Impacts of forest biomass removal on soil nutrient status under climate change: a catchment-based modelling study for Finland

Abstract

The environmental impact of different forest harvesting scenarios on soil nutrient status and water chemistry under current and future (IPCC A2) climate was evaluated for a random sample of lake catchments (n = 1066) covering Finland. Biomass removal scenarios were derived from a management-oriented large-scale forest model based on data from national forest inventories. Forest ecosystem sustainability was assessed by evaluating soil base cation balances as well as temporal changes (2010–2050) in soil base saturation and lake water acid neutralising capacity, using a dynamic hydro-geochemical model. The harvesting scenarios had very different effects on biomass and element removal as well as soil and water quality; only harvesting of above-ground woody biomass (stem-only or stem-and-branches harvesting scenarios) was predicted to be sustainable, i.e. not depleting the soil base cation pools in the long term. The most intensive scenario—whole-tree harvesting (including the removal of stumps and roots)—doubled the removal of biomass, tripled the removal of base cations from the catchment soils, and increased nitrogen removal fourfold. Climate change was predicted to have a positive impact by increasing the future supply of base cations from weathering, thus compensating their removal by biomass harvesting. However, additional inputs of nitrogen and potassium will be required to ensure sustained forest growth under intensive biomass harvesting.

Appendices

Appendix A: Forest growth under climate change

Forest growth under climate change is obtained from the reference growth G ₀ as:

$$ G = G_{0} \cdot f_{temp} \cdot f_{resp} $$

(1)

i.e. the modified growth G is obtained by multiplying the (measured) reference growth with a temperature factor f _temp and a respiration factor f _resp. Both the temperature and the respiration factor are taken from the C-Fix model (Veroustraete et al. 2002). The temperature effect on growth is modelled according to:

$$ f_{temp} = {\frac{g(T)}{{g(T_{0} )}}}\quad {\text{with}}\quad g(T) = {\frac{{\exp \left( {C_{1} - {\tfrac{{\Updelta H_{a} }}{RT}}} \right)}}{{1 + \exp \left( {{\tfrac{{\Updelta ST - \Updelta H_{d} }}{RT}}} \right)}}} $$

(2)

where ΔH _a is the activation energy (52 750 J mol⁻¹), ΔH _d the deactivation energy (211 000 J mol⁻¹), ΔS the entropy of the denaturation equilibrium of CO₂ (704.98 J K⁻¹mol⁻¹), C ₁ = 21.77, and R = 8.314 J K⁻¹mol⁻¹ the universal gas constant. T is the temperature and T ₀ the reference temperature at which G ₀ is measured (both in K). The respiration factor is modelled as the ratio of autotrophic respiration at different temperatures:

$$ f_{resp} = {\frac{{1 - A_{d} (\theta )}}{{1 - A_{d} (\theta_{0} )}}}\quad {\text{with}}\quad A_{d} (\theta ) = \left( {7.825 - 1.145 \cdot \theta } \right)/100 $$

(3)

where θ and θ ₀ are the temperatures in degrees Celsius (θ = T–273.15).

Appendix B: Base cation weathering under climate change

The temperature-dependence of base cation (Ca²⁺ + Mg²⁺ + K⁺) weathering rates is modelled as:

$$ Bc_{w} (T) = Bc_{w} (T_{0} ) \cdot A(T_{0} ,T) $$

(4)

Since the function A describes the temperature-dependence of the underlying chemical reactions, it is assumed to have the form of an Arrhenius-factor (T and T ₀ in Kelvin):

$$ A(T_{0} ,T) = \exp \,\left( {{\frac{\Updelta H}{R}}\left( {{\frac{1}{{T_{0} }}} - \frac{1}{T}} \right)} \right) $$

(5)

For the activation energy divided by R we used ΔH/R = 3600 K. This is an average of values for different mineral groups given in Sverdrup and Warfvinge (1993), which is also used as a default value in European critical load calculations (UBA 2004). For T ₀ = 274.2 K (the median temperature at the 1066 catchments; see Table 1) and T = T ₀ + 3.8 (typical increase in Finland by 2050 under A2 climate change scenario; see Fig. 1) one obtains A = 1.197, i.e. an increase in weathering by about 20%. The sensitivity to the value of ΔH/R is small: for ΔH/R = 3200 and ΔH/R = 4000 (a more than 10% variation around 3600) one obtains A = 1.173 and 1.221, respectively, i.e. less than 3% change.