H. Turner and r,J .. Tranquillini, eds.
Proc. 3rd IUFR0 \·lorkshop p 1.07-00, 1984. Eidg. Anst. forstl. Versuchswes., Ber. 270 (1985), 103-112.
ECOPHYSIOLOGICAL CHARACTERISTICS OF SEEDLING AND SAPLING SUBALPINE LARCH, LARIX LYALLII, IN THE WINTER ENVIRONMENT
J. H. Richards Utah State University
Logan, Utah, USA
ABSTRACT
The degree of winter desiccation and the limits of tolerance of seedlinRs and saplinRS of the deciduous conifer, Larix lyallii Parl., were assessed in a timber- line environment in the Rocky Mountains of Canada. Seedlings of subalpine larch carry a large component of wintergreen needles for 20-25 yr on unshaded sites.
These wintergreen needles were deter- mined, by experimental snowpack manipula- tion, to dehydrate rapidly and die when not protected by the snowpack in winter.
Protection from the snowpack appears to be necessary for establishment of this species on unshaded sites. Short shoot buds, which produced deciduous needles in the following growing season, were not damaged by twig xylem pressure potentials of <-6,5 MPa. Minimum levels of this parameter determined under natural conditions on seedling twigs, exposed above the snowpack, were -3.6 MPa and in the same year only -2,5 to -3.2 MPa in sapling twigs. The water potentials of buds dissected from these twigs were significantly different from the measured xylem pressure potentials from October through January indicating a very large resistance to water flux between the bud tissues and the xylem. The water poten- tial of the buds reached a minimum in midwinter, while measurements of xylem pressure potential were lowest in late winter and early spring. The deciduous habit appears to confer not only a great deal of desiccation resistance but also a significant advantage in desiccation tolerance. This advantage may be important for the numerous other decid- uous timberline species.
INTRODUCTION
A wide variety of tree species occurs at the northern and altitudinal limits of tree growth. They display a range of growth habits from needle-leaved evergreen to broad-leaved deciduous. At timberlines worldwide the importance of deciduous trees has often been underestimated, but is of particular significance for this study. The most important deciduous timberline trees are in.eluded in two genera: Larix ana Retula. Species of Alnus, Populus, Salix, Chosenia, Sorbus, Fagus and Nothofagus are also important deciduous timberline trees, but their distribution or dominance is somewhat restricted. The distribution of deciduous timberline
trees will be reviewed below (refer to Figure 1).
DISTRIBUTION OF DECIDUOUS TIMBERLINE TREES
Larix sibirica 1 and L. gmelini (=L.
dahurica ) dominate the northern timber- 1 in e of Eurasia. Larix gmelini is undoubtedly the most widespread and important deciduous timberline species.
It reaches farther north (72,5°N) than any other erect tree. In addition to the area where these two Larix share dominance with evergreen conifers, they form a pure larch forest which covers approximately 4.5 million km2 in N.
Eurasia. In northeastern Eurasia L.
gmelini extends more than 1000 km beyond the limit of Picea sibirica , and is joined only by the dwarf conifers Pinus pumil a and Juniper us comm uni s ( Osten feld and Larsen 1930; Tseplyaev 1965; Hustich 1966; Arno 1970). -
West of 120°E the northward extension of L. qmelini is less impressive, but even here and throughout the range of L. sibirica the larches form the northernmost forests. These two species are also important at alpine timberlines in the Ural, Verkhoyansk, Altai, and S. Khamar Dab an Mountains, and in the mountains of the northern and northeastern provinces of China (Suslov 1961; Wang 1961; Epova 1965;
Gorachakovskii 1965; Kuvaev 1965;
Tseplyaev 1965; Stanyukovich 1973).
At the northern timberline of North America L. laricina is often present, but rarely dominates or grows farther north than the evergreen spruces: Picea mariana and P. glauca ( Marr 1948; Ritchie 1959, 1960; Drew ana Shanks 1965; Larsen 1965, 1974; Hustich 1966; Rowe 1972;
El 1 i o t 197 9).
Of the seven remaining Larix species, two ( L. mastersiana and L.
occidentalis) cannot be considered timberline trees. Larix kaempferi sometimes forms timberline in Jap~n ( F r a n k 1 i n e t a 1 • 1 9 7 9 ) • La r ix griffithiana and L. potanini are important timberline trees of the northern provinces and southeastern
1Larix nomenclature follows Ostenfeld and Larsen (1930).
-EVERGREEN Sl!:~ DECIDUOUS -EVERGREEN - - - DECIDUOUS
O 4000Km
Scale on Equot?J"
Mercator Projection
Figure 1. Generalized distribution pattern of evergreen and deciduous ~rees at alpine and arctic timberlines throughout the world. Mapped locations of mountain ranges are not exact and only relative importance of deciduous and evergreens is indicated. Compiled from data in references cited in the text.
plateau regions of China, and in the bordering Himalayan valleys. Both species are restricted to high elevations and often form a distinct forest zone above the forests of other montane-boreal conifers (Ostenfeld .and Larsen 1930; Wang 1961). Larix decidua has a relatively wide altitudinal range, but is an important timberline tree in the Central and Western Alps (Ellenberg 1963), Larix lyallii is restricted to the upper subalpine forest and timberline ecotone in the N. Cascades and Central Rocky Mountains (Arno and Habeck 1972). It is readily apparent from this survey that the deciduous genus Larix dominates a wide range of timberlines in the Northern Hemisphere.
