Forest responses to future climate are highly uncertain, but critical for forecasting and managing for forest carbon dynamics. To improve ecological forecasts of forest response, we harness the strengths of two large ecological datasets: tree-ring time series data that provide annually resolved growth responses, and spatially extensive forest inventory (FIA) data. We use a Bayesian state space model to assimilate these two ecological data sets, and quantify the effects of precipitation, maximum temperature, tree size, stand density, site index and two-way interactions between these factors on tree growth. We implemented a two-stage approach to model Pinus ponderosa responses in Arizona. Stage 1 leverages tree-ring increment data and repeat diameter measurements of 515 trees to estimate effects on tree growth. Posterior parameter estimates from stage 1 were then used as priors in stage 2, where data were included from an additional 5,794 trees in the forest inventory that only have repeat bole diameter measurements.
Results/Conclusions
Precipitation has a strong positive effect on Pinus ponderosa growth in Arizona, leading to growth declines under drier future conditions. Maximum temperature does not have a strong direct effect on growth, but a positive interaction between temperature and precipitation drives decreased growth under hot and dry future conditions. Tree size, stand density, and site factors all have considerable direct effects on annual tree growth, and can modify climate responses, such that larger trees and trees with high site quality have greater growth increments, but high stand density reduces growth increments, particularly at high temperatures. Interactions between stand-level properties and climate sensitivity provide opportunities to manage for forests that optimize carbon storage and climate resilience. Fusing information from 5794 repeat diameter measurements reduces uncertainty about stand-level processes, and allows us to forecast annual growth increment in forest plots without tree ring data. Assimilating tree ring and forest inventory data can help inform current management, constrain uncertainties about the effects of climate change, and provides a framework for iterative ecological forecasts.