A possible interpretation of the fossilnext section
Although the specimen is small in comparison to other fossil larvae, at least from this Lagerstätte, and may not appear to bear many details, some of these details that are present allow a well-founded interpretation on the identity of the specimen. Texture and fluorescence capacities of the fossil resemble crustacean remains from the same deposits. Also from a structural point of view many interpretations that could come into mind, such as a fish scale, can be easily discarded. Specimens distantly resembling the fossil have been interpreted as possible remains of crustacean larvae (Haug et al., 2011a; 2014b). This seems also a likely interpretation of the new fossil.
When comparing the specimen to small-sized eucrustaceans it shows similarities to larval forms of barnacles and their relatives (Cirripedia). The pelagic larvae of cirripedes (nauplius larvae) are characterized by a pair of spine-like extensions of the anterior shield region, generally termed fronto-lateral horns (Høeg, 1987; Walker, 1992, Høeg and Møller, 2006; Pérez-Losada et al., 2009; 2012; Høeg et al., 2015). Historically, these fronto-lateral horns are an important character that was first recognized by Thompson (1830). For a long time these structures were the only argument for the monophyly of Cirripedia (Høeg et al., 2015). Shape and relative position of the two spine-like extensions of the fossil (Figs. 1D, F, H, 2A) strongly resemble these fronto-lateral horns (Fig. 2B–E).
The preserved presumed appendage remains of the fossil would also well fit into this interpretation. Cirripede nauplii have three functional pairs of appendages: antennulae, antennae and mandibles (Fig. 2B–E; Chan et al., 2014; Høeg et al., 2014a; b; Kolbasov et al., 2014).
The second structure protruding from underneath the shield of the fossil specimen (Fig. 1D, F) strongly resembles the setose swimming exopods of antennae or mandibles of modern cirripede nauplii (Fig. 2B–E; e.g. Walossek et al., 1996). Due to the number of ringlets and setae, the structure on the fossil could represent an antenna, although an interpretation as a mandible cannot entirely be excluded.
The appendage remain on the other side of the fossil specimen (third structure; Fig. 1D) could represent the less well preserved antenna of the other body side, although it remains unclear whether it could then represent the endopod or the exopod. The further anterior, very incomplete appendage (first structure, Fig. 1D) is more difficult to interpret. The distinct ringlets could be understood as another exopod. The position would argue more for an interpretation as an antennula, yet, an antennula would not be organized into such discrete ringlets. In conclusion, the observed structures are compatible with the interpretation of the fossil as a cirripede nauplius.
Difficulties with the interpretation
When interpreting the fossil as a larval form of a barnacle or one of its relatives, three possible aspects need to be discussed:
The fossil is comparably large, at least for a nauplius, as most eucrustacean nauplii are rather small. Nauplii of representatives of Cirripedia are mostly in a size range between 200 µm and 1 mm (Walossek et al., 1996; Walker, 2001; Høeg et al., 2004). Yet, also nauplii reaching astonishing sizes have been reported (Rybakov et al., 2003). In fact, shield sizes well over 1 mm seem not to be uncommon among modern forms (Fig. 2), resulting in total lengths of about 6 mm in Lepas anatifera (Moyse, 1987) or in Lepas pacifica (Ryusuke Kado, unpublished data).
The only fossil example of a possible cirripede larva is that of Rhamphoverritor reduncus (Briggs et al., 2005; see also further below). This larval specimen is not a nauplius, but may represent a settling stage, a so-called cypris, hence the stage following the last nauplius stage. Among modern forms the lengths of cypris larvae are difficult to infer from the literature. The fossil cypris has a total length of 4 mm.
Crustaceans usually increase their size by up to 30 % within a single molt (see discussion in Kutschera et al., 2012). The largest known cirripede eggs can reach up to 400 µm (Korn et al., 2004). All extant representatives of Cirripedia develop through at most six naupliar stages (nauplius I – nauplius VI; Høeg et al., 2015). By calculating this example, the possible maximum size of a nauplius VI would result in an overall size of about 2 mm.
