So again, I missed most of the morning, spending it in the UK Embassy getting an emergency passport. Unfortunately, this means most of the Romer Session, where early-career (thanks Phil for correction) students present for an award, was missed. Obviously, with the lack of Wi-Fi and live-tweeting, the session might as well have been conducted in a black hole.
I managed to catch the last talk though, on the evolutionary relationships (phylogeny) of diplodocoids sauropodomorphs. As these guys are quite an important constituent of Late Jurassic terrestrial ecosystems, figured this would be worth attending, and actually paying attention to.
A huge-specimen based character matrix was developed, which is cool as normally diplodocoid phylogenies are constructed based on genera, which often include multiple species. Unfortunately, the original tree was completely unresolved, called a polytomy, where the relationships can’t be determined.
When some ‘wildcard’ specimens were removed though, and forms of weighting applied (they change the ‘influence’ of anatomical characters on the shape of the tree), two different trees were produced, and resolved pretty well, although with a bit of conflict. Rather interestingly, this resolved several new genera and species within diplodocoids, including resurrecting Brontosaurus from its long slumber as a synonym of Apatosaurus.
A bit of a medley to kick off the afternoon. Recently, a lot of research has gone into investigating the effects of incomplete or biased sampling on our interpretations of palaeobiodiversity. Synapsids are a group of early animals that include modern mammals. Throughout the Palaeozoic and early Mesozoic. It appears that a lot of synapsid diversity can be explained by biases in the fossil record, such as the intensity of collecting efforts, as well as the number of fossil-bearing localities. Using a series of sampling-correction techniques, a severe extinction is found at the Kungurian/Roadian boundary (~269 Ma)
Also, using estimates of skeletal completeness per species, you can look at how completely known synapsids are through time. This actually is negatively corresponded to taxonomic diversity, which implies that a lot of their diversity is based on fairly incomplete specimens. The lowest point of skeletal completeness corresponds to the extinction event mentioned above, and may relate to the ascent of therapsids, the group which replaced synapsids during the Permian.
The next talk was a pretty cool topic, on using isotopes to measure the body temperature dinosaurs. Using minerals that have replaced the bones, called biominerals, you can measure the bonding of carbon and oxygen isotopes. The way in which they bond is a measure of temperature, by looking at the relative ‘clumpiness’ of the isotopes. This is a pretty new technique, and can be used to investigate past climates too, as well as body temperatures in vertebrates. Initially, this has been used to determine that the body temperature of mammoths is slightly higher than that for modern elephants, and on sauropod dinosaurs using their teeth, to show that they were around the same at around 36-38 degrees, much more than extant ectotherms. This implies that sauropods perhaps were physiologically capable of controlling their temperatures from getting too high.
We shifted over on to ear and braincase evolution in dinosaurs next, which is pretty cool as you can actually start looking at sensory evolution with new techniques like CT-scanning. There was a hell of a lot in this talk I didn’t understand, particularly about the anatomy, but the main concept was about the way in which sound waves move through different media, such as air and water, and solid bone. Turns out that, in dinosaurs, ear morphology diversified in many different ways and independently in several different lineages, and at several different times in the geological record. In ornithischian dinosaurs, this evolution may be related to a period of high diversification, which is cool. But does correlation imply causation..?
Shifting back to the Permian period (end Palaeozoic), we peered back at tetrapod (four-limbed) palaeobiodiversity in South Africa, so a really fine-scaled analysis as opposed to a lot of the more global analyses currently out and about. It turns out when you look at this kind of scale, you see a 70% genera-level extinction within about 150m of geological time, close to the end-Guadalupian, which probably classifies as a mass extinction in these animals (no-one really knows how to quantify this).
How to estimate body mass in extinct animals, particularly tetrapods, has always been a hot point of debate. It’s important to get right, as body mass is an important physiological factor when tied to biodiversity, the environment and an animal’s ecology. Bipeds and quadrupeds are difficult to compare due to their different limb proportions, measurements of which can usually be used to estimate body mass, for example in dinosaurs. A new method takes cross-sectional area of the femur, and uses this, corrected by what we already know about quadrupeds, to infer the body mass of extinct bipeds, such as Tyrannosaurus rex (“model organism”) and extant birds.
We began to wind the afternoon talks off the day by looking at a classic – the end-Cretaceous mass extinction in dinosaurs. This new analysis looked at how preservation biases effect dinosaurs coming up to the boundary in North America, by looking at completeness of their skeletons, as well as the influence of different body sizes. It was found that larger body sizes tend to lead towards better skeletal completeness, which is also reflected in the taphonomic quality (from articulated to isolated specimens). Interestingly, small taxa show an exponential rate of discovery, whereas larger ones a more logarithmic count through time, which suggests we still have lots more smaller-bodied dinosaurs to be found!
Next up, we stuck in the Mesozoic, and looked at how the completeness of dinosaur skeletons changed through time. This worked by calculating the completeness from published cladistic matrices, giving the percentage of characters which were scored against those which were unknown. A huge low in completeness was found at the end-Jurassic (great for my PhD!), with lesser troughs and peaks throughout the Mesozoic. There was also a high completeness found in mid-high latitudes in the northern hemisphere, very similar to known dinosaur diversity during the Cretaceous.
The final talk was looking at the rate of dinosaur limb evolution in a phylogenetic context, and how non-avian and avian dinosaurs differed. In forelimb morphospace
I presented my poster later on in the afternoon, which meant 2 hours of standing around trying to remember what I did two years ago. The poster(s) and original MSc thesis it was based on are both available here on Figshare (obviously).
General points learned today:
- When analysing categorical data, don’t just use arbitrary boundaries between them – they mean nothing in any scientific sense. Let the data find the gaps, before you choose randomly.
- It’s nice to actually be able to read axes on graphs; otherwise it’s just a blob of useless data.
Pingback: Green Tea