CORRELATING CHANGE IN LANDSCAPE SCALE CARBON STORAGE THROUGH SHIFTING FOREST OWNERSHIP AND MANAGEMENT GOALS

• Sponsoring Institution: National Institute of Food and Agriculture
• Project Status: COMPLETE
• Funding Source: MCINTIRE-STENNIS
• Reporting Frequency: Annual
• Accession No.: 0224849
• Project Start Date: Oct 1, 2010
• Project End Date: Sep 30, 2014
• Program Code: [(N/A)]- (N/A)

Recipient Organization
MICHIGAN TECHNOLOGICAL UNIV
1400 Townsend Drive
HOUGHTON,MI 49931

Performing Department
School of Forest Resources and Environmental Science

Non Technical Summary
Land ownership across large portions of the Upper Great Lakes region is in considerable flux. As the forest products industry shrinks, commercial timber management companies are selling their industrial timberland to predominantly non-timber industry owners with very different land management objectives, increasing the number of nonindustrial private forest (NIFP) owners across the study region. NIPF owners are also transferring their own properties to younger generations or other NIPF owners. Younger owners are often more likely to have higher incomes and live in urban areas far from their forestlands, and are less actively involved in forest management for any purpose (timber or otherwise). Commercial and NIPF land owners would be eligible to participate in carbon markets (through selling harvest rights to balance carbon-emitting activities), and these carbon permits could represent a critical source of income to the predominantly rural communities throughout the Upper Great Lakes Region. Indeed, an industry of creating carbon permits through reforestation could alleviate the economic losses due to the significant reduction of the timber and mining industries in large portions of this region. Policies encouraging carbon sequestration on privately owned land, such as the Michigan Working Forest Carbon Offset Program, could provide regional and global benefits by ameliorating climate change through increased carbon sequestration. Additionally, carbon sequestration may prove to be an important new source of income for small, rural landowners as carbon markets mature. However, ownership turnover of both commercial and NIPF forests will influence ownership characteristics and motivations, and likely have a profound, yet unknown effect on aggregate carbon storage. If programs such as the MWFCOP are going to work, an understanding of ownership turnover and parcelization rates, and their effect on forest cover, is key. The estimates generated in this project on ownership and land use change will contribute valuable information to a larger project, focused on carbon sequestration change due to ownership trends in forests throughout the Upper Great Lakes. Data on parcel ownership turnover and subdivision will be combined with land cover data to determine if ownership change is a driver of forest loss or gain. The project outputs will include spatially-explicit maps of ownership turnover and parcelization and change in forest cover. As a byproduct, the project will produce a detailed, multilayer GIS database of land ownership change at the parcel level for Houghton and Keweenaw Counties, which will be used for a cellular automata model for a larger project.

Animal Health Component

50%

Research Effort Categories

Basic

50%

Applied

50%

Developmental

(N/A)

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
123	0699	1070	100%

Knowledge Area
123 - Management and Sustainability of Forest Resources;

Subject Of Investigation
0699 - Trees, forests, and forest products, general;

Field Of Science
1070 - Ecology;

Keywords

private forests

nipf

ownership turnover

parcelization

deforestation

reforestation

carbon sequestration

forest policy

gis

remote sensing

Goals / Objectives
This project has two objectives. First, measure the rate of private forest ownership turnover and parcelization in Houghton and Keweenaw Counties. (Hypothesis: Turnover and parcelization rates are different for industrial versus nonindustrial ownership.) Second, determine if turnover/parcelization predictably affects land cover (i.e., forest, nonforest). (Hypothesis: Greater rates of turnover and/or parcelization are correlated with great loss of forest cover.) Year 1 will begin on October 1, 2010. Data collection and layer overlays will occur through the fall of 2011. Final data analysis will occur in early 2012, and manuscripts will be completed and submitted for publication by August 2012. Project completion reports will be submitted prior to September 1, 2012. Objective 1 will be completed in the first year. There are two tasks in Objective 1: Identify parcels with ownership change between GIS layer years, measure rate of ownership turnover; and Add turnover data per parcel to 2009 GIS map. The first three tasks from Objective 2 will be completed in Year 2: Identify LANDSAT/MODIS imagery for each year of parcel map availability; classify images; Pair and overlay turnover/parcelization layers onto LANDSAT and MODIS images; and Identify forested privately owned parcels for GIS parcel maps. The final task for Objective 2 (Turnover, parcelization, forest cover data analysis), as well as manuscript and report preparation and submission, will be completed in year 3. The project outputs will include spatially-explicit maps of ownership turnover and parcelization and change in forest cover. As a byproduct, the project will produce a detailed, multilayer GIS database of land ownership change at the parcel level for Houghton and Keweenaw Counties, which will be used for a cellular automata model for a larger project.

Project Methods
For Objective 1, I will purchase GIS parcel data that is commercially available (from Rockford Map Publishers) for Houghton and Keweenaw Counties. I already have the 2009 layer available, which leaves the following years: 2006 and 2003 (available as GIS vector layers); and 1997, 1994, 1991, 1988, 1985, 1981, 1978, 1974, 1970 (as TIF image files). We will generate a table with an entry for each parcel, indicating in which year (if any) that the parcel has transferred ownership and/or been subdivided into multiple parcels. Turnover in new, subdivided parcels will be tracked for the remainder of the years. We will then estimate rates of ownership change and parcelization for Houghton and Keweenaw counties over this 40 year period, based on the data table generated. We will the number of times per decade that there has been a change in ownership and, when possible, and any change in the basic characteristics of historic and current owners (e.g., timber company or investment firm, primary residence for nonindustrial owner). This will allow us to estimate the probability of ownership and management change for each land parcel in each time period (i.e., between years with available parcel maps). We will conduct a statistical analysis (e.g., Chi-squared test) to determine if NIPF parcels have greater turnover and/or parcelization rates than private industrial parcels. For Objective 2, we will overlay the 2009 GIS parcel map onto LANDSAT images from each year that we have parcel information. We will extract the portion of each image that covers the study area and classify the images, using NLCD Anderson Level 1 data for 1992, 2001 and 2006 as a guide for supervised classification methods. This will create an image for each parcel map that indicates forested and nonforested areas. Once these land cover images are created, we will add land cover information to the turnover/parcelization data. We will use statistical analyses (e.g., Pearsons correlation) to determine whether there is a relationship between turnover and parcelization rates, and loss (or gain) of forest cover at the parcel level. The results of this project will be presented at regional conferences and organizations for private forest owners. We will incorporate these results into larger projects that include interviews of regional private forest owners. These interviews will include two-way information exchange, particularly on the existence and appropriateness of state and federal forest management policies (including those emphasizing payment for carbon sequestration).

Progress 10/01/12 to 09/30/14

Outputs
Target Audience: The target audiences include other researchers (who will be addressed through peer-reviewed publications and presentations), policy-makers (who will be addressed through meetings and publications in trade journals), and private forest owners (who have been partially addressed through interviews, and will be addressed through contacts and presentations to landowner organizations). For the two latter groups, our goal is to effect a "change in knowledge" (with respect to policy impacts on landowner turnover, parcelization, and changes in harvest decision), and a "change in action and conditions" such that the forests in the UP are managed to maximize environmental, economic and social benefits. In contrast to the interaction from last year, this year has focused on using remote sensing to complement the information gathered from landowners. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided? The PhD student had the opportunity to put together a presentation of the work for the IALE conference, and he will be presenting the more complete analysis at the AGU meeting this December. He has also received training and PD opportunities for grant writing (we received several free satellite images from the company DCMI to fill a gap in the Landsat series, through a competitive proposal process) and journal article writing (we are writing two journal articles based on this work). How have the results been disseminated to communities of interest? So far our work has been disseminated to researchers through journal articles and conference publications; however, the journal article is Open Access and can be downloaded by anyone. Through our collaboration with WUPPDR we expect to reach forest owners and local and state policy makers in the upcoming year of the project. What do you plan to do during the next reporting period to accomplish the goals? In the next reporting period we will finish the following: 1. GIS shapefile attribute table of forest owner turnover (for privacy reasons, we may aggregate to means per Census Block or Township when releasing data to the public) 2. Statistical analysis relating ownership turnover to harvest occurence and intensity 3. Overlay of GIS harvest layer with carbon sequestion estimates (from models based on forest type) 4. Estimates of impact on carbon sequestration by private forest owner harvesting 5. 1-2 journal articles submitted, 2-3 presentations at scientific meetings, end-of-project briefing to WUPPDR colleagues. Project information will also be available to the public via faculty webpages hosted by Michigan Tech.

Impacts
What was accomplished under these goals? We used classifed LANDSAT imagery to measure the Normalized Difference Vegetation Index (NDVI) from one year to the next in small forested parcels in the western UP. We acquired additional harvest validation data from private foresters to calibrate the methdology and refine our sensitivity to +/- 10% basal area harvest. We are completing the automation of cloud-masking for new imagery and harvest detection using Object-Based Image Analysis. Once that is complete, we can link it to the Agent-Based Model that will produce scenarios of potential future harvests and carbon dynamics. We have also begun collaborating with the Western Upper Peninsula Planning and Development Region (WUPPDR), and they have contributed some data (parcel shapefiles) and expertise to our project that have allowed us to begin measuring ownership turnover on a roughly 5-year basis much more quickly (this task is still in progress). In turn, we will share the methdology and data produced with them. Our collaboration with WUPPDR is expected to last far beyond the conclusion of this project. PRESENTATIONS: Tortini R and AL Mayer. 2013. International Association for Landscape Ecology - US Region, Austin TX. "Remote detection of harvesting of nonindustrial private forests in the Western Upper Peninsula of Michigan (USA) using Landsat annual series."

Publications

• Type: Journal Articles Status: Published Year Published: 2013 Citation: Mayer AL and MD Rouleau. 2013. ForestSim model of impacts of smallholder dynamics: forested landscapes of the Upper Peninsula of Michigan. International Journal of Forestry Research, article ID 520207, 13 pp. http://dx.doi.org/10.1155/2013/520207.

Progress 10/01/10 to 09/30/14

Outputs
Target Audience: We presented the results of our work at two professional meetings; one to the American Geological Union and one to the US Regional Chapter of the International Association for Landscape Ecology. Scientists (faculty, government, graduate students) were the primary audience, and we had considerable interest from employees of the USDA Forest Service, particularly those working with nonindustrial forest owners and those working with GIS and remote sensing. We have submitted our work to a peer-reviewed science journal, and the primary audience is also scientists as well. However, our future work will build off of the work performed here, and this future work will include using the results for policy recommendations and educating forest owners about landscape-scale patterns in forest harvesting. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided? One graduate student (Riccardo Tortini) has received a great deal of training (e.g., writing publications, giving presentations) and professional development (e.g., travel to professional meetings, presenting work and attending training workshops related to remote sensing). How have the results been disseminated to communities of interest? Results have been disseminated to the scientific community (including USDA Forest Service researchers) through multiple presentations at professional meetings, and through a manuscript submitted to a peer-reviewed journal (Remote Sensing of Environment). What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

Impacts
What was accomplished under these goals? Objective 1: Task 1: remains incomplete. Parcel ownership data prior to the past 5-10 years remains undigitized for much of the Western Upper Peninsula, and tax records and/or plat books must be checked by hand to identify ownership turnover. This work is on-going. Task 2: Turnover data for certain counties (2006 and later) has been added to the classified imagery for forest harvests from 2000 onwards. Analysis is on-going, and final rates will be determined once the classification error (i.e., accuracy of harvests) is measured. On the other hand, with the data we are not only able to identify harvests, but the intensity of harvests, down to a removal > 20% of basal area, in increments of ~20% basal area (that is, we can differentiate between a harvest that removes 30% of the basal area of a forest, versus one that removes 50% of basal area). Therefore, our data not only reflects harvest/not harvest, but biomass removal. Objective 2: Task 1: Identification of Landsat/other imagery: completed. Task 2: pairing ownership turnover to classifed imagery: on-going Task 3: Identification of privately owned parcels: completed. Thus, the major obstacle to completion of this work has been the lack of digitized parcel owner maps, and for a few counties the lack of access to ownership information. These tasks will be completed, but not on time.

Publications

• Type: Journal Articles Status: Submitted Year Published: 2014 Citation: Tortini R, AL Mayer, P Maianti. Submitted. Using a GEOBIA approach and Landsat TM data to detect harvesting on nonindustrial private property in upper Michigan. Remote Sensing of Environment
• Type: Conference Papers and Presentations Status: Accepted Year Published: 2014 Citation: Tortini R, AL Mayer, and P Maianti. 2014. International Association for Landscape Ecology - US Region, Anchorage AK. "Geographic Object-Based Image Analysis for detection and quantification of nonindustrial forest harvests in the Western Upper Peninsula of Michigan, USA."
• Type: Conference Papers and Presentations Status: Accepted Year Published: 2013 Citation: Tortini R, AL Mayer, and P Maianti. 2013. American Geophysical Union annual meeting, San Francisco CA. "Interoperability of Landsat and DMC imagery for continuous detection and quantification of nonindustrial forest harvests in the Western Upper Peninsula of Michigan, USA."

Progress 10/01/12 to 09/30/13

Outputs
Target Audience: The target audiences include other researchers (who will be addressed through peer-reviewed publications and presentations), policy-makers (who will be addressed through meetings and publications in trade journals), and private forest owners (who have been partially addressed through interviews, and will be addressed through contacts and presentations to landowner organizations). For the two latter groups, our goal is to effect a "change in knowledge" (with respect to policy impacts on landowner turnover, parcelization, and changes in harvest decision), and a "change in action and conditions" such that the forests in the UP are managed to maximize environmental, economic and social benefits. In contrast to the interaction from last year, this year has focused on using remote sensing to complement the information gathered from landowners. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided? The PhD student had the opportunity to put together a presentation of the work for the IALE conference, and he will be presenting the more complete analysis at the AGU meeting this December. He has also received training and PD opportunities for grant writing (we received several free satellite images from the company DCMI to fill a gap in the Landsat series, through a competitive proposal process) and journal article writing (we are writing two journal articles based on this work). How have the results been disseminated to communities of interest? So far our work has been disseminated to researchers through journal articles and conference publications; however, the journal article is Open Access and can be downloaded by anyone. Through our collaboration with WUPPDR we expect to reach forest owners and local and state policy makers in the upcoming year of the project. What do you plan to do during the next reporting period to accomplish the goals? In the next reporting period we will finish the following: 1. GIS shapefile attribute table of forest owner turnover (for privacy reasons, we may aggregate to means per Census Block or Township when releasing data to the public) 2. Statistical analysis relating ownership turnover to harvest occurence and intensity 3. Overlay of GIS harvest layer with carbon sequestion estimates (from models based on forest type) 4. Estimates of impact on carbon sequestration by private forest owner harvesting 5. 1-2 journal articles submitted, 2-3 presentations at scientific meetings, end-of-project briefing to WUPPDR colleagues. Project information will also be available to the public via faculty webpages hosted by Michigan Tech.

Impacts
What was accomplished under these goals? We used classifed LANDSAT imagery to measure the Normalized Difference Vegetation Index (NDVI) from one year to the next in small forested parcels in the western UP. We acquired additional harvest validation data from private foresters to calibrate the methdology and refine our sensitivity to +/- 10% basal area harvest. We are completing the automation of cloud-masking for new imagery and harvest detection using Object-Based Image Analysis. Once that is complete, we can link it to the Agent-Based Model that will produce scenarios of potential future harvests and carbon dynamics. We have also begun collaborating with the Western Upper Peninsula Planning and Development Region (WUPPDR), and they have contributed some data (parcel shapefiles) and expertise to our project that have allowed us to begin measuring ownership turnover on a roughly 5-year basis much more quickly (this task is still in progress). In turn, we will share the methdology and data produced with them. Our collaboration with WUPPDR is expected to last far beyond the conclusion of this project. PRESENTATIONS: Tortini R and AL Mayer. 2013. International Association for Landscape Ecology - US Region, Austin TX. "Remote detection of harvesting of nonindustrial private forests in the Western Upper Peninsula of Michigan (USA) using Landsat annual series."

Publications

• Type: Journal Articles Status: Published Year Published: 2013 Citation: Mayer AL and MD Rouleau. 2013. ForestSim model of impacts of smallholder dynamics: forested landscapes of the Upper Peninsula of Michigan. International Journal of Forestry Research, article ID 520207, 13 pp. http://dx.doi.org/10.1155/2013/520207.

Progress 10/01/11 to 09/30/12

Outputs
OUTPUTS: We used classifed LANDSAT imagery to measure the NDVI from one year to the next in small forested parcels in the Western Upper Peninsula of Michigan. We used known harvest levels from the Michigan Tech Ford Forest to calibrate the methdology, and detected selective harvests down to 22% of basal area. We are now identifying harvests in private parcels and verifying harvest prediction in a parnership with a local forest consultant. We will produce annual harvest rates and types of harvests (selective to clearcut) for three different types of forests (northern harboods, conifer, mixed), and will identify the pattern of these harvests. All of these data are also being used in an agent-based model for predicting forest cover change in the Western Upper Peninsula of Michigan PARTICIPANTS: Riccardo Tortini, PhD student in Geological Engineering. Dr. Mark D. Rouleau, Assistant Professor, Department of Social Science, Michigan Technological University. The student received significant training in remote sensing methods, GIS, and modeling. TARGET AUDIENCES: The target audiences include other researchers (who will be addressed through peer-reviewed publications and presenations), policy-makers (who will be addressed through meetings and publications in trade journals), and private forest owners (who have been partially addressed through interviews, and will be addressed through contacts and presentations to landowners organizations). For the two latter groups, out goal is to effect a "change in knowledge" )with respect to policy impacts on landowner turnover, parcelization, and changes in harvest decision), and a "change in action and conditions" such that the forests in the Upper Peninsula are managed to maximize environmental, economic and social benefits. In contrast to the interaction from last year, this year has focused on using remote sensing to complement the information gathered from landowners. PROJECT MODIFICATIONS: Nothing significant to report during this reporting period.

Impacts
The remote sensing methodology developed this year can be used to measure "change in action" amongst private landowners in year 3 of the project, as they are shown the model simulations regarding different harvest rates. We will show landowners (and policy makers) not only harvest rates and patterns, but impacts on carbon sequestration (by measuring biomass removal and regrowth).

Publications

• JOURNAL Articles: Schubert JR and AL Mayer. 2012. Peer influence of nonindustrial private forest owners in the Western Upper Peninsula of Michigan. Open Jouranl of Forestry 2(3):147-155.
• PRESENTATIONS: Mayer AL. 2012. "Landscapes of prvate forest owners: dynamics, management and policy." University of Missouri, Department of Forestry.

Progress 10/01/10 to 09/30/11

Outputs
OUTPUTS: We produced an Excel file with parcel turnover (change in ownership)and parcelization (increase or decrease in number of parcels) for > 20% of parcels in Houghton and Keweenaw counties for each decade, using plat maps from 1950 to 2009. These files will be used in the next phase to identify areas in LANDSAT photos and measure rates of deforestation/reforestation. We also performed 35 interviews of private forest owners in the Western UP,to determine thir plans for selling or harvesting their properties in the near future. These interview data will be used to cross-check the plat map based data. All of these data are also being used in an agent-based model for predicting forest cover change in the Western Upper Peninsula of Michigan. PARTICIPANTS: Jillian R.Schubert-MS student in Environmental Policy, Michigan Technological University(graduated May 2011). Mackenzie Roeser;Undergraduate student in Computer Science; Michigan Technological University Dr. Mark D. Rouleau, Assistant Professor, Department of Social Sciences, Michigan Technological University. The students received significant training in ethnographic methods, GIS, and modeling. TARGET AUDIENCES: The target audiences include other researchers (who will be addressed through peer-reviewed publications and presentations), policy-makers,(who will be addressed through meetings and publications in trade journals),and private forest owners,(who have been partially addressed through interviews, and will be addressed through contacts and presentations to landowner organizations). For the two latter groups, our goal is to effect a "change in knowledge" (with respect to policy impacts on landowner turnover, parcelization, and changes in harvest decision),and a "change in action and conditions" such that the forests in the UP are managed to maximize environmental, economic and social benefits. Given that the project is in its first year, all of our work has focused on some "change in knowledge" among a small group of private forest owners, and towards a "change in knowledge" among other researchers. PROJECT MODIFICATIONS: Nothing significant to report during this reporting period.

Impacts
We induced considerable "change in knowledge" among the private forest owners whom we interviewed. This new knowledge regarding where they receive information for forest management may induce a "change in action" amongst these landowners, although, this will be difficult to verify.

Publications

• ABSTRACTS: Mayer, AL, MD Rouleau, M Roeser, JR Schubert. 2011. Modeling the impacts of communication among non-industrial private forest owners on forested landscapes. Annual meeting of the International Association for Landscape Ecology--US Region, Portland,OR.
• THESIS: Schubert, JR, 2011. Peer influence in NIPF landowner decision making in the Western Upper Peninsula, Michigan. Master's Thesis, Michigan Technological University.