However, the 30% rule seems to be less strict in certain crustaceans. The size increase between nauplius I and nauplius II in e.g. Lepas pectinata, is in average 150% (Moyse, 1987). Consequently, nauplius VI could reach overall lengths of more than 7 mm. Taking this into account, a shield length of 4.7 mm in the fossil specimen described herein is quite reasonable (but see also further below).
2) Position of the fronto-lateral horns:
In most cirripede nauplii the fronto-lateral horns arise right from the fronto-lateral corners of the shield (Fig. 2 B–C, E). This seems not to be the case in the fossil specimen. Here the shield rim is further drawn out, forming a set-off ring. Interpreting the horns differently is difficult, other possible structures such as frontal filaments, which occur within Thecostraca in all representatives (Walker, 1974; Grygier, 1987), are tiny and soft and hence unlikely to be preserved in a fossil. Also they are not horn like. In some naupliar stages, e.g. of the rhizocephalan Peltogaster paguri, the fronto-lateral horns are fully covered by a round extension of the shield (Fig. 2D; Høeg, unpublished data). These structures in the fossil specimen described herein are blunt at the tip and might therefore end in a pore as do true fronto-lateral horns. This observation supports the interpretation of the spine-like extensions as fronto-lateral horns and not as frontal filaments.
3) Interpretation of the set-off ring:
Examples of extensions of the shield, so-called ‘floating collars’, occur in some ingroups of Cirripedia, more precisely of Rhizocephala (exclusively parasitic forms). Such a floating collar has been considered as floatation device, enhancing the buoyancy of the nauplii (Veillet, 1943; Høeg et al., 2004). Such a type of floating collar (Fig. 2D) is known from the rhizocephalan ingroups Peltogastridae and Lernaeodiscidae, but could be part of the rhizocephalan ground pattern (Høeg et al., 2004; Glenner and Hebsgaard, 2006; Høeg et al., 2009).
The floating collar in rhizocephalans is shed separately from the rest of the cuticle and is made of exceedingly thin cuticle (Fig. 2D; Høeg et al., 2004). This seems to be quite different in the fossil specimen. Also in the fossil the possible floating collar seems to be positioned under the horns, while in modern forms it is over these. Still the structure and position of the ring in the fossil could still indicate an at least comparable function in the fossil. It could also be speculated that this could be indicative of a closer relationship to Rhizocephala.
Other possible interpretations:
1) Malacostracan affinity:
Most fossil larvae from the Solnhofen limestone have been identified as malacostracan larvae (see below). The fossil specimen described herein resembles in certain aspects a supposed malacostracan larva from the Solnhofen limestone (Haug et al., 2011a; 2014b, fig. 32.2K). The specimen has been suggested to represent the remains of a shield of a decapod zoea. Could this interpretation also apply to the specimen described here? This is unlikely. The supposed fronto-lateral horns could be interpreted as lateral spines for example of a brachyuran zoea. In such a case we would expect additional spines, especially a rostral spine and a postero-dorsal spine (Wear, 1968; Martin, 1984; Haug et al., 2011a; Martin, 2014a). Also in other decapod zoeas especially a pronounced rostral spine should be expected. No breakage indicators are apparent that could indicate an absence due to preservation. Also the shape of the spines and their blunt tips would be unusual for a zoea larva. Therefore an interpretation of the new fossil as a zoea appears unlikely to us. Notably, already Haug et al. (2014b, p. 176) stated that the “systematic affinities remain uncertain until better-preserved specimens are found”. The specimen from Haug et al. (2014b) could in the light of the new fossil described here also represent the conspecific cypris larva. The specimen should be reinvestigated for this aspect.
2) Branchiopod affinity:
There is also a certain resemblance of the fossil to the nauplius larva of representatives of Laevicaudata, an ingroup of Branchiopoda. These have a kind of spine-like extensions that represent the still immobile antennulae (Olesen, 2005). In contrast to larval representatives of Laevicaudata, in which these horns protrude from the ventral side of the head (Olesen, 2005; 2007), it seems that the horns in the fossil specimen described herein protrude from the dorsal side of the head shield, indicated by the relative position of the appendages (Fig. 1). Additionally, laevicaudatan nauplii have a distinct triangular shape of the anterior head which should be expected to be seen in the fossil if present. Yet, this is not the case. Also other characteristic features, such as a large, rounded labrum or caudal lobes, which are spine-like extensions posterior from the shield (Olesen, 2005; 2007) are not present in the fossil specimens. Yet, these could be more difficult to be visible, as the labrum is a soft ventral structure and the caudal lobes are comparably small. Lastly, most branhciopods are fresh water forms, only few groups of raptorial cladocerans have re-entered the marine realm, yet the original lagoons of the Solnhofen lithographic limestones must have represented a marine environment. Thus, a laevicaudatan or even a branchiopod affinity is very unlikely.
