INTRODUCTION
During the last centuries, the carbon isotopic composition of the atmosphere and biosphere has been modified due to fossil fuel (e.g., coal, petroleum, natural gas) combustion in industrial areas (Martin et al. Reference Martin, Bytnerowicz and Thorstenson1988; Farquhar and Lloyd Reference Farquhar and Lloyd1993; Vitousek et al. Reference Vitousek, Aber, Howarth, Likens, Matson, Schindler, Schlesinger and Tilman1997; Choi et al. Reference Choi, Lee, Chang and Ro2005; McCarroll et al. Reference McCarroll, Gagen, Loader, Robertson, Anchukaitis, Los, Young, Jalkanen, Kirchhefer and Waterhouse2009; Levin et al. Reference Levin, Naegler, Kromer, Diehl, Francey, Gomez-Pelaez, Schäfer, Steele, Wagenbach, Weller and Worthy2010; Pazdur et al. Reference Pazdur, Kuc, Pawełczyk, Piotrowska, Sensuła and Różański2013; Sensuła and Pazdur Reference Sensuła and Pazdur2013a; Boden et al. Reference Boden, Marland and Andres2016). CO2 from fossil fuels is devoid of any 14C (Suess Reference Suess1955). Also a progressive lowering of 13C in the air and thus in the biosphere is associated with the emission of 13C-depleted CO2 into the atmosphere from fossil fuel (Craig Reference Craig1954; Farquhar and Lloyd Reference Farquhar and Lloyd1993; McCarroll and Loader Reference McCarroll and Loader2004; Pazdur et al. Reference Pazdur, Nakamura, Pawełczyk, Pawlyta, Piotrowska, Rakowski, Sensuła and Szczepanek2007, Reference Pazdur, Kuc, Pawełczyk, Piotrowska, Sensuła and Różański2013; Keeling et al. Reference Keeling, Piper, Bollenbacher and Walker2010; Sensuła and Pazdur Reference Sensuła and Pazdur2013a, Reference Sensuła and Pazdur2013b; Leonelli et al. Reference Leonelli, Battipaglia, Siegwolf, Saurer, di Cella, Cherubini and Pelfini2012).
In Poland, according to the Statistical Review of World Energy (2016), the carbon dioxide emission in 1975 was approximately 380 million tons, whereas the carbon dioxide emission in 2014 was approximately 291 million tons. In central and east European countries, there has been a strong and durable reduction in emissions from the beginning of the 1990s and parallel to this, there has been an improvement in the health status of forests in this part of Europe (Juknys et al. Reference Juknys, Vencloviene, Stravinskiene, Augustaitis and Bertkevicius2003; Wilczyński Reference Wilczyński2006; Elling et al. Reference Elling, Dittmar, Pfaffelmoser and Rotzer2009).
The aims of this study were to determine the spatial and temporal carbon isotope fractionation associated with industrial activity, pollution from vehicles, household heating and low stack emissions in southern Poland. The purpose of this study was to extend current knowledge on interactions between trees and CO2 emissions from different sources in three forests in the Silesia Region (Poland). In the factories in Silesia many different units and production facilities have been constructed and expanded over the last century. Although numerous projects dedicated to environmental protection have been implemented in these factories over many years aimed at a reduction of the negative impact of production processes on the environment, most factories were on the list of the most environmentally noxious factories until the 1990s. The studies answer following questions: (1) how trees adapted to environmental contamination: how trees reacted to high industrial emission of carbon dioxide; (2) if the negative impact of industrial pollution on trees was stronger than the positive influence of an increase in CO2 in the atmosphere; (3) if there is any similar response in carbon stable isotopic composition of annual tree rings in sampling sites located nearby different factories; (4) if there is any spatial (taking into account a distance of sampling sites from factories) similarity in variation of carbon isotopic composition of the foliage; (5) what differences can be noted between the values of Δ14Co in atmospheric CO2 and Δ14C in tree ring cellulose and needles; (6) if there is any correlation between carbon isotopic composition of tree ring cellulose and anthropogenic CO2 emissions; and (7) how global decrease in radiocarbon (14C) concentrations in the air is recorded in pine tree rings in Silesia region over last decades.
The annual ring growth variability of Scots pine (Pinus Silvestris L.) has been used as a bio-indicator of environment change (Schweingruber Reference Schweingruber1996; Wilczynski Reference Wilczyński2006; Wagner and Wagner Reference Wagner and Wagner2006; Malik et al. Reference Malik, Danek, Marchwińska-Wyrwał, Danek, Wistuba and Krąpiec2012; Sensuła and Pazdur Reference Sensuła and Pazdur2013a, Reference Sensuła and Pazdur2013b; Pazdur et al. Reference Pazdur, Kuc, Pawełczyk, Piotrowska, Sensuła and Różański2013; Sensuła et al. Reference Sensuła, Opała, Wilczyński and Pawełczyk2015a, Reference Sensuła, Wilczyński and Opała2015b). The width of annual tree rings is considered an appropriate monitor of forest conditions (Eckstein Reference Eckstein1989; Schweingruber Reference Schweingruber1996; McLaughlin et al. Reference McLaughlin, Hellmann, Boggs and Ehrlich2002; Dobbertin Reference Dobbertin2005). Tree foliage may act as a filter, concentrating the contamination. Observations show that, due to air pollution in the 20th century, different trees species populations have demonstrated a different sensitivity to weather conditions (for example Leonelli et al. Reference Leonelli, Battipaglia, Siegwolf, Saurer, di Cella, Cherubini and Pelfini2012; Battipaglia et al. Reference Battipaglia, Saurer, Cherubini, Calfapietra, McCarthy, Norby, Cotrufo, Boettger, Haupt, Friedrich and Waterhouse2014; Sensuła and Pazdur Reference Sensuła and Pazdur2013a, Reference Sensuła and Pazdur2013b; Sensuła Reference Sensuła2016a). Carbon dioxide molecules are absorbed from the atmosphere by plants, and during photosynthesis, plants convert CO2 and H2O to cellulose which is the basic structural component of plant cell walls.
The observed anthropogenic impact on the carbon cycle, related to regional and global diffusion of industrial CO2 emissions and some other local human activities (such as, industrial factories, vehicles, housing energy, low stack emission of CO2) has led to changes in the isotopic composition of carbon in the atmosphere and biosphere (Craig Reference Craig1954; Suess Reference Suess1955; Leavitt and Long Reference Leavitt and Long1982; Martin et al. Reference Martin, Bytnerowicz and Thorstenson1988; Ehrelinger Reference Ehleringer1990; Ehrelinger and Vogel Reference Ehrelinger and Vogel1993; McCarroll and Loader Reference McCarroll and Loader2004; Sensuła et al. Reference Sensuła, Böttger, Pazdur, Piotrowska and Wagner2006; Pazdur et al. Reference Pazdur, Nakamura, Pawełczyk, Pawlyta, Piotrowska, Rakowski, Sensuła and Szczepanek2007; McCarroll et al. Reference McCarroll, Gagen, Loader, Robertson, Anchukaitis, Los, Young, Jalkanen, Kirchhefer and Waterhouse2009; Keeling et al. Reference Keeling, Piper, Bollenbacher and Walker2010; Rinne et al. Reference Rinne, Loader, Switsur, Treydte and Waterhouse2010; Savard Reference Savard2010; Sensuła et al. Reference Sensuła, Pazdur and Marais2011; Pazdur et al. Reference Pazdur, Kuc, Pawełczyk, Piotrowska, Sensuła and Różański2013; Sensuła Reference Sensuła2015; Boden et al. Reference Boden, Marland and Andres2016; Sensuła Reference Sensuła2016a, Reference Sensuła2016b, Reference Sensuła, Wilczyński and Piotrowska2016c; Sensuła and Wilczyński Reference Sensuła and Wilczyński2017). The carbon isotopic composition of trees may be affected by changes in carbon isotopic composition of atmospheric CO2 and during enzymatic reactions and due to changes in temperature, sunshine, water