Meruidae, Spangler & Steiner, 2005

Spangler, Paul J. & Steiner, Warren E., 2005, A new aquatic beetle family, Meruidae, from Venezuela (Coleoptera: Adephaga), Systematic Entomology 30 (3), pp. 339-357 : 351-355

Meruidae fam. nov.

Systematic placement

The combination of skeletal characters, habits and minute size in M. phyllisae has offered a perplexing challenge for its classification. With the fusion of the hind coxae and metaventrite, the division of the first (of six) visible abdominal sternite (ventrite II divided by metacoxae) and fusion of visible sternites 1–3 (ventrites II–IV), the large propleuron and visible notopleural and pleurosternal sutures, the broad, lobed mentum, and glabrous palpiform two-segmented galea, suggest placement of Meru in the suborder Adephaga , according to the adult synapomorphies summarized by recent phylogenetic studies ( Beutel, 1995, 1997; Beutel & Haas, 2000). Glabrous antennomeres place it among the Hydradephaga, an aquatic adephagan beetle assemblage for which monophyly is still open to question ( Beutel, 1997; Ribera et al., 2002a,b). For reasons discussed below, however, it has not been possible to place the genus in any known family taxon. We propose that it be placed in a new family, Meruidae , the ‘comb-clawed cascade beetles’, based on M. phyllisae gen. et sp.n.

Diagnosis

Meruidae can readily be identified by the following combination of characters:

1. Very small adephagan species.

2. Habitus resembling a minute haliplid.

3. Rough body sculpture with flat, wrinkled setae (possibly respiratory structures).

4. Pectinate tarsal claws.

5. Trident tibial spurs (inner metatibial spur serrate).

6. Pair of overlapping spatulate setae on the labrum.

7. Lack of swimming hairs on the legs.

8. Small, rounded scapus.

9. Alternating large and small segments of the antennal flagellum.

In the most recent key to adephagan families ( Balke et al., 2003), M. phyllisae runs to couplet 11, Dytiscidae , having obsolete metacoxal plates, but in Meru , the anterior extension of the metacoxae is obscured by complete fusion. Insertion of a couplet ‘11a’ at this point would separate Meruidae :

11a. Beetles less than 1 mm long. Tarsal claws pectinate ................................................................ Meruidae – Beetles larger than 1 mm long. Tarsal claws simple......11

Being less than 1 mm in length, M. phyllisae is among the smallest known members of Adephaga . Some Notomicrus (Noteridae) species are barely longer than 1 mm, as are the smallest Carabidae , e.g. members of Geocharidius Jeannel.

Characters of Hydradephaga coded for Meruidae

In their placement of the newly proposed family Aspidytidae, Ribera et al. (2002b) listed forty adult morphological characters used in analyses to hypothesize relationships among the family taxa of Hydradephaga. Although a full cladistic analysis to determine the sister group of Meruidae is beyond the scope of this paper (but will be the subject of ongoing and future studies), coding these characters at this point will assist in supporting the exclusion of Meru from other adephagan families, and point out some character systems that need additional scrutiny among many taxa. We were able to score most of these characters for Meruidae using character states defined by Ribera et al. (2002b), as follows, with (0) being plesiomorphic:

Character 1, body shape: with distinct pronoto-elytral angle (0); 2, head shape: not shortened and laterally rounded, eyes protruding (0); 3, compound eyes: undivided (0); 4, shape of scapus: strongly shortened but without enlarged globular part (3); 5, pedicellus: not enlarged, not strongly shortened or enclosed by scapus (0); 6, flagellomeres of males: antennomeres 5 and more than one of the following segments distinctly broadened (2); 7, galea: two-segmented (0); 8, elongate sensorial field of distal labial palpomere: absent (0); 9, sensorial field on a protuberance of the dorsal side of the distal labial palpomere: absent (0); 10, shape of the prosternal process: strongly broadened and apically truncate (3); 11, ventral procoxal joint: with distinct coxal condyle (0); 12, profemoral cleaning device: absent (0); 13, protibial burrowing spur: absent (0); 14, row of flattened thorns on apical part of protibia: absent (0); 15, outer edge of protibia: not rounded (0); 16, tibial groove or concavity for reception of protibial burrowing spur: absent (0); 17, curved spurs on ventral side of protarsomeres 1–3: absent (0); 18, prothoracic defence gland: (0, probably; needs further study); 19, mesoventrite: short, with hexagonal groove and anterolateral grooves for reception of procoxae (0); 20, mesocoxae: globular (1); 21, mesocoxal cavity: laterally bordered by mesepimeron and metathoracic anepisternum (1, but difficult to interpret with certainty due to extensive fusion); 22, proximal pro- and mesotarsomeres: not elongated and broadened (0); 23, middle and hind legs: elongate (0); 24,

swimming hairs on meso- and metathoracic legs: absent (0); 25, noterid platform of metaventrite: absent (0); 26, transverse ridge of metaventrite: absent (2); 27, contact between pro- and metasternal process: present (1); 28, metafurca: originates from intercoxal septum (1); 29, size of metafurca: narrow, with reduced lateral arms (1); 30, mesal walls of metacoxae: with extensive contact area and intercoxal septum (3); 31, anterior margin of metacoxa: undetermined due to extensive fusion; 32, metacoxal plates: largely reduced (4); 33, lateral margin of metacoxal plates: indistinct anteriorly (4); 34–35, muscle characters (not studied); 36, abdominal segments III and IV: completely fused (1); 37, bulges on anterior abdominal sternites: absent (0); 38, gonocoxosterna VIII: not exposed (0); 39–40, ramen and laterotergite: (unsclerotized; not observed).

Exclusion from other hydradephagan families

Although Meru keys to Dytiscidae , it must be excluded from that family because of the undifferentiated legs, lack of swimming fringes, and alternating leg movement during swimming ( Ribera et al., 2002b). The oarlike hind legs and often highly streamlined body of ‘diving beetles’ separate Meruidae and other aquatic taxa that are less equipped for rapid and evasive swimming. Roughley & Larson (2001) reviewed the characters, classification and literature on the Dytiscidae ; Miller (2001) provided the most recent analysis of morphology and phylogeny.

Apomorphic features of Gyrinidae ( Steiner & Anderson, 1981; Beutel & Roughley, 1988; Beutel, 1990, 1995; Roughley, 2001a) include divided eyes (for surface swimming), highly modified legs for swimming, stout and highly specialized antennae, and a single protibial spur. Meru shares none of these features but has synapomorphies with the remaining families of Adephaga , e.g. the short mesosternite, form of the male genitalia (curved asymmetric base, torsion) and other characters that would exclude it from Gyrinidae ( Beutel, 1995, 1997).

Meru phyllisae is the ‘undescribed adephagan from South America’ mentioned by Lawrence & Newton (1995) as having haliplidlike features. The general body form is reminiscent of Haliplidae , but the very small size, surface sculpture and setae of the body, the pectinate claws, the lack of expanded coxal plates, the legs lacking swimming hairs, the differing shape of the hind femur ( Kavanaugh, 1986) and trochanter, and the lack of a specialized ovipositor, readily separate Meruidae from Haliplidae . Most of these characters are autapomorphies for one or the other taxon; the greatly expanded hind coxal plates are unique to Haliplidae , not known from any other extant group of beetles ( Beutel & Ruhnau, 1990; Beutel, 1995; Roughley, 2001b).

