A new genus of mites of the subfamily Platyseiinae associated with Azteca ant galleries in Cecropia trees in Costa Rica (Acari: Mesostigmata: Blattisociidae)

The genus Calyptoseius gen. nov. of the subfamily Platyseiinae Evans is described, based on adults and nymphs of one newly described species associated with ants of the genus Azteca occupying hollow stems of Cecropia in lowland rainforests of Costa Rica. Several unusual morphological attributes are noted, particularly the autapomorphic presence of four elongated setae on each of telotarsi II to IV. Some perspectives of these mites in the Cecropia-Azteca association are discussed, including a possible dispersal link via nematoceran flies to gain access to such an unusual habitat. The definition of the previously monobasic genus Cheiroseiulus is augmented, and also modified in view of an undescribed species at hand, and the subfamily definition is modified to account for morphological aspects of the new genus. A key to the genera of the Platyseiinae is given.


INTRODUCTION
This presentation continues a series of papers centered on description of highly distinctive and biologically interesting new taxa of gamasine mites found to exist in one small area of lowland tropical rainforest of the La Selva Biological Station in Costa Rica (Lindquist and Moraza 2008, 2010, 2012, 2014Moraza and Lindquist 2011, 2015. Here, we describe a new genus and species of platyseiine mite that is unusual in having a life history perhaps confined to association with Azteca ants which are mutualistic occupants of the stem hollows of Cecropia trees (Longino 1991a). Other platyseiine mites are known as freely living occupants of a variety of soil and litter substrates worldwide, and are especially abundant in moist habitats, such as along stream and lake margins, in mosses submerged in running water, and seasonally wet substrates. However, some Cheiroseius occur in bracket fungi on damp decaying wood in tropical regions (Lindquist 2003), while others occur in drier forms of leaf litter and arable soils (Karg 1993).

MATERIALS AND METHODS
The mites were collected from a funnel-extraction of one field-collection of Azteca ant colony material at the La Selva Biological Station, Heredia Province, Costa Rica in 1995. Extracted mite material was preserved in 80 % ethanol, with specimens mounted from ethanol directly into Hoyer's medium on microslides and sealed with Glpt insulating varnish.
Taxonomic concepts of the superfamilies of mesostigmatic mites, especially the Phytoseioidea and its component families, and the genera of Blattisociidae, follow those presented by Lindquist et al. (2009) and Lindquist & Moraza (2010, 2012. Morphological observations, measurements, and illustrations were made using compound microscopes equipped with differential interference contrast and phase contrast optical systems, drawing tubes and stage-calibrated eyepiece micrometers. Setal notation for the idiosoma follows that of Lindquist & Evans (1965) as modified slightly by Lindquist (1994) and adapted for superfamilies of mesostigmatic mites in general by Lindquist et al. (2009). Measurements of structures are given in micrometers (µm) and indicate the ranges among specimens measured. Dorsal shield lengths were taken as midline length from the anterior margin anterior to the bases of vertex setae j1 to the caudal margin posterior to the bases of clunal setae J5. Lengths of ventral idiosomatic shields are midline, from the anterior margin to the posterior edge of each structure, including the postanal cribrum. Notation for leg and palpal setation follows that of Evans (1963Evans ( , 1964. Leg lengths are taken from the base of the coxa to the apex of the tarsus excluding the pretarsus. Distinction of pore-like structures on the idiosomatic integument as either poroids (lyrifissures) or glandular openings (solenostomes), as distinguished morphologically by Athias-Henriot (1969, 1975 and physiologically by Krantz & Redmond (1987), is presented stylistically in the illustrations; gland pores are shown in circular form, while poroids are shown in elliptical form as in Figure 1A. Notation for pore-like structures of the peritrematal region follows that of Johnston & Moraza (1991). Identification of the ant species was made by John Longino (see acknowledgments).

