Primary Research Projects

Intra-Skeletal, Intra-Specific, and Inter-Specific Osteohistological Variability
Determining Best Practices for of Studies Growth Inference & Body Size Evolution

Osteohistological techniques are becoming more and more common in studies examining the growth, age, and physiology of extinct vertebrates, Unfortunately, the variability in the proxies used for these inferences remains largely unquantified, and considerable debate exists regarding the optimal skeletal elements that should be used in these investigations. This project, which began through side-projects during my MSc and PhD, has expanded to form my primary postdoctoral research. Broadly, the goals of this project are to quantify the variability that exists in osteohistological features (such as growth mark count, growth zone area/spacing, vascularity and tissue structure, and osteocyte lacunar density/size/shape) at multiple scales (including within a single bone, between different bones of the same individual skeleton, between individuals of the same species, and between species), and use those data to test hypotheses of body-size evolution in non-avian dinosaurs.

While this project is ongoing, earlier components of the project have been published, and are detailed here:

In this paper we described the osteohistology of multiple hind limb elements from a series of ornithomimids and used them as a case study to assess the range of intra-specific and intra-individual variation present in LAG spacing and osteocyte lacunar density. Different patterns of LAG spacing were found between upper and lower hind limb bones, suggesting that direct comparisons of these elements may be misleading, and that LAG spacing is not a reliable proxy for individual growth rates. Osteocyte lacunar density varied more between individual bone elements than between average individual values, suggesting that the choice of sampled element can greatly influence the result, and care should be taken to not bias interpretations of the physiology of fossil tetrapods. 

This research was published in BMC Evolutionary Biology and was an Editor’s Pick and a BMC Series Highlight for November 2014.

(bone microstructure of an ornithomimid fibula, from Cullen et al 2014b)

As noted above, this project forms the core of my ongoing postdoctoral research, and thus most of it is not yet published. Stay tuned for updates as the research continues.

Ecological & Evolutionary Trends in Cretaceous Community Structure
Using VMB data to test long-term biotic responses to environmental perturbations 

The Late Cretaceous was a time of high biodiversity, relatively high global temperatures, and considerable fluctuation in regional sea levels. By quantifying the structure of Late Cretaceous ecological communities, as well as spatial and temporal changes in these systems, we can gain a greater understanding of how ecosystems respond to major environmental shifts, and should provide further insight into predicting extant ecological responses to climate change. One method for estimating Cretaceous vertebrate community structure is through the study of vertebrate microfossil bonebeds (VMBs), which are mass accumulations of small teeth, bones, and scales deposited over geologically short timescales in ancient wetland/lake/river environments, and these assemblages are thought to be representative of the average relative abundance of different clades within their contemporaneous ecological communities. A major component of my PhD thesis focused on using data from Late Cretaceous VMBs to test hypotheses of community responses to environmental perturbation.

The first of these investigations focused on describing a new VMB assemblage from the lower Belly River Group, during the regressive phase of sea level drop near the boundary between the Foremost and Oldman formations (specifically in the upper Foremost Formation, just below the Taber Coal Zone). The site preserved a transitional fauna containing a mix of marine taxa (like sharks and other chondrichthyans), and taxa with terrestrial affinities or no strong environmental affinity (turtles, dinosaurs, eusuchians, etc). In addition, the site preserved the first record of a large hybodont shark cephalic spine from the Cretaceous of Alberta, and the first record of the ratfish Elasmodus from the Foremost Formation. Comparisons of the site with other sites from the Foremost and Oldman suggested that certain groups (such as lissamphibians and shark) were integral for palaeoenvironmental inference from microsite assemblage data.

This paper was published in the February 2016 issue of the journal Palaeogeography, Palaeoclimatology, Palaeoecology.​

(A hybodont shark cephalic spine, Cullen et al 2016)

(R-vs.Q-mode cluster analysis of VMB data, Cullen et al 2016)

(Taphonomic analysis of VMB materials, Cullen et al 2016)

In the next stage of this project, we combined the data collected above with those from dozens of other VMBs (48 sites total), sampled from two regions approximately 150 km apart and spanning ~5 million years over the stratigraphic extent of the Belly River Group in Alberta, in order to analyze changes in community structure (via relative abundance among different taxa), and how these changes correlated with environmental changes related to transgressive-regressive cycles of regional sea level (as well as other factors). Specifically, we quantified changes in species composition and relative abundance between sites, and performed a series of analyses (including R- vs Q-mode cluster analysis, redundancy analysis, and pairwise relative abundance and similarity analyses) comparing the community trends against abiotic factors such as site sedimentology/deposition, sampling location, stratigraphic position, and palaeoenvironmental setting. A specific hypothesis of interest in these analyses was to test if dinosaurs were particularly sensitive to microhabitat changes driven by shifting altitudinal gradients (i.e. sea level changes), as has been hypothesized by a number of previous studies. Our analyses found that the strongest abiotic factor controlling community structure was palaeoenvironment, with site sedimentology, stratigraphic position within the Belly River Group, and sampling location having lesser impacts. On the broader community scale, the most severe changes in community structure occurred between chondrichthyans and lissamphibians, which showed a strong inverse relationship in relative abundance, particularly during shifts from marine-to-terrestrial or terrestrial-to-marine settings. While that result in of itself is not particularly surprising, it is important in establishing that community responses to environmental perturbations are detectable, and serves as a useful comparison to the dinosaur components of the community, which were largely unaffected in relative abundance by those changes (with the exception of periods at the very bottom and top of the Belly River Group where the system becomes fully marine and consequently almost all terrestrial fossils disappear). The results so far suggest that dinosaurs may not be particularly sensitive to small-scale environmental perturbations or changes in microhabitat.

This paper was published in November 2016 in the journal BMC Ecology, It was an Editor's Pick and BMC Ecology top 10 highlight for 2016, and was the subject of a BMC Series blog.

(Temporal & spatial community relative-abundance comparisons, Cullen and Evans 2016)

(Redundancy analysis of community data & environmental variables, Cullen and Evans 2016)

We are currently expanding these comparisons through the sampling, description, and analysis of additional VMB assemblages. The first step of this process involves sampling sites from the currently unsampled interval from the uppermost Oldman Formation in the Milk River / Manyberries region of southernmost Alberta, as well as collecting more detailed stratigraphic data. Going forward, I plan to expand sampling both spatially and temporally, in order to further quantify trends in community structure across environmental gradients, and more specifically to test how these systems respond to more abrupt, larger-scale perturbations such as were seen in the K-Pg mass extinction.

Stable Isotope Ecology of Extant and Ancient Coastal Floodplain Forests
Using extant models to constrain predictions & test hypotheses of dinosaur palaeoecology

Similar to the above research using vertebrate microfossil bonebeds to understand Cretaceous community structure and its response to environmental changes on broad scales, I am interested in performing research at finer temporal and spatial scales in order to understand community structure, habitat preferences and niche-partitioning, as well as diet and trophic interactions. While my interests in these questions pertain to whole communities, and all the taxa within them, I have a particular interest in the ecology of dinosaurian fauna. Considerable research has been performed attempting to characterize the habitat preferences, diet, and niches of dinosaurs, and stable isotope analyses, by virtue of recording the original diets, habitats, and behaviours of these animals, offer a potential avenue for understanding those topics. However, prior to performing detailed analyses of ancient systems, we should understand the constraints that may exist on our inferences, and where possible frame our predictions based on data from similar systems today. To that end, a large component of my PhD research focused on characterizing the stable isotope ecology of vertebrate communities in both extant and ancient systems, particularly coastal floodplain forests.

(Contrasting images of Cretaceous sedimentary rocks preserving fluvial and coastal floodplain environments, and their modern near-equivalents from the Atchafalaya River Basin of Louisiana.)

​First, stable isotope analyses were performed on vertebrate taxa sampled (both opportunistically and obtained through local collaborations and/or museum collections) from within the modern Atchafalaya River Basin of Louisiana, an analogue for the ecosystems and environments preserved across much of western North America during the Cretaceous (Cullen et al 2017 SVP abstract, Cullen et al in press). One of the major constraints limiting palaeocommunity analyses are the lack of clear modern analogues against which to test broad-scale ecological and evolutionary hypotheses, such as mechanisms of community assemblage and patterns of biogeography, within the context of deep time. Our study of the Atchafalaya Basin identified patterns of ecological resource use and aquatic-terrestrial interchange in this system, facilitated the development of improved methods of bioapatite oxygen-temperature inference, and provided an isotopic baseline for use in testing hypotheses in the fossil record. 

Nevertheless, even with well-established natural history data on ecological variables for our Atchafalaya River Basin taxa, such as trophic habits and microhabitat use, the subtleties involved in identifying these patterns in our isotopic data highlight the difficulties involved in detecting and interpreting ecological patterns in a complex C3-based system. It therefore stands to reason that for a palaeoecological system where the description of niches often requires significant inference (particularly among non-avian dinosaurs, for which there is no exact extant analogue), using isotopic data to describe fine-resolution details of niche partitioning should be done conservatively. These comparative data should assist in constraining predictions and limiting the degree to which palaeoecological-proxy data may be over-interpreted.

(Stable isotope analyses of Atchafalaya communities, Cullen et al in press)

Having established the Atchafalaya Basin as an appropriate modern comparator for a variety of Cretaceous animal assemblages opens up the potential for a variety of phenomena to be investigated such as habitat use, niche partitioning, diet, and physiology (Cullen 2018 CSVP abstract, Cullen 2018 SVP abstract, Cullen et al in prep). These fine-resolution questions are rarely accessible within a palaeontological context, and the framework developed through our extant comparisons will facilitate an expansive range of future research. Thus far, I have presented on some of the results of these fossil stable isotope analyses, where we have determined that isotopic baselines in the Cretaceous were shifted relative to the modern day (explaining why several previous studies, as well as our own, find that dinosaur carbon isotope ratios are higher than expected for animals in a C3-based system), that broad overlap exists in the isotopic distributions of large herbivorous dinosaurs (suggesting that at least in the Campanian of Alberta that these animals were not partitioning their niches through differences in microhabitat use), and that much like in extant coastal floodplain forests there was considerable aquatic-terrestrial resource interchange in these Cretaceous communities. These results, as well as some others, are not yet published, though they are detailed in several conference presentations. Additional details will be posted soon as these papers are peer-reviewed and published.

---

Secondary Research Projects

Theropod Biodiversity and Macroevolution
Combining multiple lines of evidence to aid in species discovery and understanding evolutionary patterns

In addition to my primary research, I have been involved in several studies investigating theropod (and dinosaur) evolution and/or naming new theropod species. These projects frequently make use of a combination of methods, such as morphometrics, phylogenetics, histology, and biostratigraphy.

I have been involved in a number of project investigating morphological variability in North American ornithomimid dinosaurs. This began with my undergraduate honours project, describing the first evidence of gregarious behaviour in North American ornithomimids. The evidence consisted of three individual Ornithomimus edmontonicus (or Dromiceiomimus brevitertius, depending on the status of that particular synonymization) skeletons found in close association. In addition to using these specimens to make some inferences regarding ornithomimid behaviour, the material from these specimens facilitated further collaborations wherein we reviewed of the taxonomic utility of various components of ornithomimid post-crania, particularly the pedal phalanges and unguals (McFeeters et al 2018a, McFeeters et al 2018b). As well, as noted above, the skeletons in this study were later sampled to form the core data for my research on ornithomimid osteohistovariability (Cullen et al 2014b)

(Smaller partial ornithomimid skeleton, Cullen et al 2013)

(Larger partial ornithomimid skeleton, Cullen et al 2013)

(Pedal ungual II of various ornithomimids, McFeeters et al 2018a)

(Pedal ungual III of various ornithomimids, McFeeters et al 2018a)

Another particularly interesting ornithomimid project I've been involved with, led by Brad McFeeters, focused on a partial skeleton at the Royal Ontario Museum (ROM 1790). Over the course of this research, we determined that this animal likely represents a distinct species of ornithomimid, and thus it was erected as a new genus and species (Rativates evadens). As part of this project, I performed osteohistological analyses that established that ROM 1790, despite being considerably smaller than a number of other ornithomimid skeletons known from Alberta, was in fact much older than them (determined through skeletochronological comparisons between ROM 1790 and previously sampled ornithomimids). As part of this project, a life reconstruction of Rativates was created by palaeoartist Andrey Atuchin.

This project was published in the Journal of Vertebrate Paleontology.

(Reconstruction of Rativates evadens, by Andrey Atuchin)

(Skull of ROM 1790, Rativates evadens, from McFeeters et al 2016)

(Osteohistological thin-section of femur of ROM 1790, from McFeeters et al 2016)

More recently, I was involved in a project led by my PhD supervisor David Evans that investigated putative differences between troodontid cranial material found in the Horseshoe Canyon Formation and the Dinosaur Park Formation. This project used morphometrics to analyze frontal and tooth morphology (in addition to qualitative morphological descriptions) in North American troodontids, and led us to name a new taxon, Albertavenator curriei, from the Horseshoe Canyon Formation. We also commented on the taxonomic status of Troodon, arguing that due to the inability of troodontid tooth morphology to reliably distinguish between different North American species, that Troodon formosus should be restricted to material in the Judith River Formation (pending the discovery of additional cranial material to facilitate a more thorough evaluation of T. formosus), with Stenonychosaurus inequalis being used for cranial and other skeletal material of troodontids from the Dinosaur Park Formation. My primary contributions to this project involved the frontal morphometrics and discussions of troodontid taxonomy. This is part of a broader series of projects analyzing theropod cranial morphometrics, and will also have further implications for Canadian troodontid taxonomy. Our work on Albertavenator was given a life reconstruction by Oliver Demuth.

This project was published in the Canadian Journal of Earth Sciences.

(Reconstruction of Albertavenator curriei, by Oliver Demuth)

(Geometric and linear morphometrics of North American troodontid frontals, Evans et al 2017)

(Comparison of Horseshoe Canyon and Dinosaur Park Fm troodontid frontal morphology, Evans et al 2017)

(Linear morphometrics of isolated and in-place North American troodontid teeth, Evans et al 2017)

Evolution of Sexual Dimorphism and Polygyny in Pinnipeds
Quantifying dimorphic morphology in extant and extinct taxa

My M.Sc. research used geometric morphometrics to show that males and females of sexually dimorphic, polygynous, pinnipeds can be identified based on the shape of their skulls (in addition to size differences). These extant relationships were used to hypothesize that one of the earliest fossil relatives of pinnipeds, Enaliarctos emlongi, was likely sexually dimorphic (of a degree similar to extant dimorphic otariids), and that cranial shape sexual dimorphism and a harem-based polygynous mating system are likely the ancestral condition for pinnipeds. We postulated that some of these traits may have evolved as a result of climate-driven ocean upwelling changes leading to greater mass accumulation & colony formation, and consequently higher selection for dimorphic traits and polygynous mating systems.

This research was published in Evolution and was the subject of the May 2014 issue cover.

(Comparison of diagnostic morphological characters of several fossil enaliarctine species, Cullen et al 2014)

(Geometric morphometric analysis of pinniped skull shape, incorporating both extant and extinct taxa, Cullen et al 2014)

(Ancestral character estimations of dimorphism and mating systems in pinnipeds, Cullen et al 2014)

---

Funding, Support, and Collaboration

I would like to thank the following agencies and institutions for their funding and/or support of my research:

Detailed funding information is available in my CV.