Phenology is an integrative environmental science which encompasses biometeorology, ecology, and evolutionary biology. Monitoring vegetation phenology thus helps to provide a reference framework to track vegetation dynamics related to disturbances and stress events such as drought, fire, spring frost, land use changes, climate variations, etc. Phenological studies can be performed both through small scale and large-scale investigations: the former by means of ground-based and proximal sensing studies, the latter using remotely-sensed observations. Both proximal and remote sensing techniques are based on deriving vegetation indices (VIs), e.g. Normalized Difference Vegetation Index (NDVI), Leaf Area Index (LAI), fraction of Absorbed Photosynthetically Active Radiation (fAPAR), etc., from spectral sensors. The evolution of such VIs through time exhibits a strong correlation with the typical vegetation growth stages and provides a measurement of plant phenological characteristics that are independent of taxonomic or phylogenetic linkages. This has created opportunities to expand the concept of vegetation functional types, and to investigate plant functional properties from completely new perspectives. The overall goal of this project is therefore to gain a better understanding of plant ecophysiological dynamics by measuring and monitoring the phenological responses at both leaf- and landscape-scale of different plant species through proximal and remote sensing techniques. The identification of 'plant phenological types' represents a novel way of thinking about plant categories. The major innovation of the proposed approach is the ability to go beyond the structural characteristics of plants to look instead at their functional aspects in terms of e.g. understanding of the converging climate adaptation of different species, localization of the ecological niche at large scale, or recognition of phenological responses to environmental stress.
The phenological timing of plant communities in terms of growth and reproduction is driven by synchronizing (e.g. temperature, rain, frost, drought risks, topography, latitude) and asynchronizing (e.g. resource competition, seed dispersal, pollination) factors (Wheelwright 1985; Primack 1985).
The synchronizing factors tend to homogenize the phenological behavior of different species, such that plant communities can be identified by a characteristic phenological pattern. Macroecological investigations have shown that similar phenological responses characterize species belonging to similar ecoregions (Thuiller et al. 2004; Chuine 2010) due to their plastic response to some environmental conditions such as temperature, water availability or photoperiod (Chuine 2010). The working hypothesis addressed by this project is that large-scale environmental synchronizing factors homogenize the seasonal timing of forest communities notwithstanding the fine-scale species heterogeneity. Hence, phenology represents a key trait capable to discriminate plant functional types with similar responses to similar environmental conditions.
In this view, the major innovation and potential of the proposed approach is the ability to go beyond the structural species-specific characteristics of forests, looking instead at their functional aspects.
In traditional approaches, the overall functioning of the different plant communities is generally overlooked. To the contrary, the phenological grouping of species into phenological clusters at multiple scales, from individuals to communities, represents a process-based approach aimed at identifying groups of species with similar timing. Being characterized by the same phenological moments, such groups are assumed to be driven by the same biophysical variables.
The identification of 'phenological types' thus provides valuable information on the ecological functioning of forests with direct consequences, for example, in terms of converging climate adaptation of different forest species, carbon cycling, localization of the species' ecological niche at different scales, recognition of the species' phenological responses to environmental stress, development of ad hoc practices to fight forest fires and to manage fuel load and flammability, or for building an ecological framework for management strategies.
In this view, remote sensing data provide an ideal tool to perform multiscale vegetation phenology since they consistently measure vegetation processes and functions in time and space (Wessels et al. 2009). Given the huge amount of remotely sensed data, effective computing strategies are necessary to exploit the phenological information provided by long-term time series and to reduce data redundancy and processing complexity (Siachalou et al. 2015). In the era of Big Data, this project thus promotes a new research perspective for phenological information and offers an example on how to combine the potential of coarse-scale multispectral products and fine-scale hyperspectral data into a coherent ecological framework.
References
Chuine, I. 2010. Why does phenology drive species distribution? Philos Trans Roy Soc London B Biol Sci 365, 3149-3160
Primack, R.B. 1985. Patterns of Flowering Phenology in Communities, Populations, Individuals, and Single Flowers. In: White, J. (Ed.) The Population Structure of Vegetation. Springer pp. 571-593.
Siachalou, S. et al. 2015. A Hidden Markov Models Approach for Crop Classification: Linking Crop Phenology to Time Series of Multi-Sensor Remote Sensing Data. Remote Sensing 7, 3633-3650
Thuiller, W. et al. 2004. Effects of restricting environmental range of data to project current and future species distributions. Ecography 27, 165-172
Wheelwright, N.T. 1985. Competition for dispersers, and the timing of flowering and fruiting in a guild of tropical trees. Oikos 44, 465-477