“If we do not ever take time, how can we ever have time?” said The Merovingian. Everything we do is a trade-off between what we want to do, and what we have to do. I am lucky enough to have been born in a time and place of peace and can devote time to publishing scientific research. As scientists, we are always told, “Publish or Perish.” This mantra gets so deeply rooted, that what we have to do becomes what we want to do. We perish anyway, so why not publish and acquire academic immortality.

Some of my first research papers (see here and here), were based on placoderm fishes from Middle Devonian rocks in Manitoba, published purely as a hobby. For my PhD, I worked on acanthodians and early chondrichthyans from a spectacular site known as MOTH (an acronym for “Man On The Hill”). At the MOTH locality, there is a rock pile that looks like someone sitting on the toilet – MOTH is a polite reference by anyone’s standards. The fishes from MOTH are amazingly well preserved and come from the earliest Devonian Period. By then the major groups of jawed fishes had evolved.

For many years, acanthodians were thought to be a cohesive group based on their fin spine complement, their pork-chop shaped bony shoulder girdle, and their distinctive scales. Their scales add new layers directly over top of older layers and as a result, the scale crown looks layered like an onion in cross section. It is unlikely that scales of this exact style evolved more than once. For this reason, I think that most acanthodians still hold together as a group. However others, using cladistic analysis suggested that some acanthodians are allied to osteichthyans (bony fishes) and others to chondrichthyans. If true, acanthodians are not a natural related group of fishes (in technical terms, we’d call them polyphyletic as presently classified).

Until around 1997, the earliest chondrichthyans were thought have complex scales with unique lateral growth patterns (or had simple placoid scales), had cartilaginous shoulder girdles, and only had spines preceding their dorsal fins (if they had fin spines at all). These features held true for cartilaginous fishes from the Middle Devonian to present day. Since 1997, we have published descriptions (and re-descriptions) of several fishes from MOTH which have scales showing a chondrichthyan growth pattern, and a fin spine complement like that of many acanthodians. These fishes drew instant attention.

Since recent cladistic analyses suggest that a subset of acanthodians (ischnacanthiforms and acanthodiforms) were closely related to bony fishes, and the rest grouped with chondrichthyans, I’d expect the respective groups to share more common features, perhaps becoming indistinguishable, the farther back you look in time. Instead, if you back-track in the fossil record, Early Devonian acanthodiforms like Promesacanthus eppleri, Mesacanthus mitchelli, and Melanoacanthus minutus share features of the acanthodians thought to be allied to chondrichthyans (heavy lobed scales on the head, and a pair spines along the abdomen just ahead of the pelvic spines). Moreover, the earliest known articulated ischnacanthiform lacks the plate-like bones over the head and gills, enlarged scute like scales along the dorsal midline and along the belly, large complex bony pectoral girdles, and elongate rectangular scales of the earliest bony fish. It doesn’t matter how you code individual characters, and whether you use computers and statistics to create a facade of high-tech respectability, the science still hinges on basic observations of similarity. It doesn’t matter if you look at overall similarity (Phenetics) or dissect similarity on a character by character basis (Cladistics), the earliest bony fishes and early acanthodians do not look alike. They are not closely related. Instead, I see plenty of overlap between acanthodians and the spiny chondrichthyans from MOTH.

Uraniacanthus probaton  has heavy scales for armour, spines along the abdomen, and spines preceding each fin forward of the tail.

One of the earliest-known acanthodiforms (Promesacanthus eppleri) has a pair of abdominal spines, heavy scales on its head, and unlike bony fishes and their later relatives.

Kathemacanthus rosulentus  originally described as an acanthodian based on its fin spine complement, has scales thought to show chondrichthyan growth pattern.

Lupopsyrus pygmaeus plus Lupopsyroides macracanthus and Obtusacanthus corroconis have simple shark-like scales and many fin spines. The former is considered to be an acanthodian, the latter two are chondrichthyans although few preserved characters separate them.

Essentially we have MOTH fishes which were thought to be acanthodians but have simple scales (like the placoid scales of sharks) as well as typical acanthodians like Uraniacanthus and ischnacanthids. We also have fishes like Kathemacanthus rosulentus and Seretolepis elegans with complex chondrichthyan scales, but have fin spines as you’d expect on acanthodians. Needless to say I was thrilled when an early shark (Doliodus problematicus) was found to have a pectoral fin spine (Miller et al. 2003. Nature 425: 501–504), and the shark Antarctilamna prisca also was found to have a pair of spines in a pectoral position (right behind the gill chamber). As far as I was concerned, the discovery of early sharks with paired fin spines was only a matter of time.

Here’s the biological reality that influences my thoughts: species evolve from pre-existing species. Shortly after species diverge, parent and daughter species coexist, could be almost indistinguishable (to us), and could remain that way for a long time: look at our two species of Tailed Frogs. The two Tailed Frog species are thought to have diverged over 2 million years ago but appear almost indistinguishable based on external anatomy. Sometimes parent and daughter species change more rapidly – it depends on local selective pressures. Either way, nature could easily generate a cloud of closely related species over a few million years which could persist through time, and be preserved together in the fossil record. Obvious differences within and between species-groups evolve only with increasing time – especially when we are discussing complex skeletal structures which would preserve in the fossil record.

The same applies to the origins of major groups of organisms. The process is gradual, and the founder of a new group must evolve from within pre-existing species-groups. We define larger species-groups (families, orders) after intermediate branches are pruned from the tree of life. While it’s easy to define groups from our top-down perspective, if you existed only in the Early Devonian and knew nothing of the future, I think you’d be hard-pressed to select the branches that would segregate chondrichthyans and acanthodians. In contrast, branches with bony fishes and placoderms would be distinct, like on a well-tended bonsai. Even by the Silurian their unique skull bone patterns, jaw reinforcement, fin reinforcement, shoulder girdles, and scales would be obvious.

Basically, if we look back in time, the simple classification schemes proposed by early palaeontologists were only possible because of the fossils they had in hand – and this is especially true for acanthodians. For a long while, anything with paired fin spines was shoe-horned into the Class Acanthodii, and Early Devonian chondrichthyans were only known from isolated remains. But given what we know now, the defining line between early chondrichthyans and acanthodians has become blurry. To me that suggests we are closing in on the common ancestor between those two major groups. Did chondrichthyans evolve from acanthodians? Or did the reverse happen? I suspect the latter.

Either way, we’ll never know the exact patterns of ancestry, and finding the first species of any group is about as likely as a pot of gold at the end of a rainbow. But what I see is that early ray-finned fishes and lobe-finned fishes are closer to each other than either to any other group of fishes – it’s obvious when you look at their overall structure. If you work back from acanthodian species in the Permian to Early Devonian acanthodiforms, or a Silurian ischnacanthiform, and still they are glaringly different compared to contemporary bony fishes, then the groups as a whole are not closely related. Perhaps perfect fossils will be found in a rich Late Ordovician deposit which will explain everything, and if I am lucky, it will happen in my lifetime.

The challenge we face in the immediate future is to determine how acanthodians and the earliest cartilaginous fishes interrelate. For now though, I still have five or so MOTH chondrichthyan species descriptions to publish, and one more acanthodian species description to publish. Afterwards, their data will be added to the matrix, and in time we should see patterns emerge.