) correspond with the molecular marker-based relationships of these taxa.


Introduction
The genus Dinaraea Thomson, 1858 (Staphylinidae, Aleocharinae, Athetini) includes 21 species worldwide, 12 of which are known from the Nearctic and nine from the Palaearctic; five of the latter (D. aequata Erichson, D. angustula Gyllenhal, D. arcana Erichson, D. hungarica Ádám, D. linearis Gravenhorst) occur in Europe. They are small insects (the lengths of the European species are 2.5-3.7 mm) with a subparallel, flattened body, and the integument has a distinct meshed microsculpture and distinct punctation. The head is large, subquadrate to slightly elongate, and the genae are usually longer than the eyes. Mainly saproxylic, Dinaraea species inhabit the subcortical galleries of other insects. They are also found in rotting tree trunks and in the fruiting bodies of various polypores. Because of their environmental preferences and their probably predatory mode of life, most species of this genus are potentially important as enemies of economically significant forest pests. To date, however, the diet of these rove beetles remains unknown, as do other aspects of their biology (Benick and Lohse 1974, Nikitsky and Schigel 2004, Klimaszewski et al. 2013, Löbl and Löbl 2015.
Nothing is known of the morphology of the preimaginal stages of Dinaraea. This is not particularly surprising, since very little information is available on the external structure of other Athetini taxa, just as is the case with most Aleocharinae. The larvae of only a few of the more than 170 genera classified among Athetini are known (Paulian 1941, Pototskaya 1967, Topp 1978, Ashe and Watrous 1984, Ashe 1985, Newton et al. 2000, Ashe 2005. What is more, such descriptions as do exist are usually fragmentary and relate to just a few features illustrated in diagrams. The very poor state of knowledge regarding the larvae of these staphylinids makes it almost impossible to make use of their morphologies in phylogenetic analyses. Only Ashe (2005), in a work on the phylogeny of the tachyporine group subfamilies and 'basal' lineages of the Aleocharinae, took into account features of both imagines and larvae, including three genera of Athetini. It turns out that the larval characters to a large extent stabilise the phylogenetic tree covering the taxa under consideration in the present work. In the case of other aleocharines, this same author also used larval morphologies to examine phylogenetic relationships within the subtribe Gyrophaenina (Ashe 1986). The results turned out to be confluent with the morphological analysis of the imagines of these Staphylinidae. They point to the distinct monophyletic origin of that subtribe and are strongly underpinned especially by the external features of these Aleocharinae. The greater usefulness of larval than imaginal stages in phylogenetic analyses was also demonstrated by Grebennikov and Newton (2009) in the case of ten subfamilies in the Staphylininae group. Again, Pietrykowska-Tudruj et al. (2011 highlighted the great importance of larval features in establishing the systematic membership of Quedius antipodum Sharp and the genus Astrapaeus Gravenhorst. The results substantiate data obtained from analyses of adult morphology and/or DNA sequences, and suggest the separate position of Q. antipodum in relation to the north temperate genus Quedius and the genus Astrapaeus within the tribe Quediini.
The necessity to take larval morphological features into consideration in future phylogenetic analyses and assessments of the systematic membership of Staphylinidae thus seems wholly logical. Unfortunately, a major obstacle to doing so is the insufficient and often extremely fragmentary nature of the relevant data, as mentioned above, which applies in particular to the subfamily Aleocharinae, including the tribe Athetini.
The main aim of this study is to describe in detail the external morphology, including the chaetotaxy and ultrastructure, of the early (L1) and late (L2-3) larval instars of Dinaraea based on D. aequata and D. linearis. The paper also includes data on the external appearance of the hitherto unknown pupa of this genus, as well as the feeding preferences and the life histories of both species.

Materials and methods
Larval and pupal stages of the two species were obtained by rearing five adults of D. aequata and four adults of D. linearis. Specimens of D. aequata were collected at Parchatka near Kazimierz Dolny (51°22'54.55"N, 21°59'51.53"E, SE Poland) on 11 November 2004. The insects were sifted from the remains of birch bark, in deciduous woodland growing in a shady, damp loess gully. Individuals of D. linearis were collected at Łańcuchów near Lublin (51°16'15.33"N, 22°55'20.35"E, SE Poland) on 3 rd December 2004. These beetles were sifted from pieces of bark torn off a wind-thrown ash (Fraxinus excelsior L.) in an old riparian wood of ash and alder (Circaeo-Alnetum) in the valley of the River Wieprz, a dozen or so metres from the river bank. D. aequata and D. linearis were reared from 19 November 2004 to 20 January 2005 and from 6 December 2004 to 21 February 2005, respectively, at room temperature (20 °C ± 3). Adults and larvae of both species were kept separately in plastic containers (diameter 10 cm, height 2 cm) filled with moist soil. Larvae were fed various sizes of small springtails of the family Onychiuridae. The immature stages (larvae and pupae) were killed in boiling water and preserved in ethanol (75%). The adults were identified by the first author.
The material examined for morphological study and measurements is listed in Tables 1 and 3. Chaetotaxy nomenclature, symbols, and abbreviations follow Ashe and Watrous (1984), and the morphological description style is according to . The voucher specimens are deposited in the collection of the Department of Zoology, Marie Curie-Sklodowska University, Lublin.

Morphological comments on the Dinaraea pupa (based on D. aequata)
Because of the poor state of preservation of most of the reared research material, this description covers only the ventral part of the female pupa.

Notes on the life history in the laboratory
In the rearing of adults of D. aequata (temp. 20 °C±3), started on 19 November 2004, the first larvae (L1 and L2) were observed just six days later (25 November), and eight and ten days later the first prepupae and pupae respectively appeared. Various late developmental stages (mostly L3 and pupae) were observed until 20 January 2005. Larval development in the rearing of adults of D. linearis, started on 6 December 2004, was observed from 1 January to 11 February, and pupation from 19 to 21 February 2005. The larval and imaginal forms of both species were fed exclusively with springtails from the family Onychiuridae. On several occasions foraging larvae of D. aequata were observed as they caught their victims of various sizes in their mandibles. Within the following 5 to 10 minutes they consumed most of the springtail bodies, but always rejected fragments of the carapace.

Discussion and summary
This paper describes in detail the external morphology of the hitherto unknown larval stage of the genus Dinaraea, including the chaetotaxy, using the terminology proposed for the subfamily Aleocharinae by Ashe and Watrous (1984). The description is based on individuals from European populations of D. aequata and D. linearis, the larvae of which were bred in the laboratory from imagines. The morphological differences between the mature larvae of the two species relate to: (a) the dimensions of various body parts, they are larger in D. aequata (Table 1); (b) colour of head, thoracic tergites and anterior abdominal tergites and sternites, generally darker in D. aequata; (c) degree of body flattening, somewhat more flattened in D. aequata; (d) shape of article II of antenna and Sa of article III, slightly more elongate in D. linearis; (e) structural details of the anterior margin of the labrum and the length of its setae Ld2; (f ) number of subapical teeth on both mandibles, more on the left-hand one in D. aequata, more on the right-hand one in D. linearis; (g) shape of mala, more protracted at adoral margin in D. aequata; (h) number of cuticular processes on mala, more in D. aequata (Table  4). These diagnostic features distinguish the larvae of these two species, which may co-occur under natural conditions. Particularly useful in this respect are features a-c, relating to the dimensions of the body and its general appearance. Very many more features distinguishing aleocharine larvae belonging to the same genus were established for Haploglossa Kraatz, 1865 (Staniec et al. 2010). A comparative analysis of H. picipennis (Gyllenhal, 1827) and H. nidicola (Fairmaire, 1852) revealed differences not only in larval size and colouration, but also in the chaetotaxy of abdominal segment X and epicranium (presence or absence of seta Ea1), structural details of all the mouthparts, lengths of the several leg parts, and urogomphi. It is likely that the scale of the morphological differences between the larvae of species from one genus depends largely on their different ecological preferences. The above-mentioned Haploglossa species inhabit micro-environments (H. picipennis -nests of raptors, H. nidicola -nest holes of sand martins) that are very different from those of Dinaraea species, which are usually found under bark, mainly of broad-leaved trees. Measurements of the head and pronotum of the two Dinaraea species indicate that their larval development involves three stages (Table 1): this is typical of most known aleocharines (White 1977, Ashe and Watrous 1984, Ashe 1985, Ashe 1986, Zagaja et al. 2014. Only in the case of Pella species (P. laticollis) (Lomechusini), inhabiting ants' nests, were just two larval stages found; this is due to the faster rate of development of these rove-beetles (Hölldobler et al. 1981, Staniec et al. 2009). This feature is probably an adaptation to the myrmecophilous lifestyle that aims to minimise the period during which staphylinid larvae are potentially endangered by their hosts in the anthill.
Morphological analysis of a Dinaraea larva revealed a series of differences between its first (L1), and its second (L2) and third (L3) instars, whose external structures are identical to that of L1. Apart from the clearly smaller body size (see Table 1), features exclusive to L1 include: (1) the absence of some setae on the dorsal surface of the head and thorax, and on the dorsal and ventral surfaces of the abdomen, (2) the presence of short subapical setae on the urogomphi, (3) egg bursters on some thoracic and abdominal tergites, (4) a darker terminal antennal segment than in L2-3, and (5) markedly longer urogomphi and their apical setae than in later stages. These morphological differences between the younger and older larval instars in Dinaraea are of a similar nature to those in other tribes of Aleocharinae (Table 2) (Ashe and Watrous 1984, Ashe 1986, Zagaja et al. 2014. They enable one to easily distinguish L1 from the older larval stages without recourse to metric analysis. In view of the fragmentary nature of the available information on this subject, it is not possible to state definitively whether and to what extent these differences extend across the whole subfamily.
A combination of 17 diagnostic features (see "Generic diagnosis…") have been proposed for the mature larva of Dinaraea, described above, which enable it to be distinguished from other known older larval stages of Aleocharinae (Paulian 1941, Pototskaya 1967, Topp 1975, Ashe 1981, Ashe and Watrous 1984, Ashe 1985, Ahn 1997, Jeon and Ahn 2009, Staniec et al. 2009, 2010, Zagaja et al. 2014. In this respect the genus Dalotia (Da) Casey, (Athetini) most closely resembles the genus Dinaraea (Di) (Ashe and Watrous 1984), and the small differences relate solely to: (i) body habitus -moderately dorso-ventrally flattened in Di or cylindrical in Da; (ii) shape of antennal article I longer than wide in Di or wider than long in Da; (iii) structure of anterior margin of labrum central region protruding and crenate in Di or wholly rounded and smooth in Da; (iv) seta Da1 on pronotum absent in Di or present in Da; (v) setae D3 and P5 on abdominal sternite I absent and present respectively in Di, or present and absent in Da. It should be added that the dorso-ventrally flattened body and moderately elongated antennal segment I are most probably adaptations to the under-bark lifestyle of this larva, as in the case of the mature form of Dinaraea. On the other hand, the structure of the anterior margin of the labrum, and particularly the specific features of the body chaetotaxy, may be of phylogenetic significance. That is why these structures have been included in the morphological descriptions of other Aleocharinae larvae (Ashe and Watrous 1984, Ashe 1985, 1986, Ahn 1997, Jeon and Ahn 2009, Staniec et al. 2009, 2010, 2017. Only a few of these species has the chaetotaxy not only of older but also of younger larval forms been described (Zagaja et al. 2014. A preliminary comparative analysis of the features of the chaetotaxy was carried out on the basis of well-researched larvae of three different tribes of Aleocharinae. This revealed that all the larval stages of Athetini (Dinaraea) and Oxypodini (Thiasophila) are much more similar to one another than to the larvae of Homalotini (Gyrophaena) ( Table 2). In the first two taxa the slight differences in chaetotaxy of L1-3 concern only the number and homology of the setae on the tergites and first sternite of the abdomen. By contrast, a distinctly smaller number of setae develop on all body tagmata of L1-3, especially the thoracic ones, in Homalotini than in the other two tribes. These data indicate the distinctly closer relationship of Dinaraea (Athetini) with Thiasophila (Oxypodini) than with Gyrophaena (Homalotini). This state is partially corroborated by Thomas (2009), whose research was based on the molecular analysis of two mitochondrial DNA markers. This author showed that Homalotini are a group separate from the other taxa he/she analysed. The taxa from Athetini and Oxypodini belong to sister groups, at least in some of the trees generated.
Another question relates to the significance of the features of chaetotaxy for phylogenesis depending on the larval stage. Ashe (1986) suggested that later developmental stages would be more useful at lower taxonomic levels (e.g. genus), as they exhibit more features associated with a better-developed chaetotaxy. In contrast, the first larval stages, with their smaller numbers of setae, could be phylogenetically significant in the analysis of higher systematic units. This hypothesis is convergent with that underlying our preliminary studies. These have shown that at the tribal level of Aleocharinae, distinct differences may occur in the chaetotaxy (e.g. between Athetini and Homalotini), between not only the late larval stages but also the early ones. Interestingly, in L1 of some species these differences may be even greater than in the older larval stages, e.g. the chaetotaxy of abdominal tergites I-VIII of Dinaraea and Thiasophila (Table 2).
