Guest Post: You teach science, but is your teaching scientific? (Part I)

The first in a series of guest posts about using scientific teaching, active learning, and flipping the classroom by Sarah Seiter, a teaching fellow at the University of Colorado, Boulder. 

As a faculty member teaching can sometimes seem like a chore – your lectures compete with smartphones and laptops. Some students see themselves as education “consumers” and haggle over grades. STEM (science, technology, engineering, and math) faculty have a particularly tough gig – students need substantial background to succeed in these courses, and often arrive in the classroom unprepared. Yet, the current classroom climate doesn’t seem to be working for students either. About half of STEM college majors ultimately switch to a non-scientific field. It would be easy to frame the problem as one of culture – and we do live in a society that doesn’t always value science or education. However, the problem of reforming STEM education might not take social change, but rather could be solved using our own scientific training. In the past few years a movement called “scientific teaching” has emerged, which uses quantitative research skills to make the classroom experience better for instructors as well as students.

So how can you use your research skills to boost your teaching? First, you can use teaching techniques that have been empirically tested and rigorously studied, especially a set of techniques called “active learning”. Second, you can collect data on yourself and your students to gauge your progress and adjust your teaching as needed, a process called “formative assessment”. While this can seem daunting, it helps to remember that as a researcher you’re uniquely equipped to overhaul your teaching, using the skills you already rely on in the lab and the field. Like a lot of paradigm shifts in science, using data to guide your teaching seems pretty obvious after the fact, but it can be revolutionary for you and your students.

What is Active Learning:

There are a lot of definitions of active learning floating around, but in short active learning techniques force students to engage with the material, while it is being taught. More importantly, students practice the material and make mistakes while they are surrounded by a community of peers and instructors who can help. There are a lot of ways to bring active learning strategies to your classroom, such as clicker response systems (handheld devices that allow them to take short quizzes throughout the lecture). Case studies are another tool: students read about scientific problems and then apply the information to real world problems (medical and law schools have been them for years). I’ll get into some more examples of these techniques in post II; there are lots of free and awesome resources that will allow you to try active learning techniques in your class with minimal investment.

Formative Assessment:

The other way data can help you overhaul your class is through formative assessment, a series of small, frequent, low stakes assessment of student learning. A lot of college courses use what’s called summative assessment – one or two major exams that test a semester’s worth of material, with a few labs or a term paper for balance. If your goal is to see if your students learned anything over a semester this is probably sufficient. This is also fine if you’re trying to weed out underperforming students from your major (but seriously, don’t do that). But if you’re interested in coaching students towards mastery of the subject matter, it probably isn’t enough to just tell them how much they learned after half the class is over. If you think about learning goals like we think of fitness goals, this is like asking students to qualify for the Boston marathon, without giving them any times for their training runs.

Formative assessment can be done in many ways: weekly quizzes or taking data with classroom clicker systems. While a lot of formative assessment research focuses on measuring student progress, instructors have lots to gain by measuring their own pedagogical skills. There are a lot of tools out there to measure improvement in teaching skills (K-12 teachers have been getting formatively assessed for years), but even setting simple goals for yourself (“make at least 5 minutes for student questions”) and monitoring your progress can be really helpful. Post III will talk about how to do (relatively) painless formative assessment in your class.

How does this work and who does it work for:

Scientific teaching is revolutionary because it works for everyone, faculty and students alike. However, it has particularly useful benefits for some types of instructors and students.

New Faculty: inexperienced faculty can achieve results as good or better than experienced faculty by using evidence based teaching techniques. In a study at the University of Colorado, physics students taught by a graduate TA using scientific teaching outperformed those taught by an experienced (and well loved) professor using a standard lecture style (you can read the study here). Faculty who are not native English speakers, or who are simply shy can get a lot of leverage using scientific teaching techniques, because doing in-class activities relieves the pressure to deliver perfect lectures.

Test scores between a lecture-taught physics section
and a section taught using active learning techniques.

Seasoned Faculty: For faculty who already have their teaching style established, scientific teaching can spice up lectures that have become rote or help you address concepts that you see students struggle with year after year. Even if you feel like you have your lectures completely dialed in, consider whether you’re using the most cutting edge techniques in your lab, and if you your classroom deserves the same treatment.

