Evolution experiments with microorganisms: the dynamics and genetic bases of adaptation

Elena and Lenski (2003). Nature Reviews Genetics 4: 457-469. Evolution experiments with microorganisms: the dynamics and genetic bases of adaptation

Local optima can hold back a population from doing the best that they could

Local optima can hold back a population from doing the best that they could. From Elena and Lenski (2003).

Will Pearse

Will Pearse

This paper is more than a little beyond what I usually read, but I always enjoy going to bacterial evolution talks, and I enjoyed this. This is such a rapidly-changing field that I’m certain new things have come out – please do chime in!

I was very struck by the discussion of mutators. The idea that individuals with higher rates of mutation can, in some circumstances, be of benefit to the rest of the population is really cool. It also has a lot of implications for the constant arguments I seem to have with people over what ‘evolvability’ is: as far as I’m concerned, if there’s a thing that increases the rate of adaptation of a population, that’s an increase in evolvability, and mutators seem to be that. We talk a lot in large-organism evolutionary biology and biogeography about the importance of historical accident – that fast rates of mutation only sometimes increase the evolvability of a population, and only sometimes do such individuals come to dominate, is an excellent repeatable and predictable example of the problem. It’s nothing short of amazing that we can do accurate back-of-the-envelope calculations and figure out the odds of a particular outcome in a petri dish.

One thing I want to slightly temper some of this with is the importance of natural history in these populations. It’s all very well studying the evolution of bacteria in stationary phase and discussing how this is different from what we normally find in the lab, but stationary phase is not the normal state of affairs for bacteria. In the wild, bacteria are not transferred into new media when there’s too many of them, and the same goes for large-bodied organisms too. If we want to understand natural evolutionary processes, we need more stationary phase experiments (right? Am I being ignorant?). To paraphase someone else, we also spend a lot of time examining bacteria that cause diseases in particular environments (the human gut, for instance), and modelling their evolution in situ. The problem with this is that this isn’t what those species do for their whole natural history – if you don’t consider how those species got into that environment (water, soil, kitchen sink, etc.), or what some might call a ‘fluctuating environment’ then you’re not going to get the complete picture of the evolution of that species.

Lynsey McInnes

Lynsey McInnes

This is a difficult paper to discuss because I feel like I know that there have been many advances since it came out, not least in our capacity to study mutations and their effects right down to the level of single nucleotide polymorphisms and up again to the level of whole genomes. However, my knowledge of this literature is hazy at best (but see Will’s and my first attempt at discussing experimental evolution here). I’m just going to paste in the abstract below so you know what the paper was actually about (seen as my post barely touches on it…).

Microorganisms have been mutating and evolving on Earth for billions of years. Now, a field of research has developed around the idea of using microorganisms to study evolution in action. Controlled and replicated experiments are using viruses, bacteria and yeast to investigate how their genomes and phenotypic properties evolve over hundreds and even thousands of generations. Here, we examine the dynamics of evolutionary adaptation, the genetic bases of adaptation, tradeoffs and the environmental specificity of adaptation, the origin and evolutionary consequences of mutators, and the process of drift decay in very small populations.

What am I going to write about then? Well, I think going off on a tangent seems like the best idea. My background is heavy on macroecology (patterns! scale!) and my current work centres around exploiting neutral genetic variation among populations to infer demographic history (which I am increasingly realising can get quite close to macroecology when it wants to). So I am not accustomed to thinking about mutations that affect function or adaptation in beneficial or a deleterious ways. Selected genes are in fact the bane of my data. Although I remain unconvinced that we are ever confident that a locus is ever completely neutral.

In fact, I am quite jealous of experimental evolutionary biologists. It seems unfair that they are able to watch things (really, anything!) happen in real time whereas macro-scale analyses (macroevolution, macroecology, phylogeography, biogeography) rely on sometimes shaky sets of assumptions, occasional blind leaps of faith and (more often than not) bundling a lot of unexplained variation into historical contingency. I am a big fan of comparative and meta-analytical approaches (which I’ve advocated on many a PEGE post) where generalities can emerge on diverse topics such as prevalence of niche conservatism, latitudinal richness gradients, modes of trait evolution, even of community assembly, but there is always the niggling doubt that contingency gets in the way and overrides any signal. Wouldn’t it be great if we had five Caribbeans and could throw on five identical Anolis clones and watch what happened? Bacterial experiments can do this! So jealous!

What I’d like to see is more engagement between experimental evolutionary biologists and macro-people. How close can we get to equivalent situations? Can we apply our macro approaches to bacterial setups and see what we find? What is and is not transferable? How much does asexuality mess things up? How much does single cellularity mess things up? I am quite sure that there must be some theoretical exploration of these ideas, but I’d like to see more cross-talk among empiricists. Especially now that we can sequence anything (especially easily bacteria), let’s find out more about how comparable these systems are and what the next stage might be?


How does ecological disturbance influence genetic diversity?

Banks et al. TREE 28(11): 670-679. How does ecological disturbance influence genetic diversity?

How disturbed are your genetics? From Banks et al.

How disturbed are your genetics? From Banks et al.

