Climate Change and Thresholds of Ecosystem Change: Invasibility of Tundra in the Northern Rocky Mountains

Duration:  July 1999 - June 2004

Most mountainous areas of the world have an upper elevational limit to the distribution of forests.  Above this treeline is alpine tundra, a community of non-woody plants which can survive the harsh climatic conditions found near the tops of mountains.  The alpine treeline is actually a zone of transition between continuous, mature forest and the alpine tundra and represents a threshold for ecological processes. This transition zone reflects underlying biological tension in growth, reproduction and competition because many of the plants are operating at their extreme limits.

Because the alpine treeline represents the upper threshold for tree life, a common assumption is that climate change will be readily detected at treeline by changes in tree growth and upslope establishment of seedlings.  This means that alpine tundra will be invaded by seedlings and that the overall spatial pattern of tree communities at upper elevations will change significantly.  However, there are many influences on tree growth and establishment besides climatic conditions.

Topography, soils and geology, and disturbance such as fires, avalanches and landslides all interact with climate in complex ways to create treelines.  Whether or not treelines respond sensitively to climatic change - and so could be global indicators of change - must be determined.

The aim of this research is to assess the sensitivity of alpine tundra to invasion by woody species from treeline in the northern Rocky Mountain region.  We will examine the question of whether treeline dynamics are sensitive to climatic change in a nonlinear way - and thus might show little response until a threshold is crossed, when the response would then be great.  This will be accomplished by developing simulation models of tree species establishment and growth based on a combination of more physiologically-based plant growth models and spatially-oriented forest dynamics models.

The models will be developed and validated at many spatial scales using information from satellites and by field work that will measure wind, snow, and soil factors that influence the growth of woody plants and tundra.

The results will allow the interpretation of past and ongoing changes at and above treeline in the wildlands of the western US.  The sensitivity of tundra to invasion is significant because considerable areas of natural resource - with value for wildlife, recreation, and aesthetics - exist just above the treeline ecotone and this ecotone may or may not be a sensitive indicator of climatic change.

Study Area:

 
In Glacier National Park (GNP), 80% of the transition from forest to tundra occurs over a 550 meter range; in contrast, in Rocky Mountain National Park, 80% of the transition occurs over 200 meters.

The greater variability in alpine treeline ecotone elevation in GNP may be due to combinations of variability in macro-climate, microclimate, topography, and snow and debris avalanches, and competition with tundra.  It is this variability that makes GNP an excellent place to study spatial pattern.

Approach:

Modeling

We will develop interacting models of treeline response at 3 scales.

First, we will update an existing cartographic model so that it will identify where treeline is topographically controlled at meso-scales.  Sensitivity of tundra to invasion varies as a function of both meso and micro-scale conditions.  Meso-scale conditions addressed by this model are related to topographic and lithologic (rock) position and are relatively stable over time.  Since these locations change little over long time scales, treeline will have been stable in some areas because any potential advance would be prevented by lithologic and related disturbance conditions (cliffs, rockfall, debris flow).  Two geographic Information System (GIS) data layers will be produced to serve as inputs to the other models.  They will identify areas where treeline is controlled by features such as cliffs and snow avalanche paths.  

Second, we will further integrate a mechanistic biogeochemical model with a forest gap type model into a "core" model to examine treeline dynamics at micro-scales.  Where topography and lithology are not limiting, fine scale conditions influence fine-scale tree pattern within the ecotone (transition zone).  Future changes in the position/pattern of the ecotone will mirror these conditions.  We will run the core model for selected areas at a 10 m resolution and group 10 m cells into 30 m windows to calculate tree abundance and density. 

Third, we will develop an additional model to examine patterns at 1 m resolution.  From the core model we will derive a set of rules to create this simple model for simulating micro-scale processes.  This model will be validated using digital multispectral aerial imagery, digital orthophotoquads and ground samples for a test slope.  All models will be analyzed for sensitivity.  We will simulate an array of conditions based on climate change scenarios and produce maps for the projections.