Summarizing: From the morphological point of view it seems likely that the here described fossil indeed represents a cirripede nauplius. It appears to possess a kind of floating collar that may point to a closer relationship to rhizocephalan cirripedes. The “main” shield would then measure about 3 mm and could molt into a cypris larva of the size as it is known for the fossil Rhamphoverritor reduncus with 4 mm length (Briggs et al., 2005). While the new larva is well in a possible size range for cirripede larvae, it clearly represents a giant form.
Early fossil record of Cirripedia
Cirripedes have a comparably good fossil record, at least concerning their adults. Rhamphoverritor reduncus from the Silurian (420 mya) is exceptional as only a possible cypris larva and a juvenile are known (Briggs et al., 2005). The species most likely represents the sister group to all other cirripedes (Høeg et al., 2009). There is generally a distinction of three groups within Cirripedia: Acrothoracica, Thoracica and Rhizocephala, with the latter two groups representing sister groups. The monophyly of each of the three groups is generally well supported. Yet, Thoracica is not as well characterized by morphological characters. It is therefore possible that any pedunculated fossil barnacle older than the presumed split between Rhizocephala and Thoracica (see below) might be situated phylogenetically below this point.
Representatives of Acrothoracica have been reported as trace fossils from the Devonian (380 mya; Glenner et al., 1995). Molecular analyses give support for the origin of Thoracica in the Early Carboniferous (340 mya; Pérez-Losada et al., 2008). Based on the reconstruction of a co-evolution between rhizocephalans and anomalan crabs and molecular reconstructions of thoracican barnacles, representatives of Rhizocephala have been estimated to be present also since the Carboniferous (Walker, 2001; Boyko and Williams, 2009). As a consequence, all pedunculated fossil thoracicans older than 340 million years could be considered as representatives of the unnamed sister group to Acrothoracica. The first more direct fossil indications of rhizocephalans are feminized male crabs from the Miocene (4–23 mya; Feldmann, 1998). Also important to mention in this aspect: fossils of cirripedes are well known to occur in the lithographic limestones of southern Germany (Barthel et al., 1990; Nagler et al., 2017).
With this fossil record an interpretation of the here described fossil as the nauplius of a cirripede and even as a possible relative of Rhizocephala seems reasonable; at least it is not contradicted. The fossil, therefore, most likely represents the first fossil record of a cirripede nauplius. It also follows the general pattern that we seem to be more likely able to find especially giant larval forms as fossils.
Giant larvae in metazoans
The phenomenon of oversized larval forms has been reported from various metazoan groups. Yet, in many cases ‘giant’ is a matter of relation. An overview of giant larvae can be seen in Tab. 1.
Table 1. Overview of giant larvae with larval terms and reported maximum sizes of their respective group or close relatives.
As pointed out above, larvae of flying insects (Pterygota) are in their final larval stage often as large, sometimes larger, than the adult. Yet, as almost all insects have such comparably large larvae it is somehow difficult to consider any of them as a giant. Comparably larger larval size is mostly coupled to larger adult size.
Larvae of corals, sea anemones and others (Cnidaria) – planula – have an average maximum size of about 1 mm (Leloup, 1932). Yet, also specimens of up to 11 mm have been reported. Some of the even larger specimens with larva-like morphology already possess gonads (Molodtsova 2004; Stampar et al., 2015) and are therefore no longer larvae in the meaning of being immature.
The planktic larvae of marine snails and slugs (Gastropoda) – veliger – are usually below 1 mm in size before settling to a pelagic life. Yet, in some groups significantly larger forms are known. Veliger larvae of strombiids, coniids and cypraeiids have extremely elongated structures, the velum lobes. With these structures they reach sizes of about 5 mm (Hickman, 1999). Even larger forms of about 6–7 mm have been reported by Dawydoff (1940).
The early larval stage of ringed or segmented worms (Annelida) is plesiomorphically the trochophora. These are mostly below one millimeter in size before they metamorphose into forms with few body segments that carry appendages (chaetigers; often three such segments). Exceptions are special forms of phyllodocid larvae. Here the trunk grows significantly longer from the trochophora before undergoing metamorphosis. The spherical anterior region (hence the original trochophora) can reach sizes of up to 2 mm; the trunk with up to 120 rudimentary segments can reach 10 mm. Hence, the total length of these larvae reaches up to 12 mm (Tzetlin, 1998).
Larvae of peanut worms (Sipunculida) – pelagosphaera – have an average size of 300 µm (Rice, 1967). Yet also significantly larger forms of up to 3.2 mm can sometimes be found in the plankton of open ocean regions (Rice, 1973).
Larvae of horseshoe worms (Phoronida) – actinotrocha – reach in general a maximum size of 0.7–0.9 mm. An unusually large phoronid larva has been reported by Temereva et al. (2006). This larval specimen was 3.5 mm long, thus 4–5 times larger than a “normal” actinotrocha larva.
Larvae of echinoderms (Echinodermata) are generally small, below 1 mm (e.g. Pawson, 1971). Yet, certain larvae of abyssal sea cucumbers (Holothuroidea) – auricularia – can reach sizes between 3 and 15 mm (Ohshima, 1911; Mortensen, 1913; 1921; Garstang, 1939). Also, the larva of the deep-sea starfish Luidia sarsi (Asteroidea) – bipinnaria – can reach body lengths of up to 25–35 mm (Domanski, 1984).
Larvae of acorn worms (Enteropneusta, Hemichordata) – tornaria – reach usually about 0.5–1 mm (Stiasny, 1928). Giant tornaria-like larvae (Planctosphaera pelagica) with a length of up to 28 mm have been found in the Atlantic and Pacific Ocean (Spengel, 1932; Hadfield and Young, 1983). Thus the found giant larvae are at least 20 times bigger than the “normal” larvae of Hemichordata. Yet, it is still controversial if Planctosphaera pelagica represents an ingroup of Enteropneusta (Hadfield and Young, 1983) or a separate group of hemichordates (Van der Horst, 1936).
Larvae of teleost fishes (Teleostei) are often quite large; few centimeters length is not uncommon. A very notable size is reached by larval eels (Anguilliformes) – leptocephalus – which regularly reach 300 mm in length (Miller, 2009), but sometimes even giant larvae longer than 1800 mm have been reported (Aron and McCrery, 1958; Tabeta, 1970; Kurogi et al., 2016).
Amphibian tadpoles (Lissamphibia) are all large compared to many other metazoan larvae, being in the range of several centimeters. Tadpoles of the frog Pseudis paradoxa reach sizes of up to 230 mm (Emerson, 1988). Also other species of Pseudis can reach quite a large tadpole sizes with up to 180 mm (Fabrezi et al., 2009). In these species the larva is also significantly larger than the adult. Fossil tadpoles with a size of up to 150 mm have been reported from the Miocene (Roček et al., 2006) and from the Lower Cretaceous (Chipman and Tchernov, 2002).
Giant larvae in crustaceans
Among numerous crustacean groups giant larvae have been reported, especially among decapods. Decapods usually have (at least) two larval phases: The pelagic zoea larvae swim with the outer locomotion branches (exopods) of their thoracopods. This phase may include up to ten stages. The zoea is followed by the megalopa, which mediates the transition between the pelagic larva and the benthic juvenile. Most megalopae have lost their exopods on the thoracopods and swim with their pleopods. In many groups there is only a single megalopa stage. Sometimes larvae show a kind of mixed morphologies somewhere “between” zoea and megalopa. The latest zoea as well as the megalopa usually measure only few millimeters in total length. Yet, there are quite some exceptions:
Zoea larvae of prawns (Dendrobranchiata) are usually small with shield lengths rarely reaching 1 mm. Yet, within Aristidae zoea larvae formerly addressed as “Cerataspis monstrosa” reach shield lengths of almost 12 mm (Bracken-Grissom et al., 2012).
Polychelidan lobsters (Polychelida) only have a short zoea phase (Torres et al., 2014), but have several megalopa stages that reach astonishing sizes. These eryoneicus larvae reach sizes of more than 100 mm in length (Martin, 2014b; Eiler et al., 2016). Fossil forms that show some similarities to modern forms and also an increased size have been reported from the Jurassic Solnhofen limestones (Eiler and Haug, 2016), and from the Cretaceous limestones of Lebanon (Haug et al., 2015a).
Achelatan lobsters (Achelata) develop through a characteristic type of zoea larva, the phyllosoma (Palero et al., 2014). Phyllosoma larvae have been recognized as giant larvae frequently in the literature. They can reach up to 80 mm in body length, with their thin legs extending even longer (Guérin, 1822; Richters, 1873; Johnson, 1951; Sims, 1964; Sims and Brown, 1968). Phyllosoma larvae most likely represent the largest decapod larvae (Palero et al., 2014). As a consequence, also the megalopa larvae of achelatans (nisto and puerulus larvae) are significantly larger than other types of megalopa larvae. Giant phyllosoma larvae have been reported from the fossil record with body length up to 100 mm. Besides “typical phyllosoma larvae (Polz, 1972; 1973; 1984; 1987; 1995; 1996; Haug et al., 2009; 2011a), large nisto larvae (Audo et al., 2014; Haug and Rudolf, 2015), but also transitory forms with a “mixed” morphology of phyllosoma and post-phyllosoma stages have been reported (Haug et al., 2013b; Haug and Haug, 2013, 2016).
Larvae of false sand crabs (Hippidae) usually reach a total length of 2 mm. A single specimen has been reported to have reached 15 mm in total (Martin and Ormsby, 1991). Yet recently more material turned up demonstrating that reaching such a size may be quite more common among false sand crabs than expected (Rudolf et al., 2016).
While mantis shrimps (Stomatopoda) are not decapods, they show certain similarities to them including various aspects of their larval development. Their later larva can be roughly seen as the functional equivalent to the megalopa larva in decapods. Depending on the specific ingroup these larvae are of the alima-type or of the erichtus-type. Both reach sizes of several centimeters. Alima-type larvae have been known to reach up to 5 mm (Ahyong et al., 2014). Just recently new very large erichthus-type larvae have been described (Haug et al., 2016). Erichthus-type larvae have also been described from the Jurassic lithographic limestones from Germany with up to 18 mm (Haug et al., 2008; 2015b). Notably, giant fossil larvae from the Triassic Hallstatt limestone from Austria with shield lengths of 13 mm show certain characters of the mantis shrimp larvae, and also similarities to the false sand crab larvae (Hyžný et al., 2016).
The possible function of giant larvae
Generally, we can distinguish between two types of giant larva: Type one are facultative giant larvae, type two are obligate giant larvae.
Type one giant larvae occur in species that usually have “normal-sized” larva, but in which from time to time giant individuals occur. Here ‘giant’ is meant in comparison to individuals of the same species. Such giant larvae must be understood as caused by external factors. A rather simple and probably widespread case for causing such instances is the simple absence of a settling trigger. Many larvae need specific chemical environmental cues that indicate an advantageous habitat for the benthic juvenile/adult. If such cues are absent, larvae can simply continue to grow without metamorphosing. Also other abiotic factors have been suggested to be important in this aspect. For example, temperature and shifts in photoperiod length seem to influence the development of tadpoles in the direction to giant tadpoles (Emerson, 1988; Fabrezi et al., 2010).
It has also been suggested that giant size of larvae may be a consequence of a physiological defect. Such larvae often already develop adult organs, e.g., primordial gonads (Temereva et al., 2006). A disruption in thyroid hormone production before metamorphosis has been suggested as reason for this phenomenon (Emerson, 1988; Shi and Hayes, 1994; Schreiber et al., 2001; Yun-Bo et al., 2001; Ogielska and Kotusz, 2004; Rot-Nikcevic and Wassersug, 2004; Roček et al., 2006).
