Wednesday, August 19, 2009

The Molids, part 3: Enough with Mola!

In the first two posts I've mostly discussed Mola, a clade of three genetically distinct species whose morphological and biogeographical differences have just begun to be investigated. While Mola mola was a "flagship" species which obfuscated [previously recognized] diversity, there are other molids which are effectively unknown to Western public consciousness.

The so-called sharp tail/fin mola Masturus, taken from here. The dorsal fin is of course heavily damaged and while the clavus also looks attenuated compared to those of other specimens, the rounded edge seems to suggest that it isn't damaged. Liu et al. (2009) recovered a linear relationship of total length and standard length (TL - SL = clavus length) and fairly similar proportions in males and females - the large sample of primarily immature fish may have masked differing clavus proportions in large individuals.

As the above photo illustrates, Masturus is very similar in appearance to the Mola species and can be easily differentiated by a projection of the clavus (hence the common name). Gregorova et al. (2009) found Masturus and Mola to be sister taxa with the only differences (4 characters of 57 total) relating to one less caudal vertebrae in Masturus and clavus rays terminating in bony plates exclusively in Mola. Bass et al. (2005) recovered the same topology with combined d-loop and cyt b data; d-loop data in the same study also revealed that the small sample (n = 5) of Masturus did not show clear differences across ocean basins. This seems curious since the same authors mention unpublished tracking data which showed no indication of ocean basin-scale movement or large-scale migrations in the species* (Bass et al. 2005). Also interesting was that the authors only discussed Ma. lanceolatus and didn't mention Ma. oxyuropterus; the small sample size leaves open the possibility of an additional species despite the wide geographic sampling (including the type locality for Ma. oxyuropterus). Who knows, there could be another Mola-like situation at hand here...

* Seitz et al. (2002) tagged a 1 m (TL) individual which travelled 9.7 km/day - although where it traveled was not mentioned. Contrary to prior suggestions, the fish was epi-pelagic (preferring waters <>2.7% of its time in the top 5 m of the water column and could apparently dive below 1000 m (Seitz et al. 2002). I'd assume that older individuals are more tolerant of of cold water, but due to the extreme rarity of surface behavior in Masturus it may be a while before this is confirmed.

Masturus has very recently become the target of a fishery off Eastern Taiwan and this gave Liu et al. (2009) an opportunity to examine the basic biology of the species. The population ranged in length from 42 to 192 cm (SL - the clavus was often damaged) with most individuals between 80 and 119 cm (Liu et al. 2009). All specimens over 158 cm SL were female although since only one gravid individual was recorded it appears likely that this locale is not a spawning ground for the species (Liu et al. 2009). The authors speculated that juveniles may frequent inshore nurseries (where the fisheries primarily operate) while the adults occur further offshore - they noted that specimens greatly exceeding the maxima in their study have been recorded (262.5 cm SL/313.8 cm TL - another 337 cm TL*) (Liu et al. 2009). Vertebra centra were examined for rings and working under the assumption that ring count correlated with age (it did correlate with size) this indicates that their largest female was 23 years old and longevities were estimated to be 105 and 82 years for females and males, respectively (Liu et al. 2009). The authors note that the lack of very young and old specimens can greatly affect growth estimates and of course it can't be emphasized enough that ring formation may be caused by events which do not occur annually. It seems unlikely that data pools as extensive as that of Liu et al. (2009) will be encountered elsewhere (fortunately, in a way) and it appears exceedingly unlikely that a mark-recapture study or study of captive specimens** can resolve the questions of growth and longevity in Masturus - presumably a model species can be used to answer the question that the authors brought up but couldn't answer: if the fishery will lead to a decline of stocks.

* Assuming this is a female, it would have a SL of 282 cm from the equation established in Liu et al. (2009) and 281 cm extrapolating from the largest fish with both TL/SL measurements - this seems to imply that the clavus length has the same proportion throughout life after all. A male, by the way, would have an estimated SL of 277 cm, implying that this is not a dimorphic trait.

What's especially confusing to me is that estimating the weight of the 282 cm SL individual from the equations set up by the authors yields weights of ~1000-1200 kg (the larger from a male) and scaling up from their largest individual (195 cm SL, 409) also yields ~1200 kg. This seems very strange since a 2.7 m Mola (TL?) weighs around 2.3 tonnes - how on earth can a Masturus of roughly the same linear dimensions weigh around half as much? More data on large Masturus individuals, with clearly differentiated total and standard lengths, is desirable to solve this puzzle!

** At least one Mola in an aquarium gained weight extremely quickly and possibly abnormally. If they can actually grow this quickly in the wild then perhaps the Masturus fishery won't totally wreck the population... in the near future.

It's amazing that a species for whom the discovery of a single specimen was notable several decades ago (e.g. Gudger 1935) is now being harvested by the ton - and hardly any more is known about it. While the cosmopolitan range of the species and lack of direct harvesting in most locales probably prevents it from being in any direct danger, exploiting an organism with little known about its basic biology is a dangerous game.

There's still one more molid left, the smallest, most basal, and arguably most neglected of them all!


Bass, Anna L., et al. 2005. Evolutionary divergence among lineages of the ocean sunfish family, Molidae (Tetraodontiformes). Marine Biology 148, 404-415

Gregorova, Ruzena, et al. 2009. A giant early Miocene sunfish from the North Apline Foreland basin (Austria) and its implications for molid phylogeny. Journal of Vertebrate Paleontology 29, 359–371.

Gudger, E. W. 1935. A photograph and description of Masturus lanceolatus taken at Tahiti, May, 1930. American Museum Novitates 778, 1-7. Available

Liu, Kwang-Ming, et al. 2009. Age and growth estimates of the sharptail mola, Masturus lanceolatus, in waters of eastern Taiwan. Fisheries Research 95, 154-160

Seitz, A. C., et al. 2002. Behaviour of a sharptail mola in the Gulf of Mexico. Journal of Fish Biology 60, 1597–1602.

Thursday, August 13, 2009

The Molids, part 2: Beyond Mola mola

In the previous post I argued that molids were among the strangest of fishes on the basis of peculiarities such as their loss of a caudal fin and the development of an analgous clavus from the dorsal and anal fins, non-bilaterally symmetrical flight from those dorsal and anal fins, half a trunk lateral line (related to clavus development?), a diet heavy in jellyfish despite body masses often in the hundreds of kilograms (and sometimes tonnes), and so forth. I forgot to work in the outrageous fecundity of females (300 million eggs in a 1.4 m individual) and the two larval stages which resemble pufferfish. Most of these figures pertain to Mola mola, but there are of course other species of extant molids.

It's always interesting how some clades have a "flagship" species of sorts with all the charisma and press while the majority of the diversity languishes in obscurity. Molid taxonomy has long been a mess with 54 proposed species and missing type specimens, but these days 3 "genera" and 3-5 species are typically listed (Bass et al. 2005, also citing Parente 2003, Fishbase). Bass et al. (2005) used data from the d-loop and cytochrome b of molids to establish a phylogeny of the group and were surprised that some Southern hemisphere Mola specimens were estimated to have diverged 2.8-7.5 mya from the other major clade (Bass et al. 2005). The authors resurrected the name Mola ramsayi for the divergent clade and noted that both species were recorded from South Africa (Bass et al. 2005). It appears that field workers are not able to distinguish the species (voucher specimens were lacking for Bass et al.'s study, however) which implies very similar morphology and prior workers suggested that M. ramsayi can be distinguished by more numerous fin rays and larger ossicles on the clavus (Bass et al. 2005, also citing Giglioli 1883, Fraser-Brunner 1951). Also unexpected was that the Atlantic and Indo-Pacific M. mola populations were estimated to have diverged very recently (0.05-0.32 mya) while the M. ramsayi populations in the southern portions of the same oceans apparently diverged much earlier (1.55-4.10 mya) (Bass et al. 2005). Further studies will be needed to clear up the biogeography and morphology of M. ramsayi, and things certainly haven't gotten any simpler...

Yoshita et al. (2009) examined mitochondrial D-loop data from 119 Mola specimens and found three distinct clades, two of which overlap off Japan. The group A clade (n = 20) mostly occurred off the Eastern coast of Japan (2 were from Australia) and was remarkable for being composed entirely of females (and unsexed fish) with an average total length of ~2.6 m (+/- 0.5 m) which increased with latitude - all 9 specimens from the Northeast of Japan were over 249 cm and one measured 332 cm* (Yoshita et al. 2009). The group A Mola clade also had distinctive morphological traits such as a well-developed head bump, a proportionally longer anterior portion of the body, a proportionally deeper body, 14-17 clavus fin rays, 8-15 clavus ossicles, and a clavus edge that was never wavy (Yoshita et al. 2009). It appears that the historical Fiona specimen (not the world record holder, see the footnote) displays this morphology - as opposed to this more typical aquarium specimen. The authors suggest that the infrequency, size, and possible sex bias of group A specimens caught off Japan indicates that the main population could occur near the Bonin Islands (a migration route could be present) and suggest that the clade may be primarily a Southern Hemisphere one (Yoshita et al. 2009). Alternately, it was suggested that group A could have a migration route from the Eastern or Northern parts of the Pacific to Japan since the northeastern Japanese and subtropical group A fish showed significant divergence (Yoshita et al. 2009). It is not clear if group A is synonymous with Mola ramsayi and if not, presumably either another name could be resurrected or one will be coined if none of the prior names provided an appropriate description.

* This is probably the largest reliably recorded bony fish to date. Fishbase mentions a 333 cm specimen (not the 10'2"/3.1 Fiona specimen) but I can't find the source to assess its reliability. The 332 cm female fish (caught in 2004) was no outlier as 323 cm (2004 - unsexed), 325 cm (1999 - unsexed), and 330 cm (2002 - unsexed) fish were all recently found in a small sample - all were from Japan and Group A (Yoshita et al. 2009). This fascinating article reveals that the Fiona specimen (3.1 m/2.2 tonnes) was probably never weighed since a 2.7 m Mola (from Japan, of course) weighed 2.3 tonnes - by my calculations the Fiona specimen was about 1.3 tonnes "short" and a 3.3 m Mola could weigh over, gulp, 4 tonnes.

Oh, and since Austria has a rather northerly latitude, it could be possible that the Austromola fossils represent a similar population consisting of very large animals (females?) and the actual median length could be the same as the Mola species.

Group B Mola were found to be widely distributed in the Kuroshio current and grouped with M. mola from outside Japan, including Atlantic specimens Yoshita et al. (2009). This clade was notable for having a much smaller mean size (1 m =/- 0.6 m, n = 86), a wavy clavus in larger individuals (1.9-2.7 m, n = 11), a smooth band at the base of the clavus (also present in group B), 12 rays and 8-9 ossicles on the clavus, recorded males, and no significant differences between the sexes (Yoshita et al. 2009). The morphology and genetics of this group match what was previously established for Mola mola (Yoshita et al. 2009)..

Group C Mola corresponded somewhat with the species Mola ramsayi established by Bass et al. (2005) - although one member of that clade from Australia was placed in group A by Yoshita et al. (2009) - oddly enough authors of the latter study did find a group C individual from New South Wales, Australia. The monophyly of group C was supported by a very high bootstrap by Yoshita et al. (2009) but it should be noted that the sample size was only 3 and it was found to be the sister clade to group A. While the evidence presented by the authors suggests of three Mola species, clearly group C needs many more samples in order to be convincingly demonstrated to be a distinct species. And then people need to argue over which clade gets to be M. ramsayi... there's a long way to go.

The re-recognition of multiple Mola species is going to necessitate a revision of basic biological information. Future studies will have to be careful to differentiate between populations/species as to not create chimerical data - hopefully then we'll get a clearer picture of just how these species are separated geographically and possibly behaviorally and ecologically. If large (entirely female?) group B individuals really are a minor presence in very heavily fished areas, perhaps this should be looked into as a concern for conservation.

I'm not done with molids yet - Mola is not the only one!


Bass, Anna L., et al. 2005. Evolutionary divergence among lineages of the ocean sunfish family, Molidae (Tetraodontiformes). Marine Biology 148, 404-415

Wood, Gerald. 1982. Guinness Book of Animal Facts and Feats. Third Edition.

