Open Collections

Liu, Wanyi

University of British Columbia

Forest thinning is a method to reduce competition in overstocked young forests. While the short-term effect of thinning is often studied by field thinning experiments, the long-term effect of thinning is rarely studied and remains unclear, particularly in the context of climate change. In this study, we used data from field experiment to train two Artificial Intelligence models: artificial neural network (ANN) and deep neural network (DNN) for assessing the long term effect of thinning on transpiration and water use efficiency (WUE) in young lodgepole pine forests at both tree and stand levels. We showed that DNN is more accurate than ANN in terms of model performance measures. We then used the trained DNN model to predict the transpiration and WUE under different thinning and climate change scenarios. The simulations showed that thinning has a long-lasting effect on increasing tree-level transpiration and both tree- and stand-level WUE but decreasing stand-level transpiration. Thinning effect decreases as thinning intensity decreases and trees age. Climate change reduces both tree- and stand-level WUE. Thinning mitigates the impact of climate change on stand-level transpiration. Based on the simulated long-term effects, we recommend that the thinning with the remaining of 2500–3500 stem/ha as suitable densities for a better balancing or trade-off between total carbon and water (stand transpiration) for young and overstocked lodgepole pine stands in the southern interior of British Columbia. However, selection of the final thinning densities must consider broader ecological processes and functions such as wood quality, biodiversity as well as their interactions with climate change.

Text

2023-02-23

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/83839

Earth and Environmental Sciences

University of British Columbia

UBCO

http://creativecommons.org/licenses/by-nc-nd/4.0/