Field of Science

Naviculi, Navicula

Diatoms are one of the most prominent groups of micro-algae in aquatic environments, perhaps more abundant than any other major group of aquatic organisms except bacteria. As such, they are a key component in many of the environmental processes that we ultimately depend on: food for aquatic animals, producers of oxygen, et cetera et cetera. To those who study them, they are also known for the intricate architecture of their silica walls. As well as being aesthetically pleasing, this architecture forms a key component of diatom classification. One of the most diverse groups of diatoms recognised has been the mega-genus Navicula.

Light microscope view of Navicula tripunctata, copyright Kristian Peters.

Historically, over one thousand species have been assigned to Navicula. Though more recent authors have restricted the name to a smaller, more tightly defined concept than before, it still contains some 200 or so species (Bruder & Medlin 2008). Species assigned to this genus are an elongate diamond or pill shape. Though the term 'navicula' can be translated from Latin as a small boat, and this is often assumed to be the name's origin, this is incorrect. Its original author, the French naturalist Jean-Baptiste Geneviève Marcellin Bory de Saint-Vincent, derived the name from the French term for a weaver's spindle (navette de tisserand; Cox 1999). A long fissure, the raphe, runs down the midline of each valve of the diatom wall; the diatom moves by extruding secretions through the raphe. In Navicula, the raphe is largely straight though it may be hooked at the ends of the valve. Perpendicular to or radiating from the raphe are striae formed of rows of openings (areolae); in Navicula, these areolae are more or less elongate with their long axes perpendicular to the line of the stria. In some species historically included in Navicula, the striae may be biseriate with two rows of areolae. Some authors have proposed recognising species with biseriate striae as a distinct genus Hippodonta. Cox (1999) disputed whether this distinction was enough to warrant a separate genus but Bruder & Medlin (2008) conducted a molecular phylogenetic analysis of naviculoid diatoms in which the one Hippodonta species included was placed as the sister taxon to Navicula sensu stricto. In distinguishing the genus Sellaphora from Navicula, Mann (1989) also identified a number of cytoplasmic features characteristic of Navicula sensu stricto, such as the possession of two distinct plastids per cell with rod-like pyrenoids.

SEM view of Navicula dobrinatemniskovae, from Van de Vijver et al. (2011). Scale bar = 1 µm.

Ecologically, the majority of species of Navicula sensu stricto (about 150 species) are found in freshwater environments (Bruder & Medlin 2008). In temperate and tropical regions, they are a diverse element of benthic diatom communities, but they are less predominant in coldwater habitats (Van de Vijver et al. 2011). They are most characteristic of meso- to eutrophic lakes and permanent waterways and Van de Vijver et al. (2011) therefore suggested that they might be less suited for the damp soils and temporary pools that dominate freshwater habitats in the frozen South. Nevertheless, these authors still managed to identify five previously unknown species from just this inhospitable region, giving some indication of what still remains to be discovered of this already diverse genus.


Bruder, K., & L. K. Medlin. 2008. Morphological and molecular investigations of naviculoid diatoms. III. Hippodonta and Navicula s. s. Diatom Research 23 (2): 331–347.

Cox, E. J. 1999. Studies on the diatom genus Navicula Bory. VIII. Variation in valve morphology in relation to the generic diagnosis based on Navicula tripunctata (O. F. Müller) Bory. Diatom Research 14 (2): 207–237.

Mann, D. G. 1989. The diatom genus Sellaphora: separation from Navicula. British Phycological Journal 24 (1): 1–20.

Van de Vijver, B., R. Zidarova, M. Sterken, E. Verleyen, M. de Haan, W. Vyverman, F. Hinz & K. Sabbe. 2011. Revision of the genus Navicula s.s. (Bacillariophyceae) in inland waters of the sub-Antarctic and Antarctic with the description of five new species. Phycologia 50 (3): 281–297.

The Age of Olcostephaninae

Ammonites are among the iconic fossils of the Mesozoic. These shelled cephalopods dominated the oceans during their heyday and diversified into a wide array of taxa. Many of these have become significant for recognising particular periods in the earth's history; among these are members of the Olcostephaninae of the Early Cretaceous.

Olcostephanus astierianus, copyright Hectonichus.

