Although specific controls are not well understood, the phenology of temperate forests is generally thought to be controlled by photoperiod and temperature, although recent research suggests that soil moisture may also be important. The phenological controls of forested wetlands have not been thoroughly studied, and may be more controlled by site hydrology than other forests. For this study, remotely sensed vegetation indices were used to investigate the hydrological controls on start of season timing at four bottomland hardwood sites in Louisiana, Arkansas, and Texas. A simple spring green-up model was used to determine the number of growing degree days for each site as a function of land surface temperature and photoperiod. The MODIS phenology product was used to define the start of season timing for each pixel. Timelapse cameras were installed at different topographic locations at two Louisiana sites during January 2017 to both validate satellite derived start of spring dates and further investigate how species composition may control this spatial pattern. In addition individual trees of some of the most abundant species (Carya aquatic, Celtis laevigata, Fraxinus pennsylvanica, Ulmus Americana, and Quercus lyrata) were observed throughout the spring of 2017 and 2018.
Results/Conclusions
Two of the sites showed a positive correlation between the amount of growing degree days required for spring leaf out and flood frequency as defined by potential natural vegetation classes. These sites both still receive significant surface flooding. For the other two sites, there was no apparent correlation between flood frequency and spring leaf out timing which may be explained by the lack of surface flooding at these locations due to river modifications. Preliminary results from the ground based observations indicate species typically found in higher flood frequency areas, Quercus lyrata and Carya aquatic, have later leaf out dates relative to Celtis laevigata, Fraxinus pennsylvanica, and Ulmus Americana. Phenology is correlated with hydrology however the exact mechanisms at play are still unknown. Ongoing data collection and analysis will focus on identifying uncertainty in the MODIS start of season timing estimates across gradients in flood frequency and ascertaining whether individuals of species typically found in frequently flooded areas maintain late leaf out behavior during dry years or when found in drier sites.