Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2019 Jan 15.
Published in final edited form as: Palaeogeogr Palaeoclimatol Palaeoecol. 2018 Oct 12;514:65–76. doi: 10.1016/j.palaeo.2018.10.003

An 1800-year stable carbon isotope chronology based on sub-fossil wood from Lake Schwarzensee, Austria

Marzena Kłusek a,*, Michael Grabner b, Sławomira Pawełczyk c, Jacek Pawlyta c
PMCID: PMC6287736  EMSID: EMS80146  PMID: 30546159

1. Introduction

This paper presents a new multi-century long stable carbon isotope chronology derived from sub-fossil tree trunks from Lake Schwarzensee in Austria. The new chronology has been created to reconstruct the past climate conditions in the Eastern Alps. This paper is the latest step in the palaeoclimate studies conducted by utilising the sub-fossil tree trunks from Lake Schwarzensee. The previous stages of the work focused on the analysis of the ring width and maximum latewood density (Kłusek et al., 2015). The new stable carbon isotope chronology should allow comparing, verifying, and expanding the results obtained so far.

The stable carbon isotope ratio measured in tree-rings provides an exceptional archive for reconstructing past environments. In comparison to the annual ring-width (TRW) or the maximum latewood density (MXD), stable carbon isotopes demonstrate some advantages (Gagen et al., 2011). First of all, the correlation between the isotopic ratio and climatic parameters is usually strong because this interrelationship results mainly from a direct response of a tree to a particular change in weather conditions. Therefore, unlike the classical dendrochronological variables, namely the tree-ring width and the maximum latewood density, stable isotope ratios are not only determined by tree growth limiting factors but can record additional climate information (McCarroll and Loader, 2004). In comparison with TRW and MXD, the stable carbon isotopes can be used to access a more synoptic view of the past climate and also to achieve further information on the tree physiology (Gagen et al., 2011). In contrast to this, in the case of the ring width, the variation is a general reflection of the changes in the net photosynthesis rate rather than a straight reaction to a particular climatic factor (Gagen et al., 2011). However, trees use their assimilation products in many ways, not only for the tree ring formation but also for producing leaves, shoots, roots, flowers, and seeds. As the effect, TRW responds much more variably to the temperature, precipitation, or radiation impact. Moreover, local edaphic and microclimate conditions also significantly influence the tree-ring width. Hence, a large number of trees are required to obtain the sufficiently strong common climatic signal. In turn, the stable carbon isotope composition of wood is less susceptible to between-tree individual changes and local habitat conditions than other tree-ring proxies (McCarroll and Loader, 2006). The specimen variability is, in the case of isotopic sequences, much weaker, and between-tree correlation is higher, in comparison to the wood density or the ring width. The correlations between trees from one site are usually very strong, and the signal-to-noise ratio is high. Therefore, it is not necessary to use many samples to ensure the representative average series for a stand. For this reason, the stable carbon isotopes provide a stronger climate signal than traditional proxies, particularly where the sample replication is low (Gagen et al., 2004).

Apart from the benefits that result from the application of the stable isotopes, also the chronology itself has numerous favourable features that predispose it to the dendroclimatic research. The first of these factors is the small territorial range of the area where the trees had grown. The sampled trees had originated from the mountain slopes surrounding the Lake Schwarzensee (Grabner et al., 2006). This fact secures the climatic homogeneity of the site and also ensures the small variability in the sensitivity and climate-growth relationship across the trees. Moreover, although the sampling was done at a quite low elevation, the site was located almost at the timberline. This territory is characterised by cold climate conditions and the high snow cover. It predestines the scale for the reconstruction of the temperatures. In the Alpine area, the carbon isotope research demonstrates the high correlation with the temperature, independently from the altitude of the sites (Treydte et al., 2001). Another advantage is the fact that the long-term meteorological record of this area dates back to the 18th century (Böhm et al., 2001). The access to the meteorological data encompassing the period of more than two hundred years increases the precision of the calibration and verification carried out during the dendroclimatic research. The results of the stable carbon isotope measurements presented in this article deliver new data that also have a significant impact on the quality of the palaeoclimate studies.

The combination of several physical and chemical wood parameters improves the climatic reconstructions significantly. Simultaneous analyses of the annual growth-ring width, the density, and the isotopic composition offer the possibility of a much more precise determination of the past weather conditions. They also allow obtaining stronger correlations for a higher number of climatic factors and extending the reconstruction for a more expanded period in a year (McCarroll et al., 2003). Conducting the research on the basis of multiple wood proxies is also necessary for a better understanding of a complex plant physiological response to an environmental change. A combination of all tree-ring variables might help to distinguish between climatic and non-climatic elements and might enable the proper deconvolution of the climate signal (Helle and Schleser, 2004a).

2. Materials and methods

2.1. Wood origin and sample selection

The stable carbon isotope ratios were measured on a multi-century tree-ring chronology. The sub-fossil wood excavated from a small mountain lake called Schwarzensee, and the samples obtained from living trees growing around the lake were used for the development of this chronology (Grabner et al., 2006; Kłusek and Grabner, 2016). Lake Schwarzensee (47°31′N, 13°49′E, 1450 m.a.s.l.) lies in the region of the Dachstein Mountains in the Limestone Alps.

An original master ring-width chronology was constructed with larch (Larix decidua Mill.) and spruce (Picea abies (L.) Karst.) wood from the Schwarzensee area (Grabner et al., 2006). However, different species, even the ones growing in the same locality, may not be influenced in the same way by the weather conditions. Some trees may be more sensitive to specific meteorological parameters than others, and the physiological responses may vary between different taxa of plants (McCarroll and Loader, 2006). Moreover, various rooting depths may result in diverse reactions of the trees to the long-term climatic changes (Saurer et al., 2003; Marshall and Monserud, 2006; Tsuji et al., 2006). Therefore, only spruce wood was selected for the isotopic research, in order to obtain a more uniform climatic signal. Frequently missing or extremely narrow growth-rings also caused the elimination of the larch wood. These specific wood features made obtaining a proper amount of larch wood per tree-ring too complicated.

The reaction wood and juvenile wood were also rejected during the selection of the material because the proportions of the stable carbon isotopes could be disturbed within this abnormally developed tissue (Gagen et al., 2008; Young et al., 2011). The zone of at least 30–50 growth-rings around the pith was excluded in order to remove the wood burdened with the age-related trends. It is a standard practice applied during dendroisotope analysis (Gagen et al., 2007). On the other hand, some recent studies indicate that the tree-ring δ13C series may contain the age-related trends throughout the whole lifespan of a tree. This age-dependence corresponds to an ascending linear change of ca. 0.4‰ per 100 years (Helama et al., 2015). Therefore, the removal of the juvenile part of the samples could be not sufficient for deleting these tendencies, and the presence of the age-related trends could still bias the Schwarzensee chronology.

During the selection of the research material, the samples of the sub-fossil wood with poorly preserved structure were also omitted because the stable isotope composition could be altered within them. It is the result of the partial decay taking place during the sedimentation of the wood (Loader et al., 2003; Sass-Klaassen et al., 2005; Harlow et al., 2006). In the sub-fossil samples, the mutual quantitative ratio of cell-wall components is disturbed because the disintegration of cellulose is faster than the lignin degradation. At the same time, the state of the cellulose preservation depends on the history of the particular wood. Therefore, the cellulose content could differ considerably among the specimens, even if they are equal in age (Schleser et al., 1999a; Loader et al., 2003; Sass-Klaassen et al., 2005). For this reason, the exclusion of partially decomposed material is advisable especially in the case of the samples averaged from several trees as it was done in presented research.

The samples combined from a few trees were used for the stable isotope measurements. The pooling of the samples is a common practice which allows the elimination of individual variability observed between the particular trees (Dorado Liñán et al., 2011). The differences in the isotopic composition occur mainly between different taxa but are also evident among individuals of the same species. They exist in a single tree, depending on its age, height, the organ, and the location within the trunk where the wood was obtained (Leavitt and Long, 1986; Buchmann and Ehleringer, 1998). Therefore, the averaging procedure helps to ensure the proper representativeness and the appropriate amount of the wood. It also decreases study costs and time (McCarroll and Loader, 2004).

At the beginning of the chronology, as an effect of a small number of the adequately preserved trunks, only two or three trees were available for the isotopic analysis. The oldest samples were usually significantly more degraded because the decomposition processes taking place on the bottom of the lake increased with the sedimentation time (Kłusek and Pawełczyk, 2014). In the later period, the samples were pooled from the growth-rings of four different trees (Fig. 1).

Fig. 1.

Fig. 1

Open in a new tab

The upper part of the figure shows Schwarzensee stable carbon isotope chronology and this chronology smoothed using spline fit with 50% variance cutoff at wavelength of 50 years. An inadequate sample replication (below 4 trees) which occurs in years 200–799 CE is marked with yellow colour. The lower part of the figure presents the sample depth of Schwarzensee chronology. Each from the four horizontal bars is divided into segments that correspond to individual trees. One tree is the same shade and the next tree is darker or lighter. At the beginning of the chronology only two or three trees were available for isotopic analysis in a particular year (two horizontal bars during the years 200–599 CE and three horizontal bars during the years 600–799 CE). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

The results of some studies suggest that the pooled sample should comprise more trees because a larger series replication is required to yield a representative mean signal. Small sample depth could bias the low-frequency variability in the chronology (Loader et al., 2013). Therefore, it is not precluded that the sample size may influence the long-term trends in the case of the Schwarzensee chronology, especially in the first part. However, it was also demonstrated that the averaging of the cores from four or five trees is sufficient because, in the case of the isotopic sequences, the variability occurring among the specimens is much lower in comparison to the differences observed for the growth-ring density or width (Saurer et al., 1995; Treydte et al., 2001; Dorado Liñán et al., 2011). Combining the rings from four or five trees before the analysis provides a representative expressed population signal EPS >0.85 (Leavitt and Long, 1984; Leavitt, 2010).

For the stable carbon isotope measurements, α-cellulose extracted from the whole annual growth-ring was applied. The selection between lignin and cellulose is a crucial issue during the stable isotope research because each cell wall component has a specific isotopic signature, and cellulose is significantly enriched in 13C in comparison to lignin (Leavitt and Long, 1982). Moreover, the ratio of lignin to cellulose changes within one plant – in the growth ring formed in the same year and also in the following years, as well as between particular trees (Porte and Loustau, 2001; Helle and Schleser, 2004a). Therefore, in order to eliminate the potential errors resulting from the variable proportion of lignin and cellulose within wood and to omit the troubles which could arise from their selective decomposition in the case of sub-fossil samples (Schleser et al., 1999a; Sass-Klaassen et al., 2005; Harlow et al., 2006), it was decided to use α-cellulose during the stable carbon isotope measurements.

Another problem that must be taken into account during the selection of the research material is the choice between analysing the whole growth-ring and the latewood only. The standard practice during stable isotope measurements is to use the latewood zone because the analysis of the whole growth-ring is more predisposed to the bias caused by the utilisation of reserve materials by plants (Leavitt and Long, 1982, 1991; Barbour et al., 2002; Jäggi et al., 2002; Keel et al., 2007). On the other hand, some studies proved that the application of the whole growth-ring offers a better chance of registering the weather conditions during the entire vegetation period and can record the climatic signal more comprehensively (Schleser et al., 1999b; Helle and Schleser, 2004a, 2004b). Moreover, recent analyses demonstrated that the utilisation of the stored substances for the conifer wood formation at high-mountain areas is very low and has a negligible influence on the results obtained from the stable carbon isotope research. This situation is caused by high turnover rates and small reserve pools at tree-line sites (Kress et al., 2009). Therefore, taking into account the fact that Lake Schwarzensee is located almost at an elevation of the regional timberline, it was decided to apply the whole growth-ring for the stable carbon isotope research. This choice was also motivated by the small width of growth-rings and the blurred borders between early- and latewood that did not allow the precise differentiation between these zones of wood.

2.2. Sample preparation

The samples from the sub-fossil wood and the wood extracted from living trees were selected for the stable carbon isotope measurements. In total, 51 sub-fossil tree trunks and samples from five living trees were used. Within each wood fragment, the width of growth-ring was measured to assign the tree-ring to the proper calendar year. This assignment was done on the basis of the comparison with the existing master chronology. The tree-rings were measured to the nearest 0.01 mm using the LINTAB device and TSAP Win computer program (Rinn, 2003). Next, under the magnification of a binocular microscope, every single growth-ring was split using a scalpel and then cut into thin slivers to increase the active surface of the wood that gets into contact with chemical reagents. The material obtained from a few trees was pooled for a particular year. The equal proportions by weight were maintained, and 25 mg were usually weighed out from each growth-ring. This procedure was performed although the previous research demonstrated that the bias from differing mass contributions of the raw wood towards a pool appears to be negligible (Borella et al., 1998; Leavitt, 2008). The obtained samples were chemically processed to extract α-cellulose.

The procedure of α-cellulose isolation was conducted in accordance with the modified Green (1963) method and by using an ultrasonic bath to promote a rapid and more complete penetration of the reagents into the wood tissue and to assists the disaggregation and homogenisation of the individual cellulose fibres (Loader et al., 1997). During the first phase of extraction, the water solution of acetic acid and sodium chlorite was added into each sample. The proper amounts of the reagents were measured depending on the weight of the sample. For 1 g of the sample, 175 ml of distilled water, 2.5 g of sodium chlorite, and 1.7 ml of 80% acetic acid were used. Next, the samples were placed in glass test tubes and inserted into the ultrasonic bath with the temperature of 70 °C. Then, acetic acid and sodium chlorite were added seven times, at hourly intervals. The amounts of reagents were the same as during the first step. Later, the samples were rinsed with boiling deionised water several times, until the neutral pH was obtained. This chemical pretreatment removed the majority of lignin from the sample.

During the second phase of the extraction, the sodium hydroxide solution was applied for α-cellulose isolation. In this stage of the chemical pre-treatment, the remains of lignin and hemicellulose were removed from the wood. First, 10% solution of sodium hydroxide was added to each test tube, in the quantity depending on the weight of the sample. In total, 75 ml of 10% sodium hydroxide was used for 1 g of the sample. Then, the tubes were again placed in an ultrasonic bath with the temperature of 70 °C for 45 min. Afterwards, the samples were rinsed in boiling deionised water and later, they were subjected to 45-minute reaction with 17% sodium hydroxide in the ultrasonic bath of room temperature. During this process, 67 ml of 17% sodium hydroxide were used for 1 g of the sample. Afterwards, the samples were washed in boiling deionised water until neutral pH was reached, rinsed with 1% hydrochloric acid and again washed in boiling deionised water until the neutral solution was obtained. Subsequently, the samples were ground in a mortar to separate the fibres better and to homogenise them. Finally, they were dried at the temperature of 70 °C. On average, about 40 mg of α-cellulose were obtained from 100 mg of wood.

2.3. Stable isotope measurement

Two or three sub-samples of α-cellulose, representing a particular year, were prepared for the stable carbon isotope measurements. They were precisely weighted (ca. 55 μg) into tin foil capsules, crimped, and placed into a sample tray. Next, the samples were combusted “on-line” at 1020 °C, using the elemental analyser EuroVector. In this way, the analysed wood was converted into a gaseous form of carbon dioxide. The isotopic composition of the samples was measured by the application of continuous flow Isotope Ratio Mass Spectrometer IsoPrime EA-CF-IRMS. During the measurements, the mass spectrometer was directly connected to the elemental analyser. The precision of this method is 0.1‰.

The carbon isotope measurements were expressed in the isotope delta (δ13C) notation. IAEA cellulose C3 and IAEA Two Creeks wood C5 international standards, and laboratory standard Fluka acid washed cellulose were used as the reference materials, to calibrate the sample isotopic composition versus the international Vienna Pee Dee Belemnite (VPDB). The values of δ13C were expressed in units of part per thousand and communicated in per mil shown as ‰ (Brand and Coplen, 2012).

2.4. Chronology construction

The results of the stable carbon isotope measurements obtained for the sub-samples representing a particular calendar year were averaged, and these values served directly for the raw chronology construction. The only exception was the final part of the chronology. For the period of 1850–2000 CE, an additional correction was introduced because of the anthropogenic increase of the atmospheric CO2 concentration and related lowering of δ13C of air. As the consequence of the human activity, tree-ring δ13C values exhibit a declining trend which is not related to the climate changes. This anthropogenic decline is caused by the input to the atmosphere of 13C-depleted CO2, primarily released by the burning of fossil fuels, and is also related to the land-use change. Therefore, the raw δ13C measurements were corrected to a pre-industrial atmospheric δ13C base value of −6.4‰ using simple addition. It was done on the basis of the published annual records of the stable carbon isotope ratios of atmospheric carbon dioxide (McCarroll et al., 2009).