Students also stand to gain from scientific teaching, and some groups of students are particularly poised to benefit from it:
Students who don’t plan to go into science: Even in majors classes, most of the students we teach won’t go on to become scientists. But skills like analyzing data, and writing convincing evidence based arguments are useful in almost any field. Active learning trains students to be smart consumers of information, and formative assessment teaches students to monitor their own learning – two skills we could stand to see more of in any career.

Students Who Love Science: Active learning can give star students a leg up on the skills they’ll need to succeed as academics, for all the reasons listed above. Occasionally really bright students will balk at active learning, because having to wrestle with complicated data makes them feel stupid. While it can feel awful to watch your smartest students struggle, it is important to remember that real scientists have to confront confusing data every day. For students who want research careers, learning to persevere through messy and inconclusive results is critical.

Students who struggle with science: Active learning can be a great leveler for students who come from disadvantaged backgrounds. A University of Washington study showed that active learning and student peer tutoring could eliminate achievement gaps for minority students. If you partially got into academia because you wanted to make a difference in educating young people, here is one empirically proven way to do that.

Are there downsides?

Like anything, active learning involves tradeoffs. While the overwhelming evidence suggests that active learning is the best way to train new faculty (the white house even published a report calling for more of it!), there are sometimes roadblocks to scientific teaching.

Content Isn’t King Anymore: Taking time to work with data, or apply scientific research to policy problems takes more time, so instructors can cover fewer examples in class. In active learning, students are developing scientific skills like experimental design or technical writing, but after spending an hour hammering out an experiment to test the evolution of virulence, they often feel like they’ve only learned about “one stupid disease”. However, there is lots of evidence that covering topics in depth is more beneficial than doing a survey of many topics. For example, high schoolers that studied a single subject in depth for more than a month were more likely to declare a science major in college than students who covered more topics.

Demands on Instructor Time: I actually haven’t found that active learning takes more time to prepare –case studies and clickers actually take a up a decent amount of class time, so I spend less time prepping and rehearsing lectures. However, if you already have a slide deck you’ve been using for years, developing clicker questions and class exercises requires an upfront investment of time. Formative assessment can also take more time, although online quiz tools and peer grading can help take some of the pressure off instructors.

If you want to learn more about the theory behind scientific teaching there are a lot of great resources on the subject:

These podcasts are a great place to start:
http://americanradioworks.publicradio.org/features/tomorrows-college/lectures/

http://www.slate.com/articles/podcasts/education/2013/12/schooled_podcast_the_flipped_classroom.html

This book is a classic in the field:
http://www.amazon.com/Scientific-Teaching-Jo-Handelsman/dp/1429201886

By Sarah Seiter

Null models matter, but what should they look like?

Neutral Biogeography and the Evolution of Climatic Niches. Florian C. Boucher, Wilfried Thuiller, T. Jonathan Davies, and Sébastien Lavergne. The American Naturalist, Vol. 183, No. 5 (May 2014), pp. 573-584

Null models have become a fundamental part of community ecology. For the most part, this is an improvement over our null-model free days: patterns are now interpreted with reference to patterns that might arise through chance and in the absence of ecological processes of interest. Null models today are ubiquitous in tests of phylogenetic signals, patterns of species co-occurrence, models of species distribution-climate relationships. But even though null models are a success in that they are widespread and commonly used, there are problems--in particular, there is a disconnect between how null models are chosen and interpreted and what information they actually provide. Unfortunately, simple and easily applied null models tend to be favoured, but they are often interpreted as though they are complicated, mechanism-explicit models.

The new paper “Neutral Biogeography and the Evolution of Climatic Niches” from Boucher et al. provides a good example of this problem. The premise of the paper is straightforward: studies of phylogenetic niche conservation tend to rely on simple null models, and as a result may misinterpret what their data shows because of the type of null models that they use. The study of phylogenetic niche conservation and niche evolution is becoming increasingly popular, particularly studies on how species' climatic niches evolve and how climate niches relate to patterns of diversity. In a time of changing climates, there are also important applications looking at how species respond to climatic shifts. Studies of changes in climate niches through evolutionary time usually rely on a definition of the climate niche based on empirical data, more specifically, the mean position of a given species along a continuous abiotic gradient. Because this is not directly tied to physiological measurements, climate niche data may also capture the effect of dispersal limitations or biotic interactions. Hence the need for null models, however the null models used in these studies primarily flag changes in climate niche that result from to random drift or selection in a varying environment. These types of null models use Brownian motion (a "random walk") to answer questions about whether niches are more or less similar than expected due to chance, or else whether a particular model of niche evolution is a better fit to the data than a model of Brownian motion.

The authors suggest that the reliance on Brownian motion is problematic, since these simple null models cannot actually distinguish between patterns of climate niches that arise simply due to speciation and migration but no selection on climate niches, and those that are the result of true niche evolution. If this is true, conclusions about niche evolution may be suspect, since they depend on the null model used. The authors used a neutral, spatially explicit model (known as an "alternative neutral biogeographic model") that simulates dynamics driven only by speciation and migration, with species being neutral in their dynamics. This provides an alternative model of patterns that may arise in climate niches among species, despite the absence of direct selection on the trait. The paper then looks at whether climatic niches exhibit phylogenetic signals when they arise via neutral spatial dynamics; if gradualism a reasonable neutral expectation for the evolution of climatic niches on geological timescales; and whether constraints on climatic niche diversification can arise simply through bounded geographic space. Simulations of the neutral biogeographic model used a gridded “continent” with variable climate conditions: each cell has a carrying capacity, and species move via migration and split into two species either by point mutation, or else by vicariance (a geographic barrier appears, leading to divergence of 2 populations). Not surprisingly, their results show that even in the absence of any selection on species’ climate niches, patterns can result that differ greatly from a simple Brownian motion-based null model. So the simple null model (Brownian motion) often concluded that results from the more complex null model were different from the random/null expectation. This isn't a problem per se. The problem is that currently interpretations of the Brownian motion model may be that anything different from null is a signal for niche evolution (or conservation). Obviously that is not  correct.

This paper is focused on the issue of choosing null models for studies of climate niche evolution, but it fits into a current of thought about the problems with how ecologists are using null models. It is one thing to know that you need and want to use a null model, but it is much more difficult to construct an appropriate null model, and interpret the output correctly. Null models (such as the Brownian motion null model) are often so simplistic that they are straw man arguments – if ecology isn't the result of only randomness, your null model is pretty likely to be a poor fit to the data. On the other hand, the more specific and complex the null model is, the easier it is to throw the baby out with the bathwater. Given how much data is interpreted in the light of null models, it seems that choosing and interpreting null models needs to be more of a priority.

Has science lost touch with natural history, and other links

A few interesting links, especially about the dangers of when one aspect of science, data analysis, or knowledge receives inordinate focus.

A new article in Bioscience repeats the fear that natural history is losing its place in science, and that natural history's contributions to science have been devalued. "Natural history's place in science and society" makes some good points as to the many contributions that natural history has made to science, and it is fairly clear that natural history is given less and less value within academia. As always though, the issue is finding a ways to value useful natural history contributions (museum and herbarium collections, Genbank contributions, expeditions, citizen science) in a time of limited funds and (over)emphasis on the publication treadmill. Nature offers its take here, as well.

An interesting opinion piece on how the obsession with quantification and statistics can go too far, particularly in the absence of careful interpretation. "Can empiricism go too far?"

And similarly, does Big Data have big problems? Though focused on applications for the social sciences, there are some interesting points about the space between "social scientists who aren’t computationally talented and computer scientists who aren’t social-scientifically talented", and again, the need for careful interpretation. "Big data, big problems?"

Finally, a fascinating suggestion about how communication styles vary globally. Given the global academic society we exist in, it seems like this could come in handy. The Canadian one seems pretty accurate, anyways. "These Diagrams Reveal How To Negotiate With People Around The World." 

Are we winning the science communication war?

