Research Objectives
Research Objectives
Embedded Files
The overarching objective of this research is to improve assisted migration prescription accuracy and applicability.
The first objective of this project to reach this goal is to create improved statistical species maps; first by validating and finding errors in current species distribution data.
Creating statistical species frequency maps
Creating statistical species frequency maps
The above species distribution maps have been created using plot data from the US, Alberta, and British Columbia. The maps highlight some of the inconsistencies at the border between the U.S. and Canada. Because the plot data comes from different sources with different measurement criteria these inconsistencies can be expected. However, creating maps and formulating data that is as accurate as possible is essential for relaying seed source recommendations.
The goal for this research project is to validate current species frequency data.
Eventually, the goal is to have recreated these maps using predictive modelling software in conjunction with plot data to create seamless species frequencies maps to update potentially outdated range maps, and create a cohesive and reliable software.
Determining Error in Ecosystem Delineations
Determining Error in Ecosystem Delineations
To give the most accurate recommendations on seed source locations, and to make accurate species frequency maps, having accurate ecosystem delineations is necessary. For this project the aim is to determine 'problem areas' where error is high and ecosystems may need to be redrawn to better reflect species frequencies information.
This portion of the project will inform redrawn ecosystem delineations which rely on climate, topographic, and species data to produce ecosystem delineations that vary across ecosystems as little as possible. By collecting and analyzing climate data across the U.S. and Canada the team has identified areas in the most need of effort towards recreating ecosystem boundaries.
Fig 17. MAE of species frequency per ecosystem by plot data set.
Canadian datasets show higher MAE
Canadian datasets show higher MAE
The large size of the U.S. dataset compared with the Canadian ones may explain in part the lower MAE. However, the high MAE values from Alberta and northern British Columbia are noticeably higher than those of the northwestern U.S.
The large size of ecosystem delineations in Alberta and northern British Columbia (Fig 19) display that large ecosystem delineations are less likely to accurately reflect species composition and conditions, and thus tend to have high errors.
Discontinuities can be observed at the borders between any of the 3 data sets, another factor highlighting the need to produce a uniform dataset with reliable data between political borders.
Fig 16. MAE values per ecosystem, grouped into large biome-level ecosystems.
Boreal Forest Biome shows highest MAE
Boreal Forest Biome shows highest MAE
Figure 16 displaying MAE by biome-level ecosystem delineation shows that boreal, taiga, and the northwest interior have the highest median MAE values.
In the datasets used for this study, the boreal zone was largely represented by ecosystems delineations within northern Alberta, with rather coarse ecosystem delineations. This has the potential to lead to high MAE and further highlights the need to produce more accurate and representative ecosystem delineations to better reflect species composition.
Page updated
Google Sites
Report abuse