Ecological Research

(2003)

18

, 549–558

Blackwell Science, LtdOxford, UKEREEcological Research0912-38142003 Ecological Society of JapanSeptember 2003185549558Original Article

Effect of climate on tree growthK. Takahashi

et al.

*Author to whom correspondence should beaddressed. Email: koichit@gipac.shinshu-u.ac.jp

Received 26 November 2002. Accepted 28 February 2003.

Effects of climate on the radial growth of tree species in the upper and lower distribution limits of an altitudinal

ecotone on Mount Norikura, central Japan

Koichi

Takahashi

,

1

* Hiroto

Azuma

1

and Koh

Yasue

2

1

Department of Biology, Faculty of Science, Shinshu University, Matsumoto, Nagano 390-8621, Japan and

2

Department of Forest Science, Faculty of Agriculture, Shinshu University, 8304 Minami-Minowa, Nagano 399-4598, Japan

Tree-ring width chronologies were developed for

Abies veitchii

,

Betula ermanii

and

Betula platyphylla

var.

japonica

in their altitudinal ecotone (approximately 1600 m a.s.l.) on Mount Norikura, centralJapan, to determine what climatic conditions affect the growth of tree species in the upper and lowerdistribution limits of an altitudinal ecotone. This altitude was the lower distribution limit for

A. veitchii

and

B. ermanii

in the subalpine zone, and was the upper distribution limit for

B. platyphylla

var.

japonica

in the montane zone on Mount Norikura. Tree-ring widths of the two

Bet-ula

species and

A. veitchii

were positively correlated with the August precipitation of the current andprevious years, respectively. Precipitation in August (the hottest month) was reduced compared withother months during summer. Tree-ring width of

B. platyphylla

var.

japonica

showed no correlationwith temperatures in any month in its upper distribution limit. In contrast, tree-ring widths of

B. ermanii

and

A. veitchii

were negatively correlated with the August temperatures of the current andprevious years, respectively, at the lower distribution limit of these species. Therefore, the two

Betula

species and

A. veitchii

responded to climatic conditions of the current and previous years, respec-tively. The present study also suggests that a water deficit in August reduces growth of these threespecies in this altitudinal ecotone, irrespective of the upper or lower distribution limits, and that ahigh August temperature is more detrimental to the growth of

A. veitchii

and

B. ermanii

in theirlower distribution limits. Thus, the three species with different altitudinal distributions examinedin the present study responded differently to climatic conditions in this altitudinal ecotone onMount Norikura.

Key words:

altitudinal ecotone; central Japan; climatic conditions; dendrochronology; growth; tree-ring width.

INTRODUCTION

Forest vegetation changes along altitudinal gradi-ents, accompanied by changes in air temperature.For example, in central Japan, forest vegetationchanges from evergreen broad-leaved forests at lowaltitudes to deciduous broad-leaved forests at mid-dle altitudes and to evergreen coniferous forests athigh altitudes (Yoshino 1978; Ohsawa 1995).

Vegetation patterns along altitudinal gradientschange markedly in ecotones rather than graduallywith altitude (i.e. an ecotone is a narrow transi-tional zone between adjacent communities) (e.g.Kimura 1963; Harris 1978). Therefore, altitudinalecotones are the most dynamic sites for investigat-ing the altitudinal distributions of plant speciesunder the influence of climatic conditions. How-ever, few studies have investigated ecotone dynam-ics in relation to climate, with the exception of thetimberline dynamics studies (e.g. Wardle 1968;Kullman 1993; Taylor 1995; Camarero & Gutiér-rez 1999). An increased knowledge of theresponses to climate at both the upper and lowerdistribution limits of tree species that form an

550 K. Takahashi

et al.

altitudinal ecotone is of great importance inunderstanding the effects of global warming onthe altitudinal distributions of different plant spe-cies. In the present study, we investigated how cli-matic conditions affect the growth of tree speciesin the upper and lower distribution limits of analtitudinal ecotone.

Dendrochronological analysis is often used toshow what climatic conditions affect tree-ringwidths (i.e. radial growth) and to reconstruct pastclimates (Schweingruber

et al

. 1978; Hughes

et al

.1984; Cleaveland 1986; Kienast

et al

. 1987; Sweda& Takeda 1993; Yasue

et al

. 1996, 1997; Gindl1999). For example, several studies have shownthat high summer temperatures increase tree-ringwidths at high altitudes and latitudes, and muchprecipitation increases tree-ring widths at dry sites(Fritts

et al

. 1965; Abrams & Orwig 1995; Gostev

et al

. 1996; Oberhuber

et al

. 1998; Gervais &MacDonald 2000; Rigling

et al

. 2001). Thus,climatic conditions affecting tree growth aredetectable using dendrochronological analysis.