stress, or contamination (Farquhar and Sharkey Reference Farquhar and Sharkey1982; Ehleringer Reference Ehleringer1990; Ehrelinger and Vogel Reference Ehrelinger and Vogel1993; Sensuła Reference Sensuła2016b). The studies indicated that tree ring δ13C residuals sensitively respond to CO2 emission and changing climate. Spatial difference in the long‐term declining trend in the raw δ13C may be driven by changes in stomatal conductance (i.e., supply of CO2), or in photosynthetic rate (i.e., demand for CO2), or both. Pines might show a decreasing δ13C pattern in tree rings due to increases in atmospheric CO2 concentration, to 13C and 14C-depleted CO2 emission from soil organic carbon decomposition and fossil fuel combustion or to other effects. At the same time increased atmospheric CO2 concentration can induce stomatal closure due to high intercellular CO2 concentration that leads to less C isotope discrimination and thus may mask 13C depleted CO2 effect on δ13C plant to some degree (McCarroll et al. Reference McCarroll, Gagen, Loader, Robertson, Anchukaitis, Los, Young, Jalkanen, Kirchhefer and Waterhouse2009). The effect of different sensitivity to weather conditions and masking climatic signal due to air pollution in last decades has been noted also in other parts of Europe (for example Leonelli et al. Reference Leonelli, Battipaglia, Siegwolf, Saurer, di Cella, Cherubini and Pelfini2012; Battipaglia et al. 2013; Sensuła et al. Reference Sensuła and Pazdur2013a, Reference Sensuła and Pazdur2013b).
The impact of air pollution and carbon dioxide emission on tree ring width, δ13C and 14C concentration in trees is evident (for instance, Craig Reference Craig1954; McCarroll and Loader Reference McCarroll and Loader2004; Pazdur et al. Reference Pazdur, Nakamura, Pawełczyk, Pawlyta, Piotrowska, Rakowski, Sensuła and Szczepanek2007, Reference Pazdur, Kuc, Pawełczyk, Piotrowska, Sensuła and Różański2013; Keeling et al. Reference Keeling, Piper, Bollenbacher and Walker2010; Rinne et al. Reference Rinne, Loader, Switsur, Treydte and Waterhouse2010; Savard Reference Savard2010; Sensuła et al. Reference Sensuła, Pazdur and Marais2011; Sensuła and Pazdur Reference Sensuła and Pazdur2013a, Reference Sensuła and Pazdur2013b; Sensuła Reference Sensuła2015). Elevated concentration of CO2 from industrial sources can be also associated with emission of other industrial pollutants, such as for example NOx, SO2 and O3 (Freyer Reference Freyer1979; Choi and Lee Reference Choi and Lee2012), which can also influence on carbon stable isotope composition in plant tissue. In our studies, the influence of NOx, SO2, and O3 on carbon stable isotope fractionation cannot be excluded. There is a lack of anthropogenic gaseous emission data for the Silesia region, due to the lack of access to data from factories and due to the changes in the administrative division of Poland.
MATERIALS AND METHODS
The analysis reported in this paper included 15 pine sites within three regions of significant urban areas: Dabrowa Gornicza near a steelworks Huta Katowice (HK), Kedzierzyn-Kozle near chemical factories (KK) and Laziska near a combined heat and power plant (LA) (Figure S1, Table 1).
Characteristics of the sampling sites in three regions: Dabrowa Gornicza near Huta Katowice (HK; 50°20'31''N 19°16'1''E), Kedzierzyn-Kozle near chemical factories (KK; 50°18'20''N 18°15'27''E) and Laziska near a combined heat and power plant (LA; 50°07'58''N 18°50'47.1''E).
