Columns Conversations

On evolutionary biology, and a passion for science: Interview with Amitabh Joshi

Anjali Vaidya

Thoughts on experimental evolution, problem-solving and how to pursue science with passion.

Amitabh Joshi
Amitabh Joshi  (Photo: Amitabh Joshi)

How did you become interested in evolutionary biology?

While I was doing my BSc Honours in Botany at Delhi University, I found genetics very interesting, because it brought back many of the things that I had liked about math and physics. There was a lot to understand, rather than just a lot to memorize. I remember being particularly impressed with how Jacob and Monod worked out the operon. I still think that’s one of the most beautiful things in genetics.

So after my BSc I applied for admission to MSc genetics at Delhi University. We had a course in population genetics the first year, taught by Professor C R Babu. He was quite simply the most amazing teacher I’ve ever had in my life. Many of the things he said in class I can still remember after almost thirty years.

Population genetics was just beautiful — it was cute, it was lovely. I really liked it. And so I decided to go for a PhD in evolution, and ended up working with Larry Mueller at Washington State University.

There was a certain amount of contingency in the choice of subject. I could as well have ended up a professor of Urdu literature or philosophy. I was reasonably clear that I wanted to be in academics. I couldn’t then and I can’t now imagine being in any other profession. 

Can you describe some highlights of your current research?

The approach that we take in our lab is called experimental evolution. Instead of using an existing species to infer what might have happened in the past, you work with an organism that allows you to observe several hundred generations within a few years.

What we do is set up evolutionary problems for populations of fruit flies to surmount. For example, we took one set of populations and said that only those individuals that become adults at the fastest speed are allowed to breed for the next generation. After seventeen years and 600 generations, these populations are the fastest developing line of Drosophila melanogaster that anybody has ever seen. 

When we asked what traits were sacrificed for the sake of developing fast, we found that these flies areworse than their ancestors at surviving to adulthood. We also found that faster developing flies preferentially burn lipids over carbohydrates as pupae. This gives them more energy per microgram, but it depletes their lipid reserves, which reduces fertility. In these populations you have to first be among the fastest 20% to emerge, and if you burn up your fat reserves and lower your chance of survival to develop quickly, you will probably still have higher fitness. 

Our recent work follows up on our observations that different competitive strategies can evolve under conditions of high larval density. Our evidence suggests that when a culture vial’s food column is very short, the build up of waste means that even though food is scarce, feeding faster will not help larvae survive. We are developing our results into a broader view of competitive ability that emphasizes the interaction between environmental context and larval density in the determination of density-dependent fitness. 

We have also shown that density-dependent selection can lead to the evolution of more stable population dynamics. In collaboration with Dr. Sutirth Dey (IISER Pune) we are trying to develop integrated models of the evolution of both density-dependent fitness and population dynamics. These results represent the first major conceptual advances in density-dependent evolution and population dynamics since Larry Mueller’s seminal work in this field in the 1980s and 1990s.

What is the most satisfying research problem you’ve worked on?

When I was a grad student, I became interested in the so-called cost of sex problem in evolution. I found that there was a huge debate that had been going on for twenty years in the literature about what the cost of sexual reproduction really was. John Maynard Smith argued that it was the cost of producing males, while G C Williams argued that it was the cost of genome dilution. 

I remembered from my BSc that dandelions and other asexual plants produce partially sterile pollen. But in all the models of the cost of sex, people assumed that plants that produce asexual eggs either do not produce pollen, or they produce pollen in equal amounts and with equal fertility as sexual individuals. So I built a model, which would look between these two extremes, treating the output and fertility of male gametes as separate variables. It very quickly turned out that the cost of sex has both components — cost of male function and the cost of genome dilution. If you model only these two extreme cases, you will either find that the cost of sex is due to the cost of male function or that it is due to the cost of genome dilution. Between extremes, there are both components to the cost of sex. 

This work eventually grew into several papers with Professor Mike Moody in the Journal of Theoretical Biology. In terms of satisfaction, this is one of the most satisfying pieces of work that I have done. It solved an old problem in a very clean way, by showing that what you are seeing as a problem is the fact that you’re not looking at the whole picture — you are just looking at two extremes of the spectrum.

In some sense the work was all the more rewarding because there was no need for me to do it. It wasn’t part of my thesis work, or work that somebody had told me to do. It just grew organically out of something interesting that I studied in a course, and then led to eventually something very nice.

What advice would you give to people starting out in science?

If you do your work with a certain passion, your career will take care of itself, and you will enjoy yourself. When you work, the rest of the world should stop. And I don’t mean that you should work 365 days a year. You cannot plan that Tomorrow morning from 9:30 to 10:30 am I will think of a novel hypothesis.” All my life, I have never been an organized, steady worker, but when I work, at that time I’m completely focused on what I’m doing — whether it’s writing a manuscript or analysing a particularly recalcitrant data set. At that point in some sense the rest of the world ceases to exist. 

Nowadays, people view science too much as a career. I wish people would view science the way a classical musician views music. Of course singers earn money by singing, but they don’t think of it as a profession the way working in a bank is a profession. If your only aim is to publish in high impact journals, that can give you some measure of success within the parameters of the system. But I feel there are much more interesting ways of getting one’s kicks in science. 

What are the most frequent misconceptions that you encounter regarding your work? Which questions do you dread?

The question I don’t like, but get quite often, is how is your work important to the upliftment of human society. Science is seen too much through utilitarian lenses. The purpose of science is to understand. Harnessing that understanding to practical use requires a very different mind- and skill-set. 

The most frequent misconception nowadays is that every causal explanation in biology must coalesce to one or a few genes. This position is brainwashed into students by the time they come to grad school and does a great disservice to biology.

Can you talk a little about your own approach to mentoring?

Mentoring, to me, is the most important thing any of us ever do as academicians. Mentoring is the essence of our existence as a link in the continuing chain of human knowledge. It is not about giving a grad student a good problem to work on: it is about helping the student discover his or her own good problem. As Khalil Gibran said so beautifully: If he (i.e. a teacher) is indeed wise he does not bid you enter the house of wisdom, but rather leads you to the threshold of your own mind.” 

Ultimately, mentoring is about sharing. You share the logic of how you approach a problem, probing for the right fault line at which to attack. Too often, research supervisors just tell students what to do without really explaining why, and if the student asks why, it becomes an ego problem for the supervisor because it is interpreted as having their judgment questioned. People need to be encouraged to discover their own style of doing science. It is important to transmit to the mentee a whole world view, rather than just a set of scientific techniques. 

Alongside evolutionary biology, you have found time to write Urdu poetry and study Indian history. How do your other pursuits feed into your research? How do you balance time?

I don’t think poetry feeds directly into science, or vice versa. Perhaps, at a more transcendent level, the spirit of poetry informs the spirit in which I do science. Science has its own beauty — there are models and formalisms that are astonishingly beautiful, as well as experiments that are beautiful in a way that makes you go vaah as if you had just read or heard a superb couplet. 

I don’t balance time well. I cannot do things like say I will work on a manuscript before lunch and write a poem or read philosophy after lunch. I work episodically on all things, whether science or other interests. When I am in the frame of mind for something I become immersed in it, to the exclusion of other things. 

What are your most and least favourite parts of your job?

Most favourite: teaching. Least favourite: listening in meetings to self-important people pontificate about what ails Indian science.

What is the best advice you have ever gotten?

From my mother: Jo kaam karo, theek se karo, ya phir mat karo. If you do something, do it well. Otherwise don’t do it at all.

Written By

Staff Writer, IndiaBioscience (March-August 2015)