We examined the altitudinal ecotone betweendeciduous broad-leaved forests and subalpine for-ests in central Japan. This altitudinal ecotone wasformed mainly by

Betula platyphylla

var.

japonica

Hara,

Betula ermanii

Cham. and

Abies veitchii

Lindl.We developed tree-ring width chronologies ofthese three species and examined the relationshipsbetween these species and the climatic conditionsof monthly temperature and precipitation. Thepresent study aimed to determine what climaticconditions affect the radial growth of these threespecies in this altitudinal ecotone.

METHODS

Site description

This study was carried out on the east slope ofMount Norikura (3026 m a.s.l.) in central Japan.On this slope,

B. platyphylla

var.

japonica

was dis-tributed between approximately 1100 and 1600 ma.s.l. in the upper zone of the deciduous broad-leaved forests, whereas

B. ermanii

and four conifers(

A. veitchii

,

Abies mariesii

Mast.,

Picea jezoensis

var.

hondoensis

Rehder and

Tsuga diversifolia

Mast.)dominated between approximately 1600 and2500 m a.s.l. in subalpine forests. The alpinedwarf pine scrub (

Pinus pumila

Regel) was distrib-

uted from approximately 2500 m a.s.l. to near thesummit. The forest floor was densely covered withdwarf bamboo (

Sasa senanensis

Rehder), except atthe dwarf pine scrub sites. Nomenclature followsOhwi (1961).

We selected an altitudinal ecotone (1600 ma.s.l., 36

∞

07

¢

N, 137

∞

37

¢

E) between the deciduousbroad-leaved forests and the subalpine forests, andindividuals of

B. platyphylla

var.

japonica

,

B. ermanii

and

A. veitchii

growing in the ecotonewere examined. This ecotone formed the upperdistribution limit for

B. platyphylla

var.

japonica

and the lower distribution limit for

B. ermanii

and

A. veitchii

.

Betula

species often form pure standsafter large disturbances, such as forest fires; how-ever, there was no evidence of forest fires aroundthe study area.

Mean annual precipitation was approximately1570 mm according to the meteorological datameasured at Nagawato Dam (900 m a.s.l., approx-imately 10 km from the study area). The meanannual temperature at the study area (1600 ma.s.l.) was estimated to be 4.9

∞

C from tempera-tures recorded at Matsumoto Weather Station(610 m a.s.l., approximately 35 km from the studyarea) using a lapse rate of

-

0.6

∞

C for each

+

100 maltitude. Mean monthly temperatures in the cold-est month of January and the hottest month ofAugust were estimated to be

-

7.3 and 17.7

∞

C,respectively (Fig. 1).

Sampling and measurement

For each species, 30 trees were cored at 1.3 mtrunk height. One core from each tree was takenbetween 5 and 24 November 2000 and the d.b.h.of each tree was measured. All cores were dried,mounted and sanded, and tree-ring widths weremeasured to 0.01 mm under a microscope using ameasurement stage (TA Tree-Ring System,Velmex Inc., New York, NY, USA). Tree-ringboundaries for the two

Betula

species were distin-guishable by their small diameter cells (terminalparenchyma).

Chronology development

All cores were cross dated visually by matchingcharacteristic wide and narrow rings that were syn-chronous in trees within a species. Visual cross

Effect of climate on tree growth 551

dating was statistically verified using theCOFECHA program (Holmes 1983, 1994), whichtests each individual series against a master datingseries (mean of all series) from correlation coeffi-cients. Of the 30 cores for each species, severalcores had low correlations with other trees andwere eliminated from further analyses.

Growth of trees is affected not only by climaticfactors, but also by competition between neighbor-ing trees (i.e. suppression by tall trees and releasefrom suppression). To eliminate competitiveeffects, all raw ring-width series were standardizedby fitting smoothing splines (Cook & Peters 1981)with a 50% frequency-response cutoff of 10 years.This procedure was conducted using the ARSTANprogram (Cook 1985; Holmes 1994). After stan-dardizing each individual series, the tree-ringwidth chronology for each species was calculatedby averaging the standardized individual series.We used at least 10 cores to determine the tree-ringwidth chronologies in each year.

