The evolutionary process has produced multicellular organisms with a bewildering diversity forms. I am particularly interested in understanding whether traits change in ways that are generalizable. This page includes a few examples of projects I’ve done on the topic.
Are there predictable patterns in the evolution of tissue shape and size?
SUMMARY: We assembled a dataset of egg traits from thousands of insect species, and then used phylogenetic approaches to assess long-standing hypotheses about patterns of evolution.
Over the course of evolution, the size of animals changes dramatically. Researchers have hypothesized that these changes can be explained by certain developmental, morphological, and ecological pressures that could, in principle, hold across distantly related animals. We set out to test many of those previous hypotheses using an evolutionarily conserved tissue: the insect egg.
We began by scouring the scientific record, using some custom automated tools based on natural language processing (today, using LLMs, the process could be done much more efficiently). All told, we collected more than 10,000 descriptions of eggs. These descriptions correspond to 6,706 species spanning the major groups of insects, drawn from 526 families and all extant orders. A small sample:
Eggs vary in how round, long, pointed, or curved they are, and they especially vary in size. You could fit the smallest egg inside the biggest more than 100,000,000 times. Here’s a fun example of some bee and wasp eggs at the same scale:
Eggs also vary in their surface texture & color, and they are laid in every location imaginable: in water, in plants, underground, inside other animals bodies, inside other eggs. This louse would like to lay her eggs in our hair:
So, why did we collect all these egg data points? First: to delight in all sorts of natural history gems hidden in descriptive monographs and old tomes. Second: to learn about the causes and effects of evolutionary changes to the size and shape of an essential tissue—one that is shared by a large group of related organisms.
To that end, we assembled additional datasets of developmental traits, egg-laying ecology, and others (Note: These datasets are available for others to re-use, so please reach out if you’re interested in using them for your own analyses)
We also used genetic databases to build an expanded insect phylogenetic tree, reconstructed ancestral states, and accounted for relatedness in assessing patterns. In non-technical terms: we collected info about how insects are related to one another, and used that info to uncover evolutionary patterns. Some insect lineages make unique egg shapes, but others have had tons of convergence—that is, the same size and shape evolved independently.
With these ingredients, we tested a bunch of long-standing hypotheses about egg evolution. These deal with allometry, ecology, and embryology. I’ll briefly highlight one such result. Q: Does the place where eggs are laid help to explain the evolution of egg shapes and sizes?
We tested this by first reconstructing the evolutionary history of two traits: egg-laying in water and egg-laying inside the bodies of other animals (i.e. “parasitoids”). Yes, it is a tad gnarly to hatch inside another animal and then eat your way out.
Turns out insects have switched egg-laying environment many times over the course of evolution. We then asked whether taking into account the evolutionary history of egg laying environment helped in modeling the evolution of egg shape and size. Result: yes, this was true.
We found that eggs laid in water tend to be smaller and rounder, while eggs laid inside other animals bodies tend to be smaller and more asymmetric. One take home: When studying the evolution of size, we suggest that ecological implications should be weighted strongly.
As a field we’re just scratching the surface. There are all sorts of exciting discoveries waiting to be made; in fact you can pore over the eggs yourself!
Insect egg datasets available for other researchers to use
Based on the work described above, a separate dataset paper describes the detailed methods of how we compiled the dataset, along with some now-rather-outdated recommendations for how to collect similar literature-derived, composite datasets. I have also co-authored a paper with suggestions for effectively collecting datasets for image-based (i.e. computer vision) analyses.
The dataset is available on Dryad for re-use.
Interactive data visualization made by my colleagues Sam Church and Milo Johnson.How microscopic features produce whole-tissue traits
SUMMARY: How could a dragonfly change the shape of its wings during flight even though it has no muscles within the wings? We used mechanical tests and detailed micrographs of 12 species to show that tiny flexible protein pads and mechanical linkages are built into the surface of wings. These microscopic structures affect how a wing changes shape in response to aerodynamic forces.
When birds and bats fly, they use wing muscles to actively control the wings’ shape to improve flight efficiency. Insects, whose wings are composed entirely of cuticle, do not have that option. Instead, it has been hypothesized that insects’ wings are structured such that physical forces passively change wing shape in an aerodynamically useful manner. To test this hypothesis, we surveyed the morphology and mechanical function of structural elements within the wings of 12 species of dragonflies and damselflies (order: Odonata).
We focused on two structural elements: 1) Flexible “joints” composed of the fluorescent protein resilin where two wing veins meet, and 2) Cuticular spikes that appear to locally restrict the extent of wing flexibility. We documented the distribution of these components across wings and mechanically confirmed that they have opposing effects on wing flexibility.
We found that resilin joints and cuticular spikes are widespread, yet both traits display a striking degree of morphological and functional diversity. The functional significance of resilin joints and cuticular spikes could yield insight into the evolutionary relationship between form and function of wings, as well as revealing basic principles of insect wing mechanical design.
After this article was published, Appel and Gorb (Zoologica 2014) conducted a lovely and impressively comprehensive survey of wing morphology across Odonata, and then Appel et al (Journal of Anatomy 2015) extended this work by characterizing the cross-sectional ultrastructure of vein-joints. Rajabi et al (Bioinspiration and Biomimetics 2015) modeled the effects of resilin and cuticular spikes on local wing flexibility. Most exciting of all: by carefully manipulating the wings of live bumblebees, Mountcastle and Combes (Proceedings of the Royal Society B 2014) showed that flexible resilin joints play an essential role in flight.