Behind the paper: changing forests on the US East Coast

We speak to Marcelo Ardón and colleagues about their recent PLOS Climate publication, “Coastal carbon sentinels: A decade of forest change along the eastern shore of the US signals complex climate change dynamics”.

What led you to decide on this research question?

There has been a lot of research interest – and interest from the public – about the creation of “ghost forests” along the coasts of the US. Ghost forests are upland or wetland forests that are changing to marsh or open water due to sea level rise, saltwater intrusion, and more frequent and intense storms. The research to date has either been site-based studies that focus on a single site or a couple of sites over time, or remote sensing studies that cover broad areas and rely on satellite images, or other data collected remotely. There has not been a study that has used consistent methods to measure the structure and ecosystem processes (e.g. growth or tree mortality) of forests across large scale areas. That is exactly the goal of the US Department of Agriculture Forest Service program called the Forest Inventory and Analysis (FIA) program. This program uses statistical methods to distribute plots that are measured using nationally standardized techniques, in order to provide estimates that can be extrapolated from the plot scale to larger spatial areas. This program has been used to assess the state and health of forests across the entire US, but to date had not been used to look at state of coastal forests specifically. We were also interested in looking at standing and downed dead trees, which is challenging to do with remote sensing.

How did you go about designing your study?

We wanted to use field derived data that were generated using consistent methods, timely data releases, and covered a large geographic range, which led us to the FIA surveys data covering the southeastern US. Elevation above sea level is a key indicator of coastal forest vulnerability to saltwater intrusion and sea level rise (SWISLR). We used this information to divide the dataset into low and high elevation plots that are five and 30 – 50m above sea level, respectively. Standard FIA methodologies were used to estimate county-level forest attributes from the plot network to examine patterns of both live and dead trees. We wanted this study to go beyond simple counts of live and dead trees by attempting to attribute differences in forest structural and carbon dynamics to environmental variables that can be affected by climate change. Climate data from GridMet (i.e. temperature and precipitation) and rates of sea level rise and frequency of tropical storms from National Oceanic and Atmospheric Administration (NOAA) were critical pieces to a complex story of rapid, ongoing coastal change.

Did you encounter any challenges in collecting or interpreting your data?

The first challenge was deriving population level estimates from the FIA national network, as the FIA database is large and complex to address a variety of social interests across the great diversity of forests in the US. Thankfully our co-authors, C. Woodall and K. Potter, are experts in the FIA database and were able to query the national database  to derive a study dataset optimal to meet study objectives. To get the storm data, E. White was able to download historical storm data from the NOAA website. It took some time to get all the data together in a way that we could analyze it.

What struck you most about your results? What are the key messages and who do you hope might benefit from these new insights?

The first surprising result was that forests had gained area in both low and mid elevation counties. Since we had started the research knowing that remote sensing studies had found forest loss, we were surprised to see forest area gain in this dataset. In the paper we discuss some of the reasons why the FIA data might provide different results than remote sensing studies. However, our results fit into broader land use patterns related to agricultural land abandonment. We also found that tree mortality was higher in low elevation counties, and higher in counties that have experience higher rates of sea level rise, in agreement with our initial predictions. We were also surprised to see that more frequent storms led to increased growth in both low and mid elevation counties. We think this is related to storms opening space for younger trees to grow more. We were also surprised to see that the amount of carbon stored in standing and dead trees in these forests is very large, and we understand very little about what factors control the permanence of those carbon stocks.

The results from this study will inform land managers, foresters, and people interested in climate change effects. Sea level rise and magnitude and frequency of storms are going to continue to exacerbate due to human-accelerated climate change. This study provides insights into how forest structure and function have already been affected by these ongoing changes and can help inform decisions for these rapidly changing landscapes.

What further research would you like to see in this area?

There are many questions that can be explored with FIA data and other remote sensing products. We think further studies looking at the standing and downed dead wood will be very important. Also looking at the impacts of specific storms on forest structure and function will be very important. Another key question is land use change dynamics in these increasingly disturbed coastal locations. In the near term, will forests gain as agricultural and developed land uses are inundated and abandoned? Will the structure and species composition of these coastal forests change, and will they function as carbon sources or carbon sinks?

What made you choose PLOS Climate as a venue for your article?

We wanted the results to reach a broad climate audience, to be Open Access, and to come out rapidly. We found it very easy to work with the editorial staff and reviewers at PLOS Climate.

Read the full article in PLOS Climate.

Ready to submit your own work for peer review?