Yoshita, Yukiko, et al. 2009. Phylogenetic relationship of two Mola sunfishes (Tetraodontiformes: Molidae) occurring around the coast of Japan, with notes on their geographical distribution and morphological characteristics. Ichthyological Research 56, 232-244

Monday, August 10, 2009

The Molids, part 1: Intro and Mola mola

Of all the variations on the vertebrate "fish" body plan, molids are among the most bizarre. Yes, I'm well aware of how strange some other "fish" are. Members of the already oddball clade Tetraodontiformes, molids have the general appearance of swimming heads (German: "Schwimmender kopf"), have the fewest vertebrae of any "fish" (16-18) and lack the following structures*: caudal bones, ribs, pelvic fins, [fin] spines, girdles, and swim bladders (Fishbase, Bass et al. 2005 - citing Tyler 1980). Their skeletons have to be seen in order to be believed and it's hard to imagine how reconstructions would look if these animals were only known from fossils. While other large aquatic vertebrates have generally similar tapering/streamlined body shapes (sea turtles excepted), molids have taken a more... attenuated approach:

Taken from here.

* Although Fishbase reports the lateral line to be absent, this was recently discovered not to be the case. Nakae and Sasaki (2006) noted that there are six cephalic and one trunk lateral lines; the latter structure has only 27 superficial neuromasts and is limited to the anterior portion of the fish. Interestingly, the same study discovered that what appears to be the caudal fin of molids (the clavus) is in fact derived from the dorsal and anal fins.

Mola mola is the largest extant actinopterygian ("bony fish") with a reported average length of 1.8 m (counting the clavus) and weight of one tonne (Wood 1987); they can reach 3.32 m and probably exceed four tonnes (see the next post for an explanation). Molids have a cartilaginous, weakly ossified, spongy skeleton and as a result were only known from fossil jaws and dermal plates until recently when three extraordinary upper Miocene (~22 mya) specimens were described from Austria (Gregorova et al. 2009). The Austromola angerhoferi individuals were estimated from Mola mola proportions to have total lengths of 1.5-1.7 m, 2.4 m, and 3.2 m - this would seem to suggest that Austromola had a much larger average size than the extant species (However - a Mola subpopulation off Japan has an even larger average size) (Gregorova et al. 2009). Potentially phenomenal size aside, Austromola bridges the gap between the Eocene (~42 mya) Eomola and Miocene to Pliocene fossils of extinct Mola and Ranzania species; Austromola is the sister taxa to Mola + Masturus which is in turn the sister clade to Ranzania (Gregorova et al. 2009). More on those non-Mola mola molids in the next post.

Much like the basking shark, recent tracking data demonstrates that molas are much more mobile than was previously anticipated. Despite the common name "oceanic sunfish", three small (~14-18 kg) specimens showed a general diel pattern (deeper at day and vice versa) and had a maximum recorded depth of 472 m - excluding one incident where the tracker went below 2000 m and presumably indicated the death of the fish (Sims et al. 2009). The diel pattern wasn't strictly adhered to as some fish stayed below 200 m for most of the day and at other times fish moved from their maximum depth to the surface in 4 hours (Sims et al. 2009). Vertical movements in Mola may be related to finding maximum prey concentrations (when the prey is on the move); other hypotheses for surfacing (not necessarily mutually exclusive) include warming, recovering from time spent in sub-oxic waters, or even to have parasites removed by sea birds (Sims et al. 2009). Prey densities also appear to be the driving force of Mola movements to high latitudes during summer and movements back are possibly due to thermal tolerance issues (Sims et al. 2009). Global warming could thus expand the seasonal thermal window at high latitudes for Mola and overfishing and eutrophication could increase jellyfish blooms - molids are some of the few large predators which have top-down control over the cnidarians (Sims et al. 2009). Exactly what sort of impact bycatch* has on molid populations is also an interesting question - this certainly seem like a species worth looking into.

