Water limitation can negate the effect of higher temperatures on forest carbon sequestration
- 123 Downloads
Climate change will bring about a consistent increase in temperatures. Annual precipitation rates are also expected to increase in boreal countries, but the seasonal distribution will be uneven, and several areas in the boreal zone will experience wetter winters and drier summers. This study uses the dynamic forest ecosystem model ForSAFE to estimate the combined effect of changes in temperature and precipitation on forest carbon stocks in Sweden. The model is used to simulate carbon stock changes in 544 productive forest sites from the Swedish National Forest Inventory. Forest carbon stocks under two alternative climate scenarios are compared to stocks under a hypothetical scenario of no climate change (baseline). Results show that lower water availability in the future can cause a significant reduction in tree carbon compared to a baseline scenario, particularly expressed in the southern and eastern parts of Sweden. In contrast, the north-western parts will experience an increase in tree carbon stocks. Results show also that summer precipitation is a better predictor of tree carbon reduction than annual precipitation. Finally, the change in soil carbon stock is less conspicuous than in tree carbon stock, showing no significant change in the north and a relatively small but consistent decline in the south. The study indicates that the prospect of higher water deficit caused by climate change cannot be ignored in future forest management planning.
KeywordsForest carbon stock Water deficiency Climate change Dynamic modelling ForSAFE Sweden
Climate change strongly affects the water cycle and can alter the magnitude and seasonality of precipitation (IPCC 2001). Boreal forests are expected to be particularly sensitive to climate change due to the rate of change projected at higher latitudes (Gauthier et al. 2015). A decrease in productivity and drought-induced mortality are among the main potential threats to boreal forests caused by future drier conditions (Gauthier et al. 2015; Price et al. 2013). In some areas, boreal and hemi-boreal forests are already experiencing water stress, including drought-induced tree mortality (Allen et al. 2010; Michaelian et al. 2010; Peng et al. 2011; Soja et al. 2007). Rapid changes of climate and trends towards drier conditions increase the risk of forest decline due to water stress in the boreal zone (Price et al. 2013; Wang et al. 2014). In Sweden, some studies show that water availability already limits forest growth in the southern part of the country (Bergh et al. 1999, 2005; van Leeuwen et al. 2000). In 2016–2018, groundwater levels were lower as compared to historical levels in several parts of Sweden, during the growing season. The situation was particularly alarming in southern Sweden where groundwater levels were critically low (below the 15th percentile) for more extended periods (SGU 2018). Moreover, future climate scenarios highlight that southern Sweden could experience drier summers (Ministry of Environment 2014), therefore increasing the risk of lower water availability during the growing season.
Boreal forests are important global carbon stocks and sinks (Bradshaw and Warkentin 2015) and can substantially contribute to climate change mitigation by storing carbon and by providing biomass for the production of renewable energy and materials (Canadell and Raupach 2008; Lundmark et al. 2014). Sweden has set ambitious targets to increase the production of bioenergy and renewable products from forests while preserving forest carbon stocks to achieve climate change mitigation goals (SOU 2016: 47 2016; Swedish Government 2017). However, to guarantee a long-term balance between forest production and carbon sequestration, it is necessary to account for environmental changes that can affect forest ecosystem processes, such as climate change. In evaluating the potential contribution of Swedish forests to climate mitigation goals, climate change effects were often disregarded (Cintas et al. 2017; Lundmark et al. 2015). A study by Poudel et al. (2011) included the effect of temperature increase on future forest production in Sweden suggesting a positive effect of climate change on forest production. However, this study excluded the effect of changes in precipitation, which was shown to play a significant role in future forest growth in Finland where water deficiency could reduce the growth of pine and spruce by 6–7% compared to non-water-limited forests (Kellomäki et al. 2018). The combined effect of temperature and precipitation changes and the consequences of reduced water availability on forests’ growth and carbon stocks in Sweden is yet to be investigated.
Water deficiency has the potential to cancel out the positive effect of temperature increase on tree growth.
Litterfall and decomposition will react similar to changes in temperature and precipitation, making the overall response of soil carbon inconclusive.
The study is based on the simulation of future forest growth with the dynamic ecosystem model ForSAFE and climate change assumptions based on two global climate models (ECHAM5, CCSM3) for the period 2010–2100.
Materials and methods
Summary of the input data to ForSAFE
Precipitation and temperature
Historical records from the Swedish Meteorological Institute (SMHI) and the two future projections using the ECHAM5-r3 and CCSM models following the A2 storyline of emissions (David Rayner, personal communication)
Global radiation from the NCEP/NCAR reanalysis converted into photosynthetically active radiation using SMHI’s STRÅNG model
SO42−, NO3−, NH4+ Cl−, Na+, Ca2+, K+, Mg2+
EMEP model (Simpson et al. 2012), historically according to the emission scenarios from Schöpp et al. (2003) and the future projections following the emissions scenario of the current legislation of the revised Gothenburg protocol of the Long Range Transboundary Air Pollution Convention (LRTAP)
MATCH model (Persson et al. 1996), held constant over the simulation period
Texture and mineralogy
Raw data from the Research Infrastructure National Forest Inventory (RINFI) of Sweden (Hägglund 1985). Soil mineralogy was estimated with the UPPSALA model based on soil total elemental analysis from the RINFI (Alveteg 2004)
Norway spruce, Scots pine
Based on recommendations by the Swedish Forest Agency
The model was applied to 544 productive forest sites in Sweden included in the Swedish National Forest Inventory. The database describing soil properties at the studied sites was originally compiled by Warfvinge and Sverdrup (1995) and Alveteg (2004).
Monthly averages of the historical temperature and precipitation data for the period 1961–2010 were interpolated from the Swedish Meteorological Institute’s station data for that period. Monthly averages of photosynthetically active radiation for the same period were converted from global radiation from the NCEP/NCAR reanalysis by calibrating with SMHI’s STRÅNG model.
Temperature and precipitation trends from historical data were used to extend the climate series in the past back to 1900.
A baseline scenario (BAS), which projects steady temperatures and precipitation until 2100 by repeating the values of climate variables from 1981 to 2010 in the future.
- Two climate change scenarios derived from two global climate models (GCMs), namely the Max Plank Institute’s ECHAM5 and the National Center for Atmospheric Research’s CCSM3. Both models simulated future climate changes under the SRES A2 emissions storyline. The GCM data were calibrated so that the monthly means and variances matched the observations from 1961 to 2008 period (David Ryner, personal communication). Both GCMs predict a clear increase in winter and summer temperatures (Fig. 2). While the predicted increase in winter temperatures is geographically comparable, ECHAM gives a generally higher increase in summer temperatures, with a more pronounced increase in the south and the north-east. Both models predict an increase in winter precipitation, but differ markedly on summer precipitation. With the exception of a few east-coastal sites, CCSM gives a clear distinction between a wetter north and a drier south. ECHAM predicts drier summers overall except in the western half of central and northern Sweden and the very north-east.
The atmospheric deposition of nitrogen (N, reduced and oxidized) and sulphur (S) is derived from the EMEP model (Simpson et al. 2012) by associating each site with its corresponding cell on the EMEP grid. The historical N and S depositions were based on the emission trends described in Schöpp et al. (2003). The atmospheric deposition of calcium, magnesium, potassium, sodium and chloride were derived from the MATCH model (Langner et al. 1996) and Lövblad et al. (1992). The atmospheric deposition of base cations and chloride are held constant throughout the simulation period.
In ForSAFE, the soil profile is represented as a column divided in discrete soil layers with homogeneous properties. The soil properties used as input data include information on texture, bulk density, base saturation, cation exchange capacity, hydraulic properties (field saturation, field capacity and wilting point) and mineralogy composition for each soil layer.
The soil database was originally compiled by Alveteg (2004) based on Swedish National Forest Inventory (SNFI) data (Hägglund 1985). The availability of soil base saturation (BS) in a subset of the SNFI dictated the selection of sites included in this study. Observed base saturation is necessary to back calculate the initial base saturation at the starting year of the simulations. The initial base saturation is found by iteratively and adaptively selecting starting values so that the simulated BS levels match the observed values on the observation years (Belyazid et al. 2006).
The 544 forest sites included in this study are managed forests dominated by Scots pine and/or Norway spruce. Parametrization data to the model include vegetation parameters that regulate photosynthesis, tree phenology, tree water use, nutrient allocation in the biomass and mortality rates. More detailed information on the vegetation parameters are reported in the Electronic Supplementary Material, Table S1.
As a second step, we simulated the carbon content in tree biomass (TreeC) and topsoil (SOC) over the period 1900–2100 under the baseline and climate change scenarios to evaluate the effects of climate change on forest carbon stocks (see Eq. 1). Finally, we analysed the relations between carbon stock changes and changes of temperature, precipitation and relative evapotranspiration to understand which variables better explain the changes of TreeC and SOC in the future. The analysis on the summer period refers to the months June, July and August, while the winter period includes the months December, January and February.
The simulation of future tree water deficiency (WD) under the baseline and climate change scenarios at the 544 forest sites projects that existing water limitation patterns will be exacerbated by climate change as compared to the baseline (Fig. 4b). WD will become even more severe in central and southern regions and in the coastal areas. These changes are a combination of increased temperature affecting transpiration through increasing the vapour pressure difference and changes of summer precipitation affecting water inputs during the growing season. The Swedish forest areas that will be most water-limited correspond to the areas where higher temperatures will be combined with lower summer precipitation. According to the CCSM climate model, this will cover the southern half of Sweden, while according to ECHAM, this may stretch along the coast to northern Sweden (Fig. 4b).
Discussion and conclusions
The results are able to verify the first hypothesis of the study that lower summer precipitation has the potential to override the positive effect of higher temperatures on forest growth. While higher temperatures are expected to promote growth in the north and north-western parts of Sweden, lower water availability in the south may drive a net reduction in tree carbon stocks compared to a hypothetical baseline without climate change.
The second hypothesis, the effect of climate change on soil carbon will be inconclusive, could not be verified by the study. Instead, the results show a similar pattern to tree carbon, with a possible increase in the north and a decrease in the south in response to climate change. The change in soil carbon is, nevertheless, less expressed than the change in tree carbon.
The simulations show that water deficiency is already today a limiting factor to growth in many parts of Sweden and more severe in the south-eastern part of the country. This is one of only a few studies to address water stress in Swedish forests (Bergh et al. 1999; van Leeuwen et al. 2000) and the first to analyse the possible effects on forest growth on a national scale.
The simulated lower water availability during the growing season is expected to become more severe in the future, restricting and even demoting forest growth. While this effect of lower water availability on forest growth is not confirmed by other investigations specific to Sweden, it is in line with other regional studies (Gustafson et al. 2017; Reichstein et al. 2007; von Buttlar et al. 2018). The negative effect of the simulated climate change scenarios on ecosystem productivity is more expressed in the southern Sweden where elevated temperatures coincide with lower summer precipitation, as is corroborated by the findings by von Buttlar et al. (2018). These results can be considered conservative given that the model only simulates changes in photosynthesis and respiration; droughts combined with high temperatures can additionally lead to forest dieback (Allen et al. 2010; Michaelian et al. 2010; Price et al. 2013) and more frequent and severe disturbances such as pests and fires (Astrup et al. 2018; Brecka et al. 2018; Jönsson et al. 2007), further exacerbating the effect of future climate on forest productivity.
The simulations also indicate that using annual projections of precipitation to forecast forest growth may be misleading, given that annual precipitation is expected to increase all over the country. In contrast, it is more relevant to focus on summer precipitation which more closely affects forest growth. According to our results, higher water deficiency during the growing season is likely to become a reality in all of Sweden, more so in the south and east, but even in the rest of the country. This creates new challenges that Sweden’s current forestry model is not well adapted to deal with. At the same time, yearly precipitation levels are expected to increase because of a significant increase in winter precipitation, having implications on the water-regulating function of forests. Combined with increased temperatures, higher winter precipitation may have substantial implications on the hydrological fluxes even on a landscape level. Forestry can, however, be effective in adapting to and mitigating these changes through the choice of alternative forest management regimes (Aussenac and Granier 1988; Kubin et al. 2017; Roche et al. 2018) or different tree species (Ellison et al. 2017). It remains urgent to explore the economic and technical feasibility of these measures to select appropriate management strategies in a timely manner.
Ecosystem productivity, expressed as tree carbon and soil carbon stocks, is likely to be constrained by climate change in the simulated future. This further stresses the need to review forestry practices that have traditionally been put forward to stimulate forest growth but implicitly assumed no risk of water limitation. The choice of tree species, soil preparation techniques (such as scarification and clearing of ditches) and the prospect of increasing productivity through fertilization all need to be reviewed in light of the possibility of water scarcity.
The results of this study challenge the prospect of maintaining the current levels of high forest productivity in Sweden, and points to a limited possibility of increased forest productivity with current management practices. The expectation of using forest products and services as a cornerstone in Sweden’s effort to reach climate neutrality cannot ignore the plausible possibility that climate change may constrain the supply of those services.
The authors would like to thank Formas for the financial support granted under the projects QWARTS and ForWater and the Swedish Energy Agency.
Funding was provided by the Swedish Research Council (Formas) (projet numbers 2011–01691 and 2010–89) and by the Swedish Energy Agency (Energimyndigheten).
- Alveteg M (1998) Dynamics of forest soil chemistry. Lund University, LundGoogle Scholar
- Alveteg M (2004) Projecting regional patterns of future soil chemistry status in Swedish forests using safe. In: Wieder RK, Novák M, Vile MA (eds) Biogeochemical investigations of terrestrial, freshwater, and wetland ecosystems across the globe. Springer, Dordrecht, pp 49–59. https://doi.org/10.1007/978-94-007-0952-2_4 CrossRefGoogle Scholar
- Gustafson EJ, Miranda BR, De Bruijn AMG, Sturtevant BR, Kubiske ME (2017) Do rising temperatures always increase forest productivity? Interacting effects of temperature, precipitation, cloudiness and soil texture on tree species growth and competition. Environ Model Softw 97:171–183. https://doi.org/10.1016/j.envsoft.2017.08.001 CrossRefGoogle Scholar
- Hägglund B (1985) En ny svensk riksskogstaxering (A new Swedish National Forest Survey). Swedish University of Agricultural Sciences, Report 37. Uppsala, Sweden (In Swedish with English summary)Google Scholar
- IPCC (2001) Chapter 4—hydrology and water resources. In: McCarthy JJ, Canziani OF, Leary NA, Dokken DJ, White KS (eds) Climate change 2001: impacts, adaptation and vulnerability, Contribution of Working Group II to the third assessment report of the intergovernmental panel on climate change. Cambridge University Press, Cambridge, pp 191–234Google Scholar
- Jacobson S, Sikström U, Nyström K, Övergaard B (2008) INGVAR—thinning schedules and forest management planning tool. (In Swedish with English Summary). SkogForsk, UppsalaGoogle Scholar
- Langner J, Persson C, Robertson L, Ullerstig A (1996) Air pollution assessment study using the MATCH Modelling System. Application to sulphur and nitrogen compounds over Sweden 1994. Swedish Meteorological and Hydrological Institute, Report no. 69, NorrköpingGoogle Scholar
- Lindström G, Gardelin M (1992) Chapter 3.1: hydrological modelling—model structure. In: Modelling groundwater response to acidification vol 5. SMHI, Reports Hydrology, NorrköpingGoogle Scholar
- Lövblad G, Andersen B, Joffre S, Pedersen U, Hovmand M, Reissell A (1992) Mapping deposition of sulphur, nitrogen and base cations in the Nordic countries. Swedish Environmental Research Institute. Report B 1055, StockholmGoogle Scholar
- Lundmark T et al (2014) Potential roles of Swedish forestry in the context of climate change mitigation forests 5. https://doi.org/10.3390/f5040557
- Marklund LG (1988) Biomassafunktioner för tall, gran och björk i Sverige vol Rapport 45. Sveriges lantbruksuniversitet, Institutionen för skogstaxeringGoogle Scholar
- Ministry of Environment (2014) Sweden’s sixth national report on climate change (in Swedish) vol Ds 2014:11. Swedish Ministry of Environment and Energy, StockholmGoogle Scholar
- Naturvårdsverket (2016) Climate change and the environmental objectives—CLEO (in Swedish), vol 6705. Bromma, SwedenGoogle Scholar
- Persson C, Langner J, Robertson L (1996) Air pollution assessment studies for Sweden based on the match model and air pollution measurements. In: Gryning S-E, Schiermeier FA (eds) Air pollution modeling and its application XI. Springer, Boston, pp 127–134. https://doi.org/10.1007/978-1-4615-5841-5_15 CrossRefGoogle Scholar
- SGU (2018) Grundvattennivåns avvikelse historiskt. https://www.sgu.se/grundvatten/grundvattennivaer/slider/. 07-11-2018
- SOU 2016: 47 (2016) A climate and air conservation strategy for Sweden (in Swedish). Stockholm, Sweden. https://www.regeringen.se/rattsliga-dokument/statens-offentliga-utredningar/2016/06/en-klimat–och-luftvardsstrategi-for-sverige/
- Swedish Government (2017) Environmental objectives—with a view to the future (in Swedish), Skr. 2017/18:265. https://data.riksdagen.se/fil/EDA7C87D-9121-4D0D-A03E-DC33C06E745A. Accessed 24 Jan 2019
- van Leeuwen EP, Hendriks KCMA, Klap JM, de Vries W, de Jong E, Erisman JW (2000) Effects of environmental stress on forest crown condition in Europe. Part II: estimation of stress induced by meteorology and air pollutants. Water Air Soil Pollut 119:335–362. https://doi.org/10.1023/A:1005182226292 CrossRefGoogle Scholar
- Warfvinge P, Sverdrup H (1995) Critical loads of acidity to Swedish forest soils, vol 5. Lund University, LundGoogle Scholar
- Zanchi G, Belyazid S (2019) Swedish forests are greyer than prescribed. (manuscript in preparation)Google Scholar
OpenAccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.