(Possibly) The World's Smallest Tetrapods

The subject of today's post are proturans. Proturans are rather unique little hexapods (though not, by the currently used definition, insects) that are apparently widespread despite being rarely seen (Imadaté, 1991). I have to confess that I've never seen one yet in my time in entomology - I was reminded of them because one of my co-workers briefly thought he might have found one (but alas, not to be - it was probably some form of beetle larva). The title for the post derives from the fact that, despite being hexapods, proturans are functionally quadrapedal. They lack antennae, and instead the first pair of legs is held up and in front of the body as sensory organs. Christopher Tipping has a page with some neat pictures - in particular, check out the 1907 drawing at the top of the page like something on the cover of an Edgar Rice Burroughs book.

Protura show a number of other unique features as well. They are the only insects to increase the number of abdominal segments over their life. And perhaps most notable of all, their spermatozoa are completely unlike any other hexapod (Baccetti et al., 1973). The proturan spermatozoon is non-motile, varying from a complicated twisted helical structure (Acerentulus traegardhi and Acerentomon majus) to a simple mammalian-blood-cell-like torus (Eosentemon transitorium). In those species that retain an axoneme (the flagellar 'skeleton') it shows an abnormal arrangement of microtubules. In the vast majority of eukaryotes, the axoneme has a '9 + 2' arrangement - nine pairs of microtubules around the outside and two single microtubules in the centre (this is also the same arrangement as in spirochaetes, a clade of spiral bacteria, which has lead to rather controversial suggestions that the eukaryote flagellum may be derived from symbiotic spirochaetes - a suggestion I happen to be rather skeptical of). In contrast, proturan axonemes show a whole range of arrangements - 12 + 0, 13 + 0, 14 + 0 or even 9 + 9 + 2. Such unique features have lead to suggestions that proturans may not even be related to insects (there was an article in Simonetta & Conway Morris, 1991, that I recall, but i haven't been able to find the specific reference), but as they are undoubtedly unique derived features of proturans they are completely uniformative as to outside relationships.

In terms of actual phylogenetic relationships, proturans are entognathous (that is, the mouthparts are recessed into a capsule under the head). Most authors have united them with the Collembola (springtails) in a clade called Ellipura, but most of the supposed characters of this clade reflect character losses, which are generally regarded as less trustworthy due to the higher chance of homoplasy. A couple of recent papers did not support the Ellipura grouping (Giribet et al., 2004; Luan et al., 2005), instead placing Protura with Diplura, but the former paper also found an unexpected polyphyletic arrangement of hexapods relative to crustaceans which requires further investigation.

REFERENCES

Baccetti, B., R. Dallai & B. Fratello. 1973. The spermatozoon of Arthropoda. XXII. The '12+0', '14+0' or aflagellate sperm of Protura. J. Cell Sci. 13: 321-335.

Giribet, G., G. D. Edgecombe, J. M. Carpenter, C. A. D’Haese & W. C. Wheeler. 2004. Is Ellipura monophyletic? A combined analysis of basal hexapod relationships with emphasis on the origin of insects. Organisms, Diversity and Evolution 4: 319-340.

Imadaté, G. 1991. Protura. In The Insects of Australia (CSIRO) pp. 265-268. Melbourne University Press.

Luan, Y.-X., J. M. Mallatt, R. D. Xie, Y.-M. Yang & W.-Y. Yin. 2005. The phylogenetic positions of three basal-hexapod groups (Protura, Diplura, and Collembola) based on ribosomal RNA gene sequences. Molecular Biology and Evolution 22: 1579-1592.

Taxon of the week: Marmotini

Okay, I admit, last week's entry for Taxon of the Week was a little, well, weak. Basically I didn't feel like writing on monkeys at the time. This week's entry is much better - squirrels. Specifically, the mostly North American ground squirrels - Marmotini (see here for some pictures and an article).

[As an aside, I saw Bridge to Terabithia with Jack this Saturday just been, which presented me with a bit of a biogeographic quandary. The movie looked very much like it had been filmed in New Zealand - I recognised the vegetation. Though the intended setting was not specified, I'm guessing it was somewhere in the American Midwest. A couple of scenes featured squirrels, obviously filmed in America (New Zealand animal importation laws expressly forbid importing squirrels under any circumstances). However, in one scene where the main character traps and releases an animal that has been attacking his family's greenhouse, the animal in question is a brushtailed possum - not, to my knowledge, present in North America, but very much present in New Zealand. Sometimes, I have to admit, I wish I didn't realise these things so I could just ignore it and enjoy the movie.]

The Marmotini are one of the few clades of squirrels to make it to North America, along with the flying squirrel genus Glaucomys and the genera Tamiasciurus and Sciurus*. There are five genera - the more arboreal Tamias (chipmunks) and the terrestrial Ammospermophilus, Spermophilus, Marmota (marmots) and Cynomys (prairie dogs) - though some authors divide up some of these genera, notably Tamias and Spermophilus (which is probably paraphyletic). Tamias, Spermophilus and Marmota include Eurasian members, while Ammospermophilus and Cynomys are solely North American. I haven't been able to find what are the specific morphological characters uniting this clade (reading between the lines in Callahan & Davis (1982), I suspect they're genitalic characters), but it is also recovered with molecular phylogenies (Mercer & Roth, 2003). All Marmotini hibernate during winter to some extent (A. Watts).

*There is another genus present in North America, Microsciurus, but Mercer & Roth (2003) find it to be nested within Sciurus.

Perhaps the most interesting features I've come across in a quick scan of the Marmotini are their reproductive patterns and their sociality. Marmotini produce large litters of relatively small cubs or pups or whatever a baby squirrel is called, and apparently have the smallest size at weaning relative to adult size (see here).

Marmota and Cynomys both contain species that live socially. All species of Cynomys are social and form larger colonies than Marmota, of which there are some non-social species (in line with the common name of 'prairie dog', the name Cynomys translates as 'dog-mouse'). These two genera form a clade relative to the other Marmotini, but it is debatable whether sociality has appeared multiple times within the clade or whether it has appeared once and then been lost in the non-social marmots - the nested position of the solitary Marmota monax within marmots suggests to me that the latter may be more likely (Cardini et al., 2005).

REFERENCES

Callahan, J. R., & R. Davis. 1982. Reproductive tract and evolutionary relationships of the Chinese rock squirrel, Sciurotamias davidianus. Journal of Mammalogy 63 (1): 42-47.

Cardini, A., R. S. Hoffmann & R. W. Thorington Jr. 2005. Morphological evolution in marmots (Rodentia, Sciuridae): size and shape of the dorsal and lateral surfaces of the cranium. Journal of Zoological Systematics and Evolutionary Research 43 (3): 258-268.

Mercer, J. M., & V. L. Roth. 2003. The effects of Cenozoic global change on squirrel phylogeny. Science 299: 1568-1572.

The return of Buddenbrockia

In a previous post I briefly mentioned the mysterious worm Buddenbrockia plumatellae Schröder 1910, listed by Haszprunar et al. (1991) as an "extant problematicum", and alluded to how the true identity of this critter was rather unexpectedly resolved a few years back (Okamura et al., 2002). Further work on the position of Buddenbrockia appeared this past Friday in Science and adds a further small twist to the tale (Jiménez-Guri et al., 2007).

Buddenbrockia is a parasite of freshwater bryozoans, small, sessile, colonial animals that are sometimes referred to as 'moss animals' for little apparent reason ('moss animals' happens to be the English translation of 'Bryozoa'). After Schröder first described it in 1910, he suggested two years later that it was related to nematodes due to its mesodermal muscle blocks. A relationship to trematodes (flukes) was suggested around the same time by Braem (Okamura et al., 2002).

Okamura et al. (2002) examined the ultrastructure of Buddenbrockia, and found that it possessed an inner and outer layer of cells separated by the aforementioned muscle blocks, of which there are four arranged around the body. There is no through gut. Most significantly, the outer cell layers (the mural cells) contained polar capsules. Polar capsules are rounded organelles containing a tightly coiled filament that can be ejected at great speed. The polar capsules of Buddenbrockia were very similar to those of Tetracapsula, a basal member of the Myxozoa (and also a bryozoan parasite).

