Impacts of seasonal air and soil temperatures on photosynthesis in Scots pine trees

Tree Physiology 22, 839 847 2002 Heron Publishing Victoria, Canada Impacts of seasonal air and soil temperatures on photosynthesis in Scots pine trees MARTIN STRAND, 1,2 TOMAS LUNDMARK, 3 INGRID SÖDERBERGH 3 and PER-ERIK MELLANDER 3 1 Umeå Plant Science Centre, Department of Forest Genetics and Plant Physiology, The Swedish University of Agricultural Sciences, SE-901 83 Umeå, Sweden 2 Author to whom correspondence should be addressed (Martin.Strand@genfys.slu.se) 3 Vindeln Experimental Forests, The Swedish University of Agricultural Sciences, SE-922 91 Vindeln, Sweden Received September 21, 2001; accepted March 2, 2002; published online July 2, 2002 Summary Seasonal courses of light-saturated rate of net photosynthesis (A 360 ) and stomatal conductance (g s ) were examined in detached 1-year-old needles of Scots pine (Pinus sylvestris L.) from early April to mid-november. To evaluate the effects of soil frost and low soil temperatures on gas exchange, the extent and duration of soil frost, as well as the onset of soil warming, were manipulated in the field. During spring, early summer and autumn, the patterns of A 360 and g s in needles from the control and warm-soil plots were generally strongly related to daily mean air temperatures and the frequency of severe frosts. The warm-soil treatment had little effect on gas exchange, although mean soil temperature in the warm-soil plot was 3.8 C higher than in the control plot during spring and summer, indicating that A 360 and g s in needles from control trees were not limited by low soil temperature alone. In contrast, prolonged exposure to soil temperatures slightly above 0 C severely restricted recovery of A 360 and especially g s in needles from the cold-soil treatment during spring and early summer; however, full recovery of both A 360 and g s occurred in late summer. We conclude that inhibition of A 360 by low soil temperatures is related to both stomatal closure and effects on the biochemistry of photosynthesis, the relative importance of which appeared to vary during spring and early summer. During the autumn, soil temperatures as low as 8 C did not affect either A 360 or g s. Keywords: carbon dioxide assimilation, frost, gas exchange, low temperature, Pinus sylvestris, stomatal conductance. Introduction Evergreen conifers of the boreal forest have pronounced seasonal variations in their rates of carbon dioxide assimilation, which generally decrease during autumn and early winter, depending on the severity and frequency of night frosts (e.g., Havranek and Tranquillini 1995). Winter inhibition of photosynthesis involves inhibition of photochemical and enzymatic reactions at the chloroplast level, which depresses the photosynthetic potential of the plant (e.g., Öquist and Martin 1986). In addition to low and subfreezing temperatures, winter inhibition of photosynthesis is affected by the light regime to which the needles are exposed (Strand 1995, Strand and Lundmark 1995, Lundmark et al. 1998). Recovery from winter inhibition of photosynthesis coincides with increasing temperatures during spring and early summer (Lundmark et al. 1988b, 1998, Strand and Lundmark 1995, Bergh and Linder 1999). Low daytime air temperatures and freezing nights can adversely affect the recovery process (DeLucia and Smith 1987, Lundmark et al. 1988b, 1998, Bergh and Linder 1999). Furthermore, soil frost and low root-zone temperatures during spring may influence the recovery of net photosynthesis (Troeng and Linder 1982, DeLucia and Smith 1987, Bergh and Linder 1999). Soil frost severely restricts or prevents uptake of water by roots (Tranquillini 1982). In addition, low root-zone temperatures affect the movement of water through soil and roots by increasing its viscosity and decreasing the permeability of roots to water (see Kramer 1983). Low root-zone temperatures also decrease the growth of roots (see Bowen 1991), which can reduce root water flow (Wan et al. 1999). Thus, frozen soils and low root-zone temperatures may induce water deficits in needles of evergreen conifers, manifested as low xylem water potentials, especially when the evaporative demand is high during spring (Cowling and Kedrowski 1980, Berg and Chapin 1994). However, persistent dehydration damage to needles of evergreen conifers during winter and early spring appears to be uncommon in the boreal forest because stomata remain closed throughout most of the winter (Grace 1990, Havranek and Tranquillini 1995). Furthermore, minor losses of water through the cuticle of needles can be replaced by water stored in stems during periods with above freezing temperatures (Sowell et al. 1996, Boyce and Lucero 1999). Stomatal conductance to water vapor (g s ) and net photosynthesis in evergreen conifers have been correlated with soil temperature during spring and early summer as well as during autumn (DeLucia and Smith 1987, Carter et al. 1988, Körner

840 STRAND, LUNDMARK, SÖDERBERGH AND MELLANDER 1994, Schwarz et al. 1997). In laboratory experiments, net photosynthesis and g s in seedlings of conifers decrease sharply at low root-zone temperatures (DeLucia 1986, Day et al. 1991). DeLucia et al. (1991) found that transferring Scots pine (Pinus sylvestris L.) seedlings from a soil temperature of 15 C to a soil temperature of 5 C for 24 h or more caused significant reductions in net photosynthesis and g s. The mechanism of inhibition of net photosynthesis at low soil temperatures appears to be complex. Partial stomatal closure may decrease net photosynthesis by reducing the intercellular concentration of CO 2 (Day et al. 1989, 1991, DeLucia et al. 1991). In addition, effects on the biochemistry of photosynthesis have been reported (DeLucia 1986, Day et al. 1991, DeLucia et al. 1991), especially at temperatures close to 0 C. However, relatively few experiments have been carried out in the field to determine the importance of soil temperature for gas exchange in evergreen conifers (Day et al. 1989, 1990, Bergh and Linder 1999). Soil frost in combination with a delayed snowmelt in northern Sweden might affect the recovery of gas exchange in evergreen conifers during spring either by limiting the availability of water or by delaying the onset of soil warming. To test this hypothesis, the extent and duration of soil frost as well as the onset of soil warming were manipulated in the field in a young stand of Scots pine. We examined the seasonal courses of light-saturated rate of CO 2 assimilation (A) and g s at ambient (360 µmol mol 1 ) and elevated (2000 µmol mol 1 )CO 2 concentrations in detached 1-year-old needles from early April to mid-november. Materials and methods Experimental site and plant material This study took place at Västomån (64 13 N, 19 41 E, altitude 185 m above sea level), located about 60 km northwest of Umeå in northern Sweden. The experiment was established in an open stand dominated by 20-year-old naturally generated Scots pine. The soil is a ferric podzolized sandy silt with low clay content. The climate of the area is characterized by cold winters with persistent snow cover and short growing seasons (Odin 1992). Treatments Three plots, each about 30 m 2 in area with three study trees, were selected and scaffolding was erected to provide access to the tree crowns. One plot served as a control. On the remaining plots the extent of soil frost was manipulated during the winter of 1998 1999. Soil frost was minimized on the warm-soil (WS) plot by insulating the ground with porous cloth bags filled with styrofoam pellets in late September 1998. The insulation was removed at the end of snowmelt in 1999 (April 24; cf. Figure 2A). To delay cooling of the soil during autumn 1999, the WS plot was covered with a 10-cm layer of sawdust at the end of August 1999. Trees on the cold-soil (CS) plot were subjected to a deep soil frost by removing most of the snow from February 11 to March 24, 1999. During this period, the temperature at the soil surface averaged 3.2 and 1.5 C at the CS and control plot, respectively. At the beginning of snowmelt (March 24), the snow was replaced and covered with a 10-cm layer of sawdust to delay snowmelt. The sawdust was removed in early July when the soil temperature at a depth of 10 cm had reached 6.4 C. Climatic data and soil measurements Air temperature (at 1.7 m above ground) and precipitation data were obtained from a reference weather station within the Vindeln Experimental Forests, about 5 km from the study site. The weather station was situated in an open heath of Scots pine with a climate similar to that of the study site (data not shown). Soil temperature and soil volumetric water content were measured at the soil surface and at depths of 10, 20, 40, 60 and 90 cm. Soil temperature was measured at three locations in each plot with thermistors. Soil volumetric water content was measured at one location in each plot with water content reflectometers (Model CS 615, Campbell Scientific, Logan, UT). Sensors were read automatically every 10 min and 2-hour means were calculated and stored by a data logger (Model CR10, Campbell Scientific). Gas exchange measurements Determinations of CO 2 and water vapor exchange were made with a portable computerized open-system IRGA (LI-6400, Li-Cor, Lincoln, NE). Three branches from each tree (n = 9 per treatment) were sampled from April to November 1999 between 1000 and 1600 h on clear and partly cloudy days. The CS plot was excluded from the measurement program from September to November. Gas exchange was measured on detached needles of 1-year-old sun-exposed shoots from the upper third of the crown. Each measurement was completed within about 5 min. No decrease in stomatal conductance or net photosynthesis was observed within 10 min following detachment (data not shown). Ten needles were placed across the short dimension of a2 3cmleaf cuvette. Saturating light (1500 µmol m 2 s 1 ) was provided by an LED light source (Li-Cor 6400-02B) mounted on the leaf chamber and CO 2 concentration was controlled by the Li-Cor LI-6400 CO 2 injection system. Measurements were taken at near ambient air temperatures and relative humidities. The same branches on each tree were sampled on each occasion throughout the study period. After measurements of gas exchange, the needles were kept cool and moist for up to 24 h until the projected needle area was determined, based on the mean of five measurements made with a leaf-area meter (Li-Cor LI-3000). Light-saturated rate of CO 2 assimilation at a CO 2 concentration of 360 µmol mol 1 (A 360 ) or 2000 µmol mol 1 (A 2000 ), stomatal conductance (g s ) and intercellular CO 2 concentration (C i ) were then calculated on a projected leaf area basis. TREE PHYSIOLOGY VOLUME 22, 2002

IMPACT OF SEASONAL AIR TEMPERATURES ON PHOTOSYNTHESIS OF SCOTS PINE 841 Data analyses Statistical analyses were performed with StatView 5.1 software (SAS Institute, Cary, NC). Effects of treatment on gas exchange were evaluated by one-way analysis of variance and effects of date on gas exchange were evaluated by repeated measures analysis of variance. Each branch was treated as a replicate and homogeneity of variances was assessed with Bartlett's test. Means were compared with Scheffe's F-test and differences were considered significant at P 0.01. Results According to measurements at Vindeln Experimental Forests during the period 1980 1998, air temperatures and snow depth during the winter of 1998 1999 were within the normal range for the area (Kluge 2001). However, the ground became snow-free about two and a half weeks earlier than normal. Mean daily and minimum temperatures, mean daily soil temperature at a depth of 10 cm and precipitation from April to November 1999 are shown in Figures 1A C. The weather was cold from early April to mid-may with mean daily air temperatures between 0 and 5 C (Figure 1A). Severe frosts, defined as occasions when the minimum air temperature was below 6 C, were common during this period. Mean daily air temperature increased from mid-may to mid-june and the minimum temperature was higher than 4 C during this period. Mean daily soil temperature at a depth of 10 cm was 0.3 C at both the control and WS plots from early to late April (Figure 1B). However, mean daily soil water content at the same depth was higher at the WS plot than at the control plot during this period (Figure 2B). The main increase in soil temperature at the WS and control plots from mid-may to mid-june coincided with increasing air temperature. The CS treatment strongly affected soil temperature during spring and early summer (Figure 1B). Soil temperature was slightly below 0 C and the soil water content was below 6% (Figure 2B) until mid-april, and from late April to mid-june soil temperature remained at 0.3 C because of the delayed melting of the snow (Figure 2A). However, soil temperature was only slightly lower in the CS plot than in the control plot from late April to mid-may. Because of the delayed thawing of the soil in the CS plot, soil water content was also lower than in the control plot during this period, although the availability of soil water gradually increased from late April to mid-may (Figure 2B). After the snow on the CS plot had melted in mid-june, warming of the soil was relatively fast and in mid-july soil temperatures were similar to those at the control plot. Autumn 1999 was warmer than normal. Compared with means for the period 1970 1998, air temperatures were about 3, 2 and 7 C higher during September, October and November, respectively. Daily mean temperatures in the air and soil generally decreased from early September to late October (Figures 1A and 1B). However, mean air temperatures were generally higher than 5 C during September. Furthermore, minimum temperatures were higher than 3 C during this month. The first severe frosts occurred in early October. Another period with severe frosts occurred in late October and during the cold spell in mid-november air temperatures were also lower than 0 C during the day. Differences in mean daily soil temperature between the WS plot and the control plot were small during autumn. From September 1 to October 23, the mean difference in soil temperature was only 1.1 C and during the warm spell in late October and early November soil temperatures at the control and WS plots were similar, as a result of more efficient warming of the soil in the control plot than in the WS plot with its insulating layer of sawdust. There was pronounced seasonal variation in gas exchange. Light-saturated rate of CO 2 assimilation at near ambient CO 2 concentrations (A 360 ), stomatal conductance to water vapor (g s ), and intercellular CO 2 concentration (C i ) from early April to mid-november 1999 are shown in Figures 3A and 3C. Large variations in C i were observed when A 360 was below 1 µmol m 2 s 1 in early spring and late autumn, probably because of insufficient resolution in the measurements. Therefore, these values have been omitted from Figure 3C. At the beginning of the study in early April, A 360 and g s were close to zero in needles from all plots (Figures 3A and B). Although A 360 and g s increased after the high rainfalls in mid-april, values were low from late April to mid-may. The main recovery of A 360 and g s in needles from the control and WS plots occurred between mid-may and mid-june (Figures 3A and 3B). In general, the WS treatment had no significant effect on gas exchange during spring and summer, although mean soil temperature was 3.8 C higher in the WS plot than in the control plot from late April to mid-august. In contrast, the CS treatment had pronounced effects on gas exchange. Although A 360 and g s in needles from the CS and the control plots were not significantly different on some occasions from late April to mid-may, there was a significant overall reduction in these variables for needles from the CS plot according to repeated measures analysis of variance. Furthermore, the CS treatment severely restricted recovery of A 360 and especially g s from mid-may to mid-june when mean daily soil temperature remained at 0.3 C. Warming of the soil in the CS plot from mid-june to mid-july resulted in increases in A 360 and g s. However, A 360 and g s were still significantly lower in needles in the CS plot than in the control plot in mid-july when daily mean soil temperatures at the two plots were similar. In general, the WS treatment had no significant effects on A 360, g s and C i during autumn (Figures 3A C). During September, A 360, g s and C i were relatively constant when daily mean soil temperature declined to 8.7 and 7.7 C at the WS plot and control plot, respectively. A substantial reduction in both A 360 (44 and 56% in needles from the WS and control plots, respectively) and g s (63 and 74% in needles from the WS and control plots, respectively) occurred after the first severe night frost ( 8.7 C) on October 6. The C i was also significantly reduced. On this occasion, A 360 was significantly higher in needles in the WS plot than in the control plot. However, because A 360 tended to be higher in needles in the WS plot than in the control plot on many occasions during autumn, it is possible that the significant difference in A 360 between the treatments on TREE PHYSIOLOGY ONLINE at http://heronpublishing.com

842 STRAND, LUNDMARK, SÖDERBERGH AND MELLANDER Figure 1. Seasonal courses in (A) mean daily and minimum air temperature 1.7 m above ground, (B) mean daily soil temperature at a depth of 10 cm in the control (continuous line), warm-soil (broken line) and cold-soil (dotted line) plots, and (C) precipitation. Sampling dates for measurements of gas exchange at a CO 2 concentration of 360 µmol mol 1 (see Figure 3) are indicated in (A). October 6 was associated with sampling errors rather than with differences in soil temperature. Partial recovery of A 360 and g s occurred during the week following the severe frost and, after this recovery, C i values were similar to those observed at the end of September. However, C i decreased slightly again from mid-october to mid-november. This decline was associated with a nearly complete inhibition of gas exchange. Measurements of light-saturated rate of CO 2 assimilation at an external CO 2 concentration of 2000 µmol mol 1 (A 2000 ) are shown in Table 1. No significant differences in A 2000 between the treatments were observed in early May, but in late May A 2000 tended to be lower in needles in the CS plot than in the control plot. In mid-june and early July, A 2000 was significantly lower in needles in the CS treatment than in the control treatment. However, four of the nine samples collected from the CS plot in mid-june had C i values below 800 µmol mol 1, the concentration needed to saturate photosynthesis according to in situ measurements of the relationship between light-saturated rate of CO 2 assimilation and C i in late May and early June (data not shown). Discussion Scots pine in the boreal forest typically has net photosynthetic rates close to zero during much of the winter (Troeng and Linder 1982), presumably because of low daytime air temperatures and severe frosts during the nights. However, considerable rates of net photosynthesis have been observed in this and other evergreen conifer species under favorable conditions in early spring (Troeng and Linder 1982, Wieser 1997, Schaberg et al. 1998). If the soil is frozen, the water lost in transpiration TREE PHYSIOLOGY VOLUME 22, 2002

IMPACT OF SEASONAL AIR TEMPERATURES ON PHOTOSYNTHESIS OF SCOTS PINE 843 Figure 2. (A) Snow depth and (B) soil water content at a depth of 10 cm in the control (continuous line), warm-soil (broken line) and cold-soil (dotted line) plots during April and May 1999. is assumed to come from storage in stems (Troeng and Linder 1982, cf. Sowell et al. 1996). Falls of wet snow and rain may further improve the water balance in needles under these conditions (Cowling and Kedrowski 1980, cf. Katz et al. 1989). Differences in the availability of soil water had no significant effect on gas exchange before the high rainfalls in mid-april in the present study. Lower A 360 and g s values were observed in needles in the CS treatment than in the control treatment in late April (Figures 3A and 3B), which can be attributed to the low soil water content associated with the delayed thawing of soil in the CS plot (Figure 2B). However, it is difficult to explain why differences in gas exchange between the treatments remained when the availability of soil water at the CS plot gradually improved from late April to mid-may (Figure 2B). Because gas exchange can be sensitive to changes in soil temperature slightly above 0 C (Turner and Jarvis 1975), the differences in A 360 and g s between the CS and control treatments from late April to mid-may may be associated with the slightly lower soil temperature at the CS plot than at the control plot (Figure 1B). Freezing damage to roots in the CS plot during the winter is unlikely because a frost hardiness of 15 to 20 C has been reported for Scots pine roots during this period (Sutinen et al. 1998) and 9.8 C was the lowest mean daily temperature measured at the soil surface of the CS plot. This occurred on January 30, i.e., before any removal of snow. The patterns of A 360 and g s in needles from the control and WS treatments during spring and early summer are in accordance with the view that recovery of gas exchange in evergreen conifers is strongly dependent on mean daily air temperature and the frequency of severe night frosts. From early April to mid-may, low values of A 360 and g s were associated with low mean daily air temperatures, which generally remained below 5 C (Figures 3A and 3B; cf. Figure 1A). Particularly low values of A 360 and g s were observed after severe night frosts (minimum temperature less than 6 C) during this period. Furthermore, the main recovery in A 360 and g s in needles from the control and WS plots from mid-may to mid-june coincided with increasing mean daily and minimum temperatures during this period. The direct effect of an increase in leaf temperature from 13 C in mid-may to 21 C (Figure 3D) in mid-june was probably small, because a similar decrease in leaf temperature from 24 C in early September to 16 C in late September did not have a noticeable effect on A 360, g s or C i (Figures 3A C). This relatively small temperature dependence of net photosynthesis is consistent with previous studies in conifers (DeLucia and Smith 1987, Teskey et al. 1995), especially if some acclimation to prevailing temperatures is assumed. Consequently, the similar values of C i before and after recovery in needles from both the control and WS plots (Figure 3C) indicate that recovery of A 360 from winter inhibition involved considerable increases in the capacity for photosynthesis at the cellular level. Furthermore, A 2000 increased substantially during spring recovery even when corrected for variations in leaf temperature during the measurements (Table 1), which is consistent with measurements of light-saturated rate of oxygen evolution in Norway spruce (Picea abies (L.) Karst.) in the same geographical area (Strand and Lundmark 1995). The WS treatment had little or no effect on gas exchange during spring and summer, although soil temperature was 3.8 C higher on average at the WS plot than at the control plot from late April to mid-august (Figures 3A and 3B; cf. Figure 1B). This implies that A 360 and g s in needles from control trees was probably not limited by soil temperature alone and further strengthens the view that recovery of gas exchange during spring is mainly determined by mean air temperatures and the frequency of severe frosts (cf. Bergh and Linder 1999). Responses of trees to field manipulations of soil temperature have varied, probably because of other confounding factors in the field. Day et al. (1989) found that lodgepole pine (Pinus contorta Dougl. var. latifolia Engelm.) trees had a higher midday net photosynthesis and leaf conductance in warm soil than in cold soil. Furthermore, results from the first year of soilwarming experiments in irrigated and irrigated-fertilized plots of Norway spruce in the same geographical area as our experiment (Bergh and Linder 1999), indicate that the higher soil temperature at heated plots than at unheated plots (about 5 C) increases the light-saturated rate of net photosynthesis in detached shoots during the snowmelt period in May. Daily sap flow per unit cross-sectional area of the xylem was also higher in trees in the heated plots. In contrast, soil temperature had no effect on total daily water flow in the xylem of young Engelmann spruce (Picea engelmannii Parry) trees during the snowmelt period (Day et al. 1990). However, the diurnal pattern of sap flow, which was correlated with leaf conductance, differed between trees in cold and warm soils. TREE PHYSIOLOGY ONLINE at http://heronpublishing.com

844 STRAND, LUNDMARK, SÖDERBERGH AND MELLANDER Figure 3. Seasonal courses in (A) light-saturated rate of CO 2 assimilation, (B) stomatal conductance to water vapor, and (C) intercellular CO 2 concentration (C i ) in 1-year-old needles of Scots pine (Pinus sylvestris L.) from control (, ), warm-soil (, ) and cold-soil (, ) plots during spring summer (open symbols) and autumn (filled symbols). Means and SEs are shown. (D) Means and SEs of leaf temperature during measurements of gas exchange in spring summer ( ) and autumn ( ). Leaf air vapor pressure deficit was 0.7 1.8 and 0.5 1.5 kpa during measurements of gas exchange in spring summer and autumn, respectively. In contrast to the WS treatment, the prolonged exposure of trees in the CS plot to root-zone temperatures slightly above 0 C had pronounced effects on the recovery of A 360 and especially g s from mid-may to mid-june (Figures 3A and 3B; cf. Figure 1B). Lower values of A 360 and g s in needles in the CS treatment than in the other treatments are consistent with results from short-term experiments on the effects of soil temperature on seedling conifers. Nevertheless, the pattern of gas exchange in needles in the CS treatment from late April to mid-june when soil temperature was close to 0 C suggests that increases in mean daily and minimum air temperatures also induced some recovery of A 360 and g s in needles in the CS treatment. Despite increases in A 360 and g s associated with increasing soil temperatures from mid-june to mid-july, both parameters were significantly lower in needles in the CS treatment than in TREE PHYSIOLOGY VOLUME 22, 2002

IMPACT OF SEASONAL AIR TEMPERATURES ON PHOTOSYNTHESIS OF SCOTS PINE 845 Table 1. Light-saturated rate of CO 2 assimilation at an external CO 2 concentration of 2000 µmol mol 1 (A 2000, µmol m 2 s 1 ), stomatal conductance to water vapor (g s, mmol m 2 s 1 ) and intercellular CO 2 concentration (C i, µmol mol 1 ) in 1-year-old needles of Scots pine (Pinus sylvestris L.) from control, warm-soil and cold-soil plots on selected dates during spring and summer 1999. Values of A 2000, which were assumed to be related to electron transport capacity, were corrected to 20 C based on parameters determined by Walcroft et al. (1997). Means ± SE are shown. Within rows, significant differences between treatments (P 0.01) are marked by different letters. Date Variable Treatment F-Value P-Value Control Warm soil Cold soil May 8 A 2000 4.5 ± 0.3 a 4.4 ± 0.3 a 4.5 ± 0.3 a 0.015 0.9852 g s 46 ± 3 a 61 ± 6 a 24 ± 1 b 22.493 < 0.0001 C i 1800 ± 20 a 1840 ± 20 a 1670 ± 20 b 23.677 < 0.0001 A 2000 at 20 C 9.0 8.8 8.9 May 28 A 2000 16.5 ± 0.6 a 15.3 ± 1.1 a 11.5 ± 1.4 a 5.802 0.0088 g s 120 ± 13 a 112 ± 15 a 25 ± 5 b 19.585 < 0.0001 C i 1700 ± 40 a 1670 ± 60 a 1140 ± 110 b 17.951 < 0.0001 A 2000 at 20 C 21.3 19.6 14.7 June 15 A 2000 29.7 ± 1.1 a 26.7 ± 1.5 a 17.2 ± 1.4 b 23.104 < 0.0001 g s 143 ± 17 a 137 ± 20 a 27 ± 5 b 17.701 < 0.0001 C i 1550 ± 60 a 1570 ± 50 a 730 ± 140 b 26.728 < 0.0001 A 2000 at 20 C 25.7 23.2 14.9 July 7 A 2000 24.2 ± 0.9 a 24.4 ± 0.6 a 20.3 ± 0.9 b 8.206 0.0019 g s 244 ± 12 a 200 ± 16 a 99 ± 8 b 34.978 < 0.0001 C i 1760 ± 10 a 1710 ± 20 a 1580 ± 20 b 25.890 < 0.0001 A 2000 at 20 C 23.6 23.7 19.7 July 23 A 2000 24.4 ± 0.4 a 24.1 ± 0.7 a 22.2 ± 1.1 a 2.221 0.1303 g s 244 ± 7 a 215 ± 11 ab 175 ± 13 b 10.677 0.0005 C i 1760 ± 4 a 1740 ± 10 ab 1720 ± 10 b 10.775 0.0005 A 2000 at 20 C 23.9 23.7 21.8 the control treatment in mid-july, when soil temperatures at the two plots were similar (Figures 3A and 3B; cf. Figure 1B). One explanation for the delayed recovery of gas exchange in needles from the CS treatment is that this treatment may have inhibited new root production, and thus reduced root hydraulic conductance (cf. Wan et al. 1999). Root growth in hydroponically cultured seedlings of Scots pine was severely inhibited even at root-zone temperatures of 8 C in a study by Vapaavuori et al. (1992), but in another study of the influence of root temperature on growth in Scots pine seedlings, root mass increment tended to be higher at 5 C than at 20 C (Lyr and Garbe 1995). In addition to reduced root growth, cavitation in the water-conducting system during days with a high evaporative demand may have contributed to a low hydraulic conductance in trees on the CS plot (Brodribb and Hill 2000). The reduction in net photosynthesis at low soil temperatures in conifers has been attributed to stomatal closure and effects on the biochemistry of photosynthesis (see Introduction). Stomatal closure partly explained the depression of A 360 in needles from the CS treatment, because C i was significantly lower in needles in the CS treatment than in the control treatment from early June to mid-july, and on one occasion in late April (Figure 3C). However, the similar C i values in needles in the CS and control treatments on some occasions in spring indicate that non-stomatal rather than stomatal factors were responsible for the depression of A 360 in needles from the CS treatment on these occasions. Thus, it seems possible that the relative importance of stomatal and biochemical effects of low soil temperatures on A 360 varied during spring and early summer. Measurements of A 2000 indicated that the capacity for photosynthetic electron transport was lower in needles in the CS treatment than in the control treatment from late May to early July (Table 1). However, caution should be observed in interpreting the results obtained in mid-june because some C i values in needles from the CS treatment were below the saturation level. Given that there is a linear relationship between maximum Rubisco activity and potential electron transport rate (Wullschleger 1993, Leuning 1997), we speculate that both stomatal and non-stomatal factors were responsible for the depression of A 360 in needles from the CS plot from early June to early July. Because of the small differences in soil temperature between the WS plot and control plot during the autumn and the nonsignificant effect of the WS treatment on gas exchange during spring, we expected that needles in the two treatments would have similar values of A 360 and g s. The relatively constant values of A 360 and g s during September showed that soil temperatures as low as 8 C did not affect gas exchange (Figures 3A and 3B; cf. Figure 1B). Although the relative importance of air and soil temperatures on gas exchange from early October to mid-november is unclear, substantial reductions in A 360 and g s in early October indicated that subfreezing night TREE PHYSIOLOGY ONLINE at http://heronpublishing.com

846 STRAND, LUNDMARK, SÖDERBERGH AND MELLANDER temperatures had a large impact on gas exchange during the following day (cf. Lundmark et al. 1988a, Hällgren et al. 1990). Furthermore, the partial recovery of A 360 and g s from early to mid-october can be attributed to favorable mean daily and minimum air temperatures because soil temperatures decreased slightly or did not change during this period. Reversible inhibitions of net photosynthesis and leaf conductance have been observed previously in Scots pine seedlings in the field after nights with minimum temperatures below the freezing point of needles (Lundmark et al. 1988a, Hällgren et al. 1990). Generally, these studies indicated that inhibition of light-saturated net photosynthesis after nights with minimum temperatures below the freezing point of needles is not a result of stomatal closure because C i increases. We suggest that this interpretation should be modified because the decline in C i after the first severe night frost in early October (Figure 3C), as well as the decrease in C i from mid-october to mid-november, indicated that inhibition of A 360 during autumn was partly a result of stomatal closure. However, inhibition of the biochemistry of photosynthesis either by subfreezing night temperatures (Hällgren et al. 1990) or by prevailing low leaf temperatures (Teskey et al. 1995) was probably a more important cause of the overall decline in A 360 during autumn. In contrast to some other field studies, we found that the higher soil temperature at the WS plot than at the control plot had little effect on A 360, A 2000 and g s during spring, indicating that low soil temperatures did not limit the recovery of gas exchange in Scots pine. On the other hand, the CS treatment severely restricted recovery of gas exchange. However, trees on the CS plot were exposed to soil temperatures slightly above 0 C for a relatively long time, even when differences in the onset of soil warming between years is taken into account (Kluge 2001). Because the size of our study plots was small, caution must be exercised when extrapolating our results to other stands, and additional experiments are needed to determine the importance of low soil temperatures for carbon assimilation in Scots pine. Furthermore, the mechanism whereby low soil temperatures inhibit net photosynthesis is poorly understood. In particular, the quantitative contribution of stomatal and non-stomatal limitations to net photosynthesis needs to be further investigated. Our analysis of gas exchange was largely based on calculations of C i that may be unreliable, particularly at low fluxes of CO 2 and water vapor. The possibility that C i was overestimated because of patchy stomatal closure (see Terashima 1992) was not assessed in the present study; however, Day et al. (1991) found that stomata close uniformly when loblolly pine (Pinus taeda L.) roots are exposed to low soil temperatures. Acknowledgments This investigation was part of the Research Programme for the Utilization of the Boreal Forest, and was funded by the European Regional Development Fund and the Swedish Council for Forestry and Agricultural Research. References Berg, E.E. and F.S. Chapin. 1994. Needle loss as a mechanism of winter drought avoidance in boreal conifers. Can. J. For. Res. 24: 1144 1148. Bergh, J. and S. Linder. 1999. Effects of soil warming during spring on photosynthetic recovery in boreal Norway spruce stands. Global Change Biol. 5:245 253. Bowen, G.D. 1991. Soil temperature, root growth, and plant function. In Plant Roots: The Hidden Half. Eds. Y. Waisel, A. Eshel and U. Kafkafi. Marcel Dekker, New York, pp 309 330. Boyce, R.L. and S.A. Lucero. 1999. Role of roots in winter water relations of Engelmann spruce saplings. Tree Physiol. 19:893 898. Brodribb, T.J. and R.S. Hill. 2000. Increases in water potential gradient reduce xylem conductivity in whole plants. Evidence from a low-pressure conductivity method. Plant Physiol. 123:1021 1027. Carter, G.A., W.K. Smith and J.L. Hadley. 1988. Stomatal conductance in three conifer species at different elevations during summer in Wyoming. Can. J. For. Res. 18:242 246. Cowling, J.E. and R.A. Kedrowski. 1980. Winter water relations of native and introduced evergreens in interior Alaska. Can. J. Bot. 58:94 99. Day, T.A., E.H. DeLucia and W.K. Smith. 1989. Influence of cold soil and snow cover on photosynthesis and leaf conductance in two Rocky Mountain conifers. Oecologia 80:546 552. Day, T.A., E.H. DeLucia and W.K. Smith. 1990. Effect of soil temperature on stem sap flow, shoot gas exchange and water potential of Picea engelmannii (Parry) during snowmelt. Oecologia 84: 474 481. Day, T.A., S.A. Heckathorn and E.H. DeLucia. 1991. Limitations of photosynthesis in Pinus taeda L. (loblolly pine) at low soil temperatures. Plant Physiol. 96:1246 1254. DeLucia, E.H. 1986. Effect of low root temperature on net photosynthesis, stomatal conductance and carbohydrate concentration in Engelmann spruce (Picea engelmannii Parry ex. Engelm.) seedlings. Tree Physiol. 2:143 154. DeLucia, E.H. and W.K. Smith. 1987. Air and soil temperature limitations on photosynthesis in Engelmann spruce during summer. Can. J. For. Res. 17:527 533. DeLucia, E.H., T.A. Day and G. Öquist. 1991. The potential for photoinhibition of Pinus sylvestris L. seedlings exposed to high light and low soil temperature. J. Exp. Bot. 42:611 617. Grace, J. 1990. Cuticular water loss unlikely to explain tree-line in Scotland. Oecologia 84:64 68. Hällgren, J.-E., T. Lundmark and M. Strand. 1990. Photosynthesis of Scots pine in the field after night frosts during summer. Plant Physiol. Biochem. 28:437 445. Havranek, W.M. and W. Tranquillini. 1995. Physiological processes during winter dormancy and their ecological significance. In Ecophysiology of Coniferous Forests. Eds. W.K. Smith and T.M. Hinckley. Academic Press, San Diego, pp 95 124. Katz, C., R. Oren, E.-D. Schulze and J.A. Milburn. 1989. Uptake of water and solutes through twigs of Picea abies (L.) Karst. Trees 3:33 37. Kluge, M. 2001. Snow, soil frost and springtime soil temperature regimes in a boreal forest. Licentiate thesis, Swed. Univ. Agric. Sci., Dept. For. Ecol., Umeå, Sweden, 19 p. Körner, C. 1994. Leaf diffusive conductances in the major vegetation types of the globe. In Ecophysiology of Photosynthesis. Eds. E.-D. Schulze and M.M. Caldwell. Springer-Verlag, Berlin, pp 463 490. Kramer, P.J. 1983. Water relations of plants. Academic Press, New York, 489 p. TREE PHYSIOLOGY VOLUME 22, 2002

IMPACT OF SEASONAL AIR TEMPERATURES ON PHOTOSYNTHESIS OF SCOTS PINE 847 Leuning, R. 1997. Scaling to a common temperature improves the correlation between the photosynthesis parameters J max and V cmax. J. Exp. Bot. 48:345 347. Lundmark, T., J. Bergh, M. Strand and A. Koppel. 1998. Seasonal variation of maximum photochemical efficiency in boreal Norway spruce stands. Trees 13:63 67. Lundmark, T., J.-E. Hällgren and C. Degermark. 1988a. Effects of summer frost on the gas exchange of field-grown Pinus sylvestris L. seedlings. Scand. J. For. Res. 3:441 448. Lundmark, T., J.-E. Hällgren and J. Hedén. 1988b. Recovery from winter depression of photosynthesis in pine and spruce. Trees 2: 110 114. Lyr, H. and V. Garbe. 1995. Influence of root temperature on growth of Pinus sylvestris, Fagus sylvatica, Tilia cordata and Quercus robur. Trees 9:220 223. Odin, H. 1992. Climate and conditions in forest soils during winter and spring at Svartberget Experimental Forest Station. Report 56, Swed. Univ. Agric. Sci., Dept. Ecol. Environ. Res., Uppsala, Sweden, 50 p. Öquist, G. and B. Martin. 1986. Cold climates. In Photosynthesis in Contrasting Environments. Eds. N.R. Baker and S.P. Long. Elsevier, Amsterdam, pp 237 293. Schaberg, P.G., J.B. Shane, P.F. Cali, J.R. Donnelly and G.R. Strimbeck. 1998. Photosynthetic capacity of red spruce during winter. Tree Physiol. 18:271 276. Schwarz, P.A., T.J. Fahey and T.E. Dawson. 1997. Seasonal air and soil temperature effects on photosynthesis in red spruce (Picea rubens) saplings. Tree Physiol. 17:187 194. Sowell, J.B., S.P. McNulty and B.K. Schilling. 1996. The role of stem recharge in reducing the winter desiccation of Picea engelmannii (Pinaceae) needles at alpine timberline. Am. J. Bot. 83:1351 1355. Strand, M. 1995. Inhibition of photosynthesis in current-year needles of unfertilized and fertilized Norway spruce [Picea abies (L.) Karst.] during autumn and early winter. Trees 9:332 340. Strand, M. and T. Lundmark. 1995. Recovery of photosynthesis in 1-year-old needles of unfertilized and fertilized Norway spruce (Picea abies (L.) Karst.) during spring. Tree Physiol. 15:151 158. Sutinen, M.-L., A. Ritari, T. Holappa and K. Kujala. 1998. Seasonal changes in soil temperature and in the frost hardiness of Scots pine (Pinus sylvestris) roots under subarctic conditions. Can. J. For. Res. 28:946 950. Terashima, I. 1992. Anatomy of non-uniform leaf photosynthesis. Photosynth. Res. 31:195 212. Teskey, R.O., D.W. Sheriff, D.Y. Hollinger and R.B. Thomas. 1995. External and internal factors regulating photosynthesis. In Resource Physiology of Conifers. Eds. W.K. Smith and T.M. Hinckley. Academic Press, San Diego, pp 105 140. Tranquillini, W. 1982. Frost-drought and its ecological significance. In Physiological Plant Ecology II. Water Relations and Carbon Assimilation. Eds. O.L. Lange, P.S. Nobel, C.B. Osmond and H. Ziegler. Encycl. Plant Physiol. New Ser., Springer-Verlag, Berlin, pp 379 400. Troeng, E. and S. Linder. 1982. Gas exchange in a 20-year-old stand of Scots pine. I. Net photosynthesis of current and one-year-old shoots within and between seasons. Physiol. Plant. 54:7 14. Turner, N.C. and P.G. Jarvis. 1975. Photosynthesis in Sitka spruce (Picea sitchensis (Bong.) Carr.). IV. Response to soil temperature. J. Appl. Ecol. 12:561 576. Vapaavuori, E.M., R. Rikala and A. Ryyppö. 1992. Effects of root temperature on growth and photosynthesis in conifer seedlings during shoot elongation. Tree Physiol. 10:217 230. Walcroft, A.S., D. Whitehead, W.B. Silvester and F.M. Kelliher. 1997. The response of photosynthetic model parameters to temperature and nitrogen concentration in Pinus radiata D. Don. Plant Cell Environ. 20:1338 1348. Wan, X., S.M. Landhäusser, J.J. Zwiazek and V.J. Lieffers. 1999. Root water flow and growth of aspen (Populus tremuloides)at low root temperatures. Tree Physiol. 19:879 884. Wieser, G. 1997. Carbon dioxide gas exchange of cembran pine (Pinus cembra) at the alpine timberline during winter. Tree Physiol. 17:473 477. Wullschleger, S.D. 1993. Biochemical limitations to carbon assimilation in C 3 plants a retrospective analysis of the A/C i curves from 109 species. J. Exp. Bot. 44:907 920. TREE PHYSIOLOGY ONLINE at http://heronpublishing.com