Wednesday, August 16, 2017

Evolution in a common landscape: a tale of two stickleback species

All species cope to some extent with environmental heterogeneity. How do they achieve this? Do they tolerate or avoid extreme conditions? Or do they adjust to local selective environments through adaptive evolution? Many studies in evolutionary ecology look into these questions one species at the time. Yet, species do not live in isolation, but are assembled in communities. We might thus ask if members of the same community respond to environmental conditions similarly or in species-specific ways. Answering this question is important for understanding the eco-evolutionary dynamics of communities. Since anthropogenic impact on natural systems may simultaneously put multiple species at risk, a multi-species approach is also relevant for conservation and natural resource management. In a new article in Nature Communications, we investigate to what extent two stickleback species, the threespine stickleback and its relative the ninespine stickleback, evolve “collectively” across contrasting environments. That is, a tale of two stickleback species in a common landscape.

Evolutionary biologists have accumulated ample evidence for contemporary evolution in natural populations. Meanwhile, the question of why populations do (or do not) evolve and whether they evolve in a predictable manner will still keep us busy for quite some time. There are many species and many environmental contexts in which species can evolve. For as much as we know, the way populations evolve is species- and context-dependent – and thus highly variable. This is reflected in meta-analyses such as in last month’s issue of the American Naturalist, where Krista, Gregor, Caroline and Andrew illustrated that even when we expect populations to evolve in a predictable direction (i.e., parallel evolution), the extent to which they actually do so is highly variable (Oke et al 2017). Across species and environmental contexts, populations thus show anything from “very” parallel to “not-so-parallel” evolution. Aspects of evolution are thus not very predictable.

Things become clearer when focusing on one species at the time. One of the species that has been strongly fueling the debate of the importance of parallel and non-parallel evolution is threespine stickleback. Indeed, of the 92 studies included in Krista-and-friends’ meta-analysis, 26 studies featured this very species. One of the most convincing, wide-spread and best understood cases of evolution in nature is the rapid parallel evolution of reduced body armour (e.g. from high to low numbers of lateral plates and from long to short spines) when marine threespine stickleback populations colonize freshwater. Importantly, these populations often evolve in a predictable manner, but we also have a fairly good understanding for why they sometimes don’t. Gene flow, for instance, which homogenizes the gene pool and therefore slows down or halts adaptive divergence, explains some of the limits on the evolution towards low-plated populations in freshwater (Raeymaekers et al 2014). In another threespine stickleback ecotype pair, the lake-stream system, variation in phenotypic and genomic parallelism could not only be explained by gene flow, but also by the magnitude of the difference between the lake and stream environment, which has an amplifying effect on adaptive divergence (Stuart et al 2017). Studies like this generate a better mechanistic understanding of evolution, because they show how strong selection on ecologically relevant traits and their underlying genes has to be to contribute to local adaptation, and how often there is a common genetic basis for such traits.

Threespine stickleback populations can evolve rapidly from completely plated (top) to low-plated (bottom), while homogenising gene flow can slow down this process, even when selection on plate number is evident from one generation to the next (Raeymaekers et al 2014). Photo credit: Anna Mazzarella.

Yet, single-species studies hold a major limitation for the study of contemporary evolution in nature: they do not provide insight in the generality of contemporary evolution. For instance, the evolutionary versatility of threespine stickleback may be exceptional, and thus levels of adaptation in this species may be not representative of the typical strength of adaptation in nature. And of course, species do not live in isolation but are assembled in communities. Members of the same community often face similar environmental gradients, but do not necessarily respond similarly to these gradients. From a community perspective, it is important to understand the variation in these responses, in particular because adaptation to local selective environments in one species may also influence adaptation in other species (e.g. through competition or dilution effects). So, in order to fully understand biodiversity patterns across ecologically diverse landscapes, we should consider multiple interacting species simultaneously, providing a more holistic view on the landscape processes shaping biodiversity. This also makes sense for conservation and natural resource management, since anthropogenic impact on natural systems may simultaneously put multiple species at risk.

In our new study, we performed a comparison between the threespine stickleback and its relative the ninespine stickleback (Raeymaekers et al 2017). We primarily wanted to find out to what extent both species differ in evolutionary potential to deal with challenges along the broad habitat gradient over which they coexist. In western Europe, both species co-occur frequently at the exact same spots, which includes both freshwater and brackish habitats. Yet, both species also have a wide-spread geographic distribution, are closely related (allowing us to compare homologous traits and genomic regions), show interesting differences in ecology, and are highly abundant, and thus represent an excellent pair of species for this type of study. We sampled both species at four freshwater sites and four brackish sites, and then compared them for various aspects of population divergence. We analysed 1) whether the two species show phenotypic and genomic signatures of adaptive divergence along environmental gradients, 2) to what extent both species show parallel patterns of population divergence, and 3) what are the most important spatial and environmental drivers of population divergence in each species.