In my earlier post, I referred to Myxozoa as the least animal-like of animals, and I unreservedly stand by that statement. Myxozoa are parasites and fall into two classes. The class Malacosporea contains the aforementioned Tetracapsula (and now Buddenbrockia) and are parasites of bryozoans and fish. The class Myxosporea is considerably larger and are parasites of fish and annelid worms. As an interesting aside, the Myxosporea was previously divided between two classes of superficially very distinct appearance, the fish-parasitic Myxosporea and the annelid-parasitic Actinosporea. This distinction was removed in the mid-1980s when it was shown that spores of the myxosporean Myxobolus fed to tubificid annelids developed into the actinosporean Triactinomyxon (Wolf & Markiw, 1984). The two 'classes', therefore, represent different stages in the myxosporean life cycle.

So derived are myxosporeans relative to other animals that until fairly recently they were not even recognised as animals at all, being instead classified with the parasitic protozoa (Sporozoa and Microsporidia). Myxosporeans contain very few cells, and are contained completely within the cells of the host for part of the life cycle. A connection with animals was first suggested on the basis of the presence in myxosporeans of collagen, and on the near-identical ultrastructure of the myxosporean polar capsule with the cnidarian nematocyst (stinging cell) (see here for more details).

Even after myxozoans were recognised as animals, however, their position within the animal kingdom proved very hard to establish. Obviously, the ultrastructural similarities supported a connection with the cnidarians (the parasitic cnidarian Polypodium [not to be confused with the fern genus Polypodium] was particularly suggested as a close relation). However, some molecular studies supported a connection with bilaterians, most notably the reported presence in myxozoans of bilaterian-like Hox genes.

When Okamura et al. (2002) published their ultrastructural study of Buddenbrockia, they felt that its worm-like structure supported a bilaterian relationship for myxozoans, and highlighted previous molecular studies connecting myxozoans to nematodes. Jiménez-Guri et al.'s (2007) publication, however, turns this on its head, and returns Myxozoa to a position with the Cnidaria. This was done through a Bayesian phylogenetic analysis of some 129 proteins in 47 animals (plus 13 opisthokont outgroups) in a range of higher taxa. Buddenbrockia proved to have a significant branch length (easily the longest on the tree). interestingly, parsimony analysis (which is very vulnerable to long-branch attraction) of the data set resulted in Buddenbockia associating with a clade of nematodes + platyhelminthes, the other long-branch taxa analysed. It is likely that long-branch attraction has also been the culprit for such associations in the past.

And those bilaterian-like Hox genes? Well, the authors of the current study managed to isolate the supposed myxozoan Hox genes from host species that were not even infected with myxozoans. They were unable to isolate them from myxozoan samples that had been scrupulously cleared of any host tissue. Therefore, the supposed myxozoan Hox sequences represent contamination from the hosts, and are not myxozoan at all.

REFERENCES

Haszprunar, G., R. M. Rieger & P. Schuchert. 1991. Extant "problematica" within or near the Metazoa. In The Early Evolution of Metazoa and the Significance of Problematic Taxa (A. M. Simonetta & S. Conway Morris, eds.) pp. 99-105. Cambridge University Press.

Jiménez-Guri, E., H. Philippe, B. Okamura & P. W. H. Holland. 2007. Buddenbrockia is a cnidarian worm. Science 317: 116-118.

Okamura, B., A. Curry, T. S. Wood & E. U. Canning. 2002. Ultrastructure of Buddenbrockia identifies it as a myxozoan and verifies the bilaterian origin of the Myxozoa. Parasitology 124: 215-223.

Wolf, K., & M. E. Markiw. 1984. Biology contravenes taxonomy in the Myxozoa: new discoveries show alternation of invertebrate and vertebrate hosts. Science 225: 1449-1452.

Parcelling plastids

The Table of Contents alert for a bright, shiny new issue of Protist arrived in my e-mail box today. This is a journal I can usually rely on to supply something worthy of note (often some things), and today's was no exception. Right near the beginning of the issue, we receive a new class of heterokont (Horn et al., 2007).

So many new chromist classes have appeared in recent years that they're almost becoming routine (see Filling in the Gaps for another post of mine on one, which also includes an overview of the heterokonts). Today's subject has been dubbed Synchromophyceae, and contains a single new species, Synchroma grande, scraped off marine rocks on the coast of the Canary Islands (perhaps not the most unpleasant place to do field work). Synchroma is a photosynthetic amoeboid that forms a meroplasmodium (a network of individual cells connected by branching and anastomosing [fusing together where they meet] reticulopodia). Meroplasmodial forms are extremely rare in chromists - the only other chromist that exhibits this body form is the haptophyte Reticulosphaera japonensis (Cavalier-Smith et al., 1996). The main cell body is contained in a flattened lorica, which is adpressed to the substrate and is more or less circular on a flat surface. In reproduction, one daughter cell remains sessile in the lorica, while the other is released as a migratory cell. Migratory and floating cells are fusiform, with axopodia extending from the anterior and posterior ends. Sessile cells could also spontaneously convert to migratory cells by hatching out of their lorica. At no stage in the life-cycle were flagella present.

Phylogenetic analysis of Synchroma demonstrated that it is part of the Ochrophyta, the photosynthetic heterokont clade. Both rbcL and 18S rDNA maximum likelihood trees placed Synchroma as sister to the Chrysophyceae + Synurophyceae clade, but support in both cases was relatively low. The absence of a girdle lamella in Synchroma's chloroplasts supports its exclusion from either of those two classes.

The most interesting feature of Synchroma, however, lies in the arrangement of chloroplasts in the cell. As I mentioned previously, the chromist chloroplast is believed to be derived from an endosymbiotic red alga in a secondary endosymbiosis. As a result, the chromist chloroplast is surrounded by four membranes that represent (from the inside out) the original cyanobacterium's external cell membrane, the vacuolating membrane of the primary host, the primary host's external cell membrane, and the vacuolating membrane of the secondary host. There is some debate about whether the chloroplasts of all chromist groups (and those of their putative sister group, the alveolates) derive from a single endosymbiotic event, or have been independently derived from multiple events. The majority of authors currently favour the former option, or at least that the number of events was quite few. However, Synchroma has a unique arrangement of chloroplasts in regard to surrounding membranes. Each individual chloroplast is surrounded by two membranes, and then multiple chloroplasts are clustered in packages contained by the remaining two membranes. Horn et al. suggest in passing that Synchroma may preserve an ancestral stage in the endosymbiotic process, after the loss of the eukaryotic endosymbiont's nucleus but before the separation of the endosymbiont's chloroplasts (after all, there is no reason why the eukaryotic endosymbiont would have necessarily had only one chloroplast per cell). In light of the derived position of Synchroma within the chromists, if this is indeed an ancestral state it makes the idea of a single endosymbiosis event rather unlikely, because this would require that the ancestral state was lost repeatedly in all other chromist groups. However, I feel that it is much more likely that the other explanation suggested by Horn et al. for Synchroma's unusual chloroplasts is correct - that it is a derived state resulting from an abnormal division pattern. The possibility that within chromists cryptophytes, haptophytes and heterokonts have gained their chloroplasts independently is not completely unbelievable, but the idea that multiple hterokont groups have done the same seems to be pushing it a little. Especially as all members of this undoubtedly monophyletic clade possess red algal-derived chloroplasts - if they derived them independently, why shouldn't at least some have green algal chloroplasts?

REFERENCES

Cavalier-Smith, T., M. T. E. P. Allsopp, M. M. Häuber, G. Gothe, E. E. Chao, J. A. Couch & U.-G. Maier. 1996. Chromobiote phylogeny: the enigmatic alga Reticulosphaera japonensis is an aberrant haptophyte, not a heterokont. European Journal of Phycology 31 (3): 255-263.

Horn, S., K. Ehlers, G. Fritzsch, M. C. Gil-Rodríguez, C. Wilhelm & R. Schnetter. 2007. Synchroma grande spec. nov. (Synchromophyceae class. nov., Heterokontophyta): an amoeboid marine alga with unique plastid complexes. Protist 158 (3): 277-293.

