I studied the lifecycle carbon balance (BioC-LC) of tropical forest management in Costa Rica. Until now, existing findings supported the idea that tropical logging leads to higher carbon emissions but no carbon balance analysis for these ecosystems had been done using a lifecycle approach. To quantify the effect of logging and compare it against forest ecosystem carbon balances, I used one hectare as the functional unit and defined the system’s temporal boundary as one rotation period. I show that by including all lifecycle processes, technospheric storage in combination with forest regrowth results in additional storage of carbon in the system (i.e. -2.19 Mg C ha-1 over a 15-year period with a 95% CI of -5.26 to 1.86). Just for comparison with the other results in this thesis, expressed as CO2-eq this result is equal to -8.00 Mg CO2-eq ha-1 over a 15-year period. Probabilities of a system that is a source of carbon exist, as higher harvesting intensities leading to high logging damage, insufficient recovery time, or high wood allocations into short-term uses can shift this balance. However, short-term uses increase storage in solid waste disposal sites (SWDS), and it is the combined effect from technospheric reservoirs that is important for carbon storage. Using a sensitivity analysis, I found that small changes in half-lives do not have an important effect on the stock and that only large changes such as re-allocating products from short to long-term products have substantial effects on total storage. Based on these findings we highlighted the climate mitigation opportunities of forest management for timber extend beyond the forest and that measures should be considered throughout the processes of wood transformation, use and disposal.
Then, I developed a detailed harvested wood product carbon inventory for Costa Rica. I followed IPCC Guidelines for National GHG Inventories, used country specific data and a material flow analysis, corresponding to a Tier 2 accounting level according to these Guidelines. Harvest data collected for this study is the currently best available for Costa Rica describing the evolution of wood production during the last 30 years and Chapter 3 merely scratches the surface of lessons that can be extracted from this data set. Carbon storage at the national level in 2016 (the last year of the inventory) was -412 Gg CO2 (95% CI between -447.2 and -376.4). Most of this storage was found in SWDS (77%) and was partly a consequence of a high allocation of wood production into short-term products. Given that these allocation patterns were positively correlated with planted forests becoming the country’s main wood source, I asked what have been (or will be) the effects of changes in wood source and product allocation on the carbon stock of harvested wood products. Since plantation wood tends to have a lower quality (at least lower wood densities and carbon content) and half-lives are consistently reported as drivers of carbon storage, I hypothesized that the stock must be heading towards a steady state. However, despite significant decreases in half-life and carbon content, the stock seemed unaffected. Hence, the stock of wood products appears to be characterized by a strong inertia, due to the characteristics (i.e., half-life) of the material in the stock from pervious harvests. As a result of these inherited characteristics, changing stocks of wood products may take a long time. This likely implies that the contribution of this stock to climate mitigation is smaller than commonly believed. Physical limits characterize technospheric carbon storage, and prolonging lifespan may not extend these limits much further. Thus, it is mostly through increasing harvest levels and wood use that storage can be increased. In this Chapter, I highlighted that opportunities to increase storage through increased harvests or lifespan must come from the implementation of demand-side measures.
I assessed the lifecycle climate impact of wood from natural tropical forests in Costa Rica. This work fills a gap in the understanding of the effects of logging in the tropics, where few studies have been conducted and none of these included the combined effect of biogenic and fossil emissions in a cradle to grave analysis for one rotation. Results for the harvest of wood from a hectare of tropical forests in Costa Rica show a net balance of -4.41 Mg CO2-eq ha-1 over a 15-year period (95% CI of -13.12 to 10.96), indicating that under this timeframe the system has stored more carbon than what has been released through emissions. This result was verified by studying the effect of a shorter time horizon (i.e. 20-year global warming potential (GWP)) and by extending the temporal boundaries (i.e. from 15 to 100 years). Under a 100-year system boundary, emissions increase significantly to 1.90 Mg CO2-eq ha-1 over a 100-year period (95% CI of -10.55 to 18.28) but I argue that for this functional unit (i.e. ha) this timeframe is not an appropriate boundary since not all possible rotations are taken into consideration. This boundary is useful when the functional unit is a product, or as in this case a m3 of wood used for a specific product or co-product. Results for each individual wood product and co-product were also included, but for a 100-yr system boundary only mid and long-term products show a negative GHG balance due to carbon storage. Short-term products are specially affected by a change in boundary due to EoL methane emissions. Although these require almost no manufacturing, short-term products have the highest emissions per m3, i.e. 860 Mg CO2-eq ha-1 over a 100-year period (95% CI of -1.78 to 8.28). Because of the large proportion of short-term products these have a large effect on the results per hectare or multi-functional m3. I therefore highlighted that after the evaluation of all lifecycle processes, the largest opportunities to increase the mitigation potential of forest management in the tropics is likely through reduced impact logging techniques.
Finally, I integrate the results from all chapters and discuss their implications. I first address the trade-offs from using local empirical data on the uncertainty and variability of the system. I argue that the use of local data is beneficial as it leads to an overall reduction in uncertainty, a better conceptual understanding of the system, more accurate estimation methods and adds crucial information (variability) on the system. Downsides of the use of local data include that it requires adaptation of calculation methods, it increases the risk of calculation errors, it adds unnecessary noise in the calculation process, and this may hamper interpretation of results. I continue by discussing results within the context of national policies and forestry practices. This is followed by an estimation of the potential contribution of forest management in Costa Rica to climate mitigation. Based on my own results, I provide a simple scenario analysis of opportunities to increase mitigation through increased logging intensity and increased logging area. I found that by increasing the harvest area to the maximum potential yield, a contribution of -147.0 Gg CO2-eq yr-1 and the potential substitution of -218.4 Gg CO2-eq yr-1; results in a total mitigation potential of -365.4 Gg CO2-eq yr-1. Finally, I discuss the key question of whether productive tropical forests should be managed for climate mitigation. Since the main argument against forest management is that it leads to degradation, I discuss how the results from this study provide evidence that after considering all lifecycle processes this is not necessarily always true. Finally, I argue that the main contribution from this thesis and a lifecycle approach in general, is that reveals the unintended consequences of decisions to not manage forests (i.e. indirect land use change, changes in wood production and substitution of wood). If all of these are included in the scenario against which management is usually compared, then this would more clearly show the contribution of managing tropical forests for wood production as a mitigation strategy.