Novel Tools For Community Based Assessment And Monitoring Of Forest Phenology Response To Environmental Variability And Change In The Tropics

Tropical ecosystems provide a variety of services and resources for global well-being. However, landscapes within this biome are undergoing rapid changes mostly due to anthropogenic activities. These systems are further impacted by current, extreme fluctuations in climate, including climate change and El Niño Southern Oscillation (ENSO). With predicted future increase in strength and occurrence of these climate oscillations, these landscapes and their plant communities will be impacted and changed. There is a gap in research and knowledge among tropical Indonesian landscapes in regards to the effects of global climate change at the community level. Therefore, there is an urgent need to understand how tropical plant communities respond to current environmental conditions in an effort to predict how future changes will impact these regions. The overarching focus of this Dissertation research is to determine the usefulness of digital repeat photography in monitoring tropical Indonesian plant communities and to understand how these landscapes are behaving in the face of climate change. Specifically, this study aims to: Assess the ability to remotely monitor tropical plant communities at the landscape and species level through the use of digital repeat photography. Evaluate the relationship between satellite-derived vegetation indices and ground-based calculated indices of greening. Investigate a novel approach to study the phenology, growth and expansion of a tropical mangrove tree species. Our study was carried out in a variety of landscapes in Indonesia, specifically 6 sites ranging from secondary forest with primary patches to monoculture plantations to an urban landscape in East Kalimantan, Borneo, and one mangrove forest on Karimunjawa Island, Central Java. Phenocams were established at varying times as field and research logistics provided between mid-2012 to mid-2013. To understand plant community composition at each site, ecological site descriptions using point-quarter transects were conducted at all sites except the monoculture plantations. Sites were monitored for approximately three years, which included Non-ENSO, Weak ENSO, and Very Strong ENSO years, but due to data gaps and technical issues, not all sites provided sufficient data for full analysis. The use of phenocams among a variety of plant communities provided the opportunity to convey canopy phenology in response to extreme changes in temperature and precipitation induced by ENSO events. Phenocam imagery spectral properties were analyzed to derive greening indices (the green excess index (GEI) and the green chromatic coordinate (GCC)) as a measure of plant phenology, which were compared to determine which index serves as a better metric within these landscapes. The GCC was found to be a better measure of the two, but it was also found that these plant communities express a lagged response to environmental changes of 14-21 days (site dependent) during Non-ENSO years, and an advanced zero day lag for Very Strong ENSO years. As a result of the warmer temperatures and decreased rainfall during the ENSO periods, the GCC decreased at all sites. The most impacted site was shown to be the secondary forest with primary patches, as this site may be less resistant to extreme changes in temperature and precipitation. Phenocams were also used to detect changes at the species level, through which we were able to identify key dominant species that may be driving greening seen at the landscape level. These results emphasize the effects of climate change on vegetation phenology and the importance of assessing species level response to better understand landscape dynamics. Changes and responses within plant communities effect the changes seen at landscape, regional and global levels. However, common methods used to capture these processes, such as satellite remote sensing, may not always be effective given inherent conditions of tropical environments, including abundant rain, clouds, and fog. Therefore to evaluate the effectiveness of satellite remote sensing to convey changes at the landscape level, we compared common satellite derived vegetation indices (normalized difference vegetation index (NDVI); the enhanced vegetation index (EVI)), which provide a measure of land surface phenology, to phenocam vegetation indices, the GEI and the GCC, which provided a measure of landscape phenology. For seven sites ranging from primary forest patches to secondary forests to plantations, it was shown that satellite-derived NDVI and EVI were not significantly related to landscape level GCC and GEI indices. When divided into datasets encompassing rainy and dry seasons, significant correlations between EVI and GCC were revealed during the dry season, suggesting that of the satellite-derived vegetation indices, EVI was better at conveying phenology during the dry season. Finally, when compared among ENSO events, as a result of warmer temperatures and decreased rainfall produced by ENSO events, the GCC significantly declined, but this anomaly was not conveyed in any of the other vegetation indices. These findings demonstrate that satellite remote sensing in these tropical landscapes is not as effective as ground-based monitoring, as these complex, mixed systems inherently contain variables that are challenging for remote sensing platforms. From September 2012 to May 2015, a phenocam monitored the seaward edge of a protected mangrove forest. Calculated GCC revealed seasonal greening patterns of a mangrove species, Rhizophoa apiculata, and an overall increase in the GCC, suggesting mangrove expansion. In comparing temperature and precipitation effects, it was found that this particular mangrove species had a greening optima at temperatures between 28°C and 28.5°C, and greening and canopy development response lag time of 10 weeks in response to precipitation. Tree saplings were monitored and showed to grow by 50%, mostly during a three month period during the rainy season. The establishment of new saplings was recorded, illustrating the seaward expansion of this mangrove community. Digital repeat photography appeared to be highly advantageous and an invaluable tool in assessing and monitoring mangrove phenology, growth and expansion. Combined these findings demonstrate the usefulness and benefits of using digital repeat photography in Indonesian landscapes. They also illustrate the changes that are currently undergoing within plant communities in response to extreme climate events, specifically ENSO occurrences, which have major implications for future land use/ land cover and climate change