Postdoctoral researcher position
When can we extrapolate patch size, isolation, and edge effects to infer effects of habitat fragmentation over landscapes and regions?
A two-year postdoctoral position (CDN $55,000 p.a.) is available immediately, in the Geomatics and Landscape Ecology Laboratory (GLEL) at Carleton University, Ottawa, Canada.
The PDF will work with the co-directors of the GLEL, as well as other collaborators. Major requirements for the position are a PhD with a strong background in ecology, an ability to work with large datasets, including identifying data sources and extracting and manipulating data, a solid background in statistical analysis, and strong remote sensing and GIS data processing, classification and analysis skills.
Pitfalls in extrapolation across scales (spatial extents) are well documented in ecology. For example, as first pointed out by Levine and D’Antonio (1999), diversities of native and exotic plants are often negatively correlated at a small scale, but they are usually positively correlated at larger scales. Similarly, human presence often correlates negatively with species richness at small spatial scales but positively at larger scales (Pautasso, 2007). The same is true for temporal biodiversity change: trends at small scales cannot necessarily be extrapolated to larger scales (Jarzyna and Jetz, 2018).
One situation in which cross-scale extrapolation is common in ecology is in studies of habitat fragmentation (Figure 1). Most fragmentation studies document species responses to patch size, patch isolation, or distance to patch edges, i.e. responses at small scales (Figure 1A, B, C). These responses are then often extrapolated to infer effects of fragmentation at larger landscape or regional scales. Such extrapolations may not be correct because they ignore other mechanisms, especially those acting at larger scales (e.g., spreading of risk, species interactions, landscape complementation), that can counteract effects of the documented patch-scale responses. These larger-scale mechanisms are likely the reason why landscape-scale fragmentation studies usually find either no effect or a positive effect of fragmentation (Fahrig 2017) even though patch-scale studies usually predict negative fragmentation effects.
The objective of the PDF project will be to identify the situations in which such cross-scale extrapolations are valid, and those in which they are not valid, by:
- identifying species data sets that are suitable for testing both small-scale responses to patch size, isolation, and edge, and large-scale responses to habitat fragmentation;
- identifying appropriate remotely-sensed data in the regions in which the species data sets occur, and extracting variables needed for the small- and large-scale analyses: patch size, patch isolation, distance to edge, habitat amount, habitat fragmentation;
- building statistical models relating the species data to the small- and large-scale variables and identifying the situations in which the results are congruent vs. not congruent across scales.
Figure 1. Patch-scale studies (A, B, C) document the effects of patch size, patch isolation, or distance to edge on ecological responses, by comparing these responses over multiple patches. Some authors extrapolate results of these studies to make inferences about landscape-scale habitat fragmentation effects. But these extrapolations are not evidence of landscape-scale fragmentation effects; rather they are predictions that need to be tested using landscape-scale studies (D).
Please submit your CV, a short (one paragraph) statement of research interests, and the names and contact information of two references to Lenore Fahrig (firstname.lastname@example.org) or Joe Bennett (email@example.com) before 15 November 2018.
Graduate Student positions:
Although most positions are filled for the next little while, informal inquiries are always welcome. Please feel free to get in touch via email: firstname.lastname@example.org.