Monday, April 13, 2015

Are laboratory studies useful for understanding the “real world"?

“It is interesting to contemplate a tangled bank, clothed with many plants of many kinds, with birds singing on the bushes, with various insects flitting about, and with worms crawling through the damp earth, and to reflect that these elaborately constructed forms, so different from each other, and dependent upon each other in so complex a manner, have all been produced by laws acting around us.” (Darwin 1859)
Some of what we hope to understand.
I just returned from an “eco-evolutionary interactions” workshop at Yale organized by David Post, David Vasseur, and Paul Turner. In his opening introduction to the symposium, Paul referred to research from the “real world” when speaking about something of applied relevance to humans. He then hesitated for a second and took a tangent to complain about the term “real world” as all of the work was in the “real world.” (The Turner World?) A bit later I was talking with Alvaro Sanchez about the value of lab experiments for eco-evolutionary studies – a different sort of “real world” concern. Alvaro was asking my opinion partly because he had seen an advance copy of my book in which I state:
Eco-evolutionary studies with real organisms could proceed in the laboratory or in nature. Advantages of the laboratory are manifold: populations can be genetically manipulated, environments can be carefully controlled, replicates and controls can be numerous, and small organisms with very short generation times (e.g., microbes) allow the long-term tracking of dynamics (Bell 2008, Kassen 2014). These properties dictate that eco-evolutionary studies in the laboratory are elegant and informative, yet only in a limited sense. That is, such studies tell us what happens when we impose a particular artificial environment on a particular artificial population and, hence, they cannot tell us what will actually happen for real populations in nature. Understanding eco-evolutionary dynamics as they play out in the natural world instead requires the study of natural populations in natural environments. I will therefore focus to the extent possible on natural contexts, although I certainly refer to laboratory studies when necessary.

Is this place to do it?
This exchange reminded me of a debate Andy Gonzalez and I had undertaken for the amusement of our department, and it seemed an appropriate time to revisit my side of that debate in this blog. Specifically, I would like to ask “What – if anything – can laboratory studies tell us about real world?”

What we seek to understand in any eco-evolutionary work is the chain of effects shown below: the effect of ecological conditions on organisms and the evolutionary/ecological outcome. The way laboratory studies typically address this question is to create variation in some focal ecological condition, such as temperature or carbon dioxide (or both), expose some model organisms (often only just one or a few identical clones) to those different conditions, and then measure the evolutionary/ecological outcome. The real world is much more complicated, of course; populations or species of interest might be exposed to differences in a great diversity of biotic and abiotic factors and might have large amounts of genetic and non-genetic variation in a wide array of relevant traits. These differences lead to two main limitations of laboratory studies.

1. By ignoring variation in other natural factors, laboratory studies cleverly isolate particular effects, but those effects are likely to be very different in nature.

These differences can arise in at least three ways. First, correlations can exist between explanatory variables that fundamentally shape the effect of each. Second, indirect effects cascading through other organisms or environmental conditions can amplify or offset the direct effects usually measured in laboratory studies. Third, unmeasured contributors or random noise can swamp causal effects. Interestingly, these complications are all cited as supporting the great strength of laboratory studies: by getting rid of these correlations and effects, one can get rid of the noise and get closer to the true mechanism. Sure, this is all well and good – but such studies don’t then tell you anything about the importance and nature of the actual effects IN NATURE (which is what we really care about). Instead such experiments only provide proof of concept.

2. The use of model (as opposed to natural) populations is sure to yield a misleading expectation.

Evolutionary responses of a given suite of traits will be a function of selection acting directly on the genetic variation for each trait and selection (and constraints) acting indirectly through genetic covariation with other traits. Unfortunately, genetic (co)variances in nature never equal those in nature, meaning that the observed evolutionary response will not match the response that would emerge in nature. Moreover, many laboratory studies use only a single clone of organisms so that they can assess the effects of new mutations. Wonderful – but most immediate responses to environmental change in nature will be driven by standing genetic variation, which that approach eliminates.

Similarly, ecological responses to changes in a given suite of traits will be a function of the distribution of traits in the population and the effects of that variation on ecological variables. Sadly, the distribution of traits in the lab will (for the above reasons) never match the distribution in nature. In addition, the simplified laboratory environment will never capture how a given trait distribution will shape ecological responses in more realistic scenarios.

Having just argued from first principles that laboratory studies will not be predictive of evolutionary or ecological effects in nature, I will provide a few examples by way of illustration.  

Amphibians are a classic system for studying the effects of competition. Skelly and Kiesecker (2001 – Oikos) performed a meta-analysis of how the experimental “venue” (lab, mesocosm, field) influenced estimates of how competition influences amphibian growth. Based on 227 comparisons from 52 studies, the authors showed that the effects of intra-specific competition (white bars) were reasonably consistent across venues but the effects of inter-specific competition (black bars) were dramatically different between venues. Results from the laboratory from mesocosms and from experiments in natural populations all yielded very different results.

