We used high-resolution synchrotron scans to investigate tooth development in Powichthys. This fossil fish is one of the earliest bony fishes (osteichthyans) with an advanced dentition. We found surprising parallels in the development of the dentition of Powichthys and more primitive osteichthyans such as Andreolepis. Powichthys reatins a primitive form of tooth development alongside its more advanced tooth rows.

Journal website

What exactly the ancestors of jawed vertebrates looked like has been a topic of vigorous debate. A particularly interesting question is what form the first jaws took. However, different hypotheses about the jaw bones of the placoderms complicate the picture: they are either the equivalent of the palatal bones of boney fishes, or represent maxillae and premaxillae. In this paper we introduced a method that can calculate the evolutionary tree of early jawed vertebrates while simultaneously estimating bone correspondences between placoderms and boney fishes. We found that the ancestors of jawed vertebrates most likely resembled the "maxillate placoderms", from the Silurian period of China, and had maxillae and premaxillae but no palatal bones.

Journal website

A focus of my research since my PhD is the use of a type of phylogenetic analysis (evolutionary tree building) known as tip dating. In this study, I investigated how the use of tip dating affects the placement of fossils in phylogenetic trees. In particular, I wanted to know how the use of tree models that include the ages of fossils influences results. I found that tip dating does indeed have a substantial effect, greater than the differences between other phylogenetic methods. Trees produced by tip dating analysis have a better fit to the fossil record: for example, this method will disfavour trees with a very ancient species related to very recent species. The difference in the phylogenetic relationships is most clear with highly incomplete fossils. In other words, when there is very little data to tell us the relationships of a fossil, tip dating will tend to place this fossil in a position consistent with its age. Other methods on the other hand, might suggest quite surprising trees, for example suggesting that a highly incomplete fossil is the first example of a group that otherwise does not appear until many millions of years later.

pdf Journal website

During the time of the dinosaurs, several groups of early mammals possessed a distinctive type of tooth with longitudinal rows of cusps. The relationships of these mammal groups, collectively known as the ‘allotherians’, have been debated, with one hypothesis being that they all descend from a single common ancestor. We analysed their relationships using a method known as ‘tip dating’, which analyses the age and morphology of fossils simultaneously, and found instead that ‘allotherians’ from different time periods were split into three separate groups. This suggests that early mammal evolution was complex, with different groups converging on similar lifestyles.

pdf Journal website

A major ongoing debate within palaeontology is which methods should be used to build evolutionary trees of relationships. The two major competing methods are "parsimony" which looks for the tree that explains evolution in the simplest way, and "Bayesian", which is a statistical method employing a model of evolution. In the area of vertebrate palaeontology, this argument is very important for figuring out when the osteichthyans (the bony fishes, which includes all tetrapod vertebrates as well as most fishes) first appeared. The estimated relationships of the earliest examples of bony fishes are very different under Bayesian and parsimony methods. The Bayesian result suggests a surprisingly ancient origin for the modern groups of bony fishes. However, a problem with Bayesian analysis of morphological data is the somewhat blurry connection between the final result and the anatomical data used as input. In this paper I introduced a new and simple method to figure exactly which anatomical features are driving the Bayesian analysis. I found that Bayesian analyses are overly reliant on small parts of the anatomical data, and that in the case of the earliest bony fishes, the result from parsimony may in fact be more reliable.

pdf Journal website

Brindabellaspis is a member of the placoderms, a group of fishes that went extinct 360 million years ago. It comes from the Burrinjuck fossil site near Canberra, which preserves a 400 million-year-old coral reef ecosystem. Fossils from Burrinjuck are known for their amazing preservation of braincase anatomy, and because of this Brindabellaspis is one of the best known, but also most controversial, placoderms. In this paper we describe the previously unknown front end of the skull. Brindabellaspis turns out to have a remarkable elongated bill, similar to a paddlefish or platypus, making it one of the most bizarre known placoderms. The bill also contains a unique adaptation of the lateral line sensory system, not seen in any fish, living or fossil.

pdf Journal website

Although often overlooked because we humans lack it, electroreception is a widespread and important sensory system in aquatic vertebrates. However, its origins remain mysterious. In this paper we presented the first in-depth review of possible electroreceptors in early vertebrate fossils, presenting a comprehensive suite of new data from micro-CT scans of exceptionally-preserved fossil material. Many of the possible electroreceptors from the extinct placoderms do not stand up to scrutiny, and we find no evidence for electroreception in this group. However, we found that even the earliest osteichthyans (bony fishes) had elaborate and specialised electrosensory systems.

pdf

This paper formed the centrepiece of my PhD. For several years, the extinct placoderms have been viewed as a "paraphyletic" group. This means that the modern groups of jawed vertebrates (sharks, fishes, tetrapods) evolved from within placoderms, and some placoderms are more closely related to the modern groups than others. However, various discoveries led to doubts in the minds of researchers, including the presence of claspers (mating organs) in several placoderm groups; perhaps the old view that placoderms form a coherent group of their own is true after all. To investigate this problem I compiled the largest dataset of anatomical features of early vertebrates, and analysed them with "tip-dated" methods. These methods model fossil sampling, morphological evolution and fossil ages, and I suspected they may be better able to deal with the unique problems of early vertebrates. In contrast to the currently accepted view, I found strong evidence for placoderm "monophyly" (they form a group descended from a common ancestor).

pdf Journal website

The evolution of live birth from egg laying in reptiles is a classic evolutionary case study. It has long been thought that live birth evolved many times from egg-laying, but that the reverse transition almost never occurred. However, a high-profile study published in 2013 claimed to turn the established view on its head: they suggested live birth evolved early, and there were multiple reversals to egg-laying. We re-analysed the data using multiple methods, including some newly introduced, to show that the result of the 2013 study was a statistical artifact. The problem was that their model assumed that life birth evolved at the same rate across all lizard groups. In truth some groups like Geckos (left, below) are conservative, almost all lay eggs, whereas other groups such as skinks (left, above), go crazy with closely related species differing in their reproduction. We instead found that live birth evolved 115 times, whereas egg laying evolved from live birth only 3 times. The methods we proposed are increasingly being used by evolutionary biologists, as larger and larger evolutionary trees are utilised for analysis of biological traits.

pdf

Malpighian tubules are the excretory organs of insects. The tubules of the fruit fly Drosophila have been extensively studied, but other insects far less. Here we described the embryonic development of Malpighian tubules in the beetle Tribolium. This is particularly significant because this beetle has what is known as a cryptonephridial system, a powerful water-saving device. In the cryptonephridial system, the ends of the Malpighian tubules attach to the rectum and maintain a high salt concentration, drawing water out of the excreted waste.

pdf Journal website