Bonn. zool. Beitr. Bd. 50 H. 1-2 S. 89-101 Bonn, Dezember2001 Enchytraeidae as prey of Dolichopodidae, recent and in Baltic amber (Oligochaeta; Diptera) Hans Ulrich & Rüdiger M. Schmelz Abstract. ApieceofBalticamberwithtwoinclusions,anenchytraeidwormfragmentand a dolichopodid fly, is described. The morphological account ofthe enchytraeid fragment includes details on number, shape and distribution ofthe chaetae. It is the first detailed description ofan oligochaete in Baltic amber. Field observations ofrecent Dolichopodidae preyingonEnchytraeidaearereported.Theavailableevidencesuggeststhatthefossilworm fragment was carried into the resin by the fly as its prey. Thus, this piece provides a fossil record ofpredation by Dolichopodidae on Enchytraeidae, and points to a plausible explanationastohowtheoligochaetewormsfoundasrareinclusionsinBalticambercould havebeentrappedintheresin. Key words. Baltic amber, Enchytraeidae, Dolichopodidae,predation. Introduction The present paper reports on a piece of Baltic amber with two inclusions, a dolichopodid fly and an enchytraeid worm, and on recent observations by which the common fate ofthe two animals may be explained. Recent Dolichopodidae were observed as adults preying on Enchytraeidae. As will be shown in the relevant paragraphs, the piece ofamber presents several traits which may be taken for evidence that the worm was carried into the resin by the fly as its prey. The authors of this joint paper are specialists in the families involved, the Enchytraeidae (R.M.S.) and Dolichopodidae (H.U.), respectively. Each studied the representatives ofhis group among the available material, both recent and fossil, and will discuss his observations. Thus, each is responsible forthe section concernedwith the worm and the fly, respectively. A preliminary report of our observations was presented at the 4th International Congress ofDipterology in Oxford, U.K., September 1998 (Ulrich & Schmelz 1998). Two of our photos taken from the amber piece were published in colour, with a concise explanatory text based on our interpretation, in the amber book by Weitschat & Wichard, 1998, pp. 54-55. [N.B. The author of the dolichopodid species is not Meigen but Meunier.] The amber piece The specimen (Fig. 1) was offered to the senior author and purchased by him for an amber collection of Diptera to be built up and deposited at his home institution (ZFMK). According to information from the dealer, it originated from the Eastern Baltic deposits in Kaliningrad District. 90 H. Ulrich & R. M. Schmelz The amber piece was flattened and polished to facilitate the study ofboth objects, the worm and the tly. The embedding of the inclusions is rather good. Within the piece, a curved plane roughly parallel to its upper surface marks the joint or cleft between two layers formed by successive resin flows. In the figures it is seen from its convex side, and is discernible by a delimited area around the worm (Fig. 4), reflexions, and by a stellate hair fibre (the contorted fibre left ofthe worm) and some smaller particles lying in it. The worm is lying fiat in this plane, whereas the fiy is lying on its left side and slightly oblique, with its legs partly in and its trunk and wings above the convex surface. Apparently this was the surface of the resin on which the animals were trapped. Although sticky, however, the resin must have been rather viscous, for neither ofthe two animals sank in: the worm remained lying fiat on its surface, while most ofthe fly remained outside before both were completely embedded by a sub-sequent flow. The left side of the fly's trunk, i.e. the reverse when viewed as in Fig. 1, is covered by white emulsion. Apparently this was the off-light side of the inclusion facing the substrate covered by the resin (Schlüter & Kühne 1975), just as the concave side of the curved plane mentioned above faced the substrate. So, both structures point to the same direction and confirm that Fig. 1 is a view on the outer surface ofthe resin mass as it was when the animals were trapped. The position ofboth inclusions on the same plane proves that both were trapped at the same time or, at least, in the short time between two resin fiows. Fig. 1. The dolichopodid fly, Gheynia bifurcata ?, and the enchytraeid worni (right) lying in the amberpiece. . Enchytraeidae as prey ofDolichopodidae 91 Authentic or forged? A forgery can be excluded for various reasons: - The piece contains stellate hairs and individual fibres thereof, which proves that it consists ofamber and not ofany artificial mounting medium. - It is rather clear, without any disturbances around the inclusions which could be explained as traces ofan artificial implantation. - The fly is covered on its lower side by a white emulsion produced by exudation from its body, as frequently found on amber inclusions. - It belongs to a genus, Gheynia Meunier, which is unknown from the World's recent fauna