In Noteridae , the smooth, streamlined body form and modified legs are highly specialized (except in Phreatodytes Ueńo ) for swimming and burrowing. Other derived features of the clade ( Beutel & Roughley, 1987; Beutel, 1995, 1997; Roughley, 2001c) include the form of the scape, with a short globular base and more slender distal part, dilation of antennomeres V, VII and IX, presence of a profemoral excavation and antenna cleaning organ, metacoxal platform and anterior paramedian angles. Meru shares the characters of the antennal flagellum only, but does share a derived feature with Notomicrus , considered to be sister to most remaining noterids: complete fusion of the metacoxae with the metasternum ( Beutel & Roughley, 1987). The absence of a sutural stria is also shared between Noteridae and Meru .

Although the body form of Meru is somewhat similar to that of an Amphizoa , the latter are very much larger, and the apomorphic features of Amphizoidae (absence of mental suture, one-segmented galea, and strongly reduced metafurca) preclude any close relationship. In Amphizoidae , legs are relatively unmodified for swimming and natatory abilities are poor ( Kavanaugh, 1986; Beutel, 1995; Philips & Weiping Xie, 2001) but vestigial swimming hairs are present. Hygrobiidae are also much larger, have long fringes of swimming hairs on the legs, and several unique autapomorphies as listed by Beutel (1995, 1997) that are not found in Meru . Members of the recently described family Aspidytidae ( Ribera et al., 2002b; Balke et al., 2003), also lacking swimming hairs (considered to be derived through loss) on the legs, have a smooth, noteridlike body form and unusual scapus and pedicellus very different from those of Meru . The carabidlike terrestrial Trachypachidae ( Ball, 2001) have few similarities to Meruidae except for the groundplan adephagan features. Meru lacks the protibial antennal cleaner of Trachypachidae . Relationships with this group and aquatic Adephaga have been the subject of many studies ( Bell, 1966; Roughley, 1981; Beutel & Belkaceme, 1986; Shull et al., 2001). Lastly, the characters of known fossil taxa, as discussed by Balke et al. (2003), also rule out inclusion of Meru in any of these families.

Comments and conclusions

The overall body form of Meru , with rough surfaces, protruding eyes, and unmodified legs, is typical of aquatic beetles that are weaker swimmers of lentic water, or shallow marginal stream habitats, e.g. Haliplidae and Amphizoidae , and several nonadephagan groups that have acquired aquatic habits. For aquatic Adephaga , these characters are considered to be primitive, relative to the streamlined body and specialized swimming legs and methods of Dytiscidae , Noteridae , and Gyrinidae ( Ribera et al., 2002b) . Meru has features of a beetle that has perhaps recently invaded the aquatic realm and/or has retained these primitive states, and this is in keeping with the postulated habitats of the earliest Adephaga (Beutel, 1995,1997) at the margins of ponds or rivers, from which multiple invasions of aquatic and terrestrial to arboreal habitats may have occurred.

The observed wing dimorphism, with atrophied wings in the majority of specimens, is not a common condition in aquatic beetles ( Spangler, 1979) and also not typical of obligate streamside lowland species of Carabidae ( Darlington, 1936, 1943). Maintenance of both wing forms in a population is probably advantageous for a small aquatic beetle in a habitat subject to fluctuation or sudden changes, such as flooding or drying of the stream margin habitat.

The unique apomorphic features of Meruidae are an enigmatic assemblage: the pectinate tarsal claws, the possibly respiratory structures covering the body, the odd leaflike setae of the labrum, and the trifid tibial spurs. The very small size of these beetles, also considered to be derived ( Beutel & Haas, 1998) may be a contributing factor to the assemblage of other unusual characters of this taxon. Miniaturization in Coleoptera results in alteration of other characters via fusion, reduction and loss. In Meru , fusion of thoracic sclerites, loss of the oblongum cell and binding patch in the hind wing, reduced and membranous ovipositor, and probable loss of pygidial and prothracic defence glands, etc., can probably be attributed to miniaturization; similar conclusions have been drawn for other small-sized members of other beetle taxa, e.g. Spanglerogyrus in the Gyrinidae ( Steiner & Anderson, 1981) .

