Forest dynamics in relation to meteorology and soil in the Gulf Coast of Mexico

Forest dynamics is complex, and the complexity could be a synthetic result of climate change. Specifically studying 11 forest type groups of the Gulf of Mexico coast region defined, we intended to explore and model the direct and indirect impacts of climate change on underlying forest dynamics. This study utilized normalized difference of vegetation index (NDVI) as a measurement indicator of forest dynamics, referring to the dynamics of canopy structure and phenology of forests, and for a given type of forests, seasonal and yearly NDVI values were applied to the quantification of its growth across the Gulf Coast. By utilizing geographically weighted regression (GWR) method, we related normalized difference vegetation index (NDVI) to precipitation, temperature, and silt and clay fractions in the soil. This study demonstrated an explanatory power of soil, besides the common macroclimate factors of precipitation, temperature, on explaining forest dynamics, which also revealed that the presence of spatiotemporal heterogeneity would affect model performance. Our results indicated that the model performance varied by forest type groups and seasons. The meteorology-soil model presented the best overall fit performance for White/Red/Jack Pine forests concerning R² (0.952), adjusted R² (0.905), Akaike information criterion (AIC, -1100) and residual sum of squares (RSS, 0.053) values. The comparative analysis of model performance also indicated that the meteorology-soil model has the best fit of data in summer. This study advanced the understanding of forests dynamics under conditions of climate change by highlighting the significance of soil, which is a significant confounding variable influencing forest activities but is often missed in forest-climate dynamics analysis.