In 2012, about a dozen people began meeting at the Socio-environmental Synthesis Center (or SEYSNC) in Annapolis, MD, as part of a new project: Rural Forest Communities at a Tipping Point? Trends & Actionable Research Opportunities. We were a group of researchers spanning the social and environmental sciences from Virginia, Maine, Oregon, Ohio, Wisconsin, Quebec, and elsewhere, and we were interested in a seemingly straightforward, yet in practice elusive, concept of the rural forested community. We knew from the work we conducted in forested regions of the US and Canada that these areas had experienced dramatic employment shifts, out of extractive industries like timber or mining, manufacturing jobs in textile mills, and other activities that had previously made more intensive use of the land.
We were particularly interested in forested ecosystems that had experienced some regrowth and restoration of a variety of plant and animal species, and understanding how loss of traditional industries, burgeoning new industries like tourism, “green” manufacturing, or call centers seeking cheaper land and labor costs, were intersecting with new demographic trends, like exurban or second-home migrants.
We wanted to know which rural forested communities were able to offset job losses as traditional industries waned, while managing new threats to forests in the form of parcelization and forest fragmentation, climate change, coordination of forest management, invasive species and pests, among other factors. We had big questions we wanted to answer: if rural community and forestry futures were intertwined, in what ways were communities working to maintain or enhance these joint prospects? Were certain types of economic strategies (pursuing tourism over manufacturing) better or worse for the community or the forest? Our collective experience in studying small field sites across North America, and our transdisciplinary orientation, was to provide a rich foundation for answering these questions.
What follows is a story about data. Well, it’s really about teamwork, interdisciplinarity, measurement, accuracy versus precision, and how cartoons once came to our proverbial rescue.
Lesson 1: Collaboratively define a research object
We had assembled an intrepid disciplinary team of around a dozen ecologists, remote sensing specialists, environmental economists, rural sociologists, and geographers, among others. We all agreed in principle about the threats to rural forested communities in North America, and eagerly discussed similarities and differences among our regions. However, when it came down to specifying exactly what the defining characteristics of rural forested communities might be, and what data we would use to identify them spatially (what is rural? What is a forest? What is a community??), we were in trouble.
Figure 1: Locating the rural forested counties of the U.S. (Bell et al. 2019).
We argued that the word “community” means something specific in ecology, yet remains vague or contradictory across the social sciences. We argued that the Census counts record someone’s residence, not their place of work, and commuting relationships vary regionally and across the urban-rural continuum. How do we precisely map which forest is used by which people? How much vegetation cover should there be to define a forest, in such an ecologically diverse place as the continental U.S., even? There is no clear geographic definition of what constitutes a rural forested community. But we needed to operationalize data – census data on population, economic census data on employment and occupation, land-cover datasets for information on forest extent, etc.
People are mobile and drive for miles to harvest or hike in wooded areas. Trees are stationary, yet their benefits to people and animals extend beyond the boundaries of the forest. There will always be a spatial mismatch (Bell 2005) in assimilating data from ecological and social systems, and we were likely to underbound (leave out areas that fit our conceptual understanding of these places) or overbound (include areas that didn’t really fit with what we want to study) our identification of actual rural forested communities.
The more we discussed how we were going to measure this idea we had, this concept that we knew existed in the “real world,” the more we began to struggle. We began to brainstorm terms that might capture the idea. Of course, being scientists, we wanted clever acronyms for those terms.
The suggestion came up to identify a placeholder, such as a name or object that would be easy to remember, to communicate effectively until we figured out exactly what we wanted to call this phenomenon. Voices got louder. The tone got more heated. Scribbles were madly filling up the white board. We were at an impasse. I sat back, observing some of the chaos.
“George!”, I said, “Why don’t we just call it George!”
A dozen heads snapped around to look quizzically at me. I had just thought about the Abominable Snowman in Bugs Bunny. When Bugs misses that left turn at Albuquerque and ends up in the Himalayas (neocolonial problems with this representation aside), the Snowman picks him up as a pet, and says, “My own little bunny rabbit, I will hug him and squeeze him, and name him George.”
So we agreed to use the placeholder George, to stand in for this concept that we could all imagine but could not agree how to measure, so that we could move on to develop our larger conceptual framework, and all the factors we wanted to consider (Morzillo et al. 2015, van Berkel et al. 2018): roads, governing institutions, ecological conditions, political economy, etc. We got a lot of mileage out of simply invoking the name George, over and over and over.
Lesson 2: Know when to “lump” and when to “split”
Ultimately, we realized that our big group had substantive tensions in focus. The ecologists, broadly, were most interested in forest resources in situ, focusing more directly on the people, and the houses, that were in the midst of or on the edge of those forests. The social scientists on the other hand, were most interested in the function of the community rather than individual plots of land: how were local institutions navigating big structural economic changes, and how were community strategies for development differentially experienced by residents?
Eventually, we broke the puzzle into two pieces – understanding the extent, spatial patterns and regional variations across these forested regions (van Berkel et al. 2018), and hypothesizing how differing local characteristics correspond with trajectories of change (decline, tourism, new types of production) (Morzillo et al. 2015).
There is a literature about “lumpers vs. splitters” (Endersby 2009); i.e., whether scientists tend to be more comfortable with broad generalization, or whether more specificity, and more categories or exemplars, are necessary to capture finer-scale variation. I have been in several interdisciplinary teams in my 20-year career, and have experienced this phenomenon multiple times. What’s most interesting to me as a researcher and teacher is that no perfect theory and no perfect dataset can ever solve this conundrum. It is fundamentally a social one, that requires interpersonal communication and understanding. And, like all good inquiry, it depends on the research question.
Lesson 3: Specialize as needed
Ultimately, our SESYNC group, which was assembled from several existing projects, tag-teamed the two foci: alternatively, looking at settlements within forested pixels (van Berkel et al. 2018), versus forests within 100 miles of communities (Munroe 2019).
A somewhat smaller plucky band: the rural sociologist, the human dimensions expert, the two most open-minded economists I have ever known, and myself, have continued to forge ahead. This subset, when we received funding from the USDA National Institute of Food and Agriculture, continued to refine our definition of rural forested communities, a region we now refer to as Eunice.
Figure 2: The (smaller) interdisciplinary team in Logan, Ohio. Photo Credit: Kymber Anderson
What were the lessons of this experience? Well, anyone can claim to be a quarterback on Monday morning, so take these as intended. What I personally learned is the following:
- Don’t shy away from tough conversations. It’s not impolite to disagree about concepts or data.
- Focus on what you can agree on, and what you can’t.
- Be willing to change, follow the inspiration of the moment.
- Be generous ex post (you’ll note we were generous with coauthorship).
In the endnotes of our first published paper, we have an acknowledgement to “George.”
I’d like to thank my close group of collaborators, Kathleen Bell, Chris Colocousis, Mindy Crandall, Anita Morzillo and the rest of the SEYSNC team, past and present. All errors herein are my own.
Professor & Chair
Department of Geography
The Ohio State University
- Bell, K. P. (2005). Spatial analysis and applied land use research. Land Use Problems and Conflicts, 118.
- Bell, Kathleen P., Mindy Crandall, Darla K. Munroe, Anita Morzillo, & Chris Colocousis (2019). Rural forested landscapes, economic change, and rural development. 2019 North American Meetings of RSAI, Pittsburgh, PA. 15 November.
- Endersby, J. (2009). Lumpers and splitters: Darwin, Hooker, and the search for order. science, 326(5959), 1496-1499. DOI: 10.1126/science.1165915
- Fry, J. A., Xian, G., Jin, S. M., Dewitz, J. A., Homer, C. G., Yang, L. M., … & Wickham, J. D. (2011). Completion of the 2006 national land cover database for the conterminous United States. PE&RS, Photogrammetric Engineering & Remote Sensing, 77(9), 858-864.
- Munroe, D. K., Crandall, M. S., Colocousis, C., Bell, K. P., & Morzillo, A. T. (2019). Reciprocal relationships between forest management and regional landscape structures: applying concepts from land system science to private forest management. Journal of Land Use Science, 14(2), 155-172. https://doi.org/10.1080/1747423X.2019.1607914
- Morzillo, A. T., Colocousis, C. R., Munroe, D. K., Bell, K. P., Martinuzzi, S., Van Berkel, D. B., … & McGill, B. (2015). “Communities in the middle”: Interactions between drivers of change and place-based characteristics in rural forest-based communities. Journal of Rural Studies, 42, 79-90. https://doi.org/10.1016/j.jrurstud.2015.09.007
- SEDAC, & US census Bureau (2000). Census block data on poverty, housing stock, education and key demographic. Retrieved from: http://sedac.ciesin.columbia.edu/.
- Van Berkel, D. B., Rayfield, B., Martinuzzi, S., Lechowicz, M. J., White, E., Bell, K. P., … & Parmentier, B. (2018). Recognizing the ‘sparsely settled forest’: Multi-decade socioecological change dynamics and community exemplars. Landscape and Urban Planning, 170, 177-186. https://doi.org/10.1016/j.landurbplan.2017.10.009