Parasites have also been identified as causes of suppressing a metamorphosis trigger, with this leading to giant-sized larval forms. Insect larvae infected with parasites molt more often than non-parasitized larvae and die as giant larvae (Fisher, 1963). Hormones increasing the juvenile activity of the host cause this exceptional development. In this way, the parasite gets a larger host by its hormone manipulation (Dawkins, 1990).
Type two giant larvae are cases in which representatives of all individual species (or larger group) develop through larval forms that grow significantly larger than the larvae of closely related groups (Fabrezi and Goldberg, 2009). This also leads to a prolonged larval phase. Such a prolonged larval span can enhance the capability for long-distance dispersal in the planktic phase of some species of different molluscs, echinodermatans, or achelatan lobsters (Domanski, 1984).
In this context, one could think of abyssal gigantism (Herring, 2001) also as explanation for giant larvae. Mainly crustaceans have been reported to reach a larger size in deep-sea environments than their relatives in shallow waters (King and Butler, 1985; Mauchline, 1995; Chapelle and Peck, 1999). Low temperature and restricted food availability in deep seas are thought to decrease growth rates, but to increase longevity and the time span to reach sexual maturity (Nybakken, 2001). Hence, it seems to affect juvenile instead of larval development, not necessarily leading to large larvae. Abyssal gigantism has been proposed for the loriciferan Higgins larvae by Gad (2005). Yet, these forms are in fact paedomorphic adults and therefore not larvae.
Giant larvae of type two often bear structural specializations. In many giant crustacean larvae spines or extensions of the shield are necessary to increase the buoyancy (Eiler et al., 2016; Haug et al., 2016). Eel larvae deposit large amounts of glycosaminoglycans in their musculature increasing their swimming ability due to the enhanced skeletal stability (Bishop and Torres, 1999). Giant acorn worm larvae are adapted to a prolonged larval span by relatively larger feeding structures to process more food (Damas and Stiasny, 1961; Strathmann and Bonar, 1976).
Interestingly, we can even identify combined cases of type one and type two giant larvae. Eel larvae are in some species 300 mm in average and with this significantly larger than many other fish larvae and representing cases of giant larvae of type 2. Yet, among these even larger larval individuals are known of 1800 mm, with this being cases of type 1, representing a kind of super-giant larva.
Interpretation of the present case
Cirripede nauplius larvae represent dispersal and growth stages that can last short or long (Høeg et al., 2015). A short larval span is only possible if the larvae find a suitable habitat in close distance to their parents (Buhl-Mortensen and Høeg, 2006). In environments that have a patchily distributed settlement habitat, it is more likely that larger larvae are adapted for long-distance and long-time dispersal as it has been reported for some deep-sea cirripedes (Buhl-Mortensen and Høeg, 2006; Yorisue et al., 2013). The Solnhofen limestone Lagerstätte represents a Jurassic back-reef lagoon (Barthel et al., 1990), where suitable habitats for cirripedes might have been rare and nauplii must have searched for a long time for their settlement site. Additionally, in modern cirripedes, lecithotrophic nauplii are more rounded and larger than planctotrophic nauplii, but show more simple setae and reduced development of the appendages and the labrum (Anderson, 1965; 1987; Høeg et al., 2004). However, the fossil specimen described herein is generally large and rounded, but show at the same time well developed appendages and possibly a well developed labrum (Fig. 1D). Hence, it is likely that the fossil specimen described herein could store lipids and ingest food for its metabolic needs at the same time to survive a long-term dispersal phase. As pointed out above, modern cirripedes seem to be restricted in the number of molts as a nauplius. It seems therefore most likely that the larva represents a case two, i.e. an obligate dispersal larva. This is also in accordance with a supposed floating rim of the shield.
It might be seen as special that we have a highly specialized nauplius larva as the first fossil report of a cirripede nauplius. Yet, it is in overall concordance that we tend to find giant larvae. Moreover, the finding is also important because it provides us a rare look into the Mesozoic plankton of which our knowledge is still very incomplete.