Modelling the change in tree ring 13C discrimination as a response to selection harvest in a drained peatland forest

Studies on physiological response of suppressed trees to selection harvest are scarce. Understanding how trees respond to changes in environmental factors following harvest is needed for continuous cover forestry that aims to optimize both environmental impacts and economical gain. The physiological response of the trees can be understood by measuring stable carbon isotope composition (δ¹³C) which records the changes in photosynthesis and water use of the tree. The processes that determine the response can be further elaborated by comparing the measured isotopic signal to process-level model simulations.

We studied the response of Norway spruce (Picea abies) trees to selection harvesting on a fertile drained peatland forest located in southern Finland. The studied area consisted of a control plot which was left intact and of harvested plot which was thinned in March 2016. We measured intra-annual δ¹³C from tree-rings covering the period from 2010 to 2020 at the Stable Isotope Laboratory of Luke (SILL) (Lehtonen et al., accepted). The measured δ¹³C was compared to modelled ¹³C discrimination (Δ¹³C) simulated with a vertically resolved ecosystem model describing tree photosynthesis (Launiainen et al., 2015).

The δ¹³C measurements showed that after the harvest Δ¹³C decreased already on the following growing season. The overall decrease was ca. 3.3 ‰ on average between pre- and post-harvest periods. The decrease was caused by both changes in CO₂ assimilation of the spruce trees and differences in meteorological conditions between pre- and post-harvest years. We simulated Δ¹³C with three different models with increasing number of fractionation processes considered. All three models predicted that as a response to harvest the Δ¹³C would decrease, however, none of the models could replicate the observed 3.3‰ drop in Δ¹³C. The most complex Δ¹³C model that included ¹³C fractionation in mitochondrial and photorespiration as well as transport of CO₂ from stomata to mesophyll was the closest to the measurements.

The vertically resolved model allowed us to estimate that the changes in photosynthetically active radiation, relative humidity and needle temperature following the harvest contributed the most to the observed decrease in Δ¹³C. Further, model sensitivity analysis showed that the modelled Δ¹³C is the most sensitive to g₁ parameter and mesophyll conductance. The g₁ parameter is related to calculation of stomatal conductance (Launiainen et al., 2015; Medlyn et al., 2011). By tuning the g₁ parameter and mesophyll conductance we were able to bring the modelled Δ¹³C closer to the observations.

References

Launiainen et al., Ecological Modelling, 312, 385-405, 2015.
Lehtonen et al., Forest Ecology and Management, accepted.
Medlyn et. al., Global Change Biology 17, 2134–2144, 2011