1I examined the role of species interactions on the population dynamics of the herb Thalictrum alpinum and the sedge Carex vaginata by removing the dwarf shrub Dryas octopetala over four growing seasons at Finse, in the alpine region of south Norway, Furthermore, by increasing temperature (open top chambers) and nutrient availability (nutrient addition), I assessed the effects of climate warming on biotic interactions.
2Dryas removal increased the number of Thalictrum and Carex leaves, and flowering frequency of Carex. Flower stems and leaf stalks of Thalictrum, and the leaves of Carex, became significantly shorter following Dryas removal. Warming and nutrient addition increased vegetative plant growth, whereas warming alone had positive effects on sexual reproductive effort of the target species. My results suggest that there is both competition for nutrients, and a facilitative shelter effect of Dryas on its neighbours.
3Species interactions affected population dynamics of Thalictrum and Carex. Interactions between the impacts of Dryas removal and abiotic factors on leaf production, suggested that interactions between Dryas and neighbouring species might be modified if temperature and, in particular, nutrient availability increase under global warming.
4This study shows that both biotic interactions and abiotic environmental conditions may affect alpine plant population dynamics. Furthermore, it shows that climate change may modify species interactions.
5Species interaction effects should be included in climate change experiments and in future models predicting plant community changes under global warming.
Both biotic factors, such as species properties and species interactions, and abiotic factors, such as climate and soil characteristics, affect ecosystem structure and processes (Huston & McBride 2002). However, the relative importance of the biotic and abiotic factors for community dynamics in habitats of contrasting environmental severity and productivity is far from clear (Callaway & Walker 1997). Competition appears to play an important role in structuring plant communities of high productivity (e.g. Grime 1979), but there is still a debate about the role of biotic interactions when productivity is low, such as in alpine and arctic plant communities of high abiotic stress. Competition may be strong here because resources are scarce (e.g. Chapin & Shaver 1985; Tilman 1988), or facilitation may be stronger, with neighbouring vegetation improving the survival potential for individual plants (e.g. Callaghan & Emanuelson 1985; Bertness & Callaway 1994). Alternatively, species interactions may have less impact on community composition in areas of high abiotic stress because plant growth here is primarily limited by the environmental conditions (e.g. Savile 1960; Billings & Mooney 1968; Grime 1977).
Global warming may modify abiotic conditions that influence individual plant performance, with alpine and arctic ecosystems predicted to be particularly affected (e.g. Callaghan & Jonasson 1995; Guisan et al. 1995; Körner 1995, 1999). Climate change may alter soil moisture and increase nutrient availability and, in combination with the direct warming effect, these factors may impact the growth, reproduction and resource allocation of the plants. However, functional groups (Arft et al. 1999; Dormann & Woodin 2002; Brooker & van der Wal 2003) or individual plant species (see Henry & Molau 1997) may differ in their responses to environmental changes. Furthermore, changes in the performance of one species or a functional group may change any current relationship between positive and negative plant–plant interactions, i.e. changes in competition or facilitation hierarchies between co-occurring species or growth forms.
The role of species interactions in structuring tundra plant communities differs between sites (see Hobbie et al. 1999). Moreover, effects of climate change on plant–plant interactions are rarely studied (but see Shevtsova et al. 1997; Hobbie et al. 1999), and species interactions are mostly ignored in models predicting how plant communities or ecosystems respond to climate change (Mooney 1991; Pacala & Hurtt 1993). Bertness & Callaway (1994) predicted a shift from facilitative to competitive effects of neighbouring plant species along gradients of decreasing abiotic severity and, in a study of 11 alpine sites, Callaway et al. (2002) found that species interactions shifted from positive to negative when summer temperatures increased along latitudinal and altitudinal gradients. They predicted that such shifts may speed changes in community composition and diversity compared with more physiologically based climate change models, and concluded that there is a need for more experiments on climate change effects on individual plant–plant interactions. There is also a lack of knowledge about how plant interactions affect resource allocation to reproduction vs. vegetative growth. Studies show contrasting results, which may indicate that both competition and facilitation may contribute to higher reproductive effort, though with varying responses between species (Kikvidze et al. 2001).
The aim of this study is to assess the relative impacts of biotic and abiotic factors on the growth and reproduction of two plant species in a Dryas heath community in alpine south Norway, and to gain an understanding of how interspecific interactions may be modified by abiotic conditions predicted to change under global warming. Hobbie et al. (1999) manipulated air temperature and the presence of seven dominant species in Alaskan tussock tundra, and found that the environmental conditions were more important for plant growth and biomass production than the impacts of species interactions. Moreover, as they found no interactions between warming and species removal, they concluded that global warming might not alter species interactions in tundra. In northern Finland, on the other hand, Shevtsova et al. (1997) found a number of interactions between environmental conditions (temperature and water) and species removal on the growth of sub-arctic dwarf shrubs, suggesting that community composition may modify the effects of climate change here. The contrasting results between the Alaskan tundra and the European sub-arctic site suggest that there is a need for more research on species interactions and their responses to climate change. Moreover, none of these studies assessed the impact of increased nutrient availability resulting from climate warming (but see Bret-Harte et al. 2004), although this is predicted to provide the largest effects on tundra plant community composition (e.g. Chapin et al. 1995; Shaver & Jonasson 1999; Dormann & Woodin 2002).
Dryas heaths are plant diversity ‘hot spots’ in the Scandinavian mountains, dominated by the long-lived ‘wintergreen’ dwarf shrub Dryas octopetala L. (Dryas hereafter). Dryas grows in dense clones that may have positive or negative impacts on community diversity (Klanderud & Totland, in press). The relatively simple vegetation structure, with few and compressed vegetation layers, as well as predictions of large climatic impacts on species composition in alpine areas, make this an ideal arena to test questions related to species interactions and climate change. I used experimental removal of Dryas, warming (open top chambers, OTCs) and nutrient addition, to examine the role of these factors on the population dynamics of Thalictrum alpinum L. and Carex vaginata Tausch. Furthermore, I assessed if species interactions might be modified under climate warming. Thalictrum and Carex are abundant in the alpine Dryas heaths, and they belong to different functional groups (herbs and sedges, respectively) that may contrast in their responses to other species and to the abiotic environment (e.g. Wilson & Whittaker 1995; Dormann & Woodin 2002; Brooker & van der Wal 2003). Three predictions were addressed. (i) The two target species will differ in their responses to Dryas removal and to changes in the abiotic environment. (ii) Higher productivity at our site will lead to plant–plant interactions playing a greater role in plant community dynamics than reported in the Alaskan tundra (Hobbie et al. 1999; Bret-Harte et al. 2004). (iii) Species interactions will be modified by increased temperatures and nutrient availability.
Methods
study system
The experiment started in early July 2000 and lasted until the end of the growing season (late August) in 2003. The study site was situated on a south-west exposed slope of a Dryas octopetala L. heath at c. 1500 m elevation on Sandalsnuten, Finse, northern part of Hardangervidda (60° N, 7° E) in alpine south-west Norway. The snow normally disappears in late May/early June. Mean summer (June, July, August) temperature at 1222 m elevation at Finse is 6.3 °C (Aune 1993), and mean summer precipitation is 89 mm (Førland 1993).
The dwarf shrub Dryas octopetala is widely distributed in alpine and arctic areas. At Sandalsnuten, dense Dryas mats cover c. 35% of the ground of the Dryas heath community (Klanderud & Totland, in press). Other abundant species here are the herbs Thalictrum alpinum L., Potentilla crantzii Crantz., Bistorta vivipara L. and Cerastium alpinum L., the dwarf-shrub Salix reticulata L., the grasses Festuca vivipara L. and Poa alpina L., and the sedges Carex vaginata Tausch., C. atrofusca Schkuhr, C. rupestris All. and Luzula spp. (nomenclature follows Lid & Lid 1994), in addition to a number of bryophytes and a few lichens.
Vegetative regeneration is common for alpine plants. Thalictrum alpinum and Carex vaginata grow clonally by producing new ramets from below-ground rhizomes, and seedlings of these species are rarely found at Sandalsnuten. Therefore I studied the responses of mature individuals rather than seedling responses.
experimental design
To assess the possible effects of removal of Dryas, warming and increased nutrient availability on the population dynamics of Thalictrum alpinum and Carex vaginata, and whether environmental conditions affected any species interactions, I randomly selected 20 plots within a relatively homogenous area of the Dryas heath. I placed OTCs upon 10 of these plots and left the 10 others as controls. The OTC/control plots are the main-plots in a split-plot design with 10 replicates (see, e.g. Underwood 2001). In each of the 20 main-plots, I selected eight Thalictrum and eight Carex ramets growing inside Dryas mats, and inserted half a slow-dissolving NPK-fertilizer stick into the soil c. 1 cm upslope of half of the individuals of each species immediately after snowmelt and again in late July (c. 0.2 g N, 0.04 g P and 0.17 g K per ramet/growing season). Ramets receiving fertilizer were selected so that the added nutrients could not move to unfertilized ramets, i.e. they were grouped, or situated down-slope from other ramets. Thereafter, I clipped the above-ground parts of neighbouring Dryas, and carefully removed below-ground parts that could be pulled up with minimal soil disturbance, from a diameter of c. 10 cm around half of the fertilized ramets and around half of the ramets not receiving any fertilizer. I removed re-growth twice during each of the four growing seasons. This provided a split-plot design with temperature treatment as a fixed factor conducted on main-plot level, 20 plots as a random factor nested within the main-plot factor (temperature), and nutrient addition and removal treatments as fixed factors conducted on subplot level. Thus, for each of the 10 replicates, two ramets of each of the two target species in each temperature treatment (inside and outside OTCs) received nutrient addition, two experienced Dryas removal, two nutrient addition and Dryas removal, and two were untreated controls. The responses of the two individuals receiving similar treatments were averaged prior to statistical analysis.
Removal experiments may be problematic for several reasons. Removal of above-ground biomass may leave dead roots to decompose and increase soil nutrient levels (Putwain & Harper 1970; Berendse 1983). However, these resources have most likely been obtained by competition in the past, and their release may benefit those plants that have been denied these resources earlier (Aarssen & Epp 1990). Vegetation removal may also disturb the soil, resulting in a nutrient flush. However, the primary removal in this study was of above-ground Dryas shoots, and there were probably only minor effects on below-ground processes and soil disturbance.
The OTCs are hexagonally shaped polycarbonate chambers with an inside diameter of c. 1 m, and with qualities as described in Marion et al. (1997). OTCs are commonly used to raise the temperature in climate change experiments (e.g. Marion et al. 1997; Hollister & Webber 2000). The chambers did not affect the duration of snow cover (personal observation) and were therefore left in place throughout the experiment. The site was fenced to inhibit sheep grazing.
abiotic measurements
Air (c. 5 cm above ground) and ground surface temperatures were measured with a Grant Squirrel data logger during 10 days of late July/early August 2000, and soil (c. 5 cm below ground) temperatures were measured with Tinytag 12 Plus G data loggers (Intab Interface-Teknik AB, Stenkullen, Sweden) from early June to late August 2002, inside and outside the OTCs, in plots where Dryas had been removed, and in undisturbed plots. Leaf surface temperature of Dryas inside and outside the OTCs was measured using an infrared thermometer (FLUKE 65, Fluke Corporation, Everett, USA). Soil moisture was measured at the end of the final growing season by inserting a 6 cm long sensor (Theta Probe, Delta-T Devices Ltd, Cambridge, England) into the soil immediately beneath each target ramet (averaging three measurements per ramet).
growth measurements
I measured growth and reproductive variables of Thalictrum alpinum and Carex vaginata in late August after the first (2000), second (2001), third (2002) and fourth (2003, reproductive variables only) growing seasons. To assess differences in within-season growth rates, I also measured the growth variables in late May (before any growth had started), late June, and again in late July during the second and the third growing season. However, to limit possible legacy effects of the removal treatment, only the 2002 and 2003 measurements are used in the analyses, with the 2000 measurements as covariables.
Sexually reproductive Carex ramets normally die the year after flowering, with new tillers growing out from the old ramet (Carex vaginata, K. Klanderud, personal observation; C. bigelowii, see Brooker et al. 2001). To simplify interpretation of growth and sexual reproduction independently of flowering year, I obtained target ramets at similar developmental stages by selecting flowering Carex individuals at the start of the experiment. To obtain an estimate of vegetative growth of Thalictrum and Carex, I counted the number of green leaves on each target ramet and the total number of leaves on new tillers growing out from the target ramet. The data from the target ramet and daughter tillers are pooled in the statistical analyses. As number of leaves is not reported per tiller, possible changes may be due either to increased tillering or increased size of the tillers. Furthermore, I measured the length and the width of the largest and the smallest leaf and the length of their leaf stalk (Thalictrum), and the length of the longest leaf (Carex), using a digital caliper. I calculated an approximate leaf area of Thalictrum leaves by multiplying the width by the length and then calculated the average area of the smallest and the biggest leaf. Leaf number, average leaf area of the largest and the smallest leaf, and the length of their leaf stalks, are parameters commonly used for herbs, whereas leaf number and the length of the longest leaf are commonly used for sedges to obtain non-destructive measures of vegetative growth (Molau & Edlund 1996; Arft et al. 1999). To estimate mortality I recorded dead target ramets. Measuring mortality on clonal plants is, however, not easily applied because individuals usually persist, and only parts of the plants die. The data for Carex are not analysed because mortality here coincides with sexual reproduction.
sexual reproduction measurements
To estimate the sexual reproductive effort of Thalictrum and Carex, I measured the height of flower stems and collected mature infructescences at the end of each season and recorded the number of flowers and number and weight of dried seeds. There were not enough flowering Carex individuals to test for seed production and seed weight. Arctic and alpine plants normally do not produce flowers each year (Sonesson & Callaghan 1991) and the same ramet did not reproduce more than once during the experiment for either of the two species. To increase the sample size, I therefore pooled all flowering individuals from 2002 to 2003 to estimate flowering frequency for both species, as well as height of flower stem and number of flowers and seeds for Thalictrum. There were too few mature seeds to conduct statistical tests on seed weight.
statistical analyses
Data on the number of leaves, leaf area (Thalictrum) and leaf length (Carex) were log transformed to fulfil the ancova assumptions of normality and equal variances. All graphs are shown with untransformed data. To determine if warming, nutrient addition and removal of Dryas had any impact on the vegetative growth of Thalictrum and Carex, and whether possible species interactions were affected by the environmental factors, I used general linear models (GLM, systat 10) with the temperature treatment (main-plot factor), nutrient addition, removal of Dryas, and their interactions (subplot factors) as fixed factors, and plot nested within temperature as a random factor in a split-plot ancova. I used the first year (2000) measurements as covariables in the analyses to increase the model's ability to detect treatment effects. The effect of temperature was tested over the plot error (main-plot term), whereas all other effects were tested over the model error (subplot term) (see, for example, Underwood 2001). Analyses were conducted separately for the two species. To examine if the removal of Dryas or the environmental factors had any impact on the combination of growth phenology and absolute size differences of Carex and Thalictrum during the third (2002) growing season after experimental onset, I used repeated-measures anova with the same factors as above, and with ramet identity during three measurements throughout the season (time) as the repeated measures factor. A significant interaction between time and treatment may then indicate that treatments differ in the pattern of growth during the season.
Because of low sample sizes for reproductive parameters, I used the Mann–Whitney U-test to assess significant differences in height of flower stem, number of flowers per ramet and number of seeds per flower within each treatment separately (temperature vs. control, nutrient addition vs. natural, removal vs. undisturbed). I used Pearsons chi-square test to assess significant differences in flowering frequency for Thalictrum and Carex and mortality for Thalictrum, between the same treatments as above. The small sample sizes for sexual reproduction may result in conservative results.
Results
abiotic factors
The OTCs increased mean Dryas leaf temperature and air temperature by 1.5 °C, and ground temperature by 2.5 °C (Table 1). When soil temperature was measured, the OTCs increased values in undisturbed vegetation by 1.5 °C, while Dryas removal caused a decrease of 0.5 °C inside the OTCs, and an increase of 0.5 °C outside (Table 1).
Table 1. Mean temperature of Dryas leaf surface, air (c. 5 cm above ground), ground surface and soil (c. 5 cm below ground) where vegetation was experimentally removed or intact, and soil moisture, outside (control) and inside OTCs at alpine Finse. Standard errors, minimum and maximum measures are given
Control
OTC
Mean ± SE
Minimum–Maximum
Mean ± SE
Minimum–Maximum
Dryas leaf surface (°C)
12.1 ± 0.1
9.7–14
13.6 ± 0.2
12.1–16.2
Air (°C)
7.8 ± 0.2
−2.2–28.1
9.3 ± 0.2
−2.5–36.9
Ground (°C)
10.4 ± 0.4
−2.1–45.1
12.9 ± 0.5
−2.5–55
Soil, Dryas removal (°C)
9.2 ± 0.1
2.1–18.2
9.6 ± 0.1
2.7–18.6
Soil, no removal (°C)
8.7 ± 0.1
2.3–17.0
10.2 ± 0.1
2.8–22.3
Soil moisture (%)
28.2 ± 0.6
17.9–44.2
27.9 ± 0.8
11.8–42.5
vegetative growth
Removal of Dryas octopetala increased the number of leaves of Thalictrum alpinum by 79.2% and of Carex vaginata by 56.4% across all treatments (Table 2, Figs 1a and 2a). The Thalictrum leaves became 35.1% smaller and the Carex leaves became 14.1% shorter after Dryas removal, although this was only statistically significant for Carex (Table 2, Figs 1b and 2b). The length of Thalictrum leaf stalks decreased by 49.3% after Dryas removal, although a close to significant three-way interaction indicated that temperature, nutrient addition and removal treatment all affected each other (Table 2, Fig. 1c). Significant measures from 2000 suggested treatment effects on Thalictrum leaf stalks after the first growing season.
Table 2. Fd.f. and P values (significant at < 0.05 in bold, < 0.1 in bold italics) of treatments performed in alpine Dryas octopetala heaths; T = temperature (OTC vs. control, main-plot fixed factor), N = nutrient addition (increased vs. ambient, subplot fixed factor), R = removal of Dryas (yes vs. no, subplot fixed factor), their interactions, plot nested in T (random factor), and measurements from 2000 included as covariables in split-plot ancovas on number of leaves of Thalictrum alpinum and Carex vaginata, average leaf area and length of leaf stalk of the biggest and the smallest leaf of Thalictrum, and length of the longest leaf of Carex. n = 10 for each treatment
Means and standard errors of (a) the number of leaves, (b) leaf area, and (c) length of leaf stalks of Thalictrum alpinum measured at the start of the experiment in 2000 (ancova pre-treatment values) and by the end of the 2002 growing season in a Dryas octopetala heath outside (controls) and inside OTCs, on ramets receiving nutrients (N), removal (R), nutrients and removal (NR), or no treatments (C), at Finse, alpine south Norway. n = 10 for each treatment.
Means and standard errors of (a) the number of leaves and (b) leaf length of Carex vaginata measured in a Dryas octopetala heath by the end of the 2000 and 2002 growing season outside (controls) and inside OTCs, on ramets receiving nutrients (N), removal (R), nutrients and removal (NR), or no treatment (C), at Finse, alpine south Norway. n = 10 for each treatment.
Nutrient addition increased the number of Thalictrum leaves by 32.3% and the number of Carex leaves by 59.1%, with interactions (only marginally significant for Carex) between temperature and nutrient addition suggesting that the number of leaves increased on ramets receiving both warming and nutrients, whereas there were no significant effects of either of these treatments separately (Table 2, Figs 1a and 2a). Warming and nutrient addition increased Thalictrum leaf area by 21.4% and 18.6%, respectively, and Carex leaf lengths by 8.7% and 10.8%, although these responses were not statistically significant (Table 2, Figs 1b and 2b).
growth phenology
The increased number of Thalictrum leaves after Dryas removal (Table 2, Fig. 1a) was primarily caused by production of more leaves early in the season (May, June), and again later in the growing season (August). This was particularly pronounced when nutrients were added in combination with Dryas removal, suggesting that removal of a competitor increased the benefit of the additional nutrient availability for Thalictrum (significant time × nutrients × removal, Fig. 3a). A significant interaction (time × temperature × nutrients, Fig. 3b) on Thalictrum leaf area indicated that leaves on fertilized plants, which in mid-June were bigger than leaves in all other treatments, decreased significantly in mean size from mid-July outside the OTCs, but remained bigger than all other leaves throughout August inside OTCs. Thalictrum leaf stalks, which also were longest on fertilized leaves, declined by the end of the growing season in most treatments except where Dryas was removed inside the OTCs, where they continued to grow throughout the season (significant time × removal, and time × temperature × nutrients, Fig. 3c).
Means of the number of leaves (a), leaf area (b) and leaf stalk length (c) of Thalictrum alpinum measured in a Dryas octopetala heath in late May, late June, late July and late August, outside (controls) and inside OTCs, on ramets receiving nutrients (N), removal (R), nutrients and removal (NR), or no treatments (C), at Finse, alpine south Norway in 2002. F values, degrees of freedom and P values are shown for significant time (t) × treatment interactions, as tested by repeated measures anova. To improve readability, standard errors are not shown. n = 10 for each treatment.
For Carex, the combination of Dryas removal and nutrient addition clearly increased the number of leaves, primarily because of late season growth (July, August) when leaf production in the other treatments had terminated (significant time × removal × nutrients, Fig. 4a). No other seasonal growth pattern differed significantly between treatments on the vegetative growth of Carex.
Means of numbers (a) and length (b) of Carex vaginata leaves measured in late May, late June, late July and late August, in a Dryas octopetala heath, outside (controls) and inside OTCs, on ramets receiving nutrients (N), removal (R), nutrients and removal (NR), or no treatment (C), at Finse, alpine south Norway in 2002. F values, degrees of freedom and P values are shown for significant time (t) × treatment interactions, as tested by repeated measures anova. To improve readability, standard errors are not shown. n = 10 for each treatment.
sexual reproduction and mortality
For the Thalictrum ramets, 2.5%, 5.0%, 37.5% and 3.8% reproduced sexually in 2000, 2001, 2002 and 2003, respectively. Dryas removal or nutrient addition had no effect on flowering frequency of Thalictrum (Pearsons χ2 < 0.001, P > 0.10 in both cases). Warming on the other hand, decreased flowering frequency by 32.5% (Pearsons χ2 = 8.90, P = 0.003) but increased the height of the flower stems of Thalictrum by 36.2% (Mann–Whitney U = 89, P = 0.061, Fig. 5a). There were no effects of any of the treatments on number of flowers per ramet (Mann–Whitney U < 71, P > 0.311 in all cases, Fig. 5b), whereas nutrient addition marginally decreased the number of seeds per flower by 25.5% (Mann–Whitney U = 86, P = 0.093, Fig. 5c).
Means and standard errors of (a) the height of flower stems, (b) number of flowers per ramet, and (c) number of seeds per flower of Thalictrum alpinum measured in a Dryas octopetala heath by the end of the 2002 and 2003 growing season outside (controls) and inside OTCs, on ramets receiving nutrients (N), removal (R), nutrients and removal (NR), or no treatment (C), at Finse, alpine south Norway. Numbers in the columns are sample sizes for each treatment.
For Carex, 95%, 2.5%, 2.5% and 5.0% of the target ramets reproduced sexually in 2000, 2001, 2002 and 2003, respectively. Dryas removal and warming increased flowering frequency by 13.6% (Pearsons χ2 = 5.97, P = 0.015) and 11.3% (Pearsons χ2 = 3.82, P = 0.051), respectively. Nutrient addition had no effect on flowering frequency of Carex (Pearsons χ2 = 2.22, P = 0.136). None of the treatments had any effects on the height of flower stem of Carex (Mann–Whitney < 80, P > 0.100 in all cases).
Thalictrum mortality decreased by 2.5%, 12.5% and 7.5% due to Dryas removal, warming and nutrient addition, respectively, although none of these responses were statistically significant (Pearsons χ2 < 0.001, P > 0.98 in all cases).
Discussion
biotic effects
Species interactions clearly affected plant growth, and thus could potentially influence population dynamics, of Thalictrum alpinum and Carex vaginata at alpine Finse, which is in contrast to Hobbie et al. (1999), who found no effects of removal treatments on the remaining species in Arctic tundra, and thus no evidence for species interactions.
Removal of Dryas increased the number of leaves of Thalictrum alpinum and Carex vaginata, and flowering frequency of Carex (Table 3). On the other hand, leaf stalks of Thalictrum and leaves of Carex became significantly shorter, suggesting a shift in internal resource allocation in both species after Dryas removal. The increased number of leaves (Thalictrum and Carex) and flowering frequency (Carex) after removal of Dryas, suggest competitive rather than facilitative impacts of Dryas at Finse. This is in line with Klanderud & Totland (in press), who found that Dryas decreased plant community diversity at Finse, and it is in line with other removal experiments detecting negative plant–plant interactions in alpine and subarctic sites (Aarssen & Epp 1990; Shevtsova et al. 1995). The same neighbour may, however, have multiple effects (Aarssen & Epp 1990), and the reduced length of Thalictrum leaf stalks and Carex leaves after Dryas removal may be caused by a loss of facilitative shelter against low temperature and strong winds. Bret-Harte et al. (2004) found possible facilitation of Ledum palustre on one forb in the Alaskan tundra. Moreover, Shevtsova et al. (1995) found decreased shoot length in some species after neighbour removal in sub-arctic Finland, which they suggested could be due to either a loss of facilitative protection or a release from negative shading effects. Shading may be considerable at high latitudes with low angle of sunlight and low annual solar input (Shevtsova et al. 1995), and the plants may produce larger leaves (e.g. Dormann & Woodin 2002; Totland & Esaete 2002), or longer shoots, to compensate for shading or to position flowers, seeds or photosynthetically active tissue above neighbouring plants (Grime 1979). When associated vegetation is removed, the influx of solar radiation increases, and the plants no longer need to compensate, or escape from shading, and this may result in shorter leaves and leaf stalks. However, the positive warming effects on various growth and reproductive parameters at Finse suggest that loss of facilitation effects may explain these responses, at least in part.
Table 3. Summary responses of treatments with significant effects (P < 0.05, P > 0.05 in parentheses) on growth and reproductive parameters of Thalictrum alpinum and Carex vaginata growing in alpine Dryas octopetala heaths, as tested by different methods (see text). + = positive response; – = negative response; empty cells indicate no response (P ≥ 0.1), and dots (.) indicate that no tests are done. Treatment abbreviations as in Table 2
Species and parameters
Treatments and responses
+ T
+ N
+ R
+ T + N
+ N + R
+ T + R
+ T + N + R
Thalictrum alpinum
No. of leaves
+
+
Leaf area
Length of leaf stalk
+
–
(–)
Height of flower stem
(+)
.
.
.
.
No. of flowers
.
.
.
.
No. of seeds
(–)
.
.
.
.
Flowering frequency
–
.
.
.
.
Mortality
.
.
.
.
Carex vaginata
No. of leaves
+
+
(+)
Leaf length
–
Height of flower stem
.
.
.
.
Flowering frequency
+
+
.
.
.
.
Dryas had similar effects on the vegetative growth of Thalictrum and Carex, although leaf production for Thalictrum increased proportionally more than for Carex after Dryas removal, suggesting a stronger role of competition from Dryas on Thalictrum. For sexual reproduction on the other hand, no Dryas effects were detected for Thalictrum, whereas competition from Dryas reduced the sexual reproductive effort of Carex.
abiotic effects
Warming had positive effects on sexual reproductive parameters of the two target species, such as the height of flower stems of Thalictrum alpinum and flowering frequency of Carex vaginata (Table 3). Nutrients, on the other hand, increased the vegetative growth of both Thalictrum and Carex, although in most cases only in combination with warming. This shows that both warming and nutrient availability limited vegetative growth of these species at Finse (Table 3). Increased nutrient availability had no effect on the sexual reproduction of the target species, except a close to significant negative effect on the number of seeds of Thalictrum. These results are in line with other climate change experiments, showing that nutrients limit plant growth in alpine and arctic environments, with only minor effects of temperature alone, but often with synergistic effects of warming and nutrients (e.g. Chapin & Shaver 1985; Robinson et al. 1998; K. Klanderud & Ø. Totland, unpublished data). Warming, on the other hand, has been shown to increase the reproductive effort of several alpine and arctic species (Arft et al. 1999). These results are also in line with the meta-analysis of Dormann & Woodin (2002), although they found that nutrients also limit reproduction of arctic plants.
Vegetative growth increased more after nutrient addition in Carex than in Thalictrum, which is also in line with a number of experiments showing that graminoids respond more to increased nutrient availability than other functional groups (e.g. Jonasson 1992; Dormann & Woodin 2002; Bret-Harte et al. 2004; K. Klanderud & Ø. Totland, unpublished data).
species interactions
Increased nutrient availability caused by climate change may modify the interactions between Dryas and other species at Finse. Nutrients appeared to be a limiting factor for plant growth at Finse, and the increase in late season leaf production of Thalictrum and Carex after Dryas removal combined with nutrient addition, suggests that these species benefit more from the increased nutrient availability when Dryas was removed. This is in line with Bret-Harte et al. (2004) who found increased graminoid biomass after removal treatment combined with fertilization, probably in response to increased nutrient availability when neighbours were removed. My results indicate that Dryas is a stronger competitor for nutrients than Thalictrum and Carex, which again may suggest that the role of competition from Dryas may be greater if nutrient availability increases under global warming.
Furthermore, warming reduced Thalictrum flowering frequency, while nutrient addition slightly decreased the number of seeds of Thalictrum, which may suggest that the possible negative effects of growing inside a Dryas mat may be greater when resources increase, most likely because of increased competition from Dryas. This is in line with findings of a greater role of competition under higher productivity (e.g. Bertness & Shumway 1993; Callaway 1998; Choler et al. 2001; Klanderud & Totland, in press), and it corresponds with Callaway et al. (2002), who found shifts from positive to negative species interactions of alpine plants when temperatures increased along latitudinal and altitudinal gradients.
The predicted response to increased nutrient availability varies among functional groups (e.g. Dormann & Woodin 2002). Grasses have been shown to respond strongly to nutrient addition, and increased nutrient availability caused by climate change may influence competition hierarchies, and thus species composition of plant communities. This is in line with Grime (1977) and others, who have found altered competition hierarchies in several ecosystems after changes in nutrient conditions (Austin & Austin 1986; DiTomasso & Aarssen 1989; Gurevitch & Unnasch 1989), whereas species interactions have not changed after short-term warming treatments alone (Hobbie et al. 1999). Overall, Dryas appeared to protect other species from low temperatures at Finse, but at the same time, there may be interspecific competition for nutrients. Thus, global warming may result in a shift towards increased competition, because increased nutrient availability may change competition hierarchies, resulting in more competition for light and space.
implications for climate change models
Previous climate change experiments have primarily focused on abiotic effects on plant growth, and most climate change models tend to ignore species interactions (Mooney 1991; Pacala & Hurtt 1993; Davies et al. 1998). This study clearly shows that both biotic and abiotic conditions affect alpine plant growth and possible population dynamics. Furthermore, climate change may alter species interactions, e.g. the role of competition from Dryas or other species may become greater in a future with more resources. Thus, species interactions should be considered in climate change experiments and in models predicting future plant community responses to global warming.
Acknowledgements
I thank Malcolm Press, Sarah Richardson and one anonymous referee for improving the manuscript, Ørjan Totland for having the idea for this study and for constructive comments, Maria McEvoy and Lindsay Haddon for polishing the language, Vibeke Arnesen and Ingunn Heggstad for field assistance, and Finse Research Station for hospitality. The Norwegian Research Council provided financial support.
1365-2745/images/journal_branding.png)