“Von dem Stichling”. Description of threespine and ninespine stickleback in “Fischbuch: das ist ein kurtze, doch vollkommene Beschreybung aller Fischen so in dem Meer und süssen Wasseren…” by Gessner and Forer (Zürich, 1563).With two and six dorsal spines, the drawings of both species do not look very professional. Yet, even nowadays the number of spines is a source of confusion. Threespine stickleback sometimes have four spines, and in Dutch the ninespine stickleback is called “tiendoornige stekelbaars” – or "tenspine stickleback". Based on my own counts, this is a more appropriate name.

One or two species of stickleback? Each student has to pass the test.

Here are our most important findings and some reflections:

1) Phenotypic divergence was significant for 50 % of homologous traits in threespine stickleback vs only 7 % in ninespine stickleback, while the proportion of outlier loci (SNPs which are likely genomic targets of selection) was at least 2.5 times larger in threespine stickleback. This confirms a stronger tendency to adapt in threespine stickleback. Since this is the first time that both species have been compared in exactly the same environmental matrix, we now know the effect of species-level differences in evolutionary versatility on population divergence.
2) These results do not imply that ninespine stickleback cannot adapt, since populations might already be preadapted to the environmental gradients in the study area. However, we observed a numerical advantage of the threespine stickleback in freshwater. We proposed that this relative ecological success could possibly be attributed to their evolutionary versatility. Of course, two species only represents a very small community, but it shows the potential of merging landscape genomics with community ecology to understand whether or not species evolve “collectively” across landscapes.
3) We observed substantial phenotypic, but no genomic parallelism between both species. This result demonstrates that the evolution of similar phenotypes in the same selective environments might primarily involve different genes. Based on previous comparative genomic studies, this result is not unexpected, but it is exciting to observe this in exactly the same spatial matrix. 
4) Note that we wanted our study to allow for a “fair” comparison of evolutionary versatility between the two species. We therefore compared both species for homologous traits only. Indeed, even if one species would be extremely variable for a trait which is missing in the other species, it would be hard to decide which species is most versatile. Luckily, most measurable traits in both species are homologous anyway (lateral plates, first dorsal spine, pelvic spine, gill rakers, fins, …). Non-homologous traits include dorsal spine #4, #5, #6, … #10 (guess which stickleback species is lacking those spines).
5) A reference genome is often used to facilitate SNP-typing. Yet, at present a reference genome is available for threespine stickleback, but not for ninespine stickleback. While it is possible to use the threespine stickleback genome as a reference for ninespine stickleback, we didn’t do this since this would narrow down the comparison between the two species to homologous genomic regions. Homologous traits do not necessarily have the same genetic basis, and hence the entire genome should be considered to allow for a straightforward comparison of evolutionary versatility. In addition, homologous genomic regions may already have gone through a long history of selection - perhaps pre-dating the origin of both species, and hence may bias our analyses in unexpected ways. In this respect our choice for de novo SNP typing in ninespine stickleback seemed more "safe". It is waiting now for the assembly of the ninespine stickleback genome to look into homology effects in detail.

The results are further discussed with respect to how differences in genomic architecture, gene flow and life history may induce or reflect variability in evolutionary potential and ecological success among species sharing the same landscape. Read more here.


Cited literature

Oke KB, Rolshausen G, LeBlond C, Hendry AP. 2017. How parallel is parallel evolution? A comparative analysis in fishes. The American Naturalist 190:1-16.

Raeymaekers JAM, Chaturvedi A, Hablützel PI, Verdonck I, Hellemans B, Maes GE, De Meester L, Volckaert FAM. 2017. Adaptive and non-adaptive divergence in a common landscape. Nature Communications.

Raeymaekers JAM, Konijnendijk N, Larmuseau MHD, Hellemans B, De Meester L, Volckaert FAM. 2014. A gene with major phenotypic effects as a target for selection vs. homogenizing gene flow. Molecular Ecology 23:162-181.

Stuart YE, Veen T, Weber JN, Hanson D, Ravinet M, Lohman BK, Thompson CJ, Tasneem T, Doggett A, Izen R, et al. 2017. Contrasting effects of environment and genetics generate a continuum of parallel evolution. 1:0158.

Thursday, August 3, 2017

Journal Life List

Steve Heard just posted a blog on his journal life list – all of the journals in which he has published and how his papers are distributed across those journals. Kind of like a birder’s life list of species. Dividing the number of journals by the number of papers gives a very rough JOURNAL DIVERSITY INDEX (JDI**) for an author – Steve’s is 0.61, which he expects to be high. Although I hadn’t planned to write a blog on this topic, it reminded me that I am always excited when I expand my journal life list, favoring JDI increase. Yet at the same time, I tend to target particular journals that I think are in my core research area, favoring JDI decrease. So I quickly (on the train home) did an analysis similar to Steve’s, with some additions.

At first blush, it is clear that my JDI of 0.28 is way lower than Steve’s, presumably because I tend to target core journals: Evolution, The American Naturalist, and Journal of Evolutionary Biology. But then second blush made me realize that the above contrasting motivations (increase diversity vs. focus on core journals) is a function of the joint combination of submissions and acceptances given submissions. Hence, I next did the same analysis but based not on where papers were published but on where I first submitted them.

The general trend, and indeed the JDI (0.23), are similar – suggesting that I really do focus on core journals over journal diversity. But this number is probably biased by a single journal – Nature – to which I have often submitted first but in which I have only rarely published.* Deleting Nature gives a JDF of 0.26, so I still tend to favor core journals over diversity. In fact, it is interesting that my JDF for submissions is lower than that for publications, suggesting that my rejection rate at the core journals is slightly higher than at other journals. I therefore next compared numbers of submissions to journals with numbers of publications in journals.

The two are highly correlated, as one would expect, with my core journals (apart from Am Nat) seeming to have roughly similar or higher than expected acceptance rates. Of course, the difference between these numbers are not strictly acceptance rates because I sometimes publish in journals, including my core journals, papers that I previously submitted elsewhere. At the extreme, I have published in 11 journals to which I never done a first submission of a paper. At the other extreme, I have submitted first to 4 journals in which I have never published a paper.

So, what to make of this beyond my apparent emphasis on core journals. Perhaps it is that journals should give out frequent publisher miles that might, for example, lead to waived publication fees. For instance, I have published 19 papers in Evolution, 17 papers in Journal of Evolutionary Biology, and 11 papers in Molecular Ecology. However, I don’t think I have really had to pay publication fees for most of those. Maybe better would be discounted attendance at annual meetings – yeah, I like that.


* I have included (for ease of generating this quickly), all publications in a journal even if they are introductions or notes or comments or news pieces. For example, my publications in Nature are News & Views and the like. 
** This is just a quick and simple metric - see Steve's blog for more comments on it.
*** Perhaps my favorite bird cartoon is below - simply a gratuitous blog for

Sunday, July 9, 2017

25 years later – Alaska on my mind

Several years ago, I wrote a blog post called “Trinidad on my mind”, for which I extracted a number of journal entries from my years of research in Trinidad. Today I am on a plane on my way to Haida Gwaii, British Columbia, for field work and I was debating what would be a post I would enjoy writing. I started digging through old journals on my computer – dating back to 1992, when I got my first computer. Among these were notes I had written from my pre-MSc work in Alaska – I ended up spending parts of 10 consecutive summers in Alaska. Much of the writing is a bit too flowery to put on this blog but reading it was fun enough to make me want to extract and compile some for a post “Alaska on my mind.” My edits are minimal – this is pretty much what I wrote at the time. I will presumably follow this post at some future date with other entries from field work.

Incidental catch

It was the summer of 1992 and I had a job as a deck hand on a tender converted to a gill-netter.  I was employed by the University of Washington to participate in a test fishery which was used to predict the size and timing of each year's Bristol Bay sockeye salmon run.*

It was 4:00 P.M. on June 27th.  It was miserable.  The sea was rough and the wind chill, blowing at 25 knots.  Flying spray stung cheeks, eyes smarted from the salt.  Waves broke over the back of the boat immersing us continually as we worked to strip salmon from the dripping gill net coming over the stern.  We were on our fourth set of the day.  Each of the earlier sets had broken records for numbers of salmon taken this year so far.  We were cold, wet, tired and hungry all at the same time.  I loved it.

Sadly, our Captain, Blake (shown here), perished at sea a year later.
There was something aesthetically pleasing about wading ankle deep in salmon while racing the other crew members to see who could remove the most fish from the net without breaking the mesh.  Every so often an unexpected bonus would come along in the form of a 30 pound king salmon.

On seeing the big grey form in the net a hundred feet away, my first thought was that we had an even bigger king.  A second look however, and I realized that this was much larger than any king I had ever seen. "Dolphin!"  It was an exclamation of pure horror from the entire crew when we realized what was in the net.  We had all heard stories of the horrors of drift nets but this was striking much too close to home.  Almost instantly, everyone was galvanized into frenzied action.

Mike quickly reeled the net and the surprisingly placid harbor porpoise to the back of the boat.  Chris and I leaned out and began to frantically break the mesh around the porpoise.  Blake manned the boat controls giving us as much stability as possible to work.  Initially, the porpoise was so calm, I feared it was already dead but when we had it beside the boat we could see it was moving slightly.  It didn't appear to be badly injured. and the only sign of damage was blood from a cut on the dorsal fin.  Was it drowning?

From our position at the extreme stern of the boat, we were still 4 or 5 feet from the porpoise but the holes we were ripping in the net were getting us closer.  I was leaning out so far and pulling so hard on the mesh that my legs lifted 2 feet off the deck before Mike grabbed them and held me from going overboard.  The following sea continued to wash waves through the stern opening threatening to wash everything away.  I continued to pull on the mesh until it felt like my fingers were going to be amputated.  A last convulsive heave on the net and the porpoise's body fell free.  Realizing now it was close to freedom, it began to try and swim with only its tail still caught in the mesh.  Then in a frenzy it ripped free.  As it swam slowly off, we all watched in physical and emotional exhaustion.

Bristol Bay, Alaska

The Russians Have Landed

Scholarship is a great way to cross national and international boundaries.  Scholarship brings people together where national policy and human kindness fail.  The  Bering Sea has a western coast line as well as its better known eastern shore in Alaska.  Not surprisingly, the rivers of the Kamchatka and Siberia also welcome the return of the salmon each summer.

Four Russian fisheries biologists visited us at our camp on Porcupine Island, Illiamna.  They had a translator with them and during their visit discussed common problems, ongoing studies, and issues in fisheries management with the professors from the University of Washington.  Apparently, this trip was also the first for the translator and we could usually communicate as well without her help as with it.  She did spend a long time teaching me some survival words in Russian.  Basically, "I am hungry, feed me now."  This brought my command of that phrase to a total of five languages. 

What I remember most though was that these two particular Russians were musical and, in the evening, provided rare entertainment.  One played the accordion and the other the guitar.  Picture this.  Two Russians who spoke virtually no English entertaining a bunch of American graduate students and professors in Alaska with classic American cowboy songs, sung in English.  It was great.  Their favorite, and mine, was "Ghost Riders in the Sky."    Someday, I may want to study fisheries in Kamchatka, go fishing on the other side of the Bering sea or simply to listen to a Russian concert of cowboy songs or a Beatles medley.

Alaska underwater

Research at Iliamna was really fun because most of it was in the water.  Although I had not brought my dry suit and had no initial research project, I managed to weasel my way into an average of an hour in the water each day.  It was quite cold (8-10 degrees C) considering the old wet suit I had, which was too large.  After a 1/2 hour or so I would be chilly, cold at 45 minutes and shivering by an hour.  I took to wearing socks, long underwear and two T-shirts to reduce water flow through my suit.

I had several jobs.  One was to collect sculpins.  To do this, I would simply kick my feet on the rocks in a salmon spawning area.  When the sculpins swam out I would dive down and catch them in an aquarium net.  After I had a few in the net, I would swim to shore and put them in a floating cage.  While I quite enjoyed it, several times I had to collect 100 of the little devils after I had already been in the water for over an hour at one stretch.  During these episodes, I was always amazed at how many sculpin sized crevices could for refuges for my diminutive prey. A second job was to count sculpins in a series of quadrants.  This was rather mundane and consisted of digging through all the rocks in 14, 1m2 quadrants and tallying the sculpins in four size classes.

I also had my own study, designed to determine how separate two adjacent spawning populations actually were.  Every day I would swim a 100m transect which ran across both populations and record the condition of each female and the number of attendant males in each 2.5m section.  The second part was more fun.  I used a 'Hawaiian sling', with a sharpened tube loaded with a barbed tag, to tag the right side of 22 males at one site and the left side of 22 males at the other site.  I did this twice, the second time using 1/2 sized tags.  Then, each day I would record the numbers of males of each tag type at each site along the transect.  It was a blast but occasionally I would make three or four dives in a row where every male had its wrong side toward me.  Twice, I was able to tag two fish on a single dive.**

That's all for now. I will revisit some field notes again at some future date. 


* I also worked on this test fishery in 1993. Later that summer, our boat – the 69’ Nettie H – sank with all hands lost, including our captain for my two years on the test fishery, Blake Grinstein.

** I later published this as my first first-authored paper. 

Yep - that's right - the first study I ever led was at a site called "Fuel Dump Island"

Thursday, June 22, 2017

At the Speed of Life

Twenty years ago, Mike Kinnison and I started thinking seriously about evolutionary rates. Over the years since, we – with many collaborators – have assembled a database of evolutionary rates that has been used by ourselves and many others for various purposes. We are currently in the midst a big new push to add additional studies to the database and we would like YOUR help. In short, if you have studied rates of phenotypic change in contemporary populations, we think it would be cool to have your study in the database. Information on how to participate is at the end of this blog post that traces the history of the database.

The whole thing accelerated for us in 1997, when Mike Kinnison and I were graduate students sitting in adjacent desks at the School of Fisheries in the University of Washington. At the suggestion of our prescient supervisor, Tom Quinn, we were both studying salmon introduced to new environments with an eye to how rapidly they might be evolving. Mike was working on chinook salmon introduced into New Zealand and I was studying sockeye salmon introduced into Lake Washington. And we were finding plenty of evidence that, in both cases, substantial evolutionary change had occurred since the introductions.

Mike Kinnison checking on his chinook salmon in the mid 1990s, New Zealand.
Me with a sockeye salmon the mid 1990s - here in Alaska. 
In the midst of our work in 1997, two high profile papers were published that documented similarly rapid phenotypic changes in populations experimentally introduced into new environments – guppies in Trinidad (Reznick et al. – Science) and Anolis lizards in the Caribbean (Losos et al. – Nature). (These came to our attention via a commentary in TREE written by Erik Svensson.) A few papers showing pretty much the same thing had come out previously, but these two new papers were different in one important respect – both formally quantified evolutionary rates for their studies and compared them to evolutionary rates reported in the fossil record. For both guppies and lizards, evolution in contemporary time was several orders of magnitude more rapid than evolution typically observed in the fossil record, which brought a lot of scientific and media attention to how rapid ongoing evolution seemed to be. In short, instead of the “we see nothing of these slow changes in progress until the hand of time has marked the long lapse of ages” wisdom received from Darwin, a more correct conclusion might be “we can often see rapid changes in progress even though the hand of time has barely moved.”

“Wow, that’s pretty cool”, Mike and I thought, “let’s calculate rates of evolution in our own study systems and see how fast they are – which was when the whole thing began to complexify. As we delved into the details of how evolutionary rates were calculated and interpreted, it became clear that many ambiguities, uncertainties, caveats, and subtleties were involved. This caution seemed to us important enough – and certainly relevant enough to our own work – to warrant writing something about it, which we duly published in 1998 as a short note (Taking Time with Microevolution) in TREE.

My first publication with Mike. When I asked him what his middle initial stood for, he insisted it was "Michael The Kinnison."

After publishing this note, Mike and I continued to work on the problem and it became abundantly clear that it warranted a full-length treatment about how best to estimate evolutionary rates and that also calculated and reported those rates for a bunch of studies beyond the only two that (at that time) had done so for contemporary populations. The problem, of course, was finding the time to write the paper while we were both in the throes of finishing our PhDs and starting our postdocs, which were on quite different topics.

Fortunately, right about this time, I was invited to work on a project in southern France where data collection took place mainly by camera recordings. Thus, once the experiment was set up, I had some extra time to perhaps write something else up. Although I certainly had tons of things to write up, the different venue provided mental freedom enough to say, “why not try to blast out this evolutionary rate paper.” So I very rapidly wrote up a first draft that Mike and I then edited and – without much expectation – submitted it to Evolution as a Perspective. This submission was a bit of a stretch, or at least leap of faith (or hope), for Mike and I because – at this point – neither of us had published anything in an evolutionary journal (instead only in fish journals), nor had we really had much feedback from real evolutionary biologists (although we did get good comments from the Huey/Kingsolver lab meetings).

Lo and behold, and somewhat to our surprise, Evolution was happy to publish the paper: The Pace of Modern Life: Measure Rates of Contemporary Microevolution. In the immediate aftermath, or really lack thereof, not much happened – nor had we expected it to. In fact, we kind of viewed the paper as our own little curiosity project that helped us to think more deeply about our own empirical work. We guessed it might be a bit interesting to some few other researchers but probably of little practical use or importance to most. So, once published, we continued our progress toward other topics.

Database version 1.0 - from Hendry and Kinnison (1999).
Then – out of the blue – came an invitation from a journal (Genetica), we didn’t know much – if anything – about, to edit a special issue on contemporary evolution. Being quite green postdocs who had never received such an invitation, we immediately agreed. (Nowadays, of course, the many predatory journals are constantly inviting even inexperienced, as we were at the time, people to edit special issues.) Importantly, we decided to invite every famous evolutionary biologist that we knew of (but none of which we really knew personally) who might contribute moderately relevant papers: Grant, Reznick, Losos, Sinervo, Lande, Arnold, Riechert, Wade, Kingsolver, Gingerich, Magurran, Bell, Merila, Raymond, Smith, etc. Remarkably, nearly all of them agreed, something that still amazes me today – and, in fact, which I doubt would occur today.

In our invitations, we took a strategy which I now adopt in nearly every special issue that I edit – and there have been a lot of them: invite the best people, give them the general topic (in this case, Microevolution: rate, pattern, and process), and tell them they can write specifically about whatever they want to write about in that general area. This strategy really is the best way to make sure you get great people and great topics – although the cleverness of it was not something we knew in advance. The only specific additional request we gave was that, if they were reporting phenotypic changes over relatively short time frames (several hundred years or fewer), they should formally calculate and report evolutionary rates in accordance with our suggestions from the earlier Evolution paper.

The main paper that Mike and I contributed to the Genetica special issue was a follow up of our original Pace of Life paper, wherein we now calculated evolutionary rates for more previous studies and formally analyzed the resulting database to try to answer big questions about evolution. What is the distribution of contemporary evolutionary rates and how do they compare to selection intensities? Do different types of traits evolve at different rates? How do evolutionary rates scale with time interval? In the end, the special issue had 30 contributions, a number of which have been very heavily cited (7 more than 100 times on Web Of Science).

Database Version 2.0 from Kinnison and Hendry (2001).
In 2002 or thereabouts, both Mike and I started faculty positions, me at McGill and Mike at U Maine. (I had interviewed for the same position at Maine but they chose Mike. Of course, things turned out well for the both of us.) Earlier that year, I had given a talk on contemporary evolution in the ASN Young Investigator Prize symposium and the editor of TREE, who had seen the talk, asked if I wanted to write a paper on contemporary evolution for the journal. I proposed several options, one of which was the implications of contemporary evolution for conservation biology, which was the topic the editor chose. However, I knew little about conservation and both Mike and I felt a bit overwhelmed in our new jobs, so we contacted Craig Stockwell, who worked on both contemporary evolution and conservation. This was an excellent decision as Craig very cleverly tied the two fields together, and generated a paper (Contemporary Evolution Meets Conservation Biology) that was quite influential, with 870 citations on Google Scholar.

Database Version 2.1 (really, the same as 2.0) from Stockwell et al. (2003).
But the actual database of rates was getting quite out of date. Fortunately, I taught a graduate class at McGill where one of the student task was to write a review/meta-analysis. I pitched to several students the idea of expanding the evolutionary rate database, and Thomas Farrugia – and undergrad at the time – was interested. So we then spent several months adding papers to the database and reanalyzing it to follow up the Genetic paper. A great venue for publishing this work then presented itself as Tom Smith and Louis Bernatchez were organizing a symposium at UCLA on Evolutionary Change in Human Altered Environments and were going to edit a special issue in Molecular Ecology.

Me and Mike and the mid-2000s - in this case rockin the facial hair in Trinidad.
Thomas had moved on to other things by then, and so the task fell to me to finish up the paper, which I spun into a consideration relevant to the symposium/issue: whether or not human influences increased rates of phenotypic change. Turns out they did and the paper (Human Influences on Rates of Phenotypic Change in Wild Animal Populations) has been heavily cited in that regard. A reviewer also forced me to do a quick analysis, which I felt the database was not large enough for yet, that considered which types of human disturbance led to the greatest rate of change. This last topic proved to resonate with others into the future, as we will see below.

Database Version 3.0 from Hendry et al. (2008).
A first, reviewer-insisted, analysis of rates of change associated with different types of human disturbance - from Hendry et al. (2008).
A difference this time was that we also published the existing database online, which meant that anyone could grab it and re-analyze it (usually with additions) to answer their own questions. As result, papers have now come arguing that rates of change are greatest when humans act as predators (Darimont et al. 2009 - PNAS) but are not especially noteworthy in the case of invasive species (Westley et al. 2011 – Am Nat), that evolution should be possible in response to climate change (Skelly et al. 2007 - Cons Biol), and that trait evolution should have ecological consequences (Palkovacs et al. 2012). The database has also been used to examine the links between micro- and macro-evolution (Uyeda et al. 2011 - PNAS), the pace of cultural evolution (Perreault et al. 2012 - PLoS ONE), and a wide variety of other topics.

Through the years, we have expanded the database a bit – and fixed various errors and inconsistencies – to then re-analyze and publish papers showing that (for example) body size does not evolve faster than other traits (Gotanda et al. 2015 - Evolution) and that rates of change are especially high in some urban contexts (Alberti et al. 2017 - PNAS).

Database Version 4.0 from Alberti et al. (2017).
Yet all of these analyses are still working with a database that, while large and growing, is very incomplete. A large number of relevant studies of phenotypic in natural populations have been published in the last 20 years that could be added to the database but haven’t been – simply for reasons of obscurity (to us) or time (for us). Thus, when I was asked this year to help write the Intergovernmental Platform on Biodiversity and Ecosystem Services (IPBES) Global Assessment, I realized we needed to do a major database expansion and overhaul. Hence, over the past half year or so we (especially my new student Sarah Sanderson) have been accumulating studies, extracting data on phenotypic changes, and adding them to the database. Although we have added many new studies, it is clear that many more are still out there – yet they are sometimes hard to find.

This need, then, is the impetus for this blog post – we would like your help in finding as many relevant studies as possible. Sarah generated a list of all studies in the database (we have some others still being entered), which we provide here. If you study phenotypic change in contemporary time, it would be fantastic if you could skim this list to see if your study is included. If not, we would love it if you could contact us to tell us about your study so that it can be included. Importantly, the database will – when reasonably finished – be provided online for all to use, which can only increase its (and your study’s) value to the scientific community. We hope that you will join us in this endeavor as it has been fun and profitable for me – and it can be for you to.

Cheers and thanks,


Note: we are looking for studies of changes in quantitative traits (body size, morphology, phenology, etc.) over known time frames. These can come from information on the same population in different years (allochronic) or from different populations that had a common ancestor at a known time in the past (synchronic). It is also not necessary to have confirmation of the genetic basis for the change as people are also interested in plasticity - and, regardless, these aspects are coded for each study in the database. 

Sunday, June 11, 2017

How to Tweet

I was a reluctant convert to social media. Ten years ago, I didn’t see the point of it. I didn’t have facebook, twitter, Instagram, a blog, or a youtube channel. Now I have all of these and am quite active on several. I will first outline my social media development as a background for the advice I will then suggest about social media for scientists. This advice needs to be tempered by the fact that I am not a social media icon – nor that I am especially good at it. However, I have found it very useful professionally, and often entertaining personally. (My instagram and flickr accounts are used for showing nature photos I have taken over the years, and so are not really discussed here.)

The transition started about eight years ago when my newer students convinced me that a blog would be a good idea as it was an increasingly common way for people to get scientific information. I started the blog but, at first, just kind of did it lip service without much effort and resented the need to have to generate content on a regular basis. But then people started reading and responding to some of my posts and their value became clear. Then, a few years ago, I started a series of “How To” posts that provided advice for young scientists. This series became surprisingly popular (see the table below) and useful to people to the point that I have numerous people come up to me at meetings to say they follow my blog and find it helpful. It was even featured twice in Nature magazine.

A couple of years into this blogging adventure, it became clear that twitter was a great way to promote the blog – and so I signed on in March 2014. Again, I didn’t really pay much attention at first but, as time went on, I realized its usefulness at both send and receiving information and quickly gained a reasonable number of followers. I don’t have any viral tweets – not that I seek them – but several have been retweeted more than 1000 times (an example is below). My favorite current use of twitter is to promote my book through a fun #PeopleWhoFellAsleepReadingMyBook thread, storified here.

At about the same time as I started on twitter, I initiated a YouTube channel. The goal was not to be a famed “youtuber” but rather to simply share some of my wildlife and nature videos, mostly with family and friends but also with colleagues and with students in my classes. I then added a number of teaching-oriented videos, especially the drunkard’s walk, and some personal climbing parody videos. (In retrospect, it would have been better to keep the professional and personal youtube channels separate.) Again, I am very surprised by how many times some of these quick productions have been viewed, such as those shown below.

In short, social media has become an integral part of my professional life and I think it is a great way to provide information and advice to a much broader range of people than otherwise possible. Hence, my goal in the rest of this post is to make a case for the value of social media for scientists and to provide some suggestions on how to engage, with some of these suggestions being different from those you might hear elsewhere.

Why use social media?

1. A lot of scientists not on social media think it is a triviality for which they “do not have time.” I don’t agree with this proposition; indeed, I am sure I don’t have any more time than they do. The reason I disagree is that, for me, social media has simply replaced other time-wasting activities. When working on my computer, I get bored just like anyone else and I used to procrastinate by going to cnn or espn or cbc or whatever other traditional media sites. Now, I simply replace most of that procrastination time by looking at twitter or instagram, which is much more useful as it is related to my work.

2. Some scientists are of the opinion that social media is just filled with a bunch of uninteresting stuff that will swamp them with nonsense. This proposition is certainly true if you aren’t careful and selective. However, the great benefit of social media is just the opposite – you can tailor your own news feed, making it much more targeted and useful to you than more traditional forms of media. For instance, I follow colleagues, students, journals, universities, departments, and so on. If the information I am getting isn’t useful, I simply unfollow that person. Indeed, I how have my twitter feed whittled down to a great set of complementary news sources.

3. Social media is a good way to get scientific information. If you tailor your feed appropriately, you can get lots of pointers to good new papers that are coming out that you wouldn’t otherwise see. In the old days, you could pretty much cover ALL the relevant literature by just skimming the contents of a dozen or so journals. Now there is an order of magnitude more science out there, making it impossible to find everything – but the right feed can reveal to you some of the best stuff coming out now.

4. Social media helps you reach a broader audience – and, hence, better promote your work and ideas. As just noted above, it is impossible to keep up with all the publications of others. For the same reason, your own papers are likely to be lost under in the avalanche of information that is out there. It is not enough to simply publish your work any more – you need to promote it. Social media is one way to do so as students and colleagues in your field will often follow you and see your posts.

How to use social media?

The following suggestions are based on my experiences over the past five years or so. Again, these are just my opinions, with which others will not necessarily agree. I tend to refer mostly to twitter below, but the same basic points apply to other platforms such as facebook, instragram, blogging, youtube, and so on. I also image they will be most useful to people who have not already “found their own way” on social media – but perhaps a few will be interesting even to experienced folks.

1. Be selective in who you follow. Some people will argue that the “contract” of social media is to follow back – that is, when someone follows you, you immediately reciprocate. However, this approach doesn’t work for me. To not be overwhelmed, and to be sure to see the stuff I want to see, I keep the number of accounts I follow to a manageable number – about 100 on twitter. The alternative is to follow more but “mute” them. However, the muting strategy seems to me rather dishonest. If I follow you, then I don’t mute you.

2. Be selective in what you tweet. Analyses have been done about how often one should tweet/post. I don’t know the details of those analyses but I am sure that they all say that too-frequent posting is not beneficial. Instead, retweet/post those things that are truly interesting and that you think would be interesting to your followers.

3. Re-tweet interesting posts from diverse people and feeds. Although I don’t follow everyone that follows me, I do periodically check out the feeds of my followers to find interesting posts of theirs to re-tweet. I think this help can be especially helpful to graduate students who don’t yet have a large following on social media.

4. Add a visual. Almost every one of my tweets includes an image of some sort. Images are much more likely to catch the eye of someone who is skimming quickly and thus inspire them to stop and actually read it. This comes partly from personal experience – I generally don’t pay attention to tweets that are text only. This isn’t a philosophy or anything – it is just the practicality of not having time to read everything and, while a tweet is only 140 characters, a picture is worth 1000 words!

5. Acknowledge, either by tagging or through a link, the source of the material – especially images – that you put on social media. And, along the same lines, don’t retweet people who don’t acknowledge their sources. For example, many of the aggregating feeds that post cool pictures don’t credit the original photographer. Don’t retweet them.

6. Be apolitical, at least usually, unless your goal is to be political in your professional life. I subscribe to the perspective that scientists should be as objective as possible if we are to maintain our credibility as experts. Many social media users, by contrast, spend tons of time criticizing universities, journals, funding agencies, colleagues, conference organizers, etc. I have no use or time for these diatribes. (Of course, I do retweet the occasional well-thought out or funny piece about idiotic and dangerous demagogues.)

7. Don’t troll/shame people – unless you want to be known as a troll. Sure, trolls tend to get the most followers, reminding of me of when David Houle wrote in a book review that “Negative reviews often give a frisson of pleasure to the reader.” In my opinion, however, negative comments reflect more on the commenter than the commentee. Instead, generate occasional and thoughtful positive criticism that helps move things forward in a constructive way.

8. Don’t blindly retweet everything you see about social/scientific equality. For instance, many white male profs seem to retweet everything they see about the disadvantages facing women and people of color. Those disadvantages are real, and need our recognition. However, the more someone not in those categories tweets about them, the less sincere that person seems - at least to me. Instead, periodically retweet THE BEST posts about social/scientific equality. Your sincerity and credibility will, I think, benefit from it. Here is the blog post I wrote about my own “Subtle Sexism Self-Evaluation.”

My most retweeted tweet.

And finally

If you goal is to be a scientist, then tweeting will not do the job for you. You need to publish your own research, which you can then supplement and promote with social media. A social media profile will not replace a real research portfolio – only enhance it. Of course, if your goal is to forge a career in science communication, then social media would take on greater importance.

Yikes, I (red dot) appear to be close to being a scientific Kardashian! Original paper here.