ARTICLE IN PRESS

Dendrochronologia 25 (2007) 103–112

1125-7865/$ - se

doi:10.1016/j.de

�CorrespondE-mail addr

www.elsevier.de/dendro

ORIGINAL ARTICLE

The impact of the 2003 summer drought on the intra-annual growth

pattern of beech (Fagus sylvatica L.) and oak (Quercus robur L.) on

a dry site in the Netherlands

G.W. van der Werf�, Ute G.W. Sass-Klaassen, G.M.J. Mohren

Forest Ecology and Forest Management Group, Centre for Ecosystem Studies, Wageningen University, PO Box 47,

6700 AH Wageningen, The Netherlands

Received 10 March 2006; accepted 25 September 2006

Abstract

Climate change is expected to result in more extreme weather conditions over large parts of Europe, such as theprolonged drought of 2003. As water supply is critical for tree growth on many sites in North-Western Europe, suchdroughts will affect growth, species competition, and forest dynamics. To be able to assess the susceptibility of treespecies to climate change, it is necessary to understand growth responses to climate, at a high temporal resolution. Wetherefore studied the intra-annual growth dynamics of three beech trees (Fagus sylvatica L.) and five oak trees (Quercus

robur L.) growing on a sandy site in the east of the Netherlands for 2 years: 2003 (oak and beech) and 2004 (oak).Microcores were taken at 2-week intervals from the end of April until the end of October. Intra-annual tree-ringformation was compared with prior and contemporary records of precipitation and temperature from a nearbyweather station.

The results indicate that oak and beech reacted differently to the summer drought in 2003. During the drought,wood formation in both species ceased, but in beech, it recovered after the drought. The causes of species-specificdifferences in intra-annual wood formation are discussed in the context of susceptibility to drought.r 2007 Elsevier GmbH. All rights reserved.

Keywords: Wood formation; Microsampling; Pinning

Introduction

It has been predicted that climate change will result inincreased precipitation in North-Western Europe, no-tably during the winter (Metzger, 2005; Beersma andBuishand, 2004), and more pronounced droughts duringsummer (Zinyowera et al., 2001). The year 2003 waswarm and sunny (mean annual temperature of 10.3 1Cmeasured in De Bilt, the Netherlands), with precipita-

e front matter r 2007 Elsevier GmbH. All rights reserved.

ndro.2007.03.004

ing author. Tel.: +31317 478045; fax: +31317 478078.

ess: geesje.vanderwerf@wur.nl (G.W. van der Werf).

tion during the summer months being the lowest in 100years (Fig. 1).

Many of the forested areas in the Netherlands arelocated on sandy soils with limited water-holdingcapacity. These planted forests predominantly consistof Scots pine (Pinus sylvestris L.) and other coniferousspecies; they are gradually changing into more natural,mixed forests with beech and oak, with oak predomi-nating on the drier sites (Den Ouden and Mohren,2004). On these sandy soils, water availability is crucial indetermining the competitive relations (mainly betweenoak and beech) in mixed stands.

ARTICLE IN PRESS

Fig. 1. Climate diagram, De Bilt weather station, the Netherlands: monthly mean temperature (lines) and monthly total

precipitation (bars) in 2003 (gray) compared with the means from 1905–2005 (black).

G.W. van der Werf et al. / Dendrochronologia 25 (2007) 103–112104

Dendrochronological studies on oak and beechgrowing on dry sites in Central and Southern Europehave indicated that the tree-ring width (tree-RW) ofboth the species correlates positively with early summerprecipitation and negatively with high summer tempera-ture, which implies they will suffer if there is summerdrought (e.g., Eckstein and Schmidt, 1974; Pilcher andGray, 1982; Dittmar et al., 2003; Lebourgeois et al.,2005). However, under a temperate maritime climate, asoccurs in the Netherlands, the RW of beech andespecially of oak often fails to show a clear responseto either temperature or precipitation in a particularmonth or season. The climate signal is too complex tobe captured by tree-RW, which inherently integratesboth favorable and unfavorable environmental signalsover one or more years and lacks the resolution tocapture and discriminate between different environmen-tal signals. Intra-annual higher resolution variables suchas earlywood width and latewood width (LW) (Nola,1996; Garcia-Gonzalez and Eckstein, 2003) and vessel-area variables (Sass and Eckstein, 1995; Fonti andGarcia-Gonzales, 2004; Eilmann et al., 2006) mayprovide a solution to this problem, although additionalinformation about intra-annual variation of woodformation is needed to produce more detailed physio-logical explanations.

Wood formation throughout the growing season canbe studied either by the pinning method, i.e., introdu-cing artificial markers into the wood by wounding thecambium (e.g., Wolter, 1968; Nobuchi et al., 1995;Schmitt et al., 2000), or by taking microcores (Deslaur-iers et al., 2003). In this preliminary study we testedboth the methods. The objectives of this study were(1) assessing and comparing specific differences in intra-annual growth dynamics during 2003, the year with thedriest summer since the start of instrumental recording

in the Netherlands in 1899, and (2) using the pattern ofintra-annual wood formation in oak and beech to betterexplain changes in the widths of earlywood, latewoodand rings in a temperate climate.

Materials and methods

Study site

The study area was located in Oostereng (in theVeluwe area, centre of the Netherlands, 511590N,51430E), in an approximately 80-year-old mixed forestof oak (Quercus robur L.) and Scots pine (Pinus

sylvestris L.) with beech (Fagus sylvatica L.) growingon a poor sandy soil with the water table about 12mbelow the surface. The climate is temperate maritime,with an average annual temperature of 9.5 1C and amean annual precipitation of 760mm in the period1905–2005. The climate diagram from De Bilt weatherstation (Fig. 1) indicates only little variation in the meantotal monthly precipitation and mean maximummonthly temperature during the summer months(June–August). In the summer of 2003, the temperaturewas 2 1C higher and the precipitation was only 59%compared with the averages for the period 1900–2000.

Study trees and periodic cambial sampling

In 2003, six dominant oaks (aged 50–80 years; diameterat breast height, DBH 14–37cm) and three dominantbeeches (age approximately. 80 years, DBH 37–54 cm)were selected for periodic cambial sampling. From April29 to October 17, microcores were taken at intervals of 2weeks. The small cores (1.5mm in diameter, 10mm long)

ARTICLE IN PRESSG.W. van der Werf et al. / Dendrochronologia 25 (2007) 103–112 105

were obtained using surgical bone-sampling needles(Meekers Medical, Utrecht, the Netherlands) at twoopposite sides of the tree, to take into account the radialvariation around the stem. Samples were taken starting atDBH and then spiraling up the stem but always spacedapproximately 5 cm apart in all directions, to avoid lateralinfluence of wound reactions on adjacent cores. Themicrocores were placed in tubes containing a solution ofglycerin (50%) in water.

In 2004, the wood formation of five additional oakswas investigated from April 26 to October 11 using apinning method that entailed wounding the cambiumevery 2 weeks with a tiny needle (diameter 0.8mm) tointroduce artificial markers in a spiral pattern up thestem, spaced approximately 5 cm minimum apart(Wolter, 1968). At the end of the growing season, smallwooden blocks (1 cm3) containing the needle incisionpoints were extracted with a chisel.

In October 2004, three increment cores (diameter5mm) were taken at DBH from each sampled tree fordendrochronological analysis.

Dendrochronological analysis

The increment cores from 11 oaks and three beecheswere fixed on wooden holders, air-dried and sanded.Earlywood width (EW), LW and RW of oak and RW ofbeech were measured with a precision of 1/100mm usinga LINTAB measuring device (RINNTECH) and TSAPprogram (Rinn, 1996). The time-series were cross-datedvisually and statistically using the TSAP and COFE-CHA programs (Holmes, 1983). Each radius wasdetrended by fitting a cubic smoothing spline with a50% frequency cut-off of 30 years (ARSTAN program,Cook and Holmes, 1997) in order to keep mainly thehigh-frequency variability (Cook, 1985). The time-seriesof oak (EW, LW, RW) and beech (RW) werestandardized by dividing observed by expected values,and a standard tree-ring chronology was computed byapplying bi-weight robust mean (Cook, 1985). Climate–growth relationships were calculated using the Dendro-Clim 2002 program (Biondi and Waikul, 2004), in orderto determine the climate factors mainly responsible forthe variation of the EW, LW, and RW chronologiesfrom 1950 to 2004, the predictor variables used weremonthly mean temperature and total precipitation fromthe previous July to the September of the current year.

Preparation and measurement of the micro samples

and pinning samples

A total of 288 micro samples were stored in glyceroland embedded in paraffin. Thin sections of 8 mmthickness were prepared using a rotary microscope andstained with a safranin–astrablue mixture. The staining

mixture was prepared by adding 40mg safranin and150mg astrablue to a solution of 100ml demineralizedwater with 2ml acetic acid. The samples were stained for10min and kept in glycerol after measuring. In total, 80pinning samples were glued on small wooden holderswith water-resistant glue and cut with a sliding micro-tome to prepare sections of 20 mm thickness, which werestained with the astrablue–safranine mixture and kept inglycerol after being measured.

The amount of wood formed in the developing treering was quantified by measuring the sections under thelight microscope at 100� magnifications. The incre-ment between consecutive micro samples describes thewood formation in a period of 2 weeks.

Standardization of the intra-annual increments

Cambial activity around the stem circumference maydiffer at short tangential distances, obscuring measure-ments of 2-weekly increments between adjacent samples.To overcome this problem, the measurements werestandardized, using the procedure described below. Asan example we use the year 2003; the same procedurewas also used to calculate the increment in 2004.

For the pinning samples, the amount of tissue formedat a certain time interval was calculated using thefollowing steps: (1) wound detection, (2) projection ofthe wound location left and right in radial direction intosound, undisturbed tissue, (3) measurement of theamount of tissue that had formed between the previoustree ring and the wound boundary at two positions leftand right from the wound and (4) calculation of therelative increment (Ir) by relating the mean of thesemeasurements to the mean of the total RW (RW2003) atthese two positions.

In order to estimate RW2003, the final RW at each sitewhere a microcore sample was taken, a factor (F) wascalculated (Fig. 2). F describes the ratio between the widthsof the tree rings formed in 2002 and 2003 at the last coringposition in which the tree ring for 2003 was complete:

F ¼RW2003

RW2002, (1)

where F is the standardization factor, RW2002 the width ofthe tree ring for 2002 at last microcore position (October17, 2003) and RW2003 the width of the 2003 tree ring at lastmicrocore position (October 17, 2003).

It is assumed that F stays constant even if absolutecambial activity varies around the stem circumference.RWE2003 (the estimated final RW) for each microcoreposition was then estimated by multiplying the RW2002

at these positions by F:

RWE2003 ¼ RW2002 � F (2)

Finally, the relative increment (Ir) at each microcoreposition was calculated by measuring the amount of

ARTICLE IN PRESS

Fig. 2. Visualization of standardization procedure: estimated

ring width (RWE) is calculated by a factor describing the ratio

between tree-ring widths of 2002 (RW2002) and 2003 (RW2003)

at the last coring position. M is the amount of wood formed in

2003 at one coring position. The irregular area in shaded gray

resembles the bark.

Table 1. Statistical variables of the oak and beech trees

studied (derived from COFECHA)

Mean (mm) Mean SD (mm) MS Mean AC

Oak RW 1.53 0.78 0.28 0.71

Oak EW 0.51 0.17 0.21 0.62

Oak LW 1.02 0.70 0.42 0.67

Beech RW 3.18 1.09 0.30 0.5

RW ¼ tree-ring width, EW ¼ earlywood width, LW ¼ latewood

width, Mean SD ¼ mean standard deviation, MS ¼ mean sensitivity,

Mean AC ¼ autocorrelation (lag 1).

Fig. 3. Oak chronologies: EW width (light gray), LW (dark

gray) and RW (black).

G.W. van der Werf et al. / Dendrochronologia 25 (2007) 103–112106

wood formed in 2003 (M) and relating it to theestimated RWE2003:

I r ¼M

RWE2003, (3)

where Ir is the relative increment and M the currentwidth of tree ring

Statistical analysis of the cambial measurements

Statistical analysis was done using GENSTAT 7.1(Hemel Hemstead, UK).

Profiles of intra-annual wood formation were estimatedby fitting Gompertz functions (Rossi et al., 2003) throughthe calculated relative increments, because the Gompertzfunctions has proven to be appropriate to model intra-annual tree growth (Camarero et al., 1998; Deslauriersand Morin, 2005; Rossi et al., 2003). The fitted curverepresents the growth trend, thereby smoothing short-term variations in growth rate (Rossi et al., 2003). Fiveoutliers (two for beech, three for oak in 2003) wereremoved from the dataset. In the outliers, the RW of thelast year was extremely low (less than one-third of themean RW of the last year for that tree). Comparisonswere performed between profiles of the same tree andbetween those of different trees. In order to be able torelate the trend in intra-annual ring formation with thecontemporary temperature and precipitation conditions,2-weekly as well as monthly totals of precipitation andmean temperature were calculated.

Results

Dendrochronological analysis

At the Oostereng site, the average radial growth ofbeech was greater than that of oak (3.18mm versus1.53mm: Table 1). This difference in growth cannot be

attributed to tree maturity, as the trees sampled weremature (beech approximately 60 years; oak approxi-mately 80 years). A comparison between RW, EW, andLW chronologies of oak reveals a high correlationbetween RW and LW of the same year (r ¼ 0.99; Fig. 3).For the same year, the EW is less correlated with RWthan with LW but is positively correlated with LW ofthe preceding year (r ¼ 0.68). Fig. 3 illustrates that theEW chronology varies only slightly from year to year, asis also shown by its low mean sensitivity and highautocorrelation (Table 1). The annual variation of theRW chronologies of oak and beech are similar (r ¼ 0.56;Fig. 4), but the oak chronology shows a higherautocorrelation, as can clearly be seen from theconsistent behavior of the chronology in years followingextreme small values (1959, 1976, 1996; Fig. 3).

In general, the tree-RW of beech is influenced moreby precipitation and temperature compared with oak(Fig. 5a and b). In both the species, there is generally apositive correlation between radial growth and precipi-tation. The temperature effect is less consistent, but thegrowth of both species is negatively correlated with thetemperature of the previous summer (in July andAugust). Beech and oak also respond positively to wetconditions, though beech responds best to wet conditions

ARTICLE IN PRESS

Fig. 4. Ring width chronologies of oak (gray) and beech

(black).

G.W. van der Werf et al. / Dendrochronologia 25 (2007) 103–112 107

in late autumn, whereas oak responds best to wetconditions in winter. However, the two species differ intheir response to weather during the growing season:whereas the growth of oak only correlates significantlypositively with early spring precipitation, beech growthcorrelates more strongly with the total precipitationduring the entire growing season, with significantcorrelation values in June and July. Concomitantly thereis a significant negative correlation between growth andtemperature in June (Fig. 5a).

The results of the climate–growth analysis for the LWoak chronologies are similar to those of the RWchronologies. The chronology shows the followingcorrelations with conditions at the end of the previousgrowing season: a positive correlation with above-average precipitation and a negative correlation withbelow-average temperature. It also shows a positivecorrelation with the precipitation at the beginning of thegrowing season (April–June) (results not shown).

Intra-annual growth patterns

Both the microcore and pinning methods weresuccessfully used to reconstruct intra-annual woodformation. In both the methods, but especially in themicrocore approach, difficulties were experienced be-cause of the large differences in the absolute values ofnewly formed tissue in consecutive microcore andpinning samples. This indicates that cambial activity inthe sample trees varies considerably across shorttangential distances around the stem circumference.However, by using a standardization procedure thattakes the general growth activity of a certain cambialsegment into account, we were able to normalize theseeffects. The normalization enabled us to filter out muchof the variation before constructing intra-annual growthprofiles.

The within-species intra-annual growth patternsdiffered. Of the three beech trees, two had a highergrowth rate (average RWs 2.42 and 2.05mm) comparedwith the third (1.81mm); the slower-growing treecompleted 80% of its tree ring earlier (September 18,2003, compared with October 3, 2003 for the two othertrees). However, all trees showed a slower growthpattern from the end of June until the beginning ofSeptember and there was another spurt of growth fromSeptember until mid-October.

In the five oak trees investigated, 80% of the 2003 treering was completed between July 11 and August 22 andthe tree-RW in 2003 ranged from 0.87 to 1.62mm. Thetwo oaks with the smallest tree ring ceased growing asearly as July 11.

In 2004, the RW in oak has a slightly lower range:between 1.05 and 1.58mm; 80% of the tree ring wascompleted later than in 2003 (between August 16 andSeptember 13).

A comparison of the intra-annual growth patterns ofoak and beech in 2003 (Fig. 6) shows that both speciesstarted to grow at approximately at the same time,between April 29 and May 16. Beech grew rapidly fromthe middle of May until the beginning of July, whengrowth slowed down. From mid-September to mid-October, the growth activity in all beeches increased.The oaks showed rapid growth in 2003 between themiddle of May until late July, when all trees sloweddown and almost no growth occurred until the end ofthe growing period (Fig. 6).

On the basis of our findings, the intra-annual growth ofoak and beech in the dry year 2003 can therefore besummarized as follows: at the start of the year, bothspecies started growth, but in the dry summer period(from the end of June to the end of August) woodformation stagnated. While oak had 80% of its RWalready completed at the beginning of July, beech grew forlonger, completing 80% of its RW by the middle ofSeptember (Fig. 6). In the year 2004, the oaks started togrow at approximately the same time as in 2003: betweenthe end of April and the middle of May. In general, theradial growth of oak in the beginning of the 2004 growingseason was less rapid than in 2003, as indicated by aslower increase in the Gompertz function (Fig. 6). As aconsequence, the mean RW of 2004 was slightly smaller(1.24mm) compared with 2003 (1.45mm), even thoughthe oaks continued growing for much longer in 2004 thanin 2003 and completed 80% of their RW in the middle ofAugust, 1 month later than in 2003.

Discussion

The high growth rate of both species, especially ofbeech, indicates the growing conditions on the dry sandy

ARTICLE IN PRESS

Fig. 5. Climate–growth relationships: results of the Bootstrap correlation analysis for beech (a) and oak (b). Bars ¼ rainfall,

line ¼ temperature, *significant correlations, po0.05) derived from the program DendroClim2002 (Biondi and Waikul, 2004).

G.W. van der Werf et al. / Dendrochronologia 25 (2007) 103–112108

site in Oostereng are good (Table 1). The very similargrowth pattern of the two species suggests that thegrowth of both species is influenced by the sameenvironmental factors. The climate–growth relation-ships (Fig. 5) show that precipitation plays an importantrole for the growth of both species. A significantnegative influence of temperature is evident: for bothspecies, the temperature during the summer months ofthe previous growing season is influential, but for beech,the temperature in June of the current year is influential.The correlation with precipitation is generally positive,as expected given the dry growing conditions on thesandy soil. Interesting differences occur between bothspecies in precipitation response: whereas for the growthof oaks above-average precipitation in winter and earlyspring, i.e. before the actual growing season, is crucial,

beech growth is in addition positively influenced byabove-average rainfall in combination with below-average temperature during the actual growing season,mainly in early summer. These results are in accordancewith those of other studies on oak (Rigling et al., 2001;Garcia-Gonzalez and Eckstein, 2003) and beech (Sassand Eckstein, 1995; Bouriaud et al., 2003; Dittmar et al.,2003; Lebourgeois et al., 2005) in different parts ofEurope.

Various conclusions can be drawn from the intra-annual growth in 2003 – a year with an exceptionally drysummer – in relation to differences between oak andbeech in their response to the summer drought and tothe results of the climate–growth analysis based on time-series analysis. The first conclusion is that the averagetree-RW in 2003 and 2004 was not unusually small in

ARTICLE IN PRESS

Fig. 6. Comparison of intra-annual wood formation of beech (open circles) and oak (solid circles) in 2003, means and fitted

Gompertz functions for beech (solid line) and oak (dashed line) (a). Comparison of intra-annual wood formation of oak in 2003

(open circles) and 2004 (solid circles); Gompertz functions fitted for oak in 2003 (solid line) and 2004 (dashed line) (b).

G.W. van der Werf et al. / Dendrochronologia 25 (2007) 103–112 109

the oaks and beeches studied, compared with otheryears of extremely low rainfall (e.g., 1959, 1976, and1996). This is attributable to the favorable growthconditions in 2002 and the favorable weather conditionsin autumn and winter 2002 and in early summer 2003.However, the oak and beech responded to the summerdrought differently in their intra-annual wood formation.

Both the beech trees in the sample showed the sameintra-annual pattern. Thanks to good starting condi-tions, both species, but especially beech, showed a rapidincrease in cambial activity until the beginning of July,when cambial activity diminished slowly in oak andprobably even ceased in beech (Fig. 6). Whereas thecambial activity of oak did not recover after the

ARTICLE IN PRESSG.W. van der Werf et al. / Dendrochronologia 25 (2007) 103–112110

drought, the cambial activity of beech seems to haverestarted in September and continued until the begin-ning of October. The implication is that beech on theOostereng site reacts very sensitively and immediately tochanges in water availability and records them in woodformation. In this context, it is important to mentionthat beech has a high growth rate on this site (Table 1).The high activity and sensitivity of the cambiumthroughout the growing season ensures that the drought‘‘signal’’ can be captured and reactivation is possible,even late in the growing season. This agrees with theresults reported by Schmitt et al. (2000); they found thatbeech was able to compensate for a slow start by thegrowth in the rest of the season. The immediate responseof beech to ambient weather conditions is clear from theresults of the climate–growth analysis (Fig. 5a).

Oak, however, reacts in a different way: as aconsequence of the dry conditions, cambial activityslows down, stops, and does not resume after thedrought. The finding is not unexpected, given that inring-porous trees the vessel formation, which startsbefore bud break, has first priority. The latter strategyensures both that there is a high capacity to transportwater in the outermost tree ring and that there is highphysiological activity throughout the growing season.The cambium of ring-porous species is very sensitive,especially during reactivation at the start of the growingseason, due to the accumulation of auxin precursors andcytokinin originating from the roots (Aloni, 1995).Good starting conditions lead to the formation of awide tree ring, even if conditions later in the growingseason are unfavorable, as was the case in 2003 (Fig. 6).Unfavorable starting conditions occur, however, if theconditions during the previous growing season limitedgrowth: the summer drought in 2003 resulted in a smallamount of earlywood (Fig. 3) and a lower growth rate in2004 – even though weather conditions in 2004 werefavorable (Fig. 6). Good starting conditions are mainlydefined by a high amount of available reserves stored atthe end of the previous growing season (Wareing, 1951;Barbaroux and Breda, 2002; Skomarkova et al., 2006).The high negative correlation of the RW of oak with thetemperature in the previous summer months (Fig. 5b)indicates that a high evapotranspiration rate mightreduce the amount of available reserves. Moreover, thestarting conditions are positively influenced by the highsoil moisture content that result from high precipitationin winter and early spring (Fig. 5b).

Conclusions

Differences between oak and beech were found inintra-annual wood formation in response to the summerdrought in 2003. They are attributable to species-specific

strategies to ensure a sustainable water-conductingsystem.

When the RW of the 2003 tree ring is compared withthat of other dry years (1959, 1976, 1996), it can beconcluded that neither species formed an unusuallysmall ring in 2003. Moreover, oak formed a fairly widering in 2004 too. Hence it can be concluded that bothspecies were not greatly affected by the extreme summerdrought in 2003 in terms of radial growth. Theimplication is that summer drought, i.e. a precipitationdeficit in the summer months, did not seriously affect thediameter growth of the oaks and beeches in Oostereng.Yet one Europe-wide study has reported a largereduction in primary productivity in the summer of2003 (Ciais et al., 2005). In that study, it is pointed outthat an increase in drought frequency and severity islikely to have particularly large effects on beech oftemperate climates. This might apply to oak and beechgrowing under extremely dry conditions (e.g., in theMediterranean) or to beech at high altitudes, assuggested by Dittmar et al. (2003). However, otherstudies that have examined the impact of the dry year2003 on various tree species in Central Europe found nodramatic changes in sap flow and photosyntheticcapacity in oak and beech (Leuzinger et al., 2005).Kahle’s (2006) results on radial growth of beech fromSouth Germany also indicate that although growth wasreduced at low and medium altitude, beech easilyrecovered the year after a severe summer drought. Ourfindings suggest that summer drought might producefew problems for oak and beech in the temperatemaritime climate in the Netherlands.

Acknowledgments

We are grateful to the Treeline Ecology ResearchUnit at Padova University, Italy for hospitality and forsharing their technical expertise. Special thanks go toAnnie Deslauriers for technical advice and fruitfuldiscussions. Ellen Wilderink and Leo Goudzwaardassisted in fieldwork and laboratory analysis. We alsothank two anonymous referees and the editors for theirsuggestions and comments for improving our manu-script. We thank Joy Burrough for editing the manu-script. The research is being funded by the NetherlandsOrganisation of Scientific Research (NWO/AWL836.05.030) and the European Union (ATEAMEVK2-CT-2000-00075).

