Field of Science

Glenodinium and the Horseshoe of Light

Yes, it's another dinoflagellate. The subject of the above photo (from here) is Glenodinium pulvisculus. Glenodinium is a genus of photosynthetic, mostly freshwater dinoflagellates in the family Glenodiniaceae, diagnosed by Fensome et al. (1983) by the possession of four apical plates and six postcingular plates (see near the top of this post for a brief explanation of these terms). Fensome et al. included two genera in this family, Glenodinium and Glenodiniopsis. Glenodinium has a horseshoe-shaped eyespot, but Glenodiniopsis does not. The eye-spot presumably functions in phototaxis, though it is worth noting that Glenodiniopsis is positively phototactic even without one (Highfill & Pfiester 1992).

Glenodiniopsis uliginosa, from here.

The identity of Glenodinium has been somewhat confused over the years, due in part to confusion over the identity of its type species, G. cinctum (Loeblich 1980). As a result, many of the references to Glenodinium in the literature refer to unrelated species, while true Glenodinium appears relatively little-studied. One species of Glenodiniopsis, G. steinii, has fared a little better, and its ultrastructure was described in detail by Highfill & Pfiester (1992). Among the more interesting details they noted was that instead of the multiple chloroplasts this species had originally been described as having, it really possesses a single chloroplast but one with multiple lobes, so that if it is viewed in cross-section the lobes might appear as individual plastids.


Fensome, R. A., F. J. R. Taylor, G. Norris, W. A. S. Sarjeant, D. I. Wharton & G. L. Williams. 1983. A classification of living and fossil dinoflagellates. Micropaleontology Special Publication 7.

Highfill, J. F., & L. A. Pfiester. 1992. The ultrastructure of Glenodiniopsis steinii (Dinophyceae). American Journal of Botany 79 (10): 1162-1170.

Loeblich, A. R., III. 1980. Dinoflagellate nomenclature. Taxon 29 (2-3): 321-324.

Brachythecium salebrosum: Some Like It Temperate

Brachythecium salebrosum, photographed in Slovakia by M. Lüth.

Brachythecium salebrosum is a species of moss found in many temperate regions of the world. It often grows in drier habitats than other mosses; in a study of the effects of human disturbance on forest moss communities in Estonia, B. salebrosum made up slightly less than a tenth of the moss flora in unmanaged forests, but accounted for more than a quarter of the flora in managed forests (Vellak & Paal 1999). Brachythecium salebrosum is known from Eurasia, North America, southernmost Africa and Australasia. Interestingly, it hasn't yet been recorded from South America (Delgadillo 1993), and an explanation for its distribution would need to explain how it came to disperse (or vicariate) between North America and South Africa yet bypass South America and northern Africa. Another oddity in its distribution can be seen in comparison to the similar species Brachythecium rotaeanum: while both species are found in Eurasia and North America, B. salebrosum is more common in the western part of each continent while B. rotaeanum dominates in the east (so as one travels east from Europe, the distribution bands are salebrosum-rotaeanum-salebrosum-rotaeanum) (Ignatov et al. 2008). Some authors (particularly European ones) have expressed scepticism about the distinction between these two species, but they are distinguished by both morphological and molecular characters according to Ignatov et al. (2008).

Individual leaf of Brachythecium salebrosum, photographed by Russ Kleinman & Karen Blisard.

Distinguishing Brachythecium salebrosum from related species is, admittedly, not a simple task. Specimens of this species can vary quite significantly across their range. Generally, however, B. salebrosum has plicate leaves (i.e. they are folded longitudinally like an accordion) that are more or less falcate in shape with serrated margins. There is a clearly distinct group of small subquadrate cells at the lower corners of the leaf. The spore capsules are held more or less horizontally, and the seta supporting the capsule is generally about two centimetres high (Ignatov et al. 2008).


Delgadillo M., C. 1993. The Neotropical-African moss disjunction. The Bryologist 96 (4): 604-615.

Ignatov, M. S., I. A. Milyutina & V. K. Bobrova. 2008. Problematic groups of Brachythecium and Eurhynchiastrum (Brachytheciaceae, Bryophyta) and taxonomic solutions suggested by nrITS sequences. Arctoa 17: 113-138.

Vellak, K., & J. Paal. 1999. Diversity of bryophyte vegetation in some forest types in Estonia: a comparison of old unmanaged and managed forests. Biodiversity and Conservation 8: 1595-1620.

The Ornithocheirids: Misunderstood Giants

The original specimen of Ornithocheirus simus, from Joseph Dinkel.

The pterosaurs of the Ornithocheiridae that lived between the Early and mid-Cretaceous were not the largest pterosaurs to ever live, but with a possible maximum known wingspan of about 7 m (Martill & Unwin 2011) they were certainly big enough (for comparison, an Australian pelican has a wingspan of about 2.5 metres). Ornithocheirids have been recorded from numerous parts of the world—Europe, South America, Africa and China—and, like the large seabirds that are perhaps their closest modern analogues (where 'closest' is a relative term), were probably found worldwide.

Reconstruction of Ornithocheirus mesembrinus, by Dmitry Bogdanov.

The 'misunderstood' in the title to this post refers specifically to the type genus, Ornithocheirus, which has suffered something of an identity crisis over the years. In the 1800s, 'Ornithocheirus' was used as a catch-all genus for almost all Cretaceous European pterosaurs, many known from only fragmentary remains. Eventually, the name came to be associated with a number of species centred around the British O. compressirostris. However, Unwin (2001) pointed out that, at its earliest publication, only one valid species was associated with Ornithocheirus, O. simus, and that species automatically becomes the type of the genus. Ornithocheirus simus is not currently regarded as congeneric with 'O.' compressirostris (and had for many years been treated under the name of Criorhynchus, e.g. Wellnhofer 1991), and so most species previously treated as Ornithocheirus are now treated as a genus Lonchodectes, and not ornithocheirids. Just to confuse matters further, however, some recent authors have continued to treat Ornithocheirus as typified by O. compressirostris, and refer to the 'Ornithocheiridae' of Unwin (2001) as the 'Anhangueridae'. On the other hand, some recent phylogenies have even suggested that Lonchodectes may itself be related to the Ornithocheiridae (e.g. Andres & Ji 2008).

Reconstruction of Coloborhynchus piscator, by Joseph Conway.

Ornithocheirus simus was originally described from a piece of the front of the rostrum found in the Cambridge Greensand of England. This piece was notably deep, and so O. simus was reconstructed as having a short, deep puffin-like skull. It wasn't until the later discovery of a more complete skull in a closely-related South American species, Ornithocheirus mesembrinus (alternatively known as Tropeognathus mesembrinus), that a more accurate reconstruction was possible: Ornithocheirus species had a long rostrum, similar to that found in related pterosaurs, but with prominent rounded dorsal and ventral crests at the distal end. Species of another ornithocheirid genus, Anhanguera (also known from England and South America), had rostral crests set further back and differently shaped. These crests most likely served some form of display function; suggestions that they may have aided in the capture of fish on the wing by easing the rostrum's passage through the water (Wellnhofer 1991) seem unlikely, as other ornithocheirid genera, such as Brasileodactylus and Barbosania, lacked rostral crests entirely (Elgin & Frey 2011). The possibility has been raised that some crested and crestless forms may represent different sexes of the same species, but unfortunately the fossil record of ornithocheirids may not be extensive enough to establish whether this is the case.

The skull of Ludodactylus piscator, via Darren Naish.

The phylogeny of pterosaurs remains a contentious issue but most studies have agreed that ornithocheirids form part of a clade that also includes the Istiodactylidae, Pteranodontidae and Nyctosaurus (though how exactly these taxa are interrelated has not been agreed upon). Ornithocheirids are distinguished from related forms by the arrangement of their teeth, with the first three pairs enlarged to form a terminal rosette. Most ornithocheirids are also distinguished from Pteranodontidae and Nyctosaurus by the absence of a crest on the back of the head. A noteworthy exception is Ludodactylus sibbicki, described by Frey et al. (2003) from a skull possessing a crest at least basally like that of Pteranodon (the skull is preserved on a slab of rock prepared commercially, and the distal portion of the crest [if it had been present] was removed when the slab was cut). However, it is worth noting that Ludodactylus has not (to my knowledge) been included in a formal phylogenetic analysis. Ludodactylus was identified as an ornithocheirid due to its tooth morphology, but the absence of teeth in Pteranodontidae and Nyctosaurus makes them incomparable in this regard. As the Istiodactylidae (with their broad, duck-like rostra) are also reasonably autapomorphic in their skull morphology, I can't help wondering whether the supposed ornithocheirid characters might be plesiomorphic for the larger pteranodontoid clade. But that, of course, is pure speculation on my part, and something only further study could establish.


Andres, B., & Ji Q. 2008. A new pterosaur from the Liaoning Province of China, the phylogeny of the Pterodactyloidea, and convergence in their cervical vertebrae. Palaeontology 51 (2): 453-469.

Elgin, R. A., & E. Frey. 2011. A new ornithocheirid, Barbosania gracilirostris gen. et sp. nov. (Pterosauria, Pterodactyloidea) from the Santana Formation (Cretaceous) of NE Brazil. Swiss Journal of Palaeontology 130: 259-275.

Frey, E., D. M. Martill & M.-C. Buchy. 2003. A new crested ornithocheirid from the Lower Cretaceous of northeastern Brazil and the unusual death of an unusual pterosaur. In Evolution and Palaeobiology of Pterosaurs (E. Buffetaut & J.-M. Mazin, eds) Geological Society Special Publications 217: 55-63. The Geological Society: London.

Martill, D. M., & D. M. Unwin. 2011. The world’s largest toothed pterosaur, NHMUK R481, an incomplete rostrum of Coloborhynchus capito (Seeley, 1870) from the Cambridge Greensand of England. Cretaceous Research 34: 1-9.

Unwin, D. M. 2001. An overview of the pterosaur assemblage from the Cambridge Greensand (Cretaceous) of Eastern England. Mitt. Mus. Nat.kd. Berl., Geowiss. Reihe 4: 189-221.

Wellnhofer, P. 1991. The Illustrated Encyclopedia of Pterosaurs. Salamander Books: London (reprinted 2000, in The Illustrated Encyclopedia of Dinosaurs (D. Norman & P. Wellnhofer). Salamander Books).

Callocystitids: Ambulacra Advancement and Rhomb Reduction

The Upper Silurian callocystitid Staurocystis quadrifasciata, from Museum Victoria.

The Palaeozoic echinoderms included many distinctive groups that have no close relatives among the modern fauna: blastoids, cornutes, solutes, ctenocystoids... to name just a few. From the Ordovician to the Devonian, this diverse fauna also included a hodge-podge assemblage known as cystoids. Cystoids are a grouping of mostly stalked echinoderms in which certain plates in the theca are perforated by regular arrangements of pores that probably functioned in respiration. Cystoids were not always regularly pentamerous like other echinoderms, and some were notably asymmetrical. The ambulacra were recumbent on the theca, and the feeding appendages were brachioles rather than arms (for the difference between brachioles found in many fossil echinoderms and arms found in crinoids, see the post on blastoids). Cystoids would have been filter-feeders and were probably largely sedentary. Cystoids include some very disparate forms, and many researchers have suggested that they may represent a polyphyletic assemblage. Various authors have suggested cystoid ancestry for other echinoderm groups, such as blastoids or crinoids, but this remains controversial.

The Upper Silurian Schizocystis armata, from Kesling (1967). The two pore rhombs of this species are visible just above the center and at the lower right of the theca.

The Callocystitidae were a family of cystoids that persisted over most of the total cystoid time range. Callocystitids belonged to the major cystoid subgroup called the Rhombifera, in which the diagnostic pore groups were arranged as paired assemblies, commonly called pore rhombs, that spanned the border between two thecal plates (as opposed to the remaining cystoids, the Diploporita, in which pore assemblies each occupied a single plate). Broadhead & Strimple (1978) diagnosed the Callocystitidae based on the arrangement and position of the pore rhombs, together with their possession of a relatively small periproct (the circle of plates that indicates the position of the anus) and the number of radial plates in the theca. All callocystitids possessed a stalk, often divided into a flexible proximal section and a more rigid distal section. Broadhead & Strimple (1978) recognised four subfamilies of callosystitids, but one of these, the Apiocystitinae was explicitly suggested to be paraphyletic to the Callocystitinae and Staurocystinae. This was supported by the numerical phylogenetic analysis of Sumrall & Brett (2002), who furthermore suggested that the Callocystitinae was polyphyletic.

Theca of the Upper Silurian apiocystitine Lovenicystis angelini, from Kesling (1967).

The fourth of Broadhead & Strimple's subfamilies, the Scoliocystinae, was suggested to lie outside the clade formed by the other three; Sumrall & Brett's analysis only included Scoliocystis, but does not contradict this. Scoliocystines have the ambulacra relatively short, restricted to the summit of the theca, and would have had only a small number of brachioles. The most extreme example was the Lower Silurian Osculocystis, which had only a single extremely long brachiole (Paul & Donovan 2011). Another scoliocystine, Schizocystis, had one side of the theca relatively flat and the pore rhombs reduced in number and restricted to the other side, and may have lain on its side in life rather than standing upright.

Reconstruction of Pseudocrinites together with a number of individuals of the discosorid Phragmoceras by Alison Carey.

The remaining three subfamilies had more extensive ambulacra, extending right down to the base of the theca in some species. Apiocystitines and callocystitines had four or five ambulacra, usually branched in callocystitines and unbranched in apiocystitines, that did not strongly protrude above the surface of the theca and had widely spaced brachioles. The more distinctive Staurocystinae had two to four stongly protruding ambulacra that carried tightly packed brachioles. In the staurocystine Pseudocrinites, the theca was discus-shaped with its two ambulacra running around the outer rim of the disc (Kesling 1967).


Broadhead, T. W., & H. L. Strimple. 1978. Systematics and distribution of the Callocystitidae (Echinodermata, Rhombifera). Journal of Paleontology 52 (1): 164-177.

Kesling, R. V. 1967. Cystoids. In Treatise on Invertebrate Paleontology pt. S. Echinodermata 1. General characters. Homalozoa-Crinozoa (except Crinoidea) (R. C. Moore, ed.) vol. 1 pp. S85-S267. The Geological Society of America, Inc., and The University of Kansas: Lawrence (Kansas).

Paul, C. R. C., & S. K. Donovan. 2011. A review of the British Silurian cystoids. Geological Journal 46: 434-450.

Sumrall, C. D., & C. E. Brett. 2002. A revision of Novacystis hawkesi Paul and Bolton 1991 (Middle Silurian: Glyptocystitida, Echinodermata) and the phylogeny of early callocystitids. Journal of Paleontology 76 (4): 733-740.

The Gagrella Problem Cranked Up to Eleven

A Leiobunum eating a cricket, photographed in North Carolina by Jeffrey Pippen.

In previous posts on this site, I have referred to the problem of Gagrella: a large genus of Asian harvestmen diagnosed on the basis of characters long since recognised as unreliable that remains unrevised because no-one has yet been in a position to take on the amount of work required. Well, a paper has come out in the last few weeks that indicates that the Gagrella issue is just part of a larger problem, one of truly demonic proportions.

The British Nelima gothica, photographed by Gordon.

