MASTERCHRONOLOGIES OF NORWAY SPRUCE /...I v l u I RESH FOREST SITES /.

t FORE! A . '

ETRY VITAS

Masterchronologies of Norway Spruce (Picea abies (1.) Karsten) on Fresh Forest Sites in Lithuania ADOMAS VITAS Vyta~ltas Magnus University, Faculty of Natural Sciences Vileikos g. 8, 3035 Kaunas, Lithuarzia

Vitas A. 2003. Masterchronologies of Norway Spruce (Picea ahies (L.) Karsten) on Fresh Forest Sites in Lithuania. Baltic Forestry, 9 ( I ) : 63-68.

Mastcrchronologics from living Norway sprucc trccs (Picea c~hies (L . ) Karstcn) in Piceetrrrn vaccitlio-rn,vr/illos~rrn (PC. v-rn) and Piceetltrn heputico-oxalidosurn (PC. h-ox) forcst typcs in Lithuania a rc crcatcd for thc first timc. Mastcrchronologies for each forcst typc includc data on latcwood, carlywood and annual radial growth. Mastcrchronology in PC. v-m forest typc cncompasscs thc data on thc radial growth from 18 rcscarch plots (464 trccs) and in Pc. A-ox - from 13 rcscarch plots (354 trccs). Timc span of mastcrchronology, crcated using thc annual radial growth of living sprucc trccs in PC. v-rn forcst typc is 132 ycars and in PC. h-ox - 100 ycars.

Corrclation cocfficicnts and dctected event ycars bctwccn crcatcd masterchronologics and a rcgional chronology of sprucc from Olsztyn Lakc dis tr ic t (Ziclski , Koprowski 2001) showcd a wcll cnough similar i ty bctwccn thcir growth dynamics.

Key words: clirnatc, ear lywood, cvcnt ycar, latcwood, Lithuania, mastcrchronology, Norway sprucc, radial growth

Introduction

Owing to the long-life and anatomical structure of the tree rings, conifers are one of the most impor- tant objects of the dendrochronogical research in Lithuania (Bitvinskas 1995). Due to Scots pine (Pinus sylvestris L.) domination in Lithuanian conifer forests, the main part of dendrochronological research was concentrated on pine and only a smaller part on spruce. Consequently, Norway spruce was less investigated than Scots pine (KarpaviEius, Kairaitis, Yadav 1996, Vitas, Bitvinskas 1998).

Masterchronology is a chronology averaged from any tree-ring parameter, usually combining site or lo- cal chronologies from a given region, and used to date new tree-ring series (Kaennel, Schweinguber 1995).

According to the data of National Forest Exploi- tation, spruce growing on dry sites - Piceetum vac- cinio-myrtillosum and Piceetum hepatico-oxalidosum occupies more than 6.8% of the total forest area (OzolinCius 1998) of Norway spruce stands (23.1 %) (KulieSis, Rutkauskas, Butkus 2001).

The aim of the research was to construct master- chronologies for the radial growth of living Norway spruce trees on dry forest sites in Lithuania.

Materials and methods

A total of 33 research plots on dry stands (forest types: Piceettim vaccinio myrtillosum, Piceetum he-

patico-oxalidosum) of Norway spruce (Picea abies (L.) Karsten) in Lithuania were selected. During the fieldwork, by using increment borer, samples were tak- en at the breast height. Aiming to the reliable materi- al, samples from 20 to 30 trees in each plot were taken (Fritts, Shatz 1975, Huber 1978, ~ H T B H H C K ~ C 1974). Using GPS (geographical position system) unit MAG- ELLAN 3 15 geographical coordinates for each research plot were established.

Tree ring widths (earlywood and latewood sepa- rately) were measured within 0.01 mm accuracy. For this purpose, the LINTAB tree-ring measuring table and TSAP 3.12 computer program, developed by F. Rinn and S. Jakel in Heidelberg were used.

For the dating quality control (Eckstein 1987, Kien- ast 1985, Lovelius 1997, Wendland 1975) COFECHA 3.00P computer program, developed by R. Holmes was used. Using CHRONOL 6.00P computer program de- signed by E.R. Cook, the indexing was performed. Each tree-ring series obtained from individual tree were in- dexed independently. The indexing was carried out at two stages - according to the methods, proposed by Holmes et al. in 1986. At first negative exponential curve or linear regression was used and after the spline, preserving 67% of variance at wave length 2 1 years, was fitted (Holmes 1994).

After the dating quality control and indexing, each tree-ring width series was averaged into a site chro- nology of research plot. The biweight robust mean for this purpose was used. A total of 33 site chronolo-

2003, Vol. 9, No. 1 (16) ISSN 1392-1 355 =

STRY A - . . - -

gies were constructed. Totally 3 1 site chronology were included into compiled masterchronologies. For the lnasterchronology developed for PC. v-rn forest type, site chronologies from 18 research plots (464 trees) were used and for the masterchronology representing PC. h-ox forest type, 13 site chronologies (354 trees) were used. Examination of the localities of research plots and dating quality control between the site chro- nologies showed that due to a low similarity with oth- ers two site chronologies should not be included into masterchronologies. The low similarity of Dubrava 1 (Central Lithuania) site chronology with others could be attributed to a high ground water level because stand is close to a stream. The second site chronolo- gy named Kretinga (West Lithuania) was too short (35- 40 years old stand) with asynchronous radial growth to include into masterchronologies.

Similarity between chronologies using the coeffi- cient of correlation and Gleichla.ufigkeit (GLK) coeffi- cient of similarity was evaluated. These calculations were performed using TSAP 3.12 and DPL computer programs.

The constructed ~nasterchronologies were compared with site chronologies, submitted to the ITRDB (http:/I www.ngdc.noaa.gov/paleo/treering) by E. Feliksik and F.H. Schweingruber named Gasienicowy, Hala Gasleni- cowa and Swystowko Wyznig (Southern Poland) and with a regional chronology of Norway spruce constructed for Olsztyn Lake district (Zielski, Koprowski 2001).

Event years were counted from index values: years with indices of 75 and smaller were valuated as negative event years and with indices of 120 and big- ger - positive event years.

Results

Several characteristics of research plots and site chronologies, used to develop rnasterchronologies in Table 1 are presented: name of research plot, geograph- ical coordinates (WGS 84), forest type, span of chro- nology (years), mean sensitivity of latewood, early- wood and annual ring. Time span of site chronologies varies from the shortest - 66 years to the longest - 162 years. The mean sensitivity of latewood is 0.20 - 0.43, of earlywood - 0.14 - 0.24 and of the annual ra- dial growth - 0.13 - 0.23. The average age of site chro- nologies is 104 years and the mean sensitivity of the latewood - 0.29, earlywood - 0.18 and the annual ra- dial growth - 0.17.

Nine site chronologies have time span more than 100 years with more than 10 trees and are most valu- able. The data of them was submitted to the ITRDB located at Boulder USA (http://www.ngdc.noaa.gov/ paleo1treering.html).

Table 1 . Characteristics of research plots and site chronol- ogics

Time span of conlpiled masterchronology for PC. v-m forest type covered by 10 and more trees is 132 years, and for PC. h-ox - 100 years (Table 2). Mean sensitivity of masterchronologies for the latewood is 0.22-0.23, earlywood - 0.14-0.16 and the annual radial growth - 0.12. So, the most sensitive are series of late- wood radial growth and more complacent are series created for the annual radial growth.

Correlation coefficients between two masterchro- nologies are: +0.80 (latewood), +0.78 (earlywood) and

Table 2. Characteristics of constructed niastcrchronologics for Norway spruce. Prcscnted time span of mastcrchrono- logy covcrcd by 10 or morc trces

2003, Vol. 9, No. 1 ( 1 6) ISSN 1392-1 355 =

Forcsl

type

Pr. v-m

Pr. 11-or

Wood

Late

Early

Ali11ui11

L:lle

Early

I I I U

Nul~iber

of trccs

35 1

449

46.4

156

345

354

Period (years)

1873-1999

1869- I999

1868-1999

1899-1999

1894- I999

19i3+1999

SPLIII of

~~nstcrchro~iology

(ye:~rs)

127

131

132

101

106

IW

N L I I I I ~ C ~ of

reseilrcll

plots

I4

18

18

I ?

13

13

BALTIC FORE! MASTERCHRONOLOGIES OF NORWAY SPRUCE /.../ ON FRESH FOREST SITES /..., A . VIIAb - ETRY

+0.89 (annual radial growth) and coefficients of simi- Table 3. Indices of latewood (upper valuc), earlywoocl (mid-

larity (GLK) - 71.3, 77.3 and 81.4 respectively. So, the dle valuc) and annual radial growth (lower valuc) master-

differences between the dynamics of the radial growth ~ h r o n o l o g y for Norway spruce in PC. v-m forest type

of two masterchronologies (forest types Piceetunz vaccirrio-myrtillosum and Piceetum hepatico-oxalido- sum) are insignificant (Vitas 2002). Graphs of master- chronologies in Figures 1 , 2, 3 and indices of the ra- dial growth in Tables 3 , 4 are presented.

Figure 1 . Mastercl~ronologies of Norway spruce latewood radial growth

Figure 2. Mastcrchronologies of Norway spruce earlywood radial growth

Figure 3. Masterchronologies of Norway spruce annual ra- dial growth

Discussion

Carrying out the dating quality control was es- tablished that the number of trees with dissimilar ra- dial growth or suspected with dating problems on average is 3 1% of latewood, 19% of earlywood and 16% of annual growth series, the data of which were not used for the further analysis. Poorer dating quality or greater amount of dissimilar tree radial growth was detected among the latewood series, which could be explained with reaction-compression wood. Trees of

dissimilar growth are from different age categories: significantly younger or older than the average tree in the stand. Previous researchers have also pointed out that tree number with poor dating quality in plot varies from 15% to 30% (Bitvinskas 1974) and dating of Norway spruce radial growth series between trees are often complicated (Becker 1978).

Similarity between the radial growths of site chro- nologies could be characterised by high enough co- efficients of correlation (average +0.60 - +0.62, lowest +0.26 +- +0.37 and highest +0.82 - +0.84). Similarity slightly decreases with increasing distance between research plots, but even between the farthest plots (Sventoji 2, Juodkrante and VaiSnoriSke are approxi- mately 300 km in distance), similarity is significant (coefficients of correlation varies from +0.3 1 to +0.55).

2003, Vol. 9, No. 1 (1 6 ) ISSN 1392-1 355 =

M A S l t K C H K O N U L O C j l t S Ut NVKWAY W K U L t /.../ UN t K t b H f O ~ t > l >lit> /..., A .

Table 4. Indices of latcwood (upper value), earlywood (mid- masterchronologies with previolls researchers is quite

dle value) and annual radial growth (lower valuc) master- impossible, because comprehensive analysis on the chronology for Norway spruce in PC. h-o.r forcst type latewood and earlywood widths of Norway spruce were

not performed and published in Lithuania. Any references on masterchronologies on spruce

from the neighbouring regions in the Henri's bibliog- raphy pages (ht tp: / / t ree. l t rr .ar izona.edu/cgi-bin/ bibliosearch.pl) werc not found. I found three site chronologies for Norway spruce from upland region of South Poland submitted by E. Feliksik and F. H. Schweingruber to the ITRDB (http://www.ngdc. noaa.gov/paleo/treering). Due to a long distance (ap- proximately 500 km) between the locations of re- search plots and differences in climate character s i n ~ - ilarities could not be found (correlation coefficients are -0.03 - -0.14 and coefficients of similarity - 38.5- 54.2). Mentioned coefficients indicate dissimilar growth dynamics of compared chronologies.

From Andrzej Zielski and Marcin Koprowski, I received its paper, published in the Journal Sylwan in 2001. A regional chronology of Norway spruce from Olsztyn Lake District (approximately 250 km from the nearest research plots used to compile masterchronol- ogies) in this article is presented. Its chronology has a time span of 160 years, from 1840 until 1999. My masterchronologies with the mentioned regional chro- nology shows a very good similarity in their growth dynamics (correlation coefficients are +0.68 - +0.70 and coefficients of similarity - 73.8-76.2). These coefficients are significant (p=0.05).

For a more clear view on the similarities and dif- According to the obtained research results on Nor- ferences of my masterchronologies and a regional chro- way spruce tree rings, it gives no well-grounded ar- nology of spruce from Olsztyn Lake district, event years guments to construct regional masterchronologies for of the annual radial growth in Table 5 are presented. spruce in Lithuania. From Table 5 it can be seen that the main event years

A comparison of constructed masterchronologies coincide between three chronologies well enough, e.g. with the other chronologies was not an easy task. To in 1889 (-), 1941 (-), 1946 (+), 1961 (+), 1974 (+), 1979 compare earlywood and latewood series of compiled (-), 1980 (-), 1987 (+), 1990 (+) and 1992 (-).

Table 5. Established event years of masterchronologies. Minus marks (-) in the table indicate negative events, plus (+) - positive event ycars

Yrur.r

1870 1872 1878 1879 1880 1887 1889 1903 1916 1922 1936 1940 1941 1943 1946 1950 1954

2003, Vol. 9, No. 1 (1 6 ) ISSN 1392-1 355 =

Muster- Mu.vrer- Ol,rzpr c.hronology chrono10,y.y Luke

Pi,. v-m PC. 11-ox districr

+ +

+ +

+ +

+

+ + + + + +

Yeur.r

1958 1959 1961 1962 1967 I974 1979 1980 1983 1986 1987 1988 1990 1992 1993 1999

hf[~,strr- Ma.ster- 0l.rztvn chronology chronolo,qy krke

PC:. v-m PC. h-ox (1i.rtrict + +

+ + + + +

+ + + +

+ +

+ + +

+ +

+

BALTIC FORESTRY MA~I~KLHKuNuLuc~I~~ u t I\IOKWHY 3rKULt /.../ UIU tKt3H t U K t 3 1 SITES /..., A . VITAS -

According to the results of my research carried out on the tree rings of Norway spruce in Lithuania, dry spring and summer conditions play the main neg- ative role in the formation of spruce negative pointer years in Lithuania ( I 94 1, 1979, 1980 and 1992). Posi- tive pointer years form in the response to warm spring and wet summer conditions (1946 and 1974). The im- pact of summer droughts is especially p o w e r f ~ ~ l after snow-less winter (1941) (Vitas 2002). Main results re- ceived during the research of pointer years of Norway spruce corresponds with the results obtained in oth- e r countries: spruce is very sensit ive to summer droughts and prefers cool and moist weather (Schwein- gruber 1993).

Summarizing, it could be stated that the radial growth of spruce in Lithuania and in the Olsztyn Lake District (North-east Poland) is favoured by similar cli- mate conditions. While, in the upland of Southern Poland (site chronologies by F.H. Schweingruber and E. Feliksik) Norway spruce is favoured by different climate conditions than in Lithuania and North-east- ern Poland.

Conclusions

1. Modern masterchronologies for Norway spruce on dry forest types are created and could be used by scientists in Lithuania and in the neighbouring coun- tries.

2. Coefficients of correlation and similarity, and event years showed that compiled masterchronologies have good similarity in their growth dynamic with the regional chronology for Olsztyn Lake district. It points out to the similar climate conditions in individual years and analogous reaction of Norway spruce radial growth to climate in mentioned regions.

3. Event years of masterchronologies showed that dry spring and summer conditions plays the main negative role in the formation of narrow rings of spruce in Lithuania

Acknowledgements

I am grateful to my collerzg~ies - Prof: Andrzej Zielski and Marciiz Koprolvski for sending rile his ar-

References

Bccker, B. 1978. Dcndroccological Zorics of Ccntral Europc- an Forcst Communities. In: J. Flctchcr (Editor), Dcndro- chronology in Europc. National Maritilnc Muscum, Grccn- wich, p. 101-109.

Bitvinskas, T. 1995. Norway sprucc productivity in thc Ncrnu- nas Basin 1930-1990. Dcndroclimatochronology lab. Rcport. (In Lithuanian)

Eckstein, D. 1987. Mcasurcmcnt and dating proccdurcs in dcndrochronology. In : L. Kairiukstis. Z. Bcdnarz, E. Fc- liksik (Editors), Mcthods of Dcndrochronology. IIASA, Warsaw, 3: 35-44.

Fri t ts , H.C., Shatz , D.J. 1975. Sclccting and Characterizing Trcc-ring Chronologies for Dcndroclirnatic Analysis. Trcc-Ring Bullctin, 35: 3 1-40.

Holmes, R.L. 1994. Dcndrochronology program library. Lab- oratory of Trcc-ring Rcscarch, University of Arizona: Tucson.

Hubcr, B. 1978. Annual Trcc-Rings as an Aid to Climatology and Chronolngy. In: J . Flctchcr (Ed~tor ) , Dcndroclironol- ogy in Europc. National Maritimc Muscum, Greenwich, p. 11-26.

Kaenel, M., Schrvcingrubcr, F. H. 1995. Multilingual Glos- sary of Dcndrochronology. Paul Haupt Vcrlag, Bcrnc, 467 p.

KarpaviEius, J., Kairai t is , J . , Yadav, R.R. 1996. Influcncc of Tcrnpcraturc and Moisturc on tlic Radial Growth of Scots Pinc and Norway Sprucc in Kaunas. Lithuania. Thc Korcan Journal of Ecology, 19(4): 285-294.

Kicnast, F. 1985. Dcndrookologischc Untcrsuchungcn an Hohcnprofilcn aus vcrschicdcncn Klimabcrcichcn. Inau- gural-Dissertation, UnivcrsitIt Ziirich, 129 p. (in Gcrman)

KulieSis, A., Rutkauskas , A., Butkus, A. (Editors) 2001. Lictuvos miSkq valstybinc apskaita 2001 m. sausio I d. [Statc inventory of Lithuanian forcsts, I " January 20011. LVMI, Kaunas, 76 p. (in Lithuanian)

Lovelius, N.V. 1997. Dcndroindication of Natural Proccsscs. World&Farnily-95, St. Pctcrsburg, 316 p.

OzolinEius, R. 1998. Conifcrs in Lithuania: Transformations of Morphological Structurc and Factors Inducing Thcrn. Lututc. Kaunas, 300 p . (in Lithuanian)

Schtveingruber, F.H. 1993. Jahrringc und Umwclt dcndrookol- ogic. Lis, Vologda, 474 p. (in Gcrman)

Vitas, A. 2002. Klimato vciksniq jtaka paprastosios cglcs (Picetr crbies (L.) Karstcn) radialiajam pricaugiui [Impact of climatc factors on the radial incrc~ncnt of Norway sprucc (Piceu nhies (L.) Karstcn)]. Daktaro discrtacija [Doctoral dissertation]. VDU, Kaunas, 114 p. (in Lithua- n ian)

Vitas, A., Bitvinskas, T. 1998. Dcndroclimatological sirni- laritics of Piceu ubics (L.) Karstcn and Pitllts .s,yli~cs~ri,s L. Baltic Forestry, 4(1): 24-28.

Wendland, W.M. 1975. An Objective Mcthod to Idcntify Missing or Palsc Rings. Trcc-Ring Bullctin, 35: 41-47.

Zielski A., Koprowski M. 2001. Dendrochronologiczna anali-

ticle: , ,A dendroclzronologicnl analysis o f aniziral za przyrostow rocznycli Swierka pospolitcgo na Pajczicr-

rings o f Norway sprzice of the Olsztyn Lake District ". zu Olsztyriskitn [A dcndrochronological analysis of an- nual rings in Norway sprucc of thc Olsztyn Lakc District].

Stati.stica1 analysis o f tree ring width series rising Sylwan, 7 : 65-73. (in Polish) COrllpliter program TSAP 3.12 Was performed. The ~ H T B I ~ H C K ~ C , T.T. 1974. ~ ~ C H ~ ~ O K ~ I ~ M ~ T U ~ ~ C C K M C MCCnCnOBaHMR

author wished to e- press lzis thanks to tlze owners of [Dcndroclimatological Rcscarch]. ~ I ~ A ~ O M C T C O U ~ ~ ~ T , n c -

the program - colleaglles from Department of Envi- HuHrpan, 172 p. (in Russian)

ronmerztal Sciences of VMU kindly allowed me to use his software. Received 12 Deccrnbel-2002

2003. Vol. 9. No. 1 (16) lSSN 1392-1 355 =

BALTIC FORESTRY MASTERCHRONOLOGIES OF NORWAY SPRUCE /.../ ON FRESH FOREST SITES / . . . / A . VlTAS

3TAJIOHHbIE AEHAPOIlIKAJIbI EJIEl O E ~ I K ' H O B E H H O ~ (PICEA ABIES (L.) KARSTEN) B CBEXHX MECTOIIPOET3POCTAHMtM: JIECOB JIMTBbI

~ T & I I o H H ~ I ~ AeHApOIUKUlbl U 3 H(I.lBbIX AepeBbeB enU O ~ ~ ~ K H O B ~ H H O ~ ~ (Picea abies (L.) Karsten) B TkinaX neCa (PiCeetlllll ~ a ~ ~ i l l i ~ - m ~ r t i ~ l ~ . ~ ~ l m A P ~ C E E ~ L I M h~p(lti~0-~.~fl/id~~utll) C03AaHHbl B JIATBC BnepBbIe.

K a x n a ~ UKarra BKnmuaeT B ce6e DaHHbre pawero, no3naero u ronuworo panuanbHoro npupoc-ra. U I ~ a n a B Tunc neca (Piceetum vaccinia-tn)~r.tillon.r~r~n) BKnmqaeT AaHHbIe panuarrbHoro npupocTa u3 18 I I ~ O ~ H M X nnomanefi (464 nepesbe~) a urKana B enbHuKax Tuna Piceetum he~~atico-o.mlidosrrn~ - 13 npo6~blx nnouanefi (354 nepeabea). C o s n a ~ ~ b r e AeHnpOUIK~bI panaanbHoro npupocTa enu O ~ ~ I K H O B ~ H H O ~ ~ B enbHuKax Tuna Piceeturn vnccinio mvrtillos~trn oxsaTbrsam 132 rona a B Tunax Piceet~~m hepntico-oxnlidostrm - 100 nm.

K O ~ ~ H U U ~ H T ~ I KOppenXUUU U BbIXBnRHHble KpHTUYeCKHe rOAb1 paAUanbHOr0 IIpHpOCTa CO3AaHHbIX AeHApOIUKan IlOKZ3bIBaFOT 6nu3~0e CXOACTBO C peru0HanbH0fi x ~ o H o J I o I ' u ~ ~ ~ A 3 n0nbLIl~.

2003, Vol. 9, No. 1 (1 6 ) ISSN 1392- 1355 =