Date of Report: 28 June 2000

LCMR Work Program Update Report

Date of Next Status Report: 30 June 2001

Date of Workprogram Approval:

Project Completion Date: 30 June 2001

LCMR Work Program 1999

I. PROJECT TITLE: Predicting Water and Forest Resources Health and Sustainability

Project Manager: JoAnn M. Hanowski

Affiliation: Natural Resources Research Institute

Mailing Address: 5013 Miller Trunk Highway, Duluth, MN 55811

Total Biennial Project Budget:

________________________

A. Legal Citation: ML 1999, [Chap. 231], Sec. [16], Subd. [12]. Benchmark and Indicators

Predicting Water and Forest Resources Health and Sustainability

Appropriation Language: Predicting Water and Forest Resources Health and Sustainability $150,000 the first year and $150,000 the second year are from the trust fund to the University of Minnesota, Natural Resources Research Institute, to assess ecosystem health using indicators to develop models that incorporate landscape composition change.

B. Status of Match Requirement: No match required.

II. PROJECT SUMMARY AND RESULTS:

Our objective is to compile existing or create new indicators of biodiversity, soil productivity and water quality and to incorporate these metrics into a decision support model that will predict forest and aquatic ecosystem health and sustainability under a variety of management scenarios. Ecosystem health will be quantified by an index and this index compared to a range of indices calculated for the historical condition of the landscape undergoing a natural disturbance regime. Using existing databases on forest birds, amphibians, aquatic insects and ecosystem conditions, we propose to integrate tools such as geographic information systems (GIS), satellite image analysis, multivariate statistical methods, and simulation models to create this management tool. Project investigators are currently documenting changes in terrestrial and aquatic species resulting from forest and agricultural management practices in several regions of Minnesota, therefore large data sets on birds and aquatic insects are on-hand. Other databases will be acquired and used in indicator development. Relevant GIS data layers also have been developed or acquired for Minnesota, including forest cover classification, elevation, hydrography, soils, surficial geology and wetland classification. Together the environmental and organismal data will be used to test existing indicators and develop new indicators where necessary. Specific objectives are to: 1) identify and compile existing data for specific biota in three regions of Minnesota; 2) develop metrics to measure ecosystem condition at different scales; 3) evaluate the effects of land use changes on these metrics for a variety of management scenarios and 4) develop a workable decision making system that can be used by a variety of land planning agencies, industrial groups, and private citizens.

III. PROGRESS SUMMARY:

Background Information: The first step in this study was to identify study areas. We selected three subsections of the Minnesota Department of Natural Resources Ecological Classification System, Chippewa Plains, North Shore, and Rochester Plateau. These areas were selected because: 1) relevant biological and physical data were available for all three areas; 2) work on identifying and mapping native plant communities and their range of natural variability was planned and since completed for the northern sites; and 3) these regions represent the range of forest and water resource sustainabilty issues current to the State of Minnesota.

In addition to study site selection, the other major task completed was defining "sustainability" and how it will be quantified for use in our models. We decided to use the concept of natural range of variation in native plant communities as the basis for defining sustainability. In this sense, we infer that forest and water resources will be sustainable if we manage within the range of variation that they occurred on the landscape throughout time. Two steps are required to obtain this information. The first step, calculation of the range of natural variation for native plant communities has been completed by Lee Frelich. The next step is to map locations and extent of native communities so that we are able to calculate the amount of area and location of the various native communities. This has been completed for the Chippewa Plains study area by Dave Shadis and John Almendinger. The North Shore study area is currently being mapped by personnel at NRRI with funding from the Minnesota Forest Resources Council and their assistance coordinated by Dave Miller (see detail below). We will be able to use this definition of sustainability for the two northern study areas where the map and range of variability has been completed. We are still working on a sustainability definition for the southern study area.

Native Ecosystem Mapping for the Northern Superior Uplands. In order to apply the range of natural variation calculations to determine the land area occupied by vegetation growth stages, a map showing the distribution of native ecosystem types is required. We are using a statistical based modeling approach utilizing GIS, forest inventory, classified Landsat Thematic Mapper data and other vegetation plot data along with physical data such as soils, landform, climate and topography to predict the distribution of 8 native ecosystem classes in the Northern Superior Uplands ecological section. We acquired vegetation data from the following sources: the Superior National Forest, Minnesota DNR Resource Assessment, Minnesota DNR Non-Game Heritage Program releve plots, and classified satellite data from P. Wolter from the University of Minnesota, Natural Resources Research Institute. The physical data came from a variety of sources, including: Minnesota Soil Atlas, Geomorphology of Minnesota, USGS digital elevation models, and Zedex High Resolution climate data. As of June 2000, vegetation data has been processed and integrated with physical data in GIS. Preliminary maps have been developed and are in the process of being evaluated and refined.

Result 1: We have completed the process of identifying and compiling relevant data for the biotic (amphibians, birds, fish, aquatic insects) and physical components (GIS data bases) required for the project. Data were compiled separately for each biotic subgroup and are currently processed by one individual who will be responsible for developing the models and management tools. More specific details on each biotic subgroup follows.

Amphibians: Little survey data exists for amphibians in Minnesota, therefore we have compiled data from the primary literature and other publications to try to determine species-habitat relationships for Minnesota amphibian species and 2 turtles. To date we have compiled a large bibliography and have scanned many papers and books for appropriate information. Species accounts for the amphibians which occur in our three study Subsections (8 frogs, 4 salamanders, 1 newt, 1 toad) have been completed. We have also gathered information for two turtle species of special concern in Minnesota: Wood and Blanding’s turtles. Amount and quality of data pertaining to each species varies widely.

Birds: Several relevant data exists for breeding birds in all three study regions. For this study we decided to use forest bird monitoring data that has been collected by NRRI staff over the past 5-9 years. These data were chosen because: 1) they are linked directly to forest cover type and age; 2) they represent standardized counts conducted by qualified and trained observers; 3) relative abundance and probability of occurrence of over 90 species are available; and 4) it is the largest data base available for breeding birds in the upper midwest. In addition breeding bird data for agricultural and open habitat associated species for the southern study area were available from previous studies conducted by NRRI personnel.

Aquatic insects and fish: We have been in the process of gathering existing macroinvertebrate data sets and have also gathered some stream fish data sets which we hope to use in addition to the macroinvertebrates described in the proposal. Both macroinvertebrate and fish data are less abundant for the Chippewa Plains subsection, but plentiful for the North Shore Highlands and Rochester Plateau. Macroinvertebrate data sources include: 1) North Shore study area collected by NRRI personnel; 2) Chippewa Plains study area from Leech Lake Reservation and possibly Cass Lake Lab; 3) Rochester Plateau study area collected by NRRI personnel. Fish data sources include: 1) North Shore study area from EPA (Duluth office) and MPCA; 2) Chippewa Plains study area from USFS (Chippewa National Forest) and MnDNR; and 3) Rochester Plateau study area collected by NRRI personnel. We have gathered all data that are available.

Result 2: We have made considerable progress on the development of indicators for each of the biotic sub-groups (see below). This work will continue through and be completed by December 2000.

Amphibians: We have begun to incorporate species-habitat relationships into the modeling framework. We are currently assessing how GIS data will be used to match our habitat requirements with spatial data. We are also exploring the use of species trait characteristics that reflect reproductive strategies, food preferences, and life history characteristics to quantify how land use affects the amphibian community. This type of analysis has been widely used for fish and aquatic macroinvertebrates. However, there has been little similar work done for amphibians, meaning there is little existing data to draw upon for choosing metrics for a multimetric index. Candidate metrics will be tested using a data set from central Minnesota.

Birds: Indicator species for the two northern study sites have been selected. Our intent for this exercise was to identify bird species that would best reflect changes in the amount of specific cover types in the landscape. Indicator values were calculated for all bird species by cover type and age category (when data were available). With this process we selected three bird species that had the highest indicator value for each cover type and age class. Thirty-two species were selected for the Chippewa Plains and 33 species will be used as indicator species for the North Shore study area.

Aquatic insects and fish: We plan to use a variety of published or original metrics and multimetric indices as our ecological indicators to assess the impacts of various land management scenarios on macroinvertebrates and probably fish as well. Because macroinvertebrate communities are more related to broader land cover classes (e.g. forest, agriculture, urban) rather than specific forest types (e.g. conifer, deciduous), we plan to use these broader classifications in our analyses. Because macroinvertebrates respond to conditions within individual watersheds, we will assess this community at the watershed scale, rather than whole subsection scale. We have delineated watersheds that correspond to our North Shore macroinvertebrate sample sites which will be used in a GIS with an appropriate land use coverage to determine percentages of individual land cover types. We will then relate proportions of each land cover type to scores of the multimetric index to develop relationships that predict land management effects. These relationships have recently been analyzed for our data in the Rochester Plateau.

Our general analytical approach consists of two steps. First, we will examine relationships between potential biotic metrics and watershed variables using redundancy analysis (RDA). RDA identifies the most important watershed scale variables that explain the most variation in the entire set of biotic metrics via a forward-stepwise selection procedure. This reduced set of variables is our list of ecological indicators at the watershed scale. Second, we use this reduced list of watershed variables as explanatory variables in separate multiple regression analyses for each of the potential biotic metrics. Regressions that are significant identify the biotic metrics that will be useful as ecological indicators. In addition, these significant regression models predict the direction of change in each of our biotic metrics based on alterations to watershed-scale variables. The overall results are 1) a reduced list of ecological indicators or metrics, and 2) a predictive model for each of the biotic metrics, which can be used in the decision-making models.

We will use this approach for macroinvertebrates and fish for all of the regions. However, there are some challenges regarding fish metrics for coldwater streams of the North Shore due to low taxa richness. Stream characteristics vary considerably between the three subsections and locations from which data were collected. For both macroinvertebrates and fish, the multimetric indices will have to be tailored to the specific areas. Previous work at NRRI has developed a multimetric index for North Shore macroinvertebrates. Indices developed in other Midwest areas similar to the other subsections will be tried for both macroinvertebrates and fish.

Result 3: The structure of the general model that will predict forest and water resources sustainability has been defined. The basic unit is a Populations/Guild/Metric (PGM), a single number that may represent the population of a single species, or a guild of species, or some abiotic index, for example water quality, habitat suitability. Another example of a PGM would be an amphibian habitat suitability index expressed in terms of wetland cover and road density. Rules for specifying the landscape requirements (forest cover type, stand age, landuse, etc.) of each PGM has also been defined and a prototype application has been developed. This application can interpret PGM rule files and we have used it to predict the expected trends for bird species' abundance in response to changes in covertype on a hypothetical management region. The current rules specify bird abundance in terms of cover types used by Lee Frelich in his work that calculated the range of natural variation of cover types in the two northern study areas.

Result 4: Development of the GIS interface to run models and predict sustainability is beginning. We have chosen ArcView as the required user software for this application. We will explore the possibility of a purely web based version for future development.

Result 5: As recommended by the peer review panel, we identified and convened a steering committee that would provide input at various stages of the project and as potential end users of the management tool. The steering committee met at NRRI in the fall of 1999 and provided important input to the project. Another meeting is planned for fall 2000.

IV. OUTLINE OF PROJECT RESULTS:

Completion Date: 30 January 2000

Completion Date: 30 September 2000

Completion Date: 30 December 2000

Completion Date: 30 June 2001

Completion Date: 30 June 2001

V. DISSEMINATION: Information gathered in this project will be presented to a variety of audiences including both public and private land use planners and the scientific community. The final product, a decision making model will be made available in a usable format to the public. Information will also be available as a page on the Natural Resources Research Institute Web site http://www.nrri.umn.edu/SUSTAIN/.

VI. CONTEXT

A. Significance: Minnesota’s forest and water resources are important contributors to the well-being of citizens located throughout the state. Maintaining these vital resources is a common goal among Minnesota’s public because they provide a wide variety of benefits that contribute to the economy, local community and the natural environment. Measuring current, and predicting future forest and water resource health and sustainability is difficult because they are influenced by many factors (e.g., climate, land use change, land management practices, soil type, natural disturbances) and complex interactions among these factors. In addition, coordinated planning for multiple uses of these resources including resource management, tourism and recreation is difficult due to the existence of multiple regulatory bodies, existing land ownership patterns, and the inherent complexity of these ecosystems. Decisions regarding management of these systems is difficult because all of these factors need to be considered simultaneously. This situation mandates a need to develop novel approaches to assess current, and to predict future conditions in the economy, community and natural environment under different management regimes.

We can begin to understand this complex picture by identifying the role that each factor has and the scale on which they impact water and forest resource sustainability. Some of this information exists for Minnesota including a multiscale ecological classification of Minnesota (Ecological Classification System). Additional technological developments including geographic information systems (GIS), multivariate statistical methods, and spatially-explicit simulation models provide analytical capabilities not previously available. Basic research on animal species has provided important information on distribution of these species, their response to environmental stress, and their relationship to landscape features. Therefore, this combination of spatial databases, statistical methods, and biotic databases provides the potential to develop predictive models and decision making tools for assessing the response of organisms to a variety of management strategies and a method for measuring water and forest resource health.

Information for measuring forest and water resource health can be obtained directly by conducting a large number of individual measurements (a time-consuming and expensive procedure), or we can select a few important factors that have been shown to indicate ecosystem health. A large body of work has already been undertaken to quantify the response of sentinel organisms such as birds, amphibians and aquatic insects to a range of environmental conditions. Birds are key biological indicators of the health and stability of forest ecosystems and are relatively easy to study because their ecology is well known. Breeding birds also represent 60 to 70% of the terrestrial species biological diversity in Minnesota’s forests and the highest species richness in North America occurs in Minnesota. Current knowledge indicates that birds have a strong link to the forest economy due to their insectivorous habits and declines in numbers would result in a reduction in forest productivity. Because amphibians use both land and water resources, information on this group will provide the crucial link between water and forest systems. The use of benthic invertebrates for environmental monitoring also has long history due to their reliability and effectiveness as environmental indicators. They are an integral part of riverine systems and play a large role in detrital processing, nutrient cycling and are also a food source for higher trophic levels such as fish and birds. Macroinvertebrates are being used by resource managers in environmental monitoring and as indicators of local scale site conditions.

Investigators in this study have already gathered a large data set on birds and aquatic insects in several Minnesota watersheds and are currently documenting changes in terrestrial and aquatic species due to a variety of forest and agricultural management practices. The proposed work will complement these efforts by adding species group information where it is currently lacking. In addition, several GIS data layers have been developed for Minnesota including forest cover classification, elevation, hydrography, soils, surficial geology and wetland classification. Using both existing and newly collected data, we propose to integrate tools such as geographic information systems, satellite image analysis, multivariate statistical methods, and simulation models to predict forest ecosystem health and sustainability using forest birds, amphibians and aquatic insects.

B: Time: All objectives will be completed during the 1999-2001 biennium.

C: Budget Context: Researchers at the Natural Resources Research Institute have already gathered much of the information that will contribute to the database developed and used in this project. These include (but are not limited to the following: 1) Minnesota Forest Resources Council $120,000: Some baseline data on birds and aquatic insects as well as GIS information have been gathered for three watersheds in northern Minnesota. 2) Environmental Protection Agency $925,000: A project to develop aquatic macroinvertebrate indicator species models including work in the proposed site in SE Minnesota. 3) Minnesota Forest Bird Diversity Initiative $1,200,000: GIS forest cover data for the forested area of Minnesota and landscape models for breeding birds; 4) United States Forest Service $300,000: Trend and habitat information for breeding birds in the Chippewa and Superior National Forests, and $25,000: Effect of wetlands and forest harvest on stream insects. 5) Minnesota Sea Grant $94,828. 6) Lake Superior Decision Support Systems, $515,000: Development of detailed GIS databases and decision support tools for the Lake Superior Basin. 7) USFS Great Lakes Assessment $100,000: Development and data visualization and decision support tools for Minnesota, Wisconsin and Michigan.

We have developed a good working relationship with several forest product companies and Counties in northern Minnesota over the past 10 years. The companies and Counties have cooperated with us on several projects by: 1) providing property to conduct studies, 2) conducting forest experimental manipulations on these study areas, and 3) assigning representative foresters and biologists to participate in design and implementation of experiments. For this project, all three of the major pulp and paper companies in northern Minnesota (Boise Cascade, Blandin-UPM and Potlatch) will cooperate on this project. Lake County is also participating on this project. The value of the cooperators in-kind investment in this project is about $30,000 each or a total of $120,000. This represents costs for land-use, forest inventory and other data, and staff time. Cooperators will also be part of a steering committee that will provide input to the investigators during all stages of the project. In addition, a subset of the cooperators will test the model during the development process.

The $300,000 allocation from LCMR will used to compile and analyze existing biotic data that to develop indicators to assess forest and water resources health and sustainability. This information will be applied to develop a decision support program that can be used by resource planners across the State.

BUDGET:

2. Budget detail (see Attachment A)

VII. COOPERATION:

VIII: LOCATION: Lake, St. Louis, Itasca, Cass, Carlton, Rice, Dakota, Goodhue, LeSueur Counties