Species of Betula are as important as larches at alpine timberlines, but have restricted distribution at the northern limits of tree growth. When they grow at timberline they nearly always form a zone of birch woods above or beyond the limits of evergreen trees.
At northern timberlines B. pubescens (Greenland, Iceland, Scotland, Fennos-- candia) and B. ermani (Kamchatka) form homologous vegetation types in areas with extremely oceanic climates (Walter 1968, Hamet-Ahti and Ahti 1969), One or the other of these species or their subspecies is important at timberlines in Scandinavia, the N. Urals, S. Sakhalin and Japan (Hustich 1966; Ahti et al.
1968; Troll 1973; Kullman 1979).
Betula utilis woodland forms the alpine timberline at very high elevations (3900-4200 m) in the northwest Himalaya and Karakoram ranges, and on Tirich Mir in the Hindu Kush (Schweinfurth 1957;
Walter 1968; Baig 1972; Troll 1973). It also occurs at timberline in rain shadow
areas as far east as central Nepal (Patten 1984). Further west, in the Caucasus, 13. verrucosa, B. lit,vinowii and several other species sometimes forrr. the timberline (Baig 1972; Troll 1973), Betula verrucosa is also locally important in the Cantabrian Mountains of northern Spain (Walter 1968; Troll 1973), There are many small areas where birches are found at timberlines in the mountains of Eurasia (Hustich 1966; Baig 1972). It is again apparent that deciduous species, in this case of Betula, are important timberline trees.
Fagus sylvatica may form the upper timberline in oceanic areas of Europe (W.
Pyrenees, Cantabria, S. Alps, etc.) (Baig 1972; Troll 1973). In the Scandinavian mountains, the Carpathians and Caucasus of Eurasia, and the Presidential Range of northeastern N. America, Sorbus species often grow in the timberline ecotone.
Populus, Sal ix, Chosenia and Alnus spp.
dr'e locally important in alluvial habi- tats all along the northern timberlines, and in some mountain areas (Hustich 1966;
Baig 1972).
Alnus jorulensis is presently an important timberline species on the eastern slopes of the central Andes, but this is a much disturbed timberline area and evergreen Polylepis spp. may be the
•true• timberline tree (Ellenberg 1958a, b; Troll 1973; Wardle 1974). In the southern Andes two deciduous species, Nothofagus pumilo and N. antarctica, form the timberline in rain shadow areas from 40° to 55°s (Hueck and Seibert 1972).
This distribution is analogous to the distribution of L. lyallii in the N.
Cascades, where it occurs in rain shadow areas.
A feature common to many deciduous tree-dominated timberlines is the zone of woodland above the evergreen subalpine forest. Species of Betula, Larix and Nothofaqus show this pattern distinctly at alpine timberlines (for examples see Wang 1961; Hamet-Ahti and Ahti 1969; Arno and Habeck 1972; Troll 1973; Quintanilla 1977). The pattern is also apparent with L. gmelini and species of Betula north of the boreal dark conifer forests (Tseplayaev 1965; Hustich 1966; Hamet- Ahti and Ahti 1969).
This brief review clearly shows that deciduous species occur in nearly as many timberline areas of the world as evergreens. In areas where they do occur they often occur above, or north of, evergreen species suggesting that in these environments there is some physiological advantage conferred by the deciduous habit. Deciduous trees, because of their woodiness and less exposed area, would be expected to resist winter desiccation more than even the most resistant evergreen trees. Because winter desiccation is considered the most important determinant of the alpine timberline (Wardle 1971; Tranquillini 1979) an advantage in this regard might allow deciduous species to dominate in timberline areas with particularly desiccating winter conditions.
OBJECTIVES
In this study the degree of desiccation and limits of desiccation tolerance were determined for seedlings and saplings of a representative deciduous timberline soecies, subalpine larch ( Larix l y a l l i i Parl.). This species is found in timberline habitats throughout its range in western North America (Arno and Habeck 1972).
Generally, subalpine larch forms a band of open woodland at the upper edge of the spruce-fir subalpine forest and often grows as upright trees high above the forest limit where sympatric conifers occur only as krummholz. In addition to being representative of deciduous timberline species, L. lyallii has a unique and intriguing aspect of its biology that is particularly useful in a comparison of the advantages of the deciduous and evergreen habits in winter conditions. This species maintains green, functional needles year around when young, and becomes totally deciduous only after the lowermost branches have died (Arno and Habeck 1972). These overwintering needles are wintergreen, being functional for two growing seasons (Richards 1981). Thus, the degree of desiccation resistance and limits of tolerance of wintergreen needles of small subalpine larch trees were determined and compared to the desiccation and tolerance limits of deciduous needled twigs and the buds on those twigs.
METHODS
This research was conducted at a timberline site in Middle Creek Cirque (50° 57' N, 115° 12' W, 2250 m a.s.1) of the Marmot Creek Basin Experimental
Watershed, Kananaskis Valley, Alberta, Canada. Small trees ("seedings") of L.
l y a l l i i , which averaged 0.2-0.7 m tall and 10-15 years old, were sampled on a 20° slope of SSE aspect. The mixed-age stand of subalpine larch on this site also had interspersed individuals of Abies lasiocarpa (Hook) Nutt., Picea engelmannii Parry and an occasional Pinus contorta var. latifolia Engelm. The site is exposed to the prevailing winds and is blown free of snow in patches. Some small trees were sampled on a north- facing site in an opening in a mature L.
lyallii woodland (Richards 1981). This site is well protected from wind and has very deep snow accumulation which prevented access to small trees on it during the wintertime. Saplings of L.
lyallii were sampled on both sites. More details on these sites are given by Richards (1981). Wintergreen needles are found on the lower branches of trees as old as 50 yr and 3-m tall, but are probably of l i t t l e photosynthetic importance because of their small total area and shaded position. On small trees, however, wintergreen needles appear to be much more important. To document this, 26-30 small trees on each site were cut off at their bases and sectioned for aging. These were collected in September 1977 after the deciduous needles had begun to yellow.
This made possible relatively accurate separation of the wintergreen and deciduous needles so that their contribution to the total needle biomass could be determined. The dry weights of the biomass components were determined after oven drying at 45° C for 48 hrs.
Total height was also measured.
During the 1977-78 winter, wintergreen-needled and needleless shoots from small trees were sampled monthly.
Comparable needleless shoots from saplinqs were also sampled each month.
Samples were collected between 1100-1300 hr, immediately sealed in plastic bags and kept frozen during transport to the laboratory. Xylem pressure potential
~ 1 m) was determined with a pressure cfJmber. Preliminary determinations of
~'xylE;.m made in the field and after the 1U-1o hour transport time were not significantly different. Thus, it was concluded that in winter condition twigs could be transported to the laborat,ory for determination of water relations parameters. Total water potential ('¥), osmotic plus matric potential ('¥ +-) and water content (WC) were determ"irfed on wintergreen needles and on buds dissected out of the short shoots with fine forceps. Measurements were made with chamber psychrometers constructed after Mayo (1974). The psychrometers were indi- vidually calibrated using standardized KCl solutions. Turgor ('¥
1,) was calcul- ated as the difference ·between '±' and '¥ rr +-· Water content was calculated as percent of oven-dry weight following drying for 48 hrs at 45°c.
To determine the limits of small larch trees desiccation both field and were conducted. Thirty-six
of tolerance to winter lab studies small larch
trees, 15-20 yrs-old, were carefully excavated and potted in 23 cm diameter plastic pots during the summer of 1976.
Field soil was used and the soil around the base of the tree was disturbed as l i t t l e as possible. These potted trees were maintained in the field until deciduous needle abscission occurred.
Following this, 24 of these small trees were subjected to increasingly winterlike conditions in a controlled environment chamber (EGG Inc.). Conditions in the chamber were changed at two-week intervals and included reductions in the duration of the light period and daytime and nighttime temperatures. During the last 5 weeks of the experiment the chamber was dark for 24 hours and maintained at a constant -3°C with 75%
relative humidity. These conditions simulated the temperature and light conditions at the bottom of a deep snowpack while still allowing the trees to des i cc ate. 4' x v.l em was monitored weekly as the trees c1r1ed. As controls, the remaining 12 trees were kept outside in Edmonton, Alberta during this period and were often, although not continuously covered with snow. Their 'l'x le was also monitored. At the end of t~e ~rying cycle both sets of trees were moved into a greenhouse with springlike conditions (12hr/12hr, light/dark; 16 °/8°C) and recovery was observed. Trees were considered recovered i f 'l'xvlem was similar to that of controls aho a full set of needles, which lasted one growing season, was produced.
To avoid the artificial conditions of potted trees and growth chambers, a field experiment kept ten small L.
l y a l l i i trees on the south-facing site exposed during the 1977-78 winter period.
Small snow fences (1 m2) were installed perpendicular to the prevailing wind direction and in locations where there was a larch seedling at each lower corner. Wind deflected around these barriers prevented snow accumulation over the experimental trees. Water relations parameters and visible damage (reddening of wintergreen needles and needle cast) were monitored monthly. Recovery during the following growing season was also monitored. These data were contrasted to similar data obtained on snow covered trees. In addition, needleless twigs of both small trees and saplings were collected from both above and within the snowpack to determine the effect of protection by snow on water relations parameters.
RESULTS
Height growth of small L. l y a l l i i
trees averaged approximately 1.5 cm•yr-1 during the first 20-25 years of their lives (Figure 2). There were no detectable differences in the height growth on the north- and south-facing study sites. The great varibility in the height versus age relationship may be indicative of aging errors in these small, very slow-growing trees or may indicate large microsite influences.
50
45
40
35
30
E s
l- 25 :i::
(!) w
:i::
20
15
10
5
/
0 0
y = 1.711 + 0.528x + 0.035x' r'= 0.60
0
"
"
0
°I
"
o/110"'
I"
} " 0
0 o.,
0 0
/
Ill o/ o
0 e o/ 0o
/ 0 / 0 e
"
/
/
5 10 15
AGE (YEARS)
..
I ,. I
.. I
0
I
0
/
I
i0
I
" o/
0I I
0"
0
0
20 25
Figure 2. Polynomial regression of height on age for small Larix l y a l l i i
trees from both north-facing ( o) and south-facing
<•)
sites. Data were combined for regression analysis since no differences between sites were found.After attaining 20-25 years subalpine larch individuals begin to grow more rapidly, >10cm•yr-1 on the north-facing site and >16cm·yr-1 on the south-facing site (Richards 1981). Despite no dif- ference in early height growth on the north- and south-facing sites there were large differences in total and needle biomass accumulation (Figure 3a and b).
The relative amount of wintergreen needles decreased more rapidly with age on the north-facing site than on the south-facing site. On the latter site even the 20-24 year old •seedlings' have a mean of 28% of needle biomass in wintergreen needles (Figure 3c). As a function of total tree biomass, however, the relative amount of wintergreen needles declines rapidly on both sites until total biomass is >10 g (Figure 4).
Small trees larger than this have stable relative amounts of wintergreen needles, 28% on the south-facing site but only 7%
on the north-facinis site. This
i
c:i 0 2U) U)
<(
::!!:
0 ffi Cl z
60
45
5
30a:: (.!) LIJ ~
0)
<C 15 .J ~
12
a.
s---
... - N
.... -
i
OL-...::::-~-;_-o--==Q::::::::::_ _ _ _ _ _+---
c:i 8 0 2 g:i 6
<(
::!!:
Q
0) LIJ .J Cl LIJ LIJ
4
z 2
I-' 3:
~100 LIJ Cl
~ 75 .J
1::
50~
u.. 25 0
b.
sc.
N N S N S N S
~ o._----"n~•---:-'-""-"-'--"'-'-~=-.i.:..,;"'-'-.,-L""'-""-"'-'-'-.,.-i-'-"'--'-'-!.J.-
l
AGE :I CLASS I (YEARS)
15-19
Figure 3, Total (a) and needle (b) biomass accumulation of Larix lyallii trees on north-facing (N) and south- facing (S) sites, and relative amounts of wintergreen and deciduous needles of those same trees (c). Means + SE are shown, and n is indicated at the bottom of c for each age class and site. Data in c are from Richards and Bliss (1985).
difference is significant at p<0.01 (t = 5.8, df = 15). It is apparent that at least on the south-facing site L. lyallii seedlings maintain a large commitment to wintergreen needles for 20-25 years. Any damage to these needles by winter desiccation would represent a substantial loss of carbon and photosynthetic area to these small, young trees.
When small L. l y a l l i i trees in pots were dried to low '¥.xylem ( <-6.5 MPa) in a growth chamber, all wintergreen needles were cast. After watering and warming, the trees recovered at the same pace as that of the control trees, and flushing and needle elongation rates were similar.
1.0
. 2
\
\
\ \
' '
• •
... "'(), __ Jf. __ - 0 _
s
0 '-o-~-,,,.o-~-2L0-~-3.L0_~_4_j_O_~_J_50-~---160- TOTAL ABOVEGROUND BIOMASS (g O.D.W.)
Figure 4. Relationship of relative amount of wintergreen needles to total needle biomass of small Larix l y a l l i i trees on north-facing (o) and south- facing (e) sites. Means within 5 g size classes are shown. Redrawn after Richards and Bliss ( 1985).
After one season's growth in a greenhouse, 20 (85%) of the stressed trees had produced full sets of needles, showed high '¥;xylem and appeared healthy.
Eleven (92%) of the control trees survived. Survival was independent (p>0.05) of the drying treatment, even after the damage to roots caused by potting. This experiment showed that small subalpine larch trees are tolerant of '¥ lem as low as -6.5 MPa. This is much !6wer than any 'l1 x lem measured in naturally growing trees (lowest measured value= -3.6 MPa), whether protected or exposed. The wintergreen needles, however, do not tolerate such drying.
Results of the snow fence experiment are given in Figure 5. Osmotic plus matric potential of protected (control) wintergreen needles reached a minimum of -2.9 MPa in December, while that of the exposed needles fell to -6.2 MPa in January (Figure 5a). The wintergreen needles appeared burned, but were not cast until February. Only those needles (approximately 10%) closest to the main stem and protected by the numerous short branches survived. Water content declined significantly and turgor fell to zero as the needles suffered damage and died (Figure 5a and b). The snow protected needles remained hydrated, maintained turgor and were healthy through the following ( 1978) growing season.
Although most wintergreen needles on the exposed trees were killed, the short shoot buds on those same branches were undamaged. After rehydration in spring 1978, as shown by the recovery of 'i'xylem ( Figure 5 c) , they produced a 1· u 11 complement of new needles. The 'l1 lem of wintergreen shoots on control l~ees rose sharply to -0.4 MPa in November, after the development of a protective snowpack. Water uptake from the soil
1.5
z IJJ IJJ 0::
(!) a:: 0
IJJ I-z
i
LI. 0
(/)
-1.5
I-IJJ
z 0
a.. :I:
0 0 0 -3.0 z <(
!z
~ 100IJJ .J I-0
z IJJ 0 IJJ
oz a::z
IJJ IJJ
rr-:::L::c:~.:::::::::.--~---r--.---.---~--~---,--,1.5
<f'p
1./J
1./hr+-r
-=-
SNOWPACK DEVELOPED
=-,-.
--- --- /..-4--... ..._
--- .,.,.
... --0 PROTECTED ¥'pEXPOSED cµp
_...Q--_
~¥.-
--<:, PROTECTED cµ)?- _ A
PROTECTED cµ1r+r/ ---<:>----
/ / /
100
EXPOSED
!.t
IJJ~
ffi o .__ _______________________________
(90% NEEDLE MORTALITY)..Jo
I
ffi o~---,0
1-z
i
LI.
o-
.J
/i.
-1.0<( ::!:
i=;
z l -wo bo
a.. :c -2.0
IJJ (f)
a::o
::, IJJ (/) .J
Wo ~ ~ -3.0 o..z
::!:z IJJ IJJ .J IJJ
>-a::
/ / / /
SNOWPACK DEVELOPED
-~t
PROTECTEDEXPOSED
(100% BRANCH AND BUD SURVIVAL)
-1.0
-2.0
X (!) -4.0 ' - ' - - - ' - - - ' - - - ' - - - l - - - - L - _ _ : ; : : , _ _ i _ _ _ _ ..L,_ _ _ _ i__ _ _ ___t_J_4.0
S O N
DI
J F M A M1977 1978
MONTH
Figure 5. Water relations parameters of small Larix lyallii trees which were protected by the snowpack (controls) or experimentally maintained in an exposed condition: a.) Total ('¥), osmotic plus matric ('¥ + ) and turgor ('l'g) potentials of evergreen needles (mean + SE, n = 4); bJ 'water content f wintergreen needles (mean .:1: SE, n = 4); c.T Xylem pressure potential ('¥x le ) of wintergreen-needled shoots (mean + SE, n = 7). Redrawn after Richards aia ~liss
(1985). -
probably accounts for this rise as soil temperatures at 50cm and below were greater than
o°
C (Richards and Bliss 1985). Protected shoot' s 'l'x lem then remained above -1.0 MPa throulnout the winter.These data support the conclusions reached from the growth chamber experiment. Wintergreen needles of young larch are not tolerant of exposure to winter conditions, and suffer severe dehydration and eventual mortality when not protected by snow. This damage apparently can occur after only 1-2 month's exposure. The buds and branches, however, do survive undamaged after prolonged exposure and dehydration to low ':!'
1 (-4.0 to -6.5 MPa).
xy em
The seasonal pattern of ':!'x lem of needleless twigs collected abdve the snowpack from small trees on the south- facing site and from saplings on both the north- and south-facing sites was similar, with the minimum 'l'x_ylem being observed in March (Figure 6). Twigs from small larch trees on wthe south-facing slope had much lower 'xylem than twigs from saplings on the same site and saplings had lower potentials on the north-facing site throughout winter. On small trees twigs protected by the snowpack had much higher ':!'xylem than exposed twigs. This contrasts with results from saplings, where there were no differences above and within the snowpack (Richards and Bliss 1985).
~ 0 , - - - ,
::;
~ >-
SNOWPACK DEVELOPED
x ' " 4 . 0 4 - - - S - ~ - O - - N - , - - D - + - - ~ - ~ - M - ~ - A ~ - M - - - . - ' 1977 1978
MONTH
Figure 6. Monthly, mid-day (1100-1300) xylem pressure potentials of twigs of saplings and small trees of Larix lyallii. Sapling twigs were collected on both the north-facing (N) and south- facing (S) sites and were collected above the snowpack. Twigs from small trees were collected either above or within the snowpack on the south-facing site. Mean + SE, n = 7. Data from Richards and Bliss (1985).
In s t r i k i n g c o n s t r a s t to measurements made during the summer (Richards 1981), ':!' of wintergreen needles or dissected buds was not equivalent to the 'l'gylem of the branches which bore them tF1gure 5a and c, Figure 7). These large differences between ':!' and ':!' xvlem existed over a wide range of potentlals
when the twigs and needles or buds were in winter condition. These data imply that a very large resistance to water movement forms between the xylem and the wintergreen needle or bud. The resistance appeared to form rapidly in early October and remained quite high until February or March, by which time the two potentials had returned to equilibrium (Figure 7),
A s 0 N
1977 1978
Figure 7. Xylem pressure potentials (':!'xylem) of twigs collected from saplings on Ene south-facing site and water potentials (':!') of buds dissected from those same twigs. The large resistance between xylem and bud is inferred from the large difference in potential which develops in October and lasts through January. Mean+ SE; n = 4 (':!'), n = 7 (':!'l\.Y.lem). Data from Richards and Bliss 19 ts:i •
DISCUSSION
Height growth of small Larix lyallii trees averages only 1.5 cm•yr-1 until the trees are 20-25 yr old. These growth rates are much lower than those of seedlings or young cuttings of Pinus cembra, L. decidua and Picea abies at timberlines in Europe (Oswald 1963;
Tranquillini and Unterholzner 1968;
Oberarzbacher 1977). Less intensive sampling and aging of Picea engelmannii and Abies lasiocarpa on the study sites showed that young trees of these two species also have low height growth rates which are similar to those reported here for L. lyallii. Subalpine larch trees older than 25 yr begin to grow very rapidly, at rates comparable with the maximum height growth rates of Pinus cembra, L. decidua and Picea abies reported by the authors cited above.
Young subalpine larch trees showed no difference in height growth on the two sites but large differences in total and needle biomass accumulation. On the north-facing site, where the understory is a rich forb meadow (Richards 1981), small subalpine larch trees have much greater allocation to stem tissue and are taller and thinner than are trees on the
south-facing site. This difference in allocation remains even for 30- to 40-yr old saplings, where the proportion of aboveground biomass in the trunk is 37%
on the north-facing site and only 20% on the south-facing site. These differences in allocation on the two sites are probably the result of the greater degree of shading by the surrounding mature larch woodland on the north-facing site.
Wintergreen needles constitute a large portion of the photosynthetic surface area of small trees on the south- facing s i t e , but are relatively unimportant for all but the very smallest (youngest) of trees on the north-facing site (Figure 4). Wintergreen needle branches have a much higher photosynthetic light compensation point than branches with deciduous needles (Richards 1981). Thus, again, this difference between the north- and south- facing sites may be due to different radiation environments. The importance of a large amount of wintergreen needles on the south-facing site may be related to their better drought avoidance characteristics (Richards 1981). The wintergreen needles provide a major advantage to these small trees by allowing a longer period of photosynthetic carbon gain. Small trees do not flush until about a week after snow release, which can be as late as mid-July in Marmot Creek Basin. Complete expansion of deciduous needles requires two to four weeks and thus if these small trees were totally dependent on them their photosynthetic period would be quite short. Saplings and larger trees flush deciduous needles on branches above the snowpack long before i t melts (Richards 1981).
The extremely slow growth of L.
l y a l l i i seedlings assures that the young trees, by remaining short, are rarely without the protection of the snowpack in winter. On all but the most exposed sites small topographic variations can provide adequate snowpack to protect these very short young trees.
Nevertheless, when these trees begin to grow they can grow rapidly through the abrasion zone, which can be so destructive to the leading shoots of the timberline trees ( Marchand and Chabot 1978; Tranquillini 1979). Other causes of this pattern of slow growth early in the life of the trees followed by rapid growth relate to summer conditions, such as the need to develop an extensive root system before rapid growth can begin and a large complement of deciduous needles be supported (Richards 1981).
While the wintergreen needles appear quite important to small L. l y a l l i i trees, the snowfence experiment demonstrates clearly that the wintergreen needles of these trees are not tolerant of exposure to winter conditions. It suggests that maximum leaf resistances are far less than those of needles of evergreen trees ( Marchand and Chabot 1978; Hadley and Smith 1983). These results also imply that subalpine larch establishment and growth will be better
on unshaded sites where snow cover protection is provided. Such snow cover protection may indeed be required for establishment. Because the wintergreen needles are always born on branches near the ground, this condition is not difficult to satisfy, however.
The level of desiccation which causes mort;ility of wintergreen needles of small L. l y a l l i i trees causes no damage to the short shoot buds which produce a flush of deciduous needles in the following growing season. The buds and branches are tolerant of dehydration to low xylem potentials (-4.0 to -6.5 MPa). These results taken with data from field comparisons of the amount of damage caused by winter desiccation and/or cbrasion to L. lyallii and Abies 'asiocarpa (Richards and Bliss 1985) indicate that the deciduous habit provides a relative advantage in increased desiccation tolerance under winter conditions at timberline. ThP i rnmature, non-vacuolate bud cells of L.
l y a l l i i are tolerant of extremely low water potentials. These potentials may be quite different than the water potential of the xylem of the twigs on which they occur (Figure 7). This difference suggests that a very large resistance to water movement forms between the xylem and the bud during winter hardening. This resistance may be due to an anatomical feature present in Larix, called the crown (Romberger 1963).
The result of i t is that the bud or needle behaves as a system relatively independent of the xylem. Buds are well protected from water loss by bud scales and thus show changes in water potential ( and water content) in relation to their state of cold hardiness or dormancy (Richards and Bliss 1985). Resistance to water flow, although somewhat lower than for buds, also appears to exist between the xylem and wintergreen needles (Figure 5 a and c). Whether this resistance 1s a consequence of the damage to the wintergreen needles or contributes to the rapid drying of those needles remains in question. In any case, without snow cover protection, water loss from wintergreen needles was much more rapid than uptake from the xylem and, thus, water content and water potential fell. Lethal levels of desiccation were rapidly reached.
The differences in minimum 1xylem observed in saplings and small trees from the two study sites probably relate to different amounts of water storage in main stems and differential thawing on the north- and south-facing sites.
Consistently lower 1xylem on the north- facing site may be explained by less thawing of exposed branches and twigs, preventing any movement of water to the branch tips from that stored above the snowpack in the bole (Richards 1981;
Richards and Bliss 1985). In small trees only a small volume of water can be stored in the main stems (Larcher 1963;
Tranquillini 1979). Therefore, even with thawing only a small volume of water can be supplied to the twigs. Continued water loss causes 1xvlem of exposed twigs
to fall to extremely low levels, while snow protected twigs lose l i t t l e water and thus have higher potentials.
