Researching Fossil Ungulate Communities


Alces alces (moose), Porkkala, Finland; photo courtesy of Juha Saarinen

In their paper “Patterns of diet and body mass of large ungulates from the Pleistocene of Western Europe, and their relation to vegetation,” published this past September in Palaeontologia Electronica, Juha Saarinen, Jussi Eronen, Mikael Fortelius, Heikki Seppä, and Adrian Lister investigate fossil ungulate communities found in England, Ireland, and Germany.

Not fossil ungulates, fossil ungulate communities.

The variety of fossils studied is just one of the exciting elements of their research.  Rather than focusing on a single species—which, given the limitations of the fossil record, is usually the case—they studied groups of fossils from at least 14 different ungulate species from the Middle to Late Pleistocene.

“[W]e are now at a point,” wrote Juha Saarinen, lead author of the paper, in an email, “where enough fossil material of ungulates and pollen records have accumulated to enable such a large scale quantitative comparison of body size and diets of ungulate with local vegetation patterns in the past as we did. Comparing vegetation proxies and mammal ecometrics from fossil data using such quantitative statistical analyses as we did has, to our knowledge, never been attempted before, so that is probably the most novel achievement of this study.”

The ungainly name of ‘ungulate’ refers to hooved animals: even-toed and odd-toed (Artiodactyla and Perissodactyla, respectively). Examples include horses, deer, moose, rhinoceros, bison, pigs and hippopotamuses.


Brontops tyleri (a type of brontothere and a Perissodactyl) at the Beneski Museum at Amherst College, Massachusetts.   Brontotheres survived until the Eocene, an era that ended approximately 30+ million years BEFORE the Pleistocene, so this animal–although an ungulate–was not part of this study. Picture taken by the author of this blog


Using mesowear on the fossil teeth, they were able to determine information about their diets (from browsing to grazing), and by comparing this data with the pollen record associated with the areas in which these fossils were found, they were able to tell whether they ate more browse or grass in either open or closed environments. Body mass for these fossils was calculated and then compared to the diet of these animals.

They were searching for answers to how these species adapted to the environment in which they lived.  How did their body size relate to the vegetation available? Was their body size influenced by possible predators or by other members of their species? (In other words, were they bigger to intimidate predators or were they smaller because they lived in expansive herds?) Or was thermoregulation the single determining factor in how big these animals became, as has been proposed in earlier studies?


Beneski - Irish elk

Megaloceros giganteus (otherwise known as Irish Elk and an Artiodactyl) in between a mastodon and a mammoth fossil at the Beneski Museum at Amherst College, Massachusetts; picture taken by the author of this blog


It interested me to learn that they relied on what I rather simplistically referred to as the ‘physical observation’ of fossils.

Mesowear analysis looks at the wear and shape of fossil teeth.  Various plant material affects tooth-wear in distinctly different ways, which can be seen both on the teeth themselves and in the way the teeth have evolved.

To be clear, “this is specifically wear-induced shape, not the original shape of the unworn teeth,” Juha added. “In other words, mesowear is the change in the shape of the teeth as they get worn, and different food items cause different worn shape to develop (browse maintains high and sharp features on the tooth surface, whereas grass “grinds” them down leading to them to progressively wear down lower and more blunted the more there is grass in the diet).”


Examples of a mammoth tooth — used to eat mostly grasses and sedges — and a mastodon tooth — used to eat trees and shrubs. Notice the very different shape of these teeth for very different types of vegetation. Proboscideans such as mammoths and mastodons were once grouped in with ungulates, but this has changed. Picture taken at the Harvard Museum of Natural History by the author of this blog.


Obtaining data about the pollen record (non-arboreal pollen percentages, or NAP %) meant researching published information and connecting that information with the related fossil sites.

The mathematical work behind all of this–determining mesowear, animal body size, and then relating this to the available pollen record—is staggering.

Surely, I thought, isotopic analysis would have been a much easier way to obtain information about each fossil’s diet at least.  Especially given that the pollen record isn’t always available, or—in one case—runs the risk of being skewed by the defecation of Pleistocene hippopotamuses that grazed in the area.  Why, I wondered, did they rely on methods that seemed considerably more labor-intensive and potentially (to my understanding) less accurate?

“There are a number of reasons for this,” Juha explained. “First, we wanted to obtain as much palaeodietary data as possible, comprising as complete ungulate communities as possible, and this meant dealing with very large samples of fossil molar teeth. Taking isotope samples from all those teeth would have been laborious, time consuming and expensive, not to mention also slightly destructive to the fossil specimens.


Cervus elaphus (Red Deer, Artiodactyl) at Richmond Park, London; photo courtesy of Juha Saarinen. Red Deer are one of the most extensively studied animals today. You can read about another study that references Red Deer in this post.


“Second, stable isotopes work best at resolving herbivore diet compositions in tropical areas where carbon isotope composition reflects roughly the proportions of C4/C3 –photosynthesizing plants (roughly grass vs. browse) in diet, but outside tropical areas all plants, grasses included, are C3 photosynthesizing and the carbon isotope composition varies also considerably according to so called canopy effect (open vs. closed environment), not just according to diet, and thus isotopes would not have allowed us to estimate the amount of grass vs. browse in the Pleistocene European ungulates as consistently and quantitatively as we could with mesowear analysis.

“Third, mesowear has been specifically shown to reflect average grass vs. browse compositions in the diets of ungulate populations, without being significantly obscured by other environmental variables, such as climate or environmental openness (e.g. Louys et al. 2012, Kaiser et al. 2013). Even if mesowear is a ‘physical observation’ as you say, it has been shown to specifically reflect the amount of abrasive dietary items (mostly grass) in herbivore diets.”

The authors focused on fossil-rich sites, where they could study between 3 – 10 fossils of each species.  They made sure to include species that were browsers, grazers and mixed-feeders.


Screenshot of Figure 1 from “Patterns of diet and body mass of large ungulates from the Pleistocene of Western Europe, and their relation to vegetation.” Palaeontologia Electronica19.3.32A: 1-58


“I owe thanks to my co-authors who knew much of the available European Pleistocene mammal collections already, having experience on working on them for many years,” Juha responded when asked how they knew of or had access to so many fossils.

Adrian Lister from the Natural History Museum of London in particular has a huge amount of knowledge and experience about Pleistocene mammal collections.

“I was also in contact with the curators of the museum collections, who gave me valuable information about the how much and what kind of material they have. Also, information about important fossil finds and numbers of specimens found have often been published before in scientific journals.

“The authors of this paper represent different fields of research experience on the various aspects of the study. I started to work on this research as a part of my PhD work, and I originally planned it with my PhD thesis supervisors Mikael Fortelius, Jussi Eronen and Heikki Seppä from the University of Helsinki.

“During the work, I visited the Natural History Museum of London, where I worked together with Professor Adrian Lister, whose expertise on British Pleistocene mammals, the NHM fossil mammal collections and mammal palaeoecology in general were very important for this work.”


Image of Professor Adrian Lister, Natural History Museum of London, with the mummified baby mammoth, Lyuba; photo courtesy of the Natural History Museum of London for this post.


This work was not without its challenges.  As with any study of fossils, there are limits to the number of fossils available.  While pollen record availability has increased, there is still so much more to be discovered.  And although some species–based on extant examples–do not exhibit sexual dimorphism in body size, the sex of most of the fossils they studied was indeterminate.

“Indeed, these were some of biggest challenges in this study,” Juha acknowledged, “but they were expected and nothing much could be done to completely avoid them. I would add that it was often challenging to connect the fossil mammals with associated pollen records, especially when the fossil pollen was not obtained directly from the mammal fossils. To succeed in this study, it was important to analyze lots of data in order to overcome these problems, and to ensure that the main results and conclusions of this study are robust despite of them.”