Since the time that I was a young graduate student, there have been constant calls for ecologists to communicate more with the public and policy makers (Norton 1998, Ludwig et al. 2001). The impetus for these calls is easy to understand –we are facing serious threats to the maintenance of biodiversity and ecosystem health, and ecologists have the knowledge and facts that are needed to shape public policy. To some, it is unconscionable that ecologists have not done more advocacy, while others see a need to better educate ecologists in communication strategies. While the reluctance for some ecologists to engage in public communication could be due to a lack of skills that training could overcome, the majority likely has had a deeper unease. Like all academics, ecologists have many demands on their time, but are evaluated by research output. Adding another priority to their already long list of priorities can seem overwhelming. More fundamentally, many ecologists are in the business of expanding our understanding of the world. They see themselves as objective scientists adding to global knowledge. To these ‘objectivists’, getting involved in policy debates, or becoming advocates, undermines their objectivity.

Regardless of these concerns, a number of ecologists have decided that public communication is an important part of their responsibilities. Ecologists now routinely sit on the boards of different organizations, give public lectures, write books and articles for the public, work more on applied problems, and testify before governmental committees. Part of this shift comes from organizations, such as the Nature Conservancy, which have become large, sophisticated entities with communication departments. But, the working academic ecologist likely talks with more journalists and public groups than in the past.

The question remains: has this increased emphasis on communication yielded any changes in public perception or policy decisions. As someone who has spent time in elementary school classrooms teaching kids about pollinators and conservation, the level of environmental awareness in both the educators and children surprises me. More telling are surprising calls for policy shifts from governmental organizations. Here in Canada, morale has been low because of a federal government that has not prioritized science or conservation. However signals from international bodies and the US seem to be promising for the ability of science to positively influence science.

Two such policy calls are extremely telling. Firstly, the North American Free Trade Agreement (NAFTA), which includes the governments of Mexico, Canada, and the USA, which normally deals with economic initiatives and disagreements, announced that they will form a committee to explore measures to protect monarch butterflies. They will consider instituting toxin-free zones, where the spraying of chemicals will be prohibited, as well as the construction of a milkweed corridor from Canada to Mexico. NAFTA made this announcement because of declining monarch numbers and calls from scientists for a coordinated strategy.

The second example is the call from 11 US senators to combat the spread of Asian carp. Asian carp have invaded a number of major rivers in the US, and have their spread has been of major concern to scientists. The 11 senators have taken this scientific concern seriously, requesting federal money and that the Army Corps of Engineers devise a way to stop the Asian carp spread.

There seems to be promising anecdotal evidence that issues of scientific concern are influencing policy decisions. This signals a potential shift; maybe scientists are winning the public perception and policy war. But the war is by no means over. There are still major issues (e.g., climate change) that require more substantial policy action. Scientists, especially those who are effective and engaged, need to continue to communicate with public and policy audiences. Every scientifically informed policy decision should be seen as a signal of the willingness of audiences to listening to scientists and that communicating science can work.

References

Ludwig D., Mangel M. & Haddad B. (2001). ECOLOGY, CONSERVATION, AND PUBLIC POLICY. Annual Review of Ecology and Systematics, 32, 481-517.

Norton, B. G. 1998. IMPROVING ECOLOGICAL COMMUNICATION: THE ROLE OF ECOLOGISTS IN ENVIRONMENTAL POLICY FORMATION. Ecological Applications 8:350–364

Debating the p-value in Ecology

It is interesting that p-values still garner so much ink: it says something about how engrained and yet embattled they are. This month’s Ecology issue explores the p-value problem with a forum of 10 new short papers* on the strengths and weaknesses, defenses and critiques, and various alternatives to “the probability (p) of obtaining a statistic at least as extreme as the observed statistic, given that the null hypothesis is true”.

The defense for p-values is lead by Paul Murtaugh, who provides the opening and closing arguments. Murtaugh, who has written a number of good papers about ecological data analysis and statistics, takes a pragmatic approach to null hypothesis testing and p-values. He argues p-values are not flawed so much as they are regularly and egregiously misused and misinterpreted. In particular, he demonstrates mathematically that alternative approaches to the p-value, particularly the use of confidence intervals or information theoretic criteria (e.g. AIC), simply present the same information as p-values in slightly different fashions. This is a point that the contribution by Perry de Valpine supports, noting that all of these approaches are simply different ways of displaying likelihood ratios and the argument that one is inherently superior ignores their close relationship. In addition, although acknowledging that cutoff values for significant p-values are logically problematic (why is a p-value of 0.049 so much more important than one of 0.051?), Murtaugh notes that cutoffs reflect decisions about acceptable error rates and so are not inherently meaningless. Further, he argues that dAIC cutoffs for model selection are no less arbitrary. 

The remainder of the forum is a back and forth argument about Murtaugh’s particular points and about the merits of the other authors’ chosen methods (Bayesian, information theoretic and other approaches are represented). It’s quite entertaining, and this forum is really a great idea that I hope will feature in future issues. Some of the arguments are philosophical – are p-values really “evidence” and is it possible to “accept” or “reject” a hypothesis using their values? Murtaugh does downplay the well-known problem that p-values summarize the strength of evidence against the null hypothesis, and do not assess the strength of evidence supporting a hypothesis or model. This can make them prone to misinterpretation (most students in intro stats want to say “the p-value supports the alternate hypothesis”) or else interpretation in stilted statements.

Not surprisingly, Murtaugh receives the most flak for defending p-values from researchers working in alternate worldviews like Bayesian and information-theoretic approaches. (His interactions with K. Burnham and D. Anderson (AIC) seem downright testy. Burnham and Anderson in fact start their paper “We were surprised to see a paper defending P values and significance testing at this time in history”...) But having this diversity of authors plays a useful role, in that it highlights that each approach has it's own merits and best applications. Null hypothesis testing with p-values may be most appropriate for testing the effects of treatments on randomized experiments, while AIC values are useful when we are comparing multi-parameter, non-nested models. Bayesian similarly may be more useful to apply to some approaches than others. This focus on the “one true path” to statistics may be behind some of the current problems with p-values: they were used as a standard to help make ecology more rigorous, but the focus on p-values came at the expense of reporting effect sizes, making predictions, or careful experimental design.

Even at this point in history, it seems like there is still lots to say about p-values.

*But not open-access, which is really too bad. 

How are we defining prediction in ecology?

There is an ongoing debate about the role of wolves in altering ecosystem dynamics in Yellowstone, which has stimulated a number of recent papers, and apparently inspired an editorial in Nature. Entitled “An Elegant Chaos”, the editorial reads a bit like an apology for ecology’s failure at prediction, suggesting that we should embrace ecology’s lack of universal laws and recognize that “Ecological complexity, which may seem like an impenetrable thicket of nuance, is also the source of much of our pleasure in nature”.

Most of the time, I also fall squarely into the pessimistic “ecological complexity limits predictability” camp. And concerns about prediction in ecology are widespread and understandable. But there is also something frustrating about the way we so often approach ecological prediction. Statements such as “It would be useful to have broad patterns and commonalities in ecology” feel incomplete. Is it that we really lack “broad patterns and commonalities in ecology”, or has ecology adopted a rather precise and self-excoriating definition for “prediction”? 

.

We are fixated on achieving particular forms of prediction (either robust universal relationships, or else precise and specific numerical outputs), and perhaps we are failing at achieving these. But on the other hand, ecology is relatively successful in understanding and predicting qualitative relationships, especially at large spatial and temporal scales. At the broadest scales, ecologists can predict the relationships between species numbers and area, between precipitation, temperature and habitat type, between habitat types and the traits of species found within, between productivity and the general number of trophic levels supported. Not only do we ignore this foundation of large-scale predictable relationships, but we ignore the fact that prediction is full of tradeoffs. As a paper with the excellent title, “The good, the bad, and the ugly of predictive science” states, any predictive model is still limited by tradeoffs between: “robustness-to-uncertainty, fidelity-to-data, and confidence-in-prediction…. [H]igh-fidelity models cannot…be made robust to uncertainty and lack-of-knowledge. Similarly, equally robust models do not provide consistent predictions, hence reducing confidence-in-prediction. The conclusion of the theoretical investigation is that, in assessing the predictive accuracy of numerical models, one should never focus on a single aspect.” Different types of predictions have different limitations. But sometimes it seems that ecologists want to make predictions in the purest, trade-off free sense - robustness-to-uncertainty, fidelity-to-data, and confidence-in-prediction - all at once. 

In relation to this, ecological processes tend to be easier to represent in a probabilistic fashion, something that we seem rather uncomfortable with. Ecology is predictive in the way medicine is predictive – we understand the important cause and effect relationships, many of the interactions that can occur, and we can even estimate the likelihood of particular outcomes (of smoking causing lung cancer, of warming climate decreasing diversity), but predicting how a human body or ecosystem will change is always inexact. The complexity of multiple independent species, populations, genes, traits, all interacting with similarly changing abiotic conditions makes precise quantitative predictions at small scales of space or time pretty intractable. So maybe that shouldn’t be our bar for success. The analogous problem for an evolutionary biologist would be to predict not only a change in population genetic structure but also the resulting phenotypes, accounting for epigenetics and plasticity too. I think that would be considered unrealistic, so why is that where we place the bar for ecology? 

In part the bar for prediction is set so high because the demand for ecological knowledge, given habitat destruction, climate change, extinction, and a myriad of other changes, is so great. But in attempting to fulfill that need, it may be worth acknowledging that predictions in ecology occur on a hierarchy from those relationships at the broadest scale that we can be most certain about, moving down to the finest scale of interactions and traits and genes where we may be less certain. If we see events as occurring with different probabilities, and our knowledge of those probability distributions declining the farther down that hierarchy we travel, then our predictive ability will decline as well. New and additional research adds to the missing or poor relationships, but at the finest scales, prediction may always be limited.

The lifetime of a species: how parasitism is changing Darwin's finches

Sonia Kleindorfer, Jody A. O’Connor, Rachael Y. Dudaniec, Steven A. Myers,Jeremy Robertson, and Frank J. Sulloway. (2014). Species Collapse via Hybridization in Darwin’s Tree Finches. The American Naturalist, Vol. 183, No. 3, pp. 325-341

Small Galapagos tree finch,
Camarhynchus parvulus

Darwin’s finches are some of the best-known examples of how ecological conditions can cause character displacement and even lead to speciation. Continuing research on the Galapagos finches has provided the exceptional opportunity to follow post-speciation communities and explore how changes in ecological processes affect species and the boundaries between them. Separate finch species have been maintained in sympatry on the islands because various barriers maintain the species' integrity, preventing hybrids from occurring (e.g. species' behavioural differences) or being recruited (e.g. low fitness). As conditions change though, hybrids may be a source of increased genetic variance and novel evolutionary trajectories and selection against them may weaken. Though speciation is interesting in its own right, it is not the end of the story: ecological and evolutionary pressures continue and species continue to be lost or added, to adapt, or to lose integrity.

A fascinating paper by Kleindorfer et al. (2014) explores exactly this issue among the small, medium, and large tree finches (Camarhynchus spp.) of Floreana Island, Galapagos. Large and small tree finches first colonized Floreana, with the medium tree finch speciating on the island from a morph of the large tree finch. This resulted in three sympatric finch species that differ in body and beak size, but otherwise share very similar behaviour and appearance. However, ecological and environmental conditions have not remained constant on Floreana since observations in the 1800s: a parasite first observed on the island in 1997, Philornis downsi has taken residence and has caused massive nestling mortality (up to 98%) for the tree finches. Since parasite density is correlated with tree finch body size, the authors predicted that high parasite intensity should be linked to declining recruitment of the large tree finch. If females increasingly prefer smaller mates, there may also be increased hybridization, particularly if there is some advantage in having mixed parental ancestry. To test this, the authors sampled tree finch populations on Floreana in both 2005 and 2010. Parasite numbers increase with high precipitation, and so by combining museum records (collected between 1852-1906, when no parasites were present), 2005 sampling records (dry conditions, lower parasite numbers), and 2010 sampling records (high rainfall, high parasite numbers), they could examine a gradient of parasite effects. They measured a number of morphological variables, collected blood for genotyping, estimated individual age, measured parasite intensity in nests, and observed mate choice.

Philornis downsi:
larval stage parasitizes nestlings.
(a Google image search will provide
some more graphic illustrations)

For each time period, morphological measurements were used to cluster individuals into putative species. The museum specimens from the 1800s had 3 morphologically distinguishable populations, the true small, medium and large tree finch species usually written about. In 2005 there were still 3 distinct clusters, but the morphological overlap between them had increased. By 2010, the year with the highest parasite numbers, there were only two morphologically distinguishable populations. Which species had disappeared? Although recent studies have labelled the two populations as the “small” tree finch and “large” tree finch, the authors found that the 2010 “large” population is much smaller than the true large tree finches collected in 1852-1906, suggesting perhaps the large tree finch was no longer present. Genetic population assignment suggested that despite morphological clustering, there were actually only two distinct species on Floreana in 2005 and 2010: it appeared that the large tree finch species had gone extinct, and the boundary between the small and medium tree finch species had become porous, leading to morphologically intermediate hybrids.