The folding pattern of the fully developed wing of M. phyllisae ( Fig. 12 ) is of the adephagan type ( Kavanaugh, 1986; Kukalova´-Peck & Lawrence, 1993; Lawrence & Britton, 1994; Beutel & Haas, 2000), although with the reduced venation and marginal fringe, the wing is more similar to that seen in members of Myxophaga. Similarities in the wing to that of myxophagan beetles, all of which are relatively small, may be the result of convergence, but a sister-group relationship between Myxophaga and Adephaga has been suggested by wing characters ( Kukalová-Peck & Lawrence, 1993).

Serrate tibial spurs are found in most Noteridae , especially the larger (inner) metatibial spur, which is also serrate in M. phyllisae . Some Haliplidae also bear serrations on this spur. Whether this represents any synapomorphy among these families should be a subject of future study.

The leaflike setae on the labrum, which oppose each other and overlap above a median emargination, along with the toothlike setae medial to these, may be related to a feeding specialization. The diet of these beetles is uncertain; observations on captive M. phyllisae suggest that they may be eating decaying leaf tissue or perhaps more likely, algal filaments and/or fungal hyphae growing on submerged leaves. The fact that captive beetles survived for nearly 200 days on decaying leaves, without any predatory or cannibalistic behaviour being observed during that time, supports this idea. Ancestral Adephaga were probably predators ( Beutel, 1995) as are most extant members, but predatory habits in Meru are doubtful, as it is not an agile or rapid swimmer. Although the head is prognathous, the mandibles are rather stout, less suitable for predation, but the terebral ridges and teeth in Meru mandibles would serve for cutting ( Acorn & Ball, 1991). Haliplid adults have stout mandibles and are known to be more omnivorous ( Seeger, 1971), whereas the larvae are algal specialists. Meru phyllisae could also be categorized as omnivorous, and if also an occasional predator, this would probably involve only small, perhaps sessile, prey species. Further comparisons of mandibular morphology, with scanning electron microscopy of the surfaces, will probably provide more information.

The labrum in adult haliplids ( Beutel & Ruhnau, 1990) is also distinctly emarginate and with a fringe of flattened, blunt to tapered setae directed anteromedially. There are also similarities in the labrum of some Myxophaga, which are also small animals that feed on filaments of algae, e.g. the larva of Hydroscapha ( Beutel & Haas, 1998: fig. 2A) and adult Torridincolidae ( Spangler, 1980: figs 4, 5).

Antennal characters of Meru are intriguing and show similarities to both Haliplidae and Noteridae . The globular, short scape has been considered a derived feature of haliplids, and the condition in M. phyllisae seems to be even more reduced (derived) in that there is no separate distal portion; in haliplids, there is a slight constriction between basal and distal halves. The pseudo-two-segmented state in Noteridae and Aspidytidae appears to be intermediate in form and reduction between the scape in Haliplidae and that of other Adephaga , which have the distal portion generally longer than wide.

Modification of antennomeres V, VII and IX has been considered unique to Noteridae ( Beutel, 1997) . Increased size and sensory fields of the same alternating antennomeres among the two taxa would have a very remote chance of being a result of convergence, so this appears to be a synapomorphy with Meruidae . Perhaps the miniaturization (possibly an adaptation to the stream margin habitat) of M. phyllisae has resulted in the loss of sensoria on antennomeres and reduction of the number to single organs on antennomeres VII and IX. Some Haliplidae possess vestiges of this pattern, but this character system needs more comparative studies among many taxa in order to draw conclusions on possible relationships.

The two brush-tipped setae of the anterior margin of the prosternum deserve mention, as we find similar setae in a number of Noteridae , greater in number, but with the same general form. In Notomicrus traili Sharp , twelve such brush-tipped setae occur in a row along the margin; fewer are seen in the related Speonoterus ( Spangler, 1996: fig. 58). Other larger noterids, e.g. Hydrocanthus iricolor (Say) generally have a fringe of these setae. Their function is unknown, but the distinctive form and placement are similar in Meruidae and Noteridae . Setae that line the prosternal margin in the haliplids examined appear to be of the simple, tapered form.