The authors of this paper considered numerous variables in their research, and they suggest that ungulate size has a lot to do with a number of factors.  This might seem obvious, but such has not been the result of past studies.  In particular, Bergmann’s rule, which stipulates that body size corresponds largely to thermoregulation (i.e.: big body size is the result of living in colder environments), has been supported before.


Bison bonasus (Artiodactyla), Kraansvlak, Netherlands;photo courtesy of Juha Saarinen. 


“[T]here has been a lot of discussion as to what ultimately explains the tendency of some (but not all) organisms to be larger in cold climate. This was actually one of the main questions I discussed in my PhD thesis,” wrote Juha. “Already in 1950s some researchers (e.g. Scholander 1955, Irving 1957, Hayward 1965) pointed out that increase in size alone would not give a large enough benefit for thermoregulation in cold climates, especially considering that mammals have far more effective mechanisms of keeping warm, such as thick fur.

“Since then, many authors have noted that while there is a tendency of mammals being larger in higher latitudes, there are a number of exceptions to this ‘rule’ and heat conservation alone would not explain it.

“However, body size in mammals does correlate with food quality and availability and this seems to explain most of the body size patterns observed in mammals (e.g. Rosenzweig 1968, Geist 1987, Meiri et al. 2007, McNab 2010). For example, many herbivorous mammals tend to be larger at higher latitudes because food quality is better there (e.g. because of fertile soils created by glacial erosion and because plant defense mechanisms are lower), and thus predators eating them also tend to be larger there, but for example brown bear body mass does not correlate with latitude but with distance to nearest salmon spawning areas. On the other hand, population density also affects body size through resource availability: individual body size has been noted to decrease in many species of mammals when population densities are high leading to increased intraspecific resource competition (e.g. Wolverton et al. 2009).”

The authors of this paper argue that environment–climate, open or closed vegetation, food availability and quality–and species social structure–large or small herds–affect body size.

“[T]here are many (often interconnected) factors which together affect body size,” Juha explained. “This makes it quite complicated and challenging to study what ultimately regulates body size in mammals (and other organisms).

“In fact, our results do not support Bergmann’s rule as such, because even if our analyses show that larger sizes seem to occur in some species in open environments, this is not because of low temperature, as some of the open environments were in fact quite warm. Also, we often see that when one species was particularly large in an environment, another species was particularly small under those same conditions. E.g., we found out that red deer (Cervus elaphus) tends to be large in open environments, but wild horse (Equus ferus) tends to be small in those same environments. Thus, our results do not support the assumption of Bergmann’s rule or any other “single-cause” explanation for ungulate body size variation.

“What ultimately regulates ungulate body size is primarily food quality and availability, which is affected by the interplay of vegetation structure (regulated by environmental temperature, precipitation and soil fertility), interspecific resource competition (depending on the presence of competing species) and intraspecific resource competition (depending on population density). For example, species with large population densities in open environments, such as reindeer, bison and wild horses, could be small under those conditions because of increased intraspecific resource competition, whereas species with smaller population densities in open environments, such as red deer are large under such conditions, e.g. because of abundant, high-quality food and diminished plant defense mechanics. This is also the main conclusion concerning our results of Pleistocene European ungulate body size variation.”

“I think that studying how mammals in the past interacted with their environments is important for understanding how these interactions work in general,” he concluded. “At present, environments and their mammal faunas are so heavily influenced by human activities, and they have lost so much of their original diversity, that I believe that we simply need to study fossil mammals and their palaeoenvironments to better understand how these things have worked and ‘should usually work’ in nature.”


Equus ferus (Mongolian wild horse and Perissodactyl), Lippeaue, Germany;photo courtesy of Juha Saarinen. 

It was a great honor and pleasure connecting with Dr. Juha Saarinen!  Reading this paper and gaining more insight about it from him was absolutely fascinating!  An enormous thank you to him for all of his generous help!!