SYSTEMATICS
The subfamily Platyseiinae Evans (1957) is a cladistically distinctive group, with its history and con-cepts being reviewed by Lindquist (2003). Together with the sister group Blattisociinae, they are currently thought to comprise the family Blattisociidae in the superfamily Phytoseioidea ). Members of the Phytoseioidea share what is hypothesized to be a uniquely apomorphic "phytoseioid-type" of sperm access system, which consists of a weakly sclerotized tube, the major duct, that leads from the solenostome opening externally between the bases of legs III and IV to an atrium with an embolus from which emanates a fine minor duct; the atrium connects to a usually sclerotized calyx which then leads to a blind, sac-like vesicle (Athias-Henriot 1968, Evans 1992, Alberti and Di Palma 2002. The complexity of this system makes it unlikely that it has been derived more than once from a plesiomorphic "laelapoid-type" of sperm access system (Alberti 2002, Moraza andLindquist 2011). Within this systematic framework, the possibility that platyseiines may be derived from within the Blattisociinae (e.g., from a Lasioseius stock), as noted by Lindquist (2003), is not excluded.
The discovery of Cheiroseiulus (Evans and Baker 1991) as a new monobasic genus of Platyseiinae required modification of the concepts of that subfamily as presented previously by Lindquist & Evans (1965), since adults of that genus are morphologically exceptional, as follows: The dorsal shield is markedly hypotrichous, its podonotum with 17 pairs of setae, its opisthonotum with only five pairs (absence of all but one pair of setae in each of the Jand S-series); adult females and males have an anal (rather than ventrianal) shield, and metapodal plates are rudimentary or absent; adult females lack metasternal platelets, leaving setae st4 inserted on soft cuticle; legs II to IV have the paradactyli and median lobe of the pulvillus elongate, but blunt instead of acute apically. Aside from those exceptions, the genus was readily accommodated in the Platyseiinae by sharing the following apomorphic attributes of that subfamily Baker 1991, Lindquist 2003): The fixed digit of chelicerae has a deep subapical receptacle for receiving the apex of the movable chela; the anteriormost pair of hypostomatic setae and inner palptrochanter seta are elongated and somewhat whip-like; the tarsi of legs II to IV have the median lobe of the pulvillus slender (whether blunt or acute apically), and the paradactyli flanking the claws are elongated, often more so than the median pulvillar lobe; tarsi II to IV each have anterodorsal seta ad-2 elongated and flattened, somewhat strap-like, along much of its length, tapering and curved only near its apex; tarsi II and III have a second seta (pd-2) similarly elongated and flattened; the femur of legs I and II has 11 and 10 setae, respectively, each lacking seta v-3.
One of us (EEL) has studied an adult female of an undescribed species of Cheiroseiulus, which is readily distinguished by the dorsal shield being less hypotrichous than that of the type species. This and other differences are accounted for in our augmented description of the genus, and in our key to platyseiine genera in this paper.
Although not emphasized previously, the Platyseiinae including Cheiroseiulus are also distinctive apomorphically in the form of cheliceral dentition and the prominence of the pseudosymmetric pair of apical dorsal setae (d-1) on tarsi II to IV. The fixed digit of the chelicera has a ridge between the apical hook and the pilus dentilis which bears a row of very fine or barely discernible teeth that do not (or hardly) extend proximally beyond the insertion of the pilus dentilis. The denticulate ridge is apposed at either extremity by one of the two teeth of the movable chela (see figures 40 of Lindquist and Evans 1965, 196 of Karg 1993, 2b of Evans and Baker, and 12.55H of Lindquist et al. 2009). On the tarsi of legs II to IV, the apical dorsal setae are conspicuous, longer than the pretarsus, and often reach distally to the same area as the apices of the paradactyli (see figures 2c of Evans and Baker, and 22 of Lindquist 2003).
As was the case with Cheiroseiulus, placement of our new genus in the Platyseiinae requires cladistic rationale; this is presented in the remarks following its description. In turn, the description of the subfamily must be modified slightly, to accommodate this placement (see Discussion).
Diagnosis -Adults (and deutonymphs) of the new genus are immediately distinguished autapomorphically from those of other platyseiines as well as from other members of the superfamily Phytoseioidea in having the tarsi of legs II to IV each with four elongated, strap-or whip-like setae (ad-2, pd-2, al-2, pl-2); those of other platyseiine genera have at most three setae (ad-2, pd-2, pl-2) strap-like on tarsus II and two (ad-2, pd-2) strap-like on tarsi III-IV. They are also distinguished apomorphically from other platyseines in having the third pair of sternal setae alone clearly longer than the other sternal setae, the dorsal and lateral setae of most leg segments collectively palmate-ciliated, and the male having a discrete ventral shield between the sternitigenital and anal shields. They are distinguished plesiomorphically from those of other platyseiines in having 12 and 11 setae on the femur of legs I and II, respectively (by retaining pv-2), and 11 and 9 setae on the genu of legs II and III (by retaining pv-1). Among the genera of Platyseiinae, Calyptoseius shares with Cheiroseiulus the following apomorphic attributes: opisthogaster of female and male with an anal shield; gland pores gv3 on soft integument adjacent to anal shield; and legs I to IV with paradactyli and medial lobe of pulvillus untapered, blunt-tipped. However, adults of Calyptoseius are similar plesiomorphically to those of Cheiroseius in having a holotrichous dorsal shield with 21 to 23 pairs of setae on the podonotal region and 15 to 18 pairs on the opisthonotal region, and in having 10 setae on tibia II (seta pd-2 present).
Adult male -Dorsal shield with form and sclerotization as in female except more broadly united to peritrematal shields at level of setae r4. Dorsal shield with complement and form of setae similar to that of female, except with usually two more pairs of marginal setae including r6 and sometimes R1. Lateral soft cuticle with five or six pairs of Rmarginal setae. Peritremes as in female.
Deutonymph -Dorsal shield with short lateral incisions, similarly sclerotized as in adults. Dorsal shield with complement of 30 pairs of setae, including 15 podonotal and 15 opisthonotal pairs, of similar length and shape as in adult; 14 pairs of dorsal setae on surrounding soft integument (z1, s1, s2, r2-r6, R1-R6). Dorsal complement of discernible porelike structures as in adult. Peritrematal shields free, limited to a narrow strip with two widened areas, one at anteriormost dorsal region bearing gp1, one at medial area bearing gp2, ip2; a short poststigmatic extension with ips1; gland pore gp3 and poroid ip3 on a tiny platelet on soft cuticle behind stigma; peritremes well developed, narrow, extending to level of setae s1.
Idiosomatic venter -Adult female. Tritosternum with pilose laciniae free for most of length, their fused section and base without elaborations (Fig. 1E). Ventral shields unornamented, weakly sclerotized. Presternal region without platelets. Sternal shield entire, with well developed endopodal extensions between coxae I-II bearing gland pore gvb distally, and with those poorly developed between coxae II-III; shield with three pairs of setae, st3 the longest, and two pairs of lyrifissures; setae st4 and lyrifissures iv3 on soft cuticle ( Fig.  2A). Endopodal strips between coxae III and IV free, weakly defined, distant from sternal shield. Epigynal shield with its convex hyaline anterior margin between legs II-III abutting or barely overlapping posterior edge of sternal shield, its posterior margin slightly convex; setae st5 on shield's lateral margins, paragenital poroids iv5 on soft cuticle; postgenital furrow present, its strip of transverse platelets hardly discernible. Soft opisthogastric integument with small, undivided metapodal platelets, five pairs of setae (JV1, JV2, JV5, ZV1, ZV2), and four pairs of poroids, flanked by posteriormost two pairs of R-setae. Opisthogaster with anal shield ovoid, with paranal setae inserted at midlevel of anus, and similar in size to postanal seta; anal valves with euanal poroids; shield with a well-developed cribrum behind level of postanal seta; gland pores gv3 on soft cuticle well removed from shield. Peritrematal shield connected with exopodal strips curving behind coxa IV, with two poroids (ip3, ips1) and one gland pore (gp3) in area behind stigma; shield with poroid ip2 and gland pore gp2 along widened area below level of setae r3-r4 (Figs 1A, 2B). Exopodal strip weakly developed freely alongside peritrematal shield by coxae II-IV, with small extensions between I-II, II-III, III-IV (not shown in figures). Spermathecal apparatus of phytoseioid-type, with minor duct emanating from embolus near base of sclerotized calyx, and with major duct leading from calyx to solenostome between coxae III-IV (Figs 8A, B).
Adult male.
Aspects of tritosternum and presternal area as in female. Ventral shields unornamented, weakly sclerotized. Sternitigenital shield with five pairs of setae and three pairs of poroids, and fully contiguous with well developed endopodal extensions between coxae I-II, II-III, and III-IV (Fig. 3A); relative lengths of sternal setae as in female. Opisthogaster with discrete ventral shield bearing one or two pairs of setae (ZV1 laterally and JV1 on or barely off posterior margin); soft opisthogastric integument with metapodal platelets as in female, four or five pairs of setae (JV2, JV5, ZV2, ZV3, sometimes JV1), and three pairs of poroids, flanked by posteriormost two pairs of R-setae. Opisthogaster with discrete anal shield, with aspects of its form, three circumanal setae, and cribrum as in female. Posteriorly and laterally, peritrematal shield, peritreme, and exopodal strip as in female.
Deutonymph. Tritosternum and presternal area as in female. Ventral shields unornamented, weakly sclerotized. Sternal shield with four pairs of setae and three pairs of poroids (iv3 present); setae st5 and poroids iv5 on soft cuticle; shield without endopodal extensions, but endopodal fragment well developed, apically with gland pore gvb, between coxae I and II; endopodal fragments between coxae II-III indiscernible (Fig. 6B); narrow rim of exopodal plate behind coxa IV with gland pore gv2. Opisthogaster similar to female; soft integument with five pairs of setae, three pairs of poroids, and gland pore gv3. Anal shield with aspects of its form, three circumanal setae and cribrum as in female.
Protonymph. Tritosternum and presternal area as in deutonymph. Sternal shield weakly delineated, with three pairs of setae and one pair of poroids (iv1 indiscernible); endopodal extensions between coxae I-II hardly discernible but gvb present (Fig. 7B); intercoxal soft cuticle with tiny setae st5 between bases of legs IV. Opisthogaster with well delineated anal shield surrounded by soft integument with four pairs of setae (JV1, JV2, ZV2, JV5) and two pairs of discernible pore-like structures including gv3. Anal shield with structures much as in deutonymph, except postanal seta somewhat shorter than paranal setae.
Gnathosoma -Gnathotectum with three denticulate branches (Fig. 1C), median branch not elongated. Chelicera slender, its shaft moderately elongated (but dimorphically shorter on male), without any conspicuous process along antiaxial or paraxial lateral surfaces near bases of the digits; fixed digit with setiform pilus dentilis and row of few small teeth along apical one-third of masticatory surface, and an offset tooth (gabelzahn) subapically (Figs 2C, D); movable cheliceral digit of female and nymphs bidentate, with margin of arthrodial envelope smooth; movable digit of male unidentate with spermatodactyl digitiform, free to its connection at base, directed anteriorly and somewhat ventrally. Deutosternum with usually seven (variably eight) transverse rows of denticles, anterior five rows similarly moderately wide, sixth and seventh rows somewhat wider; all rows similarly spaced from one another, multidenticulate and indistinctly connected by lateral edges. Corniculi normal in form, entire, well separated from base to apex, with blunt-tipped salivary styli appressed to dorsal cornicular faces; male with blunt lobe medial to insertion of corniculus (Fig. 3C); internal malae normal in form, somewhat longer than corniculi, fimbriated laterally, not subdivided. Subcapitular setae similarly smooth, attenuate, hp1 slightly longer, more whip-like than hp3 and pc, which clearly longer than lateral hp2 (Fig. 1D). Palpi with normal setation as described for Gamasina by Evans (1964); palptrochanter with inner seta longer, more whip-like, than outer seta; palpfemur with seta pd palmate, spiculate, like most dorsal setae on legs, and with al spatulate; palpgenual setae al-1 and al-2 more or less spatulate; palptarsal apotele two-tined (Figs 1F, G).

Etymology
The genus name is a Latinized combination of the term calypto, based on the Greek kalyptos, meaning 'concealed' or 'covered', and seius or sejus, a Roman surname used by many authors to form names for genera of mesostigmatic mites. The name is masculine in gender, and is intended to refer to the hidden niche in which these mites are found associated with ants in Cecropia trunks.

Distribution and habitats
The new genus is currently based on one newly described species. Nymphs and adults are known only from one cohabitation of a nest of Azteca ants in a stem hollow of the trunk of a Cecropia tree (see discussion).

Remarks
Placement of this new monobasic genus requires modification of the concepts of the subfamily Platyseiinae as presented previously by Lindquist (2003), since adults of this genus are morphologically exceptional, as follows. On deutonymphs and adults, autapomorphically the telotarsi of legs II to IV each have four setae modified as elongated strapor whip-like setae, while plesiomorphically, the femora of legs I and II have 12 and 11 setae, respectively, each having three ventral setae (pv-2 present), and the genua of legs II and III have 11 and 9 setae, respectively, each having two ventral setae (pv present). Also, adult males have a discrete ventral shield separate from an anal shield, an apomorphic attribute otherwise noted for a few species of the ascid genus Arctoseius, e.g., A. weberi Evans (Lindquist 1961). A basically more complete chaetotaxy of the leg femora and genua is the same as for the family Blattisociidae as a whole, thus eliminating one of the apomorphic distinctions between the Platyseiinae and its sister subfamily Blattisociinae. However, such aspects of leg chaetotaxy are prone to homoplasy, and as noted above, a conceptual rationale for the Platyseiinae remains on firm cladistic footing, with both apomorphic and plesiomorphic differences accommodated in the revised description of the subfamily in the discussion.
Aside from those exceptions, Calyptoseius is readily accommodated in the Platyseiinae by sharing the following apomorphic attributes of that subfamily Baker 1991, Lindquist 2003): The cheliceral structure of the fixed digit includes a deep subapical receptacle for receiving the apex of the movable chela, followed by a ridge between the apical hook and the pilus dentilis that bears a row of very fine or barely discernible teeth, and is apposed at either extremity by one of the two teeth of the movable chela. The anteriormost pair of hypostomatic setae and inner palptrochanter seta are elongated and somewhat strap-or whip-like. On legs II to IV, the pretarsi have the median lobe of the pulvillus slender (whether blunt or acute apically), and the paradactyli flanking the claws are elongated, often more so than the median pulvillar lobe; the apical dorsal setae (d-1) of the tarsi are prominent, longer than the pretarsi to the base of the claws; the anterodorsal seta ad-2 is elongated and flattened, somewhat strap-like, along much of its length, tapering and curved only near its apex, and the posterodorsal seta (pd-2 on tarsi II and III) is similarly elongated and flattened. A revised description of the subfamily is provided in the discussion.

Type material
All specimens extracted from one sample of a colony of Azteca constructor Emery from galleries in the hollow internode system of a large tree, Cecropia obtusifolia Bertol., felled near the comedor at the La Selva Biological Station, Heredia Province, Costa Ria (10°26'1"N, 84°1'2"W, elevation 50-150 m): HOLOTYPE: adult female, 16 May 1995, coll. J. Longino and R. Vargas C. PARATYPES: 6 adult females, 5 adult males, 3 deutonymphs, 3 protonymphs, with same data as holotype. The holotype and one paratype male deposited in the Instituto National de Biodiversidad (INBio) of Costa Rica, Santo Domingo de Heredia; other paratypes are deposited in the Canadian National Collection of Insects and Arachnids (CNCI), Science and Technology Branch, Agriculture & Agri-Food Canada, Ottawa, and the Museum of Zoology, University of Navarra (MZU-NAV), Pamplona, Spain.

Etymology
The specific epithet honors our colleague John (Jack) T. Longino, a world authority of the systematics and ecology of ants, and one of the key initiators of the Arthropods of La Selva Project (ALAS), Costa Rica. He collected the sample from which mites were extracted and used to describe this new taxon, and he continues to support a wide variety of investigations relevant to Project ALAS.

Remarks
Adult males of Calyptoseius longinoi have a peculiarly formed pair of structures associated with the corniculi. Each is a short, blunt lobe arising medially, closely beside where the corniculus inserts into the subcapitulum (Fig. 3C). Unlike the process that arises from the corniculus base in both sexes of some melicharine mites (Moraza & Lindquist 2015), this one does not appear to emanate from the corniculus and is not evident in females. It is not to be confused with the longer, somewhat blunt apex of the salivary stylet, which may be projected into that area from distortion in slide preparations (as shown in Fig. 7C of a protonymph).

