Climatic Change

, Volume 118, Issue 2, pp 259–273

Impacts of climate change on primary production and carbon sequestration of boreal Norway spruce forests: Finland as a model

Authors

    • State Key Laboratory of Estuarine and Coastal ResearchEast China Normal University
    • School of Forest SciencesUniversity of Eastern Finland
  • Seppo Kellomäki
    • School of Forest SciencesUniversity of Eastern Finland
  • Heli Peltola
    • School of Forest SciencesUniversity of Eastern Finland
  • Xiao Zhou
    • School of Forest SciencesUniversity of Eastern Finland
  • Hannu Väisänen
    • School of Forest SciencesUniversity of Eastern Finland
  • Harri Strandman
    • School of Forest SciencesUniversity of Eastern Finland
Article

DOI: 10.1007/s10584-012-0607-1

Cite this article as:
Ge, Z., Kellomäki, S., Peltola, H. et al. Climatic Change (2013) 118: 259. doi:10.1007/s10584-012-0607-1

Abstract

The aim of this study was to estimate the potential impacts of climate change on the spatial patterns of primary production and net carbon sequestration in relation to water availability in Norway spruce (Picea abies) dominated forests throughout Finland (N 60°–N 70°). The Finnish climatic scenarios (FINADAPT) based on the A2 emission scenario were used. According to the results, the changing climate increases the ratio of evapotranspiration to precipitation in southern Finland, while it slightly decreases the ratio in northern Finland, with regionally lower and higher soil water content in the south and north respectively. During the early simulation period of 2000–2030, the primary production and net carbon sequestration are higher under the changing climate in southern Finland, due to a moderate increase in temperature and atmospheric CO2. However, further elevated temperature and soil water stress reduces the primary production and net carbon sequestration from the middle period of 2030–2060 to the final period of 2060–2099, especially in the southernmost region. The opposite occurs in northern Finland, where the changing climate increases the primary production and net carbon sequestration over the 100-year simulation period due to higher water availability. The net carbon sequestration is probably further reduced by the stimulated ecosystem respiration (under climate warming) in southern Finland. The higher carbon loss of the ecosystem respiration probably also offset the increased primary production, resulting in the net carbon sequestration being less sensitive to the changing climate in northern Finland. Our findings suggest that future forest management should carefully consider the region-specific conditions of sites and adaptive practices to climate change for maintained or enhanced forest production and carbon sequestration.

1 Introduction

Boreal forests play an important role in the global carbon balance, and their carbon stores need better management (Pimm et al. 2009). The effects of expected climate change on forest productivity and carbon sequestration have become a subject of debate. Potter et al. (2009) reported that global warming stimulates the carbon sequestration rate at the small-scale in northern terrestrial ecosystems (including boreal forests). However, increasingly more studies based on experiments and models have demonstrated that the capacity of forest carbon storage under climate change is dependent on the conditions of water and nutrients (Pussinen et al. 2009; McCarthy et al. 2010). The changes in the frequency of occurrence of weather extremes (e.g. wind storm, spring temperature backlashes and summer drought) in response to climate change are explored (Schlyter et al. 2006; Gastineau and Soden 2009; Seidl and Blennow 2012), which would decrease the vitality and production of the boreal forests. In other ways, the global warming would drive the insect outbreaks with deleterious effects on the boreal forests (Jönsson et al. 2007; Komonen et al. 2011).

Norway spruce (Picea abies) is a representative tree species throughout Europe, especially in the boreal zones. In Finland, Norway spruce represents, on average, approximately 40 % of the national forest resources with primary importance for carbon stores and the wood-processing industry (Finnish Forest Research Institute 2007). However, the growth of Norway spruce is water-limited (Roberntz and Stockfors 1998; Phillips et al. 2001; Bergh et al. 2005; Schlyter et al. 2006), relative to other boreal tree species (Kellomäki et al. 2005). Moreover, Norway spruce favors organic soil, which is expected to lose more carbon under warming climate (Ise et al. 2008).

According to the FINADAPT climate change scenario (following the IPCC SRES A2 scenario) of Finland, the annual mean temperature is predicted to be higher by 4–6 °C in different regions (from southern to central and northern Finland), with a concurrent elevation of CO2, by the end of the 21st century (Carter et al. 2005). Precipitation may increase in wintertime while staying the same as currently in summertime. Accordingly, the climate change would modify the atmospheric and edaphic conditions of the forest sites, causing the response of tree growth and the carbon budget to differ from that observed under the current climatic conditions. However, few studies have analyzed the impacts of changing climate on the forest production and carbon sequestration distributing along the continental zones with a wide-range gradient climatic characteristics (Bergh et al. 2005; Pussinen et al. 2009).

Within this context, a detailed ecosystem model was used to simulate the impacts of climate change on the spatial patterns of primary production and carbon sequestration of Norway spruce-dominated forests in relation to hydrological budget and water availability throughout Finland over a 100-year period (2000–2099).

2 Material and methods

2.1 Model description

The process-based ecosystem model (FinnFor) utilized in this study was initially developed by Kellomäki et al. (1993). The structural and functional properties of the FinnFor model have been presented by Ge et al. (2010, 2011a, b, 2012). The model is composed of the object-oriented submodels of stand development and seasonality, canopy photosynthesis and autotrophic respiration, biomass allocation and tree growth, soil efflux, stand self-thinning, nutrition cycle, evapotranspiration, soil profile water flow and snow dynamics (see Supplementary material for details).

The performance of the FinnFor model on stand development, leaf area expansion and tree growth of Finnish forests has been estimated in detail by contrasting the calculated values against corresponding measurements available from the 245 permanent sample plots (1,191 trees) of the Finnish National Forest Inventory (NFI) (see Supplementary material for details), which covers the 13 Forest Centers throughout Finland (60°–70° N). Recently, based on the model validation work on the gross primary production (GPP), net primary production (NPP), total ecosystem respiration (TER), net carbon sequestration (also net ecosystem exchange, NEE) and evapotranspiration (ET) components of a Finnish coniferous stand (see Supplementary material for details), the modeled day-to-day variations of carbon and water fluxes showed relatively good agreement with the eddy covariance measurements over a period of 10 years (1999–2008) under climatic variability.

2.2 National forest inventory data used for simulation inputs

The simulations used the initial status of the tree stands representing the permanent sample sites of the 9th Finnish National Forestry Inventory (NFI) during 1999–2002 (Fig. 1). A total of 707 Norway spruce-dominated (in terms of the stem volume) plots on upland mineral soils were initialized in this study (Fig. 1). All of the spruce forests are fertile and medium fertile sites, i.e., the Oxalis-Myrtillus (OMT type, 272 plots) and Myrtillus type (MT, 435 plots) types (Tamminen 1991). The amount of soil organic matter (SOM) of litter and humus on the plots was defined on the basis of thickness of the organic layer measured in the national inventory. The thickness was converted into the mass of SOM, using the bulk density of SOM considering the site type and tree species dominating the plot (Tamminen 1991). Thereafter, the mass of SOM was regressed against the prevailing temperature sum (Day-Degree, d.d.) of the plot by the site types as presented in Fig. 1. These values are also used in initializing the simulations for a specific site type.
https://static-content.springer.com/image/art%3A10.1007%2Fs10584-012-0607-1/MediaObjects/10584_2012_607_Fig1_HTML.gif
Fig. 1

Locations of the Norway spruce-dominated plots used in the calculation from the 9th National Forest Inventory (NFI). The shaded and unshaded circles indicate the site types of OMT and MT, respectively. The climate zones I–IV were identified on the basis of the temperature sum distribution. The northernmost region with an extremely low temperature sum (< 600 d.d.) was excluded in this study (not favorable for Norway spruce). The grey boundary lines divide Finland into 13 Forest Centers. Embedded chart (bottom left): amount of soil organic matter as a function of site type and temperature sum

As the regional temperature sum affects the growth and management recommendation in Finnish forests (Tapio 2006), the entirety of Finland is divided for further analysis into four climatic zones on the basis of the temperature sum and site types: zones I and II (> 1,000 d.d.) are located in southern Finland, and zones III and IV (< 1,000 d.d.) a located in northern Finland (Fig. 1, Table 1). To initialize the simulation, the site characteristics of the 9th NFI are represented in Table 1.
Table 1

The site distribution in different climatic zones (I–IV) and site types (OMT and MT) with ranged initial leaf area index (LAI) and stand stocking for the year 2000

Site type

Stand layout

Site distribution in the climatic zones

I

II

III

IV

OMT

Number of plots

158

109

5

0

LAI (m2 m−2)

0.1–5.5

0.1–8.3

1.9–6.6

Stocking (m3 ha−1)

24–352

23–357

57–379

MT

Number of plots

168

171

74

22

LAI (m2 m−2)

0.1–5.4

0.2–5.1

0.2–6.4

0.2–4.8

Stocking (m3 ha−1)

24–277

21–283

19–355

16–176

2.3 Climate scenarios

The climate and weather input data for the whole of Finland used in this work represent the grid-based current climate (CUR) and changing climate (CC) scenarios compiled by the Finnish Meteorological Institute for the FINADAPT project (Carter et al. 2005). The spatial resolution of the grid for the current climate is 10 km × 10 km, while for the climate change scenario it is 50 km × 50 km. In the simulations for a given sample plot, the calculation algorithm uses the climate for the closest grid point of the climate data.

The CUR scenario, used for the period 2000–2099, represents the mean data of the period 1971–2000 repeated over the total simulation period, with a constant atmospheric CO2 concentration of 352 ppm. In the CC scenario based on the IPCC A2 emission scenario, the annual mean temperature increases by 4–6 °C throughout the country. The atmospheric CO2 concentration is 352 ppm at the start of the simulation in 2000 and increases up to 836 ppm at the end, in 2099. The annual precipitation amount increases by 10–15 % in the southern region and by 15–25 % in the north. However, the increases in precipitation occur mainly in wintertime, and in summertime it stays mostly the same as currently.

The original data for the CUR and CC scenarios represented the daily values. For model initialization, the time resolution of the climate data will be broken down to an hourly basis in the weather simulator of the FinnFor model (Kellomäki et al. 1993). Other climatic inputs of radiation, relative humidity, cloudiness and windiness for the CUR and CC scenarios were assumed the same by means of weather statistics for the period 1971–2000.

2.4 Analyses of model simulation outputs

Over the 100-year (2000–2099) simulation period, the simulation was conducted without management to identify the pure climate change effects. In the current simulations, we excluded the abiotic (storm, frost and fire) and biotic (insect and fungal pests) damage on the growth and life cycle of trees because the focus of this study was on the pure impacts of changes in the main climatic factors (temperature, precipitation and atmospheric CO2) on hydrological budget, primary production and carbon balance in the boreal spruce forests. The ratio of annual evapotranspiration to precipitation amount and the soil moisture were calculated for water availability estimation. The annual gross photosynthesis of stands (understory vegetation was not taken into account) was quantified as GPP, and NPP was calculated as the difference between GPP and autotrophic respiration. Meanwhile, NEE was identified as the net carbon sequestration, with the difference value between GPP and total ecosystem respiration. To scale up the modeling outputs (Appendix Table 5) from the plots (individual-to-stand) to the regional and national level, we employed the spatial interpolating program (Kriging method) for geographical information using ARC/INFO (ESRI ArcMap 9.2).

3 Results

Under CUR, the average annual snow depth (total annual accumulation/365 days) is 25–50 mm in the southern (zones I and II) and 50–100 mm in the northern (zones III and IV) regions over the 100-year simulation period. Under CC, the snow depth is substantially reduced due to earlier snowmelt and higher evaporation and to the decreased fraction of precipitation as snow (Fig. 2a). Over the simulation period, the average water depletion of forest (identified as evapotranspiration) increases up to 10–15 % in the southern region and 2–5 % in the northern region under CC compared with CUR. The average ratio of evapotranspiration to precipitation in zones I and II is higher, by an average of 10 %, while it is slightly lower in zones III and IV under CC than that under CUR (Fig. 2b). The soil water content is identified by the balance between the input of precipitation with snowmelt and the depletion of evaporation and water use for tree growth. Consequently, the average annual soil water content is lower in zones I and II and higher in the zones III and IV under CC than under CUR (Fig. 2c).
https://static-content.springer.com/image/art%3A10.1007%2Fs10584-012-0607-1/MediaObjects/10584_2012_607_Fig2_HTML.gif
Fig. 2

Distribution of the average annual mean snow depth (a), the average ratio of annual evapotranspiration to precipitation (ET/P, b) and (c) the average annual soil moisture (volume water content) over the 100-year simulation period under the current (CUR) and changing climate (CC). The solid lines divide Finland into four climatic zones (I–IV). The shaded area is outside the Norway spruce occurrence limit

Figure 3 shows the spatial distribution of the cumulative GPP, NPP and NEE over the 100-year simulation period. Compared with CUR, CC reduces the cumulative GPP and NPP in most of the southern region, while it increases the primary production in all the northern forests. The cumulative NEE over the 100-year simulation period also distinguishes between the southern and northern regions under CUR compared with CC (Fig. 3).
https://static-content.springer.com/image/art%3A10.1007%2Fs10584-012-0607-1/MediaObjects/10584_2012_607_Fig3_HTML.gif
Fig. 3

Distribution of the cumulative gross primary production (GPP, a), net primary production (NPP, b) and net carbon sequestration (NEE, c) over the 100-year simulation period under the current (CUR) and changing climate (CC). The solid lines divide Finland into four climatic zones (I–IV). The shaded area is outside the Norway spruce occurrence limit

To separate the gradual effect of the changing climate, the primary production and net annual carbon sequestration are considered according to three simulation periods ending in 2030, 2060 and 2099. The average annual GPP and NPP under CC are higher by 5–15 % than that under CUR until 2030 regardless of the climate zones. In the middle period, 2030–2060, CC decreases the average annual GPP and NPP by 0.1–0.8 % in zones I and II, with a further decline during the period 2060–2099 (Tables 2 and 3). The decrease rate of primary production in zone I is greater than that in zone II. In zones III and IV, the average annual GPP and NPP are higher over the whole 100-year simulation period under CC compared with CUR. In the last of simulation period, 2060–2099, CC increases the primary production by 17–25 % over CUR.
Table 2

Average annual gross primary production (GPP) during different periods (2000–2030, 2030–2060 and 2060–2099) under the current (CUR) and changing climate (CC) in different climatic zones (I–IV) and site types (OMT and MT). The values in parentheses are the percentage change (%) from that under the current climate

Climatic zone

Site type

GPP (Mg C ha−1 year−1)

2000–2030

2030–2060

2060–2099

CUR

CC

CUR

CC

CUR

CC

I

OMT

13.5

14.4 (6.8)

14.3

14.2 (−0.5)

12.8

11.8 (−7.5)

MT

13.2

13.9 (4.8)

15.7

15.6 (−0.4)

13.3

12.1 (−9.4)

Total

13.3

14.1 (5.8)

15.0

14.9 (−0.4)

13.1

11.9 (−8.5)

II

OMT

12.3

13.3 (8.4)

13.3

13.4 (−0.1)

11.5

10.6 (−7.3)

MT

12.4

13.3 (7.0)

14.2

14.1 (−0.1)

12.3

11.4 (−7.6)

Total

12.4

13.3 (7.7)

13.7

13.7 (−0.1)

11.9

11.0 (−7.5)

III

OMT

10.7

11.6 (9.1)

11.0

12.5 (13.0)

11.4

13.6 (19.4)

MT

9.7

10.6 (9.0)

10.6

11.8 (11.5)

10.0

11.9 (19.2)

Total

10.2

11.1 (9.0)

10.8

12.1 (12.3)

10.7

12.7 (19.3)

IV

OMT

MT

3.7

4.2 (15.1)

3.9

4.6 (18.6)

4.5

5.7 (25.1)

Total

3.7

4.2 (15.1)

3.9

4.6 (18.6)

4.5

5.7 (25.1)

Table 3

Average of annual net primary production (NPP) during different periods (2000–2030, 2030–2060 and 2060–2099) under the current (CUR) and changing climate (CC) in different climatic zones (I–IV) and site types (OMT and MT). The values in parentheses are the percentage change (%) from that under the current climate

Climatic zone

Site type

NPP (Mg C ha−1 year−1)

2000–2030

2030–2060

2060–2099

CUR

CC

CUR

CC

CUR

CC

I

OMT

4.6

4.9 (5.2)

4.9

4.9 (−0.8)

4.4

3.9 (−12.7)

MT

4.6

4.8 (4.4)

5.4

5.4 (−0.7)

4.6

4.0 (−13.9)

Total

4.6

4.8 (4.8)

5.2

5.1 (−0.7)

4.5

4.0 (−13.3)

II

OMT

4.2

4.5 (6.6)

4.6

4.6 (−0.2)

4.0

3.6 (−10.1)

MT

4.3

4.6 (6.5)

4.9

4.9 (−0.1)

4.2

3.8 (−11.3)

Total

4.3

4.5 (6.6)

4.7

4.7 (−.02)

4.1

3.7 (−10.7)

III

OMT

3.7

4.0 (8.7)

3.8

4.2 (10.1)

3.9

4.6 (17.1)

MT

3.4

3.6 (8.6)

3.6

4.0 (8.8)

3.4

4.0 (16.7)

Total

3.5

3.8 (8.6)

3.7

4.1 (9.5)

3.7

4.3 (17.0)

IV

OMT

MT

1.3

1.4 (12.6)

1.3

1.6 (18.0)

1.6

1.9 (23.8)

Total

1.3

1.4 (12.6)

1.3

1.6 (18.0)

1.6

1.9 (23.8)

The average annual NEE is higher by 4–12 % under CC than that under CUR during the period of 2000–2030, regardless of the climates in zones I and II (Table 4). However, CC decreases the average annual NEE from the middle period of 2030–2060 (by 0.3–0.8 %) until the final simulation period of 2060–2099 (by 11–14 %) compared with CUR. The decrease rate of NEE in zone I is greater than that in zone II. In zones III and IV, the average annual NEE is higher under CC over the whole simulation period compared with CUR. In the last of simulation period, 2060–2099, CC increases the net carbon sequestration by 16–23 % compared with CUR in zones III and IV (Table 4). As further shown in Fig. 4, the percentage change of the average annual NEE is lower by 2–10 % in the southern region but is higher by 8–16 % in the northern region over the 100-year simulation period under CC compared with CUR.
Table 4

Average of annual net carbon sequestration (NEE) during different periods (2000–2030, 2030–2060 and 2060–2099) under the current (CUR) and changing climate (CC) in different climatic zones (I–IV) and site types (OMT and MT). The values in parentheses are the percentage change (%) from that under the current climate

Climatic zone

Site type

NEE (Mg C ha−1 year−1)

2000–2030

2030–2060

2060–2099

CUR

CC

CUR

CC

CUR

CC

I

OMT

2.9

3.0 (5.1)

3.1

3.0 (−0.8)

2.7

2.4 (−13.2)

MT

2.8

2.9 (3.9)

3.3

3.3 (−0.8)

2.8

2.5 (−13.8)

Total

2.8

3.0 (4.5)

3.2

3.2 (−0.8)

2.8

2.4 (−13.5)

II

OMT

2.6

2.8 (6.3)

2.8

2.8 (−0.4)

2.4

2.2 (−11.3)

MT

2.6

2.8 (6.2)

3.0

3.0 (−0.3)

2.6

2.3 (−12.0)

Total

2.6

2.8 (6.2)

2.9

2.9 (−0.4)

2.5

2.3 (−11.7)

III

OMT

2.3

2.4 (7.5)

2.3

2.6 (9.3)

2.4

2.8 (16.2)

MT

2.1

2.2 (7.3)

2.2

2.4 (8.6)

2.1

2.5 (16.1)

Total

2.2

2.3 (7.4)

2.3

2.5 (9.0)

2.3

2.6 (16.2)

IV

OMT

MT

0.8

0.9 (12.0)

0.8

1.0 (17.1)

1.0

1.2 (23.2)

Total

0.8

0.9 (12.0)

0.8

1.0 (17.1)

1.0

1.2 (23.2)

https://static-content.springer.com/image/art%3A10.1007%2Fs10584-012-0607-1/MediaObjects/10584_2012_607_Fig4_HTML.gif
Fig. 4

Change of the average annual net carbon sequestration (NEE) over the 100-year simulation period under changing climate from that under the current climate

Figure 5 shows the percentage change of the primary production and net annual carbon sequestration under CC compared with that under CUR as a function of the temperature sum at the end of the simulation period for the southern and northern regions. The changes of GPP, NPP and NEE correlate positively with the increase in temperature sum in zones III and IV. The higher temperature sum does not enhance the primary production and net annual carbon sequestration in zones I and II; instead, there is a slightly negative effect by the elevated temperature. The regression slopes of the functions indicate that the changes of decrease (southern) in NEE are slightly higher, and increase (northern) lower than those of GPP and NPP under climate warming (Fig. 5).
https://static-content.springer.com/image/art%3A10.1007%2Fs10584-012-0607-1/MediaObjects/10584_2012_607_Fig5_HTML.gif
Fig. 5

For southern (a, n = 605) and northern (b, n = 102) Finland, percentage change of the average annual gross primary production (GPP), net primary production (NPP) and net carbon sequestration (NEE) under changing climate from that under the current climate as a function of the change of the temperature sum of 2099 compared with 2000

4 Discussion and conclusions

To predict the effects of climate change on the primary production and carbon balance of forest, process-based system models are useful scientific tools, providing formalized statements of hypotheses (Pussinen et al. 2009; Luo et al. 2011). In this study, the process-based FinnFor model is utilized for the first time to estimate the impacts of climate change on a boreal dominant species (Norway spruce) over the whole country.

For this study, the A2 scenario of FINADAPT project was selected instead of the B1 scenario in order to have relatively larger changes in the climate. This may result in uncertainties regarding details of future weather patterns. In general, the A2 scenario could be preferable in the climate change impact and adaptation studies because it will better tackle the risks to forests, through drought for example (Carter et al. 2005; Kellomäki et al. 2005), as associated with the climate change. We also did not predict the effects of extreme weather and biotic damage under climate change on the growth and life cycle of trees, due to unavailability of model establishment and calibration. Nevertheless, the important issues on impact of extreme weather and insect outbreaks in Finland under climate change had been thoroughly stated by Kellomäki et al. (2005), beyond the modeling procedure.

During the early simulation period, a moderate increase in temperature and CO2 will most likely lead to increased carbon uptake and tree growth in both the southern and northern regions, which is in agreement with the results of BIOMASS model on Norway spruce (Bergh et al. 2005). McCarthy et al. (2010) also reported that elevated CO2 led to a greater primary production and leaf area of the boreal coniferous trees from the Duke FACE (free-air CO2 enrichment) site by several years.

As further simulated for the future of Finnish forests, emphasis is on the effects of water availability on the regional primary production and carbon balance of spruce forest. The changing climate will create an environment with more evaporation due to a higher vapor pressure deficit and lower diffusive resistance. Also in our previous simulation in three southern case sites (Ge et al. 2012), the canopy size will be more expansive under CC than that under CUR, resulting in a further higher water consumption of canopy evaporation and transpiration demand. At the same time, the anticipated higher temperatures will likely lead to a substantial reduction in snow accumulation due to a decreased fraction of precipitation as snow and to later snowfall and earlier snowmelt (Kellomäki and Väisänen 1996), which could reduce the recharging of soil water in the spring and early summer. These factors led to a reduction in the infiltration of water into the soil profile and thus increased the water deficit in the rooting zone.

During the middle and last simulation periods, the much increased evapotranspiration and reduced water infiltration into the soil profile increased the occurrence of drought periods and decreased GPP and NPP in the Norway spruce forests, as with our simulation in southern Finland. The ecophysiological core of the FinnFor model is the Farquhar-type carbon assimilation model and Jarvis-type stomatal conductance model (Ge et al. 2010), allowing the simulation of how varying atmospheric and edaphic conditions affect stomata-photosynthesis interactions. For instance, high vapor pressure deficit and soil water stress will limit stomatal behavior and the consequent carbon uptake and tree growth. The conclusion has been supported by the previous field experiments in the Swedish (Asa and Flakaliden) and Finnish (Heinola and Sahalahti) research sites by Phillips et al. (2001) and Jyske et al. (2010), in which it was found that the growth of Norway spruce is water-limited regarding decreased carbon uptake and wood properties.

Although the increase in atmospheric CO2 is expected to enhance the primary production in forests, this study is based on the hypothesis that the enhancement of carbon uptake and tree growth by elevated CO2 is dependent on the availability of growth resources such as water and nutrients (Ainsworth and Long 2005; Ge et al. 2010; McCarthy et al. 2010). According to the FACE meta-analysis, the stimulation of tree photosynthesis and gross productivity under CO2 enrichment was short-lived, probably due to light limitation (i.e., canopy closure), water and nitrogen deficit and down-regulation of photosynthesis (Ainsworth and Long 2005). Furthermore, the predicted soil drought will result in the reduction of litter and humus production, simultaneously affecting the long-term nitrogen availability for tree growth (Ge et al. 2010).

As reported by Battles et al. (2008) with the CACTOSclim model, the intensity and extent of the summer drought and moisture deficit under climate change are considered to be limiting factors in the gross productivity and viability of the coniferous forests. Loustau et al. (2005) also used the process-based models (i.e., CASTANEA, GRAECO and ORCHIDEE) to study the impact of climate change on GPP of the pine forests in France. It was predicted that the expected positive effect of CO2 elevation on forest growth and primary production was overcompensated by the increasing number of frequent and severe droughts during the growing season (because of a pronounced shift in seasonal rainfall from summer to winter). At the regional scale, Eastaugh et al. (2011) applied the species-specific adaptation of the biogeochemical model BIOME-BGC to Norway spruce across a range of Austrian climatic change zones, using input data from a number of national databases (NFI). The results showed that climate change has negligible effect on Norway spruce productivity, due to a warming trend over the past 50 years and little overall change in precipitation (Eastaugh et al. 2011).

Based on the FINADAPT climate scenario we used (Carter et al. 2005), the increase in precipitation will be higher in the north compared with the south. We found that there are differences in climate change response in the southern (zones I and II) and northern (zones III and IV) regions, especially for southernmost zone I and northernmost zone IV. In northern Finland, the GPP and NPP are higher under CC compared with CUR over the 100-year simulation period due to the increased total photosynthesis over the years. Photosynthetic production was further increased by the atmospheric CO2 enrichment, where soil moisture seldom limits forest growth.

Temperature sum is one of the most important climatic factors for decision-making of the forest management in Finland because the growth and development of forests is greatly limited by low air temperatures and a short growing season. As simulated in northern Finland under CC, the increased temperature sum is positively correlated with primary production and carbon sequestration of the forests due to lengthened growing season, although the relationship was not very strong. This on the other hand indicated that the growth of forests is limited by multi-resource factors (availability of water, nutrient, light and so on) in relation to age class and resource demand of tree. However in southern Finland, higher temperature sum under CC do not enhance the primary production and carbon sequestration in the Norway spruce forests, even with weak negative effects. This unusual output suggested that the adverse effect of drought might counteract the stimulation of forest growth under warming environment.

NEE is quantified as the net increment of carbon based on the difference between GPP and TER. Establishing relationships between GPP and TER is important for predicting the impacts of climate change on the future net sequestration of carbon by the Norway spruce forest ecosystem. In this study, the changes of NEE are also distinguished by the southern and northern regions according to the results of GPP and NPP under climate change. Nevertheless, NEE is relatively more and less sensitive to climate warming in the southern and northern regions, respectively. The probable reason is that the stimulated TER under climate warming (data not presented) further reduces NEE in zones I and II. Higher carbon loss of ecosystem respiration also partially offsets the increased GPP in zones III and IV. Therefore, the changes of NEE are insensitive to climate change in northern Finland.

In southern Finland, climate change may create a suboptimal environment for Norway spruce, which may maintain the forest productivity on the most fertile sites (e.g. sites of OMT) with sufficient water supply. As reported by Kellomäki et al. (2008), the drought-tolerant boreal species of Scots pine and birch may increase on less fertile sites (e.g. sites of MT) currently occupied by Norway spruce. Zang et al. (2012) also suggested that spruce is adapted less efficiently to the increasing water stress, compared to pine and other deciduous species. In this condition, Norway spruce is probably less competitive with other boreal species under climate change.

In conclusion, the expected changing climate will affect the water and nutrient budgets along with the consequent forest growth and carbon sequestration. Finnish forests have been intensively managed for timber production for a long time, but future decision-making of forest managers may have to change from that under the current climatic conditions. The problems faced in forest management under the climate change are how to maintain and enhance the capacity to sequester and store carbon in the ecosystems, and at the same time to meet the needs of timber production. The most important suggestion is how to optimize the regional suitability of management regimes for Norway spruce from southern to northern Finland for adaption to climate change. For instance, wider spacing and shorter return periods between thinnings could facilitate sustainable growth and the avoidance of a higher water demand in south (e.g. Bréda et al. 1995; Misson et al. 2003). An alternative cost-effective way would be to choose the drought-tolerant ecotypes, in order to adapt the forest to climate change as regards the potential condition of frequent drought.

Acknowledgments

This work was funded through the Finland Distinguished Professor Programme (FiDiPro No. 127299–A5060–06), the National Technology Support Program (No. 2010CB951204) and the National Natural Science Foundation of China (No. 41201091). Dr. David Gritten and the American Journal Experts are gratefully acknowledged for revising the language of this paper.

Supplementary material

10584_2012_607_MOESM1_ESM.doc (253 kb)
ESM 1(DOC 253 kb)

Copyright information

© Springer Science+Business Media Dordrecht 2012