The sampling sites (Table 1, Figure S1) were located at different distances: between approximately 5 and 20 km from industrial factories. Needles were collected from 15 sampling sites. Samples of needles collected in January 2013 (11 samples) had been created in the previous year, i.e. 2012. The needles collected in September 2013 (15 samples), grew during 2013, and the needles collected in July 2014 (15 samples) had also been created during 2014. The difference in number of samples collected in winter 2013 and in summer 2013 and in summer 2014 was due to banning entry into the part of the forests (Sensuła Reference Sensuła2015).
Tree rings were collected from 11 sampling sites (Table 1, Figure S1). In each of 11 selected sites, 20 pines, aged between 80 to 100 years, were sampled by taking one increment per tree at a height of 1.3 m above ground. All 220 sampled trees were dominant and codominant individuals without damage. Dendrochronological analysis has been described in detail already (Sensuła et al. Reference Sensuła, Opała, Wilczyński and Pawełczyk2015a, Reference Sensuła, Wilczyński and Opała2015b, Reference Sensuła, Wilczyński, Monin, Allan, Pazdur and Fagel2017). The average ring width was computed for 3 time intervals: low level of emissions of industrial pollution (1940–1960), culmination of emissions (1961–1990) and reduction of industrial emissions (since 1991). Dendrochronological analyses (Sensuła et al. Reference Sensuła, Opała, Wilczyński and Pawełczyk2015a, Reference Sensuła, Wilczyński and Opała2015b, Reference Sensuła, Wilczyński, Monin, Allan, Pazdur and Fagel2017) enable the selection of sites for analysis of the stable isotopes and 14C concentration in trees. From each region, one sampling site per stand, where the strongest and the longest tree ring width reductions were observed due to an increase in industrial pollution emission, was chosen for isotopic investigation (LA_11 in Laziska region, KK_3 in Kedzierzyn-Kozle region, and HK_14 in Dabrowa Górnicza region, respectively).
The dendrochronologically dated annual tree rings were manually separated, pooled, homogenized, and cut into small pieces. α-cellulose samples were extracted (from 10 trees per site) by applying procedures based on Green’s method (Reference Green1963) used in the mass spectrometry laboratory of the Silesian University of Technology (Pazdur et al. Reference Pazdur, Nakamura, Pawełczyk, Pawlyta, Piotrowska, Rakowski, Sensuła and Szczepanek2007, Reference Pazdur, Kuc, Pawełczyk, Piotrowska, Sensuła and Różański2013; Sensuła et al. Reference Sensuła, Pazdur and Marais2011; Sensuła and Pazdur Reference Sensuła and Pazdur2013a, Reference Sensuła and Pazdur2013b; Sensuła et al. Reference Sensuła2016a, Reference Sensuła2016b).The stable isotopic compositions of annual tree rings were analyzed with annual resolution for the period 1975–2012.
The 14C isotopic compositions of annual tree rings α-cellulose were analyzed with annual resolution for the period 1990–2012 and for each 5th year resolution for the period 1975–1990. Additionally, 14C isotopic composition was measured also in tree rings created in 1986 to evaluate the effect of the accident at the Chernobyl nuclear power plant in Ukraine.
STABLE ISOTOPES
δ13C were determined at the mass spectrometry laboratory of the Silesian University of Technology using an Isoprime continuous flow isotope ratio mass spectrometer (GV Instruments, Manchester, UK).
The isotope values were reported in the delta notation (Equation 1, in ‰)
$$\delta ^{{13}} {\rm C}{\equals}\left( {{\rm R}_{{{\rm sample}}} \,/\,{\rm R}_{{{\rm standard}}} -1} \right) \hskip-2pt \cdot \hskip-2pt 1000$$
relative to the international V-PDB (Vienna Pee Dee Belemnite) standard. Rstandard and Rsample are the molar fractions of 13C/12C for the sample and the standard, respectively.
The standard deviation for the repeated analysis of an internal standard (C-3 and C-5, IAEA) was better than 0.2‰.