Taxon of the Week: Anthropoidea

I hardly feel that this group needs much introduction. The Anthropoidea (the name means 'man-like') are the monkeys, and everyone knows what a monkey is (Groucho Marx supposedly once claimed that the word 'duck' was the funniest in the English language, but I'm sure that the word 'monkey' mustn't be far behind). Modern-day anthropoids fall into two clades, the New World Platyrrhini (New World monkeys) and the Old World Catarrhini (Old World monkeys and apes, including humans). There is also a handful of fossil taxa that probably fall outside the crown group such as Parapithecidae and Eosimiidae. It seems pretty clear that the Anthropoidea were Old World in origin, possibly Asian (Beard & Wang, 2004). How the ancestors of the Platyrrhini managed to turn up in South America is something of a mystery - it seems most likely that they dispersed there from Africa, but how they managed to cross the Atlantic remains an open question (Schrago & Russo, 2003). However they did it, it is possible that the ancestors of the South American caviomorph rodent radiation arrived the same way at the same time (Schrago & Russo, 2003).

Anthropoids are fairly easy to recognise, or at least the modern ones are. The eyes are more vertically-positioned than in other primates, among other features. At some point, the ancestors of anthropoids also lost the ability to produce ascorbic acid (vitamin C), which is why we have to receive it from our diet (apparently the only other animals to suffer from this deficiency are guinea pigs). It has even been suggested that this loss has been a major factor in anthropoid evolution, as mutation increases in the absence of ascorbate's antioxidant effect (Challem, 1997). However, the rate of evolution of the hominoid (humans and apes) branch, at least, seems to have been reduced relative to other primates (Tëtushkin, 2003).

REFERENCES

Beard, K. C. & J. Wang. 2004. The eosimiid primates (Anthropoidea) of the Heti Formation, Yuanqu Basin, Shanxi and Henan Provinces, People's Republic of China. Journal of Human Evolution 46 (4): 401-432.

Challem, J. 1997. Did the loss of endogenous ascorbate propel the evolution of Anthropoidea and Homo sapiens? Medical Hypotheses 48: 387-392.

Schrago, C. G., & C. A. M. Russo. 2003. Timing the origin of New World monkeys. Molecular Biology and Evolution 20 (10): 1620-1625.

Tëtushkin, E. Ya. 2003. Rates of molecular evolution of primates. Genetika 39 (7): 869-887 (transl. Russian Journal of Genetics 39 (7): 721-736.

Love Hurts

People generally know distressingly little about insects. The gossip section of the weekly telly guide with the paper on Saturday mentioned a comment by Jerry Seinfeld on an upcoming movie of his featuring bees: "Other insects are just kind of crawling around. They don't have the sophistication of the bee. They have no crime, they have no drugs, they have no rape. A little rape, but it's not that bad." Leaving aside the question of the propriety of these comments, I have one word to say about their accuracy: bedbugs. Welcome to the world of Traumatic Insemination.

Male bedbugs (Cimicidae) have a sharpened intromittent organ (or if you prefer, a great spike on their knob). The usual means of entry is ignored in mating - rather, the female genital tract is only used in egg-laying (Stutt & Siva-Jothy, 2001). Instead, the male uses his sharpened organ to pierce through the female's body wall and inject semen directly into the body cavity.

Males and females do not always have the same aims in sexual reproduction. Because the limits on reproduction rates for males are relatively minimal, it is generally in the interest of males to maximise their insemination rate, in order to maximise the number of offspring they produce. Females, on the other hand, are more likely to have a maximum reproductive rate limited by the number of offspring they can safely produce in a given time period. Therefore, it is more advantageous for them to limit fertilisation to the best males to maximise the health of their offspring. As well as the obvious restrictions a female can place on fertilisation by limiting the ability of males to mate with her, the genital tract of females often has adaptations to 'test' sperm. The genital tract itself may be hostile to sperm survival (by being highly acidic, for instance). There may be structures such as a bursa copulatrix that store and/or digest sperm, limiting their access to further parts of the tract.