The Olcostephaninae, as recognised by Wright et al. (1996), are known from the Valanginian and Hauterivian epochs of the Early Cretaceous, disappearing from the fossil record some time during the earlier part of the latter. The Valanginian ran from about 140 to 133 million years ago; the Hauterivian lasted for about three and a half million years after that. A brief reminder here: the Cretaceous lasted for a bloody long time, with more time separating the beginning and end of the Cretaceous than separates the end of the Cretaceous and today. One genus described from Pakistan, Provalanginites, has been supposed to come from the latest Jurassic but, as this is at least five million years earlier than any known olcostephanine anywhere else, its age is regarded as questionable. Olcostephanines can be very abundant in formations of the right age. A mass occurrence in the latest Valanginian of northwestern Europe has long been recognised as a geological marker, dubbed the 'Astierien Schichten' (Astieria being a synonym of Olcostephanus; Lukeneder 2004).

Saynoceras verrucosum, from here.

Olcostephanines are small to moderate-sized ammonites. Lukeneder (2004) refers to macroconches* of Olcostephanus guebhardi up to about ten centimetres in diameter. The olcostephanines pictured in Wright et al. (1996) seem to indicate an average size smaller than this and the group also includes a number of dwarf genera that look to only be a bit over one centimetre in diameter. The shell of olcostephanines is usually characterised by a pattern of transverse ribs coalescing in bundles to meet tubercles on the inner margin of the whorl. One dwarf genus, Saynoceras, has a stronger ornamentation of two rows of tubercles near the midline and outer margins of the whorls.

*A common pattern in ammonoids is the co-occurrence within a formation of distinct forms, termed 'macroconches' and 'microconches', that are broadly similar except in size and the configuration of the aperture (generally simple in macroconches but with protruding lappets in microconches). The most popular interpretation of this phenomenon is that the forms represent sexual dimorphism. Obviously which sex is which can't be known at this time though comparison with living cephalopods suggests that the macroconches may be female.

Valanginites nucleus, from here.

Olcostephanines are very similar in external appearance to the earlier subfamily Spiticeratinae (known from the very earliest part of the Cretaceous) and are likely to be descended from among that group. Though the Olcostephaninae themselves as currently recognised disappeared during the Hauterivian epoch, this may not have been the actual end of the olcostephanine lineage. The slightly later Holcodiscidae are very similar to the olcostephanines and some have questioned whether they even warrant separation. There is also a strong similarity between early members of the superfamily Desmoceratoidea and species of Olcostephanus (Wright et al. 1996). If this similarity also indicates ancestry, then the family line of the olcostephanines would continue right until the final extinction of the ammonites at the end of the Cretaceous.


Lukeneder, A. 2004. The Olcostephanus level: an Upper Valanginian ammonoid mass-occurrence (Lower Cretaceous, Northern Calcareous Alps, Austria). Acta Geologica Polonica 54 (1): 23–33.

Wright, C. W., J. H. Calloman & M. K. Howarth. 1996. Treatise on Invertebrate Paleontology pt L. Mollusca 4, revised vol. 4. Cretaceous Ammonoidea. The Geological Society of America, Inc.: Boulder (Colorado), and The University of Kansas: Lawrence (Kansas).

The Pireninae

The chalcidoid wasps are truly a remarkable array: tiny wonders coming in a bewildering variety of forms. For this post, I'm looking at the members of the chalcidoid subfamily Pireninae.

Macroglenes sp., copyright Charley Eiseman.

The Pireninae are currently recognised as one of the subfamilies of the Pteromalidae, a chalcidoid 'family' that is long overdue for reclassification as phylogenetic studies have agreed that it is extensively polyphyletic* (e.g. Heraty et al. 2013). The pirenines are very small wasps, about one to two millimetres in length. They've always struck me as having a fairly fly-like habitus: they lack the metallic coloration and strong sculpturing of many other pteromalids, often being uniformly black or yellow, and carry upright bristle-like setae on the head and mesosoma. Characteristic features of the Pireninae also include antennae inserted low on the face, reduced numbers of antennal segments (and hence often rather short antennae), a large clypeus that often protrudes ventrally, and a dorsally rounded mesosoma often with deeply impressed notauli (longitudinal grooves on the mesoscutum) (Bouček 1988). About ten genera are currently recognised in the subfamily. Perhaps the most remarkable is the genus Zebe, named by John La Salle in 2005 from a single female that he says, at the time, had stymied multiple hymenopterists as to what it might be for two decades. Zebe has legs with four-segmented tarsi, instead of the five-segmented tarsi of other pteromalids, and the female has a long horn extending forward from the mesoscutum and hanging over the head. As in other micro-wasps with comparable structures, this horn probably provides space for the retraction of an extraordinarily long ovipositor.

*But just in case anyone who stands to have some influence is reading, I will point out that the number of subfamilies that need to excised from the Pteromalidae could be kept to a minimum if the family is expanded to include the Ormyridae, Torymidae, Eucharitidae and Perilampidae. Bonus points that these are the families that are the hardest to distinguish from pteromalids to begin with.

Female Zebe cornutus, from Mitroiu (2011).

The life habits of most pirenines are very little known. Those whose hosts are known develop as parasites of Cecidomyiidae, gall midges, and may be found in association with the galls produced by those flies on various plants. Species of the genus Macrogelenes attack cecidomyiids associated with grasses, and some have been investigated as control agents for midges on commercial crops. So these tiny little wasps could prove themselves very valuable to humans.


Bouček, Z. 1988. Australasian Chalcidoidea (Hymenoptera): A biosystematic revision of genera of fourteen families, with a reclassification of species. CAB International: Wallingford (UK).

Heraty, J. M., R. A. Burks, A. Cruaud, G. A. P. Gibson, J. Liljeblad, J. Munro, J.-Y. Rasplus, G. Delvare, P. Janšta, A. Gumovsky, J. Huber, J. B. Woolley, L. Krogmann, S. Heydon, A. Polaszek, S. Schmidt, D. C. Darling, M. W. Gates, J. Mottern, E. Murray, A. D. Molin, S. Triapitsyn, H. Baur, J. D. Pinto, S. van Noort, J. George & M. Yoder. 2013. A phylogenetic analysis of the megadiverse Chalcidoidea (Hymenoptera). Cladistics 29: 466–542.

La Salle, J. 2005. Zebe cornutus gen. et sp. nov., a new Pireninae (Hymenoptera: Pteromalidae) with 4-segmented tarsi and a mesoscutal horn. Acta Societatis Zoologicae Bohemoslovenicae 69: 193–197.

Snakes and Lace

The holometabolous insects—that is, the clade containing most insects with a complex life cycle including differentiated larval and pupal stages—is one of the most extensive radiations of animals on this planet. Much of this diversity is assigned to four major orders: wasps, moths, beetles and flies. But there are also a number of smaller lineages making up the holometabolous insects. Among these are the lacewings and their relatives in the clade Neuropterida.

Female snakefly Puncha ratzeburgi, copyright Hectonichus.

Modern members of the Neuropterida are generally recognised as belonging to three orders—the lacewings and ant-lions in the Neuroptera, the snakeflies in the Raphidioptera, and the alderflies and dobsonflies in the Megaloptera—though go back a few decades and you may find texts referring to a single order Neuroptera. A number of authors have advocated for use of the name 'Planipennia' for the lacewing order to avoid confusion with the broader sense of Neuroptera but, while a case could certainly be made for this usage, it's just never really caught on. Most neuropteridans are fairly similar in overall appearance: long-bodied insects with well developed wings with numerous crossveins. Of the living holometabolous insects, they probably bear the greatest overall resemblance to the clade's ancestors and hence they are commonly thought of as 'relicts'. However, they do possess their own specialisations and are not primitive in every regard (for instance, the most primitive egg-laying apparatus among holometabolous insects belong to wasps). Species of Neuropterida are mostly predators as larvae. The larvae of the lacewing family Ithonidae may possibly feed on decaying plant matter though we don't know for certain (Grimaldi & Engel 2005). Adults are predators and/or pollen-feeders, or may not feed at all in some short-lived forms.

Male (above) and female dobsonflies Corydalus cornutus, copyright Didier Descouens.

The exact relationships between the neuropteridan orders have been debated over the years. Though most of their obvious similaities to each other represent shared ancestral features, there is a broad consensus that they do indeed form a clade. There has also been little, if any, question of the monophyly of the Raphidioptera and Neuroptera; the monophyly of Megaloptera has been more debated but seems more likely than not. Most recent studies have suggested that the Raphidioptera are the sister group to a clade of the other two orders (Engel et al. 2018). Raphidioptera are the least diverse of the generally recognised living orders of insects with about 250 known species. They are found in cooler regions of the Northern Hemisphere—in the temperate zone or at higher elevations in lower latitudes—and are completely absent from the Southern Hemisphere (Aspöck & Aspöck, 1991, refer to a failed attempt to introduce them to Australia and New Zealand but provide no details why such a thing was tried in the first place). They are characterised by a notably elongate prothorax (the first segment of the thorax) which explains the vernacular name of 'snakefly'. Larvae live under bark or in litter and moult into pupae with the onset of cold weather. The pupae of Raphidioptera and Megaloptera are primitive in aspect, with legs separate from the body wall, and are highly mobile. Engel et al. (2018) even refer to the pupae of Raphidioptera as 'active predators' but I've not been able to find corroborating details for that remarkable description.

The Megaloptera are often particularly large neuropteridans, reaching up to twenty centimetres in wingspan, and comprise a bit less than 400 species worldwide, mostly found in temperate regions. Larvae are aquatic, living under rocks and debris, and characterised by the presence of filamentous lateral gills on the abdomen. Adults are short-lived and feed little if at all. Male dobsonflies (of the subfamily Corydalinae) possess spectacularly large, curved mandibles of largely unknown purpose; certainly they do not seem to use them for biting.

Mantisfly Mantispa styriaca, a raptorial lacewing, copyright Gilles San Martin.

The largest of the three orders, by a considerable margin, is the Neuroptera with over 5700 known species. Needless to say, this level of species diversity is associated with a high diversity of appearances and lifestyles, too many to cover adequately here. The larvae of two families of Neuroptera, the Nevrorthidae and Sisyridae, are aquatic and there has been a long-running debate whether this aquatic habit is an ancestral feature of the order shared with the Megaloptera (Nevrorthidae larvae are generalist predators, Sisyridae are specialised feeders on freshwater sponges and bryozoans). However, recent phylogenetic studies (e.g. Vasilikopoulos et al. 2020) do not agree with earlier hypotheses that the Nevrorthidae represent the sister taxon of the remaining Neuroptera. Instead, Nevrorthidae and Sisyridae may form a clade with the Osmylidae, a family whose larvae are not aquatic but often inhabit damp stream banks. The aquatic Neuroptera probably entered the water independently of the alderflies. The current favourites for the sister clade of other neuropterans are the dustywings of the Coniopterygidae, a group of small neuropterans with reduced wing venation that have historically been difficult to place owing to their derived features.

An unidentified dustywing, Coniopterygidae, copyright Katja Schulz.

A fourth order has often been associated with the Neuropterida, the extinct Glosselytrodea. Glosselytrodeans are small insects known from the Late Permian to the Jurassic, characterised by wings bearing dense cross-veins of which the fore pair would have had a leathery appearance in life (not dissimilar in texture to the fore wings of grasshoppers and other Orthoptera). Other than the wings, the features of glosselytrodeans are poorly known: they seem to have been hypognathous (i.e. had the head directed downwards) with slender legs (Grimaldi & Engel 2005). Connections to Neuropterida are based on features of the wing venation but cannot be considered strongly supported. Other authors have regarded them as of uncertain position within the broader holometabolous clade, or even as more closely related to the Orthoptera than any Holometabola. Unless more complete remains should come to light, it seems likely that the question will remain open.


Aspöck, H., & U. Aspöck. 1991. Raphidioptera (snake-flies, camelneck-flies). In: CSIRO. The Insects of Australia: A textbook for students and research workers 2nd ed. vol. 1 pp. 521–524. Melbourne University Press.

Engel, M. S., S. L. Winterton & L. C. V. Breitkreuz. 2018. Phylogeny and evolution of Neuropterida: where have wings of lace taken us? Annual Review of Entomology 63: 531–551.

Grimaldi, D., & M. S. Engel. 2005. Evolution of the Insects. Cambridge University Press: New York.

Vasilikopoulos, A., B. Misof, K. Meusemann, D. Lieberz, T. Flouri, R. G. Beutel, O. Niehuis, T. Wappler, J. Rust, R. S. Peters, A. Donath, L. Podsiadlowski, C. Mayer, D. Bartel, A. Böhm, S. Liu, P. Kapli, C. Greve, J. E. Jepson, X. Liu, X. Zhou, H. Aspöck & U. Aspöck. 2020. An integrative phylogenomic approach to elucidate the evolutionary history and divergence times of Neuropterida (Insecta: Holometabola). BMC Evolutionary Biology 20: 64.

Mites of Southern Sediment

Water mites of the clade Hydrachnidiae are one of the few groups of arachnids that have not only adopted an aquatic lifestyle but have thrived and diversified there. Over fifty families are currently recognised within this clade, some of which can be found in almost every body of fresh water worldwide. Others, however, are notable for their restricted ranges. One of these latter examples is the Omartacaridae.

Ventral view of female Omartacarus elongatus, from Cook (1963).

Omartacaridae is a small family currently recognised as including only two genera, Omartacarus and Maharashtracarus. They have a somewhat elongated body with a soft integument, contrasting with the more globular form of many other water mites. They are also distinguished by the arrangement of the coxae (the basal segments of the legs on the underside of the body) which are clustered together with the medial edges of the anterior pairs much longer than those of the posterior pairs (Walter et al. 2009) so the third pair of coxae are triangular in shape. As far as is known, omartacarids are restricted to interstitial habitats or the hyporheic zone of sediment beneath and alongside stream beds. I am unaware of any direct observations of omartacarid behaviour but they are presumably predators like other water mites. Most of the (rather limited) attention that has been given to omartacarids has focused on discussions of their distribution. Species of Omartacarus are found in South and southern North America, as well as in Australia. Maharashtracarus species are known from India and Costa Rica. It has been presumed that this reflects an ancestral Gondwanan distribution, spreading into North America from South America as the continents joined.

The larval stage of omartacarids is, to date, unknown. Larvae of other water mites live as parasites of water-associated insects such as midges and omartacarid larvae are presumably also parasitic. But in what capacity? Do mature omartacarids emerge from their subterranean habitats at some particular time of year in search of a host for their eggs? Do they somehow manage to find a host while remaining safely sequestered underground? The secret remains to be uncovered.


Walter, D. E., E. E. Lindquist, I. M. Smith, D. R. Cook & G. W. Krantz. 2009. Order Trombidiformes. In: Krantz, G. W., & D. E. Walter (eds) A Manual of Acarology 3rd ed. pp. 233-420. Texas Tech University Press.

Arrow Poison and Arrow without Poison

Central and South America are home to a remarkable diversity of frogs, coming in nearly all the shapes and sizes a frog can possibly come in. Among this diversity, probably the most famous representatives are the arrow-poison frogs of the Dendrobatidae.

Two dendrobatid frogs of two different subfamilies: dyeing dart frog Dendrobates tinctorius (Dendrobatinae, left) and phantasmal poison frog Epipedobates tricolor (Colostethinae, right), copyright H. Krisp. Offhand, has someone been playing silly buggers with dendrobatid species names? Dendrobates auratus is green and black, not gold, and I'm sure I only see two colours on that E. tricolor.

The Dendrobatidae are themselves a diverse family, with somewhere in the area of two hundred currently recognised species. Many of these have only recently been recognised: nearly half of the currently known species have been named since 1985 (Grant et al. 2006). There are also ninety or so species in the closely related family Aromobatidae that were historically treated as dendrobatids and still may be in some sources. The Dendrobatidae are currently divided between three subfamilies: about half the species belong to the subfamily Dendrobatinae, a bit less than a quarter to the Colostethinae, and close to sixty species are placed in the genus Hyloxalus that forms its own subfamily (Grant et al. 2006).

Panama rocket frog Colostethus panamensis, copyright Brian Gratwicke.

Members of the Dendrobatidae are best known, of course, for their remarkable toxicity, associated with bright, striking warning colours. The name 'arrow-poison frog' reflects this trait as an arrow scraped across a frog's skin would pick up some of the frog's own lethality. The toxin, comprising various alkaloids, is not produced directly by the frog itself but is instead acquired through its arthropod diet. Most of the alkaloids sequestered by arrow-poison frogs come from ants (Darst et al. 2005) but other potential sources include beetles, millipedes and oribatid mites. However, not all dendrobatids are toxic and colourful. In fact, these features are largely characteristic of the Dendrobatinae only. Members of the Colostethinae and Hyloxalus are mostly cryptic in coloration and largely do not sequester alkaloids. The distinction is not an unshakeable rule: some non-dendrobatine dendrobatids are quite colourful in their own right and a handful of colostethines (members of the genus Epipedobates) are toxic, having seemingly evolved the ability to secrete alkaloids independently of the dendrobatines. Laboratory studies indicate that at least some non-toxic colostethines are able to consume alkaloid-bearing prey without ill effects, suggesting that alkaloid resistance is ancestral for the family as a whole.

Male Hyloxalus nexipus carrying tadpoles, copyright Santiago Ron.

More characteristic of dendrobatids as a whole is their breeding behaviour. As a rule, dendrobatids are more or less terrestrial, not habitually living in water, though many species are found alongside the margins of water bodies and may dive into the water to escape danger. Others will be found among leaf litter or be completely arboreal. Eggs are laid in damp terrestrial locations such as under leaves; males may deposit their sperm before or after the female deposits her eggs. Hatching tadpoles are then carried on the back of one of the parents to a suitable body of water such as a pool or stream. In members of the Dendrobatinae, tadpoles are deposited in phytotelmata, water-filled hollows in vegetation (such as in the cenre of bromeliads or holes in trees). Adelphobates castaneoticus, found in Pará in Brazil, has a habit of using the fallen husks of Brazil nuts. In some species, tadpoles are transferred one at a time; in others, groups of tadpoles will be carried en masse. In most genera, the male parent is the primary or sole transporter of tadpoles. Females of some species may also carry tadpoles; in others, a female finding an unattended cache of eggs will simply eat them. In the dendrobatine genus Oophaga, tadpole transport is the sole responsibility of the female. Following deposition, developing tadpoles of many species live on a diet of detritus. Others, particularly among the phytotelm-inhabiting species, are carnivorous, feeding on insects and other aquatic vertebrates, or even on their own siblings. In the aforementioned Oophaga, the transporting female will also lay a deposit of unfertilised eggs at the same time as she drops off the tadpoles. As well as providing food for the developing larvae, these eggs may also carry a shot of alkaloids to provide a head start in developing their defenses.

Strawberry poison-dart frogs Oophaga pumilio, two different colour morphs, copyright Pavel Kirillov.

Despite their often bright colours, many dendrobatids are poorly known due to cryptic habits and many species are only found in restricted ranges. As well as the usual threats to their survival from habitat destruction and the like, many dendrobatid species are threatened by collection for the pet trade. Their bright colours make dendrobatids popular specimens and captive individuals lose their toxicity if not provided with the prey from which alkaloids are derived. Unfortunately, about a quarter of dendrobatid species are currently recognised as endangered, many severely so. The highest diversity of endangered species is in the northern Andean region, in Venezuela, Colombia and Peru, which is also the centre of diversity for the family as a whole (Guillory et al. 2019). Urgent action may be required if we are to preserve these tiny, shiny, toxic beauties.


Darst, C. R., P. A. Menéndez-Guerrero, L. A. Coloma & D. C. Cannatella. 2005. Evolution of dietary specialization and chemical defense in poison frogs (Dendrobatidae): a comparative analysis. American Naturalist 165 (1): 56–69.

Grant, T., D. R. Frost, J. P. Caldwell, R. Gagliardo, C. F. B. Haddad, P. J. R. Kok, D. B. Means, B. P. Noonan, W. E. Schargel & W. C. Wheeler. 2006. Phylogenetic systematics of dart-poison frogs and their relatives (Amphibia: Athesphatanura: Dendrobatidae). Bulletin of the American Museum of Natural History 299: 1-262.

Guillory, W. X., M. R. Muell, K. Summers & J. L. Brown. 2019. Phylogenomic reconstruction of the Neotropical poison frogs (Dendrobatidae) and their conservation. Diversity 11: 126.