Dynamic factor analysis of long-term growth trends of the intertidal seagrass Thalassia hemprichii in southern Taiwan

We examined environmental factors which are most responsible for the 8-year temporal dynamics of the intertidal seagrass Thalassia hemprichii in southern Taiwan. A dynamic factor analysis (DFA), a dimension-reduction technique, was applied to identify common trends in a multivariate time series and the relationships between this series and interacting environmental variables. The results of dynamic factor models (DFMs) showed that the leaf growth rate of the seagrass was mainly influenced by salinity (Sal), tidal range (TR), turbidity (K), and a common trend representing an unexplained variability in the observed time series. Sal was the primary variable that explained the temporal dynamics of the leaf growth rate compared to TR and K. K and TR had larger influences on the leaf growth rate in low- than in high-elevation beds. In addition to K, TR, and Sal, UV-B radiation (UV-B), sediment depth (SD), and a common trend accounted for long-term temporal variations of the above-ground biomass. Thus, K, TR, Sal, UV-B, and SD are the predominant environmental variables that described temporal growth variations of the intertidal seagrass T. hemprichii in southern Taiwan. In addition to environmental variables, human activities may be contributing to negative impacts on the seagrass beds; this human interference may have been responsible for the unexplained common trend in the DFMs. Due to successfully applying the DFA to analyze complicated ecological and environmental data in this study, important environmental variables and impacts of human activities along the coast should be taken into account when managing a coastal environment for the conservation of intertidal seagrass beds.