740 Research Paper

The Impact of Altitude and Simulated Herbivory on the Growth and Carbohydrate Storage of albus

U. Scheidel and H. Bruelheide Institute of Geobotany and Botanical Garden, Martin-Luther-University Halle-Wittenberg, AmKirchtor 1, 06108 Halle, Germany

Received: February 9, 2004; Accepted: September 1, 2004

Abstract: We tested the hypothesis that higher respiratory Printz (1933) proposed that the warmer lowland climate leads losses caused by higher temperatures in the lowlands, com- to higher respiratory losses of storage carbohydrates, especial- pared to montane sites, prevent growth of the montane hemi- ly during the winter (Dahl, 1992, 1998). This hypothesis has cryptophyte Petasites albus (). In addition, we tested been tested on several species (e.g., Stewart and Bannis- whether increased levels of herbivory enhanced carbon losses at ter, 1974; Crawford and Palin, 1981; Graves and Taylor, 1988; lower elevations. Rhizomes of Petasites albus were transplanted Arnone and Körner, 1997; Bruelheide and Lieberum, 2001). to a montane and a lowland site. In the subsequent three grow- Carbon losses from various sources could result in a less posi- ing seasons the were artificially defoliated to simulate tive or even negative carbon balance. The final effect would be mollusc herbivory. Whereas there were no altitudinal differ- less available carbon for growth in the next spring and reduced ences in the leaf number per plant, the leaf area was higher at flowering. In the long term, such plants may produce less bio- the montane site. At the montane site, the leaf number and leaf mass at low as compared to high altitudes. Altitudinal limita- area decreased with increasing damage, and the rhizome dry tion may also be caused by biotic interactions that are modi- weight in the third year was much higher in the undamaged fied indirectly by climatic effects. In the lowlands, high-alti- plants. In contrast, fructan concentrations in the rhizomes that tude plants may suffer from competition with lowland spe- were harvested at the end and at the beginning of the growing cies (e.g., Woodward and Pigott, 1975; Woodward, 1988) or seasons were generally higher at the lowland site. No clear de- from more severe herbivory compared to higher altitude sites foliation effects were observed on most harvest dates. The re- (e.g., Galen, 1990; Kelly, 1998; Suzuki, 1998; Bruelheide and sults indicate that the lower altitudinal limit of Petasites albus Scheidel, 1999). The objective of this study was to investigate cannot be explained by the negative effects of higher tempera- the direct effects of climate combined with the effects of arti- tures or more leaf damage by herbivores in the lowlands, either ficial leaf damage on the growth and carbohydrate storage of alone or in combination. An explanation will require considera- the montane hemicryptophyte Petasites albus (Asteraceae). tion of other site factors such as competition and possibly inter- Like many Asteraceae and several other plant species (Meier actions with herbivory and carbohydrate storage. and Reid, 1982; Hendry, 1987), P. albus accumulates fructans instead of starch as its major reserve carbohydrate. We trans- Key words: Defoliation, fructan storage, geographical range planted P. albus individuals to a lowland site and a montane limits, plant distribution, transplantation. site, and hypothesized that: (1) lowland transplants grow less vigorously than highland transplants with respect to above- ground and belowground biomass, and (2) this reduced growth is due to a reduced fructan content. Introduction An alternative explanation would be that P. albus suffers more The existence of plant range boundaries towards lower alti- from herbivory at low altitudes. Leaves of P. albus are readily tudes in temperate humid climates is a phenomenon that is eaten by molluscs (Hägele and Rahier, 2001; Scheidel and not easily explained by plant physiological requirements, giv- Bruelheide, 2001), and leaf damage to native P. albus popula- en the fact that growth conditions are generally considered to tions in the Harz Mountains increased with decreasing ele- be more favourable in the lowlands. To explain lower altitudi- vations (Scheidel et al., 2003). We simulated leaf herbivory nal boundaries, two hypotheses have been proposed: a direct to test the hypothesis (3) that increasing levels of defoliation effect of lowland climate on plant performance and effects of result in reduced growth of P. albus. biotic interactions that are indirectly influenced by climate. Materials and Methods

Study species

Plant Biol. 6 (2004): 740± 745 The Asteraceae species Petasites albus (L.) Gaertn. is a hemi-  Georg Thieme Verlag KG Stuttgart ´ New York cryptophyte that occurs in Europe from Spain to Russia and DOI 10.1055/s-2004-830352 further east to the Altai Mountains (Hegi, 1987). In Central Eu- ISSN 1435-8603 rope, except for the Baltic Sea coast, most populations are Impact of Altitude and Herbivory on Fructan Storage Plant Biology 6 (2004) 741 found at altitudes higher than 200 m a.s.l. In the Harz Moun- At each site the plants were randomly assigned to one of three tains in Central Germany this species occurs mainly between levels of artificial damage treatments, 50 controls, 50 with 370 and 470 m (Dierschke et al., 1983), with a lower distribu- low-level damage, and 50 with high-level damage. For the tion limit at about 300 m. P. albus is found on nutrient-rich for- low level, all completely unrolled leaves were cut perpendicu- est slopes and along forest edges, in high montane vegetation lar to the midrib at about 50% of the midrib length, resulting in dominated by tall herbs, and is a pioneer species on disturbed a mean leaf area loss of 22%, calculated from 30 randomly cho- soils of montane road edges and river banks. Most populations sen leaves whose area was measured. For the high-level dam- have a large range due to clonal fragmentation of the rhizomes age, unrolled leaves were cut at about 75% of the midrib that can reach a length of 20±40 cm before basal decay after length, resulting in a mean leaf area loss of 44%. This was in some years. In summer, P. albus forms dense stands of circular the range of the maximum loss found by Scheidel et al. (2003) 15±30 cm tall leaves which are continuously produced and in native P. albus populations at their lower altitudinal distri- senesce throughout the entire growing season. In addition to bution limit in the Harz Mountains. In 1999, the damage treat- the damaging effects of molluscs and insects on the leaves, ment to all new leaves was applied on four dates between the other herbivores may cause biomass losses, in particular larvae beginning of June and the middle of August; whereas in the of Lepidoptera mining into the rhizomes (Scheidel et al., 2003). growing season of 2000 the plants were damaged on five dates between the beginning of May and the middle of August. In Experimental design 2001, the plants were damaged on four dates between the end of May and the beginning of August. At the beginning of March 1999, rhizomes were collected from several genets of a large submontane population along a 1-km To determine leaf production, the number of leaves on each road edge near the village of Zorge in the Southern Harz Moun- plant was assessed every two to four weeks in all three grow- tains, Lower Saxony, Germany, at about 440 m. Apical rhizome ing seasons. All leaves were considered that had a developed segments, without inflorescence buds and about 5 cm long, (unrolled) lamina and were non-senescent. At the time, when were cut, washed, and weighed. Ten rhizomes were frozen at the maximum leaf number was attained in each of the three about ± 188C for fructan analysis. Another 300 rhizomes were growing seasons, between the end of July and the beginning planted individually into 4-l plastic pots filled with loamy gar- of August, the maximum width of each unrolled non-senes- den soil, and placed outdoors in the Experimental Botanical cent leaf was measured. After an allometric regression using Garden at Göttingen. For about three weeks after potting, the 30 randomly chosen leaves from a submontane P. albus pop- plants were repeatedly watered. ulation whose maximum width and area were determined (r2 = 0.832; p < 0.001), the total leaf area per plant, including Using a two-factorial design, the pots were randomly assigned the removed leaf area of the damaged plants, was calculated to two transplantation sites and three defoliation levels. In the from each leaf width. middle of April, 150 pots each were distributed to two loca- tions, a lowland site and a montane site. The low elevation site The plants were harvested on five successive dates, after se- was in the Botanical Garden of Göttingen at about 175 m, nescence in autumn (November) and before emergence in where the pots were buried in a 6 ” 25 grid in the sandy sub- spring (March): in the autumn of 1999, in the spring and au- strate of a lysimeter basin. The montane site was a fenced area tumn of 2000, and in the spring and autumn of 2001. On each in the courtyard of the Road Service Station near Braunlage harvesting date, 10 plants from each of the damage levels were in the Harz Mountains at about 600 m, where the pots were randomly selected. The plants were washed carefully to re- buried in the same pattern in a regularly mown lawn area. Sin- move all soil and senescent plant material from the rhizomes gle slugs and insect larvae, which caused only small leaf area and roots and were frozen at approximately ± 188C. losses of less than 5%, were removed when found on the leaves or the soil surface on the sampling dates. In addition, at both In the summer of 2002, the fructan content of the rhizomes sites some rhizomes were found on the harvesting dates to be was analysed. The plants were allowed to thaw, dissected into partly attacked by mining Lepidoptera larvae. rhizome and non-rhizome parts (buds, roots), and from each plant three rhizome sections with a length of about 1cm were Temperature and precipitation differ clearly between the two cut at randomly chosen age stages. Rhizome and non-rhizome transplantation sites. The mean air temperature in January parts were weighed, dried at 808C for about 24 h, and weighed (average values from 1961 to 1990; Deutscher Wetterdienst, again. Since age definition from internode lengths was uncer- 1999±2001) is 0.38C in Göttingen (weather station at 167 m) tain due to interrupted rhizome growth during dry periods in and ± 2.28C in Braunlage (weather station at 697 m). The summer, and since preliminary fructan analyses gave no clear mean temperature in July is 17.18C in Göttingen and 14.28Cin differences in the fructan contents of one-, two-, or three- Braunlage. The mean annual precipitation amounts to 698 mm years-old rhizome parts, the three 1-cm rhizome sections were in Göttingen and 1265 mm in Braunlage. All three years of the analysed together for fructan using an enzymatic Fructan As- experiment were characterized by higher temperatures than say Kit (Megazyme Ltd.). It can be assumed that, in fructan- the mean of 1961 to 1990 (Deutscher Wetterdienst, 1999± accumulating plant species, other polysaccharides, such as 2001). starch, play only a minor role as reserve carbohydrates (Hen- dry, 1987) and can therefore be neglected. The principle of the At both sites the plants were watered during longer dry peri- Megazyme procedure is to first hydrolyse sucrose, starch, and ods, which occurred more often in Göttingen than in Braun- maltosaccharides. The resulting reducing sugars were then lage. During the first hot period in May 1999 six plants died at reduced to their sugar alcohols. Thereafter, the fructans were the Braunlage site; they were not replaced. During the course hydrolysed to fructose and glucose with purified fructanase, of the experiment the plants were not fertilized. Individuals of and the reducing sugars were measured colorimetrically. other plant species germinating in the pots were removed. 742 Plant Biology 6 (2004) U. Scheidel and H. Bruelheide

