Réf. Paulsen & al 2000 - A

Référence bibliographique complète

PAULSEN, J., WEBER, U.B., KÖRNER, C. 2000. Tree growth near treeline: abrupt or gradual reduction with altitude? Arctic, Antarctic and Alpine Research, 32, 14–20.

Abstract: Natural climatic treelines are relatively discrete boundaries in the landscape established at a certain elevation within an otherwise continuous gradient of environmental change. By studying tree rings along elevational transects at and below the upper treeline in the European Alps, we (1) determine whether radial stem growth declines abruptly or gradually, and (2) test climatic influences on trees near treeline by investigating transects for climatically different historical periods. While tree height decreases gradually toward the treeline, there is no such general trend for radial tree growth. We found rather abrupt changes which imply threshold effects of temperature which moved upslope in a wave-like manner as temperatures increased over the past 150 yr. Currently radial tree growth at treeline in the Alps is the same magnitude as at several hundred meters below current treeline. Over short intervals, tree-ring width is more dependent on inter-annual climatic variability than on altitudinal distance to treeline. We conclude that (1) the elevational response of tree-rings includes a threshold component (a minimal seasonal temperature) and that (2) radial growth is more strongly correlated with year to year variation in climate than with treeline elevation as such. Our data indicate that the current treeline position reflects influences of past climates and not the current climate.




Organismes / Contact

Institute of Botany. University of Basel. Schönbeinstrasse 6, CH-4056 Basel. Switzerland. jens.paulscn@unibas.ch


(1) - Paramètre(s) atmosphérique(s) modifié(s)

(2) - Elément(s) du milieu impacté(s)

(3) - Type(s) d'aléa impacté(s)

(3) - Sous-type(s) d'aléa


Vegetation : trees near treeline




Pays / Zone

Massif / Secteur

Site(s) d'étude



Période(s) d'observation

- Switzerland

- Austria

- Cantons: Valais, Bern, Vaud

- Land: Tirol

9 sites [see the study]



The past 150 yr


(1) - Modifications des paramètres atmosphériques




Compared to the second half of the 19th century present mean temperatures represent a warming of 1.5 K for the growing period (June to September) al treeline elevations in the Alps. However, this long-term trend is superimposed on more pronounced short-term fluctuations. At Grand St- Bernard Station (200 m above outpost treeline), within any period of 20 yr the coolest year was between 2.5 to 4.0 K cooler than the warmest. An extreme example is the growing seasons in 1911 and 1912, in which the mean temperature was 7.1°C in 1911, but only 2.9°C the following year.

These interannual temperature fluctuations correspond to a difference in elevation of several hundred meters (e.g., 3 K / 0.6 K 100 m-1 = 500 m). On the other hand, the elevational range across which radial tree growth rapidly declines within the ecotone is not wider than 100 m of elevation at all studied sites, which corresponds to a temperature difference of 0.6 K. This small difference (much smaller than interannual differences) suggests a signal that becomes critical for treeline formation only when integrated over many years.

A general warming of 1.5 K (which corresponds to 250 m of elevation) during the growing period since 1820 was measured at a meteorological station 200 m above climatic treeline (Grand St. Bernard).






Informations complémentaires (données utilisées, méthode, scénarios, etc.)

Climatic data were taken from Grand St. Bernard Station, the only alpine station of the Swiss Meteorological Institute (SMI) which provides monthly mean air temperatures since 1818. The station is situated at 2419 m a.s.l., approximately 200 m above the "outpost treeline" and at 60 to 80 km distance from the Swiss coring sites.


(2) - Effets du changement climatique sur le milieu naturel





The closer trees were situated to the outpost treeline the more the age of the sampled, dominant trees decreased. However, with reference to their 50- to 80-cm coring height, the authors found I00-yr-old or older non-hollow trees within the uppermost 50 m of the ecotone at all sampling locations. Even 200-yr-old individuals (pith age al coring position) were found 50 m below the present “outpost treeline” (roughly corresponding to the current treeline). If one accounts for the years elapsed between tree germination and coring height, this indicates that there was no major shift in treeline elevation within the last ca. 250 yr.


Tree height of dominant trees decreases as one approaches the outpost treeline. With an error probability of < 10-4 this effect cannot he attributed to the somewhat lower age of trees at outpost treeline only, that is, trees in the upper zone are not just smaller because they are younger. The reduction of tree height with increasing elevation was sites-specific and varied between 2 and 17 m per 100 m of elevation. At a given tree age and position in the ecotone, Picea abies is significantly taller than Pinus cembra.