Gagrella belongs to a family of harvestmen called the Sclerosomatidae. Sclerosomatids are distinguished by having a more heavily sclerotised dorsum than most other long-legged harvestmen, and by the morphology of the male genitalia, with many (but not all) species having lateral extensions (like wings) on the shaft just behind the glans. Within the Sclerosomatidae, most authors have recognised four subfamilies: Sclerosomatinae, Gagrellinae, Leiobuninae and Gyinae. The boundaries between the subfamilies have long been realised to be a bit fuzzy, particularly between the Leiobuninae and Gagrellinae. A few molecular studies (for instance, Giribet et al. 2010) that have included more than one representative of more than one subfamily have failed to resolve them as separate. Nevertheless, the subfamilies have served as a convenient way to divide what is a quite large family of over 1200 described species, at least until a more extensive analysis can be conducted.

Enter Hedin et al. (2012) who take molecular data for six genes from seventy-odd sclerosomatids, mostly of the Holarctic genera Leiobunum and Nelima (both Leiobuninae). One suspects that the primary focus of the authors was originally just the phylogeny of these two genera, with the smattering of other sclerosomatids in the mix primarily intended as outgroups. But then they got a result that looks like this (click on the image for a higher resolution):

Yikes. That is a mess. The two main genera have not only failed to resolve as monophyletic, they have completely exploded. Gyas (the type species of the Gyinae) has apparently been so terrified by this show of opilionid pyrotechnics that it has scurried off to join the Phalangiidae. The Gagrellinae have also undergone something of a breakdown, appearing in four separate places on the tree (two of those including species supposedly of a single genus). But this is not simply a matter of poor resolution: many of the incongruent clades are reasonably supported. It is also worth noting that the results are not inconsistent with those recovered by Giribet et al. (2010) who included seven sclerosomatid species in their analysis.

As noted by the authors, the distribution of taxa in the results is not entirely random. There is a certain amount of biogeographic patterning, with the European and Japanese taxa included forming distinct clades, while the New World taxa form three clades. Within the Japanese taxa (which are some of the best studied sclerosomatids), there is some correlation with species groups recognised on morphological grounds, such as monophyly of the Leiobunum curvipalpe group. This latter example is interesting, as it represents a group of species that exhibit conservative features of external morphology despite showing broad variation in genitalic characters (Tsurusaki 1985).

The sclerosomatine Astrobunus laevipes, photographed by Ch. Komposch.

Draining the morass of sclerosomatid taxonomy and phylogeny will doubtless be an arduous process, with all the large genera currently recognised likely to be polyphyletic. The support by Hedin et al. for previously recognised groups such as the Leiobunum curvipalpe complex suggest that morphological features will not be irrelevant in revising the family, but they will have to be analysed in their proper context. For instance, Hedin et al. suggest that the recovered polyphyly of Nelima may indicate that this genus, primarily defined by the absence of ornamentation found in other genera, may represent convergence through paedomorphosis (the acquisition of sexual maturity in a juvenile stage of development). In particular, this could explain how Giribet et al. (2009) found Nelima silvatica nested within two species of Sclerosomatinae, a primarily Mediterranean group of flattened, particularly heavily sclerotised species that might have been thought better supported than the other sclerosomatid subfamilies (unfortunately, Hedin et al.'s analysis includes only a single sclerosomatine). The view of the road ahead might be daunting, but it promises to be a memorable journey.


Giribet, G., L. Vogt, A. Pérez González, P. Sharma & A. B. Kury. 2010. A multilocus approach to harvestman (Arachnida: Opiliones) phylogeny with emphasis on biogeography and the systematics of Laniatores. Cladistics 25: 1-30.

Hedin, M., N. Tsurusaki, R. Macías-Ordóñez & J. W. Shultz. 2012. Molecular systematics of sclerosomatid harvestmen (Opiliones, Phalangioidea, Sclerosomatidae): geography is better than taxonomy in predicting phylogeny. Molecular Phylogenetics and Evolution 62: 224-236.

Tsurusaki, N. 1985. Taxonomic revision of the Leiobunum curvipalpe-group (Arachnida, Opiliones, Phalangiidae). I. hikocola-, hiasai-, kohyai-, and platypenis-subgroups. Journal of the Faculty of Science of the Hokkaido University VI, Zoology 24: 1-42.

Marginal Limpets

Shells of Emarginula solidula, photographed by Jan Delsing.

For today's subject taxon, I've drawn the Emarginulini. This is a tribe within the gastropod family Fissurellidae, members of which are commonly known as keyhole or slit limpets. They get these names because of openings in their shells: keyhole limpets have a distinct hole at the apex of their shell, while slit limpets have a longitudinal slit running back from the front of their shell. In both groups, the slit or 'keyhole' functions in excretion. Gastropods undergo a process early in development known as torsion: the viscera of the embryo twists around so that it reverses its original direction (you can see a basic diagram of the process here). The reasons why this happens remain somewhat uncertain (one early suggestion was that it is what allowed the larval gastropod to retract into its shell and close the shell opening with an operculum) but a potentially negative side-effect of the process is that the anus comes to open directly above the mouth. Unless you want to go through your life with a bit of a funny taste in your mouth, this is not ideal. Therefore, many torted gastropods develop some sort of sinus or recess in their shell so the anal opening can be moved rearwards, away from the mouth.

Live individual of Tugali parmophoroidea, from here. Tugali has an expanded mantle, but is still able to retract it body underneath the shell for protection.

The members of the Emarginulini as recognised by Bouchet et al. (2005) are slit limpets rather than keyhole limpets, though in some species the slit has become much reduced and may only be visible on the underside of the shell (i.e there is a ventral groove rather than a full slit). Bouchet et al. (2005) divided the Fissurellidae between the Fissurellinae and Emarginulinae, a classification based on the structure of the radula and shell muscles (Aktipis et al. 2011). The Fissurellinae are all keyhole limpets, but Bouchet et al.'s Emarginulinae included (in addition to the Emarginulini) the tribe Diodorini, whose members are keyhole limpets like the Fissurellinae but have emarginuline internal anatomy. Other authors have recognised this group as a third intermediate subfamily. Two further tribes of Emarginulinae, the Scutini and Fissurellideini, include species with expanded mantles and reduced shells that may be entirely concealed within the soft body of the slug-like animal.

An elephant snail or shield limpet Scutus sp., from here. This genus has a greatly enlarged mantle, which usually folds over to conceal the reduced shell.

However, this classification of the fissurellids was challenged by the molecular analysis of Aktipis et al. (2011). The results of these authors indicated that the Emarginulini of Bouchet et al. (2005) is para- or polyphyletic. A clade of Fissurellinae with Diodorini indicates a single origin of keyhole limpets, with the emarginuline radula and muscle structure being ancestral for fissurellids as a whole. This result is also consistent with the fossil record: 'emarginulines' are known as early as the Triassic, but fissurellines and diodorines have not been found earlier than the Caenozoic. Aktipis et al. (2011) therefore recognised a more restricted monophyletic Emarginulinae containing the genera Emarginula, Montfortula, Tugali and Scutus, while more basal forms (sister to all other fissurellids) were separated as the Hemitominae (Aktipis et al. did not analyse the position of the Fissurellideini). Of Aktipis et al.'s Emarginulinae proper, only Emarginula has a well-developed slit (the others have ventral shell grooves; Scutus has a quite reduced shell), but this genus was also not monophyletic. Instead of aligning by morphology, the species analysed formed clusters corresponding more to their biogeography: a Mediterranean Emarginula clade, a Pacific clade of Emarginula and Montfortula species, and an Australian clade of Scutus and Tugali. The relationships between these three clades varied by analysis method. The Australian Scutus clade was not necessarily sister to the remaining Emarginulinae, so it may not be worthwhile at this point in time distinguishing the tribes Emarginulini and Scutini.


Aktipis, S. W., E. Boehm & G. Giribet. 2010. Another step towards understanding the slit-limpets (Fissurellidae, Fissurelloidea, Vetigastropoda, Gastropoda): a combined five-gene molecular phylogeny. Zoologica Scripta 40 (3): 238-259.

Bouchet, P., J.-P. Rocroi, J. Frýda, B. Hausdorf, W. Ponder, Á. Valdés & A. Warén. 2005. Classification and nomenclator of gastropod families. Malacologia 47 (1-2): 1-397.