Fig.1 Mean total area of the unrolled, non-senescent leaves per sites in Göttingen (left) and Braunlage (right). Only those plants har- plant, calculated by allometric regression from leaf width measured vested in the autumn of 2001 (n = 9 to 10) were included in the calcu- on one sample date in late July or the beginning of August 1999, lations. Error bars indicate standard deviation. 2000, and 2001 for three damage levels and the two transplantation

Data analysis Table 1 Results of repeated measures two-way ANOVA on the effects of transplantation site (Braunlage and Göttingen) and damage level All calculated parameters were tested for normal distribution (0%, 22%, and 44%) on total leaf area per plant on a sample date (proc univariate, Shapiro-Wilk-statistics; SAS Institute, 2000). in the summer of 1999, 2000, and 2001. * p < 0.05, ** p < 0.01. The Effects of transplantation site and damage level on maximum post-hoc tests refer to analysis of the temporal contrasts between suc- leaf number and total leaf area were measured repeatedly on cessive years the same plants and tested with a repeated measures two- Source of variation df F way ANOVA. Effects of transplantation site, damage level, and harvesting date on rhizome dry weight and on relative and ab- Site 1 24.01** solute fructan content refer to different plants and were tested Damage 2 2.87 with a three-way ANOVA (proc glm; SAS Institute, 2000). Site ” damage 2 5.80** Date 2 57.00** Results Date ” site 2 16.38** Date ” damage 4 2.49* Leaf number and leaf area Date ” site ” damage 4 5.73**

The maximum leaf number and the total leaf area per plant in- Post-hoc tests creased during the course of the experiment (Fig.1, Table 1). 1999 ® 2000 Date 1 47.22** Whereas the leaf number did not differ between the transplan- Site 1 23.73** tation sites or among the damage levels, the Braunlage plants Damage 2 6.43** produced a significantly larger total leaf area than the Göttin- gen plants. Only in the Braunlage plants was the leaf area re- 2000 ® 2001 Date 1 43.72** duced with increasing damage level, resulting in a significant Site 1 9.58** interaction of transplantation site and damage level, but no Damage 2 1.30 overall damage effect.