It should also be borne in mind that the development of setae on the different body parts of the larvae during ontogenesis is uneven. In the taxa we are analysing here, the fewest setae appear on the head -two pairs at most, if any at all (Gyrophaena). In contrast, the changes in the chaetotaxy are the greatest on the pronotum, and somewhat less so on the abdominal tergites and sternites (Table 2) . Being homologous aspects of the chaetotaxy, they could be of phylogenetic importance, especially at lower systematic levels, e.g. subtribe or genus, but this would have to be confirmed by further research.
As a representative of Athetini, the pupa of Dinaraea possesses general structural features, such as an exarate body type, lightly sclerotised, with numerous setae growing from basal, cuticular protuberances and double gonotheca on the ventral surface of the final segment in the female, characteristic of the pupal stages of other aleocharines from Homalotini, Lomechusinii, and Oxypodini (Ashe 1981, Staniec et al. 2009, 2010, Zagaja et al. 2014. Its specific appearance relating to the body outline, width of head, shape of pronotum, length of legs and antennae, shape, and length of mouthparts, resembles the features of the adult form.
As in Dinaraea, the production of a pupal, silken cocoon, into which particles of the surrounding substrate are often woven, has been observed in numerous tribes of Aleocharinae, such as Athetini, Aleocharini, Corotocini, Falagriini, Homalotini (including members of subtribes Bolitocharina, Gyrophaenina and Homalotina), Hypocyphtini, Liparocephalini, Lomechusini, Oxypodini, and Placusini (Ashe 1982, Frank and Thomas 1984, Thayer et al. 2004, Staniec et al. 2010, Zagaja et al. 2014. Some authors assume that this behaviour may be at least a basal condition of the higher classification of Aleocharinae. Topp (1975) even suggested that this feature occurs exclusively in this subfamily of Staphylinidae. However, this statement seems controversial in the light of reports from other researchers, who described a similar structure in taxa from the subfamilies Steninae and Staphylininae, although in the latter case, it is made mostly from soil (Weinreich 1968, Staniec 2004, Pietrykowska-Tudruj and Staniec 2007, Staniec et al. 2008. The pupal cocoon undoubtedly plays a protective role. That is why it is probably so common in the Aleocharinae, including Dinaraea, in which the delicate exarate pupae are enclosed in a weakly sclerotised cuticle (Staniec et al. 2010, Zagaja et al. 2014, the present study). Likewise, in the case of some Staphylininae, whose larvae spin cocoons (e.g. Gabrius splendidulus, Rabigus tenuis), their pupae are covered by an exceptionally thin cuticle compared with other members of this subfamily (Pietrykowska-Tudruj and Staniec 2007, Staniec et al. 2008. Presumably, then, pupation within a cocoon could have evolved independently in members of many different subfamilies of a range of rove-beetles, as an adaptation associated with the pupal structure and/or the biotic conditions of the environment, e.g. pressure on the part of predators. Nevertheless, knowledge of the life history of Staphylinidae, including the occurrence of a pupal cocoon in this, the largest family of beetles, remains fragmentary and is restricted to a very small number of taxa summarised for Staphylinidae by Frank and Thomas (1984). Thus, more data are required before more binding conclusions regarding this structure can be drawn.
In Europe, both Dinaraea species are quite widespread saproxylic rove-beetles, although both in Poland and some other countries D. linearis appears to be far less common than D. aequata. The flattened and parallel-sided body of adults and larvae (the present study) of these staphylinids are probably a consequence of their mode of life under tree bark; they do not display any particular preferences as regards the species of tree they colonise. D. aequata is also encountered in various arboreal fungi. In Poland both these rove-beetle species can be observed in nature all the year round, along with some 30 other species of saproxylic Coleoptera, from six families (Nikitsky andSchigel 2004, Alexander andAnderson 2012, Melke, oral information;authors' unpublished data).
In the rearing, the development of D. aequata and D. linearis took place in the autumn and winter months (November-February), as in the case of Phloeonomus punctipennis Thomson (Omaliinae) -another saproxylic staphylinid, whose larval stages were caught in the field in the second half of November ). It is not exactly known, however, whether the above-mentioned reproductive period of Dinaraea in the laboratory coincided in time with its reproduction in nature, or whether it was more the effect of the suitable ambient temperature at which the rearings were performed. Certain information was also obtained as regards the diet of these rove beetles, which was hitherto completely unknown (Klimaszewski et al. 2013). On the basis of laboratory observations, these rove beetles are presumably predators, which in natural conditions hunt for various tiny arthropods with delicate cuticles and consume their soft tissues.