This is an evolutionary tree of vertebrates highlighting the bichirs, a basal lineage of ray finned fishes (thick blue line above). The Senegal bichir Polypterus senegalus was used as a model for the evolution of terrestrial locomotion in a recent study (reference below). This tree is meant to illustrate the phylogenetic perspective I thought was missing a bit from that study, and especially from a write-up of the study in National Geographic - Evolution’s Baby Steps.
The simple narrative to present this story is that bichirs represent a sort of stop in the transition between aquatic vertebrates, “fish”, and us terrestrial vertebrates. We all know the trope of the fish that crawled up from the sea. On the surface it seems reasonable to assume that bichirs are an evolutionary remnant of that primordial land-living fish; a primitive evolutionary relic somehow stuck in the middle to this day. After all, bichirs can survive out of the water and use their pectoral fins to help propel themselves on land for short distances. They even have simple lungs to help them breathe air. Carl Zimmer’s write-up for National Geographic seems to go straight for this simple narrative, reproducing the trope that “fish” are a primitive stage of vertebrate evolution with the “potential” to one day become us land-living vertebrates.
This is unfortunate. The reality is a bit more complicated and far more exciting. To start with, bichirs are completely removed from the emergence of four-limbed land vertebrates. Bichirs are ray-finned fishes (in blue above) and the transition to land occurred in the lobe-finned fish lineage (in pink above). The time period for the transition to land in the late Devonian, around 370 million years ago, is roughly marked with a grey box in the tree. Bichirs are more closely related to teleost fishes - what we usually think of when we think of a “fish”.
If you locate four-limbed vertebrates or tetrapods in the tree you can see that we are surrounded by “fish” on all sides. In fact, it is entirely possible to argue that we are “fish” as well.
Until we tend to call tetrapods (including ourselves) lobefin fish and accept that all living vertebrates are fish, the term ‘‘fish’’ should be used with caution. - Source
From a phylogenetic perspective, a lungfish would be a better model for the evolution of terrestrial locomotion. However, they don’t use their fleshy fins for terrestrial locomotion, even though they are able to survive and move on land.
So the straightforward narrative of the progression from “fish” to land living vertebrates is, at best, incomplete. It leaves out most of the fascinating diversity of vertebrates, the majority of which are teleost fish, and it ignores all the complex and not at all primitive adaptations found in now living “fish”.
The Nature paper describing the study does a better job at justifying why bichirs can be a good model for the evolution of terrestrial locomotion. The authors highlight a series of features that are similar to those of early lobe-finned fishes, like "an elongate body form, rhomboid scales, ventrolaterally positioned pectoral fins and functional lungs", as well as tetrapod-like features like terrestrial locomotion using the pectoral fins. Usually bichirs use their pectoral fins to prop themselves up on the bottom of the lakes and rivers where they live, almost looking as if they were doing pull-ups.
The authors also make a sort-of argument based on the phylogenetic position of bichirs, mentioning that they are the now-living fish closest to the common ancestor of both ray-finned fishes and lobe-finned fishes. This is of course an exciting perspective, but the point is not developed further in the paper. You are left to speculate that perhaps the adaptations that allowed early lobe-finned fish to adapt to terrestrial life were present already in a much earlier ancestor. Although I am not aware of any other line of evidence that would support this. It’s also worth mentioning that the presence of lungs and terrestrial locomotion is not necessarily a primitive feature, as demonstrated by the mudskipper, a teleost fish.
In any case, the aim of the study was to provide experimental evidence for the proposed role of phenotypic plasticity in the evolution of terrestrial locomotion. Both the morphologic and behavioural changes associated with terrestrial locomotion could have arisen relatively quickly through the fine-tuning of the phenotypic variation already found in the population. If it can happen in bichirs today, it could have happened in the late Devonian.
Standen, E., Du, T., & Larsson, H. (2014). Developmental plasticity and the origin of tetrapods Nature, 513 (7516), 54-58 DOI: 10.1038/nature13708