Erin L. Koen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Erin L. Koen has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Jeff Bowman
Trent University
4 shared publications - Colin Garroway
Department of Biological Sciences University of Manitoba Winnipeg Manitoba Canada
3 shared publications Paul J. Wilson
Department of Biology, Trent University, Peterborough, Ontario, Canada
3 shared publicationsAaron A. Walpole
Wildlife Research & Development Section, Ontario Ministry of Natural Resources, 2140 East Bank Drive, Peterborough, Ontario K9J 7B8, Canada
1 shared publicationStephen C. Mills
Northeast Science and Information Section, Ontario Ministry of Natural Resources
1 shared publication
Latest Publications
- Open Access
- 8 February 2013
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 8 (2) , e56204
The Sensitivity of Genetic Connectivity Measures to Unsampled and Under-Sampled Sites
Landscape genetic analyses assess the influence of landscape structure on genetic differentiation. It is rarely possible to collect genetic samples from all individuals on the landscape and thus it is important to assess the sensitivity of landscape genetic analyses to the effects of unsampled and under-sampled sites. Network-based measures of genetic distance, such as conditional genetic distance (cGD), might be particularly sensitive to sampling intensity because pairwise estimates are relative to the entire network. We addressed this question by subsampling microsatellite data from two empirical datasets. We found that pairwise estimates of cGD were sensitive to both unsampled and under-sampled sites, and FST, Dest, and deucl were more sensitive to under-sampled than unsampled sites. We found that the rank order of cGD was also sensitive to unsampled and under-sampled sites, but not enough to affect the outcome of Mantel tests for isolation by distance. We simulated isolation by resistance and found that although cGD estimates were sensitive to unsampled sites, by increasing the number of sites sampled the accuracy of conclusions drawn from landscape genetic analyses increased, a feature that is not possible with pairwise estimates of genetic differentiation such as FST, Dest, and deucl. We suggest that users of cGD assess the sensitivity of this measure by subsampling within their own network and use caution when making extrapolations beyond their sampled network.
- 8 February 2013
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 8 (2) , e56204
Open AccessThe Sensitivity of Genetic Connectivity Measures to Unsampled and Under-Sampled Sites
Landscape genetic analyses assess the influence of landscape structure on genetic differentiation. It is rarely possible to collect genetic samples from all individuals on the landscape and thus it is important to assess the sensitivity of landscape genetic analyses to the effects of unsampled and under-sampled sites. Network-based measures of genetic distance, such as conditional genetic distance (cGD), might be particularly sensitive to sampling intensity because pairwise estimates are relative to the entire network. We addressed this question by subsampling microsatellite data from two empirical datasets. We found that pairwise estimates of cGD were sensitive to both unsampled and under-sampled sites, and FST, Dest, and deucl were more sensitive to under-sampled than unsampled sites. We found that the rank order of cGD was also sensitive to unsampled and under-sampled sites, but not enough to affect the outcome of Mantel tests for isolation by distance. We simulated isolation by resistance and found that although cGD estimates were sensitive to unsampled sites, by increasing the number of sites sampled the accuracy of conclusions drawn from landscape genetic analyses increased, a feature that is not possible with pairwise estimates of genetic differentiation such as FST, Dest, and deucl. We suggest that users of cGD assess the sensitivity of this measure by subsampling within their own network and use caution when making extrapolations beyond their sampled network.
- 22 February 2012
- Published by Wiley in Molecular Ecology Resources
- Vol. 12 (4) , 686-696
The effect of cost surface parameterization on landscape resistance estimates
A cost or resistance surface is a representation of a landscape’s permeability to animal movement or gene flow and is a tool for measuring functional connectivity in landscape ecology and genetics studies. Parameterizing cost surfaces by assigning weights to different landscape elements has been challenging however, because true costs are rarely known; thus, expert opinion is often used to derive relative weights. Assigning weights would be made easier if the sensitivity of different landscape resistance estimates to relative costs was known. We carried out a sensitivity analysis of three methods to parameterize a cost surface and two models of landscape permeability: least cost path and effective resistance. We found two qualitatively different responses to varying cost weights: linear and asymptotic changes. The most sensitive models (i.e. those leading to linear change) were accumulated least cost and effective resistance estimates on a surface coded as resistance (i.e. where high‐quality elements were held constant at a low‐value, and low‐quality elements were varied at higher values). All other cost surface scenarios led to asymptotic change. Developing a cost surface that produces a linear response of landscape resistance estimates to cost weight variation will improve the accuracy of functional connectivity estimates, especially when cost weights are selected through statistical model fitting procedures. On the other hand, for studies where cost weights are unknown and model selection is not being used, methods where resistance estimates vary asymptotically with cost weights may be more appropriate, because of their relative insensitivity to parameterization.
- 22 February 2012
- Published by Wiley in Molecular Ecology Resources
- Vol. 12 (4) , 686-696
The effect of cost surface parameterization on landscape resistance estimates
A cost or resistance surface is a representation of a landscape’s permeability to animal movement or gene flow and is a tool for measuring functional connectivity in landscape ecology and genetics studies. Parameterizing cost surfaces by assigning weights to different landscape elements has been challenging however, because true costs are rarely known; thus, expert opinion is often used to derive relative weights. Assigning weights would be made easier if the sensitivity of different landscape resistance estimates to relative costs was known. We carried out a sensitivity analysis of three methods to parameterize a cost surface and two models of landscape permeability: least cost path and effective resistance. We found two qualitatively different responses to varying cost weights: linear and asymptotic changes. The most sensitive models (i.e. those leading to linear change) were accumulated least cost and effective resistance estimates on a surface coded as resistance (i.e. where high‐quality elements were held constant at a low‐value, and low‐quality elements were varied at higher values). All other cost surface scenarios led to asymptotic change. Developing a cost surface that produces a linear response of landscape resistance estimates to cost weight variation will improve the accuracy of functional connectivity estimates, especially when cost weights are selected through statistical model fitting procedures. On the other hand, for studies where cost weights are unknown and model selection is not being used, methods where resistance estimates vary asymptotically with cost weights may be more appropriate, because of their relative insensitivity to parameterization.
- 5 November 2011
- Published by Springer Nature in Landscape Ecology
- Vol. 27 (1) , 29-43
Landscape resistance and American marten gene flow
- 5 November 2011
- Published by Springer Nature in Landscape Ecology
- Vol. 27 (1) , 29-43
Landscape resistance and American marten gene flow
- Open Access
- 26 July 2010
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 5 (7) , e11785
The Effect of Map Boundary on Estimates of Landscape Resistance to Animal Movement
Artificial boundaries on a map occur when the map extent does not cover the entire area of study; edges on the map do not exist on the ground. These artificial boundaries might bias the results of animal dispersal models by creating artificial barriers to movement for model organisms where there are no barriers for real organisms. Here, we characterize the effects of artificial boundaries on calculations of landscape resistance to movement using circuit theory. We then propose and test a solution to artificially inflated resistance values whereby we place a buffer around the artificial boundary as a substitute for the true, but unknown, habitat. We randomly assigned landscape resistance values to map cells in the buffer in proportion to their occurrence in the known map area. We used circuit theory to estimate landscape resistance to organism movement and gene flow, and compared the output across several scenarios: a habitat-quality map with artificial boundaries and no buffer, a map with a buffer composed of randomized habitat quality data, and a map with a buffer composed of the true habitat quality data. We tested the sensitivity of the randomized buffer to the possibility that the composition of the real but unknown buffer is biased toward high or low quality. We found that artificial boundaries result in an overestimate of landscape resistance. Artificial map boundaries overestimate resistance values. We recommend the use of a buffer composed of randomized habitat data as a solution to this problem. We found that resistance estimated using the randomized buffer did not differ from estimates using the real data, even when the composition of the real data was varied. Our results may be relevant to those interested in employing Circuitscape software in landscape connectivity and landscape genetics studies.
- 26 July 2010
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 5 (7) , e11785
Open AccessThe Effect of Map Boundary on Estimates of Landscape Resistance to Animal Movement
Artificial boundaries on a map occur when the map extent does not cover the entire area of study; edges on the map do not exist on the ground. These artificial boundaries might bias the results of animal dispersal models by creating artificial barriers to movement for model organisms where there are no barriers for real organisms. Here, we characterize the effects of artificial boundaries on calculations of landscape resistance to movement using circuit theory. We then propose and test a solution to artificially inflated resistance values whereby we place a buffer around the artificial boundary as a substitute for the true, but unknown, habitat. We randomly assigned landscape resistance values to map cells in the buffer in proportion to their occurrence in the known map area. We used circuit theory to estimate landscape resistance to organism movement and gene flow, and compared the output across several scenarios: a habitat-quality map with artificial boundaries and no buffer, a map with a buffer composed of randomized habitat quality data, and a map with a buffer composed of the true habitat quality data. We tested the sensitivity of the randomized buffer to the possibility that the composition of the real but unknown buffer is biased toward high or low quality. We found that artificial boundaries result in an overestimate of landscape resistance. Artificial map boundaries overestimate resistance values. We recommend the use of a buffer composed of randomized habitat data as a solution to this problem. We found that resistance estimated using the randomized buffer did not differ from estimates using the real data, even when the composition of the real data was varied. Our results may be relevant to those interested in employing Circuitscape software in landscape connectivity and landscape genetics studies.