Will Pearse

Will Pearse

Disturbance is a topic very close to my heart (that’s meant to be a physiology joke), mostly because I get very annoyed when people don’t define precisely what they mean by it. So I was very heartened to read this review, where the authors discuss the various temporal and spatial scales of disturbance, and also because it’s a very nicely written paper.

Disturbance, within certain conditions, can be part of the background homogeneity of a system, and the authors are keen to stress that in this paper. I was a little surprised to not find mention of the intermediate disturbance hypothesis (even though some find it controversial), since it’s so appropriate in this context. I found figure 1 (partially reproduced above), where the authors go through some case studies of what different kinds of disturbance look like, quite helpful in reminding me that disturbance can be lots of different things, and it can have lots of different effects (not always bad). However, that figure 1 is made up of case studies reflects our lack of a coherent framework to structure how we think about disturbance. Moreover, the right hand side of the figure (which I cropped out, sorry!) talks about two case studies that involve “metapopulation” and “patch dynamics”; this makes a lot of intuitive sense to me, but on reflection I find that kind of weird. Metapopulation theory is a concept humans have generated, it’s not a thing that biological systems recognise, and I think it might be better to categorise systems on the basis of properties they share rather than how we find it easiest to model them.

So what would such a categorisation look like? After reading this paper I think disturbance severity, duration, and extent (bear with me) are three important axes. With ‘extent’ I want to incorporate the ability to temporally and spatially escape a disturbance; spatially means whether the disturbance is everywhere and whether you can move to avoid it, and temporally that means whether the disturbance happens very often or very infrequently and would probably incorporate seed bank effects. I’m sorry ‘extent’ is such a poor descriptor; I’m decaffeinated and would appreciate better suggestions! I’ve very deliberately chosen to put space and time on the same axis; you might prefer to split them. You might also prefer to add predictability as another axis; I don’t, not because I don’t think it’s important, but because I think a system’s history (which, in turn, incorporates predictability) affects quite a lot and the other axes mostly capture what the system has been doing in the past. Not a lot about genetics in this post (sorry!), and instead a framework that almost certainly already exists somewhere and I’ve forgotten I’ve read it. Please do tell me where!

Lynsey McInnes

Lynsey McInnes

I had high hopes for this paper. I’m attracted to any paper that deals with intraspecific variation head-on and am well aware that intraspecfic variation affects and is affected by processes occurring on varying spatial and temporal scales. So, a paper dealing with how disturbance affects genetic diversity seemed right up my street. I was curious about the direction the paper would take as my feeling was genetic diversity is generally quite hard to measure particularly in non-equilibrium populations (such as those that have been disturbed) and assigning particular genetic signatures to historical events (‘disturbances’) is notoriously difficult as not only can a range of different events leave the same genetic signature, the same event can leave different signatures depending on the ecology and population structure of the species involved.

It was good for my ego to find that the authors largely confirmed my suspicions of these issues, but sad for the paper that there seems no easy way out.

It seems that the current state of understanding is that we live in an increasingly ‘disturbed’ world . Events such as tsunamis, fires and grazing impact nearby populations, reducing the number of individuals and thus most likely (at least) point estimates of genetic diversity and the challenge is to recognise the types of populations/species that will find recovery from such impacts difficult or impossible (if one is interested in conserving viable populations, otherwise all impacting populations are interesting, for instance, what kinds of species can you bombard with disturbances and they bounce right back to pre-disturbance levels of abundance and genetic diversity?). It seems however, that little research has focussed on the relationship between disturbance and genetic diversity and that there are many outstanding questions.

The second half of this paper gave a helpful overview of these outstanding questions and laid out some helpful ways forward. Namely, and understandably, the integration of multiple sources of data (event type, species’ traits, samples across the range and through time, etc.) will help to unravel the impact, or non impact, of putative disturbances on genetic diversity and, more importantly, what these effects mean for the longer term survival of species and/or communities. In fact, the paper lists FOURTEEN outstanding questions linking disturbance and genetic diversity and all of these are interesting. It would have been nice if these had been dealt with in more detail in the paper, perhaps focussing on a couple and on real routes forward to addressing them.

Maybe I missed this in the paper, but I also felt that what was missing was strong evidence that one expects any general link between disturbance and genetic diversity. As next gen sequencing gets cheaper and more accessible for non-model organisms, it will become trivial to look for these links, but, I feel, we need to know what we are looking for before we go looking for it. The general view is that more genetic diversity per population is better to ensure buffering against a variety of disturbances, but the authors show this is not always the case. Individuals can come from beyond the disturbance centre to make up for lost individuals and/or diversity. To predict this rescue effect one has to have a bigger picture encompassing knowledge of the genetic diversity of multiple populations within and beyond the disturbance centre. Are there enough individuals for recovery and do these individuals possess the desired adaptations? (So, I might differ from Will in thinking metapopulation theory might be helpful here).

I absolutely believe that intraspecific variation within and between populations in terms of genes and ecology must be considered if we hope to understand how populations will cope in the face of point disturbances and longer term environmental fluctuations. This paper drove home to me quite how difficult this endeavour is going to be.

%d bloggers like this: