A framework for assessing and projecting climate change effects on forest communities across the U.S.
•Download as PPTX, PDF•
1 like•248 views
Presented at US-IALE annual meeting in Baltimore, MD. We are using hierarchical classification to produce an empirical set of forest tree assemblages for use in projection, assessment, and monitoring of global change effects on forest communities.
Recommended
Recommended
More Related Content
What's hot
What's hot (20)
Similar to A framework for assessing and projecting climate change effects on forest communities across the U.S.
Similar to A framework for assessing and projecting climate change effects on forest communities across the U.S. (20)
More from Jennifer Costanza
More from Jennifer Costanza (9)
Recently uploaded
Recently uploaded (20)
A framework for assessing and projecting climate change effects on forest communities across the U.S.
- 1. A framework for assessing and projecting climate change effects on forest communities across the US Jennifer Costanza John Coulston, Dave Wear April 10, 2017
- 2. Climate and land use change Ordoñez et al. 2014 Nature Clim. Ch.
- 3. Macrosystems ecology From Heffernan et al. 2014
- 4. Forest species composition is important and can be affected by global change drivers.
- 5. Establish an empirical, hierarchical, classification of forest community composition in the continental US A baseline characterization to facilitate monitoring, assessment, and projection of global change effects on forest communities
- 6. USDA Forest Service Forest Inventory and Analysis data (FIA) 130,000+ forest plots across the continental US 108 m
- 7. Hierarchical clustering of FIA plots by tree species composition • Relative importance value by species (abundance and basal area) • For every cluster: Indicator species analysis = high fidelity to and representative of a cluster according to permutation test (Dufrene and Legendre 1997)
- 8. Hierarchical clustering of tree species composition with indicator species analysis 147 clusters specific assemblages 29 clusters broad assemblages Pick levels of hierarchy that maximize: Number of significant indicator values Sum of significant indicator values
- 9. Slash pine-longleaf pine Balsam fir-quaking aspen Common persimmon Butternut-sweet birch Sourwood-scarlet oak Sugar maple-red maple Loblolly pine-sweetgum Hawthorn spp.-American plum Black willow Green ash-American elm Velvet mesquite Chittamwood Honey mesquite-Pinchot juniper Cedar elm Live oak-Ashe juniper California live oak-California laurel Blue oak-interior live oak Black cottonwood-bigleaf maple Alligator juniper-Arizona white oak Utah juniper-two needle pinyon Gambel oak Western juniper-curlleaf mountain mahogany Lodgepole pine-subalpine fir Rocky Mountain Douglas-fir-ponderosa pine Chokecherry-Pacific dogwood Oregon white oak Canyon live oak-California black oak Mountain hemlock-Pacific silver fir Coast Douglas-fir-western hemlock 29 broad assemblages
- 11. Slash pine-longleaf pine 2,934 plots Examples of 147 specific assemblages Coast Douglas-fir-western hemlock 10,093 plots
- 12. Hierarchical characterization Plot Multiple levels of classification Nation-wide forest composition Assess vulnerability Monitor change Project dynamics
- 13. Example: assessing climate change impacts using dominant species • Dominant species are likely to be ecologically important (Hildebrand et al. 2008) • Dominance structure is likely to be altered by global change drivers • Changes in dominance can be an early warning of impacts • Assessing potential impacts to dominant species is important • Within a cluster, species dominance index (SDI – Frieswyk et al. 2007): • High mean importance across all plots in the cluster • Tendency to occur with few other species
- 14. For 5 clusters in the East US: extract projected change in habitat suitability for dominant species e.g., slash pine-longleaf pine assemblage: slash pine longleaf pine pondcypress turkey oak Potential climate suitability change at plot locations and for the community Overlay modeled habitat suitability from Climate Change Tree Atlas (Climate Change Research Group 2014)
- 15. Balsam fir-quaking aspen Avg: -74.5% change Slash pine-longleaf pine Avg: 3.3% change Sugar maple-red maple Avg: -48.2% change Loblolly pine-sweetgum avg: 1.6% change Green ash-American elm avg: 19.3% change Projected suitability change: Hadley High scenario
- 16. Predicting broad assemblages based on ancillary environmental data Slash pine-longleaf pine Balsam fir-quaking aspen Common persimmon Butternut-sweet birch Sourwood-scarlet oak Sugar maple-red maple Loblolly pine-sweetgum Hawthorn spp.-American plum Black willow Green ash-American elm Velvet mesquite Chittamwood Honey mesquite-Pinchot juniper Cedar elm Live oak-Ashe juniper California live oak-California laurel Blue oak-interior live oak Black cottonwood-bigleaf maple Alligator juniper-Arizona white oak Utah juniper-two needle pinyon Gambel oak Western juniper-curlleaf mountain mahogany Lodgepole pine-subalpine fir Rocky Mountain Douglas-fir-ponderosa pine Chokecherry-Pacific dogwood Oregon white oak Canyon live oak-California black oak Mountain hemlock-Pacific silver fir Coast Douglas-fir-western hemlock Using Random Forests and bioclimate (8), soil (8), and topographic (5) variables
- 17. Planned next steps • Use re-measured plots to examine disturbance, land use, and climate drivers of past changes in cluster (among-community change) and dominance (within-community change) • Project observed changes into the future under climate and land use change scenarios as part of the USDA Forest Service 2020 Resources Planning Act (RPA) Assessment • Link with spatial projections of landscape change to project spatial patterns of forest species composition nationwide
- 18. Summary • We developed a hierarchical, empirical classification of inventory plots for the continental US • Leveraging information at multiple levels of organization fits in a macrosystems framework • Dominant species provide one example of how the classification can inform assessment of potential climate change impacts on forest communities • The classification can be used as the basis for monitoring, assessment, and projection of global change effects
- 19. Thank you • L. Iverson and M. Peters – USDA Forest Service Climate Change Atlas projections • R. Li – FIA database queries • K. Riitters – spatial data extraction Contact me: Jennifer Costanza jennifer_costanza@ncsu.edu
Editor's Notes
- A macrosystems framework is ideal for answering questions about changes in forest species composition Draws on landscape ecology, hierarchy theory, coupled human-natural systems, telecoupling (Rose et al. 2016) At regional to continental scales Using large data sets
- Important for: Wildlife habitat Forest structure Ecosystem services like carbon, water quality
- We wanted to accompany the classification with some metrics that could be used to monitor observed changes, assess which places may be vulnerable to change, and project potential future dynamics Example: assessment of potential climate change effects on assemblages via impacts to dominant tree species
- FIA is ideal for use in macrosystems ecological research. Brief description Systematic sampling across space and through time, so can facilitate hierarchical, cross-scale analysis and scaling up. Lots of people have used FIA data to answer macrosystems questions about forests and global change
- FIA is ideal for use in macrosystems ecological research. Brief description Systematic sampling across space and through time, so can facilitate hierarchical, cross-scale analysis and scaling up. Lots of people have used FIA data to answer macrosystems questions about forests and global change
- I will mostly focus on the 29 broad assemblages, with a mention of just a few of the specific assemblages
- We labeled each assemblage with the top two indicator species Clusters toward the top tend to occur in east (with at least 1 exception) Clusters toward the bottom tend to occur in west
- First one correpsonds to 10 specific assemblages Second one corresponds to 4 specific assemblages
- For Hadley High scenario Further work should examine local effects at smaller extents; because we know which plots may be impacted more and where they are, we can inform local scale efforts.