Judges’ Queries and Presenter’s Replies

  • May 20, 2013 | 10:26 p.m.

    Please comment on sample size. My concern is that no one type of scratching on teeth would characterize the modern human population, i.e., compare a vegan to one that chews on bones. Does your Miocene sample come from one or several different populations?

  • Icon for: Brian Shearer

    Brian Shearer

    May 21, 2013 | 12:31 a.m.

    Thanks for your interest, Dr. Yavitt! To answer part of your question: modern humans are weird. We’re the most generalist of the Primates (and of many mammals, really), so you’d be hard pressed to find an overall microwear signature that characterizes the species of present day Homo sapiens. However, if you look at more traditional populations (i.e., hunter-gatherers) you can detect microwear differences between populations that relates to levels of meat in the diet (see El-Zaatari, 2008 for more details). As we go, modern humans aren’t characteristic primates when it comes to these types of data, and therefore aren’t included in our extant data set.

    Regarding sample size: yes, it is very important for this type of research, although depending on the primate species, a sample of around 10 individuals may be sufficient to characterize the dietary preferences of a species in a given season. Many primates (especially some of the Old and New World monkeys) have very specialized diets and dental characteristics specifically adapted to exploit particular niches, which they rarely stray from (e.g., the genera Alouatta or Theropithecus are nearly completely folivorous and granivorous, respectively). Therefore we tend to find very consistent microwear signatures for most extant primates, and where we don’t, we know from behavioral observation data that it’s because those primates are more generalist in their diets. Luckily, our extant sample size is well over 400 individuals, which gives us confidence in the robusticity of our results involving the living primates.

    Regarding the early Miocene sample, many of the species come from what were probably several populations that experienced slight-to-moderately different habitats, though given that all specimens were excavated from neighboring localities in Kenya we can be fairly certain that many of our species were very closely related (this is supported by morphological and cladistic data). Given the often unsatisfying nature of the fossil record, we lack the time resolution to determine whether or not animals in the same geological layer were a population of interbreeding individuals (i.e., that they lived at the same time), but we treat each species as a continuous group, assuming that morphological continuity suggests a stable population structure.

    Please let me know if any of my points need clarification.

  • May 21, 2013 | 11:00 a.m.

    Can you combine your work with an analysis of available flora and fauna data to corroborate the findings or more speculative, can you reconstruct broad aspects of the flora and fauna by looking at dental wear patterns?

  • Icon for: Brian Shearer

    Brian Shearer

    May 21, 2013 | 11:46 a.m.

    Absolutely, Dr. Bhattacharya! This type of research is just a component in characterizing the landscape of past environments, and is most useful when paired with disciplines like macro- and micro-paleontology, palynology, and geology in order to give us a more holistic answer to our questions.

    Microwear studies most commonly paired with stable isotope analysis of the tooth enamel, which gives us a more complete representation of the environment that these animals were living in. For instance, while microwear studies may be able to tell us that an animal ate mostly tough, fibrous foods in the weeks before death, by adding isotope data we can determine the proportion of C3 vs C4 plants in the diet, which gives us a better idea of what the environment could have been like in the past. Add to that evidence from geological analyses and the ratio of known forest or savannah dwelling micro- and macro-mammals, and we start to really get a good idea of what types of environments our target taxa lived in.

    Reconstructing broad aspects of the flora and fauna with dental microwear textures is a bit more tricky. If we were to analyze animals from many different orders with known modern analogs, we would be able to build a decent picture of how dry a locality was, what general types of vegetation were present (i.e., grasses, scrub brush, etc), though we would never be able to completely reconstruct an environment with microwear alone. A good example of this is Ungar et al, 2012, where the authors show that microwear analyses of small tragulids corroborate previously proposed ideas about how wet or dry neighboring island environments were in the Miocene. Excellent question!

  • May 21, 2013 | 10:41 p.m.

    Brian, you should know that you did a great job because I know nothing about dentary analysis and had no trouble following your materials. My question is quite general: you originally sought to determine whether dentary fossil data (microwear texture) can be used as a diagnostic character in phylogenetic analysis of a group of primates. Along the way, you learned something interesting about feeding behavior, but the answer to your original question was essential “no” – microwear texture is not an informative character because the subjects were more-or-less eating the same things. Where to from here? Would you argue that molecular data is the primary way forward, or does behavior and ecology still have an essential role to play in reconstructing historical lineages?

  • Icon for: Brian Shearer

    Brian Shearer

    May 21, 2013 | 11:24 p.m.

    Thank you for the compliments, Dr. McGarvey! Your question touches on a common misconception regarding microwear data: we’re not necessarily trying to directly understand the phylogeny of these critters through microwear (after all, very unrelated animals can eat the same foods!), but rather we’re trying to assess whether or not they fit any particular known feeding niche (e.g., hard object specialist, granivore, frugivore etc) so we can better assess the role they played in their environments, and thereby understand what kind of primates they were on a finer scale.

    Because we have so many of these fossil primates found in the same deposits, it was a bit of a mystery as to how they were partitioning their niches, especially as some of them had very similar body masses. Different species of extant primates that live sympatrically tend to specialize in their diets to the point where there is very little conflict, and that’s what we expected to see with the Miocene fossil groups. However, our data are showing that our focal taxa were not doing that, which while technically a “no” to our original question, is still super interesting because we now need to figure out how they were managing to do it!

    Regarding the where-to-from-here part: I would say that molecular studies are definitely an informative avenue of future research for many things primate. However, as molecules don’t preserve in fossils, the most it can inform us of at this time-depth is potential divergence dates of some of these groups (and that assumes an ancestor-descendant relationship between our fossil taxa and some modern groups, something that not everyone in my field is convinced of!). Morphological studies and traditional cladistics will probably continue to be the most informative avenues of research for determining relationships for our fossil groups, and direct studies of ecology through microwear or stable-isotope analysis will definitely help us determine what our taxa were doing, not just who they are related to. Thanks for the great question!

  • May 23, 2013 | 09:54 a.m.

    No sweat. Thanks for the great answer.

  • May 21, 2013 | 11:02 p.m.

    Dear Brian,

    Truly enjoyed your presentation! I have used texture analysis to look at patterns of wildlife habitat in satellite imagery. Very different scale! Curious though to learn more about the way you calculated texture exactly. Software programs you used, etc. Wondering what the remote sensing community could learn from you?


  • Icon for: Brian Shearer

    Brian Shearer

    May 21, 2013 | 11:55 p.m.

    Thank you very much Dr. Radeloff! I think it’s great that this type of analysis is viable on scales from entire habitats to microscopic wear on enamel, and I’d love to hear from you if you have any suggestions on how to improve the technique or on any updates in your field that might not necessarily have trickled down the paleontology pipeline yet!

    For this study I used a Sensofar white-light scanning confocal microscope (Solarius, Inc.) to create three-dimensional point cloud data. I used a 100x objective lense to generate 4 adjacent fields of 102 × 138µm with a lateral sampling interval of 0.18 µm and vertical resolution of 0.005 µm. I then analyzed these 3-D point clouds using software packages called Toothfrax and Sfrax (Surfract Corp, Worchester, MA), which use scale-sensitive fractal analysis to generate values for an enamel surface’s anisotropy, complexity, scale of maximum complexity, texture fill volume, and two levels of heterogeneity (on both a 3×3 and 9×9 grid density). For primates, we tend to find that anisotropy and complexity have a very strong inverse correlation, and therefore are usually our most informative variables. I’m not sure how it is with larger scales; I would imagine that data acquisition is very much the same, although I would suspect that you would find variables other than complexity and anisotropy very informative, yes? It’s very interesting to think about this trying to characterize a larger scale surface…

    I’d be happy to refer you to some of the seminal literature on the subject in my field that will do a much better job of explaining it that I can here. Thanks again for your question!

  • May 22, 2013 | 02:00 a.m.

    What an excellent video! The question I would like to ask is how long does it take for teeth to become marked enough to be classified in this method?

  • Icon for: Brian Shearer

    Brian Shearer

    May 22, 2013 | 11:41 a.m.

    Thank you very much for the praise, Dr. Anderson! Because enamel is constantly being worn away by food particulates (in non-human primate mouths as well as yours and mine!), toothed animals generally have their microwear signature completely re-written every 2-3 weeks (depending on the abrasiveness of the diet). In my field, this is known commonly as the “Last Supper Effect”, as microwear that we find on fossils in only indicative of the last few weeks of life, though if sample sizes are good, we can presume that these data represent a normal distribution of texture patterns for a species (or at least a regional population). Very good question!

  • May 22, 2013 | 05:21 p.m.

    That is just a fascinating answer -complete with a marvelous expression to seal it in my memory! Thank you, Brian.

  • Icon for: Brian Shearer

    Brian Shearer

    May 22, 2013 | 06:10 p.m.

    You are very welcome!

  • Further posting is closed as the competition has ended.

Presentation Discussion

  • Icon for: Joni Falk

    Joni Falk

    May 23, 2013 | 04:53 p.m.

    Thought this video was great in explaining your research to the public. You’ve made your work very accessible. Thanks for submitting this!

  • Icon for: Brian Shearer

    Brian Shearer

    May 24, 2013 | 10:50 a.m.

    Thanks for the complements! I’m very happy you enjoyed it!

  • Further posting is closed as the competition has ended.

Icon for: Brian Shearer


City University of New York
Years in Grad School: 3

Dental microwear profilometry of African non-cercopithecoid catarrhines of the Early Miocene

The early Miocene of Kenya, roughly 23-16 million years ago, has yielded the remains of many important fossil species that provide a glimpse of the higher primate radiation at a time of major faunal turnover in Africa. These taxa, commonly called the Miocene apes, have been subject to innumerable studies, yet there is still no consensus on their diets, which is a major component in reconstructing their environmental niches. In this study, I analyzed the dental microwear textures of fossil primates from the Early Miocene of Kenya. I collected high-resolution casts of all molar specimens at the National Museums of Kenya, Nairobi, and inspected them with scanning confocal profilometry, a method that distinguishes between patterns in enamel damage caused by mastication of different types of food. This revealed 83 individuals with microwear of the fossil genera Dendropithecus, Micropithecus, Limnopithecus, Proconsul, Kalepithecus, Nyanzapithecus, and Rangwapithecus. Scale-sensitive fractal analysis was used to generate texture attributes, which aims to quantify enamel damage patterns for each specimen, and the fossil taxa were compared to each other and to living primate groups. My analyses revealed no significant variation in microwear texture among the fossil taxa, which suggests the fossil taxa consumed foods with similar mechanical properties despite striking morphological differences. However, further analysis separates the Miocene fossil sample from several living primate genera, which suggests that the primates of the African Early Miocene had generalized diets and had not yet specialized to the degree of many modern taxa, despite variation in size and tooth form.