Rhizome biomass an extremely high rhizome mass of the undamaged Braunlage The rhizomes contributed about half of the total dry mass of plants on the final sample date, whereas the differences were the harvested plants, with no consistent differences in the bio- smaller or contrary on earlier dates. Therefore, the effects of mass partitioning into rhizomes, buds, and roots among the site, damage, and sample date interacted significantly. sites or damage levels. The rhizome dry weight increased dur- ing the growing seasons and remained constant or decreased Fructan content slightly over winter (Fig. 2A), resulting in a highly significant effect of the sampling date (Table 2). The highest dry mass in- The relative and absolute fructan contents increased during crement was found in the growing season of 2001(Fig. 2A), the growing seasons and decreased over winter (Figs. 2B,C), which was characterized by a lower mean temperature and a resulting in highly significant effects of the sample date. Dur- higher sum of precipitation than in 1999 and 2000 (Deutscher ing the relatively wet and cold growing season of 2001, when Wetterdienst, 1999±2001). Overall, the rhizome dry mass was the rhizomes showed the highest growth, the relative fructan significantly higher in Braunlage than in Göttingen, and sig- content increased only slightly. Overall, the relative fructan nificantly higher in the undamaged individuals compared to content was significantly higher in Göttingen than in Braun- the damaged plants (Table 2). These effects largely result from lage (Table 2). It was not affected by the damage level in gener- Impact of Altitude and Herbivory on Fructan Storage Plant Biology 6 (2004) 743

Fig. 2 Mean total rhizome dry weight per plant after removal of buds centage fructan content of the rhizomes (C) for three damage levels and roots, harvested in the spring or autumn in the course of the and the two transplantation sites in Göttingen (left) and Braunlage experiment (A), mean fructan concentration [% dry weight] of the rhi- (right); n = 9 to 10. The spring 1999 values result fromten rhizomes zomes (B), and mean absolute fructan content [g] of the whole rhi- randomly chosen at the start of the experiment, before treatments zome per plant, calculated from the rhizome dry weight and the per- were applied. Error bars indicate standard deviation.

al. However, a significant interaction effect of sample date and Discussion damage showed different damage effects on certain sample dates. Due to the higher rhizome mass at the Braunlage site, The first hypothesis, better growth at the higher site, was sup- the absolute fructan content had no site effect. In contrast to ported by the larger total leaf area per plant at the montane the relative fructan content, the absolute content was depend- site. Additionally, only in the undamaged plants, the rhizome ing on the damage level, mainly resulting from the extremely mass on the harvesting dates was increasingly larger in the high fructan content of the undamaged plants at the montane montane plants than in the lowland plants. In contrast, other site on the final sample date, indicated by an interaction be- parameters did not differ between sites or were even greater tween the damage level, site and sample date. at the lowland site, such as rhizome mass in the spring of 2000.

Our second hypothesis clearly must be rejected, as the per- centage fructan content was consistently higher at the lowland site. In general, the high amounts of storage carbohydrates 744 Plant Biology 6 (2004) U. Scheidel and H. Bruelheide

Table 2 Results of a three-way ANOVA on the effects of transplanta- The third hypothesis is not unequivocally confirmed. The ex- tion site (Braunlage and Göttingen), damage level (0%, 22%, and 44%) pected negative effects of defoliation were mainly found at and harvesting date (autumn of 1999, spring and autumn 2000, spring the montane site. Therefore, the results provide no indication and autumn 2001) on rhizome dry weight, percentage fructan con- that the lower altitudinal range limit of P. albus might be the tent, and absolute fructan content per plant result of herbivory. In general, the influence of defoliation on Rhizome Percentage Absolute subsequent growth is difficult to predict. Compensation and dry fructan fructan even overcompensation can be assumed to be a frequent phe- weight content content nomenon (e.g., Paige, 1999), beginning in the remaining leaves df F F F that show enhanced photosynthetic activity or delayed senes- cence (e.g., Prins et al.,1989; Morrison and Reekie,1995). Wyka Site 1 5.93* 102.96** 0.28 (1999) found a significant reduction in non-structural carbo- Damage 2 10.06** 0.02 5.19** hydrates some days after complete defoliation of Oxytropis ser- Date 4 70.35** 128.26** 57.17** icea. In Trillium grandiflorum defoliation reduced the replen- Site ” damage 2 7.43** 0.91 4.60* ishment of storage organs to support reproduction function Site ” date 4 2.96* 0.66 0.77 (Lubbers and Lechowicz, 1989). In order to clearly demonstrate Damage ” date 8 5.90* 2.08* 3.00** detrimental effects on P. albus, it may be necessary to increase Site ” damage ” date 8 8.28** 1.80 7.12** the defoliation to a level higher than that applied in the pres- ent experiment. However, such a strategy would defoliate the plants to an even higher level than the maximum found in natural mollusc-damaged leaves at the lower altitudinal dis- found in Petasites albus are typical for plants growing at high tribution limit of P. albus (Scheidel et al., 2003). The natural altitudes, which start growth early in the spring (Russell, herbivore damage to P. albus is, however, not restricted to the 1940; Mooney and Billings, 1960; Bruelheide and Lieberum, leaves. In particular, larvae of Lepidoptera can be found mining 2001). Carbohydrate depletion in the spring and replenish- the rhizomes (Scheidel et al., 2003). Therefore, the experiment ment during the growing season, as exhibited by P. albus only allows conclusions to be drawn on the tolerance of P. al- throughout the experiment, have been found in several mon- bus to the experimental leaf herbivory but not to herbivory in tane and alpine plant species (e.g., Mooney and Billings, 1960; general. Stewart and Bannister, 1973; Zachhuber and Larcher, 1978). In high-altitude plants transplanted to lowland sites, a more rap- In summary, lowland climate and defoliation by 44% of the leaf id over-winter reduction of storage carbohydrate has been re- area do not prevent P. albus from growing well at the experi- peatedly reported (e.g., Mooney and Billings, 1965; Skre, 1993; mental sites over three growing seasons. An alternative expla- Bruelheide and Lieberum, 2001). In P. albus the lowland plant nation for this species altitudinal range limit might be the fructan content was higher than at the montane site, even in unexpected lower leaf area in the lowland site. In general, it is the spring. Poor regulation potential of montane or alpine spe- assumed that montane plant species make less use of more cies exhibiting unfavourably high respiration rates at higher favourable growth conditions in the lowlands, which by itself temperatures, as found, for example, by Stewart and Bannister would be a competitive disadvantage (Bruelheide and Liebe- (1974), Bell and Bliss (1979) or Crawford and Palin (1981), ob- rum, 2001). As the leaf area in P. albus was even reduced under viously does not apply to P. albus. Larigauderie and Körner lowland conditions, the competitive disadvantage should be (1995) reported a large variation in the acclimation potential much severe than generally assumed for montane and alpine of alpine plant species. Moreover, according to Körner (1999: plants (Woodward, 1988). However, the competition effect 201, 208), alpine plant growth is generally not limited by car- was not tested in our experiment as competitors were artifi- bon supply because low temperatures prevent carbon invest- cially excluded from the pots. Further experiments should ad- ment in growth. Wyka (1999) found that reductions in carbo- ditionally include the manipulation of competition. hydrate storage did not reduce winter survival of Oxytropis ser- icea, and rodent herbivory on the upper alpine Ranunculus gla- Acknowledgements cialis had no obvious effects on plant vigour (Diemer, 1996). If the assumption of low demand is also correct for P. albus, there The staff of the Experimental Botanical Garden of Göttingen must be an unknown limiting factor at the lowland site, other University helped with potting the rhizomes in the spring of than low temperatures, so that the plants were not able to in- 1999 and watering throughout the course of the experiment. vest in structural compounds and, instead, formed non-struc- The Road Service Station of Braunlage kindly allowed us to tural carbohydrates. However, fructan storage reflects not only use their courtyard. Maria Anna Scheidel helped with the bio- the output rate but also the input. In addition to the effect of metrical measurements. Marianne Gscheidlen performed the low temperatures (Chatterton et al., 1989; Jeong and Housley, laboratory fructan analyses. A graduate grant from Göttin- 1990; Thorsteinsson et al., 2002), fructan production may also gen University is gratefully acknowledged by U. Scheidel. The be enhanced by drought stress (De Roover et al., 2000). Dry pe- manuscript benefited from the constructive comments of riods occurred more often in the lowland climate and could J. M. H. Knops and three anonymous referees. therefore be considered as a component of the altitudinal dif- ferences. To distinguish between the altitudinal effects of tem- References perature and precipitation in future experiments, moisture conditions would have to be manipulated, e.g., by watering Arnone, J. A. and Körner, C. (1997) Temperature adaptation and accli- the pots, or sites would have to be repeated along precipitation mation potential of leaf dark respiration in two species of Ranun- gradients at the same altitude. culus from warm and cold habitats. Arctic and Alpine Research 29, 122±125. Impact of Altitude and Herbivory on Fructan Storage Plant Biology 6 (2004) 745

Bell, K. L. and Bliss, L. C. (1979) Autecology of Kobresia bellardii: why Mooney, H. A. and Billings, W. D. (1960) The annual carbohydrate winter snow accumulation limits local distribution. Ecological cycle of Alpine plants as related to growth. American Journal of Monographs 49, 377±402. Botany 47, 594±598. Bruelheide, H. and Lieberum, K. (2001) Experimental tests for deter- Mooney H. A. and Billings, W. D. (1965) Effects of altitude on carbo- mining the causes of the altitudinal distribution of Meum atha- hydrate content of mountain plants. Ecology 46, 750±751. manticum Jacq. in the Harz Mountains. Flora 196, 227±241. Morrison, K. D. and Reekie, E. G. (1995) Pattern of defoliation and its Bruelheide, H. and Scheidel, U. (1999) Slug herbivory as a limiting effect on photosynthetic capacity in Oenothera biennis. Journal of factor for the geographical range of Arnica montana. Journal of Ecology 83, 759±767. Ecology 87, 839±848. Paige, K. N. (1999) Regrowth following ungulate herbivory in Ipo- Chatterton, N. J., Harrison, P. A., Bennett, J. H., and Asay, K. H. (1989) mopsis aggregata: Geographic evidence for overcompensation. Carbohydrate partitioning in 185 accessions of Gramineae grown Oecologia 118, 316±323. under warm and cool temperatures. Journal of Plant Physiology Prins, A. H., Verkaar, H. J., and van den Herik, M. (1989) Responses of 134, 169±179. Cynoglossum officinale L. and Senecio jacobaea L. to various degrees Crawford, R. M. M. and Palin, M. A. (1981) Root respiration and tem- of defoliation. New Phytologist 111, 725±731. perature limits to the north-south distribution of four perennial Printz, H. (1933) Granens og furuens fysiologi og geografiske utbre- maritime plants. Flora 171, 338±354. delse. Nytt Magasin for Naturvidenskapene 73, 167±219. Dahl, E. (1992) Relations between macro-meteorological factors and Russell, R. S. (1940) Physiological and biological studies on an Arctic the distribution of vascular plants in Northern Europe. Universitet vegetation. III. Observations on carbon assimilation, carbohydrate Trondheim Vitenskapsmuseet, Rapport Botanisk Serie 1992, 31 ± storage and stomatal movement in relation to the growth of plants 59. on Jan Mayen Island. Journal of Ecology 28, 289±309. Dahl, E. (1998) The Phytogeography of Northern Europe (British Isles, SAS Institute (2000) SAS Procedures Guide, Version 8. Cary, NC. Fennoscandia and Adjacent Areas). Cambridge: Cambridge Univer- Scheidel, U. and Bruelheide, H. (2001) Altitudinal differences in her- sity Press. bivory on montane Compositae species. Oecologia 129, 75±86. De Roover, J., Vandenbranden, K., Van Laere, A., and Van den Emde, W. Scheidel, U., Röhl, S., and Bruelheide, H. (2003) Altitudinal gradients (2000) Drought induces fructan synthesis and 1-SST (sucrose:su- of generalist and specialist herbivory on three montane Astera- crose fructosyltransferase) in roots and leaves of chicory seedlings ceae. Acta Oecologica 24, 275±283. (Cichorium intybus L.). Planta 210, 808±814. Skre, O. (1993) Photosynthesis and respiration in leaves of mountain Deutscher Wetterdienst (1999±2001) Witterungsreport express. birch (Betula pubescens Ehrh.) and related tree species. Meddelel- Offenbach: Deutscher Wetterdienst. ser fra Skogforsk 45, 1±23. Diemer, M. (1996) The incidence of herbivory in high-elevation pop- Stewart, W. S. and Bannister, P. (1973) Seasonal changes in carbohy- ulations of Ranunculus glacialis: a re-evaluation of stress tolerance drate content in three Vaccinium species with particular