It is suspected that traumatic insemination evolved in males to bypass these restrictions on the part of the females. A good demonstration of this is seen in bugs of the family Nabidae, where entry into the female is still by the genital tract, but the sharpened intromittent organ pierces the wall of the bursa copulatrix (Tatarnic et al., 2006). Experimental evidence has shown that in traumatic inseminations the fertilisation advantage is held by the last males to mate with a female (Stutt & Siva-Jothy, 2001).

Nevertheless, females are not without defenses. While female bedbugs show a surprising lack of behavioural defenses against mating, they have developed an entirely novel paragenital system called the spermalege. In the common bedbug (Cimex lectularius), the external part of the spermalege is a special notch and a thickened part of the cuticle. Internally, there is a pocket filled with cells that receives the male ejaculate. Morrow & Arnqvist (2003) demonstrated that traumatic insemination through the spermalege had relatively little effect on the health of the recipient females. However, if the body wall was pierced anywhere else the female's survival rate was severely compromised. The internal part of the spermalege probably fulfils the sperm-killing role of the usual genital tract, as well as reducing the direct exposure of the female body cavity to potentially harmful ejaculate (Morrow & Arnqvist, 2003).

Different genera of bedbugs show a wide range of variation in the complexity of the paragenital system, from species entirely lacking one to species in which the system is exceedingly complex. In at least one genus, Afrocimex, a spermalege is present in both males and females (Tatarnic et al., 2006). It has been suggested that the presence of a spermalege in males defends against damage from homosexual matings - male bedbugs apparently tending to stab first and check suitability afterwards.

REFERENCES

Morrow, E. H., & G. Arnqvist. 2003. Costly traumatic insemination and a female counter-measure in bed bugs. Proceedings of the Royal Society of London Series B - Biological Sciences 270: 2377-2381.

Stutt, A. D., & M. T. Siva-Jothy. 2001. Traumatic insemination and sexual conflict in the bed bug Cimex lectularius. Proceedings of the National Academy of Sciences of the USA 98 (10): 5683-5687.

Tatarnic, N. J., G. Cassis & D. F. Hochuli. 2006. Traumatic insemination in the plant bug genus Coridromius Signoret (Heteroptera: Miridae). Biology Letters 2: 58-61.

The fall of Rafflesiales

I have to make a confession - I probably won't be covering a huge number of plants (in the sense of embryophytes) on this blog. This is, I admit, a pretty huge omission. After all, plants are a pretty significant chunk of phylospace, and pretty damn influential in shaping our environment. But compared to animals, plants are scary and complicated and I just don't know all that much about them. That said, I thought I should rectify my previous deficiencies by giving some plants the time of day. I'm not going to completely break with precedent, though - I'm looking at another group of parasites, the "Rafflesiales".

Until very recently, Rafflesiales was an order of holoparasites (i.e. they derive all their nutrition from their host, as opposed to hemiparasites such as mistletoes that still have leaves and produce some of their own nutrients) on roots of other flowering plants. Their main claim to fame lies in the genus Rafflesia, well-known as producers of the largest flowers in the world - Wikipedia gives the largest as over 100 cm in diametre and up to 10 kg in weight. I would be interested to know why they produce such ridiculously huge flowers. There just doesn't seem to be much point. When not flowering or fruiting, Rafflesiales are pretty much invisible, as they are otherwise entirely contained within the host.

I say "until very recently" because the Rafflesiales has, of recent years, fallen apart. This is not particularly surprising. Parasitic taxa of all varieties often show a great reduction in complexity, as organs related to nutrient gathering, production and processing lose their function - the host supplies all those things ready-made. Internal parasites are particularly notable in this regard. No need for a protective dermis of your own when you're safely contained within another organism's. No need for eyes, ears, nostrils - what are you going to see, hear, smell (that you would want to smell, at least)? Pretty much the only thing that the devoted endoparasite needs to do for itself is reproduce, and so many become little more than balls of gonad.*

*Seriously, no pun intended. I didn't even realise I'd written that at first.

Reduction in complexity often means loss of the characters that unite a parasitic clade with its non-parasitic sister group. And simplified characters come to resemble each other despite their different origins. Hence the previous unification of the families of 'Rafflesiales', and their sometimes-suggested connection with the Balanophorales, another group of reduced root-parasites (James Reveal seems to have phrased it best, though I'm not sure he meant quite the same thing as I do - "wherever go the Rafflesiales so go the Balanophorales").

The first molecular study that broke apart the Rafflesiales was Barkman et al. (2004). Barkman et al. looked at two genera of Rafflesiaceae (Rafflesia and Rhizanthes) as well as Mitrastema, another genus previously included in Rafflesiales. The two Rafflesiaceae clustered together, and appeared closely related to the Malpighiales, a large order of the flowering plant subclass Rosidae. In contrast, Mitrastema appeared as a member of the Ericales, in the subclass Asteridae. The remaining two families, Apodanthaceae and Cytinaceae, were added to the investigation by Nickrent et al. (2004). Cytinaceae was associated with Malvales (Rosidae). Apodanthaceae differed in its placement within Rosidae depending on which gene was examined, but it was never close to Rafflesiaceae (they may yet be related to Cytinaceae).

Nickrent et al. found that the peculiarities of 'rafflesialean' molecular evolution interfered with phylogenetic resolution. All the 'Rafflesiales' showed whacking great branch lengths, and when analysed under maximum parsimony, which is rather vulnerable to long-branch attraction (the tendency of long branches to randomly attach to each other), the Rafflesiales appeared as a near-monophlyetic clade. In one of the genes used by Nickrent et al., atp1, the Apodanthaceae clustered with Leguminosae, the host family of one of the Apodanthaceae. Because this result was in conflict with those from other genes, Nickrent et al. suggested that it represented lateral gene transfer from the host to the parasite. The occurrence of lateral (aka horizontal) gene transfer in eukaryotes is a subject of much debate, but it has been demonstrated to occur at least occasionally in flowering plants (Bergthorsson et al., 2003).

Most recently, Davis et al. (2007) investigated the specific position of Rafflesiaceae proper within Malpighiales, and found it to be actually nested within the Euphorbiaceae. This is a decidedly unexpected result, as Euphorbiaceae as a whole produce relatively minute flowers. Davis et al.'s result was unlikely to be simply long-branch attraction, because the other cluster on their tree to show accelerated evolutionary results was not Euphorbiaceae but the branch including Clusiaceae and Podostemaceae (the latter are aquatic plants that also show extreme reduction in complexity, to the point where they barely resemble flowering plants at all)*. I'm waiting in anticipation for Davis et al.'s results to be tested further.

*In recent years, there has been an increasingly distressing trend for articles to appear in high-impact journals with only the briefest of summaries of results in the actual printed journal, with the greater proportion of hard data, methods, etc. buried in online supplementary material (where it generally becomes inacessible after a couple of years). Davis et al. (2007) is actually one of the most extreme examples of this, with the printed article only a single page, with the supplementary info 424 pages long (fairness does compel me to admit that 415 pages of that are sequence alignments).

REFERENCES

Barkman, T. J., S.-H. Lim, K. Mat Salleh & J. Nais. 2004. Mitochondrial DNA sequences reveal the photosynthetic relatives of Rafflesia, the world's largest flower. Proceedings of the National Academy of Sciences of the USA 101 (3): 787-792.

Bergthorsson, U., K. L. Adams, B. Thomason & J. D. Palmer. 2003. Widespread horizontal transfer of mitochondrial genes in flowering plants. Nature 424: 197-201.

Davis, C. C., M. Latvis, D. L. Nickrent, K. J. Wurdack & D. A. Baum. 2007. Floral gigantism in Rafflesiaceae. Science 315: 1812.

Nickrent, D. L., A. Blarer, Y.-L. Qiu, R. Vidal-Russell & F. E. Anderson. 2004. Phylogenetic inference in Rafflesiales: the influence of rate heterogeneity and horizontal gene transfer. BMC Evolutionary Biology 4: 40. (online here).