Field of Science

Taxon of the Week: Protoperidinium grande

I can't really award anyone the prize for identifying yesterday's image; identifying it as a "dinoflagellate" doesn't really cut the mustard considering how many thousands of dinoflagellate species are known.

Protoperidinium grande. From Steidinger & Williams (1970).

Many references describe dinoflagellates as photosynthetic; this is wrong, in the same way as describing mosquitoes as feeding on blood is wrong. In terms of number of species, there are probably more non-photosynthetic than photosynthetic dinoflagellates. Protoperidinium is a genus of more than 200 species of mostly non-photosynthetic marine dinoflagellates, many of which possess a single apical horn and two antapical horns as seen in the photo above. Features distinguishing P. grande include the reticulate theca and the compressed cingular area (the cingulum is the groove around the midline; one of the dinoflagellate's two flagella wraps around the cingulum). Unlike some other Protoperidinium species, P. grande does not produce resting cysts; as a result, it is found only in warmer waters around the world. As non-photosynthetic heterotrophs, Protoperidinium species obtain their nutrition by feeding on other micro-organisms such as diatoms, cyanobacteria or other dinoflagellates. Rather than directly engulfing their prey in the way of an amoeba, Protoperidinium extend a large pseudopodial extension called a pallium from their theca's antapical pole to envelop it. The food organism is digested by the pallium, which is then withdrawn back into the dinoflagellate theca carrying a load of nutrients with it. This feeding behaviour was first 'discovered' in the late 1990s, but ironically it had actually been illustrated as long ago as 1895 with later researchers failing to recognise earlier records for what they were (Jacobson 1999*).

*Jacobson's comment on this re-discovery are worth repeating: "the brilliant, detailed observations of Kofoid and Swezy, Schütt, and Biecheler remain a humbling reminder to those of us working in a highly capitalized, high-tech environment that important work can arise from a simple light microscope, coupled with patience, luck and the appropriate search image".

Protoperidinium depressum feeding on diatoms. Figure from Jacobson (1999).

Until relatively recently, Protoperidinium species were included in the genus Peridinium along with a number of freshwater dinoflagellates. The taxonomy of Recent dinoflagellates* has traditionally been dominated by a small number of what might be termed 'super-genera' of hundreds of species that between them encompass the great majority of living taxa. It is probably not surprising that phylogenetic analyses have suggested that many of these super-genera are polyphyletic, but most of those analyses have tended to return very poorly supported results and attempts to subdivide the super-genera have not been entirely successful. The division of Peridinium is one of the more successful examples, based on ecology (Peridinium sensu stricto is freshwater, Protoperidinium is marine), fine details of the arrangement of plates in the theca (Peridinium has five plates around the cingulum; Protoperidinium has four) and the features of cysts produced in some species (Dale, 1978). Molecular analyses have supported the monophyly of Protoperidinium (Yamaguchi & Horiguchi, 2005). The division on ecological grounds has been a common pattern in studies on protists; molecular analyses of a number of other micro-eukaryotic groups such as myxozoans have also produced results that contradict traditional morphological classifications but correlate strongly with ecological features.

*The taxonomy of fossil dinoflagellates is an entirely separate pot of evil.


Dale, B. 1978. Acritarchous cysts of Peridinium faeroense Paulsen: implications for dinoflagellate systematics. Palynology 2: 187-193.

Jacobson, D. M. 1999. A brief history of dinoflagellate feeding research. Journal of Eukaryotic Microbiology 46 (4): 376-381.

Steidinger, K. A., & J. Williams. 1970. Dinoflagellates. Memoirs of the Hourglass Cruises 2: 1-251.

Yamaguchi, A., & T. Horiguchi. 2005. Molecular phylogenetic study of the heterotrophic dinoflagellate genus Protoperidinium (Dinophyceae) inferred from small subunit rRNA gene sequences. Phycological Research 53 (1): 30-42.

Name the Bug # 22

Another ID challenge as a preview for tomorrow's post. I know at least one of my regular readers will be able to get this easily if she sees it, so from her (she knows who she is) I'm going to require a species-level identification or close to. Everyone else, see what you can do:

The figure is in ventral view and the specimen is 130 µm wide. Attribution to follow.

Update: Identity now available here. Figure from Steidinger & Williams (1970).

