Near-surface remote phenology and the leaf exchange patterns along a gradient of seasonality

Author(s)

Bruna de Costa Alberton

• Document type: Doctoral Thesis
• Press: Rio Claro. 2018-09-13.
• Institution: Universidade Estadual Paulista (Unesp). Instituto de Biociências. Rio Claro
• Defense date: 2018-08-03
• Advisor: Leonor Patrícia Cerdeira Morellato; Ricardo da Silva Torres

Abstract

Plant phenology is a traditional science focused on monitoring, understanding, and predicting recurrent life cycles events, which are mainly related to climate Leaf development stages are essential plant phenophases for the better understanding of ecosystems processes such as carbon and water fluxes, regulation of productivity, and nutrient cycling. Through the investigation of plant responses to climate and phenological shifts prediction, we can better forecast climatic change effects on vegetation dynamics and prevent loss of ecosystems functionalities. Aiming to become the phenological collection wider and more feasible worldwide, the seek for new technologies has stimulated several research centers of plant phenology monitoring. Here, we incorporated a new technology of field phenological observations using digital cameras for the monitoring of leaf exchanges in tropical vegetations. On this work, divided into four sections, we aimed to: (i) compile information about concepts and properties of the repeated photograph technic, create guidelines for the phenocams setup in tropical vegetation sites, and to provide key contributions of daily imagery monitoring on biological conservation; (ii) to stablish a monitoring of different seasonal vegetations, describing the phenological trends, and identify the environmental cues which are triggering the leaf flushing and senescence for each vegetation type; (iii) analyze the canopy greenness obtained from digital cameras in relation to gross primary productivity measurements, to better understand the role of leaf phenology controlling ecosystem productivity in the tropics; and (iv) present some results of phenocam image analysis research initiatives and tools devised in the context of e-Science collaborations and built in the framework of the e-phenology project and the e-phenology network of phenocams. The selected study sites belong to different seasonal biomes, which comprehends areas from caatinga, savanna grasslands, savanna woodlands and Atlantic rainforest. Temporal series representing foliar phenology were extracted from the data imagery of each vegetation site and were analyzed into the context of each section of this work. We demonstrated that the establishment of phenocam networking is a powerful tool for biological conservation through its capability of fine temporal resolution data associated with wide spatial monitoring coverage. Besides, phenocams applications can bring new information for management and restoration practices at several sites and environments and contribute for the education for conservation and citizen science initiatives. We observed that water and light were the most important predictors for the leaf phenological patterns across seasonal vegetation communities. Water-plant relationships were more important for the caatinga community, and light, through day-length seasonality, had more influence in the leafing patterns of the cerrado communities. Regarding the ecosystem, we demonstrated a novel approach to relate leaf phenology to seasonality of tropical ecosystems productivity. The phenological dynamics regarding the variability of species phenological signals, and how they are built in into each contrasting vegetation communities explains drivers of leaf phenology and productivity. A compilation of articles developed through the e-science collaboration, presented in Section 4, were of great importance for the generation of methods and analytical work, as well as for the results achievements in this thesis. This work will offer a new tool for the phenological monitoring in the tropics, and suggests next challenges to be addressed and the continuity of the e-phenology network and the spread of new cameras covering new vegetation types; the development of bottom-up studies, integrating on-the-ground observations, cameras, drones, and satellites, inter-comparing them and placing camera-derived phenology in its own scale, by understanding how much and what kind of information can be retrieved from ecosystems. (AU)

• FAPESP's process: 14/00215-0 - Remote phenology and leaf exchange patterns towards a sazonality gradient
• Grantee: Bruna de Costa Alberton
• Support Opportunities: Scholarships in Brazil - Doctorate