The exposure of trees to the increasing level of atmospheric CO2 also leads to the changes in the internal CO2 concentrations of the needles. This situation results in the adaptations of the stomatal conductance, water-use-efficiency, and photosynthetic assimilation rate (Farquhar et al., 1982; Konter et al., 2014). The effect of elevated CO2 on the tree growth and physiology may involve the simultaneous reduction of the stomatal conductance and the stimulation of the photosynthesis rate and biomass production, as well as the increase in leaf internal discrimination against 13C, which leads to a decrease in tree-ring δ13C values. This isotope discrimination might be particularly strong for high-elevation tree habitats, as they represent lowered CO2 partial pressure environments. As a result, many tree-ring δ13C series, especially over the recent decades, have shown an additional decline for which there is no evidence of a climatic cause (Saurer and Siegwolf, 2007; Treydte et al., 2009). In order to compensate the change in the ecophysiological response to the rapidly rising concentrations of CO2, the second correction, known as the PreINdustrial or PIN correction, was applied to the fossil fuel corrected data in the period of 1820–2000 CE (McCarroll et al., 2009; Konter et al., 2014).

2.5. Correlations with climate parameters

The chronology developed from the stable carbon isotopes was compared with the meteorological records. For this purpose, monthly averaged data for temperature (1780–2000 CE) and the monthly sums of precipitation (1800–2000 CE) were tested. These time series were gridded for the area of Lake Schwarzensee by the application of data from the 1/6th of a degree grid box centred at 47°35′N, 13°45′E (Auer et al., 2005; Efthymiadis et al., 2006; Chimani et al., 2013). The measurements of solar radiation (1884–2000 CE) originated from the meteorological station of Kremsmünster (48°06′N, 14°13′E, 389 m.a.s.l.) (Auer et al., 2007). These climatic data were compared with the results of isotopic measurements by the application of the DENDROCLIM 2002 software (Biondi and Waikul, 2004). The correlation coefficient (r) and response function values were calculated for all time intervals from the previous August to October of the current year. In each case, only one instrumental parameter was included in the response function analysis.

Moreover, the years characterised by the extreme values of the stable carbon isotope content in the wood were selected in order to obtain additional information about the relationship between the climate and the stable carbon isotopes. These pointer years were specified by means of a 5-year running mean (RM) and the standard deviation (SD) values as well as the arbitrary established critical level of standard deviation (CL). As a threshold for defining extreme events, the single standard deviation of a 5-year period was accepted, so in this case, the critical level equalled to 1. If the central ring was smaller than (RM − (CL * SD)), then this year was assigned to be the negative pointer year. The positive pointer year was classified if the central ring was larger than (RM + (CL * SD)) (Cropper, 1979; Treydte et al., 2001). This technique had previously been used in the case of the stable isotope chronologies consisting of pooled trees (Treydte et al., 2001; Battipaglia et al., 2007). The obtained pointer years were compared with the climate data of the corresponding calendar years. For this purpose, the meteorological measurements of temperature, precipitation, and radiation were separated into three groups. The first of them consisted of the positive pointer years, the second group contained the negative ones, and the third group was composed of the remaining normal years. Then, the negative pointer years were compared with the remaining normal years and the positive pointer years, while the positive pointer years were compared with the remaining normal years and the negative pointer years. The Student t-test was applied to verify if the difference between the mean values of the meteorological data was significant for the paired groups of the pointer and all the remaining years.

3. Results

Our measurements resulted in the long stable carbon isotope chronology from Lake Schwarzensee. This chronology covers the period 200–2000 CE (Fig. 1) and is much shorter than the master ring-width chronology constructed originally for the Lake Schwarzensee. This chronology reaches back to 1475 BCE. The new stable carbon isotope chronology is reduced in length because only the properly preserved spruce samples were utilised in the measurements. For the same reason, in the period 200–799 CE, the sample depth is lower than four trees.

The stable carbon isotope chronology was constructed directly on the basis of the raw measurement data obtained from the mass spectrometer and calibrated versus the reference materials. The additional corrections were introduced only for the period of 1820–2000 CE, to compensate for the anthropogenic increase of the atmospheric CO2 concentration (Fig. 2). However, despite the application of these corrections, a δ13C decreasing tendency was observed for the years 1800–1950 CE. Moreover, a downward trend was also visible during the periods of 450–800 CE and 1450–1650 CE. In contrast to this, the increase in δ13C was noticeable for the time spans 800–1450 CE and 1650–1800 CE.

Fig. 2.

Fig. 2

Open in a new tab

Schwarzensee chronology constructed on the basis of raw stable carbon isotope measurements (black line), the chronology corrected to a pre-industrial atmospheric δ13C base value (red line), and PIN correction applied to the fossil fuel corrected data (brown line). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Unfortunately, some part of the long-term variation observed in the Schwarzensee chronology could be due to the sample depth. An inadequate sample replication occurred in years 200–799 CE. During this period, only two or three trees were used for the stable isotope analysis. The varying sample size (from 2 to 4 trees) and relatively low maximum sample replication (4 trees) could bias the low-frequency signal in the Schwarzensee chronology.

The potential for reconstructing past climate changes was determined using the correlation coefficients calculated between the δ13C values and the meteorological data from the area of Schwarzensee. A positive relationship was determined between δ13C and the temperature in August (r = 0.28), September (r = 0.19), and October (r = 0.16) of the previous year as well as February (r = 0.15), May (r = 0.21), June (r = 0.14), July (r = 0.34), August (r = 0.42), and September (r = 0.18) of the current year. In turn, negative correlations were obtained between δ13C and the monthly precipitation sum during the summer months of July (r = −0.13) and August (r = −0.15) of the current year, and a weak positive link was noticed with snow precipitation in February (r = 0.14) of the current year. The analysis of the solar radiation level demonstrated a significant positive association with October (r = 0.25) of the previous year and April (r = 0.24), July (r = 0.39), and August (r = 0.33) of the current year (Table 1).

Table 1.

Correlations with climate parameters. Bootstrap correlation and response function values were calculated in DENDROCLIM 2002 between mean monthly temperature (1780–1950 CE), precipitation (1800–1950 CE) and radiation (1884–1950 CE), and stable carbon isotope chronology. Significant values at the 0.05 level are marked in bold. The asterisks indicate months in the year previous to growth-ring formation.

Months Temperature Precipitation Radiation
Correlation coefficient Response function Correlation coefficient Response function Correlation coefficient Response function
August* 0.28 0.14 0.00 –0.01 0.02 0.00
September* 0.19 0.07 0.04 0.01 0.02 –0.02
October* 0.16 0.05 –0.14 –0.09 0.25 0.11
November* –0.01 –0.09 0.10 0.55 0.02 –0.01
December* 0.05 –0.01 0.09 0.09 0.05 0.06
January 0.08 0.06 –0.02 –0.04 0.18 0.09
February 0.15 0.09 0.14 0.12 –0.11 –0.13
March 0.07 0.00 0.04 0.06 0.02 0.01
April 0.09 –0.01 –0.02 0.00 0.24 0.16
May 0.21 0.11 –0.06 –0.05 0.17 0.18
June 0.14 0.08 –0.04 –0.05 0.18 0.14
July 0.34 0.17 –0.13 –0.13 0.39 0.26
August 0.42 0.24 –0.15 –0.11 0.33 0.19
September 0.18 0.06 –0.06 –0.03 –0.04 –0.12
October 0.12 0.04 –0.02 –0.01 –0.03 –0.01
Open in a new tab

These correlation coefficients between δ13C and the weather parameters were calculated excluding the period of 1950–2000 CE. This circumstance resulted from the fact that during the last decades the correlation coefficients were distinctly reduced and ceased to be significant (Fig. 3). The mutual interdependencies between δ13C and the climate variables do not disappear entirely, and the lack of correlation is connected with a low-frequency component (the increasing trend of chronology), and not with a year-to-year variability (Fig. 4). We hypothesise that the anthropogenic effects may have weakened these relationships.

Fig. 3.

Fig. 3

Open in a new tab

Moving interval correlation coefficients computed using DENDROCLIM 2002 software between mean monthly values of solar radiation and stable carbon isotope chronology. A base length of 25 years was progressively slid through the years from 1884 CE to 2000 CE. Figure presents significant values at the 0.05 level. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4.

Fig. 4

Open in a new tab

Comparison of normalised series of mean temperature (A), precipitation (B) and solar radiation (C) averaged for July and August (red lines) with normalised Schwarzensee stable carbon isotope chronology (black lines). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

In addition to the calculations of the correlation coefficients, the pointer year analysis was performed to describe the climate influence on the δ13C ratio in the wood. For the whole chronology range, 250 negative and 252 positive pointer years were obtained (Table 2). The negative and positive pointer years were compared with the remaining years. This comparison was carried out during the period common for the pointer years and the meteorological records. As a result of this analysis, a significantly higher temperature was seen for June, July, and August during the positive pointer years, and a significantly lower temperature was determined for July and August in the negative pointer years. A relationship with the precipitation sum was observed in February for the positive pointer years and in June and July for the negative pointer years. The sunshine values showed a significant link for October of the positive pointer years as well as for July of the negative pointer years (Table 3).

Table 2.

Negative and positive pointer years calculated for stable carbon isotope chronology. These pointer years were specified by means of a 5-year running mean (RM) and the standard deviation (SD) values as well as the arbitrary established critical level of standard deviation (CL). For more information, see Materials and methods.

Negative pointer years Positive pointer years
205 583 979 1417 1805 204 557 977 1360 1761
213 591 989 1420 1813 208 564 982 1367 1767
218 594 995 1435 1821 212 570 987 1380 1770
223 597 1002 1444 1828 216 575 993 1385 1774
228 603 1020 1448 1833 219 580 996 1394 1778
237 612 1029 1452 1838 224 587 1009 1408 1781
244 615 1036 1457 1844 233 592 1015 1415 1803
256 627 1056 1462 1851 236 599 1026 1419 1807
259 644 1062 1465 1860 242 611 1038 1423 1846
263 652 1065 1468 1879 247 616 1044 1436 1849
269 663 1068 1478 1883 254 619 1050 1447 1868
280 667 1072 1485 1888 258 624 1055 1451 1874
283 671 1076 1489 1898 267 626 1059 1458 1877
287 677 1080 1492 1906 291 637 1066 1473 1882
298 696 1088 1497 1910 294 640 1074 1476 1885
305 704 1095 1516 1913 306 643 1092 1479 1892
313 707 1105 1527 1920 309 650 1097 1487 1904
323 716 1112 1530 1924 312 656 1100 1494 1928
334 721 1121 1550 1933 316 668 1103 1498 1935
339 727 1130 1554 1940 320 675 1111 1503 1939
349 733 1133 1563 1948 328 683 1115 1507 1944
356 754 1136 1567 1951 331 687 1123 1513 1950
369 760 1147 1580 1955 336 694 1126 1517 1958
374 767 1152 1586 1967 342 701 1131 1528 1962
384 773 1158 1593 1985 348 712 1135 1531 1976
396 779 1164 1596 1999 355 719 1146 1535 1983
401 787 1171 1606 360 724 1149 1540 1991
402 794 1174 1609 367 728 1150 1545 1998
407 798 1179 1628 371 737 1153 1551
414 803 1196 1632 377 750 1161 1569
417 815 1204 1636 382 755 1173 1572
427 819 1207 1643 387 756 1176 1584
431 824 1224 1648 391 759 1180 1587
437 827 1227 1657 394 764 1184 1590
442 831 1230 1663 397 778 1188 1594
451 834 1237 1668 405 795 1200 1599
457 843 1242 1672 421 801 1203 1605
464 846 1256 1677 433 804 1206 1611
473 852 1261 1682 440 818 1212 1616
486 856 1269 1685 449 823 1217 1625
493 866 1286 1690 461 829 1221 1635
501 873 1290 1703 465 836 1228 1651
507 892 1299 1707 475 845 1244 1662
511 895 1303 1710 480 851 1254 1666
517 902 1310 1716 484 857 1262 1673
522 909 1314 1721 495 861 1265 1676
525 913 1324 1733 503 886 1273 1680
529 918 1328 1747 506 893 1279 1692
537 924 1335 1752 513 898 1288 1702
541 930 1340 1753 518 906 1301 1711
545 944 1350 1757 521 912 1309 1715
549 947 1364 1765 524 917 1312 1719
555 958 1392 1777 528 941 1326 1726
562 959 1399 1786 532 945 1332 1729
568 967 1404 1795 542 948 1337 1739
579 972 1407 1800 546 965 1346 1745
Open in a new tab

Table 3.

The negative and positive pointer years compared with all the remaining years. Mean monthly temperature, precipitation and radiation were averaged for negative and positive pointer years, and all remaining years. Student–t-tests were applied to verify significance level of differences between those groups of years. For calculations monthly averaged data for temperature (1780–1950 CE) and for solar radiation (1884–1950 CE), as well as monthly sums of precipitation (1800–1950 CE) were used. Significant values at the 0.05 level are marked in bold.

Months Mean temperatures and in parentheses standard deviations Student t-test Mean temperatures and in parentheses standard deviations Student t-test Mean precipitations and in parentheses standard deviations Student t-test Mean precipitations and in parentheses standard deviations Student t-test Mean radiations and in parentheses standard deviations Student t-test Mean radiations and in parentheses standard deviations Student t-test
Remaining years Negative Pointer Years Remaining years Positive Pointer Years Remaining years Negative Pointer Years Remaining years Positive Pointer Years Remaining years Nrgative Pointer Years Remaining years Positive Pointer Years
January –5.48 (2.44) –5.45 (2.60) –0.05 –5.54 (2.51) –4.94 (1.82) –0.95 103.64 (59.28) 102.68 (58.52) 0.07 104.56 (59.37) 94.50 (56.54) 0.64 45.35 (17.42) 39.20 (13.26) 1.06 44.81 (16.98) 41.63 (17.33) 0.50
February –4.08 (2.32) –4.68 (1.64) 0.54 –4.55 (2.26) –3.61 (1.87) –1.66 102.33 (62.82) 104.23 (57.05) –0.13 97.80 (56.56) 143.19 (87.75) –2.84 91.28 (31.65) 99.80 (44.72) –0.74 93.85 (34.17) 83.00 (29.60) 0.85
March –1.74 (2.03) –1.95 (1.66) 0.49 –1.71 (1.96) –2.31 (2.11) 1.18 111.53 (57.37) 110.05 (57.14) 0.11 111.01 (58.28) 113.81 (48.01) –0.18 131.70 (37.08) 145.60 (41.38) –1.08 135.39 (36.11) 121.88 (49.47) 0.95
April 2.22 (1.77) 2.43 (1.83) –0.54 2.20 (1.81) 2.71 (1.42) –1.13 116.22 (49.74) 116.73 (51.13) –0.04 116.28 (50.46) 116.44 (45.02) –0.01 151.75 (42.50) 146.50 (45.02) 0.36 151.17 (44.31) 149.50 (28.63) 0.10
May 7.20 (1.77) 7.10 (1.90) 0.28 7.23 (1.73) 6.84 (2.24) 0.85 127.88 (43.66) 140.36 (49.88) –1.21 130.33 (44.04) 124.38 (50.88) 0.50 221.51 (44.36) 211.90 (35.42) 0.65 221.81 (41.66) 207.25 (53.61) 0.90
June 10.26 (1.45) 9.85 (1.25) 1.30 10.10 (1.40) 11.14 (1.32) –2.93 194.43 (55.61) 219.73 (48.17) –2.01 198.26 (52.81) 196.88 (74.34) 0.09 234.35 (48.40) 209.00 (39.50) 1.56 227.54 (48.06) 252.88 (41.74) –1.42
July 12.37 (1.29) 10.65 (1.19) 6.11 12.04 (1.39) 12.97 (1.35) –2.64 225.33 (59.54) 263.45 (49.71) –2.84 233.38 (60.10) 209.81 (52.33) 1.50 231.54 (39.45) 195.70 (22.52) 2.78 222.85 (36.84) 250.88 (51.29) –1.92
August 11.98 (1.36) 11.03 (1.39) 3.15 11.76 (1.34) 12.67 (1.67) –2.60 211.82 (62.73) 227.14 (53.62) –1.08 215.39 (61.10) 202.81 (66.33) 0.77 243.93 (43.60) 228.70 (50.40) 1.00 238.02 (44.71) 268.50 (35.45) –1.85
September 8.99 (1.50) 8.69 (1.31) 0.93 8.96 (1.50) 8.79 (1.28) 0.47 153.45 (69.56) 175.41 (57.91) –1.40 156.79 (68.68) 155.50 (66.65) 0.07 181.79 (44.71) 173.10 (34.96) 0.58 183.10 (44.09) 161.25 (32.76) 1.35
October 4.67 (1.68) 4.69 (1.45) –0.06 4.69 (1.68) 4.49 (1.28) 0.46 114.64 (55.97) 97.18 (51.43) 1.37 110.79 (56.60) 123.06 (45.33) –0.84 106.02 (37.03) 116.50 (37.63) –0.82 111.56 (37.06) 78.25 (20.71) 2.48
November –0.58 (1.62) –0.60 (2.00) 0.08 –0.59 (1.71) –0.47 (1.36) –0.28 106.72 (53.94) 105.50 (75.50) 0.09 104.39 (57.70) 124.75 (51.67) –1.35 39.47 (12.87) 49.90 (9.69) –2.44 41.69 (13.35) 36.13 (8.46) 1.15
December –4.15 (2.23) –4.22 (2.72) 0.15 –4.12 (2.34) –4.48 (1.90) 0.60 120.43 (67.08) 105.27 (62.02) 0.99 115.64 (67.71) 140.00 (50.54) –1.39 35.95 (13.76) 38.40 (14.14) –0.52 36.49 (13.48) 35.00 (16.48) 0.29
Open in a new tab

4. Discussion

The strongest relationship between the Schwarzensee stable carbon isotope chronology and the weather parameters was found for the summer months (July and August). Notably, the temperature and solar radiation have a strong positive impact on the δ13C in the Schwarzensee wood. It is a typical situation observed in many Alpine stands (Treydte et al., 2001; Levanič et al., 2009; Kress et al., 2010, 2014; Hafner et al., 2011, 2014). However, the Schwarzensee chronology differs considerably from the other stable carbon isotope chronologies constructed for this area, especially within the time span of the last two hundred years. Therefore, the detailed analysis of the connection between the climate impact and the stable isotope content in the wood was conducted to explain the causes of this phenomenon.

In general, thermal and sunshine changes are the dominant controls influencing the stable carbon isotope content of the wood in the Alpine areas (Treydte et al., 2009). The increases of the solar radiation and temperature stimulate the rate of photosynthesis (Farquhar et al., 1982; Leal et al., 2007). Under these conditions, carbon dioxide diffusion into the leaf is insufficient in the relation to the assimilation level, and, as a consequence, the internal leaf concentration of CO2 is reduced. Therefore, photosynthetic enzymes must utilise more 13C, and the newly developed wood cells are enriched with this isotope. In turn, the diminished temperature and sunshine levels result in low photosynthesis intensity and a small amount of 13C in the organic matter formed under these conditions.

Thermal and solar factors determine the photosynthesis rate, but the temperature can also indirectly affect stomatal closure and opening. The high temperature increases the evapotranspiration processes. In such circumstances, trees have to counterbalance the enhanced transpiration by closing their stomata, in order to maintain their water regime. This physiological response leads to a decrease of the leaf internal partial pressure of CO2, and consequently to an enrichment of the wood in 13C. In contrast to this, low temperatures allow stomata to remain open and result in the wood depleted in 13C (Farquhar and Lloyd, 1993; McCarroll and Loader, 2006). Such a mechanism of stomatal conductance could share the responsibility for the observed stable isotope content in the case of the Schwarzensee wood.

However, the primary controlling factor that determines stomatal closure and opening is not the temperature but the air humidity. Unfortunately, the precipitation level, relative humidity and water supply to plants are not tantamount to each other. Therefore, it is difficult to conclude about the stomatal conductance on the basis of the precipitation data obtained from meteorological measurements. Nevertheless, in the case of the inadequate water supply, plants close their stomata, but when water resources are not limited, stomata remain open. Thus, the moisture stress is reflected in the stable carbon isotope content of the wood, especially in dry site conditions (McCarroll and Loader, 2004).

Although the water availability, in general, is adequate in the Schwarzensee area, δ13C correlates with the precipitation of February as well as July and August of the growing season. Most probably, the negative relationship observed between δ13C chronology and the monthly precipitation during July and August results from stomatal closure during the hottest summer period which could be caused, at least partially, by the reduced water supply to the plants. As suggested by the analysis of the pointer years, the link between δ13C and the precipitation during June and July was seen only in the case when the rainfall was significantly lower than for the other years. In contrast, the pointer years analysis implies that no relation was observed between significantly higher rainfall and δ13C during the summer months. Therefore, the precipitation significantly influences the stable carbon isotope ratio in the wood during noticeably drier seasons. On the other hand, the connection with a snowfall level was significant for February and was visible in the correlation coefficient and the pointer year analyses. The cause of this positive correlation is not apparent. It could possibly arise from the photosynthesis enhancement. Evergreen plants may potentially photosynthesise during the winter season. In this case, the low rate of assimilation is associated with low minimum temperature and moisture deficit, since the root system is unable to cope with larger evaporative demand (Pallardy, 2008). As a result, during the winter months, the assimilation level is significantly restricted by water availability. Thus, the delivery of water resources in the form of snow could stimulate the photosynthesis process and increase the δ13C of wood (Bowling et al., 2018). This hypothesis is confirmed by the positive correlation coefficient obtained for the February temperature, which indicates that during this month higher temperature directly raises the rate of photosynthesis, but also causes the melting of snow, and as a result of both these processes elevates the δ13C values in the wood.

These observations point to the precipitation as the other climatic factor that could contribute to the final stable carbon isotope ratio in the wood. The occurrence of the strongest relationship between δ13C and weather parameters in the hottest summer season, the period which is most susceptible to moisture deficit, supports the hypothesis that stomatal conductance can play a role in the isotope composition of the Schwarzensee wood. Moreover, it must be taken into account that spruce trees have a shallow root system and therefore are more exposed to water stress (Schweingruber, 1993).

Nevertheless, the observed relationship between δ13C and precipitation is rather weak and, as mentioned above, is also burdened with the additional uncertainty because the major climatic factor determining stomatal conductance is not rainwater but the relative humidity level (McCarroll and Loader, 2004). Unfortunately, there is no linear dependency between these two parameters in the Alpine region (Auer et al., 2007; Treydte et al., 2009). It means that the monthly sum of precipitation does not reflect the level of relative humidity during a particular month. During a mostly dry month, a few major precipitation events with rapid overland flow could result in the same total monthly precipitation as more evenly distributed rain in a generally humid month (Treydte et al., 2009).

Moreover, the obtained temperature signal is at least partially indirect because the temperature strongly and directly governs neither stomatal conductance nor photosynthetic rate. In general, the dominant control for stomatal conductance is the vapour pressure deficit, and the assimilation rate is mostly related to photosynthetically active radiation (McCarroll and Loader, 2004). Therefore, the link observed between the temperature and δ13C is the consequence of the relationships between thermal conditions and other controlling variables, such as irradiance, sunshine duration, or vapour pressure deficit (Treydte et al., 2009; McCarroll et al., 2011).

The temperature has a direct impact only on the dark phase of the photosynthesis when the reactions catalysed by enzymes take place irrespective of sunlight. The increase of the temperature stimulates these processes and enhances the production of the photosynthetic enzymes. On the other hand, the sunshine level determines the light phase of photosynthesis during which light-dependent reactions use the energy of the sunlight. This process is not affected by the changes in temperature (Pallardy, 2008). As a result, the photosynthesis is controlled most strongly and directly by the supply of energy in the form of photon flux from sun radiation (McCarroll and Loader, 2006; Hafner et al., 2014).

The correlation between δ13C and sunlight is relatively high for the Schwarzensee wood. However, this relationship is weaker than the temperature signal. It could be explained by more precise and accurate measurements of the temperature in comparison with the irradiance level, which is the rule in long meteorological measurements. Some care is also required because the temperature changes smoothly across space, whereas sunshine and precipitation are more spatially heterogeneous. It means that the inter-annual temperature change is likely more similar to δ13C record than the sunshine, even if the real influence of temperature and sunshine on a proxy data is equally strong. Therefore, the temperature usually gives higher correlation values due to the superior instrumental robustness and the spatial homogeneity of the temperature variable (Hafner et al., 2014). Additionally, the sunshine data used during this research originated from a more distant location in comparison with the temperature record which was extrapolated precisely for the Schwarzensee area. The measurements of solar radiation were obtained from the meteorological station of Kremsmünster which is located 96 km away, at a much lower altitude. Therefore, these data should be treated with caution because the insolation level measured for this station may considerably differ from the real situation in the Schwarzensee area. Consequently, sunlight may have a more significant impact on the stable carbon isotope contents in the wood than the obtained results suggested.

Taking into account all the factors mentioned above, it is impossible to distinguish unequivocally which signal – thermal or sunshine – is prevailing in the Schwarzensee chronology. However, the question which physiological mechanism – photosynthesis rate or stomatal conductance – decides about the final isotopic composition of the wood can be answered more unambiguously. The stronger high-frequency correlations to the summer temperature and solar radiation, and the weaker relationship with the precipitation level suggest that the leaf internal CO2 concentration is mainly controlled by the carboxylation efficiency of the photosynthetic active enzyme RuBisCo rather than by the variations in stomatal conductance. In the habitats where the assimilation rate is the dominant factor determining the stable carbon isotope content, solar radiation usually has a decisive influence. In principle, sunshine has the strongest control on carbon isotope fractionation under cool, moist conditions (McCarroll and Loader, 2004). Most probably, this is also the case of the Schwarzensee area. This area is characterised by long, cold winters, and short, cool to mild summers. Only three months of the year (June, July, August) have an average temperature above 10 °C, and five coldest months (November, December, January, February, March) have the average temperature below 0 °C. Simultaneously, the Schwarzensee area distinguishes itself by a high precipitation value due to the orographic lift. This region receives an average rain-equivalent of 1720 mm of precipitation per year. Rainfall occurs mostly in the warmer months (June, July, and August) (Fig. 5).

Fig. 5.

Fig. 5

Open in a new tab

Mean monthly temperature (bar graph) and precipitation (line graph) averaged over the years 1800–2000 CE for Schwarzensee area. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Presumably, in the area of Schwarzensee Lake, the increase of the temperature strongly stimulates the evaporation and water transpiration from plants. As an effect, there could be more cloud condensation, which would lead to the increased development of cloud cover. The elevated air humidity could promote the formation of fog as well. These processes might also be enhanced by the vicinity of the water reservoir – Lake Schwarzensee. The described weather conditions limit the amount of solar radiation reaching the surface. If we assume that sunshine has the primary impact on the stable carbon isotope ratio and the photosynthesis rate, then this situation manifests itself in the reduced concentration of 13C in the wood. The 13C content could be further diminished if stomata remained open. Both stomatal conductance and evaporation can remain intense because moisture resources are sufficiently large in this area. However, in this case, the mentioned physiological phenomena (weak assimilation and strong stomatal conductivity) take place despite the high-temperature level. This hypothesis points to solar radiation as the climatic variable which is responsible for the final ratio of the stable carbon isotopes in the Schwarzensee wood.

When assessing the impact of environment on the stable carbon isotope content in the wood, the comparison with the previous research, conducted on the basis of the maximum density measurements, may be helpful as well. The climatic response of MXD chronology, constructed for the same wood from Schwarzensee Lake (Kłusek et al., 2015), is very similar to the results demonstrated in this paper. This similarity is evident in the correlation coefficients calculated between MXD and δ13C chronologies, and the weather parameters. The correlation coefficients have the highest values during the summer months for both of these wood proxies. The best results for MXD chronology were observed for the temperatures of July, August, and September, similarly as in the case of δ13C chronology. The relationship with the temperature was also significant for April, May, and June. These results are also comparable to δ13C chronology. For solar radiation, positive relation with MXD was noted for April, July, August, and September of the current year. It also coincides with the outcomes obtained for δ13C chronology. In spite of such similarities, the distinct long-term trends observed within both of these chronologies differ considerably. For MXD chronology, the positive growth tendencies were visible especially within the time range of about 400–670 CE. The higher MXD values were also observed for the periods of 1120–1250 CE, 1300–1420 CE, and 1640 CE. Nevertheless, the period of 670–1800 CE was characterised by a rather low MXD data with only momentary fluctuations towards the increase of density. The largest index values occurred during the 20th century. The increase in the density began ~1820 and climbed steeply since ~1900. In turn, the time intervals marked by strongly diminished MXD values covered the years of 680–780 CE, 880–980 CE, 1270–1310 CE, and 1420–1500 CE (Kłusek et al., 2015). These differences observed in the long-term trends prove that MXD and δ13C chronologies reflect various climate variables, despite the similar results obtained during the correlation and response function analyses.

However, the influence exerted by the weather conditions on the stable isotope content varies over the time. The relationship reported for δ13C chronology and climate parameters is clear for most of the calibration period, but it was distinctly reduced and ceased to be significant in the time span of the last 50 years approximately. The lack of the correlation is connected to an increasing trend which is visible in the last decades of the chronology. This increasing tendency is partially caused by the introduction of PIN and fossil fuel corrections that upraised the chronology values and diminished the correlation with climate parameters for the recent 50 years. Nevertheless, the application of these corrections does not explain such a significant upward trend. Most probably, this trend results also from the climate changes induced by human activity (Waterhouse et al., 2004).

A similar unstable relationship with climate parameters was previously described for the carbon isotope measurements carried out for Norway spruce trees from the Swiss Alps. The temperature signal obtained for this area was substantially enhanced during the cooler and wetter summers of the first half of the 19th century, whereas in the warmer late 20th century, precipitation became more influential (Esper et al., 2017). Presumably, the photosynthetic rate (forced by temperature and irradiance) controlled the fractionation during the less favourable thermal conditions of the early 19th century, while stomatal conductivity (forced by soil moisture and air humidity) governed the discrimination against 13C during the calibration period of the late 20th century (Esper et al., 2017).

The Schwarzensee chronology also reveals significant similarities in climate response with other stable carbon isotope time-series constructed in the Alpine area. Many studies confirm the positive relationship with the temperature and sunshine level of the mid-summer period (July and August) and the negative correlation with the precipitation in the same months (e.g. Treydte et al., 2001; Levanič et al., 2009; Kress et al., 2010, 2014; Hafner et al., 2011, 2014). This resemblance, visible in the Schwarzensee and other Alpine stable carbon chronologies, results from the analogous physiological response of the plants relative to climatic impact, which stimulates or limits tree growth and influences the signature of the stable carbon isotopes in the wood.

5. Conclusions

Our research demonstrates that the meteorological factors distinctly affect the stable carbon isotope content in the Schwarzensee wood. Stronger high-frequency correlations with the summer temperature and sunshine level and weaker relationship with the precipitation value indicate that the stable carbon isotopes in the Schwarzensee wood are mainly governed by the carboxylation efficiency of the photosynthetic active enzyme. The variations in stomatal conductance have a lower impact. Water resources, which decide about the stomatal closure and opening, are sufficient in this region.

The final ratio of the stable carbon isotopes in the Schwarzensee wood is the result of interdependences that exist between several inter-related weather parameters. Unfortunately, it is impossible to conclude unequivocally which signal – thermal or sunshine – is prevailing in δ13C chronology. More evidence indicates that the dominant control on stable carbon isotopes in the wood is, most probably, solar radiation. Further research is necessary to confirm this hypothesis. For this purpose, the measurements of the stable oxygen isotopes are carried out on the base of the same material from Schwarzensee Lake. A comparison between these tree-ring proxies may clarify some peculiarities of the mixed climatic record incorporated in the δ13C chronology.

Supplementary Material

Chronology

Acknowledgements

We would like to express our gratitude to the anonymous Reviewers and Editor – Doctor Paul Hesse for the detailed, very precise revision and constructive comments that greatly improved the manuscript. This work was supported by the Austrian Science Fund FWF [grant numbers M 1127-B16, P 23998-B16].

Footnotes

Data availability

Datasets related to this article can be found at https://data.mendeley.com/datasets/65pzb3j9dh/draft?a=ac55b943-167c-450bb-b3a-094356ff0188, an open-source online data repository hosted at Mendeley Data.

References

  1. Auer I, Matulla C, Böhm R, Ungersböck M, Maugeri M, Nanni T, Pastorelli R. Sensitivity of frost occurrence to temperature variability in the European Alps. Int J Climatol. 2005;25:1749–1766. [Google Scholar]
  2. Auer I, Böhm R, Jurkovic A, Lipa W, Orlik A, Potzmann R, Schöner W, Ungersböck M, Matulla C, Briff K, Jones P, et al. HISTALP - historical instrumental climatological surface time series of the Greater Alpine Region. Int J Climatol. 2007;27(1):17–46. [Google Scholar]
  3. Barbour MM, Walcroft AF, Farquhar GD. Seasonal variation in δ13C and δ18O of cellulose from growth rings in Pinus radiata. Plant Cell Environ. 2002;25:1483–1499. [Google Scholar]
  4. Battipaglia G, Cherubini P, Saurer M, Siegwolf RT, Strumia S, Cotrufo F. Volcanic explosive eruptions of the Vesuvio decrease tree-ring growth but not photosynthetic rates in the surrounding forests. Glob Chang Biol. 2007;13:1–16. [Google Scholar]
  5. Biondi F, Waikul K. DENDROCLIM2002: a C++ program for statistical calibra-tion of climate signals in tree-ring chronologies. Comput Geosci. 2004;30:303–311. [Google Scholar]
  6. Böhm R, Auer I, Brunetti M, Maugeri M, Nanni T, Schöner W. Regional temperature variability in the European Alps: 1760– 1998 from homogenized instrumental time series. Int J Climatol. 2001;21:1779–1801. [Google Scholar]
  7. Borella S, Leuenberger M, Saurer M, Siegwolf R. Reducing uncertainties in δ13C analysis of tree rings: pooling, milling, and cellulose extraction. J Geophys Res. 1998;103:19519–19526. [Google Scholar]
  8. Bowling DR, Logan BA, Hufkens K, Aubrecht DM, Richardson AD, Burns SP, Anderegg WRL, Blanken PD, Eiriksson DP. Limitations to winter and spring photosynthesis of a Rocky Mountain subalpine forest. Agric For Meteorol. 2018;252:241–255. [Google Scholar]
  9. Brand WA, Coplen TB. Stable isotope deltas: tiny, yet robust signatures in nature. Isot Environ Health Stud. 2012;48(3):393–409. doi: 10.1080/10256016.2012.666977. [DOI] [PubMed] [Google Scholar]
  10. Buchmann N, Ehleringer JR. CO2 concentration profiles, and carbon and oxygen isotopes in C3 and C4 crop canopies. Agric For Meteorol. 1998;89:45–58. [Google Scholar]
  11. Chimani B, Matulla C, Böhm R, Hofstätter M. A new high resolution absolute temperature grid for the Greater Alpine Region back to 1780. Int J Climatol. 2013;33:2129–2141. [Google Scholar]
  12. Cropper JP. Tree-ring skeleton plotting by computer. Tree-Ring Bull. 1979;39:47–59. [Google Scholar]
  13. Dorado Liñán I, Gutiérrez E, Helle G, Heinrich I, Andreu-Hayles L, Planells O, Leuenberger M, Bürger C, Schleser G. Pooled versus separate measurements of tree-ring stable isotopes. Sci Total Environ. 2011;(11):2244–2251. doi: 10.1016/j.scitotenv.2011.02.010. [DOI] [PubMed] [Google Scholar]
  14. Efthymiadis D, Jones PD, Briff KR, Auer I, Böhm R, Schöner W, Frei C, Schmidli J. Construction of a 10-min-gridded precipitation data set for the Greater Alpine Region for 1800–2003. J Geophys Res. 2006;110:D01105. doi: 10.1029/2005JD006120. [DOI] [Google Scholar]
  15. Esper J, Carnelli AL, Kamenik C, Filot M, Leuenberger M, Treydte K. Spruce tree-ring proxy signals during cold and warm periods. Dendrobiology. 2017;77:3–18. [Google Scholar]
  16. Farquhar GD, Lloyd J. Carbon and oxygen isotope effects in the exchange of carbon dioxide between terrestrial plants and the atmosphere. In: Ehleringer JR, Hall AE, Farquhar GD, editors. Stable Isotopes and Plant Carbon-Water Relations. Academic Press; New York: 1993. pp. 47–70. [Google Scholar]
  17. Farquhar GD, O'Leary MH, Berry JA. On the relationship between carbon isotope discrimination and the intercellular carbon dioxide concentration in leaves. Aust J Plant Physiol. 1982;9:121–137. [Google Scholar]
  18. Gagen M, McCarroll D, Edouard J-L. Latewood width, maximum density, and stable carbon isotope ratios of pine as climate indicators in a dry subalpine environment, French Alps. Arct Antarct Alp Res. 2004;36(2):166–177. [Google Scholar]
  19. Gagen M, McCarroll D, Loader NJ, Robertson I, Jalkanen R, Anchukaitis KJ. Exorcising the ‘segment length curse’: summer temperature reconstruction since AD 1640 using non-detrended stable carbon isotope ratios from pine trees in northern Finland. The Holocene. 2007;17(4):435–446. [Google Scholar]
  20. Gagen M, McCarroll D, Robertson I, Loader NJ, Jalkanen R. Do tree ring δ13C series from Pinus sylvestris in northern Fennoscandia contain long-term non-climatic trends? Chem Geol. 2008;252:42–51. [Google Scholar]
  21. Gagen M, McCarroll D, Loader N, Robertson I. Stable isotopes in dendroclimatology: moving beyond potential. In: Hughes MK, Swetnam TW, Diaz HF, editors. Dendroclimatology, Progress and Prospects Development in Paleoenvironmental Research. Vol. 11. Springer Verlag; Dordrecht, Heidelberg, London, New York: 2011. pp. 147–174. [Google Scholar]
  22. Grabner M, Klein A, Geihofer D, Reschreiter H, Barth FE, Sormaz T, Wimmer R. Bronze age dating of timber from the salt-mine at Hallstatt, Austria. Dendrochronologia. 2006;24:61–68. [Google Scholar]
  23. Green JW. Wood cellulose. In: Whistler RL, editor. Methods of Carbohydrate Chemistry. Academic Press; New York: 1963. pp. 9–21. [Google Scholar]
  24. Hafner P, Robertson I, McCarroll D, Loader NJ, Gagen M, Bale RJ, Jungner H, Sonninen E, Hilasvuori E, Levanič T. Climate signals in the ring widths and stable carbon, hydrogen and oxygen isotopic composition of Larix decidua growing at the forest limit in the Southeastern European Alps. Trees. 2011;25(6):1141–1154. [Google Scholar]
  25. Hafner P, McCarroll D, Robertson I, Loader N, Gagen M, Young G, Bale R, Sonninen E, Levanič T. A 520 year record of summer sunshine for the eastern European Alps based on stable carbon isotopes in larch tree rings. Clim Dyn. 2014;43(3-4):971–980. [Google Scholar]
  26. Harlow BA, Marshall JD, Robinson AP. A multi-species comparison of δ13C from whole wood, extractive-free wood and holocellulose. Tree Physiol. 2006;26:767–774. doi: 10.1093/treephys/26.6.767. [DOI] [PubMed] [Google Scholar]
  27. Helama S, Arppe L, Timonen M, Mielikäinen K, Oinonen M. Age-related trends in subfossil tree-ring δ13C data. Chem Geol. 2015;416:28–35. [Google Scholar]
  28. Helle G, Schleser GH. Interpreting climate proxies from tree-rings. In: Fischer H, Floeser G, Kumke T, Lohmann G, Miller H, Negendank JFW, Von Storch H, editors. The KIHZ Project: Towards a Synthesis of Holocene Proxy Data and Climate Models. Springer Verlag; Berlin: 2004. pp. 129–148. [Google Scholar]
  29. Helle G, Schleser GH. Beyond CO2-fixation ation by Rubisco - an interpretation of 13C/12C variations in tree rings from novel intra-seasonal studies on broad-leaf trees. Plant Cell Environ. 2004;27:367–380. [Google Scholar]
  30. Jäggi M, Saurer M, Fuhrer J, Siegwolf R. The relationships between the stable carbon isotope composition of needle bulk material, starch, and tree rings in Picea abies. Oecologia. 2002;131:325–332. doi: 10.1007/s00442-002-0881-0. [DOI] [PubMed] [Google Scholar]
  31. Keel SG, Siegwolf RT, Jäggi M, Korner C. Rapid mixing between old and new C pools in the canopy of mature forest trees. Plant Cell Environ. 2007;30:963–972. doi: 10.1111/j.1365-3040.2007.01688.x. [DOI] [PubMed] [Google Scholar]