Many laboratory studies have shown that exposure to novel parasites causes the evolution of resistance to those parasites. Similarly, releasing populations from parasite pressures leads to the evolution of reduced resistance, presumably because resistance is costly. We tested the latter expectation in nature through assays of resistance to a common parasite (Gyrodactylus) in guppies from a highly parasitized population that was introduced into four replicate environments that lacked the parasite. We found – completely contrary to lab experiments – that all four of the introduced populations rapidly evolved INCREASED resistance to the parasite they no longer experienced. Presumably this result arose because other factors in nature caused selection on other traits that were pleiotropically coupled to resistance. The lead author (Felipe Dargent) wrote about this work in a previous post and the work has since been the subject of a comment-response that specifically contrasted the complexity of nature as a weakness or strength of experiments in nature.

The populations not exposed to parasites (LL, UL, T, C) are now MORE resistant to the parasite than is the ancestral population (S) that remains exposed to parasites.
The position I have taken in this post is: if we are to understand nature we must work in nature. Of course, this doesn’t mean that we shouldn’t also work in the lab. In particular, laboratory experiments provide critical proof of concept that a given ecological factor can have a particular evolutionary effect under a given set of conditions. Similarly, they provide proof of concept that a given evolutionary change can have a given ecological effect under a given set of conditions. The critical next step, however, is to then test key aspects of those outcomes in the messy natural world.

So – in the end – the study of a given eco-evolutionary effect would ideally couple laboratory and field (and theoretical) studies. Reassuringly, this was a major theme of the working group – designing questions and experiments that can be explored simultaneously in theory, in the lab, and in nature. I look forward to seeing what emerges.

The crew!

Wednesday, April 1, 2015

Jokes Go In The Acknowledgements - or Anywhere on April 1

Much ink has been spilled over the dry and boring nature of scientific writing. This style stems, of course, from the desire to communicate information in the fewest possible words with the greatest possible clarity. I generally agree with this goal and sentiment. At the same time, however, the quality of the writing itself can increase the appreciation of science and make you WANT to read that paper that is sitting there on your desktop. 

So how best to accomplish this balance between clarity/brevity and an appreciation of the writing itself. Two approaches are common. First, in what we might call the "subtle" (or even subversive) approach, authors outwardly adhere to the formal scientific style but sprinkle fun details into key parts, most commonly the title, the figures, or the acknowledgements. Second, in what we might call the "overt" approach, authors throw out the rules and find a good venue to just write a fun paper from start to finish. In the spirit of Steve Heard’s wonderful “On Whimsy, Jokes, and Beauty: Can Scientific Writing Be Enjoyed,” I wanted to here provide some examples of these approaches, both my own attempts and the successes of others.

In the subtle approach, jokes and whimsy often appear in titles or figures. The title option is most common, where one simply adds a colon between the serious and funny parts of the title. This approach is very common and sometimes works and it sometimes doesn’t. Here are a few examples and I will let you judge which work well and which would best have been avoided.

The good and the bad of colon-based title jokes.
The figure option is also somewhat common, and I have already posted a “Top Ten best scientific figures” to outline some great examples, with a few shown here. 

Probably by far the most common way to sneak some surprises into a paper is in the acknowledgements, as these are either not carefully read by editors or simply a place where they are willing to allow some lee-way. Posts have already been written about how acknowledgements are ways to insult or criticize reviewers or colleagues – but they are also good places to follow your whimsy muse. Here are some of my own attempts.

From a paper in Conservation Genetics in 2000, titled "Questioning Species Realities"

From a paper in Journal of Evolutionary Biology - the unnamed Deciding Editor with the great sense of humor was Rhonda Snook.

In the subtle approach, one has to hope that either the reviewers or editors don’t read carefully enough to catch your jokes – or that they catch the jokes and have enough of a sense of humor to allow them to continue regardless of you thumbing your nose at convention. Sometimes it works and sometimes it doesn’t - an instance is noted below for a "Whither Adaptation" paper I published.

In the overt approach, one really needs an editor who buys into the whole thing by either allowing you to produce a joke paper from start to finish or by allowing you throw out the norms of writing even if the paper and topic itself is not a joke. Even serious journals will sometimes allow joke papers to be published, with three examples given below.

Probably the best tongue-in-cheek paper ever writen.

For serious papers that are written in an unconventional way, whether tongue in cheek or self-mocking, you often have to get out of the serious scientific journals and into “lower-tier” journals or journals that allow some “philosophy.” Here are a couple of examples. 

Much of the paper was written in this style but we also wanted to add an asterisk to the author names with the indication "each author things to other contributed less." However, it was excised despite repeated requests.
Of course, another solution is to publish your paper on April Fools, such as this paper published today in Nature.

Happy writing!

Friday, March 20, 2015

Eco-evolutionary Island Biogeography

(This post is by Tim Farkas. I am just putting it up. Andrew)

If pressed to list the most influential paradigms of the last century, few ecologists would forget MacArthur and Wilson’s (1967) theory of island biogeography, and even fewer would intentionally exclude it. Since its inception, island biogeography has served as a powerful neutral model to explain patterns of biodiversity across space, and has seen its share of modifications, amendments, and criticisms, as befits any theory so prominent and enduring. Even when challenged to compete with metacommunity theory – its freshly minted and highly comprehensive contemporary – island biogeography performs surprisingly well, boasting a highly intuitive framework within which to develop new theory.

The late Robert H. MacArthur (left), a more recent E. O. Wilson, and their iconic 1967 book.
Perhaps it was only a matter of time, then, as the eco-evolutionary synthesis came into focus, that island biogeography was reconsidered in the light of rapid and ecologically powerful evolutionary dynamics. In the March issue of Trends in Ecology and Evolution, my colleagues and I (Farkas et al. 2015) point out that habitat area and isolation, central variables that influence equilibrium species richness in island biogeography theory (MacArthur and Wilson 1967), can also influence the fundamental processes of ecological genetics: gene flow, mutation, genetic drift, and natural selection. Hence, while area and isolation determine colonization and extinction rates through neutral processes, they can also cause rapid evolution.

MacArthur and Wilson’s (1967) original model. Small islands experience higher rates of extinction than larger islands, and isolated (far) islands experience less colonisation than well-connected islands (near), driving differences between islands in species richness at equilibrium.
We go on to lay down a general framework for how rapid evolution itself can influence equilibrium species richness, though effects on colonization and extinction. The key is that different evolutionary processes either promote or break down local adaptation, so habitat isolation and area can determine the location of a species on the “(mal)adaptation” continuum (Hendry and Gonzalez 2008). Building off the trophic theory of island biogeography (Gravel et al. 2011), (mal)adaptation in a single species can ripple through food webs and impact community-level patterns of colonization and extinction, ultimately influencing species richness at equilibrium.

The take-home: habitat area and isolation can have effects on equilibrium species richness mediated by both ecology and evolution, and those effects might reinforce or oppose one another.

Gene flow is perhaps the best process with which to illustrate eco-evolutionary island biogeography, because it is dependent on dispersal, which is heavily influenced by habitat isolation. A highly isolated habitat is expected to have low species richness at equilibrium, because colonization events will be rare, relative to a well-connected habitat (MacArthur and Wilson 1967). However, isolation will also reduce gene flow. Gene flow can have a diversity of consequences, but if gene flow is strong, and comes from populations locally adapted to divergent habitats, it breaks down local adaptation. Supposing gene flow causes local maladaptation, what may be predicted for colonization and extinction throughout the community? It depends on the role of the maladapted species in the food web. If maladaptation in a generalist pollinator reduces its abundance, the likelihood of extinction for plant mutualists might increase, reducing their species richness. On the other hand, if maladaptation reduces the abundance of a dominant consumer (e.g., a granivorous rodent), it could increase the species richness of competitors (other rodents).

In one of the examples above, the effects of isolation on species richness mediated by ecology and evolution oppose one another. In an extreme, where the evolutionary effect outweighs the ecological effect, isolated habitats could in theory have higher species richness than well-connected habitats, at least for particular guilds. This outcome can be illustrated using MacArthur and Wilson’s (1967) equilibrium figure of crossing extinction and colonization curves.

Reprinted from Farkas et al. (2015). Notice, in our extension of island theory, how an opposing influence of (mal)adaptation can lead to inverted predictions compared to island theory (arrows), such that highly connected patches have lower species richness at equilibrium (bottom). C = connected, I = isolated.
We draw three primary conclusions in this article. First, eco-evolutionary dynamics research would benefit from the explicit inclusion of gene flow, mutation, and genetic drift alongside natural selection. Second, a study of (mal)adaptation is likely a profitable means by which to accomplish that goal.  Third, rapid evolution can have a strong influence on species richness, and in particular can modify the core predictions of island biogeography. We hope this perspective encourages evolutionary ecologists to focus on (mal)adaptation and to incorporate evolutionary dynamics into their studies of biogeographic patterns. 


Farkas, T. E., A. P. Hendry, P. Nosil, and A. P. Beckerman. 2015. How maladaptation can influence biodiversity: Eco-evolutionary island biogeography. Trends in Ecology & Evolution 30.

Gravel, D., F. Massol, E. Canard, D. Mouillot, and N. Mouquet. 2011. Trophic theory of island biogeography. Ecology Letters 14:1010–1016.

Hendry, A. P., and A. Gonzalez. 2008. Whither adaptation? Biology & Philosophy 23:673–699.

MacArthur, R. H., and E. O. Wilson. 1967. The Theory of Island Biogeography. Princeton University Press, Princeton, NJ.

Monday, March 16, 2015

How to Succeed in Graduate School - Part 2

Almost immediately after posting “How to Succeed in Graduate School” 10 days ago, I started receiving comments reminding me of other tips that I should have included. Now 1300+ views later, the original post clearly has to be Part 1 and I here provide Part 2. (Don't worry, I won't test your patience further with a Part 3.) As in Part 1, this post applies most directly to students who wish to make a career of research in academia or, to some extent, in government/industry/NGO. In addition, all of the suggestions apply to PhD students, whereas only some apply to MSc students. Finally, my experience - and therefore advice - relates most directly to students in ecology and evolution, although I am sure much of it applies more broadly.


Driven in part by the requirements of committee meetings and qualifying exams, students try to map out their thesis in precise detail: chapters, papers, time lines, sampling/experimental design, stats, etc. Doing so is all well and good, but 12 years of experience on student committees has made clear that such plans are NEVER realized. I would suggest a crude estimate that  only 1-2 proposed chapters actually make it into a thesis, and each of those that do ultimately look quite different from what was proposed. In short, planning is great but flexibility and opportunity are just as critical. Keep your eye out for exciting new ideas even if they weren’t in your original thesis plan; these inspirations often pan out as well or better than the original plan. And don’t stress out too much when your careful plan implodes – just accept from the beginning such an outcome is likely. As a specific implementation of this general suggestion, never apologize for failing to realize your original plan when discussing your work in talks or presentations (and try to avoid it in committee meetings). In reality, no one cares what you didn’t do, they only care what you did do. Talking about what you didn’t do just distracts and (sometimes) annoys the listener. In short, you should focus on what you have actually achieved and what you plan to do next.

From The DogHouse Diaries


Students are often encouraged to provide a strong a priori prediction for a given study, which can lead to several problems. First, it is usually easy to make a reasonable prediction that is directly contrary to the prediction advanced by the student: increasing A could just as easily lead to decreasing B as increasing B. Second, in correlative studies, one can nearly always conceptually invert the x-axis (presumed to be the cause) and the y-axis (assumed to be the effect) and yet still have a perfectly reasonable interpretation: see my post on “Faith’s Conjecture.” Third, when the study is actually conducted, a priori predictions are often NOT confirmed. Instead, negative (non-significant) or contrary results frequently emerge. Fourth, a strong a priori prediction can lead a student to assert support for the hypothesis when, in reality, the data more strongly support an alternative. This disconnect is very common in papers that I edit/review: in essence, the inferences aren't supported by the data. (Many journals even have a check-box in their reviewer forms for just this outcome.) For all of these reasons, a single prediction is usually not optimal. Instead, it is much more useful generate plausible alternative predictions that correspond to alternative mechanisms/processes/effects. That way, no matter what the outcome, you already anticipated it and you have an explanation for it. Moreover, you are less likely to be disappointed when your data don’t support a prediction, and disappointment in such cases has a huge influence on how enthusiastic and confident you are in presenting your results. In closing this point, I need to note that your supervisor might have a particularly strong opinion about predictions and, if so, you obviously need to take that into consideration.

An example of Faith's Conjecture: gene flow can constrain adaptive divergence and adaptive divergence can constrain gene flow. From Rasanen and Hendry (2008 - Ecology Letters).


Following from the above point, a given project always looks best before the actual work starts. After that, entropy happens and things start to fall apart. Targeted populations can’t be collected. Permits can’t be obtained. PCRs don’t work. DNA degrades. Funds run out. Hurricanes or floods destroy replicates. Hard drives crash (whatever else you do, make frequent backups of everything). And so on. Yet the project often turns out OK anyway. The problem is that most students lose excitement and interest as their ideal visualization transforms into messy reality – and this disillusionment increases with time, which decreases the motivation and drive to publish the work. Then they move on to something new – like a postdoc – without having published their previous work and the “grass is greener” syndrome kicks in. Now you are planning a new project, which as noted above looks ideal in concept, whereas the older reality is tarnished. As a result, efforts are often directed toward the newer work and the older stuff languishes. Yet it remains valuable to publish the older work as soon as possible. Stated in a more concrete fashion, you shouldn’t neglect your unpublished PhD work after you graduate. One reason is that your new project will also pick up considerable tarnish before you get around to publishing it. Another reason is that most people get their 2nd post-PhD position on the merits of their PhD work, not on work conducted during their 1st post-PhD position. Timeframes for transitions between most positions (e.g., postdocs) are simply too short to get much on your CV from your current position before moving on to the next one. A third (related) reason is that following through on your PhD work is usually by far the quickest route to a new publication, and lots of publications produced quickly will help your career. (Great advice on manuscript necromancy – breathing life into a “dead” manuscript – can be found here.)


Despite external evidence of their competence, those with the [imposter syndrome] remain convinced that they are frauds and do not deserve the success they have achieved. Proof of success is dismissed as luck, timing, or as a result of deceiving others into thinking they are more intelligent and competent than they believe themselves to be. (via WIKIPEDIA) A number of students fall into this trap and yet I can assert with confidence that success in graduate school comes FROM YOU. Certainly it helps to have a supportive university, department, supervisor, and lab, but research success requires you. If you have a cool research result, you obtained it. If you publish a paper, you did it. If you get an award, it is because you deserved it. None of this would have happened without you. Note that acknowledging your own abilities  doesn’t mean you shouldn’t get help from others – that is often essential too. No one can be an expert in everything, and it makes the utmost sense to get help from experts. I am reminded of the pithy qualifier offered at the start of his defense presentation by a fellow graduate student, Andy Dittman: During my talk, when I say “I”, I really mean “we”; and when I say “we”, I really mean “they.”

21 ways to overcome the imposter syndrom from Startup Brothers


What examiners want to see at your defense is a colleague rather than a student. The whole point of being awarded a PhD is that you are being recognized as intellectually equivalent to the people sitting in judgement on whether or not you should enter that club. Thus, your defense will always go most smoothly if you imagine yourself as a professor giving a seminar in another department, rather than a graduate student humbly seeking approval. Importantly, this advice is not an encouragement for you to be arrogant or dismissive, as I suppose sometimes happens with seminar speakers. With this in mind, emulate seminar speakers that have come to your department and whose talks (and answers to questions) you, and others in the department, reacted to best. Be confident but not arrogant. Be assertive but not abrasive. Admit when you don’t know something but don’t apologize for it. Listen respectfully to questions but don’t be cowed by them. Those people out there want to see you as their equal and you should proceed accordingly. (Here is some related advice from The Professor Is In)



The maximum length of your degree is often set by your university, department, or supervisor – and you need to meet those deadlines. So you can’t take too long. However, you also want to be to quick about it. Trying to finish too quickly will generally decrease your publication quality, quantity, or both. These outcomes are problematic because getting your next position depends largely on your PhD publications. Thus, you shouldn’t rush yourself out the door unless it is required. Note that following this advice will also help to fix problems that can arise from the “grass is greener” syndrome noted above. However, you certainly shouldn’t PLAN to take a long time, nor PLAN to exceed the time that you initially schedule. In such cases, you will find yourself running afoul of Hofstadter’s Law:  It always takes longer than you expect, even when you take into account Hofstadter's Law. Thus, make a realistic time plan and try to meet it but don’t rush yourself at the end if opportunities permit. Moreover, it is critical that you use such extra time productively – you need to get papers out!!!!!!

From A Reasoners Miscellany


In direct contrast to the above encouragement, some students drag on way too long. They insist on getting everything perfect. They want to publish every paper before submitting their thesis. Or they simply procrastinate endlessly. This strategy is also a very bad idea. You need to work productively and efficiently and submit your thesis and move to the next position without worrying about publishing everything.  I suggest you start planning your next position, such as a postdoc, at least a year in advance. Networking is critical: start talking to profs at meetings. Grants are great: apply for postdoctoral fellowships as they will give you the most flexibility for the future. But, critically, don’t spend so much time networking and writing grants that you diminish your publications, because these are by far the most important predictor of success in obtaining your next position. Thus, you need to find the sweet spot between not rushing and not dragging. My former student Ben suggests a nice analogy:I find myself with the mental image of a pendulum swinging; when you graduate, you want to be on the upswing, but not yet to that pause at the top of the swing where you no longer have any momentum.

From NIH's Rock Talk


I am have encouraged to also talk about personal interactions, such as when you and your supervisor don’t get along. Or you hate your lab/office/field mates. Much has been written on these points by others and I am not going to get into them here. They are simply too context-specific to fit into this rapid point-by-point style of advice. If you really want to work through these issues, I suggest you first talk to the people with whom you are having trouble and then talk confidentially to a neutral third party (another prof, for example). And always be respectful, regardless of how you feel about someone.

From Nash Turley 

Previous posts in this "How to" series

4.     How to choose a journal (+ part 2)


LINKS THAT MIGHT OR MIGHT NOT AGREE WITH MY ADVICE (thanks to Kiyoko Gotanda for the list)

From the Bruna Lab
From Dynamic Ecology
From the Hall Lab
From Aerin Jacob
From the Keogh Lab

Friday, March 6, 2015

How to Succeed in Graduate School - Part 1

My last “how to” post was about getting into graduate school, so it seems logical to follow up with a post about how to kick ass once you get into graduate school. Of course, the advice that follows won’t apply in all instances, and some of it is rather obvious, but I hope that I can provide some unique insights that might help you on your way. (Advice from other sources are provided by links at the end.) I also need to note that this post applies most directly to students who wish to make a career of research in academia or, to some extent, in government, industry, or NGOs. In addition, all of the suggestions apply to PhD students, whereas only some apply to MSc students. As an aside, I think MSc projects are great (I got a lot out of mine) but many institutions, including mine (McGill), and funding agencies put the major emphasis on PhD students. As a result, it is more and more common to simply skip the MSc and go straight from undergrad into PhD.

This post will be part 1 of 2 as the topic is super important and too much needs to be said to fit into a single post. Part 2 is here.


We had better get the obvious out of the way first – your success in graduate school is mostly determined by the papers you publish. You want quantity. You want quality. You want quality in quantity. You want it early and often and well. So how to achieve this?

Get started early. The earlier you start publishing the better. I strongly suggest identifying a definable project very early on that you can submit fairly quickly – in your first year. It doesn’t have to be earth-shaking or ground-breaking. A (hopefully novel) review paper on your proposed research area is a good option.) It just gets you started. You learn how the process works. You establish yourself as someone who can see research through to completion, someone who has something to contribute, someone who can get the job done. It will make everything easier: committee meetings, qualifying exams, defenses, future positions, everything. If you conducted undergraduate research, write it up early in your graduate degree (the longer you wait the harder it will be). 

 Shoot high at some point. It is so much easier to dump your work in open-access journals than to work your way through the more picky alternatives. A few papers in open-access journals are fine (especially early on) but you also need some papers in picky journals, ideally society-based journals like Ecology or Evolution or PRSB. These journals are a critical stamp of approval that says your work isn’t just “solid” but also “important” and “interesting.” Of course, the glamour journals (Science, Nature, PNAS) are even better career-wise if you can get into them, but they are a big risk. I have known a number of students who have essentially put their entire thesis into one paper trying to get it into those journals only to have it rejected from all of them after years of trying. The student has then been in such a hurry to publish before graduation that they dumped it in open access. And then they had to scramble to come up with a few more chapters. So it is a risky strategy to shoot for the glamour. Note that your supervisor will likely have strong opinions on these points, and her/his suggestions come before mine. 

First authored papers are by far the best. Collaborations with other students, postdocs, and profs are an extremely rewarding part of graduate school (more on this below) but what is most critical for your future career opportunities is first authored papers. Don’t make the mistake of thinking that a bunch of co-authored papers are equivalent to a first-authored paper. Moreover, co-authored efforts can still take an immense amount of time and effort, potentially reducing your ability to produce first-authored papers. So collaborate on other projects but remember that your success will be determined much more strongly by your first-authored publications.


Very early on as a supervisor, I realized that students (in fact, everybody) can be ordered along a continuum from “thinker” to “doer." Thinkers read all the relevant literature. They debate and puzzle and work over every concept and idea in great detail and at great length. Doers see the task and get it done – quickly and efficiently. The truth is that the current competitive academic environment is geared to reward doers given that the sheer volume of papers matters. Or, stated more precisely, a thinker who has very few papers will have trouble moving on, even if those few papers are great. And yet, at the same time, a doer who has tons of papers will sometimes be deemed superficial or a dilettante if those papers are deemed too minor – or folks can be suspicious of people who publish what they deem as “too many papers to do a good job on any of them.” The first step is to realize where you fall on the continuum: “My name is Andrew, and I am doer.” If you are a thinker, you must force yourself in the doer direction. You have to set yourself goals and deadlines and you must meet them. Stop thinking quite so much – and do some more doing. You will publish more papers, which is good. Of course, the opposite applies to doers – slow yourself down and do more reading and thinking. You will publish better papers, which is good.

Key: There are two types of people: those that talk the talk and those that walk the walk. People who walk the walk sometimes talk the talk but most times they don't talk at all, 'cause they walkin'. Now, people who talk the talk, when it comes time for them to walk the walk, you know what they do? They talk people like me into walkin' for them.


Publishing good work (early and often) depends first and foremost on data, which depends on the best possible sampling/experimental design and implementation. Nothing, simply nothing, can fix a poor sampling design or insufficient/inappropriate replication. The subtext here is that stats are great, and need to be done as well as possible, but the data come first and foremost. Data are real and exist without stats. Stats do not exist without data. Thus, don’t get too hung up early with trying to be a #STATSHero – that can come later. Of course, knowledge of the stats you will apply will greatly aid decisions about experimental design and effort. It is just that you have to inflate your football before you can throw it. As an aside, it is true that a #STATSHero will often get lots of publications by helping other students with their stats and will be a desirable commodity as a postdoc and in many other contexts. This is all well and good but remember that first-authored publications are considered much better than a ream of co-authored publications, and also that effort helping others with their stats can reduce the time you can devote to your own work. Don’t let good stats come between you and good data.


Many students agonize over the precise details on drafts of proposals and papers that they are producing before showing them to their supervisor. The undercurrent seems to be that they want to impress their supervisor (a reasonable sentiment) and thus want their writing to be perfect before she/he sees it. The reality is that you will fail in this effort. Your early drafts will NEVER come close to satisfying your supervisor – but that is the point, really. If your supervisor doesn’t trash your drafts, then what good are they to you really? The truth is that your supervisor is best viewed as an experienced and invested collaborator. They are there to help and that help often involves an entire restructuring or reframing of your paper, an effort in which they will usually (but not always) be well informed. Thus, don’t worry so much about making your first (or second or third) draft perfect because, no matter how hard you try, it won’t look anything like what you originally wrote after your supervisor gets a hold of it. Instead, it is much more efficient and less frustrating to get your drafts to your supervisor in the early stages while they are still quite rough. It will save everyone a lot of time and stress and will reduce the chances you will be offended or hurt or crushed by massive changes to something you thought was great. (Even if your paper is great at the start, it can always be better – and that is what supervisors try to do: no matter how good the paper is, they want to make it better.) Before moving on, I need to point out that you should not carry this suggestion to far. That is, supervisors also can get annoyed if your document is too cursory or incomplete. Do a good job, just don’t strive for perfection – that comes later.


Qualifying exams in their various forms are often terrifying for students, who have to stand at the front of the room playing “guess what’s in the professor’s head.” Here are some suggestions that can really help get you through it. First, make sure (if the regulations allow) to talk to all of the committee members well before the exam. Tell them about your project and ask them what they suggest you should read before the exam. This is essential because different profs with different backgrounds often have very different ways of viewing the world and think different ideas and papers are critical for you to know. No matter how much your supervisor helps you prepare, they simply can’t intuit what the other examiners might be thinking. Second, realize from the beginning that you will NOT know the answer to most of the questions. In fact, the way these things play out is that professors are specifically seeking to find out what you DO NOT know. Thus, as soon as it is clear you know the answer to a particular question, they immediately want to move on to something else. They are instead probing for what you DON’T know – and they will find it. Think of it as their attempt to circumscribe the hypervolume of your knowledge so they can decide if it is large enough to proceed toward a PhD. The definition of this hypervolume requires, by definition, the exploration of its edges rather than its center. So relax, you won’t know the answer to many of the questions, which is fine and the point, really. Third, admit when you don’t know the answer (don’t just blather on about stuff that you hope is related) but then, whenever possible, say you are willing to try to figure it out on the fly. Fourth, failure (of the temporary kind) is often a good thing. It gives you an opportunity to improve your knowledge in key areas and can lead to papers on its own. In fact, evolutionary biologist Tim Mousseau turned his remedial paper following a qualifying exam “failure” at McGill into a citation classic.

My Dad's frustration with qualifying exam preparation led him to blast out this poem one afternoon. He posted it on the departmental bulletin board and, apparently, the professors weren't impressed. I would have been.


Like it or not, success in today’s scientific enterprise requires (or at least greatly benefits from) not just good science but also a good “presence.” In fact, after publishing papers, the next most important skill – yes, even more than becoming a #STATSHero – is to give a great research presentation. This is true at every level of your career: undergrad, grad, postdoc, prof. The best way to learn how to give a good presentation is to practice, practice, practice. By this I mean at conferences, retreats, lab meetings, etc. – any opportunity you can get. It doesn’t matter so much what you talk about (don’t worry if you don’t have data) but how you talk about it. So do it early, do it often, and don’t quit doing it. Before each major presentation, practice it for fellow students and, if possible, your supervisor. Lab meetings are great for this. Immediately after a presentation, ask fellow students and, ideally, your supervisor how you can make it even better next time. Some of the advice will be technical (less text, simpler figures, better colours, etc.) and you will gradually adopt what works and jettison what doesn’t. However, the greatest improvement can come simply from increasing your comfort level, which increases with the number of times you give presentations. (If you are petrified by public speaking, think of joining a debating club or taking acting/theater/performance classes – I think my comfort in speaking stems in part from performing arts classes taken in high school.)

For evidence that it doesn't have to matter what you say - but how you say it - check out this Chicken Chicken Chicken: Chicken Chicken presentation.


Following from the above idea of a “presence”, participation in free-form discussions is very important. By this I mean speaking out in discussions and asking questions in lab meetings and seminars and at conferences. Becoming comfortable in such contexts is an extremely valuable skill and it becomes increasingly important as one moves up the career path. For instance, professors are often expected to have a strong opinion about almost everything. Moreover, being animated in discussions is a great way to establish collaborations, develop new ideas, and generally feel connected to your lab, fellow students, and your department. I have a colleague (Mike Kinnison) who challenged his students to ask at least one question per talk they saw at a conference. I think this is a great idea – not just the actual asking but the preparation for it. If you tell yourself that you are going to ask a question at the end of a talk, then you find yourself paying much closer attention and being more critical in your evaluation of the work. Doing so helps you gain a lot more intellectually from a talk. I suggest you write down at least one question per talk and try to ask it at the end. I realize that some people are mortified about speaking out but I can first reassure you that no one ever thinks anything negative of a student who asks a question – regardless of the specific question. (Unless it is rude, which you should never do, or unless you ask too many per talk!) In fact, they are nearly always impressed. So, if you are in this shy category, find a context where you can feel somewhat comfortable asking questions, ask as many questions as possible in that context, then move up to the next context, and keep going.


Side projects can be great. They are an opportunity to do something a bit different from your thesis, often with fellow students or with postdocs/profs. And, returning to the point about publishing early, they are often more defined - and therefore easier to publish quickly than your %&#%^$ thesis that has to be integrated across multiple chapters. Of course, too many side projects can delay your progress and drive your supervisor nuts, so don’t get too carried away. Indeed, side projects are the bread-and-butter of “doers” but, for the reasons discussed earlier, doers need to restrain themselves from too much temptation. The funny part is that side projects are often what people become known for rather than their major thesis focus, so do follow your muse when it strikes.


A social media presence through blogging and Twitter is certainly a way to increase your profile, meet a community of like-minded scholars, and get tips to important papers. I think every student should be on Twitter and be a part of some blog (not an administrator but rather a contributor). However, social media is also a big time-suck and so it is critical to not let it detract from your research. Of course, some people become more famous for their social media presence than for their science, which feels good but (usually) does not translate into a career in research. So focus on the research and use the social media to supplement/promote it and yourself. Some ways to minimize the negative aspects are to only follow a modest number of accounts and restrict yourself to only a few tweets per day and posts per month.


Teaching is a valuable part of any graduate program and you should be certain to do it a bit. Moreover, it can help distinguish you on the job market from someone with an equally-good publication record but no teaching experience. But I suggest not doing it too much (unless you have to) as I can assure you it comes at a trade-off with progress in research.


From an extremely timely tweet by @phoebemaund
Graduate school should be one of the best times of your life. You can help this to happen by being successful, as per the above comments, and by making sure to have fun. Indeed, many of the greatest joys of graduate school come from stuff unrelated to the research. Hell, I learned to fly airplanes while in graduate school and I had a long series of very late nights playing Doom with my office mate (I kicked your butt, Mike!) And the parties. And the fishing and diving trips. And I found my future wife! You obviously need to embrace these activities; and, yet, of course, you can’t get too carried away. You need to make sure that you are progressing in your research. So I guess the other way to say it is “work hard, play hard.”