The question then, is whether the extinction of the large tree finch and the collapse of the boundary between small and medium tree finches can be attributed to the parasite, and the changing selective pressures associated with it. Certainly there were clear changes in size structure (from larger birds to smaller birds) and in recruitment (from few young hybrids to many young hybrids) between the low parasite year (2005) and the high parasite year (2010). Strikingly, parasite loads in nests were much lower for hybrids and smaller-bodied populations than for the larger-bodied population (figure below). Compared to their large-bodied parents, hybrids somehow avoided parasite attack even in years with high parasite densities (2010). When parasite loads are high, hybrid offspring have a fitness advantage, as evidenced by the large number of young hybrids in 2012. The collapse of the large tree finch population is also likely a product of parasite pressures as well, as females selected smaller mates with comparatively lower parasite loads. Despite the apparent importance of the parasites in 2010, the existence of only a few older hybrid individuals, and greater morphological distance between populations seen in the 2005 survey (a low parasite period) suggests that selection for hybrids varies greatly through time. Though the persistence of the Philornis parasite on Floreana may prevent re-establishment of the large tree finch, changing parasite densities and other selective pressures may continue to cause the boundaries of the remaining finch populations to overlap and retract in the future. The story of Darwin's finches is even more interesting if we consider that it doesn't stop at character displacement but continues to this day.

From Kleindorfer et al 2014: Philornis parasite intensity in nests sampled in 2005 (lower parasite) and 2010/2012 (higher parasite), for nests of the small-bodied (population 1), intermediate hybrid, and larger-bodied (population 2) individuals.

EEB & Flow inclusion in Library of Congress Web Archives

I just received this email the other day, and nearly deleted it as another spam email (along with fake conference invites, obscure journal submission invites, and offers to make millions). But apparently it's legit, and the US Library of Congress has been archiving web sites for some time. They are now building a collection of science blogs (link, link), which is a pretty cool idea, and we're excited to be part of it :-)

Evolution at smaller and smaller scales: a role for microgeographic adaptation in ecology?

Jonathan L. Richardson, Mark C. Urban, Daniel I. Bolnick, David K. Skelly. 2014. Microgeographic adaptation and the spatial scale of evolution. Trends in Ecology & Evolution, 19 February 2014.

Among other trends in ecology, it seems that there is a strong trend towards re-integration of ecological and evolutionary dynamics, and also in partitioning ecological dynamics to finer and finer scales (e.g. intraspecific variation). So it was great to see a new TREE article on “Microgeographic adaptation and the spatial scale of evolution”, which seemed to promise to contribute to both topics.

In this paper, Richardson et al. attempt to define and quantify the importance of small-scale adaptive differences that can arise between even neighbouring populations. These are given the name “microgeographic adaptation”, and defined as arising via trait differences across fine spatial scales, which lead to fitness advantages in an individual’s home sites. The obvious question is what spatial scale does 'microgeographic' refer to, and the authors define it very precisely as “the dispersal neighborhood … of the individuals located within a radius extending two standard deviations from the mean of the dispersal kernel of a species”. (More generally they forward an argument for a unit--the ‘wright’--that would measure adaptive divergence through space relative to dispersal neighbourhoods.) The concept of microgeographic adaptation feels like it is putting a pretty fine point on already existing ideas about local adaptation, and the authors acknowledge that it is a special case of adaptation at scales where gene flow is usually assumed to be high. Though they also suggest that microgeographic adaptation has received almost no recognition, it is probably fairer to say that in practice the assumption is that on fine scales, gene flow is large enough to swamp out local selective differences, but many ecologists could name examples of trait differences between populations at close proximity.

From Richardson et al. (2014). One
example of microgeographic adaptations.

Indeed, despite the general disregard to fine-scale evolutionary differences, they note that there are some historical and more recent examples of microgeographic variation. For example, Robert Selander found that despite the lack of physical barriers to movement, mice in neighbouring barns show allelic differences, probably due to territorial behaviour. As you might expect, microgeographic adaptations result when migration is effectively lower than expected given geographic distance and/or selection is stronger (as when neighbouring locations are very dissimilar). A variety of mechanisms are proposed, including the usual suspects – strong natural selection, landscape barriers, habitat selection, etc.

A list of the possible mechanisms leading to microgeographic adaptation is rather less interesting than questions about how to quantify the importance and commonness of microgeographic adaptation, and especially about its implications for ecological processes. At the moment, there are just a few examples and fewer still studies of the implications, making it difficult to say much. Because of either the lack of existing data and studies or else the paper's attempt to be relevant to both evolutionary biologists and ecologists, the vague discussion of microgeographic differences as a source of genetic variation for restoration or response to climate change, and mention of the existing—but primarily theoretical—ecological literature feels limited and unsatisfying. The optimistic view is that this paper might stimulate a greater focus on (fine) spatial scale in evolutionary biology, bringing evolution and ecology closer in terms of shared focus on spatial scale. For me though, the most interesting questions about focusing on smaller and smaller scales (spatial, unit of diversity (intraspecific, etc)) are always about what they can contribute to our understanding. Does complexity at small scales simply disappear as we aggregate to larger and larger scales (a la macroecology) or does it support greater complexity as we scale up, and so merit our attention? 

P-values, the statistic that we love to hate

P-values are an integral part of most scientific analyses, papers, and journals, and yet they come with a hefty list of concerns and criticisms from frequentists and Bayesians alike. An editorial in Nature (by Regina Nuzzo) last week provides a good reminder of some of the more concerning issues with the p-value. In particular, she explores how the obsession with "significance" creates issues with reproducibility and significant but biologically meaningless results.

Ronald Fischer, inventor of the p-value, never intended it to be used as a definitive test of “importance” (however you interpret that word). Instead, it was an informal barometer of whether a test hypothesis was worthy of continued interest and testing. Today though, p-values are often used as the final word on whether a relationship is meaningful or important, on whether the the test or experimental hypothesis has any merit, even on whether the data is publishable. For example in ecology, significance values from a regression or species distribution model are often presented as the results. 

This small but troubling shift away from the original purpose for p-values is tied to concerns about false alarms and with replicability of results. One recent suggestion for increasing replicability is to make p-values more stringent - to require that they be less that 0.005. But the point the author makes is that although p-values are typically interpreted as “the probability of obtaining a test statistic at least as extreme as the one that was actually observed, assuming that the null hypothesis is true”, this doesn't actually mean that a p-value of 0.01 in one study is exactly consistent with a p-value of 0.01 found in another study. P-values are not consistent or comparable across studies because the likelihood that there was a real (experimental) effect to start with alters the likelihood that a low p-value is just a false alarm (figure). The more unlikely the test hypothesis, the more likely a p-value of 0.05 is a false alarm. Data mining in particular will be (unwittingly) sensitive to this kind of problem. Of course one is unlikely to know what the odds of the test hypothesis are, especially a priori, making it even more difficult to correctly think about and use p-values. 

from: http://www.nature.com/news/scientific-method-statistical-errors-1.14700#/b5

The other oft-repeated criticism of p-values is that a highly significant p-value make still be associated with a tiny (and thus possibly meaningless) effect size. The obsession with p-values is particularly strange then, given that the question "how large is the effect?", should be more important than just answering “is it significant?". Ignoring effect sizes leads to a trend of studies showing highly significant results, with arguably meaningless effect sizes. This creates the odd situation that publishing well requires high profile, novel, and strong results – but one of the major tools for identifying these results is flawed. The editorial lists a few suggestions for moving away from the p-value – including to have journals require effect sizes and confidence intervals be included in published papers, to require statements to the effect of “We report how we determined our sample size, all data exclusions (if any), all manipulations and all measures in the study”, in order to limit data-mining, or of course to move to a Bayesian framework, where p-values are near heresy. The best advice though, is quoted from statistician Steven Goodman: “The numbers are where the scientific discussion should start, not end.”