Methods Beyond the Population

Post Provided by SEAN MCMAHON and JESSICA METCALF

Demography Beyond the Population” is a unique Special Feature being published across the journals of the British Ecological Society.  The effort evolved from a symposium of the same name hosted in Sheffield, UK last March. Both the meeting and the Special Feature were designed to challenge ecologists from a range of fields whose research focuses on populations.

The participants were charged with sharing how they are pushing the work they do beyond the stage where the population is the focus into research where the population is just the beginning and the focus spans scales, systems and tools. This encompasses a broad suite of biological research, including range modelling, disease impacts on communities, biogeochemistry, evolutionary theory, and conservation biology. The meeting was a great success, and this Special Feature should be equally valuable to the broad readership of the BES journals.

Methods in Ecology and Evolution has a special place in the Special Feature, hosting four papers. These papers not only introduce new efforts in population biology, they provide the methods that other scientists can use to implement them. With the tools provided by these four papers, researchers will be able to advance forest modelling, evolutionary theory, climate change biology and statistical inference of hidden population parameters.  Seriously good stuff!

The Scale of Climate Change

The impacts of climate change on terrestrial communities are difficult to quantify

The impacts of climate change on terrestrial communities are difficult to quantify. © High Contrast

How will communities and biomes change due to climate change? Which features of climate change are likely to influence populations locally, at the landscape scale, or across species ranges? These questions are central to inferring climate change impacts on terrestrial communities, but are difficult to quantify. It is generally easier to bound problems a priori for climate problems — choose a scale and move forward.

Brittany Teller and colleagues provide a spline method that can help quantify decisions about variable selection and scales of importance in studying population impacts of climate. Their approach opens the way to identifying important climate drivers of demographic rates but also the impact of competition. These are both features known to affect source-sink dynamics, but they’re rarely included together in population-climate analyses.

Forgetfulness: No Longer a Problem in Population Biology

How many baby butterflies do you need to take over the world? ©Lokal Profil

How many baby butterflies do you need to take over the world? ©Lokal Profil

How do you recover important data about a population you plumb forgot to collect? Well, forgetfulness isn’t usually the problem, but something like it comes up in many studies. Edgar González and colleagues present an excellent introduction to inverse modelling in population biology.

This work follows an important and increasingly popular approach to estimating the unmeasurable in ecology (see Hartig et al. Statistical inference for stochastic simulation models – theory and application (2011), and Forest community response to invasive pathogens: the case of ash dieback in a British woodland by Needham et al. in this feature). Using integral projection models (IPMs) González and colleagues demonstrate how population models can be used to infer (and quantify) information by simulating how the population might behave were it to have a specific value for that unmeasured quantity.

Say you measured growth and survival of individuals, but forgot to see how many babies they had. Oops. You know they live and die by rules derived in big fun complex statistical models! But critters are zeroes. Well do you think the population is stable? Let’s say it is.

Propose a number of critters each adult might have. Now simulate the population using an IPM and see if what happens to the population makes even a bit of sense. Do they take over the world? Do they disappear in an instant? Keep the estimates that make sense.  Throw out the ones that don’t. But no matter what, keep proposing new numbers of critters, keep running your IPMs, and keep filtering out the bad ideas from the good. Repeat until you get values that make the population simulation behave reasonably. Ta Da!

Forgot to count seeds but want to know how many annual seeds produced would keep your butterfly garden happy? Just read Edgar’s article (and spend a lot of time measuring growth and survival in your butterfly garden).

Which Trait Did What?

An earlobe (not Mark’s). © Maksim

An earlobe (not Mark’s). © Maksim

We all have lots of traits. I do. You do. Mark Rees and Steve Ellner do. Reproducers among us can give some of those traits to their offspring. Are those traits beneficial to these new critters (babies, butterflies…)? You can measure the trait, and check fitness using a population model.  Fine. New generation, traits are tested, some advance. Trait by fitness found.

But this classic population model misses the fact that traits don’t just matter in a population, they matter in a part of a population. Rees and Ellner, in their contribution, Evolving Integral Projection Models: Evolutionary Demography meets Eco-Evolutionary Dynamics, show that knowing the structure of the population (for example, the size or age of individuals with a trait) is essential to understanding whether or not that trait really and truly confers fitness. Or does it just confer random earlobes? Mark does have nice earlobes, must be said…

Simulating a Kingdom

…or explore a range of ecological and evolutionary questions in forest biology

…or explore a range of ecological and evolutionary questions in forest biology

Do you want to simulate a tree population? How about a forest community? A meta-community? What about ALL plants?! Dan Falster and colleagues give you the R-package plant (seriously, no one takes the names that don’t have “i” or “e” attached to the organism of choice – well done here, Dan). plant may not actually simulate Plantae, but it can explore a range of ecological and evolutionary questions in forest biology.

By incorporating demography and trait values in an efficient algorithm that follows individuals through time, life-history strategies can be made silicate and ‘grown’ (see figure below). The fact that these populations can be given trait values (or strategies), allows many scales of inference to be simulated. Again, this is a package. Download and play. There are a lot of data out there, many of which are in the public domain, that could be used in plant. So come up with an idea, get the data, and get plant.

A forest in flux. Each line is a cohort of trees whose size is on the y-axis and whose number is the thickness of the line. Pioneer species are orange and late-successional species are steel-blue. From an open field, the pioneers fill the space and then the late successional individuals come in to play. Gaps keep things interesting so that both species can co-exist. Cool forest. Thank you, plant! ©Daniel Falster

A forest in flux. Each line is a cohort of trees whose size is on the y-axis and whose number is the thickness of the line. Pioneer species are orange and late-successional species are steel-blue. From an open field, the pioneers fill the space and then the late successional individuals come in to play. Gaps keep things interesting so that both species can co-exist. Cool forest. Thank you, plant! ©Daniel Falster

In addition to these wonderful methods pieces, the Special Feature offers many further interesting advances in population research. In all, there are 21 contributions that can be found in the Journal of Ecology, the Journal of Animal Ecology, Functional Ecology, the Journal of Applied Ecology and Ecology and Evolution. This is a unique approach to producing a special feature, but a great example of how important and broad demographic research has become in ecology. For an overview of the Special Feature, enjoy the editorial by Griffith et al. in the Journal of Ecology.

Advertisements

2 thoughts on “Methods Beyond the Population

  1. Pingback: Inverse Modelling and IPMs: Estimating Processes from Incomplete Information | methods.blog

  2. Pingback: Issue 7.2: Demography Beyond the Population | methods.blog

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s