Some morphological aspects of Calyptoseius longinoi
Weak sclerotization of idiosomatic shields -The dorsal shield and the ventral shields of the coxisternal and opisthogastric regions of adult Calyptoseius longinoi are similar in being relatively weakly sclerotized and nearly unornamented. This condition contrasts with the typically well sclerotized and conspicuously ornamented dorsal and ventral shielding found among adults of species of Cheiroseius and Platyseius, and with the dorsal shield of Cheiroseiulus. Strong shielding may serve as both physical armor and protection against water loss in these free-living mites, which (as mentioned above) occur in a wide variety of habitats, often subject to considerable swings in moisture content. Weakly sclerotized shields seem to correlate with adaptations to living in sequestered habitats, in this case perhaps an existence largely confined to the hollow spaces in nodes of Cecropia stems, where both the host plant and perhaps even the tolerance of the ant associates provide protection.
Epigynal shield proportions -The hyaline flap of the epigynal shield is expansive in area, its length occupying nearly forty percent of the shield's entire length, its anterior margin broadly rounded, and its width exceeding that of the more sclerotized posterior portion. Such a configuration would seem a modification to accommodate either larviparity or extrusion of a very large egg (none of the females examined were gravid).
Sperm access system configuration -Although components of this system are difficult to discern among the few females of Calyptoseius at hand, they are clearly of the phytoseioid type, with a long minor duct emanating from an embolus in the calyx region. As in the new taxon, a weak or indiscernible sclerotization of the calyx region, along with a diverticular lobe where the minor duct is connected, has been noted or illustrated for species of Platyseius (Lindquist 2003), and also for Cheiroseiulus reniformis (Evans & Baker 1991, Lindquist 2003; although a minor duct wasn't discerned for the latter, one is present in the undescribed form of Cheiro-seiulus at hand. Other than the description and illustrations for two species of Cheiroseius from South Africa by Jordaan et al. (1987) and one from South America by Mineiro et al. (2009), the spermathecal apparatus has been ignored in descriptions of well over 100 other species in other parts of the world. Typically, among females of many species of Cheiroseius we have examined, a sclerotized calyx is present, from which a long, minor duct emanates and extends into a mass of convolutions, as shown to some extent by the figures of Jordaan et al. (1987). In this respect, the configuration of the system in Calyptoseius longinoi is more similar to that of Platyseius and Cheiroseiulus than to Cheiroseius.
Structural peculiarities of leg tarsi II to IV -While seta pl-3 is simple and attenuated on basitarsi II-III, it is bushy and short, similar to the other three setae of that segment on basitarsus IV. Strangely, this contrasts with the form of the pseudosymmetric dorso-apical pair (d-1), which is relatively short, blunt on telotarsi II-III, yet attenuate on telotarsus IV. Scant attention has been given to these aspects in previous descriptions of platyseiine mites. A somewhat similar pattern of differential elongation of seta pd-3 on basitarsi II-III versus basitarsus IV was noted and illustrated by Lindquist (2003) for two species of Platyseius, yet setae (d-1) were consistently/uniformly attenuated on telotarsi II-IV. Whether this pattern of differential elongation of pd-3 on basitarsi II-IV generally holds for members of the genus Platyseius is uncertain, and the literature provides no information in this respect for members of the more speciose genus Cheiroseius, leaving no basis for phylogenetic considerations.
Although the subfamily Platyseiinae is defined in part by the apomorphic modification of setae (d-2) being elongated and straplike on tarsi II and III, there is some convergence evident among some species of Lasioseius (Naeem et al. 1985, our personal observations). The general pattern among those species is that seta ad-2 is modified on tarsi II and III, seta pl-2 is similarly so (or less straplike) on tarsus II, and with or often without such a modified seta (either al-2 or pl-2) on tarsus IV. Such a convergence was part of the rationale for Krantz to place his newly proposed genus Hyattella in the Platyseiine, rather than as an aberrant subset of Lasioseius (Krantz 1962). However, the pattern is not only sporadic among the over 100 species we have studied, but varies in several respects. For examples, tarsus IV may have a similarly modified seta, but it is al-2, as in Lasioseius (=Hyattella) epicrioides (Krantz, 1962) and L. (=H.) americanella (De Leon, 1964) (EEL personal observations), and L. elegans Naeem et al. and L. chelaserratus Naeem et al., 1985), or that seta may be whip-like and pd-2, as in L. peterfuldi (Ohmer et al. 1991); tarsus II may be the only tarsus with a modified (whip-like) seta, but it is pl-2, as in several other species of Lasioseius (Mineiro et al. 2009, Brito et al. 2011; or setae (d-2) may be long and attenuated but not strap-like, not distinctly different from adjacent tarsal setae (EEL, unpublished observations). A similar convergence is seen in the monobasic blattisociine genus Arrhenoseius, in which the elongated strap-like setae are ad-2 on tarsi II to IV and pl-2 on tarsus IV (there being only one such seta on tarsus III) (Walter & Lindquist 2001). These examples demonstrate the plasticity in form and homologies of such modified setae among blattisociine taxa, yet emphasize the homogeneity of expression of these setae among platyseiine taxa.
Apart from being autapomorphic for the genus Calyptoseius, the striking elongation of four setae (d-2, l-2) on each of tarsi II to IV is an enigmatic attribute in terms of its functional significance. In general, the legs of platyseiine mites have two such elongated setae on telotarsi II-III and one on telotarsus IV, along with at least a somewhat elongated setal pair (d-1), elongated paradactyli, and an acuminate medial lobe of the pulvillus, all of which are thought to be adaptations for movement in moist substrates (Evans & Hyatt 1960, Karg 1981. Perhaps, the presence of four elongated setae on tarsi II-IV may be a further adaptation to moving in moist or sticky substrates. The stem hollows of Cecropia plants invaded by Azteca ants contain highly moist mounds of matter (see below) (Longino 1991a(Longino , 1991b. Strangely, however, in contrast with most other platyseiines, the pretarsal structures (paradactyli, pulvillar lobes) of Calyptoseius are less strongly modified.

