| Climate
Change and Garry Oaks |
 Figure: Pollen sample sites in the BCMPD. This map shows only a limited number of sites. A full site list is available in our paper, and a Google Map representation exists here. |
British Columbia
Modern Pollen Database
We have assembled a database of over 250 modern pollen samples from British Columbia collected from the literature. Of great assistance in this task is the database compiled by Whitmore et al (2005) [here]. From this starting point I have been able to add and revise the dataset to increase its predictive power. Using a combination of simple distance metrics and more complex ordination and clustering methods (code here) we have developed a scalable technique for determining the ecological setting of "unknown" pollen spectra. This method is published as the NMDS/GAM method in Goring et al. (2009) |
|
|
|
| Climate Change and Suitability With Marlow Pellatt and others I have been using randomForest models to predict site suitability for Garry oak in the Pacific Northwest. Our results suggest interesting patterns of change in the future. While many sites may be suitable in the long term, short term changes in precipitation and temperature may render many current sites unsuitable for regeneration. This work is on-going and results will be presented at the Conservation Biology conference in Edmonton, July 2010. |
 Figure: Garry oak site suitability for the
Pacific Northwest for climate change scenarios in 2010, 2040 and
2070. Models were constructed using downweighted PRISM models for
a number of climate parameters. Suitability was determined using
randomForest models, implemented in R based on actual presence-absence
data. |
|
|
|
 Figure: Lines represent zones with significant differences in tree shape (measured using Iw) based on a phylogeny constructed using pollen taxa from modern pollen assemblages. |
Current intrest in the stucture of community phylogenies led me to examine whether we can obtain information on short-term changes in community structure through BC pollen records. Community phylogenies allow researchers to make inferences about the shared evolutionary history of species within a community. British Columbia has a geological heritage that includes long periods of glaciation followed by rapid colonization, volcanism, mountain building and collapse, all within the past 50MYa. In particular we have found that pollen-inferred phylogenies appear to show patterns of difference between ecological communities at a provincial scale. In addition, we detect patterns of change in pollen-inferred phylogeny shape over time. Preliminary results were presented in a poster at the Canadian Society for Ecology and Evolution conference in Vancouver BC. |
|
|
|
| Tree Shape and Mammal
Extinctions
As part of my work at Simon Fraser University I have been fortunate to work with Tyler Kuhn and Arne Mooers, examining the effect of Holocene extinctions of mammals on phylogeny shape. In a recently published book chapter we show that extinctions through the Holocene have resulted in significantly greater imbalance in the mammal phylogeny, compared to the null expectation of random loss. This fascinating result depends on the effects of non-random extinction events throughout the Holocene. It is clear that Holocene extinctions, especially anthropogenic-related extinctions targeted specific groups of species, for example, small taxa associated with isolated islands (cf. Manne et al., 1999). During the course of this work I also wrote some code to calculate Iw, a measure of imbalance first proposed by Fusco and Cronk (1995), and later modified by Purvis et al. The code is available as a text file for use with R here. |
Figure: An extinct mammoth raises its mighty head. Credit: Soul of Beer |
|
|
|