Additionally, Dr. Saarinen was extraordinarily kind and helpful in clarifying points about the research that I had misunderstood.  That is always appreciated.  THANK YOU!!


  1. Saarinen, Juha, Eronen, Jussi, Fortelius, Mikael, Seppä, Heikki, and Lister, Adrian M. 2016. Patterns of diet and body mass of large ungulates from the Pleistocene of Western Europe, and their relation to vegetation. Palaeontologia Electronica 19.3.32A: 1-58

Color Vision Discovered in 300 Mya Fish

According to Dr. Gengo Tanaka, the fossil below was found about 50 years ago.

“I have a friend [who owns a] fossil shop,” he wrote in an email. “I bought this specimen from him.”

Dr. Tanaka explained that his friend attributed the fossil find to his father, who discovered it in a quarry five decades ago.

It’s a small fish known as Acanthodes bridge, and it is thought to have lived in shallow waters 300 million years ago in what is now Kansas.


Acanthodes bridge - Tanaka, et al

[image of fossilized Acanthodes bridge, courtesy of the Natural History Museum, London]


It might be a little fish, but it is providing enormous and exciting information about the evolution of color vision.

In their recent paper, Dr. Gengo Tanaka of Kumamoto UniversityProfessor Andrew Parker of the Natural History Museum, London and 13 other scientists describe evidence of color vision 100 million years earlier than previously known.  They are the first to record fossil rods and cones—the cells responsible for enabling sight.

“The soft tissue of eyes are usually the first to decompose when an animal dies, and before they are fossilized. In our fish, however,” wrote Professor Parker,  “the soft tissue was preserved before burial (by sediment) and turning to rock. The original organic material has been altered (but in some cases not too much), although remains relatively soft.”

Using SEM (scanning electron microscope) and TEM (transmission electron microscopy), the scientists studied the fossil eyes in more depth.  They compared rods and cones of 509 retinal cells, obtained from the fossil itself and from existing freshwater fish.  Cones within the eye are the key to color vision, although they assert that the discovery of opsins in the fossil record would provide “conclusive evidence” for such vision.

“Cone cells are those responsible for colour vision in extant animals,” Professor Parker further explained. “They contain the opsins that react to different wavelengths of light.”

When asked if he expected to find color vision in this fossil, Dr. Tanaka wrote, “I have discovered fossilized rod and cones in several Cenozoic fishes. So, I expected that we could discover fossilized rod and cone cells in other specimens.”

Similarly, Professor Parker wrote, “I would expect to find colour vision in the geological record at some point, but I did not expect when.”

Acanthodes bridge - eye detail


[details of Acanthodes bridge, courtesy of the Natural History Museum, London; a: Complete dorsoventrally compressed specimen, b: details of the head region, c: details of right eye]

“We were a team interested in the emergence and history of vision, when Gengo [Tanaka] found the fossil fish,” he continued, describing how it came to pass that these 15 scientists collaborated on the paper.

But why would color vision be significant for a species such as Acanthodes bridge?

“That such ancient fish had colour vision tells us that the type of ecologies and behaviours that exist today, where light plays a major role, were also in place 300 million years ago. For the fish, they could distinguish predators and prey with greater accuracy and in some cases crack the camouflage of these animals.”

“This is the first time that colour vision has been identified in any extinct animal, regardless of geological age,” he wrote. “It suggests that our modern behavioural system, or way of living, where colour plays a major role, has been in place over at least 300 million years.”

“This can explain,” he continued, “why things have changed little over that period: predators and prey have changed form in some ways, but the balance of the different types of animals and plants living together has remained similar.  Triceratops has been replaced with rhinos, ichthyosaurs replaced by dolphins, but their roles in the ecosystem are similar. That they saw in the same way helps us to understand this.”


An enormous thank you to Chloe Kemberry, Dr. Gengo Tanaka and Dr. Andrew Parker!  What a great pleasure connecting with you about such an exciting discovery!

Paper referenced: