STUDY SITES

The East Brøgger Glacier is located near Ny-Ålesund in the northwestern part of Svalbard, Norway (78°55′N, 11°51′E). The annual mean air temperature and amount of precipitation in this area are about −5.7°C and 490 mm, respectively. The maximum monthly mean air temperature is about 5°C in July and August, and the minimum monthly mean air temperature is about −15°C in January. The snow-free period ranges from 1 to 2 mo and varies considerably year by year.

We set up 4 study sites (Sites 1, 2, 3, and 4) along a primary successional series of the glacier foreland (Fig. 1). Site 1 was the youngest, situated on a newly deglaciated moraine, about 100 m from the tip of the glacier. Only isolated plants such as S. oppositifolia were observed. The total coverage of vascular plants was less than 1%, and black crusts of cyanobacteria and several species of bryophytes partially covered the ground (Minami et al., 1996). Site 2 was about 300 m from the glacier, and the coverage of S. oppositifolia was greater. Poa alpina var. vivipara and Draba alpina were observed. The coverage of the plants was less than 10%. Site 3 was on a small moraine where approximately 17% of the ground was bare, 30% is covered with black crust, and 53% is covered with mosses and vascular plants (mainly S. oppositifolia, Salix polaris, Luzula confusa, and Sanionia uncinata). Site 4 was on the oldest moraine about 1200 m from the glacier. The ground at Site 4 was patterned by small polygons. Almost all the ground was covered with black crusts, lichens, mosses, and vascular plants (mainly Salix polaris, Luzula confusa, and Sanionia uncinata). The coverage of Saxifraga oppositifolia decreased, and Salix polaris was dominant. These sites were the same used by Nakatsubo et al. (1998) and Bekku et al. (1999). Nakatsubo et al. (1998) measured soil and root respiration and soil water content at each site. Bekku et al. (1999) measured microbial biomass at each site. Throughout all successional stages, there was a positive high correlation between soil C or N content and microbial biomass with correlation coefficients of 0.91 and 0.85, respectively. Those studies also measured vertical profiles of C, N, and microbial biomass in the soil. With the progress of succession, C and N contents increased in the soil surface layer (0–1 cm).

Because the height of most of the vegetation was lower than 10 cm, the δ¹³C value might be affected by soil respiration. Nakatsubo et al. (1998) showed soil respiration rates at Site 1 to 4 and seasonal variation of soil respiration rates early July to mid August at Site 2 and 3. The soil respiration rates in the latter stages of succession (Site 3 and 4) were usually higher than those in the earlier stages (Site 1 and 2) but the maximum rate is quite small (less than 60 mg CO₂ m⁻² h⁻¹ at most). Thus, the effects of soil respiration can be assumed negligible.

To compare the effects of the different moisture environments on the growth form, Site 5 was set up on a riverbank about 600 m from the glacier. There were large patches of S. oppositifolia, and P-form and C-form plants grew in close proximity within the same homogeneous habitat (cf. Kume et al., 1999). Site 6 was set up on a dry moraine ridge, a polar semidesert ecosystem about 3.5 km north of the glacier containing many small patches of S. oppositifolia and Dryas octopetala. The snow-free period was about 2 wk longer than that of other sites.

SAMPLES

In the Ny-Ålesund area, two morphs of Saxifraga oppositifolia were recognized: the prostrate form (P-form) and the cushion form (C-form). Although the existence of intermediate growth forms was pointed out by Brysting et al. (1996), typical P-form plants have long-internode shoots. We used the third internode length of shoots as a key for distinguishing between the two forms. A plant having shoots of the third internode length longer than 6 mm was categorized as P-form (Kume et al., 1999). The distributional pattern of the two morphs in the glacier foreland was described in Kume et al. (1999). On 30 and 31 July 1999, which is near the end of the growing season in Ny-Ålesund, 3 to 5 individuals of each growth form of S. oppositifolia and other species growing on each site were arbitrarily selected. These plants were immediately put in a deep freezer and then freeze-dried. The green parts of the samples were milled with a coffee mill modified for small samples (KTG-0001, Koizumi-Seiki Co., Osaka, Japan). Carbon isotope ratios in plant dry matter were measured for subsamples containing 100–200 μg of carbon with a system combining an elemental analyzer (EA1108, Carlo-Erba, Italy) for combustion of the samples, an interface (Finnigan MAT Confro, Bremen, Germany), and a stable isotope ratio mass spectrometer (Finnigan MAT 252, Bremen, Germany). The mean reproducibility for isotope ratio measurement was ±0.08‰ (n = 26).

STATISTICS

Data are presented as mean ± one standard error (SE). The statistical significance in the effects of sites and growth forms on δ¹³C values was assessed the two-way analysis of variance. All statistical analyses were carried out using the StatView 5.00 (SAS Institute, Inc.).

δ¹³C OF SAXIFRAGA OPPOSITIFOLIA

In all study sites, δ¹³C of S. oppositifolia was distinctly different between the two growth forms (Fig. 2, Table 1). At each site, δ¹³C differed by about 2‰ between the two growth forms. The δ¹³C values of the C-form were always more negative and ranged from −28‰ to −30‰, while those of the P-form ranged from −26.5‰ to −28‰. With the progress of succession (from Sites 1 to 4), the water and nitrogen content in the soil increased distinctly, but δ¹³C of both growth forms significantly increased (Fig. 2, Table 1). At Site 5, where the two growth forms grow sympatrically on the moist riverbank, the δ¹³C value of the C-form was 3‰ more negative than that of the P-form. The C-form plant was hardly observed at Site 4, which was the latest successional site and the wettest site, while the P-form plant was not observed at Site 6, which seemed to be the driest site.

δ¹³C OF EACH SPECIES AT EACH SITE

The δ¹³C values for the dominant vascular plants, mosses and lichens at each site are shown in Figure 3. At Site 1, only S. oppositifolia was present in sufficient amounts for sampling. At all other sites, the δ¹³C values of the P-form were more positive than the δ¹³C values of all other species. On the other hand, δ¹³C of the C-form tended to be more negative than the δ¹³C values of the other species. Narrow-leaved grasses (Poa and Luzula) had small δ¹³C values. Salix polaris and mosses (Sanionia uncinata and Aulacomnium turgidum) had intermediate δ¹³C values regardless of the site. The lichen Cetrariella delisei had quite positive δ¹³C values at Site 6, a dry site, and negative values at Site 4, a moist site. This result suggested that C. delisei was strongly affected by the moisture conditions on the soil surface, but more samples may be required to confirm this hypothesis.

Most of the variance in δ¹³C values was due to the species or growth form. Even though the soil mass water contents of the study sites varied from 10% to 140% and the soil nitrogen contents varied from 0.02% to 0.85% (Fig. 2), the variance among individuals and the variance in the same species among the sites were both small. The average δ¹³C values of the observed species at Sites 1–6 were −28.5, −28.5, −28.4, −28.6, −28.6, and −28.0‰, respectively. The difference among the sites was less than −0.6‰ and no significant correlation was detected. The δ¹³C values for a given species or growth form varied only slightly among the study sites.

WATER UTILIZATION OF THE TWO MORPHS OF SAXIFRAGA OPPOSITIFOLIA

Our results show that δ¹³C value of the P-form plants was always the most positive among all species, indicating severe physical restrictions (including stomatal restriction) of CO₂ uptake in photosynthesis. The δ¹³C value of the P-form plants also suggested the highest water use efficiency. On the other hand, δ¹³C value of the C-form of S. oppositifolia was always more negative than that of the P-form, indicating that the physical restrictions in photosynthesis were small, at least at the individual leaf level. These results show that δ¹³C depends on the growth form rather than on the soil conditions. Though Körner et al. (1991) concluded that a range of 4‰ is to be expected within a single community of C₃ plant species, the difference in δ¹³C between C-form and P-form was significant and consistent.

When the water utilization of a whole plant is considered, the transpirative water consumption (TWC) is determined by the product of surface area per biomass (SA) and average conductance to water vapor of the leaf surface (gl). Both SA and gl may be important for TWC. Because the two growth forms had similar root mass fraction and root types (taproot or fibrous root) when growing in the same habitat (Kume et al., 1999), the manner of water utilization of S. oppositifolia is determined by stomatal openness and biophysical characteristics of aboveground growth forms. C-form plants only have shoots with a short internode length (<6 mm), and the upper and the lower leaves imbricate each other. Some leaves of P-form plants are thin and flat shaped, but all the leaves of C-form plants are thick and fleshy. According to Kume et al. (1999), C-form plants are able to adjust their SA to soil moisture conditions by changing the degree of leaf overlap: SA has a value of about 50 cm² g⁻¹ in dry environments and 100 cm² g⁻¹ in damp environments. P-form plants have long internode shoots with thin, detached leaves. The SA of P-form plants is about 100 cm² g⁻¹. At Site 5, a damp riverbank, where both growth forms grow sympatrically, the C-form and the P-form had similar SAs (about 100 cm² g⁻¹). However, the difference in the δ¹³C values between the two growth forms was the largest among the sites. These facts suggest that the P-form leaf's innate physical restriction of CO₂ uptake is larger than that of the C-form.

In the net photosynthesis rate per SA, Kume et al. (1999) could not detect any differences among the single shoots of different forms when they were measured under field conditions with an open-flow infrared gas analysis system (LI-6400, LI-COR, Inc., Lincoln, Nebraska, U.S.A.). However, the Ci/Ca ratio of the P-form was about 0.72, and that of the C-form was about 0.82. These results agree with the δ¹³C analysis and raise the possibility that the P-form had lower water consumption than the C-form under the damp conditions at Site 5, even though their net photosynthetic rates were similar. However, this assumption must be incorrect. Water consumption at the whole plant level must be less in the high-density cushion form than in the low-density prostrate form. The ground cover area per biomass of P-form plant is the largest among the plants growing in Ny-Ålesund. The average ground cover area per biomass (GA) of the P-form was about 240 cm² g⁻¹, whereas GA of the C-form was at most 70 cm² g⁻¹ in the moist sites and less than 30 cm² g⁻¹ in the dry sites (Kume et al., 1999). The conductance of water vapor of the boundary layer of the plant surface can be assumed equal because all of these colonies spread on the ground and have similar aerodynamic resistance. Therefore, the P-form's effective evaporative surface area is over 4 times larger than that of the C-form, and the demand of transpiration per biomass of P-form plants may be markedly larger than that of C-form plants. At the whole plant level, the C-form plant is self-shaded and has small conductance per biomass because of its high plant (area) density.

Nakatsubo (1994) obtained similar results in moss colonies of different area densities and pointed out the importance of GA in restricting water loss from moss colonies. TWC is more accurately given by GA × gl than by SA × gl, and δ¹³C is mostly affected by GA. Despite the small differences among different shoot forms in the net photosynthesis rate per SA, it is expected that the whole-plant photosynthesis per biomass of P-form plants is markedly larger than that of C-form plants. This conclusion is in accordance with the results of Crawford et al. (1993). They showed that the two growth form plants had different photosynthesis rates per biomass (the P-form was about 2.3 μmolCO₂ g⁻¹ min⁻¹, and the C-form was about 1.3 μmolCO₂ g⁻¹ min⁻¹). Crawford and Wolf (1999) also provided a value of the ¹³C isotope discrimination ratio (Δ) for S. oppositifolia in the High Arctic (–19.5‰ ± 0.32, n = 8). If the ¹³C value of surrounding environment (δ¹³C of air) was assumed −8‰, their δ¹³C will be −27.5‰ and well accord with our results.

Körner (1999) showed that the differences in life form had negligible effects on δ¹³C in alpine plants. One reason is that, he did not consider the relationships among SA, GA, gl, root mass fraction, and life forms. More appropriate life form evaluation might be needed to understand the effects of different life form on δ¹³C.

ADAPTIVE SIGNIFICANCE OF GROWTH FORMS OF SAXIFRAGA OPPOSITIFOLIA

In the High Arctic, the growing season in areas where snowbeds occur, about 30 days, is significantly shorter than that on snow-free ridges, about 50 days (Crawford 1997). Usually P-form plants are dominant in snowbeds, that is, damp areas with a short growing season. In such environments, P-form plants have an advantage because of their high productive efficiency, while C-form plants are not able to produce enough photosynthate to survive in the short growing season. Nonetheless, at Site 6, which was the driest site but also the one with the longest expected growing season, the number of C-form plants increased significantly. This result may be because the GA of P-form plants is so large that it is difficult to decrease TWC by decreasing gl alone in extreme dry conditions. At the base of the East Brøgger Glacier, there is plenty of meltwater from the glacier, and the growing season is relatively long compared to that in areas where snowbeds occur. These conditions may enable the sympatric growth of C- and P-form plants.

Teeri (1972) showed that the uptake of dissolved oxygen during flooding was faster in the P-form than in the C-form. This difference may be because of the large GA of the P-form and suggests that another advantage of the P-form is its greater conductance at the whole-plant level.

The biophysiological characteristics of the whole-plant conductance may be significant only in the high arctic environment, where the competition for light is negligible during the growing season and where the plants spread only in two-dimensional space. Therefore, environmental changes such as global warming will affect the balance between the two growth forms by increasing three-dimensional competition (e.g., Stenström et al., 1997).

EFFECTS OF PRIMARY SUCCESSION

In both growth forms of S. oppositifolia, the δ¹³C values increased slightly with increasing soil water content with the progress of succession from Sites 1 to 4 (Fig. 2). These results were puzzling because they suggested that the physical restrictions in photosynthesis of S. oppositifolia increased with increasing soil water availability. Even if the soil respiration affected the δ¹³C values, the values should be decreased with increasing soil respiration. Saxifraga oppositifolia is the only vascular plant that occurs in all successional stages in Ny-Ålesund, although the population size decreased at the later successional stages, especially in C-form plants (Kume et al., 1999). One reason for the decrease in population size is that competition from other vascular plant species and mosses makes it more difficult to establish new seedlings at the later successional stages. Another possible reason is that there is an increase of GA per root mass of S. oppositifolia because of aboveground competition with other species. In later successional stages, the centers of the developed patches of S. oppositifolia were invaded by mosses, mainly Sanionia uncinata, which tended to cover the patches. To escape the mosses, Saxifraga oppositifolia might have to grow in the outer areas of a patch that had not yet been invaded by other species. Such activities should force S. oppositifolia to increase GA per biomass, hence decreasing the water supply per GA and increasing stomatal restriction. Therefore, S. oppositifolia might increase δ¹³C with the progress of succession in spite of the increase in the soil water content.

GROWTH FORM VS. ROOT MASS FRACTION

The differences in the average δ¹³C values of all the species growing at each site were quite small among the sites (Fig. 3), but the reason was not clear. Narrow-leaved grasses and Dryas octopetala had small δ¹³C values, which suggested a small stomatal restriction in photosynthesis. The grass species had quite large SA, and D. octopetala grew in the polar semidesert, a dry environment. We inferred that the small δ¹³C of these plants was caused by their greater root mass fraction, which was about 2 to 5 times larger than the root mass fraction of S. oppositifolia. The deeper, well-developed roots of these grasses can provide enough water for their large TWC.

In our study sites, all vascular plants except S. oppositifolia seemed to have different root mass fraction in different growth environments. Kume et al. (1999) showed that the root characteristics of S. oppositifolia were little affected by the growth form. In addition, the root mass fraction of S. oppositifolia was relatively small. For example, the root mass fraction of S. oppositifolia was about 0.20, and its coefficient of variation (CV, SD/mean) was 0.2. The mean ratio of Draba alpina was 0.35 and CV was 0.23; that of Poa alpina was 0.40 and CV was 0.25. For a plant to be established in the early stage of primary succession, it is essential for the roots to reach the safe layer in which the soil water content is stable, but root elongation is often disturbed by physical and meteorological events (Maruta, 1976). Thus, in newly deglaciated areas, where root development is often restricted, it may be beneficial for first colonizers, such as S. oppositifolia, to have two morphs with different aboveground growth forms or to be able to control the biophysical characteristics of growth forms in response to the variable severe growing conditions.

Even though stomatal restriction does not occur in mosses, the δ¹³C values of mosses were similar to those of vascular C₃ plants, and the variances in the δ¹³C values within individual species of mosses were relatively small. These results were similar to those of previous studies (Rundel et al., 1979; Teeri, 1981; Proctor, 1992). In Ny-Ålesund, high photosynthetic activities of mosses were observed only on rainy days or soon after rainfall, when moss water content was high (Uchida et al., 2002). Therefore, whole-plant conductance should be kept within a certain range during photosynthesis, and this should result in steady δ¹³C values. As mentioned earlier, Nakatsubo (1994) pointed out the importance of differences in GA of each species in restricting water loss from moss colonies. The difference in GA causes the difference in the drying rate of the plant colonies and can readily affect the δ¹³C values. The difference in δ¹³C between the moss species may be caused by different colony densities, as is the case for S. oppositifolia. In Ny-Ålesund, the mosses, which do not have roots, tended to be more dominant than the vascular plants, which do. These facts emphasize the importance of the aboveground growth form in the high arctic environment.

ADAPTIVE SIGNIFICANCE OF POLYMORPHISM

Saxifraga oppositifolia had one of the broadest distributions near the lower end of the East Brøgger Glacier. It is also distributed widely in areas of receding glaciers in alpine regions of Europe, Asia, and North America (Hultén and Fries, 1986). In high arctic regions, contrasting extreme microclimatic environments are adjacent to each other. Polymorphism may be one of the adaptive characteristics for surviving under such a severe, mosaic environment. The photosynthesis, transpiration, and many other physiological and reproductive characteristics were significantly different between the two growth forms (Teeri, 1972; Crawford et al., 1995; Kume et al., 1999). Crawford et al. (1995) suggested that S. oppositifolia adapts to changing climates and survives glaciation by changing the frequencies of the two morphs. Various types of polymorphism are reported in Dryas octopetala (McGraw, 1995; Max et al., 1999), which showed much visible ecotypic variation throughout its wide polar range distribution (Elkington, 1971). Morphological variability or polymorphism within a species appeared to be one of the adaptive characteristics of plants growing in the high arctic, as well as genetic flexibility (Brochmann and Håpnes, 2001), where the forces of selection vary spatially and temporally from short-term annual disturbances to long-term climatic oscillations.