Previous posts in this "How to" series

4.     How to choose a journal (+ part 2)


LINKS THAT MIGHT OR MIGHT NOT AGREE WITH MY ADVICE (thanks to Kiyoko Gotanda for the list)

From the Bruna Lab
From Dynamic Ecology
From the Hall Lab
From Aerin Jacob
From the Keogh Lab

Wednesday, February 25, 2015

Incorporating animal mass mortality events in ecology and evolutionary biology

[ This post is by Samuel Fey, Stephanie Carlson, and Adam Siepielski; I am just putting it up.  –B. ]

Many thanks to Andrew for the invitation to write a post on our recent article published in the January 27 issue of PNAS on recent shifts in the occurrence, magnitude, and cause of animal mass mortality events. We thought we would use this opportunity to write about the motivations for this study, our basic findings, and how we hope this study improves our ability to better understand the ecological and evolutionary importance of large animal die-offs.

Fig. 1.  A sunfish and largemouth bass mass mortality event (MME) following a severe winter on Wintergreen Lake, Michigan.  Photo: Gary Mittelbach.

We conceived of this project after reading and listening to a series of media stories on several different animal mass mortality events (MMEs). Such media coverage, whether documenting the death of thousands of birds or a large fish die-off (Fig. 1), understandably attracts attention from readers and listeners. While the circumstances surrounding such events seem largely idiosyncratic, one similarity in how such events were reported is that they are rarely placed into any larger context. As such, we were surprised to learn that no quantitative synthesis of MMEs across animal taxa existed. Thus began the process of combing the scientific literature to generate a database that we hoped could shed light on how the frequency, magnitude, and causes of MMEs have changed through time.

Years later, after reading thousands of morbid accounts of animal morbidity, and with the support of a large research team, we finally analyzed the patterns present in our database (Fig. 2). Briefly, our data suggested that (1) the frequency of large animal die-offs has been increasing through time; (2) that the magnitude of events has been increasing for birds, fishes, and marine invertebrates, invariant for mammals, and has been decreasing in magnitude for amphibians and reptiles; and (3) that events caused by multiple stressors, biotoxicity, and disease have been increasing most rapidly through time.

Fig. 2.  Changes in MME magnitudes through time.  (Click for a larger view.)

Lest we give the impression that this effort represents a perfect approach to understanding the true dynamics of animal die-offs, we try to be as clear as possible in our paper about the limitations surrounding such a synthetic approach. Documenting trends in the frequency of events raises the question of detecting an “epidemic” versus “an epidemic of awareness”. Understanding trends in causes is additionally complicated by methodological improvements allowing greater detection of certain causes. For example, techniques to detect the occurrence of disease as a cause of death have advanced rapidly. Additionally, large animal die-offs reported in the peer-reviewed literature represent only a fraction of the actual events that are reported by state or federal agencies and media organizations, which themselves represent a fraction of events observed by any human anywhere. These factors mean that we do not know the true frequency of MMEs in nature.

Fig. 3.  MMEs most frequently occurred in aquatic ecosystems, where stressors such as dessication contributed to causing MMEs. Photo: Adam Siepielski.

Despite these limitations, we believe that collecting the best available data on such events, and analyzing patterns in these events to the best of our ability, was a worthwhile endeavor. Clearly, there is a need for an improved understanding of mortality events intermediate in scale between background mortality (e.g., a house cat eating a few red-winged blackbirds) and species-level extinctions. This paper represents an early step toward establishing this research program, so many important questions remain unanswered: What does it mean that the magnitudes of large animal die-offs have been increasing for certain taxa but decreasing for others? What is the relationship between large animal die-offs and human-induced changes to the planet at local and global scales? Do the patterns of animal die-offs we found represent an early signal of greater ecological consequences from such human-induced changes?

Although our focus was primarily on the demographic effects of MMEs, the occurrence of these events may also have considerable consequences for microevolutionary processes.  For example, removal of 90% or more of a population may generate extremely strong survival selection and thus the potential for rapid evolution. At the same time, however, such strong selection, combined with the resulting small number of survivors, may place populations at risk of local extinction through demographic stochasticity. Even if an evolutionary-rescue-type scenario occurs, there may be a cost to such adaptation if the MME is a one-time event. As a result, any adaptive evolution could actually be deleterious if the phenotype favored by the MME event is maladaptive once conditions return to “normal”. On the other hand, it may be that such large-scale mortality events do not typically favor any particular phenotype. We simply do not know.

It’s our (possibly naïve) hope that publishing this paper will lead to three outcomes: (1) improved data collection, (2) improved sharing of existing resources on animal MMEs, and (3) generation of interest in the importance of rare events on ecological and evolutionary dynamics. We note that animals are of course not the only organisms subject to MMEs. Whether or not other groups display the kinds of trends we documented is presently unknown.

Given the surprising nature of the shifts in the patterns surrounding MMEs, especially the trends in magnitudes, an adequate research program is needed to better understand and monitor such events. Yet, our database shows that the proportion of an animal population that is affected by such large die-offs – perhaps the single most important metric for contextualizing such events – was reported in only about 10% of all studies. In addition to encouraging better data collection, we also hope to help centralize the collection of data on animal mass mortality events by government agencies, media organizations, and citizen scientists to gain additional insight into the mechanisms underlying these events. Ultimately, we hope that this paper is one of many steps towards better understanding the ecological and evolutionary causes and consequences of rare demographic events.