Name the Bug: Psectra diptera

The points for this one go to Kai, who successfully identified this as a flightless hemerobiid lacewing:

Psectra diptera, from MacLachlan (1868).

Psectra diptera is found in both Europe and North America but is apparently nowhere very common (a live photo can be seen here). Gunnar pretty much gave the reasons for identifying this insect as a lacewing in his comment: the high density of veins (particularly cross-veins) in the wings and the long, filamentous antennae (but lost out on the marks by assuming that it must be a fossil).

Of course, lacewings normally have four wings, not two. Many basic references on insects will indicate that the only group of insects with a single pair of developed wings* is the Diptera, the true flies, but diptery is also found among Strepsiptera, mayflies, scale insects, lacewings, Psocodea and snowfleas** (and probably a couple of others that I've forgotten). Different processes seem to have resulted in wing loss in different lineages—among lacewings and psocops, for instance, it is associated with loss of flying ability, but Diptera and mayflies are capable of aerobatics that few other insects can rival.

*Technically, Diptera do still have their hindwings: they've been altered into small knobbed rods called halteres. Hindwing halteres are unique to flies among living insects. Strepsiptera have the forewings altered into halteres; scale insects have knobless filaments called hamulohalteres; and hindwing halteres were present in the Cretaceous lacewing Mantispidiptera.

**Though the appendages of male snowfleas are just so freakishly bizarre (as described in the linked post) that I rather feel calling them wings is a little like calling a whale's flippers 'legs'. It may be correct in an evolutionary sense but it's a little ridiculous by any other measure.

As noted by the commenter 'Reprobus', Strepsiptera are a little different from the other entries in the list in that they have lost the forewings instead of the hindwings. Relatively few groups of insects have larger hindwings than forewings: Orthoptera (crickets and grasshoppers), stick insects, earwigs, some beetles. All of these have the forewings protectively hardened to some degree, and in earwigs and beetles the forewings have become entirely hardened into protective cases called elytra. The evolutionary affinities of Strepsiptera are decidedly contentious (see the post linked to above) but some authors have suggested a relationship to beetles (and one recent molecular study has provided further support for this relationship: Longhorn et al., 2010). If so, strepsipterans may have been derived from ancestors that also had hardened forewings that were no longer functional when flying, potentially explaining why that was the pair that was lost.


Longhorn, S. J., H. W. Pohl & A. P. Vogler. 2010. Ribosomal protein genes of holometabolan insects reject the Halteria, instead revealing a close affinity of Strepsiptera with Coleoptera. Molecular Phylogenetics and Evolution 55 (3): 846-859.

MacLachlan, R. 1868. A monograph of the British Neuroptera-Planipennia. Transactions of the Royal Entomological Society of London 1868: 145-224.

Name the Bug # 21

Yes, the ID challenge today is an actual insect, for once. And, as a clue to identity, consider first the point that it only has a single pair of wings, rather than the more usual complement of two.

Attribution to follow.

Update: Identity now available here. Figure from MacLachlan (1868).

Hello to my Chinese Readers

A little over a week ago, I got an e-mail from Paul Selden saying that he was on a trip to China and had discovered that Catalogue of Organisms is seemingly on the list of sites blocked by the Chinese filtering system. Is mediocre rambling on biology that much of a threat to the Communist Party?

Still, with that information in mind, I'd like to extend a warm welcome to any readers from China. And say how privileged I feel that you have dedicated your hacking skills to reading my posts, instead of just looking up porn.

If you were looking for porn and just came here by accident, no need for disappointment. Enjoy this completely gratuitous photo of a penis:

Left lateral view of penis of Megalopsalis linnaei, from Taylor 2008.

Some Like It Cold (Taxon of the Week: Saccogynidium vasculosum)

I haven't introduced the Taxon of the Week post with a Name the Bug challenge this week because (a) even I'm not evil enough to make you try and identify liverworts, and (b) I haven't been able to find any illustrations of the specific liverwort concerned. The figures below from Gao et al. (2001) show other species in the same genus from China:

Leafy liverworts are small plants that are superficially similar in appearance to mosses. Like mosses, they grow in moist localities and lack well-developed supporting vascular tissue. Leafy liverworts can often be distinguished from mosses by having a different arrangement of leaves (liverwort leaves often grow in lateral rows, moss leaves in spirals), lacking a median vein in the leaf and potentially having teeth or lobes on the edge of the leaf. Liverworts also have different reproductive structures from mosses; instead of opening with a cap, liverwort spore capsules usually split down the sides.

Saccogynidium vasculosum is a species of liverwort restricted to the Falkland Islands and the very southernmost part of South America (Engel, 1990; Frey & Schaumann, 2002). Earlier authors referred to it as Lophocolea vasculosa but this was due to confusion with a different species, L. elata, from which it can be distinguished by the presence of small papillae (bumps) covering the leaves, a feature of the genus Saccogynidium (Engel, 1978). Saccogynidium is also distinguished from related genera by producing the female reproductive organs inside a fleshy protective covering called a marsupium (one is shown in the lower part of the figure above). Saccogynidium vasculosum is distinguished from other species in the genus by having finer papillae on the leaves, and having the tips of the leaves narrowly rounded rather than two-pointed.

Whar's really notable about Saccogynidium is its distribution (Schuster, 1972). As well as S. vasculosum, the Falkland Islands are home to S. australe, a species also found in New Zealand. Other species are found in Tasmania and south-east Asia. Interesting questions could be asked whether the current distribution of Saccogynidium is due to Gondwanan ancestry (in which case the disjoint distribution of S. australe might argue for incredibly slow rates of evolution) or to more recent dispersal, something some authors seem to have dismissed out of hand.


Engel, J. J. 1978. A taxonomic and phytogeographic study of Brunswick Peninsula (Strait of Magellan) Hepaticae and Anthocerotae. Fieldiana: Botany 41.

Engel, J. J. 1990. Falkland Islands (Islas Malvinas) Hepaticae and Anthocerotophyta: a taxonomic and phytogeographic study. Fieldiana: Botany, new series 25.

Frey, W., & F. Schaumann. 2002. Records of rare southern South American bryophytes. Studies in austral temperate rain forest bryophytes 18. Nova Hedwigia 74 (3-4): 533-543.

Gao, C., T. Cao & M.-J. Lai. 2001. The genus Saccogynidium (Geocalycaceae, Hepaticae) in China. Bryologist 104 (1): 126-129.

Schuster, R. M. 1972. Continental movements, "Wallace's Line" and Indomalayan-Australasian dispersal of land plants: some eclectic concepts. Botanical Review 38 (1): 3-86.

The Sad, Sad Story of Physeter

Original drawing of the sperm whale stranded near Berkhey, the Netherlands, in 1598. Reproduction from Husson & Holthuis (1974).

In yesterday's post on the sperm whales, I alluded to the long and reprehensible debate over the name of the great sperm whale Physeter macrocephalus. Reprehensible because for at least the last hundred years there has been absolutely no disagreement over the nature of the animal concerned; the conflict has purely been concerned with what to call it.

When Linnaeus discussed the genus Physeter in the 1758 Systema Naturae, he referred to four species: P. macrocephalus, P. catodon, P. tursio and P. microps. Most authors now treat these names as synonyms of the great sperm whale*. Normally, when two or more names are available for the one species, the oldest name automatically becomes the correct one. However, because Linnaeus' 1758 publication is the official starting point for zoological nomenclature, none of these names count as the oldest. In such a case, the general rule is that the first person to treat the names as synonymous and pick one of them to be the correct name establishes which has priority (the principle of the First Reviser).

*Physeter tursio and P. microps were both described as having high dorsal fins, something the great sperm whale completely lacks, leading to considerable confusion over the identity of the animals concerned. Modern authors tend to assume they were based on distorted or mistaken accounts of ordinary sperm whales; this is not really a satisfactory explanation, but the true identity will probably never be establishable (killer or pilot whales seem not entirely unlikely to me), and there would be little to be gained from trying.

During the 19th Century, most authors knew the great sperm whale as Physeter macrocephalus while the name P. catodon was less often referred to (and sometimes thought to refer to something like the beluga or pilot whale). It wasn't until the beginning of the 20th Century that Oldfield Thomas (1911) asserted the synonymy of the species and selected P. catodon as the correct name. However, in 1938 Hilbrand Boschma noted that Murray had treated the names as synonymous in 1866 and selected P. macrocephalus, pre-dating Thomas' selection. This was countered in 1966 by Philip Hershkovitz who claimed that Murray's selection was invalid.

