Abstract
Vegetation phenology is an integrative indicator of environmental change, and remotely–sensed data provide a powerful way to monitor land surface vegetation responses to climatic fluctuations across various spatiotemporal scales. In this study, we synthesize the local climate, mainly temperature and precipitation, and large-scale atmospheric anomalies, El Niño-Southern Oscillation (ENSO)-connected dynamics, on a vegetative surface in a subtropical mountainous island, the northwest Pacific of Taiwan. We used two decadal photosynthetically active vegetation cover (PV) data (2001–2020) from Moderate Resolution Imaging Spectroradiometer (MODIS) reflectance data to portray vegetation dynamics at monthly, seasonal, and annual scales. Results show that PV is positively related to both temperature and precipitation at a monthly timescale across various land cover types, and the log-linear with one-month lagged of precipitation reveals the accumulation of seasonal rainfall having a significant effect on vegetation growth. Using TIMESAT, three annual phenological metrics, SOS (start of growing season), EOS (end of growing season), and LOS (length of growing season), have been derived from PV time series and been related to seasonal rainfall. The delayed SOS was manifestly influenced by a spring drought, <40 mm during February–March. The later SOS led to a ramification on following late EOS, shorter LOS, and reduction of annual NPP. Nevertheless, the summer rainfall (August–October) and EOS had no significant effects on vegetation growth owing to abundant rainfall. Therefore, the SOS associated with spring rainfall, instead of EOS, played an advantageous role in regulating vegetation development in this subtropical island. The PCA (principal component analysis) was applied for PV time series and explored the spatiotemporal patterns connected to local climate and climatic fluctuations for entire Taiwan, North Taiwan, and South Taiwan. The first two components, PC1 and PC2, explained most of data variance (94–95%) linked to temporal dynamics of land cover (r > 0.90) which was also regulated by local climate. While the subtle signals of PC3 and PC4 explained 0.1–0.4% of the data variance, related to regional drought (r = 0.35–0.40) especially in central and southwest Taiwan and ENSO-associated rainfall variation (r = −0.40–−0.37). Through synthesizing the relationships between vegetation dynamics and climate based on multiple timescales, there will be a comprehensive picture of vegetation growth and its cascading effects on ecosystem productivity.
Highlights
The dynamics of vegetation growth or phenology, which plays a crucial role in regulating the water cycle, carbon cycle, energy balance, biomass accumulation, and productivity largely depends on key climatic factors, temperature, precipitation, and radiation [1,2,3]
The monthly photosynthetically active vegetation cover (PV) had a significant log-linear relationship with one-month lagged of precipitation for all vegetation/land cover types, in which the higher PV was associated with greater rainfall within the range of 0–100 mm of precipitation, above 100 mm of precipitation the relationship leveled off (Figure 3c)
Taiwan using PV calculated from Moderate Resolution Imaging Spectroradiometer (MODIS) reflectance data to delineate vegetation growth at monthly, seasonal, and annual scales
Summary
The dynamics of vegetation growth or phenology, (i.e., the timing and duration of vegetation activity across a year) which plays a crucial role in regulating the water cycle, carbon cycle, energy balance, biomass accumulation, and productivity largely depends on key climatic factors, temperature, precipitation, and radiation [1,2,3]. Many studies have shown that temperature is the dominant control of plant growth in high latitudes and cold regions, precipitation is the dominant factor in arid and semiarid areas, whereas radiation is key to plant growth in rainforest [4,5,6,7]. More efforts are required to provide a better understanding of vegetation-climate dynamics for isolated islands in tropical and subtropical regions, where there are hotspots of biological diversity that are more susceptible due to their unique environment and limited area [12].
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call