Cecropia-Azteca niche dynamics
In view of many hundreds of samples taken from a wide variety of substrates at the La Selva Biological Station, the finding of mites of Calyptoseius only in a single sampling of a colony of Azteca ants in a trunk hollow of Cecropia indicates a particular association with this niche. The sample size is small (a total of 20 specimens were retrieved), but its inclusion of protonymphs and deutonymphs as well as both sexes of adults indicates that the life history of this mite occurs in that niche. Although very little is known about the feeding habits of platyseiine mites, the rather uniform morphology of their cheliceral structure (as noted above) and limited feeding experiments indicate they prey on nematodes and soft-bodied microarthropods, often in regularly or periodically moist substrates (Karg 1971, Hinton 1971. The hollow internodes of Cecropia stems occupied by Azteca ants are very damp, fully humid, but not inundated, as they are well above any part of the tree that would be flooded. Nesting behavior and structure of Azteca ants differs among different species of the ants. In the case of Azteca constructor Emery, the so-called "carton nest" is typically a spindle-shaped mass, some 30cm long, inside a swollen internode section of the central bole of the tree. The carton consists of a crusty black material altered by the ants from the inside parenchyma pith walls of the stem and shaped into a honeycomb of passages and chambers. The carton nest is well separated from a variable series of domatia in internodes of higher branches of the tree. The domatia contain regular mounds of brown, sticky, bran-like material called "knollen", composed primarily from pith scraped by the ants from the internode walls. Internal passages through the internodes are made and maintained by the ants to the knollen, which writhe with numerous nematodes and fly larvae (Longino 1991a). The primary fly that undergoes its life history in the knollen is Alepia longinoi Quate andBrown, 2004 (Longino, personal note, 22 November 2015). Whether Calyptoseius mites may access all of these components of the ant colony as sources of prey via the ants' passageways is uncertain. Mites were observed, both in the nest and amidst the knollen, but their identities were undetermined (Longino, personal note, 16 May 1995, personal correspondence 19 May 2015Ronald Vargas C., personal correspondence 23 November 2015).
The Cecropia-Azteca association has been found to be an interacting community of species, even in a local area like the La Selva Biological Station in Costa Rica, where disturbed lowland rainforest with second-growth trees includes two or three species of Cecropia growing sympatrically and supporting two to four species of Azteca, all obligate inhabitants of Cecropia (Longino 1991a(Longino , 1991b. As there is no absolute host plant specificity, with each Azteca species being found in a variety of Cecropia species, competition may be intense, with multiple queens and multiple species initiating colonies in individual saplings and vying for dominance. On one hand, there are more than 100 species of Cecropia in the Neotropical region, and on the other, Azteca is a diverse genus of ants, with over 150 species, many of which are associated with myrmecophytes other than Cecropia. Beyond that, there are other obligate ant inhabitants of Cecropia, including several species of the ponerine genus Neoponera. Although never as abundant as the dominant Azteca specialists, they are widespread in the Neotropics, and almost certainly represent multiple independent colonizations of Cecropia from within the genus Neoponera (Longino, pers. obs.). Thus, the extent or range of association of Calyptoseius longinoi within this plant-ant association, or whether there may be a variety of such congeneric mites, is completely problematical.
Adaptation by these mites to a coexistent way of life with ants as part of the Cecropia niche dynamic involves another challenge -their dispersal between such specialized niches. Wingless working caste Azteca ants restrict their movement to the host tree itself; they do not forage off the host tree (Longino 1991a). Thus, any dispersal of the mites would seem to depend on winged insects, whether the alate ants or associated flies like Alepia longinoi that come to coinhabit the ants' nests and knollen. Based on limited observations, a general pattern among platyseiine mites that occupy transient niches may be their behavioral preference for dispersal on nematoceran flies, whose larvae live in a remarkable diversity of damp to fully aquatic habitats, some in unusual substrates (Whitsel & Schoeppner 1973, Rack 1976, Ishikawa 1979, Walter & Lindquist 1995. Perhaps such dispersals offer opportunistic access to niches available for specialized colonization. If this scenario is the case with Calyptoseius longinoi, then access to an association with ants in Cecropia trees would be secondary to the mites' colonization of such a habitat by way of dispersal on Alepia flies. Dispersal to such a particular niche would depend on a phoront with such preference for this niche. Alepia longinoi is known only from this niche and from two widely separated and ecologically distinct localities in Costa Rica, one in lowland rainforest, and one in cloud forest at 1190 m (Quate & Brown 2004).

Descriptive notes for the genus Cheiroseiulus Evans & Baker
The original descriptions of this monobasic genus and its type-species did not provide a variety of morphological details significant for comparisons with other platyseiine genera. Also, our study of an adult female of an undescribed species of this genus has facilitated some distinction between genus-and species-group attributes. Thus, the following descriptive notes are based on observations by one of us (EEL) of the type material of Cheiroseiulus reniformis, as well as of the undescribed form, along with further observations by Ann Baker (personal correspondence, June 1994).
Tritosternum with base elongated (two to four times as long as wide), with pilose laciniae free for most (70 -80 %) of length. Ventral shields unornamented or weakly so; presternal region without platelets. Female sternal shield with well developed endopodal extensions between coxae I-II and moderately so between II-III, with three pairs of attenuated setae of similar length; lyrifissures iv3 and adjacent setae st4 (sometimes absent) on soft cuticle. Female epigynal shield with convex hyaline anterior margin abutting or barely overlapping posterior edge of sternal shield, its posterior margin slightly convex; paragenital poroids iv5 on soft cuticle; postgenital furrow present, its strip of transverse platelets hardly discernible. Male sternitigenital shield with four pairs of setae (st4 absent) and three pairs of poroids, and fully contiguous with endopodal extensions between coxae I-II, II-III and III-IV. Opisthogaster of female and male with metapodal plates rudimentary or absent; with anal shield oval or oboval, its paranal setae inserted at midlevel of anus, and clearly (ca 3x) longer than postanal seta; soft cuticle with four pairs of setae, and gland pores gv3 well removed from anal shield. Peritrematal shield well connected with exopodal strips curving behind coxa IV. Components of female spermathecal apparatus difficult to discern, with a weakly or indiscernibly sclerotized calyx region with a diverticular lobe where minor duct emanates (if discernible).
Gnathotectum with three denticulate branches, median one slightly longer. Chelicerae slender, its shaft elongated (excluding basal section, ca 0.6 dorsal shield length); fixed digit with short file of fine to hardly discernible teeth on ridge flanked by two teeth in apposition on movable digit. Deutosternum with seven transverse rows of denticles, distance between sixth and seventh rows less than that between fifth and sixth or other successive rows. Subcapitulum with setae hp1 longer, more strap-or whip-like than hp3, pc sometimes also attenuated but not whip-like. Palptrochanter with inner seta longer, sometimes more whip-like, than outer seta; palpfemur with dorsal setae smooth, attenuate, and with seta al slightly spatulate; palpgenual setae al-1, al-2 slightly spatulate.

Genus-group affinities among the Platyseiinae
In his review of generic concepts in the subfamily Platyseiinae, Lindquist (2003) concurred with Evans and Baker (1991) that the genus Cheiroseiulus may have phylogenetic affinities with Platyseius, in distinction to Cheiroseius. The former two genera share the putatively apomorphic attributes of an incomplete series of J-setae on the opisthonotum, nine setae on tibia II (pd-2 absent), and a dissimilar spacing of the proximal rows of hypognathal denticles. In contrast, the more speciose genus Cheiroseius was not readily definable apomorphically, as it is distinguished by a series of attributes, all of which are plesiomorphic within the subfamily. A previous review of Cheiroseius by Karg (1981) included recognition of two subgenera; however, it presented both nomenclatural difficulties and systematic inconsistencies, with attributes used to distinguish each subgenus not holding for its constituent species (Lindquist 2003). Therefore, Karg's subgeneric concepts were not accepted, and the genus Cheiroseius, with its disparity of nearly 120 described species (personal records, Erika Britto and Gilberto de Moraes, December 2015), still remains in need of revision on a cladistic basis.
The new genus Calyptoseius shares with Cheiroseiulus the following attributes: the opisthogaster of both the female and male has an anal shield (rather than an expansive ventrianal shield as in other platyseiines), which may be an ontogenetic retention of the deutonymphal form of the anal shield; the paranal setae are inserted at mid-level of the anal opening (rather than more posteriorly); the female has sternal setae st4 (when present) and lyrifissures iv3 inserted freely on soft cuticle (rather than on metasternal platelets); and tarsi II to IV have the paired pulvillar lobes blunt (rather than more broadly rounded). However, in other respects, particularly differences in dorsal idiosomatic chaetotaxy, leg chaetotaxy, and different form of certain setae of tarsi II to IV (as presented in the following key to genera), a sister-group relationship between Cheiroseiulus and Calyptoseius is highly problematical. Also, mites of these taxa were collected from such completely different habitats as to leave it difficult to link the two genera ecologically. While a small population of Calyptoseius longinoi was extracted from the gallery system of Azteca ants in a Cecropia tree bole, four females and one male of Cheiroseiulus reniformis were collected from two water pools on epiphytes in Jamaica, and a single female of the undescribed form of Cheiroseiulus was extracted from an unspecified substrate in Crystal Cave, Pierce County, Wisconsin, USA (although this taxon was not included in the published survey of invertebrate cave faunas by Peck & Christiansen, 1990, they noted mites of a variety of oribatid taxa extracted from litter from the same cave, with same collection date). A basal link between Cheiroseiulus and Platyseius on one hand, and a separate link between Calyptoseius and some element or subset of Cheiroseius on the other, may be more probable, and a hypothesis worth testing, eventually with the help of molecular analyses.

Revised description of subfamily Platyseiinae (apomorphic attributes italicized)
Gnathosoma -Gnathotectum triramous, each branch usually denticulate, median projection usually not greatly elongated. Fixed digit of chelicerae with setiform pilus dentilis, with few to many small to fine teeth restricted to a ridge on apical one-third of masticatory surface, and with deep subapical receptacle for receiving apex of movable chela; fixed digit without hyaline rim near base on paraxial surface. Movable digit of adult female and nymphs usually bidentate, occasionally with one or three teeth, without ventral mucro; arthrodial envelope at base of movable chela small, not fringed. Movable digit of male unidentate, with spermatodactyl freely extending anteriorly or anteroventrally, its form highly differing between species. Infracapitulum with usually seven transverse rows of deutosternal denticles, each row wide, finely denticulate, sixth row usually slightly widest; lateral lines connecting rows arcuate or weakly formed or obliterated; anterior pair of hypostomatic setae hp1 elongated, whiplike or flattened strap-like along part of length; corniculi simple, widely spaced basally, never forked or convergent apically. Palptrochanter with inner seta elongated, similar in form to hp1; palptarsal apotele two-tined.
Idiosomatic dorsum -Dorsal shield entire in adult (opisthonotal region rarely with incisions between setae S2 and S3), with lateral incisions in deutonymph; podonotal region usually with 19 to 22 pairs of setae, but exceptionally, vertical setae j1 suppressed to alveolar vestiges in adults though well developed in immature instars (Platyseius italicus species-group), or setae z1, z3 and either or both of s1, s2 lacking (Cheiroseiulus); opisthonotal region usually with 12 to 15 pairs of setae in the J-, Z-Sseries, occasionally lacking one to all of J1-J3 (Platyseius spp.), rarely more hypotrichous with 5 to 9 pairs, lacking J1-J4 and several of Zand S-series (Cheiroseiulus). Of marginal series, r2-r5 usually inserted on dorsal shield; r6 and R-series on soft cuticle in adult females but variably on margin of dorsal shield in adult males. Submarginal URsetae usually absent.
Idiosomatic venter -Female: sternal shield with three pairs of setae and two pairs of lyri-fissures, fully consolidated with endopodal strips alongside coxae I-III, and with apices of strong endopodal extensions bearing a gland pore gvb between coxae I and II; third pair of lyrifissures associated with fourth pair of sternal setae usually on small metasternal plates, or rarely inserted in soft cuticle, and with st4 present or absent; epigynal shield with lateral margins usually only slightly widened posteriorly, with posterior margin gently convex or truncate, with sternal setae st5 inserted on shield margins but paragenital poroids on soft cuticle. Male: sternitigenital shield fully consolidated with endopodal strips alongside coxae I-IV, the endopodal extensions between coxae I-II and II-III as in female; setae st4 present or reduced or absent (sometimes ontogenetically suppressed); sternitigenital shield abutting but not coalescing with ventrianal shield. Female and male usually with well-developed ventrianal shield or rarely anal shield present in both sexes (Cheiroseiulus, Calyptoseius); insertions of paranal setae usually even with or behind posterior margin of anal opening, exceptionally at midlevel of opening (Cheiroseiulus, Calyptoseius); postanal seta usually shorter than paranal setae (subequally as long in Calyptoseius). Peritrematal shield broadly united anteriorly to dorsal shield, well united posteriorly to exopodal strip beside coxa IV, free or confluent with anterolateral margins of ventrianal shield in male; peritremes fully developed, extending anteriorly to vertex, and with or without poststigmatic projection; exopodal shield developed as continuous strip beside coxae II-IV, often confluent along most of its extent with peritrematal shield. Metapodal plates usually present as a small pair, free or sometimes appended to (but not integrated with) peritrematal shields on females, integrated in ventrianal shield on males, but free on males of Calyptoseius and exceptionally absent in both sexes of Cheiroseiulus.
Legs -Tarsus I with a usually conspicuously clavate-tipped seta s (as in Fig. 5A), and usually with ambulacrum and paired claws; rarely, ambulacrum and claws absent, or claws sessile. Tarsi of legs II to IV with well developed ambulacrum with paired claws and lobate pulvillus; median lobe of pulvillus slender, acuminate apically (exceptionally, apically blunt in Cheiroseiulus); paradactyli flanking claws of tarsi II-IV elongated, often more so than median lobe of pulvillus, acuminate apically (but apically blunt in Cheiroseiulus). Tarsi II to IV with apical dorsal setae (d-1) conspicuous, longer than pretarsus, often reaching distally to same area as apices of paradactyli; tarsi II-IV each with anterodorsal seta ad-2 elongated and flattened, strap-like, along much of its length, tapering and curved only near apex; tarsi II and III with a second seta (pd-2) similarly elongated and flattened; femora of legs I and II maximally with 12 and 11 setae, respectively, but usually with 11 and 10, each lacking a third ventral seta, pv-2; genua of legs II and III with 11 and 9 setae, respectively, but usually with 10 and 9, each lacking seta pv-1. Leg II of male never crassate or armed with spurs, and with minimal dimorphism in form of setae.