The most detailed discussion of the matter was by Husson & Holthuis (1974) who discussed each of the records cited by Linnaeus for the names Physeter catodon and P. macrocephalus, selecting a lectotype for the former and a neotype for the latter that confirmed both as sperm whales. They also established that Blasius had treated the names as synonyms in 1857 and selected P. macrocephalus as the valid name, meaning that P. macrocephalus had priority even if Murray was disqualified as an authority.

However, the validity of Physeter catodon was again championed by Schevill (1986) on the basis that P. macrocephalus was supposedly invalid from the get-go. Linnaeus had distinguished the two species on the basis that P. macrocephalus supposedly had its blowhole on its neck while P. catodon had it at the front of the head; the correct position in the sperm whale is, of course, the latter. Schevill claimed that Husson & Holthuis' examination of the earlier records to correct Linnaeus' description was invalid as the concept of type specimens did not exist in Linnaeus' time, making the printed description the only judge of the species identity. Because the description of P. macrocephalus did not agree with a real sperm whale, it could not be used as the valid name.

As pointed out by Holthuis (1987), Schevill's latter argument was simply wrong. If the original author did not explicitly nominate a type specimen for a new species, then all specimens considered in the original description automatically become the type series*. To claim that the concept of types is inapplicable to Linnaeus is to ignore a fundamental aspect of the nature of the Systema Naturae, which did not spring ex nihilo but was in many places an index to the work of earlier naturalists, tying their descriptions into Linnaeus' new nomenclatural system. In the case of the sperm whale, Linnaeus was mislead by the faulty descriptions provided by others (Linnaeus himself had never seen a sperm whale). Examination of these earlier records allows the error to be recognised. Husson & Holthuis (1974) chose as lectotype of P. macrocephalus a specimen stranded in the Netherlands in 1598; while the specimen has not been preserved anywhere, illustrations of it leave no doubt that it was a sperm whale.

*Though it is true that the type specimen did not exist as a formal concept in 1758, it was not long afterwards that naturalists were finding it useful to examine earlier authors' specimens to determine their intention. Exactly when the type concept became formalised, I'm not sure.

So, in summary, both P. catodon and P. macrocephalus are available names for the great sperm whale; Blasius as First Reviser established the priority of the latter in 1857. The correct name for the great sperm whale is therefore Physeter macrocephalus.


Holthuis, L. B. 1987. The scientific name of the sperm whale. Marine Mammal Science 3 (1): 87-88 (reply by W. E. Schevill, pp. 89-90).

Husson, A. M., & L. B. Holthuis. 1974. Physeter macrocephalus Linnaeus, 1758, the valid name for the sperm whale. Zoologische Mededelingen 48 (19): 205-217, pl. 1-3.

Schevill, W. E. 1986. The International Code of Zoological Nomenclature and a paradigm: the name Physeter catodon Linnaeus 1758. Marine Mammal Science 2 (2): 153-157.

More than Just Moby (Taxon of the Week: Physeteridae)

Not surprisingly, it did not take readers long to reach a consensus about the identity of yesterday's ID challenge: the lower jaw of Kogia breviceps, the pygmy sperm whale. Kogia is distinguished from other toothed whales by its relatively small number of noticeably slender teeth; K. breviceps is usually distinguishable from the other species in the genus, the dwarf sperm whale K. sima, by having 12-16 pairs of teeth in the lower jaw (K. sima usually has 8-11).

Dwarf sperm whale, Kogia sima. The two Kogia species are externally very similar; indeed, Watson (1981) noted that their status as separate species had not yet gained universal acceptance. Photo by Robert Pitman.

The family Physeteridae as used in this post includes just three living species, the great sperm whale Physeter macrocephalus* and the two Kogia species (alternatively, many authors refer to this groups as the superfamily Physeteroidea, separating the living genera and their respective stem groups between families Physeteridae and Kogiidae). Though superficially distinct in appearance, the two genera share a number of unusual characteristics including enamel-less teeth (that are restricted to the lower jaw) and an externally squared head with anteriorly placed blowhole. The distinctive profile is due to the spermaceti organ, a pair of gigantic oil-filled sacs that fill the inside of the head. The upper sac, the spermaceti, contains a less dense oil than the lower sac, the junk. Just exactly what the spermaceti organ does is still uncertain; proposed uses (which are not all mutually exclusive) include vocalisation, echolocation, use as a battering ram (Carrier et al., 2002) and for buoyancy control (with the whale controlling the temperature of the organ by controlling the bloodflow to it and hence controlling the specific gravity of the contained oil) (Clarke, 1978). All living species feed primarily on squid though the Kogia species eat more fish than does P. macrocephalus**.

*The scientific name of the sperm whale is renowned for being one of the most prolonged, contentious and utterly pointless conflicts in zoological nomenclature. The needless complexity of this argument is such that I'm going to farm it out to a separate post rather than try to stuff it into this one.

**I've come across a number of references to the sperm whale as the largest ever carnivore. This isn't actually true. Even if one excludes the blue whale Balaenoptera musculus as a 'carnivore' due to its diet of planktic crustaceans (still technically animals), P. macrocephalus is just edged out of the top spot by the fin whale Balaenoptera physalus which, in at least some parts of its range, feeds primarily on fish (Watson, 1981).

The great sperm whale, Physeter macrocephalus. Source of image uncertain: I got it via Google Images from here, but the actual link appears to be broken.

However, as is not uncommon in the world of biology, the most familiar members of this family are far from being the most typical. The fossil record of Physeteridae sensu lato extends back to the Oligocene; for most of that time, fossil sperm whales had teeth far more intimidating than those possessed by any living species, indicating diets potentially more rapacious* (take a look at the gnashers in the specimen pictured at the top of a Tet Zoo post on the subject). Just this year, of course, we saw the publication of 'Livyatan', the largest of these killer sperm whales, with a skull some 3 m in length (Lambert et al., 2010; unfortunately, the original name bestowed on this animal, Leviathan, has turned out to be preoccupied, admittedly under pretty goofy circumstances). Bianucci & Landini (2006) suggested that the raptorial sperm whales may have been edged out by the evolution of large predatory delphinids (the lineage including the modern killer whale Orcinus orca) during the Pliocene, leaving only the squid-eaters behind.

*Interestingly, while both genera of living sperm whales lack teeth in their upper jaws, the loss of upper teeth seems to have happened independently; the stem lineages for both genera include taxa with teeth in both upper and lower jaws (Lambert et al., 2010).

Rather than the usual reconstruction of 'Leviathan' melvillei, I thought I'd show you this one by Hodari Nundu. The shark is supposed to be a juvenile.


Bianucci, G., & W. Landini. 2006. Killer sperm whale: a new basal physeteroid (Mammalia, Cetacea) from the Late Miocene of Italy. Zoological Journal of the Linnean Society 148 (1): 103-131.

Carrier, D. R., S. M. Deban & J. Otterstrom. 2002. The face that sank the Essex: potential function of the spermaceti organ in aggression. Journal of Experimental Biology 205: 1755-1763.

Clarke, M. R. 1978. Buoyancy control as a function of the spermaceti organ in the sperm whale. Journal of the Marine Biological Association of the United Kingdom 58: 27-71.

Lambert, O., G. Bianucci, K. Post, C. de Muizon, R. Salas-Gismondi, M. Urbina & J. Reumer. 2010. The giant bite of a new raptorial sperm whale from the Miocene epoch of Peru. Nature 466: 105-108.

Watson, L. 1981. Sea Guide to Whales of the World. Hutchinson.

Name the Bug # 20

Tomorrow's Taxon of the Week post will be related to the owner of this impressive array:

Any takers?

Update: Identity now available here. Photo from here.

Ammonites of the Arctic (Taxon of the Week: Arctocephalitinae)

Ammonites are one of the classic animal groups of the Mesozoic. These coil-shelled cephalopods are guaranteed a mention in almost every popular book alluding to that time period. But what is often glossed over in popular accounts is that ammonites were an extremely speciose group, making them one of the best-studied groups in understanding marine fossil diversity.

Specimens of the arctocephalitine ammonite Arcticoceras harlandi. This species probably reached a diameter of around 10 cm though most preserved specimens are smaller as the large body chamber tends to break apart before preservation. Figures 5 and 6 show a microconch (see below). Figure from Rawson (1982).

Diagram showing the internal septa of a mature whorl from Arcticoceras harlandi. Figure from Poulton (1987).