The hidden scutellum is a feature common to all known Haliplidae and Noteridae except Phreatodytes ( Beutel, 1997) , Speonoterus Spangler (1996) , and Notomicrus species ( Spangler, 1996); the latter have an exposed, but very small, scutellum, but according to Beutel (1997), concealed. Several unrelated tribes of smaller-sized Dytiscidae also have the scutellum concealed; the character may have arisen independently in some cases and may be another feature of convergence, related to the reduction in body size.

The form of the proventriculus, with its eight lobes of alternating large and small size, is typical for all Adephaga ( Balfour-Browne, 1944; Smrzˇ, 1982) but the backward pointing fringe of hairs on the smaller lobes is considered to be derived. If a proventriculus with lobes bearing sclerotized teeth is primitive compared with the purely hair filter type ( Balfour-Browne, 1944), Meru shares the derived state with Gyrinidae , Noteridae , Haliplidae , and some hydroporine Dytiscidae .

The simple, membranous ovipositor with reduced gonocoxae in M. phyllisae indicates that these beetles are probably depositing eggs on surfaces rather than inserting them in plant tissue, as is the case for some Haliplidae ( Beutel & Ruhnau, 1990; Roughley, 2001b) and Noteridae ( Beutel, 1995) . This appears to be the result of reductions of the sclerotized gonocoxae and setose appendages seen in most hydradephagan taxa.

We believe that the dytiscoid type of metacoxal fusion and metafurca, along with the other synapomorphies with Noteridae , suggest a sister-group relationship between Meruidae þ Noteridae þ the remaining Dytiscoidea. The globular antennal scape in Meruidae and Haliplidae , along with the nonpredatory feeding and associated features, could also indicate some relationship among these two taxa. A working hypothesis of a relationship among Haliplidae , Noteridae and Meruidae needs consideration. The position of Haliplidae among the other families of Adephaga has long been controversial and we hope that the discovery of Meruidae will lead to clarification of this. The results of molecular analyses in progress, using fresh material of M. phyllisae (M. Balke, pers. comm.), are eagerly awaited.

Early aquatic Adephaga were probably nonswimming forms that lived at the edges of rivers or ponds ( Beutel, 1995). Hygropetric habitats have existed as ‘edges’ for millennia, as the exposed bedrock has never been under forest canopy, perpetuating the sunlit microhabitats for surface algal growth and the assemblage of often relictual insects associated with them. The recently discovered ‘cliff water beetles’, Aspidytidae , demonstrate other unusual taxa to be found at these sites ( Ribera et al., 2002b; Balke et al., 2003). If there was a common ancestor to the modern aquatic Adephaga (excluding Gyrinidae ), it probably had the general appearance of Meru , lacking modifications for swimming, but was probably larger, perhaps resembling an Amphizoa . From this form radiated the diverse and independent adaptations for different swimming, feeding and oviposition strategies. Future studies on additional characters of M. phyllisae (parsimony analysis including musculature, presence of defence glands via thin sectioning, molecular data, etc.) and the discovery of the elusive larval stages will further clarify these notions.

Figures

Fig. 17. Biotope at El Tobogán, the type locality of Meru phyllisae. A, The upper cascade looking upstream from the dam; B, the upper cascade looking downstream from the upper edge. C, The edge of the upper cascade showing the wet rock face and stream edge; D, the stream edge at the upper cascade where the root mat habitat was sampled; E, Biotope at El Tobogán, the type locality of Meru phyllisae. A, The dam and pool below the upper cascade looking upstream; F, the pool and forested stream below the cascade looking downstream; G, Section of forested stream below the cascade, looking downstream; H, the open rock surface and edge of a rapid section of the cascade.

Fig. 12. Meru phyllisae, hind wing, macropterous condition, in folded position.