Speciation of Asian Drosophilids

Life on this planet is one key feature that sets it apart from other planets we know. This life not only exists, but organisms on this planet are diverse; compare a slime mold traversing a rotting log to giant elephant cactus towering the desert sands to a great white shark agilely hunting in the ocean.

How did this biodiversity as we know it come about? A fundamental part of generating biodiversity is speciation. Closely related species that have diverged into different niches were originally part of the same ancestral population. Over time, these putative species diverge genetically through neutral or selective processes. We can use these differences in their genomes to understand their divergence.

During the early stage of divergence, however, putative species may still be able to hybridise. The consequence of this would be homogenisation of the diverging genomes and a collapse or at least a deceleration of the speciation process. Yet we now know many examples of reticulate evolution i.e. introgression events in species trees (Figure 1A), and even examples where hybridisation encourages speciation such as in African cichlid fish. How does this happen?

To explore this topic, I am studying a closely-related pair of Drosophila species D. lutescens and D. takahashii. The two species are often noted for their distribution differences within Japan, with D. lutescens being found in the colder northern parts of Japan, and absent in the subtropical islands in southern Japan (Figure 1B). D. takahashii, on the other hand, is absent in the north but present in the south. The northern range of D. takahashii overlaps with the southern range of D. lutescens; at this overlap in southern Honshu, there is thought to be hybrid populations. The two species are hybridisable to F1 in the lab. We want to determine if this hybrid zone does exist and the extent of hybridisation occurring there.

I have so far generated a long read reference genome for D. lutescens, and have used this and other published reference genomes to generate a multi-species alignment to start answering questions on their speciation; when did the species diverge? What areas of the genome are more diverged than others? What is the level of introgression that we can detect between the two species?

Figure 1. (A) *Hypothetical trees, on the left we have a standard hierarchical tree, whereas on the right in green we have a similar tree with reticulation events.* (B) *Map of Japan with rough estimate of distributions of* D. lutescens *in blue and* D. takahashii *in orange. On the right there are admixture plots, we expect individuals in the north to be of* D. lutescens *ancestry, the south to be of D. takahashii, and a mix in the middle.*

Evolution of temperature-related behaviours

D. lutescens and D. takahashii experience very different temperatures in their respective ranges. D. lutescens is found as far north as Hokkaido in Japan, where winter temperatures are often sub-zero. D. takahashii is found in the subtropics, where the winter temperatures are similar to the summer temperatures of Hokkaido. Therefore, it is logical that D. lutescens is more cold-tolerant than D. takahashii, and this is indeed the case. Our interest moves away from physiology, and we would like to see if there are any behavioural differences in relation to temperature between these two species.

This is just one such example of closely-related drosophilids adapted to very different temperatures, other examples include D. santomea and D. yakuba, and D. auraria and D. rufa. D. yakuba is a fruitfly found across Africa, whereas its sister species, D. santomea is endemic to the island of Sao Tome. D. santomea is primarily found at higher altitudes where the temperature is cooler. In adult flies, D. santomea has a colder temperature preference than D. yakuba. Understanding behavioural differences in these different species pairs will let us start to understand the contribution of behavioural divergence when adapting to new niches.

Fruitfly thermal sensing is pretty well described in D. melanogaster at both the adult and larval stages. We have a good idea of the sensory proteins involved, where the sensory cells are and even where they project to in the brain. Ultimately, we would like to know what is changing between species to give differences in temperature-related behaviours.

For a first glance, I have developed an assay to determine larval temperature preferences (Figure 2). This assay consists of a continuous patchy temperature gradient that we allow larvae to explore. I have developed an analysis pipeline to extract temperature preferences from the assay. The primary measurement is the amount of time larvae spent at different temperatures during a run. We are also able to measure finer scale behaviours such as the magnitude of head sweeps, where larger sweeps indicate lower preference. We will use this assay to determine temperature preferences of late third instar larvae of several different species, including D. lutescens, D. takahashii, D. santomea and D. yakuba.

Figure 2. (A) Cartoon illustration of the arena for behavioural assays. It consists of a metal plate and four peltier elements, which either generate hot or cold. The gradient is carried on the metal plate. The temperature of the peltier is controlled by an Arduino. Assays are carried out in evenly illuminated red light. (B) *Runs of the assay are recorded by a thermal camera to measure the temperature gradient, and a camera to record larval movement (white paths).*

Other

During my master’s degree at Imperial College London, I completed a project on the divergence of two Metrosideros tree species. These species are found on a remote and minute island roughly 600km away from the coast of Australia, and I focussed on identifying the genes that enabled their divergence in the face of gene flow. Have a read of it as a part of this publication: Sympatric speciation in mountain roses (Metrosideros) on an oceanic island

I have written a small wiki on the species of the D. takahashii species subgroup, which you can have a read of by clicking this link:

Contact

Please don’t hesitate to get in touch if you want to know more about my projects, or if you would like to find out more about working in Lausanne.

You can contact me through twitter: @tane_kafle

My work email is tkafle@unil.ch and I also have a website: https://tanekafle.gitlab.io/