Ammonite identification is, by all accounts, a tricky beast. Donovan et al. (1980) admitted that "Students tell us in their essays that one of the desirable attributes of a zonal fossil is that it should be easily recognizable. Most ammonites are not". Ammonites as a whole are readily distinguished from other shelled cephalopods by the ridiculously complex sutures separating chambers. However, lineages of ammonites in different periods and times often converged with each other in their morphology and successful identification often requires, in addition to simple morphology, consideration of such matters as geographical provenance and the nature of forms found in contiguous strata. And quite frankly, I'll be buggered if I've got the intellect to distinguish most of them.

That carping aside, the Arctocephalitinae were a subfamily of ammonites restricted to the region of the modern Arctic Ocean during the middle part of the Jurassic. Arctocephalitines are represented by an extensive fossil record found in localities such as Greenland, northern Canada and Siberia which have allowed a reasonable degree of success in tracing their lineages. The Arctocephalitinae are the basal radiation of the family Cardioceratidae, arising from early Sphaeroceratidae during the latter half of the Bajocian epoch; the subfamily Cadoceratinae was derived from within the Arctocephalitinae during the succeeding Bathonian and would itself give rise in turn to the Cardioceratinae (Donovan et al., 1980; ammonite researchers have so far been unimpressed by arguments for strict monophyly as a guiding principle in classification). The cadoceratines would outdo their arctocephalitine forebears by spreading beyond the Boreal region.

Specimen of 'Costacadoceras'; the asterisk indicates the start of the body chamber. This 'genus' includes the microconches of Arctocephalitinae. Microconches were much smaller, morphologically distinct forms of ammonite that were found alongside the usually more abundant and more characteristic larger forms (macroconches). It is now universally accepted that microconches and macroconches represent distinct sexes of a single species (with, by analogy to modern cephalopods, microconches probably being male and macroconches female) but matching a particular microconch with a particular macroconch is often not possible. Figure from Mitta (2005).

During the period of the earliest two genera of Arctocephalitinae, Cranocephalites and its descendant Arctocephalites, the subfamily had the Arctic to itself; no other ammonite families had reached the largely isolated ocean (Navarro et al., 2005). The arctocephalitines were largely laterally compressed with deep angular whorls (discocones). Things changed with the arrival of another family, the similarly discoconic Kosmoceratidae, in the Arctic Basin around the time of the origin of the third main arctocephalitine genus, Arcticoceras. The arrival of the kosmoceratids seems to have provided a competitive impetus to arctocephalitine evolution: the overall disparity in the family decreased and they were pushed out of the discocone niche. Instead, the succeeding cadoceratines were initially cadicones with broad shallow whorls though some cadoceratines returned to a discocone form after leaving the Arctic Basin.


Donovan, D. T., J. H. Callomon & M. K. Howarth. 1980. Classification of the Jurassic Ammonitina. In: House, M. R., & J. R. Senior (eds) The Ammonoidea pp. 101-155. Academic Press: London & New York.

Mitta, V. V. 2005. Late Bathonian Cardioceratidae (Ammonoidea) from the middle reaches of the Volga River. Paleontological Journal 39 (Suppl. 5): S629-S644.

Navarro, N., P. Naige & D. Marchand. 2005. Faunal invasions as a source of morphological constraints and innovations? The diversification of the early Cardioceratidae (Ammonoidea; Middle Jurassic). Paleobiology 31 (1): 98-116.

Poulton, T. P. 1987. Zonation and correlation of Middle Boreal Bathonian to Lower Callovian (Jurassic) ammonites, Salmon Cache Canyon, Porcupine River, northern Yukon. Geological Survey of Canada Bulletin 358: 1-155.

Rawson, P. F. 1982. New Arctocephalitinae (Ammonoidea) from the Middle Jurassic of Kong Karls Land, Svalbard. Geological Magazine 119: 95-100.

Name the Bug # 18

Here is an assortment of figures of various specimens, all belonging to a species pertaining to tomorrow's Taxon of the Week (attribution to follow):

In the lower figure the whorl height is 3.2 cm. Because the general kind of organism involved is fairly obvious, I'm going to ask for a reasonable amount of specificity to count as a successful ID. But to give you a hand with being specific, I'll at least tell you that the age of the specimens is lower Upper Bathonian (during the Jurassic).

Update: Identity now available here. Upper figure from Rawson (1982); lower figure from Poulton (1987).