There is no difference in tree-ring width between Picea and Pinus but a distinct historical trend exists in the uppermost 250 m of the treeline ecotone. In the first part of the 19th century, annual increments linearly decreased with increasing elevation. After 1940, average tree-ring width within the uppermost 250 m below the outpost treeline was similar, irrespective of the e1evation of the tree location. The time period between 1860 and 1920 represents a period of transition. At the lower end of the ecotone, 250 m below outpost treeline, mean ring widths have not changed since 1840. Therefore, whether one detects a historical trend in tree-rings depends on the distance of the coring site to the outpost treeline.

The annual radial stem increment within the uppermost 100 m below outpost treeline did not increase continuously over the considered time period. The increase of tree-ring width occurred rather abruptly and with an elevation-specific time lag. Within only 15 to 20 yr, average radial increment doubled from 0.4 mm to 0.8 mm yr-1 across sites. Since then, ring width has never fallen below these means. This phenomenon started around 1850 for trees 50 to 100 m below present outpost treeline. Around 1870, it occurred for trees 30 to 50 m below the present outpost treeline, and for trees at present outpost treeline, the major increase is apparent after 1940.


A recent stimulation of radial tree growth near cold climate treelines has been documented previously [see references in the study]. (…) Tree-ring widths from the coring sites appear to reflect a warming in the upper treeline ecotone, but in a peculiar way. Currently, ring width at outpost treeline in the Alps is as wide as it was 250 m lower in the first part of the 19th century. A 1.5 K warming is analogous 10 an ascent of treeline of ca. 250 m. Recent shifts of treeline have been suggested by several authors (e.g., Griggs, 1946; Innes, 1991). However, the present data do not indicate a significant upslope migration over the last 180 yr, given the great abundance of 150 to 200-yr-old trees in the upper part of the ecotone. From the decrease of mean tree age with increasing elevation of 0.28 meters of elevation per one year, a theoretical ascent of outpost treeline of only 50 m since 1820 could be assumed (180 yr x 0.28 m a-1). If there is an upslope move of the outpost treeline of 50 m and a warming of the growing period of 1.5 K since 1820, the mean air temperature at the outpost treeline during the growing period is theoretically 1.2 K higher than it was 180 yr ago (50 m of elevation correspond to 0.3 K). The observed "conservative" behavior of treeline elevation in the Alps is well evidenced for most of the Holocene (Burga, 1988). Similarly, Petersson (1998) reports changes in treeline elevation of less than 100 m over palaeoperiods differing by 2 to 3 K. (…)

Il is unknown whether enhanced nitrogen deposition contributes to tree stimulation at treeline. The present data do not support such a critical role of nutrients. According to the authors’ analysis, the growth increase in the Alps occurred with a systematic time-lag with increasing elevation. A difference in e1evation of 50 m (corresponding to a horizontal distance of ca. 100 m) corresponds to a time-lag of 20 to 30 yr. However, increased nutrient input is uniform over large areas (Gäggelcr, 1997) and, thus, the observed dramatic changes in growth cannot be explained by a nutrient effect. Time-series analysis also led Nicolussi et al. (1995) to conclude that high-elevation trends in tree-ring width do not match nitrogen input dynamics of the 20th century in the Alps. Similarly such a threshold response of tree growth which is delayed with elevation does not strongly favour a stimulation of tree growth by atmospheric CO2-enrichmcnt alone. During the years when this upslope wave of growth stimulation occurred (in the 2nd half of the 19th century). CO2-enrichment was minute. It has been explained elsewhere (see discussion in Körner 1998) that it is unlikely that treeline trees have a carbon balance problem. They rather have a carbon investment problem associated with periodically low temperatures.

If temperature dominates growth responses of trees near treeline, the present data suggest that growth-limiting threshold temperatures for shoot or root meristems exist, because the stimulation of radial growth exceeds rates of any known proportional metabolic temperature response. Radial growth responses that occurred over the past 150 yr may reflect the operation of a "critical" temperature, above which responses are less temperature sensitive. Il remains to be resolved whether seasonal means, temperature sums above certain thresholds or other measures of direct or indirect temperature influences on meristems provide the best explanation.

The highest correlation between tree-ring width and monthly or seasonal mean air temperatures from a meteorological station was r = 0.52 (mean temperature of July). Remarkably. Peterson (1998) also found the best correlation (ca. 0.5) with summer temperature in high-elevation forests in the Olympic Mountains of western North America. This supports the view that temperature is the major driver of radial stem increment, but the way this signal is integrated over very long periods is not well understood and predictions for shorter periods seem to be very unprecise. During some periods the direction of trends in air temperature al Grand St-Bernard and radial stem increment near treeline were even opposite (e.g., 1851–1863, 1985–1992). The period of increased growth between 1855 and 1870 is not reflected in the air temperature data, and tree ring width between 1912 and 1920 was more than twice as great as between 1840 and 1850, although both periods exhibited the same temperature sum during the growing period. The authors do not understand why the observed growth increase is delayed with increasing elevation and why a difference of 50 m elevation corresponds to a delay of 25 yr. A narrow threshold temperature for growth-related processes appears to be most plausible. Factors other than temperature such as local soil moisture or nutrition arc unlikely causes of the general growth increase, because growth was measured at so many locations, with different local hydrology across the Alps.

Individual trees appear to respond very sensitively to temperature, but treeline does not. Sub- and supra-optimal years affect growth, but do not influence treeline position, except if they persist over very long periods of time (Peterson. 1998). Currently, there arc no trees in the Alps al elevations high enough where the temperature would limit radial tree growth to near zero increments. In order to detect early indications of treeline responses to global warming, the study of growth dynamics within the treeline ecotone, rather than at its upper boundaries (Slatyer and Noble, 1992), appears most promising.





(…) The authors dismiss the view that treeline formation is a question of seedling establishment, because almost everywhere seedlings may be found above outpost treeline. The mere existence of a krummholz bell illustrates that sporadic seedling establishment of woody plants is possible above treeline (except where tire is prohibitive, which is not the case at our sites). The critical phase for tree establishment is the stage when a young tree reaches a height where needles become closely coupled to the thermal and radiative conditions of the free atmosphere and do not continue to profit from radiative heating near the ground. [See references in the study] An advance of treeline could take place only if a long-term warming trend stimulates growth frequently enough (also in cooler yean), or, in other words, if low-temperature events/periods which limit growth are so rare that they are insignificant. (…)


Sensibilité du milieu à des paramètres climatiques

Informations complémentaires (données utilisées, méthode, scénarios, etc.)

Climatic high-elevation treelines are conspicuous vegetation boundaries that occur over a wide range of latitudes. Highest treeline positions (>4000 m) are found in subtropical Mountains of the northern hemisphere. At high latitudes, the elevational "alpine" treeline merges with the 1ow-atitude arctic treeline. Regionally a maritime climate can suppress treelines and a continental c1imate can facilitate higher treeline elevations (the so called "Massenerhebungseffekt").

Many explanations for the treeline phenomenon have been attempted in the past. Some correlative approximations such as the 10°C isotherm of the warmest month have been found to have predictive value on a local but not on a global scale. A worldwide consideration of treeline positions does not support the usefulness of notions such as 'treeline is a complex phenomenon and results from interactions of many different (!) environmental drivers". If this were true, natural alpine treeline positions would not correlate with one specific component of the mountain climate across the globe, which they do (e.g., seasonal mean temperatures of 5 to 7°C). Many phenomena (photosynthesis, winter water relations, branch maturation, snow and ice injury, seedling survival), which may be of local importance for treeline formation, are less significant on a global scale.

The slowing of tree growth within the treeline ecotone is a well acknowledged phenomenon in the forest literature (…). Because mean air temperature during the growing period decreases by ca. 0.6 K per 100 m elevation, a reduction of tree-ring width towards treeline may be interpreted as a direct reaction to decreasing temperature. In contrast, variation in ring width in trees at equal elevational distance from the treeline between different years or decades can be assumed to reflect climatic variation. The reconstruction of past climates from tree rings is a major topic in dendrochronology, but not the aim of the present study. Instead, the authors were interested in the growth response itself.

[See references in the study]

Across the treeline ecotone, which is the transition from tall closed forest “timberline” to the upper limit of the “krummholt” hell, stand density and tree vitality decreases rapidly with increasing elevation. The response of the radial growth of trees across a 250-m elevational gradient within the treeline ecotone of central Europe is the focus of the present study. Elevational trends in radial growth are explored in order to understand the upper limits of tree distribution.

In their analytical approach the authors used a curvature analysis of the elevational decline of tree-ring width in the uppermost zone of tree growth. They expected "initial slope" and "sharpness" of curvature (decline of ring width with elevation) near treeline to provide a "fingerprint" of the current and past dependence of tree growth on climate. The elevational response may be a linear or nonlinear decline, or its shape may vary over the years, depending on concurrent climate.

They intended to deduce early signals of a potential advance or retreat of high elevation forest boundaries from growth dynamics of mature trees within the ecotone, rather than from actual recruitment success beyond the current limit. They believe that seedling success has low predictive value as long as individuals are small and nested in the aerodynamic boundary of low-stature vegetation (Körner, 1998).


Cores were collected along straight elevational transects from the uppermost outpost tree downwards across the treeline ecotone at each of nine locations in the Swiss and Austrian Alps. Areas of obvious human influence on treeline position or abrupt changes in slope or soil properties were avoided. At all sampling locations the tree height continuously decreased with increasing elevation. At each location, dominant trees were sampled along one discrete transect. Only undamaged and undeformed trees were considered.

Tree elevation was determined by an atmospheric pressure altimeter, which was calibrated several times per day with topographic maps 1:25.000 (contour interval 20 m). Tree heights were estimated by eye. Cores were taken with a 5-mm increment borer al 50-80 cm height, parallel to the slope contour, thus minimising the influence of compression wood. Ring width (±0.01 mm) was measured and recorded with an electronic analysis bench and a microscope. (…)


[In the study], the authors describe how they standardized the raw data obtained from wood cores for differences in tree age, absolute elevation of outpost treeline and region specific tree vigor. (…)


(3) - Effets du changement climatique sur l'aléa










Paramètre de l'aléa

Sensibilité des paramètres de l'aléa à des paramètres climatiques

Informations complémentaires (données utilisées, méthode, scénarios, etc.)





(4) - Remarques générales


For practical reasons, in this analysis a tree is defined as an upright woody plant with a single above-ground stem that reaches a height of al least 3 m, independently of whether reproduction occurs or not. This height assures that such a tree would have its crown closely coupled to prevailing atmospheric conditions and protrudes above deep snow where snow occurs.

The definition of an upper boundary of tree occurrence is more delicate, because such a line refers to the boundary of a vegetation type and, as formulated so elegantly by Armand (1992), "any natural boundary is in reality a transition zone, which has its own two boundaries. They are, in turn, also transition zones with their own boundaries, and so on endless. So localisation of a natural border is in principle inexact and therefore determined by convention."

Tree height decreases continuously with elevation at allocations. The uppermost individuals are shrub-like. For practical reasons, the position of the uppermost trees (~3 m) was chosen as the upper end of the sampling transects. This elevation is referred to by the term outpost treeline because these trees were often, but not always, isolated "outposts." This simply turned out to be the most useful and practical convention for this analysis. All calculations and analyses refer to this elevation as the (theoretical) point of minimum growth. Since these outposts hardly ever exactly met the "3-m-tree" criterion, the actual zero-point of transects was obtained by linear interpolation of tree heights. At all transects, this point differed very little (less than 20 m of elevation) from the position of the uppermost >3-m outpost tree.

The treeline marks a line connecting the highest patches of forest within a given slope or series of slopes of similar exposure. (…) At the sampling sites the treeline is roughly 50 m below the outpost treeline. The authors refrain from using "timberline" in this study, which would refer 10 the upper limit of closed forest.

Tree age is always the age of the pith of the tree al the coring height (50 to 80 cm above the ground). If the pith was missed only slightly (as easily happens with cores in asymmetric or large trees) the number of missing tree-rings was calculated from the diameter of the innermost intact tree-ring and the average width of the 5 following rings. If the pith was missed by more than 4 cm or if the tree was hollow, the absolute tree pith age remained unknown. The sampled trees may well be 30 yr (or more) older than the pith hit by the core because it may take a tree 30 or more years to attain a height of 80 cm at such high elevations. So the true absolute age counted from the year of germination can not be obtained without felling the tree, but this problem should not affect the analysis in a systematic way.


(5) - Syntèses et préconisations


Références citées :

Briffa K. R., Bartholin, T. S., Eckstein D., Jones P. O., Karlén, W., Schweingruber F. H., and Zetterberg P., 1990: A 1.400year tree-ring record of summer temperatures in Fennoscandia, Nature. 346: 434-439.


Briffa K. R., Schweingruber F. H.. Jones P. D., Osborn T. J., Harris I. C., Shiyatov S. G., Vaganov E. A., and Grudd H., 1998: Trees tell of past climates: but are they speaking less clearly today? Philosophical Transactions Royal Society London Series [Biol], 353: 65-13.


Gäggeler H. W.. Stauffer B., Döscher A., and Blunier T., 1997: Klimageschichte im Alpenraum aus Analysen von Eisbohrkernen. Bern: VDF Hochschulverlag AG. 97 pp.


Griggs, R. F., 1946: The timberlines of Northern America and their Interpretation. Ecology, 27: 275-289.


Innes, J. L., 1991: High-altitude and high-latitude tree growth in relation to past, present and future global climate change. The Holocene, 1: 168-173.


Körncr. Ch., 1998: A re-assessment of high elevation treeline positions and their explanation. Oecologia, 115: 445-459.


Nicolussi, K., Bortenschlager, S., and Körner. Ch., 1995: Increase in tree-ring width in subalpine Pinus cembra from the central Alps that may be CO2-related. Trees, 9: 181-189.


Petersan. D.. 1998: Climate, limiting factors and environmental change in high-altitude forests of Western North America. In Beniston. M. and Innes, J. L., (eds.), The Impacts of Climate Variabilty on Forests. New York: Springer, 191-208.


Slatyer and Noble, 1992 (…)