reference in Alpine environments. Oikos 75, 486±492. to V. uliginosum and its distribution in the British isles. Flora 162, Dierschke, H., Otte, A. and Nordmann, H. (1983) Die Ufervegetation 134 ±155. der Flieûgewässer des Westharzes und seines Vorlandes. Natur- Stewart, W. S. and Bannister, P. (1974) Dark respiration rates inVacci- schutz und Landschaftspflege in Niedersachsen, Beiheft 4, 1±83. nium ssp. in relation to altitude. Flora 163, 415±421. Galen, C. (1990) Limits to the distributions of Alpine tundra plants: Suzuki, S. (1998) Leaf phenology, seasonal changes in leaf quality and herbivores and the Alpine skypilot, Polemonium viscosum. Oikos herbivory pattern of Sanguisorba tenuifolia at different altitudes. 59, 355±358. Oecologia 117, 169±176. Graves, J. D. and Taylor, K. (1988) A comparative study of Geum rivale Thorsteinsson, B., Harrison, P. A., and Chatterton, N. J. (2002) Fructan L. and Geum urbanum L. to determine those factors controlling and total carbohydrate accumulation in leaves of two cultivars of their altitudinal distribution. II. Photosynthesis and respiration. timothy (Phleum pratense Vega and Climax) as affected by temper- New Phytologist 108, 297±304. ature. Journal of Plant Physiology 159, 999 ±1003. Hägele, B. F. and Rahier, M. (2001) Determinants of seasonal feeding Woodward, F. I. (1988) Temperature and the distribution of plant of the generalist snail Arianta arbustorum at six sites dominated by species. In Plants and Temperature (Long, S. P. and Woodward, F. . Oecologia 128, 228±236. I., eds.), Cambridge: Society for Experimental Botany, pp. 59±75. Hegi, G. (1987) Illustrierte Flora von Mitteleuropa, Vol. VI/4. Mün- Woodward, F. I. and Pigott, C. D. (1975) The climatic control of the al- chen, Hamburg: Compositae II, 3rd edn. Parey. titudinal distribution of Sedum rosea L. Scop. and S. telephium L. I. Hendry, G. (1987) The ecological significance of fructan in a contem- Field observations. New Phytologist 74, 323 ±334. porary flora. New Phytologist 106 (Suppl.), 201 ±216. Wyka, T. (1999) Carbohydrate storage and use in an Alpine popula- Jeong, B. and Housley, T. L. (1990) Fructan metabolism in wheat in al- tion of the perennial herb, Oxytropis sericea. Oecologia 120, 198± ternating warm and cold temperatures. Plant Physiology 93, 902 ± 208. 906. Zachhuber, K. and Larcher, W. (1978) Energy content of different al- Kelly, C. A. (1998) Effects of variable life history and insect herbivores pine species of Saxifraga and Primula depending on their altitudi- on reproduction in Solidago macrophylla (Asteraceae) on an eleva- nal distribution. Photosynthetica 12, 436±439. tional gradient. American Midland Naturalist 139, 243±254. Körner, C. (1999) Alpine Plant Life: Functional Plant Ecology of High Mountain Ecosystems. Berlin, Heidelberg, New York: Springer. H. Bruelheide Larigauderie, A. and Körner, C. (1995) Acclimation of leaf dark respi- Institute of Geobotany and Botanical Garden ration to temperature in alpine and lowland plant species. Annals Martin-Luther-University Halle-Wittenberg of Botany 76, 245±252. Am Kirchtor 1 Lubbers, A. E. and Lechowicz, M. J. (1989) Effects of leaf removal on 06108 Halle reproduction vs. belowground storage in Trillium grandiflorum. Germany Ecology 70, 85±96. Meier, H. and Reid, J. S. G (1982) Reserve polysaccharides other than E-mail: [email protected] starch in higher plants. In Encyclopedia of Plant Physiology,Vol.13A (Loewus, P. A. and Tanner, W., eds.), Berlin: Springer, pp. 418±471. Editor: J. Knops