  32. Kłusek M, Grabner M. X-ray densitometry of Norway spruce subfossil wood from Austrian Alps. Tree-Ring Res. 2016;72(1):23–34. doi: 10.3959/1536-1098-72.01.23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Kłusek M, Pawełczyk S. Stable carbon isotope analysis of subfossil wood from Austrian Alps. Geochronometria. 2014;41(4):400–408. doi: 10.2478/s13386-013-0168-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kłusek M, Melvin TM, Grabner M. Multi-century long density chronology of living and sub-fossil trees from Lake Schwarzensee, Austria. Dendrochronologia. 2015;33:42–53. doi: 10.1016/j.dendro.2014.11.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Konter O, Holzkämper S, Helle G, Büntgen U, Esper J. Climate sensitivity and parameter coherency in annually resolvedδ13C and δ18O from Pinus uncinata tree-ring data in the Spanish Pyrenees. Chem Geol. 2014;377:12–19. [Google Scholar]
  36. Kress A, Young GHF, Saurer M, Loader NJ, Siegwolf RTW, McCarroll D. Stable isotope coherence in the earlywood and latewood of tree-line conifers. Chem Geol. 2009;268:52–57. [Google Scholar]
  37. Kress A, Saurer M, Siegwolf RTW, Frank DC, Esper J, Bugmann H. A 350 year drought reconstruction from Alpine tree ring stable isotopes. Glob Biogeochem Cycles. 2010;24:GB2011. doi: 10.1029/2009GB003613.2010. [DOI] [Google Scholar]
  38. Kress A, Hangartner S, Bugmann H, Büntgen U, Frank DC, Leuenberger M, Sieg-wolf RTW, Saurer M. Swiss tree rings reveal warm and wet summers during medieval times. Geophys Res Lett. 2014;41:1732–1737. doi: 10.1002/2013GL059081. [DOI] [Google Scholar]
  39. Leal S, Melvin T, Grabner M, Wimmer R, Briff K. Tree-ring growth variability in the Austrian Alps: the influence of site, altitude, tree species and climate. Boreas. 2007;36:426–440. [Google Scholar]
  40. Leavitt SW. Tree-ring isotopic pooling without regard to mass: no difference from averaging δ13C values of individual trees. Chem Geol. 2008;252:52–55. [Google Scholar]
  41. Leavitt SW. Tree-ring C-H-O isotope variability and sampling. Sci Total Environ. 2010;15:5244–5253. doi: 10.1016/j.scitotenv.2010.07.057. [DOI] [PubMed] [Google Scholar]
  42. Leavitt SW, Long A. Stable carbon isotopes as a potential supplemental tool in dendrochronology. Tree-Ring Bull. 1982;42:49–55. [Google Scholar]
  43. Leavitt SW, Long A. Sampling strategy for stable carbon isotope analysis of tree-rings in pine. Nature. 1984;311:145–147. [Google Scholar]
  44. Leavitt SW, Long A. Stable carbon isotope variability in trees foliage and wood. Ecology. 1986;67:1002–1010. [Google Scholar]
  45. Leavitt SW, Long A. Seasonal stable-carbon isotope variability in tree rings: possible paleoenvironmental signals. Chem Geol. 1991;87:59–70. [Google Scholar]
  46. Levanič T, Gričar J, Gagen M, Jalkanen R, Loader NJ, McCarroll D, Oven P, Robertson I. The climate sensitivity of Norway spruce [Picea abies (L.) Karst.] in the southeastern European Alps. Trees. 2009;23:169–180. [Google Scholar]
  47. Loader NJ, Robertson I, Barker AC, Switsur VR, Waterhouse JS. An improved technique for the bath processing of small wholewood samples to α-cellulose. Chem Geol. 1997;136:313–317. [Google Scholar]
  48. Loader NJ, Robertson I, McCarroll D. Comparison of stable carbon isotope ratios in the whole wood, cellulose and lignin of oak tree-rings. Palaeogeogr Palaeoclimatol Palaeoecol. 2003;196:395–407. [Google Scholar]
  49. Loader NJ, Young GHF, McCarroll D, Wilson RJS. Quantifying uncertainty in isotope dendroclimatology. The Holocene. 2013;23:1221–1226. [Google Scholar]
  50. Marshall JD, Monserud RA. Co-occurring species differ in tree-ring δ18O trends. Tree Physiol. 2006;26(8):1055–1066. doi: 10.1093/treephys/26.8.1055. [DOI] [PubMed] [Google Scholar]
  51. McCarroll D, Loader NJ. Stable isotopes in tree rings. Quat Sci Rev. 2004;23:771–801. [Google Scholar]
  52. McCarroll D, Loader NJ. Isotopes in tree rings. In: Leng MJ, editor. Isotopes in Palaeoenvironmental Research. Developments in Paleoenvironmental Research Springer Verlag; Netherlands, Dordrecht: 2006. pp. 67–116. [Google Scholar]
  53. McCarroll D, Jalkanen R, Hicks S, Tuovinen M, Gagen M, Pawellek F, Eckstein D, Schmitt U, Autio J, Heikkinen O. Multiproxy dendroclimatology: a pilot study in northern Finland. The Holocene. 2003;13(6):831–840. [Google Scholar]
  54. McCarroll D, Gagen M, Loader NJ, Robertson I, Anchukaitis KJ, Los S, Young GHF, Jalkanen R, Kirchhefer A, Waterhouse JS. Correction of tree ring stable carbon isotope chronologies for changes in the carbon dioxide content of the atmosphere Geochim. Cosmochim Acta. 2009;73:1539–1547. [Google Scholar]
  55. McCarroll D, Tuovinen M, Campbell R, Gagen M, Grudd H, Jalkanen R, Loader N, Robertson I. A critical evaluation of multi-proxy dendroclimatology in northern Finland. J Quat Sci. 2011;26(1):7–14. [Google Scholar]
  56. Pallardy SG. Physiology of Woody Plants. third ed. Academic Press; San Diego: 2008. [Google Scholar]
  57. Porte A, Loustau D. Seasonal and interannual variations in carbon isotope discrimination in a maritime pine (Pinus pinaster) stand assessed from the isotopic composition of cellulose in annual rings. Tree Physiol. 2001;21:861–868. doi: 10.1093/treephys/21.12-13.861. [DOI] [PubMed] [Google Scholar]
  58. Rinn F. TSAP-win Professional Timeseries Analysis and Presentation for Den drochronology and Related Applications, Version 0.3. Heidelberg: 2003. Quick Reference. [Google Scholar]
  59. Sass-Klaassen U, Poole I, Wils T, Helle G, Schleser GH, Van Bergen PF. Carbon and oxygen isotope dendrochronology in sub-fossil bog oak tree rings - a preliminary study. IAWA J. 2005;26:121–136. [Google Scholar]
  60. Saurer M, Siegwolf RTW. Human impacts on tree-ring growth reconstructed from stable isotopes. Stable Isotopes as Indicators of Ecological Change. Terrestrial Ecology. 2007;1:49–62. [Google Scholar]
  61. Saurer M, Siegenthaler U, Schweingruber F. The climate-carbon isotope relationship in tree rings and the significance of site conditions. Tellus. 1995;47B:320–330. [Google Scholar]
  62. Saurer M, Cherubini P, Bonani G, Siegwolf R. Tracing carbon uptake from a natural CO2 spring into tree rings: an isotope approach. Tree Physiol. 2003;23:997–1004. doi: 10.1093/treephys/23.14.997. [DOI] [PubMed] [Google Scholar]
  63. Schleser GH, Frilingsdorf J, Blair A. Carbon isotope behaviour in wood and cellulose during artificial aging. Chem Geol. 1999;158:121–130. [Google Scholar]
  64. Schleser GH, Helle G, Lucke A, Vos H. Isotope signals as climate proxies: the role of transfer functions in the study of terrestrial archives. Quat Sci Rev. 1999;18:927–943. [Google Scholar]
  65. Schweingruber FH. Trees and Wood in Dendrochronology: Morphological, Anatomical, and Tree-ring Analytical Characteristics of Trees Frequently Used in Dendrochronology. Springer Verlag; Berlin, Heidelberg: 1993. [Google Scholar]
  66. Treydte K, Schleser GH, Schweingruber FH, Winiger M. The climatic significance of δ13C in subalpine spruces (Lotschental, Swiss Alps) Tellus. 2001;53B:593–611. [Google Scholar]
  67. Treydte KS, Frank DC, Saurer M, Helle G, Schleser GH, Esper J. Impact of climate and CO2 on a millennium-long tree-ring carbon isotope record. Geochim Cosmochim Acta. 2009;73:4635–4647. [Google Scholar]
  68. Tsuji H, Nakatsuka T, Takagi K. δ18O of tree-ring cellulose in two species (spruce and oak) as proxies of precipitation amount and relative humidity in northern Japan. Chem Geol. 2006;231:67–76. [Google Scholar]
  69. Waterhouse JS, Switsur VR, Barker AC, Carter AHC, Hemming DL, Loader NJ, Robertson I. Northern European trees show a progressively diminishing response to increasing atmospheric carbon dioxide concentrations. Quat Sci Rev. 2004;23:803–810. [Google Scholar]
  70. Young GHF, Demmler JC, Gunnarson BE, Kirchhefer AJ, Loader N, McCarroll D. Age trends in tree ring growth and isotopic archives: a case study of Pinus sylvestris L. from northwestern Norway. Glob Biogeochem Cycles. 2011;25 doi: 10.1029/2010GB003913. GB2020. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Chronology
NIHMS80146-supplement-Chronology.xlsx (46.3KB, xlsx)

RESOURCES