Responses to climatic conditions

Simple correlation analyses were used to show theeffects of monthly mean temperature and total pre-cipitation on tree-ring widths of the three species.Before the correlation analysis, we compared mete-orological data from Nagawa (1068 m a.s.l.,

approximately 7 km southeast of the study site),which was the nearest weather station to the studysite, with data from Nagawato Dam (900 m a.s.l.,approximately 10 km east of the study site) andMatsumoto (610 m a.s.l., approximately 35 kmeast of the study site). The recording period for themeteorological data at Nagawa was comparativelyshort (from 1979 to 2000;

n

=

22), whereas therecording periods at Nagawato Dam and Matsu-moto were from 1969 to 2000 (

n

=

32) and from1898 to 2000 (

n

=

103), respectively. At Naga-wato Dam the records were for precipitation only.Monthly temperatures at Matsumoto were highlycorrelated with those at Nagawa (

r

=

0.89–0.98).However, monthly precipitation at Matsumoto didnot correlate well with that at Nagawa (

r

=

0.58–0.95), whereas monthly precipitation was highlycorrelated between Nagawato Dam and Nagawa(

r

=

0.75–0.98). The low correlations for precipi-tation between Matsumoto and the other two sta-tions were probably the result of local variations inrainfall. Thus, we used temperature data fromMatsumoto and precipitation data from NagawatoDam for further analyses because of the shortrecording period at Nagawa. A calibration periodfrom 1969 to 2000 (

n

=

32) corresponded to therecording period at Nagawato Dam.

Correlation analyses between ring widths andclimatic data were conducted from the start of theprevious growing season to the end of the currentgrowing season because the growth of many treespecies is affected not only by the climatic condi-tions of the current growing season, but also by theclimatic conditions of the previous growing season(cf. Fritts 1962; Sano

et al

. 1977; Takahashi

et al

.2001). The approximate growth period at thestudy site (1600 m a.s.l.) was estimated to be fromMay to October (Fig. 1) because the mean monthlytemperatures exceeded 5

∞

C, effective heat for plantgrowth (Kira 1949), during this period. Thus, thecorrelation analysis was performed from May of theprevious year to October of the current year(18 months in total).

RESULTS

Tree-ring width chronology

Of the 30 core samples for each species, 27, 24 and23 cores were successfully cross dated for

Fig. 1.

Monthly mean temperatures (

�) at 1600 ma.s.l. in Mount Norikura and precipitation ( ) at900 m a.s.l. (Nagawato Dam). The temperature at1600 m a.s.l. was estimated from the temperature at610 m a.s.l. (Matsumoto) using the lapse rate of -0.6∞Cfor each +100 m altitude. The growing season of plantsat this study site is defined as May to October withmonthly mean temperatures >5∞C (---).

Pre

cpit

atio

n (

mm

mo

nth

–1)

Air

tem

per

atu

re (

°c)

552 K. Takahashi et al.

B. platyphylla var. japonica, B. ermanii andA. veitchii, respectively, and were used to develop amaster chronology with good correlations betweentrees of each species (Table 1). Although we sam-pled trees that appeared to be older in the studyarea, the ages of the sampled trees were quiteyoung. Tree-ring width chronologies for the threespecies were only for several decades (Table 1,Fig. 2). The mean tree-ring widths of the threespecies differed significantly from one other(Tukey HSD test, P < 0.05). Abies veitchii had thelargest tree-ring width among the three species(Table 1). The mean tree-ring width ofB. platyphylla var. japonica at the upper distribu-tion limit was greater than that of B. ermanii at thelower distribution limit. The mean correlationcoefficient between trees of B. ermanii was lowercompared with the other two species (0.21 com-pared with approximately 0.4, Table 1). The first-order autocorrelations of standardized chronolo-gies, as measures of the influence of the previousyear’s growth on the growth in the current year,were lower than 0.05 in all three species (Table 1).The mean sensitivities and standard deviations, asmeasures of interannual variation in tree-ringwidth, were also low (<0.15) for all three species.These values indicate that individual trees of eachspecies grew synchronously in this altitudinal eco-tone, and that interannual variation in tree-ringwidth was not high.

Each chronology showed several characteristicwide and narrow rings. The tree-ring width indi-

ces of the two Betula species decreased in 1978and 1995 (Fig. 2). The indices of B. ermanii andB. platyphylla var. japonica also decreased in 1973and 1988, respectively, in addition to 1978 and1995. The characteristic declines of ring widthsin A. veitchii (1974, 1979 and 1989) were 1 yearlater than those of the two Betula species, exceptfor 1995 which showed a strong decline in allspecies. Furthermore, A. veitchii showed character-istic wide rings in 1977, 1993 and 1999 (Fig. 2).These wide rings of A. veitchii were also 1 yearlater than the wide rings recorded in the twoBetula species in 1976, 1992 and 1998. In addi-tion, the index of B. ermanii increased in 1980and 1991, whereas the index of B. platyphylla var.japonica increased in 1980 and 1982. The tree-ring width index of A. veitchii was positivelycorrelated with the 1-year lagged index ofB. ermanii (r = 0.38, P < 0.01) and the index ofB. platyphylla var. japonica (r = 0.39, P < 0.01) for1949–2000.

Responses to climatic conditions

Temperatures were not significantly correlatedwith the tree-ring width index of B. platyphyllavar. japonica at its upper distribution limit (Fig. 3).In contrast, the tree-ring width indices ofB. ermanii and A. veitchii at their lower distribu-tion limit were negatively correlated with theAugust temperature of the current year and previ-ous year, respectively (Fig. 3).

Table 1 Basic statistics of tree-ring width chronologies for Abies veitchii (Av), Betula ermanii (Be) and Betula platy-phylla var. japonica (Bp) in central Japan

Av Be Bp

No. samples 23 24 27Range of d.b.h.† (cm) 25–45 21–42 19–40Range of chronology (years) 65 77 52Mean tree-ring width‡ (mm) 3.10 2.51 2.93Standardized chronologyMean sensitivity 0.105 0.128 0.150Standard deviation 0.093 0.113 0.135First-order autocorrelation 0.042 0.023 -0.043Mean correlation between trees§ 0.404 0.211 0.381Signal-to-noise ratio§ 15.6 3.5 11.1

†Diameter at breast height.‡Mean tree-ring widths of the three species were significantly different from each other (Tukey HSD test, P < 0.05).§Calculated for the common intervals.

Effect of climate on tree growth 553

The tree-ring width indices of the two Betulaspecies were positively correlated with precipita-tion in March and August of the current year(Fig. 3). In addition, the tree-ring width index ofB. platyphylla var. japonica was positively correlatedwith precipitation in October of the current year.Abies veitchii responded to precipitation in a differ-ent way to the two Betula species: its tree-ringwidth index was positively correlated with theAugust precipitation of the previous year, whereasthe two Betula species responded precipitation inthe current year (Fig. 3). Therefore, climatic con-ditions of the previous year affected the tree-ringwidths of A. veitchii, whereas climatic conditionsof the current year affected the two Betula species.This difference between the two genera corre-

sponded to the 1-year difference in the character-istic wide and narrow rings (Fig. 2). August wasthe hottest month, but precipitation in this monthwas reduced compared to other months betweenJune and September (Fig. 1).

DISCUSSION

One-year difference in tree-ring width variations between Abies and Betula

We found a 1-year difference in the variations intree-ring width chronologies between A. veitchiiand the two Betula species. Abies veitchii respondedto climatic conditions in the previous year and thetwo Betula species responded to conditions in thecurrent year. Clearly, the responses to climatic con-ditions in radial growth differed between the twogenera. Kikuzawa (1983) categorized deciduousbroad-leaved tree species in northern Japanbetween two extremes according to leaf phenol-ogy: leaves emerging simultaneously during ashort period at the beginning of a growing season(flush type) or continuously one by one over a longperiod (succeeding type). He also suggested thatflush-type species grow using the photosyntheticproduction of the previous year, whereas succeed-ing-type species are strongly dependent on thephotosynthetic production of the current year.Species in the genus Betula, includingB. platyphylla var. japonica and B. ermanii, are cat-egorized as intermediate between flush and suc-ceeding types (Kikuzawa 1983). In Betula species,early leaves, usually one or two, emerge at first, andthen late leaves emerge one by one. Thus, it seemsthat the two Betula species are largely dependenton the photosynthetic production of the currentyear. Unfortunately, we do not have enough infor-mation on the relationship between growth traitsand photosynthetic production in A. veitchii. How-ever, several researchers have reported significantcorrelations between tree-ring width (or shootelongation) and the climatic conditions of the pre-vious year (Fritts 1962; Sano et al. 1977; Eshete &Ståhl 1999; Takahashi et al. 2001; Takahashi2003), which is similar to the results reported forA. veitchii in the present study. For example, shootelongation in the alpine dwarf pine (Pinus pumila)in Japan is positively correlated with summer tem-peratures in the previous year (Sano et al. 1977;

Fig. 2. Standardized tree-ring width chronologies andsample depths of (a) Abies veitchii, (b) Betula ermanii and(c) Betula platyphylla var. japonica in the altitudinal eco-tone (1600 m a.s.l.) on Mount Norikura, central Japan.Each chronology was constructed using at least 10 cores.Arrows indicate the characteristic years in which thegrowth of many trees decreased or increased synchro-nously.

554 K. Takahashi et al.

Fig. 3. Correlation coefficients between the standardized tree-ring width chronologies and the monthly climatedata (precipitation and temperature). (�), Significant correlation (P < 0.05).

Effect of climate on tree growth 555

Takahashi 2003). Kibe and Masuzawa (1992)examined seasonal changes in the amount of car-bohydrates in P. pumila, and suggested that carbo-hydrate gained from photosynthesis in summer isrestored in branches and needles as sugar in winter,and is used for growth in the following year. Theirsuggestion is consistent with the growth traits ofP. pumila (i.e. shoot elongation is affected by cli-matic conditions in the previous year). It is likelythat the same is true for A. veitchii, that is, the pre-vious year’s photosynthetic production is used forthe current year’s growth. Therefore, the 1-yeardifference in tree-ring width chronology betweenthe two genera (Abies and Betula) is probablyrelated to whether the current year’s photosyn-thetic production is used for the current year’sgrowth or is restored for the following year’sgrowth.

Responses to climatic conditions

The tree-ring width indices of the two Betula spe-cies were positively correlated with the amount ofprecipitation in March of the current year. Precip-itation in March took the form of snow because thetemperature was below zero (Fig. 1), but the rea-son that a large amount of snow in March had apositive effect on the growth of the two Betula spe-cies is unknown. The tree-ring width index ofB. platyphylla var. japonica was positively correlatedwith precipitation in October of the current year.The reason for this result is also unknown becauseleaves start to fall in October (the end of the grow-ing season) and, therefore, radial growth in thismonth is likely to be negligible.

Tree-ring width indices of the three species werepositively correlated with precipitation rates in thehottest month, August, which had less precipita-tion compared with the other summer months.Low precipitation coupled with a high tempera-ture causes drought stress in plants, which in turnreduces the photosynthetic production of plants(Havranek & Benecke 1978; Cienciala et al. 1998)because of stomatal closure in response to a limi-tation in the amount of soil water available(Cienciala et al. 1994; Granier & Loustau 1994;Granier & Bréda 1996; Miller et al. 1998). There-fore, precipitation in August is likely to increasethe growth of the three species examined in thepresent study in this altitudinal ecotone, irres-

pective of the upper or lower limits of theirdistributions.

Temperature in the hottest month of Augustwas negatively correlated with tree-ring widthindices in B. ermanii and A. veitchii at their lowerdistribution limit. However, these negative corre-lations do not appear to be independent of the pre-cipitation in August. If high temperatures reducegrowth of B. ermanii and A. veitchii, a negative cor-relation would have also been recorded for Julytemperatures because the temperatures in Julywere as high as the August temperatures (Fig. 1);however, a negative correlation was not detected.High temperature combined with reduced precip-itation in August appears to increase droughtstress in these two species at their lower distribu-tion limit. In contrast, B. platyphylla var. japonicashowed no significant correlation with Augusttemperature at its upper distribution limit, whichis attributable to its adaptation to high-tempera-ture conditions. The optimal temperature for pho-tosynthesis is higher for B. platyphylla var. japonica(10–24∞C) than B. ermanii (8–20∞C) (Koike1987). These different optimal temperaturesbetween the two Betula species correspond to theiraltitudinal distribution (i.e. B. platyphylla var.japonica is distributed more in the lower altitudi-nal zone than B. ermanii). Therefore, high Augusttemperatures have less of an effects on the growthof B. platyphylla var. japonica compared withB. ermanii in this altitudinal ecotone.

Although the upper distribution limit ofB. platyphylla var. japonica was approximately1600 m a.s.l. on Mount Norikura, Takahashi(1962) observed that B. platyphylla var. japonicawas distributed up to approximately 2000 m a.s.l.on Mount Yatsugatake in central Japan. Themechanisms determining the upper distributionlimits of this species are unknown. However, it issuggested that the upper distribution limit ofB. platyphylla var. japonica on Mount Norikuramay not be determined by temperature alone.

In the present study, a positive correlation wasnot found between summer temperatures and tree-ring widths of the three species examined in thisaltitudinal ecotone (1600 m a.s.l.). However, Fuji-wara et al. (1999) found positive relationshipsbetween summer temperatures and tree-ringwidth for Abies mariesii at approximately 2000–2200 m a.s.l. on Mount Norikura, the site of our

556 K. Takahashi et al.

study. It is likely that the discrepancy between theresults of our study and those of Fujiwara et al.(1999) occur because the study site used byFujiwara et al. (1999) was at a higher altitude thanour study site. Precipitation is greater at a higheraltitude in this region, and is associated with adecrease in air temperature (Nagano Meteorologi-cal Observatory 1998), suggesting that treegrowth at Fujiwara et al.’s (1999) study site is lim-ited more by low temperature than by precipita-tion. A similar result was also reported forDahurican larch (Larix cajanderi) on theKamchatka Peninsula in the Russian Far East. Thetree-ring width of larch at a low altitude is posi-tively correlated with summer precipitation,whereas tree-ring width at a high altitude is pos-itively correlated with early summer temperature(Solomina et al. 1999; Takahashi et al. 2001).Thus, the relative importance of summer temper-ature for tree growth may increase with altitude(cf. Wilson & Hopfmueller 2001).

CONCLUSION

The present study showed the effects of climaticconditions on radial growth in three species attheir upper and lower distribution limits in analtitudinal ecotone between deciduous forests andsubalpine forests on Mount Norikura using a den-drochronological technique. Radial growth in thethree species was affected by drought stress in thehottest month of August in this ecotone. A waterdeficit combined with high temperatures inAugust were more detrimental to the growth ofB. ermanii and A. veitchii at their lower distribu-tion limit than B. platyphylla var. japonica at itsupper distribution limit. Therefore, environmen-tal fluctuations are suggested to exert differenteffects on the growth of these three species, withdifferent altitudinal distributions, in this ecotoneon Mount Norikura.

ACKNOWLEDGEMENTS

We are grateful to the Tokyo Electric Power Com-pany for providing the weather records of Naga-wato Dam. This study was partially supported bygrants from the Ministry of Education, Science,

Sports and Culture of Japan (No. 13760128,13780418) and by a grant for EnvironmentalResearch Projects from the Sumitomo Foundation(No. 003280).

REFERENCES

ABRAMS M. D. & ORWIG D. A. (1995) Structure,radial growth dynamics and recent climatic vari-ations of a 320-year-old Pinus rigita rock outcropcommunity. Oecologia 101: 353–360.

CAMARERO J. J. & GUTIÉRREZ E. (1999)Structure and recent recruitment at alpine forest-pasture ecotones in the Spanish central Pyrenees.Ecoscience 6: 451–464.

CIENCIALA E., KUCERA J., RYAN M. G. & LIN-

DROTH A. (1998) Water flux in boreal forest dur-ing two hydrologically contrasting years; speciesspecific regulation of canopy conductance and tran-spiration. Annales des Sciences Forestieres 55: 47–61.

CIENCIALA E., LINDROTH A., CERMAK J.,HALLGREN J-E. & KUCERA J. (1994) The effectsof water availability on transpiration, water poten-tial and growth of Picea abies during a growing sea-son. Journal of Hydrology 155: 57–71.

CLEAVELAND M. K. (1986) Climatic response ofdensitometric properties in semiarid site treerings. Tree-Ring Bulletin 46: 13–29.

COOK E. R. (1985) A time series analysis approachto tree ring standardization. PhD dissertation,University of Arizona, Tucson (unpubl.).

COOK E. R. & PETERS K. (1981) The smoothingspline: a new approach to standardizing forestinterior tree-ring width series for dendroclimaticstudies. Tree-Ring Bulletin 41: 45–53.

ESHETE G. & STÅHL G. (1999) Tree rings as indi-cators of growth periodicity of acacias in the RiftValley of Ethiopia. Forest Ecology and Management116: 107–117.

FRITTS H. C. (1962) The relation of growth ringwidths in American beech and white oak to vari-ations in climate. Tree-Ring Bulletin 25: 2–10.

FRITTS H. C., SMITH D. G., CARDIS J. W. &BUDELSKY C. A. (1965) Tree-ring characteristicsalong a vegetation gradient in northern Arizona.Ecology 46: 393–401.

FUJIWARA T., OKADA N. & YAMASHITA K.(1999) Comparison of growth response of Abiesand Picea species to climate in Mt. Norikura, cen-tral Japan. Journal of Wood Science 45: 92–97.

GERVAIS B. R. & MACDONALD G. M. (2000) A403-year record of July temperatures and treeline

Effect of climate on tree growth 557

dynamics of Pinus sylvestris from the Kola Penin-sula, northwest Russia. Arctic, Antarctic and AlpineResearch 32: 295–302.

GINDL W. (1999) Climate significance of light ringsin timberline spruce, Picea abies, Austrian Alps.Arctic, Antarctic and Alpine Research 31: 242–264.

GOSTEV M., WILES G., D’ARRIGO R., JACOBY

G. & KHOMENTOVSKY P. (1996) Early summertemperatures since 1670 A.D. for Central Kam-chatka reconstructed based on a Siberian larchtree-ring width chronology. Canadian Journal ofForest Research 26: 2048–2052.

GRANIER A. & BRÉDA N. (1996) Modelling can-opy conductance and stand transpiration of an oakforest from sap flow measurements. Annales des Sci-ences Forestieres 53: 537–546.

GRANIER A. & LOUSTAU D. (1994) Measuring andmodelling the transpiration of a maritime pinecanopy from sap-flow data. Agricultural and ForestMeteorology 71: 61–81.

HARRIS S. A. (1978) Vertical zonation of landsnailsin the Iraqi slopes of the Persian mountains and inthe Rocky Mountains of Alberta, Canada. Arcticand Alpine Research 10: 457–463.

HAVRANEK W. M. & BENECKE U. (1978) Theinfluence of soil moisture on water potential, tran-spiration and photosynthesis of conifer seedlings.Plant and Soil 49: 91–103.

HOLMES R. L. (1983) Computer-assisted qualitycontrol in tree-ring dating and measurement. Tree-Ring Bulletin 43: 69–78.

HOLMES R. L. (1994) Dendrochronology programlibrary, Version 1994. Laboratory of Tree-RingResearch, University of Arizona, Tucson.

HUGHES M. K., SCHWEINGRUBER F. H., CART-

WRIGHT D. & KELLY P. M. (1984) July–Augusttemperature at Edinburgh between 1721 and1975 from tree-ring density and width data.Nature 308: 341–343.

KIBE T. & MASUZAWA T. (1992) Seasonal changesin the amount of carbohydrates and photosyn-thetic activity of Pinus pumila Regel on alpine incentral Japan. Proceedings of NIPR Symposium ofPolar Biology 5: 118–124.

KIENAST F., SCHWEINGRUBER F. H., BRÄKER

O. U. & SCHÄR E. (1987) Tree-ring studies onconifers along ecological gradients and the poten-tial of single-year analyses. Canadian Journal ofForest Research 17: 683–696.

KIKUZAWA K. (1983) Leaf survival of woody plantsin deciduous broad-leaved forests. 1. Tall trees.Canadian Journal of Botany 61: 2133–2139.

KIMURA M. (1963) Dynamics of vegetation in rela-tion to soil development in northern Yatsugatake

Mountains. Japanese Journal of Botany 18: 255–287.

KIRA T. (1949) Forest Zone of Japan. Ringyogizyutu-kyokai, Tokyo (in Japanese).

KOIKE T. (1987) The growth characteristics in Jap-anese mountain birch (Betula ermanii) and whitebirch (Betula platyphylla var. japonica), and their dis-tribution in the northern part of Japan. In: HumanImpacts and Management of Mountain Forests (eds T.Fujimori & M. Kimura) pp. 189–200. Forestry andForest Products Research Institute, Ibaraki.

KULLMAN L. (1993) Pine (Pinus sylvestris L.) tree-limit surveillance during recent decades centralSweden. Arctic and Alpine Research 25: 24–31.

MILLER B. J., CLINTON P. W., BUCHAN G. D. &ROBSON A. B. (1998) Transpiration rates andcanopy conductance of Pinus radiata growing withdifferent pasture understories in agroforestry sys-tems. Tree Physiology 18: 575–582.

NAGANO METEOROLOGICAL OBSERVATORY (1998)Annual Report on Weather Conditions of Nagano Pre-fecture in 1998. Nagano Branch of Japan Meteoro-logical Society, Nagano (in Japanese).

OBERHUBER W., STUMBÖCK M. & KOFLER W.(1998) Climate-tree-growth relationships of Scotspine stands (Pinus sylvestris L.) exposed to soil dry-ness. Trees 13: 19–27.

OHSAWA M. (1995) Latitudinal comparison of alti-tudinal changes in forest structure, leaf-type, andspecies richness in humid monsoon Asia. Vegetatio121: 3–10.

OHWI J. (1961) Flora of Japan. Shibundo, Tokyo (inJapanese).

RIGLING A., WALDNER P. O., FORSTER T.,BRÄKER O. U. & POUTTU A. (2001) Ecologicalinterpretation of tree-ring width and intraannualdensity fluctuations in Pinus sylvestris on dry sitesin the central Alps and Siberia. Canadian Journal ofForest Research 31: 18–31.

SANO Y., MATANO T. & UJIHARA A. (1977)Growth of Pinus pumila and climate fluctuation inJapan. Nature 266: 159–161.

SCHWEINGRUBER F. H., FRITTS H. C., BRÄKER

O. U., DREW L. G. & SCHÄR E. (1978) The X-ray technique as applied to dendrochronology.Tree-Ring Bulletin 38: 61–91.

SOLOMINA O. N., MURAVYEV Y. D., BRAEUN-

ING A. & KRAVCHENKO G. N. (1999) Two newring width chronologies of larch and birch fromthe Kamchatka peninsula (Russia) and their rela-tionship to climate and volcanic activities. In: Cry-ospheric Studies in Kamchatka II (ed. R. Naruse) pp.111–124. Institute of Low Temperature Science,Hokkaido University, Sapporo.

558 K. Takahashi et al.

SWEDA T. & TAKEDA S. (1993) Construction of an800-year-long Chamaecyparis dendrochronologyfor central Japan. Dendrochronologia 11: 79–86.

TAKAHASHI K. (1962) Studies on vertical distribu-tion of the forest in middle Honshû. Bulletin of theGovernment Forestry Experiment Station 162: 1–171(in Japanese).

TAKAHASHI K. (2003) Effects of climatic condi-tions on shoot elongation of alpine dwarf pine(Pinus pumila) at its upper and lower altitudinallimits in central Japan. Arctic, Antarctic, and AlpineResearch 35: 1–7.

TAKAHASHI K., HOMMA K., SHIRAIWA T.,VETROVA V. P. & HARA T. (2001) Climatic fac-tors affecting the growth of Larix cajanderi in theKamchatka Peninsula, Russia. Eurasian Journal ofForest Research 3: 1–9.

TAYLOR A. H. (1995) Forest expansion and climatechange in the mountain Hemlock (Tsuga mertensi-ana) zone, Lassen Volcanic National Park, U.S.A.Arctic and Alpine Research 27: 207–216.

WARDLE P. (1968) Engelmann spruce (Picea engel-mannii Engel.) at its upper limits on the frontrange, Colorado. Ecology 49: 483–495.

WILSON R. J. S. & HOPFMUELLER M. (2001) Den-drochronological investigations of Norway sprucealong an elevational transect in the Bavarian For-est, Germany. Dendrochronologia 19: 67–79.

YASUE K., FUNADA R., FUKAZAWA K. &OHTANI J. (1997) Tree-ring width and maxi-mum density of Picea glehnii as indicators of cli-matic changes in northern Hokkaido, Japan.Canadian Journal of Forest Research 27: 1962–1970.

YASUE K., FUNADA R., KONDO T., KOBAYASHI

O. & FUKAZAWA K. (1996) The effect of cli-matic factors on the radial growth of Japanese ashin northern Hokkaido, Japan. Canadian Journal ofForest Research 26: 2052–2055.

YOSHINO M. M. (1978) Altitudinal vegetationbelts of Japan with spatial reference to climaticconditions. Arctic and Alpine Research 10: 449–456.