Fourth, temporal scale will be addressed by running models at an annual iteration to examine decadal-scale variability.  The underlying core model will provide information that can be analyzed for annual variability.  We will develop a tree ring chronology at one site where it is apparent that trees successfully invaded tundra since the Little Ice Age, and we will assess a longer term trend by radiocarbon dating fossil wood that appears to represent the ecotone in this area prior to the Little Ice Age.

Remote Sensing

Remotely sensed data will be used for validation of the patterns of treeline vegetation generated from the various modeling approaches.  We will use:

• Landsat Thematic Mapper imagery
• ADAR 5500 airborne high-resolution multi-spectral imagery, and
• Digital orthophotography from the U.S. Geological Survey (below)

Combined, the three data sets will facilitate the mapping of leaf area index (LAI), vegetation classes, and tree cover at resolutions of 30 m, 10 m, and 1 m.  We will measure the reflectance and transmittance properties of the vegetation canopy in the field to calibrate relationships between vegetation properties and the remotely sensed images.

    
Field parameterization and ground truthing

Field data are needed for model development and to test remote sensing observations.

Microclimate - We will monitor micrometeorological conditions year-round for 2 years with 2 ridge top stations at Lee Ridge, our primary site.  Additionally, data on wind speed, temperature within the canopy, radiation transmission through the canopy, snow depth, relative humidity, and soil moisture and temperatures within the rooting zone will be collected within krummholz and patches of trees.

Tree chronology - Cores will be taken from trees to reconstruct the chronology of tree invasion on Lee Ridge and to validate model outputs.  A sample of tree rings other than that collected for parameterization will be used to validate the model simulations.

Soil - On Lee Ridge we will sample soil along gradients of tree establishment and occupation to relate soil development, specifically nitrogen and organic matter, to tree occupancy.  We will sample soil along several invading tree fingers, in tree patches, at lone trees, and in tundra.  

Application of Results:

It has been asserted that global warming in mountain regions will be evident from upward migration of alpine treelines and that this measure can be used globally to track rates of climate change.  Unfortunately, all the processes controlling treeline have not been defined well enough to support this assertion.  When completed, the integrated predictive models we will have developed should be applicable to treelines worldwide and can be used to assess if, and where, monitoring treelines as evidence of climate change is appropriate.

Products:

Information will be made available to Department of the Interior (DOI) and other agencies, to the general public, and to the scientific community. This project will produce a variety of digital data, including raw imagery, interpreted imagery, and computer simulation projections.  All of these will be collected in matching georeferenced files and made available as data files to DOI agencies, primarily the USGS-BRD and the National Park Service.  These data will also be reproduced as image files and posted on a web page for public access.

Specific products will include:  

• Georeferenced data sets for interpreted Landsat, ADAR, DOQ and DEM data
• Simulation results in full detail, including maps from sensitivity analyses
• Field data on aspects of the potential positive feedbacks expected at the ecotone in measures of microclimate, soil and sediment, and vegetation growth
• Maps of land-cover, environmental gradients, and results of simulations in digital form
• Peer-reviewed publications, and publications in regional outlets, in cooperation with the educational mission of the Glacier Institute and the Glacier Natural History Association
     

Collaborators:

George Malanson, Dept. of Geography, University of Iowa, Iowa City, IA 52242; David Butler, Dept. of Geography and Planning, Southwest Texas State University, San Marcos, TX 78666;
Stephen J. Walsh, Dept. of Geography, University of North Carolina, Chapel Hill, NC 27599;
David Cairns, Dept. of Geography, Texas A&M University, College Station, TX 77843;
Daniel Brown, Dept. of Geography, University of Michigan, Ann Arbor, MI.

	Primary Contact:  Dan Fagre, USGS Northern Rocky Mountain Science Center, Glacier Field Station E-mail Dr. Fagre