but rather common in Baltic amber. - Since predation on worms by Dolichopodidae is not widely known, it is hard to imagine that a forgercould have known about it and made an artificial mount ofpre- dator and prey. The dealer was not aware either ofthe biological context when he offered the specimen, and although it was explained to him, he still offered it at a moderate price which would have been too cheap for a perfect forgery. -Lastbutnot least, the followingparagraphs will showthat all evidence fitstogether to allow a reconstruction ofwhat happened in the last hours ofthe animals' lives so well, that it can hardly be a product offorgery. The fly (H. Ulrich) Recent observations in the field Dolichopodidae with Enchytraeidae as prey were observed repeatedly since 1968 in woods in southern and western Germany (Mumauer Moos, Upper Bavaria; Härts- feld, Swabian Jura; Eifel). Some ofthe flies were photographed, some were taken together with their prey and preserved in ethanol. The observations include the fol- lowing species and sexes (classification after Ulrich 1981): Rhaphiinae: Rhaphium crassipes (Meigen) ?, Argyra auricoUis (Meigen) ?, and Argyra sp. indet. ?; Dolichopodinae: Dolichopus atripes Meigen ?, D. nigricornis Meigen sensu Parent [=discifer Stannius] ? and cT, D.popularis Wiedemann ? and cT, D. ungulatus (Lin- naeus) ?, Gymnoptermis aerosiis (Fallén) ?, Hypophyllus crinipes (Staeger) ?, and H. obscurellus (Fallén) ? Most ofthe flies were seen perching on leaves oflow herbage with a worm held by the proboscis (Figs 2-3), but some were observed on the soil searching for prey and catching worms. Preying on Enchytraeidae appears to be a common habit of adultDolichopodidae, whichcanbe observedwhereverthe flies are abundantandthe worms are accessible to them. Thenormal procedure is as follows: A tlyroves about on the soil in search ofprey, walking with some quick and short steps and stopping, walking again, sometimes leaping a few centimetres or flying to another place. It inspects small hollows and cracks in the soil, finds at last an accessible worm in its tube, stoops down and apparently seizes the worm with its proboscis. The Dolichopodidae are known to use their labella for grasping prey and for holding it firmly while sucking it out. This was first described and figured for Medetera Fischer and a psychodid fly as prey, by Fischer (1819) who, however, misunderstoodhis observations as swallowing. This errorwas correctedby Macquart 92 H. Ulrich & R. M. Schmelz Fig. 2. Dolichopus iingulatus ? with an enchytraeid. The worm has been grasped at some distance from its end which is folded back. Swabian Jura, Härtsfeld, 10 July 1970. (1828a: 15-16, 1828b: 227-228). The mouthparts of Dolichopodidae have been studied repeatedly. A classical account oftheir structure and function was published by Snodgrass (1922), extensive comparative descriptions by Cregan (1941) and Sato (1991). Having seized the worm, the fly tries to pull it out from its tube. This, ifsuccess- ful, is normally done in two steps: first the worm is pulled forth forpart ofits length, subsequently a second attempt is made to pull it out as a whole. Ifthe fly succeeds and the worm is exposed for all its length, a heavy struggle follows: the worm squirms vigorously, beats around and has a good chance to be released by the fly and not to be attacked again. The larger the worm is in proportion to the fly, the better is its chance to get free. During the fly's second attempt it may happen that the worm is torn in two pieces and the fly gets a fragment only to feed on. Ifthe fly keeps hold ofthe worm, it will normally take offand fly to a leafto feed on its prey there. The worm is hanging down from the fly's proboscis, held by the labella. It squirms as long as it can, while it is rhythmically moved up and down by the fly, apparently by alternate retraction and protraction of the labium. By these movements the end ofthe prey held between the labella is probably pushed against the epipharyngeal armature which serves to lacerate its body wall (see Snodgrass 1922). In all cases observed by me, the flies did not use their legs to support the labella in seizing orholding the prey. However, a female ofArgyra auricoUis did use a fore leg : Enchytraeidae as prey ofDolichopodidae 93 Fig. 3. Argyra auricollis ? with an enchytraeid. The fly is warding offthe squirming worm with its left fore leg. Note the soil particles adhering to the worm. Upper Bavaria, Mumauer Moos, 23 July 1968. tokeep a squirmingworm offits head(Fig. 3). SimilarobservationsbyCregan (1941 14) on Dolichopus ramifer Loew may be interpreted so. In other cases {Hypophyllus obscurellus) it seemedthatthe flytriedtopress the worm againstthe soil with its legs during the struggle after capture. In contrast, Hydrophorus Fallén has been observed holding insect prey with the fore legs (Aldrich 1911: 45, Williams 1939: 307, Peterson 1960: 270), which