Evolutionary Ecology

For me, the most important word in the biological sciences is: divergence. I like that word because it suggests the beginning of evolution and also the relentless continuation of it. Divergence can refer to speciation but is ongoing before, during, and after the speciation event. It can refer to morphology, ecology, or behavior, without having to progress all the way to speciation. It can be observed at the level of genes, genomes, populations, taxa, or even biota. Divergence is both a pattern and a process. The result of divergence is diversity and that is usually the starting point of my research.
But the divergences that I am most interested in are the ones that happen in ecological time. By that I mean, I like to study evolution in action, as it happens between sister species, populations, or even individuals sharing the same habitat.

I wouldn’t call myself an evolutionary biologist, but I’m not an ecologist either sensu stricto—I’m something in between. I am a marine evolutionary ecologist, and I use molecular tools to study divergence in marine organisms (mostly fish).

The genomic basis for reproductive barriers

A) Naso hexacanthus B) Naso hexacanthus nuptial coloration. C) Naso caesius (bottom) and Naso hexacanthus (top) D) Naso caesius nuptial coloration.

An important question is: what processes maintain divergence? The two unicornfishes, Naso hexacanthus and caesius, are close sister species. So close, in fact, that they are often observed foraging in mixed schools, and even mixed spawning aggregations. Moreover, they share a lot of the same genetic variants in both nuclear and mtDNA (Horne et al. 2013. J. Evol. Biol.). So what keeps these fish from mating together and eventually becoming one species?

One clue is that both species have specific “nuptial coloration.” These visual patterns probably help each species reproduce with like mates, even though they spawn at the same time, in the same place. Some hybridization likely occurs, but the hybrids may not be as fit as purebred individuals… otherwise there would have been no need to evolve such nuptial coloration. N. hexacanthus in particular has some of the most striking nuptial color changes in any fish, and my hypothesis is that this evolved as a form of enhanced isolation.

My ongoing research continues to look into the genomic processes that maintain species boundaries.

Demographic genetics

Population genetic theory suggests that demographic processes, such as competition, can drive divergence when limited habitat and/or resources exclude migrants. But empirical examples of this are few and somewhat controversial because intra-specific competition is hard to demonstrate conclusively.

However, one hallmark of competition is contrasting population growth trajectories among different lineages, which can be inferred from genetic data.

A good system for investigating this are anemone fishes.


Anemone fishes are sequential hermaphrodites—starting out as males and then becoming females later in life. But only the oldest fish in each anemone gets to reproduce as a female… which means that only the oldest and most successful individuals will get to pass on their mtDNA to offspring.

So do mitochondrial lineages show contrasting population growth trajectories in populations of anemone fishes?

Sort of.

Quartz %d
Significantly negative Tajima’s D values indicate population expanison, as do significantly low R2 values

If you just look at the above haplotype network you might be convinced that I’m onto something, but not every population in this species looks so clear cut. Also sample sizes are a bit too small to be confident (clade II only has six individuals).

Capturing a true signal of competitive exclusion will depend on getting the spatial and temporal scale of sampling just right, and having enough data to draw robust conclusions.

This is still a work in progress.

Comparative genomics meets population genetics

There are many different kinds of genetic divergence. As a molecular ecologist, it has been most easy to look at changes in nucleotide composition, or length polymorhpism in microsatellite loci. But recent studies suggest that large structural genomic variation is where the action is (see Berg et al. 2016. Nature Scientific Reports – DOI: 10.1038/srep23246).

Finding large structural genomic rearrangements lies in the realm of comparative genomics, but like others are starting to do I want to study comparative genomics in the context of population genetics.
For me, the taxon of choice for this type of research would be pufferfishes.
Canthigaster valentini
Moorea, French Polynesia. Photo credit: Giacomo Bernardi

Not only do pufferfishes have small compact genomes (an eighth the size of a human genome) with little non-coding DNA, and two model reference genomes in the family, they are also one of the most rapidly diverging groups of teleosts (Santini et al. 2016. J. Evol. Biol.). Canthigasterine pufferfishes (above) in particular have undergone recent evolutionary radiations in coral reef environments.

Canthigasterine pufferfishes are to coral reefs what poison dart frogs are to Neotropical rain forests. There are a number of divergences in this family that could be investigated using this group. And it would be interesting to compare their ecology to that of their terrestrial counterparts that are much better studied.

Expression quantitative trait loci

What would be even cooler than doing a genome-wide association study with ecological phenotypes? A genome-wide association study with gene expression phenotypes.

Back when I was a student at Johns Hopkins, I took a class where we did this, we associated 300,000 human SNPS with RNA-seq data from the same individuals. The result: we were able to identify quantitative trait loci for expressed genes. In other words, these SNPs were associated with promoters and enhancers and other regions of epigenetic significance.

When I first did this I knew that someday I’d have to do it on fish. EQTL analysis would enable us to link genomic and epigenetic variation. EQTL would enable us to understand the genomic basis for phenotypic plasticity. And that is particularly important if we want to understand how flexible species are in their tolerances to a changing environment.

The only problem is that for statistical power you need lots of loci and lots of individuals in your sample. This can get expensive. But hopefully someday I’ll be able to scrounge up a grant to do work in this area.