EDI Staging Environment
These reconstructions of forest composition in the northeastern US for the last 8,000 years help establish natural baselines, variability, and trajectories of forest dynamics before and during the emergence of intensive anthropogenic land use. These reconstructions are based on 1) the pollen–vegetation model (PVM) STEPPS, 2) a network of fossil and modern pollen data mostly drawn from the Neotoma Paleoecology Database (www.neotomadb.org), and 3) a vegetation calibration dataset based on a spatial statistical model of relative tree abundances from the Township Proprietor Survey (TPS) from the early EuroAmerican settlement period. STEPPS is a process-based Bayesian Hierarchical Model that is run in two stages: a parameterization stage based upon spatial data layers of forest composition and pollen assemblages, and a prediction stage based on fossil pollen assemblages. The statistical modeling of the TPS forest data and relative abundances is described by Paciorek et al. (2016, PLoS One, doi: 10.1371/journal.pone.0150087) and is available at EDI as PalEON Product msb.paleon.1, doi: 10.6073/pasta/8544e091b64db26fdbbbafd0699fa4f9. The parameterization of STEPPS and comparison to REVEALS, a different widely used PVM, is described by Trachsel et al. (2020, Quaternary Research, doi:10.1017/qua.2019.81). Both PVMs predict the observed macroscale patterns of vegetation composition in the NEUS; however, reconstructions of minor taxa are less accurate and predictions for some taxa differ between PVMs. These differences can be attributed to intermodel differences in structure and parameter estimates. STEPPS parameter estimates are similar between the UMW and NEUS, suggesting that STEPPS parameter estimates are transferable between floristically similar regions and scales. The parameterized STEPPS model was then run for a network of fossil pollen records from the Neotoma Paleoecology Database to produce posterior estimates of forest composition at 100-year intervals for the last 8,000 years. The work from this prediction stage was not published at the time of deposition of these records with EDI. Hence, at this time, this EDI data deposition is the primary citation endpoint. The prediction methods generally follow those described by Dawson et al. (2019, Ecology, doi: 10.1002/ecy.2856). This material is based upon work supported by the National Science Foundation under grants #DEB-1241874, 1241868, 1241870, 1241851, 1241891, 1241846, 1241856, 1241930.
EDI is a collaboration between the University of New Mexico and the University of Wisconsin – Madison, Center for Limnology:
