Friday, February 24, 2017

A Tale of Two Thousand Cities - by Charles Darwin

The number of cities in the world depends on how you count but it’s a big number. Brilliant Maps says more than 4000 cities have more than 100,000 inhabitants. The UN says 1692 cities have more than 300,000 inhabitants and 512 cities have at least 1,000,000 inhabitants (totalling 23% of the World’s population). And, hey, helpfully narrows the number of cities and towns down to somewhere between 600,000 and (apparently) 4,784,754,000. No matter how you slice it, I am certain there are 2,000 major urban areas with lots of people in them.

"4,037 cities in the world that have over 100,000 people" SOURCE

Cities change everything for the organisms that live in them. Temperature changes. Noise changes. Available habitat changes. Prey changes. Predators change. Food changes. Pollution. Eutrophication. Invasion species. For years, the temptation was simply to write these areas off from the perspective of biodiversity and nature; but – over many years – a shift has occurred to establish a vigorous field of “urban ecology”. The idea is that cities are ecosystems too and we should manage them and their biodiversity as such. And where ecology goes, evolution follows. That is, any sort of environmental change is expected to impose selection on the organisms that remain in that environment, which should lead to evolutionary adaptation to urban environments. This post is about how urbanization dramatically shapes the evolution of many species. I might have called it “Darwin Comes To Town” but Menno Schlithuizen’s forthcoming book has already appropriated that wonderful title.

The past few decades saw a smattering of studies demonstrating evolution in response to urban environments. Byrne et al. (1999) showed that a new form (species?) of mosquitoes had evolved in the London Underground. Cheptou et al. (2008) showed that plants evolved reduced dispersal in cities because dispersers were likely to end up in the in hospitable “concrete matrix.” Following from these earlier, somewhat sparse demonstrations, studies of urban evolution have really heated up recently: cool new papers are coming out, books are being written, grants are under review, symposia are being organized, and working groups are being convened. Inspired by this recent enthusiasm, I want to highlight some of my own work in this area, some of the exciting new work that has come out this year, and some attempts to tie it altogether through meta-analysis.

Darwin's finches of multiple species near Puerto Ayorra pigging out on rice provided by humans. 

My own foray into urban evolution started with coincidental discoveries in Darwin’s finches of Galapagos. Up to the 1970s, medium ground finches (Geospiza fortis) at Academy Bay, beside the small town of Puerto Ayora on Santa Cruz Island, were bimodal in beak size: many large or small birds with relatively few intermediates. By the time we started working there in the 2000s, Puerto Ayora had grown dramatically, and the collection of new beak size data did not reveal the same bimodality as in the past. Yet at the same time, we uncovered bimodality at a site (El Garrapatero) well removed from the town where finches were not exposed to urban conditions. Compiling data from 1964 to 2005, we confirmed that beak size bimodality was lost the finches living in and around Puerto Ayora coincident with the dramatic human population increase. We then showed in later work that this collapse of diversity was associated with a degradation of the diet differences that normally differentiate the species. In short, all the finches are now feeding on human foods, which has removed the selection pressures formerly favoring diversification in this group. Indeed, additional work we currently have in review shows that urban finches are actively attracted to humans and their foods, whereas finches outside of the city are not.

Darwin's finch beak size distributions, with the arrow showing situations tending toward bimodality.

Acorn ant colonies are entirely contained within acorns, which is pretty darn cool – and makes for a wonderful experimental system. One can pick up an acorn and move it to a new site, or to the lab, and thereby test for thermal tolerances and local adaptation. And – conveniently for the question at hand – oak trees producing lots of acorns are found both inside and outside cities. Sarah Diamond and colleagues tested whether urban acorn ants had different temperature tolerances than rural ants. Consistent with the “urban heat island” effect (temperatures are higher inside cities than outside), the authors found that city acorn ants had higher thermal tolerances, and that this difference could be attributed to a complementary combination of plasticity (warmer rearing temperatures increased thermal tolerance) and genetic differences (city ants had higher tolerances for a given rearing temperature). But the temperature effects of cities might not always be so straightforward.

From Diamond et al. (2017)

One of the most ubiquitous plants in urban environments is clover – as a kid, I spent many hours searching for 4-leafed versions. Clover is also abundant outside cities, and so might be a good model for understanding how evolution proceeds in response to urban conditions. Marc Johnson, Ken Thompson, and colleagues hypothesized that the urban heat island effect should lead to the evolution of reduced freeze tolerance in clover, which is controlled by a known genetic polymorphism for hydrogen cynanide. Surprisingly, they found exactly the opposite – freeze tolerance genotypes were more common inside Toronto than outside. The same result was obtained for New York and for Boston, whereas no pattern was evident for Montreal. After a long trip down the rabbit hole, the authors showed that, because snow cover is less common in cities than without, some cities are actually “urban cold islands” in winter that favor the evolution increased – rather than decreased – cold tolerance in plants. (Montreal has so much snow both in and out of the city than it doesn’t matter.)

The use of multiple urban-nonurban gradients, as above, allows greater insight than only a single gradient. Also this year, Liam Revell, Kristin Winchell, and their collaborators studied Anolis lizards on Puerto Rico, comparing those in three cities to those just outside the cities. In forests, these lizards are commonly found on branches that can be quite narrow, whereas in cities they tend to occur on the much broader substrates of walls. Previous work showed that hindlimb length tends to evolve according to substrate size – being longer on broader substrates. That was just what the authors found here: city lizards have longer legs and a common-garden rearing environment confirmed that at least some of this difference was genetic.


The above examples are just a few studies from this year. Many other studies are also demonstrating trait responses to urbanization, although, in some cases, it isn’t yet clear if the change is genetically based. City birds sing different songs, appear smarter, have different behaviors and stress responses, sometimes have different clutch sizes, and so on. City mice differ in key genes that might reflect adaptation, Daphnia evolve to be smaller in urban ponds, and so on. These wonderful examples of phenotypic changes (at least some evolutionary) in urban environments raise the question: are they exceptional? Humans influence evolution in all sorts of contexts apart from cities (hunting/harvesting, fragmentation, climate change, pollution, eutrophication, invasive species, etc.), as we recently reviewed in a special issue of PTRSB. Are urban environments any different, such as by driving faster rates of change than in other contexts?

Marina Alberti has led the recent charge in reviewing work on urban evolution and contacted me with an idea to use our database of rates of phenotypic change to quantitatively ask if changes were greater in cities than in “natural” or other human-disturbance contexts. The same database had previously been used to show that – among other things – human disturbances accelerated rates of change, that the most dramatic effects were evident when humans acted as predators, but that evolution was not exceptionally rapid in the context of invasive species. Georeferencing all the observations in this database and linking them to urbanization estimates, the study – a collaboration among many people – showed that adding information on urbanization substantially improved the ability to predict rates of change – a number of which are confirmed to be genetically based. I speculate that the main reason is that urbanization is associated with many forms of environmental change occurring all together (a subset are listed at the outset of this post), which should impose particularly strong and diverse selection on the organisms that persist.

Locations of rate data used in our analysis.

The next time you walk through a city, take a look past the steel, glass, and concrete, to see the plants and animals that live there. (And, of course, to not see all the microbes.) Each of these organisms is experiencing selective pressures that simply didn’t exist in most places until relatively recently. Selection is the engine of evolution and – indeed – many of these organisms have evolved to better suite them for urban conditions. Indeed, some of those organisms might not exist in cities were it not for adaptive evolution keeping pace with increasing urbanization.

Urban evolution is the new hot Broadway (and off-Broadway) play in the evolutionary – and eco-evolutionary – cannon. See it now.

Here are some Darwin's finches pigging out in the Baltra Airport, Galapagos.

Saturday, February 11, 2017

On Sabbatical

Sabbaticals might seem a strange thing to students, administrators, politicians, the general public, and – well – everyone who doesn’t take them. A common perception is that professors who take a sabbatical are “taking a year off” – and certainly that sometimes happens. As a result of perceptions such as this, some countries don’t allow paid sabbaticals, some states within countries don’t allow paid sabbaticals, and some particular universities don’t allow paid sabbaticals. In many other cases, only partial support is provided or the time between sabbaticals increases beyond the normal every-7th-year. In this post, I make the case for fully paid sabbaticals every 7th year as the greatest benefit to everyone.

About the above: I started my Eco-Evolutionary Dynamics book on my first sabbatical and finished it on my second sabbatical! Only sabbaticals made it possible. For more see

Teaching (and service) improves

Most people who do not attend university – and even many people at universities – think that what professors are for is teaching (and various committee-style “services”). Certainly, most professors do a lot of teaching, which is how most students know them. So, if the role of professors is to teach, and they don’t teach on sabbatical, then they aren’t doing their job on sabbatical – so they shouldn’t be paid. This logic is precisely why legislators in some countries and states forbid paid sabbaticals. Professors have other important jobs besides teaching and service – and those other jobs (research!) benefit dramatically from sabbaticals. However, I first want to make the point that even teaching benefits from sabbaticals. The main reason is that: “The biggest thing for the professors is they get the chance to refresh themselves and to escape. They come back … invigorated.”

Teaching the same course year after year after year (or even different courses year after year after year) can whittle away at enthusiasm and the motivation to make major improvements. A year away can completely re-invigorate a professor’s motivation to teach, teach often, and teach well. (Part of this motivation comes from the guilt a professor feels when his/her colleagues have to teach those courses for a year.) From my own experience, I definitely feel this benefit is critical. Just this fall – right after my sabbatical – I taught three courses: my graduate class in Advanced Evolutionary Ecology, an undergrad class in Evolution, and our Introductory Biology class. I also took over coordination of the last of these and gave guest lectures in a number of other classes. Teaching was exciting again – fun again – motivating again. I wanted to do new things, exciting things, more things. This sort of excitement and motivation really improves with a year away from teaching.

Importantly, classes rarely suffer from sabbaticals in the sense that most of the classes are taught anyway – just by other professors. Hence, the long-term benefit to teaching does not come with any major short-term costs. Sabbaticals are good for teaching!

Research improves

The primary thing that many professors do is research. In fact, research at many universities is what professors are supposed to spend most of their time doing. This is critical. Universities are not just about the transfer of information and ideas from experts (professors) to trainees (students), they are just as much about the generation of new ideas and new knowledge. Moreover, this generation of knowledge benefits the transfer of knowledge because students respond much more strongly to professors who are speaking from their own experience – and often injecting examples from their own work. And then undergraduate (and graduate) students can become involved in the research and thereby have real “hands-on” training. In my lectures, I specifically emphasize research conducted by McGill undergraduates who were sitting in the same seats as the current crop of students in the class. Research benefits teaching!

Sabbaticals have a HUGE effect on research because they afford the time and motivation to learn new methods, write new grants, publish that backlog of papers, do intensive field or lab work, etc. Some professors travel to places where they can get training in new technologies. Some professors travel to places where they can be close to their field work, or their collaborators, or important infrastructure. Some professors remain local and focus on publishing papers. On sabbatical, professors have the time to think about science, do science, write science, learn science. Sabbaticals are critically important for research success, particularly “taking it to the next level.”

Apparently not everyone (or every study) finds that average research productivity goes up after sabbaticals. This doesn’t mesh with my experience. Some years ago, Keith Crandall was telling me a story about how he was fighting to convince the administration of a university of the value of sabbaticals. Among other things, he showed a graph of his publication rate in relation to the timing of his sabbaticals. When preparing this post, I asked Keith about graph and he was able to recreate it from Web Of Science – showing big jumps in publication productivity with each sabbatical. 

Keith: thanks for the idea and the graph!

I did the same calculations for myself and found the same thing – big jumps in productivity with each sabbatical. Beyond benefits accruing to the professor and the people influenced by his/her research, universities are often ranked based largely on professor research productivity – and these rankings can have major consequences for funding, recruitment, and continued success of a university.

As an aside, you will see another message in the graph – starting a faculty position is often coincident with a big drop in productivity. For all you new profs out there worried about your slow start, take heart, it is only temporary. It takes time to build up a lab and a research program – and this is the case for EVERYONE.

Sabbaticals rule

In summary, sabbaticals are good for everyone involved. Ok, fine, a politician might say, but “we don’t need to pay the full salary – go out and get some yourself.” To those people, I would say: “Sabbaticals when you travel are extremely expensive, particularly if you have a family.” If you don’t provide full pay to professors, they are much less likely to go to new places, which is of great benefit to many. (Of course, a great sabbatical can also be had while staying in the same location.) My own university provides full support for sabbatical every 7th year (or 6 months support after every 3 years) – THANKS MCGILL – KEEP IT UP. However, even then, I lose money. The only way I can make it work is because I can stay almost for free with family in California and, most recently, the wonderful Miller Institute for Basic Research helped fund my sabbatical at UC Berkeley.

So, please everyone, from someone who has now had two sabbaticals, keep full support for sabbaticals every 7th year. Everyone wins – except those countries, states, and universities who don’t have them. 


To be honest, some graduate students might not benefit so much from their professor going away on sabbatical. Physical proximity between a professor and his/her students is more conducive than is skype to progress on their thesis. Of course, skype, joint field work, and visits can help minimize the cost to these students. Personally, I need to be better in this area on my next sabbatical.

Friday, February 3, 2017

Yes, humans influence evolution. Yes, that has consequences for humans!

Some might think it would be empowering to earn the moniker of "world's greatest evolutionary force" where it would seem some superhero can, Superman style, rapidly make something "evolve". Like tossing a common ancestor into a phone booth and out pops all the species of Darwin's Finches. Whelp, the reality is that this might not be such a great thing. What if this greatest evolutionary force might be causing detrimental things to happen such as biodiversity loss? What if this greatest evolutionary force is affecting human society?

Well, what might this greatest evolutionary force be? For better or for worse, it's humans. For most of us living in our privileged world with a roof over our heads and food in our stomach, it is easy to become ensconced in our little bubble of consumables and disposables while hiding behind some electronic device, and not think about how we, humans, are affecting evolutionary processes. For example, when something is domesticated for human consumption, be it crops or pets, what happens when those domesticated individuals intermix with wild individuals? What happens when we put antibiotics into the wild? What happens when we want to put taxidermy heads on our walls? What happens when we move something from one place to another, intentionally or by accident (does it matter?)? So, perhaps it's time to come out from behind your screen to think about these questions.

All of these questions are important, but it is not enough to just ask how we are influencing evolution, but to also ask what are the consequences of this? How is this feeding back on to humans and our societies? Well, in working with Andrew Hendry from McGill University and Erik Svensson from Lund University, we set out to agglomerate a special issue of Philosophical Transactions of the Royal Society - B focusing on just this question. We came up with the unimaginative, but descriptive title of "Human Influences on evolution, and the ecological and societal consequences".

We structured this issue in a slightly different way where we considered individual contexts of human influence as a 'topic'. These topics include things such as domestication, habitat fragmentation, hunting, urbanization, medicine, disease, and more. For most of the topics, we opted to have two manuscripts related to the topic: a review type article, focusing on a particular context, how that affects evolution, and then in turn, how this might affect humans, and an empirical paper that set out to test some of the ideas laid out in the review.

We considered everything within two theoretical frameworks. The first is the phenotypic adaptive landscape, where a three dimensional surface is pictured with the peaks representing high fitness phenotypes and the valleys representing low fitness phenotypes. Selection would then be acting to push a population up the adaptive peak. If that landscape is altered (say by humans introducing a novel predator), however, then the landscape, and thus selection will shift. Eco-evolutionary dynamics would then consider that change in a distribution of phenotypes and how that would affect the ecology of that population, including changes at the community and ecosystem level.

 As you can imagine, trying to understand how everything is connected can be rather confusing, so we modified the "traditional" eco-evolutionary framework to incorporate all the different parts together.

Using this, we then set out to try and predict which human influence might have the strongest effects on evolutionary and ecological processes. By no means is what we've done comprehensive, it's merely a tiny stepping stone to fully comprehending the impacts that we humans are having on evolution and how that is feeding back to human society. One of the reasons making predictions is so difficult is because evolution can be influenced by a myriad of factors. For instance, predicting how invasive organisms will respond, as well as how the native populations will respond, are dependent on so many biotic and abiotic interactions, it's extremely difficult to predict if the invasion will succeed, and if so, what will the consequences of it be and then how that would subsequently affect human populations.

Let's take a look at one of our thoughts about human influences on evolutionary processes. Depending on the context, specific components can affect selection itself, or other components of evolution such as standing genetic variation or as we put it, evolutionary potential. In some contexts,  a strong effect is quite obvious - hunting and harvesting are usually size specific, which will result in evolutionary changes in size. However, the hunting can not persist forever at high rates as the population would eventually go extinct. Perhaps more important is our efforts to control for pests or perceived "enemies" because it will result in increased tolerance and resistance, which then means we need stronger/better ways to control the enemies, and then they will again evolve increased tolerance. This perpetual "arms race" could, and has, led to things like superbugs that cannot be controlled with any current medicine.

What about potential ecological consequences as a result of this human induced evolutionary change? If we again consider hunting and harvesting, perhaps we are reducing the population size of or even removing a keystone species. If certain species are targeted, where individuals have a strong role in the community structure, their removal will have obvious cascading effects. For example, otters are essential in maintaining the giant kelp ecosystem in the Pacific Ocean, and the loss of otters in this ecosystem, perhaps due to pollution, then cascades into human society as important fisheries and a carbon sequestration source are subsequently lost.
I just wanted an excuse to post a photo of cute otters. 

In our introductory article, we actually ask a total of eight questions relating to human influences on evolution and the consequences they have on human society. We know you will have opinion and agree or disagree with us, so we would love to hear from you in the comments. In a nutshell, we hope this issue make you realize just how much humans are influencing evolution, and that these human induced shifts in evolution have societal consequences. Unfortunately, a large number of those consequences are detrimental to humans. And we're not alone on this planet, so those negative consequences are also affecting everything else!!

Link to special issue:
Link to our introductory article:

A 25-year quest for the Holy Grail of evolutionary biology

When I started my postdoc in 1998, I think it is safe to say that the Holy Grail (or maybe Rosetta Stone) for many evolutionary biologists w...