Phenological indicators of vegetation recovery in restored wetland

Landscape phenology is increasingly used to measure the impact of major disturbances on vegetation productivity and functions. However, few studies have used phenology as a tool to monitor wetland recovery following restoration. As plants show a rapid phenological response to abiotic conditions, studying long-term variation in site phenology could help assess restoration success or detect unexpected fluctuations in vegetation productivity. To provide a low-cost approach to measuring restoration progress, we used free data from the Landsat and MODIS archives to characterize long-term phenological trends in 20 restored wetlands of California, USA. By tracking phenological changes across a 20-year period, we sought to identify a set of phenological metrics most responsive to time and restoration treatments. Comparing phenological metrics derived from the Landsat (16-day revisit) and MODIS (daily revisit) archives enabled us to measure the sensitivity of these phenological metrics to data frequency, scale, and quality. Preliminary results indicate that the length of growing season and the timing of the 50^th percentile and maximum greenness show a rapid response to restoration treatments and a greater consistency between the Landsat and MODIS datasets. Substantial variability among the phenological responses of our study sites suggest that landscape context modulates the time needed to reach phenological stability. Well-connected sites seem to benefit from an increased availability of propagules enabling them to reach peak greenness and extend their growing season more rapidly. These results demonstrate the potential of phenological analyses to measure restoration progress and detect landscape factors promoting wetland recovery. A thorough understanding of wetland phenology is key to quantifying ecosystem functions including carbon sequestration and habitat provisioning.