Addressing context dependence in ecology

A phrase that you are bound to hear many times at any ecology conference is “it depends”. We see context dependence – variation in the magnitude or sign of ecological relationships depending on the conditions under which they are observed (Fig. 1) – in just about every study and every system. Such variation, especially when unexplained, can lead to spurious or seemingly contradictory conclusions across studies, which can limit understanding and our ability to transfer findings across studies, space, and time. Because of the wide prevalence of observed context dependence and the critical need to tackle it, a group of us recently knocked heads (and read lots of fabulous papers!) about how it can be addressed. Our reading, thinking, talking, drawing and writing culminated in this open access paper in TREE

Figure 1: Context dependence may be invoked when the observed relationship between two variables varies in (a) magnitude (strength), (b) sign (direction), and (c) uncertainty, applied here to hypothetical examples from plant invasions.

In the paper, we identify two types of context dependence resulting from four sources (Fig. 2). Mechanistic context dependence occurs when a relationship, say between variables X and Y, fundamentally differs under different ecological and spatiotemporal conditions. Such relationships arise from (i) interaction effects of another variable, Z, which modifies the effect of X on Y, reflecting ecological processes. Apparent context dependence occurs when the relationship between variables X and Y does not differ but appears to due to: (ii) the presence of confounding factors that are either unaccounted for or are measured and accounted for in some studies but not others; (iii) problems of statistical inference where studies differ in sampling accuracy and precision, statistical power, or interpretation of statistical measures; and (iv) methodological differences among studies whereby studies observe and measure variables or relationships in different ways.

Figure 2: Four sources of variation in the relationship between independent variable X and dependent variable Y, with illustrative examples and actions that can reduce unexplained variation and the likelihood of apparent context dependence.

We illustrate our typology using examples from biological invasions, a field where context dependence is prominent and widely discussed, but we propose that the typology is applicable across all areas of ecology (and it may well extend to all natural and biological sciences…). We conclude the paper by outlining steps for addressing the different types and sources of context dependence, and provide a decision tree that outlines key actions likely to be helpful. We believe that by recognising the different ways in which context dependence can arise, we can better account for context dependence and reduce the prevalence of unexplained variation in ecology.

Full paper: Catford, J.A., Wilson, J.R.U., Pyšek, P., Hulme, P.E. & Duncan, R.P. (in press) Addressing context dependence in ecology. Trends in Ecology & Evolutionlink (open access)

Sunshine and visitors

Esti, Luis and Alan make the most of the sunshine at New Forest NP

Long days, warm weather, dry roads and visitors. I love summer!

Esti Palma from the University of Melbourne and Dr Luis Mata from RMIT University (and, the all-important, Alan Mata) came to visit for a couple of weeks recently. They both gave excellent talks to our ecology group in Biological Sciences.

Luis spoke about “The Little Things That Run the City”, drawing on the pioneering work in Melbourne where Luis and other folk from RMIT are joining forces with Melbourne City Council to bring biodiversity into the city – and to make people aware of, and value, it. They’ve produced a beautiful children’s book as part of this work. I’m looking forward to seeing the other things that this innovative and productive group produce.

Esti focused on her invasive species traits work where she is using 80 plant species to test whether it is more informative to separate invasive species into different “types” based on their dimension of invasiveness, or whether it doesn’t matter if all species are lumped into one category. Early results seem to point to the former – but watch this space!

Esti also spoke briefly about her work that shows trait-based trends in the types of species that are being both lost from and gained in cities. As a bit of extra excitement, Esti’s paper featured on the cover of Ecography.  

john dwyer and bjorn robroek.jpeg
John and fellow Southamptonite, Bjorn Robroek, enjoying a cleansing ale after a walk along the River Itchen (we did talk ecology, promise!).

While talking about sunny visitors, it would be remiss of me not to mention Dr John Dwyer from the University of Queensland who stopped to say hello in June.

John also educated us with a talk, this time on how trait covariance can help us understand biodiversity trends along environmental gradients. Some really lovely work by John and Daniel Laughlin.


Flow regulation and drought drive riparian plant invasion

Ecologists, like epidemiologists, are often confronted with the challenge of trying to determine causality by piecing together bits of information observed in nature. When the presence or absence of a species at a site is affected by the characteristics of the environment and community, the availability and dispersal success of propagules, stochastic events and the peculiarities of the species itself, it can be very difficult to isolate the likely mechanisms that lead to the occurrence – or lack thereof – of a particular species, especially when the influential factors are highly correlated.

The joys of being in control.  (Source:
The joys of being in control.

Experiments are obviously made for getting around such problems; by controlling and isolating one factor at a time, the relative importance of different factors can be quantified. However, experiments are not always possible, desirable or ethical. Take plant invasions along rivers, for example: they occur at large spatial and temporal scales; many factors may drive the invasion process; introducing and augmenting the supply of invasive species is unpalatable and likely prohibited; plus, river environments are very hard to control and manipulate, as any manager will tell you. So, if we are limited to potentially confounded survey data, how can we more effectively identify the drivers of plant invasion so that we know which factors to target in weed management?

In a paper recently published in Diversity and Distributions, my colleagues and I contend that incorporating data about species characteristics into survey-based approaches provides an additional line of evidence that can be used to improve inferences drawn from patterns. We illustrate how using information about environmental gradients, species distributions and species characteristics can increase understanding of ecological phenomena – here, riparian plant invasion, which can help inform management responses.

Using this approach, we find that, of four hypotheses examined, hydrological modification (indicated by flood magnitude) most likely drives invasion in River Murray wetlands. Flow regulation may inhibit native species adapted to the historical hydrological regime, facilitating exotic species with different environmental ranges. A symptom of environmental change, invasion may have been exacerbated by drought, although it is unclear why.

By hitching a ride on walkers' shoes and boots, exotic plants may even invade places like this.  Columbia River Valley, Oregon. (Source: JAC).
By hitching a ride on walkers’ boots, exotic plants may even invade places like this.
Columbia River Valley, Oregon. (Source: JAC).

There was no indication that human-increased propagule pressure or colonisation ability facilitated invasion. Exotic cover was unrelated to proximity to towns, recent flood frequency and cattle grazing intensity. Additionally, similar proportions of exotic and native species were used in cultivation and, despite a higher proportion of exotics being known weeds, weed status was unrelated to exotic species occupancy. Overall, colonisation ability was unrelated to species’ origin or response to water depth and hydrological change. Although exotics had higher specific leaf area and shorter longevity (indicative of higher colonisation ability), they had heavier (not lighter) seeds and did not differ in height from natives.

Based on our findings, we conclude that (i) using environmental flows to reinstate mid-range floods and (ii) augmenting the propagule supply of native species with characteristics suitable for modified conditions may help limit invasion in these wetlands.

For more, have a look here or drop me a line and I’ll send you a copy. I’d be delighted to hear any thoughts, comments or queries that you may have.

Land of 10,000 Lakes

Located on the Canadian border, smack bang in the middle of the US, Minnesota is indeed a land of many lakes. While this  postcard does sell them a bit short (there are in fact about 15,000 lakes and that doesn’t include those with more pond-like dimensions), it communicates the basic facts: there are a LOT of water bodies in this part of the world. And for someone who is partial to water, that is a pretty cool thing.

As part of my DECRA research, I am spending a couple of months in Minnesota this year.  I am planning to conduct an invasion-style experiment at Cedar Creek Ecosystem Science Reserve in 2013 and 2014, so this is the essential reconnaissance trip.

So, what have I discovered so far?

1). My University of Minnesota mentor (Dave Tilman), his lab and the Dept of Ecology, Evolution and Behavior (EEB) are all tops. Widely recognized as a place of (i.e. people that do) cutting edge ecological research, it is also proving to be a really fun and dynamic place to work.

Me obscuring a beautiful Bur Oak (Quercus macrocarpa) in sand savanna. Sand savanna now only makes up 0.00005 of its former range, so this patch of habitat is pretty precious.

To illustrate: within 10 days of arriving, I have had dinner at Dave’s house with his lab group and Simon Levin (another super high-echelon ecologist visiting from Princeton), gone to an EEB summer barbeque, spent a day out at Cedar Creek and to top it off, I’ve just returned from a “lab get-away” at the Tilman family cabin that is situated on a lake (of course!) ~3 hours north of Minneapolis. A pretty good start, I would say. 

2). A staggering amount of high quality research has taken place at Cedar Creek. Owned and operated by the University of Minnesota in cooperation with the Minnesota Academy of Science, Creek Creek is a large ecological research station that is part of the Long-Term Ecological Research Network that exists in the US. Established in 1940, Cedar Creek is located in central Minnesota and contains natural habitats that represent the entire state. To cite Nee and Lawton’s (1996) Nature paper, Cedar Creek “is rapidly becoming one of ecology’s classic localities” (p. 672). Well, not wanting to assume that I am qualified to make the call, but I would say that Cedar Creek is one of ecology’s classic localities. If you’re in any doubt, then check out the publications that have been generated from work at Cedar Creek (be prepared to scroll!).

Dave Tilman chatting to some interns who are weeding a plot of the big Biodiversity Experiment at Cedar Creek

Like all things that work well, the strength of Cedar Creek comes down to its workforce. 24 Faculty members work at Cedar Creek, as well as 15 support staff, 10s of postdocs and postgrads and even more summer interns. This summer there are about 60 people working at Cedar Creek and – in two days time – I’ll be one of them. Yippee!


3). As an upshot of (1) and (2), some incredibly exciting opportunities are available for my research. I can take a fresh look at past experiments and surveys, tweak experiments that are already up and running and then there is the option of setting up an experiment myself. The world is most definitely my oyster – I can hardly believe my luck!

While being completely spoilt for choice is a wonderful situation to find oneself in, it does present the inevitable challenge of narrowing things down. So, in the next little while I plan to do some general reading, some Cedar Creek-themed reading, some Cedar Creek fieldwork and then a generous dose of coffee-fuelled and bike-inspired thinking. At the end of that, and the end of my two months here, I hope that some sparks will have flown, some light bulbs lit and some cool ideas arisen. Exciting times!

And now, off for a bike ride and a swim… You probably have an inkling where those activities will take place!

My ride home from work: Lake Calhourn with downtown Minneapolis in the background

New paper: The intermediate disturbance hypothesis and plant invasions: implications for species richness and management

Jan 2012 – Some colleagues and I have recently written a paper that examines the relationship between the intermediate disturbance hypothesis (IDH) and alien plant invasions.  Published in Perspectives in Plant Ecology, Evolution and Systematics, the paper is structured around two questions: in accordance with IDH, 1) at what disturbance frequencies is alien plant colonisation most likely and why, and 2) where along the disturbance continuum (at which successional stage) are alien plants likely to reduce community diversity and why?  We use understanding of community and invasion ecology to answer these questions, drawing on empirical evidence from a variety of terrestrial ecosystems.  We conclude the paper by discussing implications and strategies for managing plant communities and how patterns of invasion might change in the future.

You can find a summary of the paper on our lab website.

If you’d like a copy, please shoot me an email.

%d bloggers like this: