This post is by MSU grad student Emily Dolson
Imagine that an alien species arrives on earth. It happens to be able to live and reproduce in any part of the world, and, over successive generations, it begins to adapt to its new environment. Among other things, it adapts to eat new kinds of food. Where in the world would you expect the first alien capable of eating papayas to be born?
Most people would probably say South or Central America, since papayas live there. When asked why they said that, there are two potential explanations they might give:
- The ability to eat papayas is more beneficial in regions where papayas actually grow.
- Regions where papayas grow are also populated by similar plants, like guavas.
The first of these explanations is based on either a misunderstanding of the question or a misunderstanding of evolution (but don’t worry, it’s a common one!). We didn’t ask where aliens that eat papayas are most likely to survive long term; we asked where the very first alien that could eat a papaya would be born. Recall how natural selection works: mutations occur randomly, some happen to be useful, and organisms that have mutations that happen to be useful usually go on to be successful. The first alien to be born with the ability to eat papayas has presumably had some sort of mutation that (in the context of the rest of its genome) gave it this ability. That ability hasn’t yet had a chance to prove to be useful or not. This alien may never even encounter a papaya in its life, in which case having the ability to eat them will have no effect. The presence or absence of papayas can have no influence on the birth of the first alien that can even interact with them.
The second explanation, however, is plausible. South and Central America are home to other tropical fruits, such as guavas. Tropical fruits have similar physical and chemical properties. If an alien can eat a guava, it probably only needs a few subtle mutational tweaks to be able to eat a papaya (assuming that papaya-eating even requires further adaptation). Once a subset of the aliens in the tropics gain the ability to eat some sort of fruit, their progeny will likely go on to be very successful in the regions where these fruit grow. These aliens will be well-positioned, in both physical and mutational space, to begin eating other fruits, such as papayas.
While aliens are unlikely to invade earth and eat our tropical fruits, the phenomenon of a population encountering a set of entirely new challenges as it moves into a new geographic region (or as the geographic region in which it currently lives changes) is common. And it’s common for these new challenges to be related to each other. It’s no accident that plants with similar fruit live in the same region; plants in the same region experience the same selective pressures and may also have a shared evolutionary history. These factors generalize across a wide variety of scenarios. Thus, if some of the challenges of thriving in a new area are easier to solve than others, the easy ones can serve as “evolutionary building blocks” for the harder ones. The presence of such building blocks, i.e. simple adaptations that provide a good starting point for more complex adaptations, has been shown in previous work to be important to the evolution of complex traits (Lenski et al., 2003).
If spatial layout does indeed impact the ease of adaptation, it would be useful to understand, both for evolutionary biology and evolutionary computation. As species shift their ranges in response to climate change, they will traverse regions where different traits are advantageous. Predicting how the positioning of these regions impacts evolution will help us predict whether the species will be able to survive. Evolutionary computation often takes advantage of evolutionary building blocks by rewarding solutions to different problems over time, but this is an imprecise art. If we could understand how to reward them differently across space to promote evolution of an overall solution, we could more easily generalize evolutionary computation to more problems.
Of course, it’s also entirely possible that these spatial effects are too small to care about. Recently, I’ve been trying to figure out whether or not that’s the case (Dolson and Ofria, 2017). Since this would be an incredibly labor intensive question to address in the lab or field, I’m using the Avida Digital Evolution platform to perform preliminary experiments on the computer. Once we know more, I’d be very interested in collaborating with wet lab biologists to see if our digital results are consistent with results from DNA-based systems.
In Avida, I created eight different environments with different patterns of resources across space. I then let 100 different populations of digital organisms evolve independently in each environment. Within each of these runs, I found the location of the first organism with the ability to use each resource in each run. From these data, for each resource and each environment, I determined which regions appeared more often then we would expect to see by chance. These regions are “evolutionary hotspots.” Sure enough, each environment (except the control, which had all resources everywhere) had at least some hotspots (see Figure 1). In some environments, most of the hotspots overlap. In others, they are largely in different regions of the environment.
I’m now working on trying to predict why hotspots are where they are. Surprisingly, a number of seemingly obvious explanations (e.g. the number of resources present, the presence or absence of specific resources, and local diversity) do not appear to explain the pattern. Currently, it appears that the sequence of environments that a lineage experiences over evolutionary time may be a key variable (see Figure 2). I’m looking forward to understanding more soon!
Lenski, R. E., Ofria, C., Pennock, R. T., & Adami, C. (2003). The evolutionary origin of complex features. Nature, 423(6936), 139-144.
Dolson, E. and Ofria, C. (2017). Spatial resource heterogeneity creates local hotspots of evolutionary potential. In Proceedings of the 14th European Conference on Artificial Life. Vol. 14. pp. 122 – 129. MIT Press.