Do landscape processes predict phylogeographic patterns in the wood frog?
Article first published online: 19 MAR 2009
© 2009 Blackwell Publishing Ltd
Volume 18, Issue 9, pages 1863–1874, May 2009
How to Cite
LEE-YAW, J. A., DAVIDSON, A., MCRAE, B. H. and GREEN, D. M. (2009), Do landscape processes predict phylogeographic patterns in the wood frog?. Molecular Ecology, 18: 1863–1874. doi: 10.1111/j.1365-294X.2009.04152.x
- Issue published online: 17 APR 2009
- Article first published online: 19 MAR 2009
- Received 14 October 2008; revision received 4 February 2009; accepted 7 February 2009
- gene flow;
- genetic structure;
- isolation by distance;
- isolation by resistance;
- landscape genetics;
- spatial scale
Understanding factors that influence population connectivity and the spatial distribution of genetic variation is a major goal in molecular ecology. Improvements in the availability of high-resolution geographic data have made it increasingly possible to quantify the effects of landscape features on dispersal and genetic structure. However, most studies examining such landscape effects have been conducted at very fine (e.g. landscape genetics) or broad (e.g. phylogeography) spatial scales. Thus, the extent to which processes operating at fine spatial scales are linked to patterns at larger scales remains unclear. Here, we test whether factors impacting wood frog dispersal at fine spatial scales are correlated with genetic structure at regional scales. Using recently developed methods borrowed from electrical circuit theory, we generated landscape resistance matrices among wood frog populations in eastern North America based on slope, a wetness index, land cover and absolute barriers to wood frog dispersal. We then determined whether these matrices are correlated with genetic structure based on six microsatellite markers and whether such correlations outperform a landscape-free model of isolation by resistance. We observed significant genetic structure at regional spatial scales. However, topography and landscape variables associated with the intervening habitat between sites provide little explanation for patterns of genetic structure. Instead, absolute dispersal barriers appear to be the best predictor of regional genetic structure in this species. Our results suggest that landscape variables that influence dispersal, microhabitat selection and population structure at fine spatial scales do not necessarily explain patterns of genetic structure at broader scales.