References

Aloni, R., 1995. The induction of vascular tissue by auxin and

cytokinin. In: Raghavendra, A.D. (Ed.), Physiology of

Trees. Wiley, New York, pp. 175–197.

ARTICLE IN PRESSG.W. van der Werf et al. / Dendrochronologia 25 (2007) 103–112 111

Barbaroux, C., Breda, N., 2002. Contrasting distribution and

seasonal dynamics of carbohydrate reserves in stem wood

of adult ring-porous sessile oak and diffuse porous beech

trees. Tree Physiology 22, 1201–1210.

Beersma, J.J., Buishand, T.A., 2004. Joint probability of

precipitation and discharge deficits in the Netherlands.

Water Resource Research 40, 1–11.

Biondi, F., Waikul, K., 2004. DENDROCLIM2002: A

C++ program for statistical calibration of climate signals

in tree-ring chronologies. Computers & Geosciences 30,

303–311.

Bouriaud, O., Breda, N., Le Moguedec, G., Nepveu, G., 2003.

Modeling variability of wood density in beech as

affected by ring age, radial growth and climate. Trees 18,

264–276.

Camarero, J.J., Guerrero-Campo, J., Gutierrez, E., 1998. Tree-

ring growth and structure of Pinus uncinata and Pinus

sylvestris in the central Spanish Pyrenees. Artic and Alpine

Research 30, 1–10.

Ciais, Ph., Reichstein, M., Viovy, N., Granier, A., Oglee, J.,

Allard, V., Aubignet, M., Buchmann, N., Bernhofer, Chr.,

Carrara, A., Chevallier, F., De Noblet, N., Friend, A.D.,

Friedlingstein, P., Grunwald, T., Heinesch, B., Keronen, P.,

Knohl, A., Krinner, G., Loustau, D., Manca, G., Mat-

teucci, G., Miglietta, F., Ourcival, J.M., Papale, D.,

Pilegaard, K., Rambal, S., Seufert, G., Soussana,

J.F., Sanz, M.J., Schulze, E.D., Vesala, T., Valentini, R.,

2005. European-wide reduction in primary productivity

caused by the heat and drought in 2003. Nature 437,

529–533.

Cook, E.R., 1985. A time-series analysis approach to

tree-ring standardisation. Dissertation. University of

Arizona.

Cook, E.R., Holmes, R.L., 1997. ARSTAN: chrono-

logy development. In: Grissino-Mayer, H.D. (Ed.), The

International Tree-Ring Data Bank Program Library,

Version 21, User’s Manual. University of Arizona,

Laboratory of Tree-Ring Research, Tucson, AZ, USA,

pp. 75–92.

Den Ouden, J., Mohren, G.M.J., 2004. Waldokosysteme

unter gesellschaftlichem Druck. Forst und Holz 59,

376–381.

Deslauriers, A., Morin, H., 2005. Intra-annual tracheid

production in balsam fir stems and the effect of meteoro-

logical variables. Trees 19, 402–408.

Deslauriers, A., Morin, H., Begin, Y., 2003. Cellular pheno-

logy of annual ring formation of Abies balsamea in the

Quebec boreal forest (Canada). Canadian Journal of Forest

Research 33, 190–200.

Dittmar, C., Zech, W., Elling, W., 2003. Growth variations of

common beech (Fagus sylvatica L.) under different climatic

and environmental conditions in Europe – a dendrochro-

nological study. Forest Ecology and Management 173,

63–78.

Eckstein, D., Schmidt, B., 1974. Dendroklimatologische

Untersuchungen an Stieleichen aus dem maritimen

Klimagebiet Schleswig-Holsteins. Angewandte Botanik 48,

371–383.

Eilmann, B., Weber, P., Rigling, A., Eckstein, D., 2006.

Growth reactions of Pinus sylvestris L. and Quercus

pubescens Willd. to drought years at a xeric site in Valais,

Switzerland. Dendrochronologia 23, 121–132.

Fonti, P., Garcia-Gonzales, I., 2004. Suitability of chestnut

earlywood vessel chronologies for ecological studies. New

Phytologist 163, 77–86.

Garcia-Gonzalez, I.G., Eckstein, D., 2003. Climatic signal of

earlywood vessels of oak on a maritime site. Tree

Physiology 23, 497–504.

Holmes, R.L., 1983. Computer-assisted quality control in

tree-ring dating and measurement. Tree-ring Bulletin 43,

69–78.

Kahle, H.P., 2006. Impact of the drought in 2003 on intra- and

inter-annual stem radial growth of beech and spruce along

an altitudinal gradient in the Black Forest, Germany. In:

Gaertner, H., Esper, J., Schleser, G. (Eds.), TRACE-Tree

Rings in Archaeology, Climatology and Ecology, vol. 4.

Proceedings of the DENDROSYMPOSIUM 2005, April

21–23, 2005, Fribourg, Switzerland. Julich: Forschungszen-

trum Julich, 2006. (Schriften des Forschungszentrums

Julich-Reihe Umwelt; 61, pp. 151–163.

Lebourgeois, F., Breda, N., Ulrich, E., Granier, A., 2005.

Climate–tree-growth relationships of European beech

(Fagus sylvatica L.) in the French Permanent Plot Network

(RENECOFOR). Trees 19, 385–401.

Leuzinger, S., Zotz, G., Asshoff, R., Koerner, C., 2005.

Responses of deciduous forest trees to severe drought in

Central Europe. Tree Physiology 25, 641–650.

Metzger, M.J., 2005. European vulnerability to global change:

a spatially explicit and quantitative assesment. Dissertation,

Wageningen University.

Nobuchi, T., Ogata, Y., Siripartanadilok, S., 1995. Seasonal

characteristics of wood formation in Hapea odorata and

Shorea henryyana. IAWA Journal 16, 361–369.

Nola, P., 1996. Climatic signal in earlywood and latewood of

deciduous oaks from northern Italy. In: Dean, J.S., Meko,

D.M., Swetnam, T.W. (Eds.), Tree Rings, Environment,

and Humanity. Radiocarbon, pp. 149–158.

Pilcher, J.R., Gray, B., 1982. The relationships between oak

tree growth and climate in Britain. Journal of Ecology 70,

297–304.

Rinn, F., 1996. TSAP Reference Manual. Frank Rinn,

Heidelberg, Germany.

Rigling, A., Waldner, P.O., Forster, T., Braker, O.U., Pouttu,

A., 2001. Ecological interpretation of tree-ring width and

intraannual density fluctuations in Pinus sylvestris on dry

sites in the central Alps and Siberia. Canadian Journal of

Forest Research 31, 18–31.

Rossi, S., Deslauriers, A., Morin, H., 2003. Application of the

Gompertz equation for the study of xylem development.

Dendrochronologia 21, 33–39.

Sass, U., Eckstein, D., 1995. The variability of vessel size in

beech (Fagus sylvatica L.) and its ecophysiological inter-

pretation. Trees 9, 247–252.

Schmitt, U., Moller, R., Eckstein, D., 2000. Seasonal wood

formation dynamics of beech (Fagus sylvatica L.) and

black locust (Robinia pseudoacacia L.) as determined

by the ‘‘pinning’’ technique. Journal of Applied Botany

74, 10–16.

Skomarkova, M.V., Vaganov, E.A., Mund, M., Knol, A.,

Linke, P., Boerner, A., Schulze, E.-D., 2006. Inter-annual

ARTICLE IN PRESSG.W. van der Werf et al. / Dendrochronologia 25 (2007) 103–112112

and seasonal variability of radial growth, wood density

and carbon isotope ratios in tree rings of beech (Fagus

sylvatica) growing in Germany and Italy. Trees 20, 571–586.

Wareing, P.F., 1951. Growth studies in woody species. IV. The

initiation of cambial activity in ring-porous species.

Physiologia Plantarum 4, 546–562.

Wolter, E.K., 1968. A new method for marking xylem growth.

Forest Science 14, 10–103.

Zinyowera, R.T., Moss, M.C., Watson, R.H., 2001. IPCC

Special Report on the Regional Impacts of Climate

Change: an Assessment of Vulnerability. IPCC secretariat,

Geneva.