Spatiotemporal patterns and drivers of Spartina alterniflora phenology and belowground biomass

Detailed information on vegetation phenology in salt marshes is useful for understanding patterns in plant production and carbon sequestration. I used high-frequency PhenoCam imagery to characterize the phenophases of Spartina alterniflora, the dominant macrophyte in eastern U.S. salt marshes. These observations showed differences of > 1 month in the timing of spring green-up in plants located only meters apart. I found that green-up could be explained by elevation-driven variation in soil temperature, and used this insight to develop the first empirical spring green-up model for S. alterniflora. Next, I built on this work by creating an algorithm that estimates S. alterniflora belowground biomass from aboveground proxies. This analytical tool is important because the belowground production of salt marsh plants contributes to coastal marsh resiliency by stabilizing soils and maintaining vertical marsh elevation. Belowground production also contributes to soil organic matter, helping to store “blue carbon.” However, estimates of belowground biomass are difficult to obtain because field data are laborious to collect. The algorithm that I developed used Random Forests to combine information on phenology and elevation with leaf area index, leaf chlorophyll concentration, and land surface temperature, each of which can contribute to variation in root:shoot ratios and biomass allocation differences. The model explained 77% of belowground biomass variance and had a Root Mean Squared Error (RMSE) of 267 g m-2 across testing data