The better to bite fruit with: Natural selection shapes mechanical advantage in bats

Glossophaga soricina, a nectar bat, feeding on the flowers of a banana plant. Nectar feeding bats comprised one of three evolutionary optima for mechanical advantage among New World Leaf-nosed bats. Photo credit: Dr. Beth Clare, Queen Mary University.

"The key finding is that in a highly diverse group -- New World Leaf-nosed bats -- selection for mechanical advantage has shaped three distinct optimal skull shapes that correspond to feeding niches," Dr. Dávalos explains. "The key development is an engineering model of a very complex structure -- the skull -- that can morph into both known, observed skulls, as well as into forms that do not exist, but nonetheless would be biomechanically feasible. This opens the door to future studies with other highly diverse groups."

The research team investigated adaptive radiation -- the explosive evolution of species into new ecological niches, powered by natural selection -- of New World Leaf-nosed bats. These bats, of which there are almost 200 species, eat a variety of foods, including insects, frogs, lizards, fruit, nectar, and even blood. Their skulls mirror the variety of their diets -- bats with long and narrow snouts eat nectar; Short-faced bats have short, wide palates and eat hard fruits; snouts of species that eat other foods are intermediate in shape.

The team's approach to identifying natural selection for mechanical function combined both evolutionary and engineering analyses. Dr. Dávalos and her Stony Brook University students generated the evolutionary hypotheses, evolutionary trees, and tests of selection reported in the study; Drs. Dumont and Grosse developed the engineering model. The interdisciplinary effort involved evolutionary biology, biomechanics, and mechanical engineering. The research was funded in part by National Science Foundation grants.

The researchers first built the three-dimensional, finite element model to simulate bat skulls with myriad combinations of snout length and width.

Visual comparison of the shapes of micro-CT scans (light blue) and engineering (dark blue) models for the base model of the omnivorous bat Carollia perspicillata (B), and the morphed models for the nectar-feeding Glossophaga soricina (A) and the specialized fig-eating Short-faced bat Centurio senex (C).

They then analyzed the models to determine structural strength and mechanical advantage -- the efficiency and hardness of the bats' bite. Finally, they studied the engineering results across hundreds of evolutionary trees of the bats to uncover the three optimal snout shapes favored by natural selection. Selection favored the highest mechanical advantage in Short-faced bats, which gives them the high bite forces needed to pierce through the hardest figs. Nectar feeders have very low mechanical advantage -- a trade-off for having long, narrow snouts that fit into the flowers in which they find nectar.

Morphological diversity among New World Leaf nosed bats with different diets. Nectar: A) Platalina genovensium, B) Glossophaga soricina; generalists: C) Carollia perspicillata, D) Vampyrum spectrum; fig-eating frugivores: E) Artibeus jamaicensis, F) Chiroderma villosum; and short- faced bats: G) Phyllops falcatus, H) Centurio senex. Crania are not shown to scale.

"This means that even though these bats have been diverging for millions of years, we can still find the signatures of natural selection in their current diversity," says Dr. Dávalos. "This would not have been achieved if it weren't for the development of the engineering model, which allows us to assess the biomechanical characteristics of very different bats. Adaptation shaped by natural selection is the key mechanism that explains diversity, but in many cases it is difficult to find the 'signatures' of selection. This kind of engineering model may illuminate many other adaptive radiations, and the origin of so much diversity on Earth."

By coupling a flexible engineering model with analyses based upon evolutionary trees, the study opens the possibility of discovering evidence for selection in other very diverse organisms, and of reconstructing skull shapes in long-extinct ancestral species. The researchers are now applying these models to uncover why certain forms that are feasible mechanically do not exist. They are also applying new evolutionary models to find out if these limits may arise through natural selection that is not solely focused on the biomechanics.

Distribution of hypothetical species based on snout length and width. A single model was morphed to represent species within the entire space in order to calculate the engineering performance of different skull shapes.

"My goal as a scientist is to uncover the evolutionary forces that have shaped biodiversity," says Dr. Dávalos. "This work advances our knowledge by showing that natural selection can generate diversity. It also highlights the growing role of evolutionary trees in testing longstanding hypotheses on adaptation that could not be tested even ten years ago, and certainly not without the engineering model."