The goal of this project is to take a model used for numerous biomes and apply it to the growing human-dominated exurban landscape. One challenge is that most ecosystem process models are developed for homogeneous wildland systems and in the exurban landscape there are heterogeneous mixtures of vegetation types that have been heavily modified by humans. For this analysis, we are modifying the ecosystem process model Biome-BGC so that it can better simulate ecosystem processes in landscapes that that have been heavily modified by humans.
We collected field measurements, along with vegetation and soil samples in a 10 county study region of southeastern Michigan to parameterize the model. We have determined ecophysiological parameters for different plant functional types (PFTs) from laboratory analysis of carbon and nitrogen content of leaves, stems, roots, litter and soil. Initial site conditions that represent the transition from agriculture to residential development were determined by grouping data by PFT and graphing carbon pools (above ground and soil) against time since development. Using a linear trendline, the value of carbon calculated at time zero will be used as the initial value for each carbon pool. Based on the C:N ratios we established, the initial nitrogen pools were determined. Once the model was parameterized simulations were run to see variations in net primary productivity (NPP).
Results/Conclusions
We found that an important element of this approach is to decompose residential neighborhoods and parcels into cells with different PFTs because the vegetation is very heterogeneous. Two of the most common PFTs in the landscape are deciduous trees and turfgrass. Ecophysiological parameters and initial site conditions were determined for each of these vegetation types and model simulations were run. Model simulations of deciduous trees in the landscape show NPP ranges from 411.6 to 796 gC m-2 yr-1. Simulations of turfgrass show NPP ranges from 147.6 to 253 gC m-2 yr-1. Ongoing analysis includes additional plant functional types as well as incorporation of human management behaviors in our simulations.