* So exactly how numerous are molids? Aerial observations in the Irish and Celtic seas observed small individuals (0.5 to 0.7 m) at a density of about 1 individual per 100 square kilometers working under the assumption that they are not at the surface 3/4 of the time (Houghton et al. 2006). The authors note that the largest Mola specimen from British waters only weighed 363 kg (slightly above average?) which would either suggest that adult individuals don't bask often or are rarely present - or both. It's worth noting that molids are very commonly caught as bycatch (29-93% of all catch, in some instances), and while fish are released alive, this activity must have substantial (but as yet unknowable) impacts on the fish (Houghton et al. 2006). It takes an estimated 20 years to grow a large 3 m Mola mola in the wild, but one aquarium fish went from 26 kg to 400 kg in only 14 months.

Another Mola tracking study took a novel approach by including the leatherback turtle Dermochelys coriacea. Unexpectedly, the giant species have converged to feed on low-nutrient jellyfish prey - however Mola can feed extensively at depths around 500 m while the turtles spent time almost exclusively in the upper 200 m (Hays et al. 2009). Unlike Sims et al. (2009) the Mola specimens were closer to the reported average (1.08-1.60 m, 67-213 kg) - the turtle were fair-sized as well (~1.7 m and ~450 kg for 2 specimens) (Hays et al. 2009). It is interesting that in South Africa at least, molas stay broadly in the same location and leatherback turtles migrate extensively despite having a similar diet (Hays et al. 2009). The reason why there are multiple large jellyfish feeders (there are other molids after all...) wasn't explored by Hays et al. (2009) but this is of course an area of ecology that we are just beginning to understand. I'm going to assume that the turtles have some sort of advantage in shallower waters (capable of handling larger prey?) and that an equilibrium between the species probably varies between geographical locations (if oxygen levels at depth play a role, for instance).

But on a different note, as tracking studies imply, molids are not the planktonic weak swimmers they were long assumed to be. Watanabe and Sato (2008) used larger Mola specimens for their study (48, 59, 153 kg) and found that the tracked fish swam actively by stroking their fins and had cruising speeds roughly comparable to sturgeons, salmon, marlin, and blue sharks; larger fish swam slower to maintain the same Reynold's number (Watanabe and Sato 2008). Mola acceleration data indicated a one-stroke cycle with lift-based thrust similar to that previously reported in penguins - this is the only known instance of "wings" which are not bilaterally symmetrical (Watanabe and Sato 2008). The dorsal and anal fins are symmetrical in shape with identical muscle mass although the muscle morphology differs markedly; the aspect ratio of the fin/wings decreases possibly due to reasons of mechanical strength (Watanabe and Sato 2008). Watanabe and Sato (2008) also determined that Mola is neutrally buoyant due to a layer of gelatinous tissue (similar to deep-sea fishes); the incompressible tissue gives stable buoyancy at every depth unlike fish which rely on swim bladders for neutral buoyancy (Watanabe and Sato 2008).

More molids soon!


Hays, Graeme C., et al. 2009. Vertical niche overlap by two ocean giants with similar diets: Ocean sunfish and leatherback turtles. Journal of Experimental Marine Biology and Ecology 370, 134–143

Houghton, Jonathon D. R., et al. 2006. The ocean sunfish Mola mola: insights into distribution, abundance and behaviour in the Irish and Celtic Seas. J. Mar. Biol. Ass. U.K. 86, 1237-1243

Gregorova, Ruzena, et al. 2009. A giant early Miocene sunfish from the North Apline Foreland basin (Austria) and its implications for molid phylogeny. Journal of Vertebrate Paleontology 29, 359–371.

Nakae, Masanori and Sasaki, Kunio. 2006. Peripheral nervous system of the ocean sunfish Mola mola (Tetraodontiformes: Molidae). Ichthyological Research 53, 233-246. Available

Sims, David W., et al. 2009. Satellite tracking of the World's largest bony fish, the ocean sunfish (Mola mola L.) in the North East Atlantic. Journal of Experimental Marine Biology and Ecology370, 127–133

Watanabe, Yuuki and Sato, Katsufumi. 2008. Functional Dorsoventral Symmetry in Relation to Lift-Based Swimming in the Ocean Sunfish Mola mola. PLoS ONE. Available

Wood, Gerald. 1982. The Guinness Book of Animal Facts and Feats. Third Edition.