• Growth and nitrogen (N) economy of mountain birch are reported here in response to temperature change. Mechanisms of temperature effects on plant growth in temperate–arctic regions are discussed in the light of decreasing growth rates and increasing leaf-N contents along altitudinal and latitudinal temperature gradients.
• Mountain birch ( Betula pubescens ssp. czerepanovii) seedlings were grown at two soil temperatures, air temperatures and nutrient concentrations in a full-factorial experiment during one growing season in northern Sweden.
• Changes in air and soil temperature affected aboveground growth more than belowground growth. An increase in air temperature increased leaf area ratio and plant-N productivity while decreasing plant-N concentration and leaf-N content. A change in soil temperature affected root-N uptake rate and plant-N concentration, similar to the effect of a change in nutrient supply. Air and soil temperature had interactive effects on growth rate, N productivity and leaf-N content.
• The results indicate that increasing leaf-N content with increasing altitude and latitude is not only a passive consequence of weaker N dilution by reduced growth, but also a physiological acclimation to lower air temperature.
Low temperatures are prominent factors determining plant growth in temperate and boreal regions, especially near altitudinal and latitudinal distribution limits (Chapin, 1983; Körner & Larcher, 1988; Körner, 1998). Temperature affects many different processes involved in plant growth. Some of these are probably predominantly influenced by belowground conditions, such as root growth and nutrient uptake, while others could be expected to be affected by aboveground temperature such as leaf growth and carbon fixation. Where these processes act in an additive way, the qualitative effect of increased temperature on plant growth might be simple to predict and similar for raised air and soil temperature. We expect that increases in both air and soil temperature will result in an increase in aboveground growth rate, because raised air temperature might particularly favour shoot growth and higher soil temperature might stimulate nutrient acquisition and thereby growth rate. Since improved nutrient supply commonly accompanies a greater allocation of biomass to above- than belowground plant parts (cf. Ericsson, 1995), increases in both air and soil temperatures should enhance biomass allocation to shoots and result in a similar pattern to that seen when nutrient supply is increased.
Many effects of temperature on plants might act in a complex manner. Major effects of raised air temperature are enhanced leaf expansion rate (Ong & Baker, 1985; Woodward et al., 1986) and shoot growth rate (Larigauderie et al., 1991), whereas elevated soil temperature primarily might increase root growth (Tryon & Chapin, 1983; Larigauderie et al., 1991) and nutrient uptake rate (Chapin, 1974; Kummerow & Ellis, 1984; Karlsson & Nordell, 1996). Since plant processes above- and belowground are intimately linked, the ultimate effects of increased temperature on whole-plant growth and single growth parameters often are not obvious. For example, an increase in temperature might cause the plant N concentration to increase or decrease, depending on the relative effects of increased soil temperature on root N uptake and of raised air temperature on aboveground growth and nutrient dilution. It is therefore difficult to discuss the influence of temperature on plant growth processes and/or landscape patterns of plant nutrient concentrations without separating air and soil temperature effects on various plant growth parameters. In addition, an analysis of the differential effects of soil and air temperature may help to resolve the question whether a decrease in plant N uptake with lower temperature is mainly caused by either weaker nutrient demand of reduced tissue growth under lower temperature (Körner, 1998), or stronger limitation of root N uptake in colder soils (Weih & Karlsson, 1999a). Also, leaf N concentrations have been discussed in the literature with respect to altitudinal and latitudinal temperature gradients (Körner, 1989; Sveinbjörnsson et al., 1992). However, it is unclear whether higher leaf N concentration along with lower temperature reflects only a passive consequence of reduced leaf growth and weaker dilution of the nutrient pool (Körner & Larcher, 1988) or also a physiological adaptation to low-temperature environments (Weih & Karlsson, 1999b). In order to better understand the mechanisms of temperature limitation of plant growth in high-altitude/latitude environments, experimental designs are needed where air and soil temperature are manipulated separately and using plants that are frequently growth-limited by low temperature in their natural environments.
Most studies of the effects of increased temperature on plant growth have used grasses or plants that have their growing points in contact to the soil surface (Kummerow & Ellis, 1984; DeLucia et al., 1992). A direct effect of soil temperature on shoot growth cannot be ruled out in experiments involving plants that have their growing points not spatially separated from the soil. We designed an experiment to investigate the effect of separately varied soil and air temperature on whole-plant growth of a species that has its growing points above the ground surface and is adapted to low-temperature environments. Seedlings of mountain birch (Betula pubescens ssp. czerepanovii), a common tree-line species in North-west Europe, were chosen for this experiment. Earlier studies using this species have shown that an increase in soil temperature results in increased root N uptake, plant N accumulation rate, whole-plant growth rate and, finally, winter survival of young seedlings (Karlsson & Nordell, 1996; Weih & Karlsson, 1999a). In addition, decreased temperature (of both air and soil) was associated with decreased leaf N productivity and increased leaf N content (area basis), suggesting a physiological link between leaf N concentration, N use efficiency and leaf temperature (Weih & Karlsson, 1999b). However, air and soil temperatures were not manipulated separately in any of these experiments.
The objective of this study was to investigate the independent effects of air and soil temperature on various growth parameters of young birch seedlings, and to compare these effects with those caused by increased nutrient supply. We tested the following specific hypotheses: both air and soil temperature might increase aboveground more than belowground growth rate; root N uptake rate should be mainly affected by soil temperature, whereas leaf N content should be more influenced by air temperature; soil and air temperature should affect whole-plant N concentration in opposite ways.
Materials and Methods
Location, plant material, experimental design and growth conditions
The experiment was conducted at the Abisko Scientific Research Station in northern Sweden (68°21′ N, 18°49′ E, 380 m above sea level (asl)) during the 1994 growing season. Seeds of the mountain birch (Betula pubescens Ehrh. ssp. czerepanovii (Orlova) Hämet-Ahti, also called B. p. ssp. tortuosa (Ledeb.) Nyman) were collected from the local population near Abisko. Seeds were sown in peat on 25 May 1994, and the seedlings were pregrown in a tent for 2 wk, until the beginning of the experiment on 9 June 1994. At this time, the mean seedling biomass (±1 SD) was 0.4 ± 0.1 mg (n = 15) and the mean leaf area (cotyledons) was 0.3 cm2. The experiment lasted for 11 wk, until 24 August 1994. During the experiment, seedlings were grown separately in 0.4-dm3 plastic containers of peat. The experimental treatments were four regimes differing in air and soil temperature (Table 1); and two fertilization regimes corresponding to 2 (FL fertiliser low) and 5 g N m−2 season−1 (FH high). The temperature treatments were arranged in four blocks. Each block consisted of two boxes filled with sand (each sandbox approx. 3.0 × 1.6 m in size and 0.5 m in depth), one with ambient and one with increased soil temperature. Each sandbox was divided into two parts assigned to ambient and increased air temperature treatments, respectively. The pots containing the seedlings were placed in the sandboxes and sunk in the sand bed of seven rows formed by the space between heating cables (see below). The replicates of the two fertilization treatments were randomly positioned within the four temperature treatments. Air temperature was increased by using tents, which were covered with clear 0.05-mm polyethylene plastic sheeting. The tents were ventilated and open in order to keep the humidity and CO2 concentration inside the tents similar to the conditions outside. The photosynthetically active radiation (PAR, 400–700 nm), the humidity and the CO2 partial pressure were regularly recorded inside and outside the tents. The PAR was between approx. 12% (at high sun elevation) and 20% (at low sun elevation) lower inside the tents than outside. Soil temperature was increased by heating cables placed at a distance of 23 cm within the sandboxes. The heating cables were standard Devi warming cables that have a heat level of 11 W/m of lineal heat and an overall heat output of roughly 70 W/m2. There were three levels (−1, −12, −35 cm) of cable to provide a uniform soil temperature. Air and soil temperatures were recorded every 10 min with thermistor sensors (Campbell Scientific type 107, Campbell Inc., Logan, UT, USA) connected to a datalogger. The temperature control worked by reading the temperature from four thermistors – 3 cm belowground – and two thermistors – 8 cm aboveground – in each sandbox. The temperature difference between each control and heated sandbox-pair was used to control soil heating. The temperature difference was controlled every five min. The nutrient solutions were prepared using a balanced, complete fertiliser (Rika, Weibulls, Sweden), with NH4+ and NO3− in equal proportions and N, P, K, and Mg in the proportions 5 : 1 : 4 : 0.4. Nitrogen concentrations in the solutions were 2.6 (FL) and 6.5 mM (FH). The fertiliser solutions were supplied in 16-mL portions at weekly intervals, and seedlings were watered daily. The fertilization levels supplied each seedling with approx. 6.8 (FL) and 17 mg N (FH) during the experimental period of 11 wk. The fertiliser regimes simulated nutrient conditions in a mountain birch woodland poor and rich in nutrients (Weih, 1998; Weih & Karlsson, 1999a).
Table 1. Mean air and soil temperatures (± SE) recorded for various treatments in a growth experiment where air and soil temperature were manipulated separately. Air temperature was measured 8 cm aboveground and soil temperature was recorded 3 cm belowground
Treatment A, air temperature; S, soil temperature; subscript L, low: H, high; n = 4.
12.6 ± 0.3
9.5 ± 0.2
17.8 ± 0.2
9.9 ± 0.2
13.7 ± 0.3
12.9 ± 0.2
18.8 ± 0.3
13.6 ± 0.3
Harvest procedure and chemical analysis
Twenty-five seedlings were used for an initial harvest, and between 12 and 14 replicates per treatment were harvested at the end of the experiment. All harvested plants were fractionated into leaves, stems, and roots. Leaf area was determined on fresh leaves using an image analysis system (DIAS, Delta-T Devices, Cambridge, England). Dry weight was determined after 36 h of freeze drying (Lyovac GT2, Leybold-Heraeus, Köln, Germany). Total N content of the whole plants (initial harvest) or of the different plant fractions (leaves, stems and roots; final harvest) was analysed after a Kjeldahl digestion. Nitrogen concentrations were subsequently determined photometrically by flow injection analysis (FIAstar 5012, Perstorp AB, Höganäs, Sweden).
In the experiment described here, the period between the initial and the final harvest represented a period of continuous growth, and we used a functional growth analysis approach to compare seedling growth and N economy between the two harvests (cf. Hunt, 1982). Thus, possible differences in relative growth rate (RG,% d−1) between seedlings were related to differences in relative N accumulation rate (RN,% d−1), leaf area productivity (PLA, gm−2 d−1; also called the unit leaf rate), whole-plant and leaf N productivity (PN and PLN, g [mol N]−1 d−1), root N uptake rate (UN, mmol N [g root]−1 d−1), leaf area ratio (LAR, cm2 g−1), specific leaf area (SLA, cm2 g−1), plant N concentration (PNC,%), leaf N content (LN/LA, µmol N m−2) and leaf N allocation (LN/N,%). The following relationships among traits of growth and N economy were utilized (cf. Lambers et al., 1990; Garnier et al., 1995):
ANOVA (GLM procedure, SPSS Release 9.0, SPSS Inc., Chicago, Illinois) was used to assess the effects of the fixed factors air temperature, soil temperature and fertilization, the random factor block, and of interactions between them. We assumed that the seedlings in all experimental treatments had a similar biomass and total N pool at the initial harvest in this experiment. Therefore, the loge-transformed biomass and N pool data of the final harvest were used to assess differences in RG and RN between the treatments.
Seedling height, aboveground biomass and number of buds were increased by both elevated air and soil temperature (Tair and Tsoil, respectively) and increased nutrient supply (Fig. 1, Table 2). Whole-plant relative growth rate (RG) was mostly affected by Tsoil, whereas increased Tair resulted in an increase in RG only at low soil temperature (Fig. 2a and the A × S term in Table 2).
Table 2. ANOVA for effects of fixed factors air temperature (A), soil temperature (S), fertilization (F) and random factor block on various growth traits and N economy of young mountain birch seedlings experimentally grown at Abisko, northern Sweden during the 1994 growing season. Interaction effects including the factor block contributed very little to variation and were suppressed in the ANOVA model
† log e transformed. d.f., degrees of freedom; SS, sums of squares; P, significance level.
Within + residual
Air temp. (A)
Soil temp. (S)
A × S
A × F
S × F
A × S × F
Increase in Tsoil and Tair and nutrient supply all affected RG through increased allocation to leaf area, both on a whole-plant basis (LAR) and leaf basis (SLA) (Fig. 2, Table 2; Eqn 1). The productivity per unit leaf area (PLA) was increased by elevated Tsoil, but unaffected (SL) or decreased (SH) by elevated Tair (Fig. 2). Thus, increase in Tair at the high soil temperature resulted in increased LAR in parallel with decreased PLA and, finally, similar RG.
Similar to the pattern in RG, whole-plant N accumulation rate (RN) was increased by higher Tsoil but also nutrient supply, whereas raised Tair resulted in increased RN only at low soil temperature (Fig. 3 and the A × S term in Table 2). The soil temperature effect was stronger on RN than on RG and, consequently, increased plant N concentration (PNC; Fig. 3, Table 2). In contrast, raised Tair resulted in nutrient dilution and decreased PNC. The pattern in PNC was different from that in leaf N content (area basis, LN/LA). Thus, LN/LA was strongly decreased by elevated Tair, but either increased (AL) or decreased (AH) by elevated Tsoil (Fig. 3 and the A × S term in Table 2). In particular at the higher Tsoil the strong decrease in LN/LA caused by raised Tair cancelled out its positive effect on productivity per unit N (PN) (Fig. 3). Ultimately, this effect was responsible for the pattern of increased PLA and unchanged RG seen in the Tair contrast at the elevated soil temperature.
Whole-plant PN and leaf-level PLN were significantly correlated (Pearson r = 0.89, P = 0.003, n = 8; Eqn 2). The relationship between PLN and LN/LA varied among seedlings grown at different air temperature (Fig. 4). Thus, increase in Tsoil resulted in higher LN/LA at the low air temperature and the PLN was positively related to LN/LA up to an optimum LN/LA and decreased thereafter. In contrast, increase in Tsoil resulted in lower LN/LA at the high air temperature and PLN generally was decreased with increased LN/LA.
In this study we chose to use first-year seedlings of mountain birch that were growing close to their natural growth rate. The first growing season is more critical than older seedling stages for mountain birch survival (cf. Sveinbjörnsson et al., 1996), which strongly depends on N uptake and soil temperature (Weih & Karlsson, 1999a; Karlsson & Weih, 2001). Therefore, temperature effects on mountain birch growth and allocation patterns are of particular interest during the critical juvenile stage. In addition, many of the results seen in this study might reflect temperature response patterns of various plants growing in temperate-arctic regions.
Considering only aboveground seedling growth, an increase in temperature – irrespective of air or soil temperature – always resulted in higher plant growth (cf. Figure 1). Thus, increase in air and soil temperature affected some growth parameters in an additive way and these temperature effects were similar to the growth response to elevated nutrient supply. Specifically, stem height, leaf area allocation (LAR) and aboveground biomass allocation were enhanced by increases in both air and soil temperature and no significant interaction effect was found. In addition, elevated air and soil temperature increased the number of leaf buds (Fig. 1b) and should also affect seedling growth rate in future, since the number of buds appears to be a major determinant of growth rate of mountain birch seedlings during the subsequent growing season (Weih, 2000a). The results are in agreement with those of many field experiments simulating the effects of climate warming on aboveground growth of various plant species (Parsons et al., 1994; Graglia et al., 1997; Hobbie et al., 1999). Nevertheless, field studies of that type commonly deal with single plant parts and the type of experimental manipulation often affects soil temperature only marginally. Therefore, the effects of increased air and soil temperature on whole-plant growth might go undetected, especially when they interact with each other as was partly shown in this study.
Soil temperature effects
Increase in soil temperature affected growth of the seedlings in a similar way to the increase in nutrient supply, owing to a great influence of soil temperature on total plant N accumulation rate (RN). Strong effects of soil temperature on root N uptake rate and internal nutrient transport rate have been reported for many plant species (Chapin, 1974; Kummerow & Ellis, 1984; DeLucia et al., 1992) including the mountain birch (Karlsson & Nordell, 1996; Weih & Karlsson, 1999a). Weih & Karlsson (1999a) used a simple model to predict root N uptake rate (UN) of experimentally grown mountain birch seedlings as a function of the mean soil temperature. The data of this study fit this model well, up to a mean soil temperature of approx. 15°C, which was the upper limit of the valid range for model application (Fig. 5a) and also near the highest soil temperatures recorded in the field (Karlsson & Weih, 2001). Increase in soil temperature above this range resulted in weaker increase of UN or even its decrease, which was reflected by a similar pattern in plant N accumulation rate (RN).
Several studies pointed on the importance of a pronounced soil temperature effect on root growth rate (Tryon & Chapin, 1983; Larigauderie et al., 1991). We also found higher root growth rate in response to elevated soil temperature, but shoot growth always was stimulated more than root growth and root biomass allocation was therefore decreased.
Air temperature effects
Increase of air temperature by using tents was inevitably associated with a small rise in soil temperature by 1 K (Table 1). The difference in soil temperature could explain the small increase of root N uptake rate seen in the tented treatments (Fig. 5; Weih & Karlsson, 1999a). Nevertheless, pronounced effects of air temperature were found on leaf area allocation (LAR and SLA), N productivity (PN and PLN) and leaf N content (LN/LA). Thus, increase in RG in response to elevated air temperature was accomplished primarily by larger leaf area and higher productivity per unit N, and was associated with a dilution of the internal N pool. This dilution effect of increased air temperature could, in part, explain the decrease in leaf N content (LN/LA).
Increased leaf N content at a lower temperature might, however, also reflect acclimation with respect to optimal N use. Hikosaka (1997) studied the physiological processes involved in the photosynthetic acclimation of leaves of C3 plants to temperature with respect to N use. According to a model applied by Hikosaka (1997), the optimal leaf N content (area basis) increases with decreasing air temperature. Also for mountain birch, Weih & Karlsson (1999b) provided evidence that the optimal leaf N content should be lower at a higher air temperature than at a lower air temperature. The results of this study are consistent with the general model by Hikosaka (1997) and support the arguments by Weih & Karlsson (1999b). Thus, under a higher air temperature, the maximum leaf N productivity (PLN) was achieved at lower leaf N content (LN/LA) than under a lower air temperature (Fig. 4). Moreover, there was a remarkable similarity in the PLN– LN/LA relationship between the two different studies (see Fig. 4b in Weih & Karlsson, 1999b). Therefore, the results of this study provide further evidence that higher leaf N concentration along with increasing altitude or latitude (Körner, 1989; Sveinbjörnsson et al., 1992) is not only a passive consequence of reduced leaf growth but, at least in part, a physiological adaptation to low-temperature environments.
It should be remembered that an increased air temperature has effects on leaf phenology and, thereby, the duration of the period of active plant growth (Pop et al., 2000). The design of the present study does not allow to evaluate any effects of temperature on seedling phenology. However, mountain birch growth rates are generally low and the growth conditions during the growing season possibly have larger effects on annual biomass production of seedlings than the prolongation of the growth period by one or two weeks (Weih, 2000b).
Air and soil temperature effects on all whole-plant growth parameters interacted with each other. These results are in line with those of Larigauderie et al. (1991). Interaction effects were thus found on RG, N accumulation rate (RN) and all productivity traits (PLA, PN, PLN). Apparently, increase in air and soil temperature alone had a greater effect on total biomass production than in combination. This interaction effect was caused by: a weaker effect of elevated soil temperature on root N uptake rate (UN) at soil temperatures > 15°C, and; a great effect of elevated air temperature on leaf N content, and, thereby, leaf area productivity (PLA), especially at a high soil temperature.
Within the natural distribution range of the mountain birch in northern Europe, soils are likely to be colder than 15°C in most cases (Karlsson & Weih, 2001), except for early successional patches covered by little vegetation (Weih, 2000b). Growth reductions caused by a diminished effect of soil temperature on root N uptake rate might therefore be less frequent for mountain birch seedlings growing under natural conditions.
Mean air temperature recorded by the meteorological observatory at Abisko varied between 8 and 12°C during the last years (June to August mean 1961–90). The air temperature near ground experienced by young tree seedlings grown in the field might be somewhat higher than those measured meteorologically, making them more similar to those used in this study (Table 1). Therefore, growth reductions caused by the effect of low air temperature on leaf N content and leaf N productivity might represent a pattern seen in mountain birch seedlings growing under natural conditions. If this pattern is generalized, a hypothesized physiological link between air temperature, leaf N productivity and leaf N content (Weih & Karlsson, 1999b) could partly explain why low growth rate of high-altitude and high-latitude plants is associated with high leaf N content, although these parameters commonly are positively correlated.
In accordance with the hypotheses of this paper, increased air and soil temperature enhanced aboveground growth rates more than belowground growth rates. Therefore, temperature effects on plants seen aboveground could be a combination of changed allocation and production and might not necessarily be correlated with whole-plant production. The major effect of elevated soil temperature on seedling growth was increase in root N uptake, which was followed by increase in plant N concentration and, finally, relative growth rate. Increase in soil temperature affected seedling growth in a similar way as increase in nutrient supply does. Thus, high growth rate and also winter survival of young seedlings could be achieved by either high nutrient availability in combination with low soil temperature or low nutrient availability in combination with high soil temperature (cf. Weih & Karlsson, 1999a). The major effects of increase in air temperature were increases in shoot growth and leaf area allocation, and decreases in plant N concentration and leaf N content (area basis). The decrease in leaf N content was particularly great at a high soil temperature. The interactive effect of air and soil temperature on leaf N content corresponded to an interaction effect seen on whole-plant relative growth rate. Leaf N content was primarily affected by air temperature and unrelated to plant N concentration, root N uptake rate and plant N accumulation rate. Therefore, the results suggest that increasing leaf N content along with cooler environments is not only a consequence of weaker N dilution by reduced growth, but also reflects a physiological adaptation to low-temperature conditions.
We are grateful to Francis P. Bowles and Bengt Wanhatalo, who constructed the growth facilities. Financial support was obtained from the Swedish Natural Science Research Council and the Royal Swedish Academy of Sciences.