To describe the variation of the carbon isotope composition of tree ring cellulose caused by climate changes and anthropogenic emission of CO2, we used a model (Sensuła et al. Reference Sensuła, Pazdur and Marais2011) based on multiple regressions (Equation 2) using Statistica 12 (Statsoft Inc. 2014):
$$\delta {\equals}{\rm R{\plus}}\mathop{\sum}\limits_{{\rm M}{\equals}{\rm Oct}}^{{\rm Sep}} {{\rm b}_{{{\rm MT}}} {\rm T}_{{\rm M}} } {\rm {\plus}}\mathop{\sum}\limits_{{\rm M}{\equals}{\rm Oct}}^{{\rm Sep}} {{\rm b}_{{{\rm MT}}} {\rm P}} {\rm {\plus}}\mathop{\sum}\limits_{{\rm M}{\equals}{\rm Oct}}^{{\rm Sep}} {{\rm b}_{{{\rm MT}}} {\rm S}_{{\rm M}} {\plus}{\rm b}_{{\rm E}} {\rm E}_{{{1 \over {\rm CO}_{{\rm 2}}}} } $$
where R corresponds to the interdependences between the monthly climate factors and other environmental changes (for instance: pollutant emission, “potential for growth,” carbon flux from biosphere to atmosphere or error connected to the assumptions (Pazdur et al. Reference Pazdur, Kuc, Pawełczyk, Piotrowska, Sensuła and Różański2013, Sensuła et al. 2013a), M is the month (from October of the previous year to September of the given year), b is the regression coefficient for the following variables: T (average of the monthly temperatures), P (total monthly precipitation), S (monthly hours of sunshine), E1/CO2 =emission of CO2was used in statistical analyses.
Additionally, raw δ13C data can be mathematically corrected to a preindustrial atmospheric δ13Ccor using the data from McCarroll and Loader (Reference McCarroll and Loader2004).The correction reduces the decline of the last decades (Rinne et al. Reference Rinne, Loader, Switsur, Treydte and Waterhouse2010).
Radiocarbon
Samples of α-cellulose were converted to graphite for AMS 14C measurements. The process was performed in an AGE automated graphitization system (Wacker et al. Reference Wacker, Nemec and Bourquin2010). Subsamples of ca. 3 mg of cellulose were weighed to tin capsules, combusted in a VarioMicroCube elemental analyzer and the CO2 was reduced with hydrogen over the Fe catalyst. Coal was used as a blank material and was prepared with the standard ABA treatment (Piotrowska Reference Piotrowska2013). Oxalic Acid II (NIST SRM4990C) was used as a reference material, without any chemical pretreatment. The blank and OxII samples were combusted and graphitized as described above. 14C concentration was determined at the DirectAMS laboratory, Bothell, WA, USA (Zoppi et al. Reference Zoppi2010). Three to four graphites of blank and OxII were measured in the same run as unknown α-cellulose samples. The average blank for these series was 0.3 pMC, which corresponds to 46.6 ka BP, and was subtracted from measured concentrations.
The needles were prepared using a standard acid-alkali-acid treatment and converted to benzene for LSC measurements. Measurement of 14C concentrations in pine needles was performed with a ultra-low liquid scintillation spectrometer of the type Quantulus 1220 (Pawlyta et al. Reference Pawlyta, Pazdur, Rakowski, Miller and Harkness1998).
The measured 14C concentration or activity was corrected for δ13C and normalized to the standard of modern biosphere, resulting in the value of F14C. The 14C content in a tree ring in a year of its origin was calculated according to the formula (Equation 3; van der Plicht and Hogg Reference van der Plicht and Hogg2006):
$$\rDelta ^{{14}} {\rm C}{\equals}\left( {{\rm F}^{{14}} {\rm C}{\rm \cdot e}^{{{\minus}\lambda ({\rm Ti}{\minus}1950)}} {\minus}1} \right)\hskip-2.5pt\cdot\hskip-2.5pt 1000$$
where: F14C-normalized 14C concentration, λ-decay constant for 14C isotope, equal to 8267 yr−1; Ti – calendar year of the tree ring or formation of the needles.