Saplings, and larger trees, of Larix l y a l l i i are subjected to relatively low levels of water s~ress, as indicated by 'l\\l'..l.em during winter. This results from bu!Tering of twig ~xylem by bole water storage (Richards and Bliss 1985) and low rates of water loss from leafless twigs. Twigs on small trees of L.
l y a l l i i , when exposed, experience greater stress, but their limits of tolerance are so great (<-6.5 MPa) that little damage occurs to branch tips and short shoot buds. Both the resistance to water loss and the tolerance of extremely low water potentials in buds appear to be related to the deciduous habit. Wintergreen needles, which are important for small L.
l y a l l i i trees on unshaded sites, rapidly dehydrate and die when exposed during winter. Thus, their maximum diffusion resistance appears to be much lower than that of needles on evergreen timberline trees. The results also suggest that some snow cover protection will be required for establishment of L. l y a l l i i
seedlings on unshaded sites. This snow cover need only be 10-20 cm deep, however, because of the very low growth rates of young L. l y a l l i i trees.
Since the great resistance to, and tolerance of, winter desiccation shown by subalpine larch appears to result from the deciduous habit, it is hypothesized that other typical timberline deciduous trees should show advantages similar to those of L. l y a l l i i . Certainly, L.
decidua does not exhibit this advantage,
but the long shoots of that species are produced much later in the growing season than those of L. l y a l l i i . This phenological difference may explain the difference in winter desiccation resistance (Richards 1981). Winter water relations studies of other deciduous timberline species are needed to better understand the importance of the deciduous habit in timberline environments.
ACKNOWLEDGMENTS
This research was conducted at the Botany Department, University of Alberta, Edmonton, Alberta, Canada and was supported by grants from the National Research Council of Canada to L. C. Bliss and from the Boreal Institute for Northern Studies to J. H. Richards and L.
C. Bliss. Logistic support provided by D. Fisera, R. Swanson, and A. Legge through the Northern Forest Research Center and the Kanasaskis Environmental Sciences Center of the University of Calgary are much appreciated. I would also like to thank R. A. Black for help with the winter sampling and M. M.
Caldwell for review of the manuscript.
LITERATURE CITED
Ahti, T., L. Hamet-Ahti, and J. Jalas.
1968. Vegetation zones and their sections in northwestern Europe.
Annal. Bot. Fenn. 5:169-211.
Arno, S. F. 1970. Ecology of alpine larch (Larix lyallii Parl.) in the Pacific Northwest. Dissertation.
University of Montana, Missoula, Montana, USA.
Arno, S. F., and J. R. Habeck. 1972.
Ecology of alpine larch (Larix l y a l l i i Parl.) in the Pacific Northwest. Ecol. Monog. 42:412-450.
Baig, M. N. 1972. Ecology of timberline vegetation in the Rocky Mountains of Alberta. Dissertation. University of Calgary, Calgary, Alberta, Canada.
Drew, J . V. , and R. E. Shanks . 1 9 6 5 . Landscape relationships of soils and vegetation in the forest-tundra ecotone, upper Firth River Valley, Alaska-Canada. Ecol. Monog. 35:285- 306.
Ellenberg, H. 1958 a and b. Wald oder Steppe? Die n a t u r l i c h e
Pflanzendecke der Anden Perus, I.
and II. Wiss. Techn. 645-648, and 679-681.
Ellenberg, H. 1963. Vegetation Mitteleuropas mit den Alpen in kausaler dynamischer, und historischer Sicht. In: Einfuhrung in die Phytologie IV: 2; ed. by H.
Walter, Stuttgart, Federal Republic of Germany.
E l l i o t t , D. 1979. The current regenerative capacity of the northern Canadian trees, Keewatin, N.W.T., Canada: some preliminary observations. Arct. Alp. Res.
11 :243-251,
Epova, N. A. 1965. Tentative subdivision of phytogeographical zones of the Khamar-Daban Mountains (south- central Siberia). In: Studies of the flora and vegetation of high mountain areas; ed. by V. N.
Sukachev, Academy of Sciences of the USSR All-Union Botanical Society, Volume 5. Translated by Israel Program for Scientific Translations, Jerusalem, Israel.
Franklin, J. F., T. Maeda, Y. Ohsumi, M.
Matsui, H. Yagi, and G. M. Hawk.
1979. Subalpine coniferous forests of central Honshu, Japan. Ecol.
Monog. 49:311-334.
Gorachakovskii, P. L. 1965. An essay on the phytogeographical subdivision of the Uralian Highlands. In: Studies on the flora and vegetation of high mountain areas; ed. by V. N.
Sukachev, Academy of Sciences of the USSR All-Union Botanical Society, translated by Israel Program for Scientific Translations, Jerusalem, Israel.
Hadley, J. L., and W. K. Smith. 1983.
Influence of wind exposure on needle desiccation and mortality for timberline conifers in Wyoming, U.S.A. Arct. Alp. Res. 15:127-135.
Hamet-Ahti, L., and T. Ahti. 1969. The homologies of the Fennoscandian mountain and coastal birch forests in Eurasia and North America.
Vegetatio 19:208-219.
Hueck, K., and P. Seibert. 1972.
Vegetationskarte von Sudamerika.
Gustav Fischer, Stuttgart, Federal Republic of Germany.
Hust i c h , I. 19 6 6. On the forest - tundra and the northern tree-lines. Annal.
Univ. Turkuensis, Series A. II, 36:7-47.
Kullman, L. 1979. Change and stability in the altitude of the birch tree- limit in the southern Swedish Scandes 1915-1975, Acta Phytogeographica Suecica 65, Uppsala, Sweden.
Kuvaev, V. B. 1965. Distribution pattern of the vegetation in the western Verkhoyansk Mountain area.
In: Studies of the flora and vegetation of high mountain areas;
ed. by
v.
N. Sukachev, Academy of Sciences of the USSR All-Union Botanical Society, translated by Israel Program for Scientific Translations, Jerusalem, Israel.Larcher, W. 1963. Zur spatwinterlichen Erschwerung der Wasserbilanz von Holzpflanzen an der Waldgrenze.
Ber. Naturwiss. Med. Ver.
Innsbruck 53:125-137,
Larsen, J. A. 1965. The vegetation of the Ennadai Lake area, Northwest Territories: studies in subarctic and arctic bioclimatology. Ecol.
Monogr. 35:37-59,
Larsen, J. A. 1974. Ecology of the northern continental forest border.
In: Arctic and Alpine Environments;
ed. by J. D. Ives and R. G. Barry, pp. 341-369, Methuen and Co., Ltd., London, England.
Marchand, P. J., and G. F. Chabot. 1978.
Winter water relations of tree-line plant species on Mt. Washington, New Hampshire. Arct. Alp. Res. 10: 105- 11 6.
Marr, J. 1948. Ecology of the forest- tundra ecotone of the east coast of the Hudson Bay. Ecol. Monogr.
18:118-144.
Mayo) J. M. 1 9 7 4 . A therm O CO u D le psychrometer chamber for field ·and laboratory use. Can. J. Plant Sci.
54:597-598,
Oberarzbacher, P. 1977, Beitrage zur physiologischen Analyse des HHhenzuwachses von verschiedenen Fichtenklonen entlang eines HHhenprofils im Wipptal (Tirol) und in Klimakammern. Dissertation.
University of Innsbruck, Innsbruck, Austria.
Ostenfeld, C. H., and C. S. Larsen.
1930. The species of the genus Larix and their geographical distribution. Det Kgl. Danske videnskabernes Selskab, Biologiske Meddelelser IX 2:1-106.
Oswald, H. 1963. Verteilung und Zuwachs der Zirbe (Pinus cembra L.) der subalpinen Stufe an einem zentralapinen Standort. Mitteilungen der Forstlichen Bundesversuchsan- stalt Mariabrunn 60:437-499-
Patten, D. T. 1984. Timberline and forest gradients in the upper Marsyandi River Valley, Nepal.
Abstract. Bull. Ecol. Soc. Am.
65:63,
Quintanilla, V. 1977. A contribution to the phytogeographical study of temperate Chile. In: Ecosystem research in South America; ed. by P.
MUller, p. 31-41. Biogeographica Vol. VIII, Dr. W. Junk, The Hague, Netherlands.
Richards, J. H. 1981. Ecophysiology of a deciduous timberline tree, Larix lyallii Parl. Dissertation. Univ.
of Alberta, Edmonton, Alberta, Canada.
Richards, J. H., and L. C. Bliss. 1985.
Winter water relations of a decid- uous timberline conifer, Larix lyallii Parl. Oecologia (in press).
Ritchie, J. C. 1959. The vegetation of northern Manitoba, III. Studies in the subarctic. Arctic Institute of North America Technical Paper No. 3, University of Calgary, Calgary, Alberta.
Ritchie, J. C. 1960. The vegetation of northern Manitoba. VI. The lower Hayes River region. Can. J. Bot.
38:769-788.
Romberger, J. A. 1963. Meristems, growth and development in woody plants. USDA, Forest Service, Technical Bulletin No. 1293, Washington, D.C., USA.
Ro we, J. S. 1 9 7 2. F-::, rest regions of Canada. Canadian Forestry Service Publication No. 1300. Department of the Environment, Ottawa, Canada.
Schweinfurth, U. 1957, Die horizontale und vertikale Verbreitung der Vegetation im Himalaja. Bonner Geogr. Abh., Heft 20., Bonn, Federal Republic of Germany.
Stanyukovich, K. V. 1973. Vegetation of the mountains of the USSR. A botanical-geographical outline.
Izd. Donish, Dushanbe, USSR.
Suslov, S. P. 1961. Physical geography of Asiatic Russia. W. H. Freeman and Co., San Francisco, California.
Tranquillini, W. 1979. Physiological ecology of the alpine timberline.
Ecol. Studies 31:137 pp. Springer New York.
Tranquillini, W., and R. Unterholzner.
1968. Das Wachstum zweijahrigar Larchen einheitlicher Herkunft in verschiedener Seehohe. Centralbl.
Ges. Forstwesen 85:43-59,
Troll, C. 1973, The upper timberlines in different climatic zones. Arct.
Alp. Res. 5:A3-A18.
Tseplyaev, V. P. 1965. The forests of the USSR. Israel Program for Scientific Translation, Jerusalem, Israel.
Walter, H. 1968. Die Vegetation der Erde II. Die gemassigten und arktischen Zonen. Gustav Fischer, Stuttgart, Federal Republic of Germany.
Wang, C. W. 1961. The forests of China.
Maria Moors Cabot Foundation Publication No. 5, Harvard Univer- sity, Cambridge, Massachusetts.
Wardle, P. 1971. An explanation for alpine timberline. New Zealand J.
Bot. 9:371-401.
Wardle, P. 1974. Alpine timberlines.
In: Arctic and Alpine Environments;
ed. by J. D. Ives and R. G. Barry, p. 371-402. Methuen and Co., Ltd., London, England.