are used both for predation and, by the male in a pair, for holding the female (Dyte 1988; Lunau, unpublished photo; inferred from sexual dimorphism by Aldrich 1911: 48). Use ofthe fore legs as predatory legs has also been reported for other hydrophorine genera, viz., Scellus Loew, Hydatostega Philippi, Thinophilus Wahlberg, and Aphrosylus Haliday (Doane 1907, Harmston 1948, Roubaud 1903). Most ofthe flies met with a worm were females. This conforms to the common experience that female flies are more voracious as predators than males, and can be explained by the need ofprotein food for production of eggs. However, males did feed on Enchytraeidae too, as observed in Dolichopus nigricornis andD. popularis. While searchingforEnchytraeidae, some ofthe flies met insectswhichwouldhave been a suitablepreytoo, butapparently didnottake notice ofthem. Itappears thatthe worms were more attractive. Nevertheless, the published record of Dolichopodidae preying on oligochaete worms is scanty, whereas preying on insects has been recor- ded frequently. Wahlberg (1845, quoted in English by Lundbeck 1912: 15) described predation on Naididae by Dolichopodidae of various subfamilies on a muddy 94 11. Ulrich & R. M. Schmelz seashore. Aldrich (1922) mentions oligochaete worms besides dipterous larvae as prey ofthe adults, Cregan (1941) observed Dolichopus Latreille extracting annelids from the soil, Lunau (1993) mentions Enchytraeidae and Tubitlcidae among the prey of Poecilohothriis Mik. Observations by Negrobov and Pogonin on oligochaete worms, apparently Enchytraeidae, as predominant food of Dolichopodidae are reported and demonstrated by photos in an unpublished thesis (Pogonin 1984). (I wish to thank A. Stark who called my attention to this account, and O.P. Negrobov who sent me copies ofthe relevant paragraph.) The fossil fly mm The tly is a female and was identified Gheynia bifurcata Meunier. It is 2 long. Gheynia can be recognized by the shape ofthe third antennal joint (postpedicel, Stuckenberg 1999), with projections above and below the insertion of the arista (stylus) which are longer in the male, short but still discernible in the female (see figures in Meunier 1908a: 8 and 1908b: 57). A similar shape is found in some recent species ofChiysotus Meigen, particularly so in C. furcatus Robinson and C. hilbur- ni Woodley (see Robinson 1964 and Woodley 1996). Meunier(1907-1908) regarded Gheynia (wrongly emended to Gheynius) as an aberrant representative ofhis genus Palaeochrysotus [now recognized as ajunior synonym oíPaleomedeterus Meunier]. He placed both close to Chiysotus and included them in a common key to species (1907: 199, 209-210; 1908a:^ 8-9; 1908b: 9, 13-15, 56-59). However, the specimen ofGheynia discussed here, as well as other specimens examined from Baltic amber, have an anterior pre-apical bristle on each mid and hind femur as in Sympycninae, whereas the recent Chiysotus species cited above agree, according to their de- scriptions, with the ground pattern ofChiysotus and the Diaphorinae in lacking this bristle and having instead an anteroventral row ofbristles before the end ofthe hind femur. So, Gheynia cannot be a close relative of Chiysotus and even less of C C. furcatus and hilburni, and the peculiar antennal shape must have evolved independently. It may be closer to the recent sympycnine Scotiomyia Meuffels & Grootaert, whose antennae resemble those ofthe female (see Meuffels & Grootaert 1997). Seiivanova & Negrobov (1997) placed Gheynia in the Peloropeodinae. As stated inthe description ofthe amberpiece, the fly is lying on its left side above the surface on which it was trapped, and its left (lower) side is completely covered with white emulsion. In addition, there are small globularmasses ofemulsion on the right (upper) side above both thoracic spiracles and a thin layer on the right pleural membrane and around the tip ofthe abdomen. Evidently the thoracic and abdominal spiracles and the anus were the openings through which fluids oozed from the body. The fly is lying at some distance fromthe womi with its legs pointingto the worm. The right hind leg is broken between the metatarsus and the second tarsomere, and the distal fragment is separated from the proximal part by a distance equalling the length ofthe tibia. All these observations can be explained by the following scenario: The flywith its prey, alighting on the resin, touched the surface with its legs, got stuck and dropped theprey which fell onto the sticky surface. The tlythen tilted overto its left side and, pushed by the legs, drifted away from its original position. The right hind leg may have broken when the fly tried to pull it out from the resin. When a tarsus is torn in a living fly or a recently killed one, the unguitractor tendon may adhere to the distal Enchytraeidae as prey ofDolichopodidae 95 fragment and be drawn out from the proximal part to connect both. Apparently this did not happen in the fossil fly, since no trace oftendon could be discerned in this position. It is likely that the tendon would have been preserved in the amber and should be visible ifit was lying exposed. The worm (R. M. Schmelz) Material from recent observations Recent enchytraeid worms preserved in ethanol were identified as far as possible. They had been collected by the senior author at one of the localities of his field observations, in Upper Bavaria, Mumauer Moos, in wood with Impatiens noli- tangere as the dominating element of the herbal layer, 1 and 8 July 1968. The identifications are given below. A sample collected from the upper stratum of the soil near the surface where Dolichopodidae were searching for prey contained species ofHenlea Michaelsen, 1889 (23 specimens), Fridericia Michaelsen, 1889 (3 specimens) and Buchholzia Michaelsen, 1887 (1 specimen). Six worms preserved together with their predators {Dolichopuspopularis 1 3?, Hypophyllus crinipes ?,Argyraaiihcollis ?) w^ere all Henlea sp., including 1 complete adult worm with 40 segments and two fragments with hind end. It is uncertain though unlikely that dolichopodids specifically select Henlea species as prey. The selection goes possibly rather by size and occasion. Many Henlea species live especially close to the soil surface, up to the Of-horizon. Furthermore, the presence of Impatiens noli-tangere indicates a habitat where Henlea sp. can be found in high numbers (Schmelz, pers. obs.). In the collection of enchytraeids from the Mumauer Moos, Henlea sp. is the dominant genus, but the sample is not representative as the worms were eye-picked and not sampled quantitatively. The fossil worm The object is observable from both sides of the amber piece, but the chaetal arrangement is clearly visible only from one side, as viewed in Fig. 4, whereas refractions blur the microscopical details on the other side. The worm is not complete; one end forms a wound (lower end in Fig. 4, left in Fig. 5). The wound is not closed, shreds oftissue project outside. It is surrounded by a small and highly refractile field, consisting ofa mixture oftissue and body fiuid. This field is surrounded by a much larger quasi-circular field lying in the same curved plane as the worm fragment and the fly, which was apparently foiTned by coelomic fluid that had oozed from the wound and spread on the surface of the resin. Its texture is grainy with some scattered granular concentrations, possibly coelomocytes. About halfofthe worm fragment lies within this field, the wounded end in its centre (Fig. 4, lower half), whereas the other halfis outside with its free end rounded and uninjured. As no prostomium and no internal differentiation ofthe terminal seven segments are discernible, the intactend is most likelythe rearendand the fragment the posterior part ofthe worm. 96 H. Ulrich & R. M. Schmelz Fig. 4. The worm fragment lying in the field of putative coelomic fluid. Same view as in Fig. 1, highermagnification in darkfíeld illumination. Intact hind end ofworm above, injured end below. The worm fragment is c. 1400 //m long and 100-200 ¡Lim wide. It consists of31 chaetigerous segments, considerably contracted towards the injured end. There are fourchaetal bundles per segment, two lateral and two ventral ones. The dorsal halfof the body surface is without chaetae. When viewed from the side with the best optical resolution (Fig. 5), the dorsal body surface is visible in the posterior half of the fragment. Only one lateral chaetal bundle per segment is visible here. Due to a successive torsion ofthe object, three chaetal bundles are discernible on the opposite side ofthe fragment, two ventral and one lateral. There are three chaetae per bundle in all observable cases, in fan-shaped arrange- ment. The distal part of each chacta is strongly curved, with a simple, pointed tip (Fig. 5, bottom right).The proximal part inside the body is not observable. The chaetae are fairly stout,estimated lengths and diameters are 40-50 //m and approx. 3 jum respectively. The lateral and ventral chaetae are alike. The chaetae appearto be thicker in lateral view than in top view. Each segment has a strong secondary annulation. There are 5-8 annuli per segment. The annuli on chaetal level are most prominent. Intersegmental furrows are observable in some segments as clearly marked furrows exactly halfway between the chaetae. The internal organization of the worm is indistinguishable. Dark brownish food pellets are visible inside the body overthe entire length ofthe fragment including the posterior end. Enchytraeidae as prey ofDolichopodidae 97 Systematic position A posterior end ofan oligochaete worm that shows not much more than the chaetal pattern cannot be identified with certainty, because usually mature specimens with complete anterior ends are needed to identify the species. This applies to recent specimens and all the more to a c. 40 million year old fragment. However, if a similar recent worm fragment from a terrestrial subtropical habitat was found it would be determined as an enchytraeid. All observable characters except the strong intersegmental annulation - possibly an artefact due to desiccation before complete inclusion - agree with what is only found in Enchytraeidae. Earthworms (e.g. Lum- bricidae, Megascolecidae) are excluded by the presence ofthree chaetae per bundle and by the size ofthe fragment which is very small for an earthworm, even ifit was a hatchling. Other 'microdrile' oligochaete taxa, most ofthem aquatic, are excluded by the following combination of characters which is found only in Enchytraeidae: presence of lateral (instead of dorsal) chaetal bundles, lateral and ventral chaetae alike, three per bundle, chaetae curved with simple-pointed tips. Furthermore, the peculiardistribution pattern ofthe chaetae indicates that the worm belongs to atruly- soil-dwellingratherthan an aquatic species: the lateral (ratherthan dorsal) position of the uppermostchaetal bundles is not found in aquatic microdrile oligochaetes andcan be explained as an adaptationto locomotion interrestrial habitats (Schmelz, inprep.). Fig. 5. Details oftheworm fragment. Sameviewas inFig. 4, butturnedsothatthe intacthind end points to the right. The irregular vertical lines above and below the mid-region ofthe fragmentmarktherefractileperipheryofthe fieldofcoelomic fluid. Bottomright: Distalparts ofthe chaetae ofone bundle, lateral view. 98 H. Ulrich & R. M. Schmelz Among all known recent genera of Enchytraeidae, the specimen fits best in Bitchhohia, the only genus with the character combination of lateral (rather than dorsal) chaetal bundles, morethan two chaetae perbundleandcurved chaetaltips, but the specimen may also belong to an extinct or a yet unknown recent genus. It should also be stressed that the key criterion that decides on the identity ofan enchytraeid, the location ofthe reproductive organs, is unknown here; so the possibility remains that the specimen belongs to a different family, unknown and extinct, of terrestrial oligochaetes. Since Menge's description of Enchytraeus sepultus from Baltic amber (1866), small and undifferentiated vermiform inclusions have repeatedly been identified as oligochaetes and more specifically as enchytraeids (Bachofen-Echt 1949, Larsson 1978), although this was mainly inferred from recent observations in habitats comparable to the Baltic amber forest. Wunderlich (1996: 205) points at the easy confusion of these "enchytraeids" with nematodes or dipteran larvae. Rarely were characters observed in the inclusions that would prove the oligochaete nature ofthe object. Chaetae are mentioned only once, but not described (Menge 1866). A clitel- lum is seen in Bachofen-Echt (1949, fig. 17. p. 22). The present amber inclusion of an oligochaete fragment is the first one from which not only the existence ofchaetae is recorded but also their number, shape and distribution is described. What happened to the worm? To ask how the worm could be embedded as a fragment, means to ask if it was wounded after being trapped in the resin, or before. Generally, an animal embedded in a fragmentary state might have been partly exposed before it was completely surrounded by resin, and its exposed parts eaten by a predator. This is unlikely in the present case since the worm, lying flat in the surface, did not offer a point ofattack to predators, or otherwise the surface ofthe resin would have been stirred up which would have left traces. So we must assume that the worm was wounded before it got in contact with the resin. This, however, would be unlikely ifthe worm was either surprised by resin flowing over it or actively crept into the resin, as discussed by Larsson (1978: 117). And, if so, it would lie deep in the amber near the lower surface of the flow. The dolichopodid fly lying beside the fragment offers the explanation that the wonn was torn by the fly when it was pulled out from the soil. Shortly after, both must have been trapped on the resin, otherv\'ise there would not have been much coelomic fluid left to spread on the surface. It may be remarkable that the fragment is the posteriorpart ofthe worm. This con- formstothe observation thatthetwo available fragments ofrecent Enchytraeidae pre- served togetherwith theirpredators are posteriorparts, too. Ifthe few cases allow the conclusion that the fly normally gets the posterior end, this is another support for our assumption that the worm was the fly's prey. The worm and the fly - concluding remarks If our experience from recent field observations is applied to the present amber inclusions and all detailed evidence to be drawn from the amber piece is taken into consideration, everything fits together to suggest that the worm was in fact carried into the resin by the fly as its prey, after it had been torn when pulled out from the