Pollution and Meteorological Data
The emission data were obtained from the Carbon Dioxide Information Analysis Center (Boden et al. Reference Boden, Marland and Andres2016). The δ13C changes in the isotopic ratio of atmospheric CO2 due to global fossil fuel emissions ranged from 0.37‰ (in 1975) to 1.97‰ (in 2012) (McCaroll and Loader Reference McCarroll and Loader2004; McCaroll et al. Reference McCarroll, Gagen, Loader, Robertson, Anchukaitis, Los, Young, Jalkanen, Kirchhefer and Waterhouse2009). The experimental data were extrapolated to the present.
Mean Δ14Co concentrations (Δ14Co) in the atmospheric CO2 from April to September (Table S1) of the given year (reference values, commonly called “background”) were calculated on the basis of the data presented by Hammer et al. (Reference Hammer and Levin2017) and Hua et al. (Reference Hua, Barbetti and Rakowski2013). The former Δ14Co values (from Jungfraujoch and Schauinsland) cover the period from 1986 until the present, whereas the latter Δ14Co values (for Northern Hemisphere Zone 1) cover the period from 1975 to 1986, due to the lack of Δ14Co data for Jungfraujoch for the period prior to 1986.
The data for temperature, humidity and precipitation come from the meteorological station in Katowice and Opole. The meteorological data were obtained thanks to the Polish Institute of Meteorology and Water Management (IMGW-PIB).
RESULTS AND DISCUSSION
Dendrochronology
Previous studies showed that the pollutant emissions weakened the vitality of the pines, especially those exposed to their direct impact (Eckstein Reference Eckstein1989; Juknys et al. Reference Juknys, Vencloviene, Stravinskiene, Augustaitis and Bertkevicius2003; Wilczyński Reference Wilczyński2006; Elling et al. Reference Elling, Dittmar, Pfaffelmoser and Rotzer2009; Sensuła et al. Reference Sensuła, Opała, Wilczyński and Pawełczyk2015a, Reference Sensuła, Wilczyński and Opała2015b). The strongest and longest-drawn reductions of radial growth were recorded in the stands located in the paths of the dominant winds. A rapid decrease in the annual increment of pines in all pine stands can be observed after 1960, when the industrial pollution strongly increased (Figure S2).
The decrease in the annual radial increment was not dependent on the location of the pollution sources. All sites were located in the path of winds carrying pollutants (Figure S1). The duration of the radial growth depression period in individual stands was similar. It lasted 30 years (Figure S2). The recovery started in the 1990s along with the reduction of pollutant emissions (Figure S2). At the beginning of the 1990s the weakened pines, which greatly reduced their growth, began to rapidly increase their radial increment. This revitalization was very dynamic (Figure S2). This is also confirmed by the results shown in Figure S3.
The results of the correlation shown a very strong and significant (p<0.01) dependence between the radial increment values in the pre-depression, depression and recovery periods (Figure S3). A positive relationship has been found between the radial increment reduction and the degree of recovery. The greater the radial increment of pines in the pre-depression period, the greater the growth reduction during periods of increased pollution, and the more dynamic the incremental recovery (Figure S3).
Stable Isotopes in Tree Ring Cellulose
Spatial difference in long‐term trend in the δ13C chronology has been observed in all investigated sites (Figure S4).
Despite the differences in δ13C chronology trends in each sampling site, the average value of pine δ13C in each sampling site was about −24‰.
In the Kedzierzyn-Kozle (KK) pine δ13C series data range from −24.6‰ to −23.2‰ (uncorrected) and from −23.1‰ to −21.9‰ (corrected), in the Laziska (LA) pine δ13C series data range from −24.4‰ to −23.1‰ (uncorrected) and from −23.4‰ to −21.6‰ (corrected), in the Dabrowa Gornicza (HK) pine δ13C series data range from −25.3‰ to −22.9‰ (uncorrected) and from −23.6‰ to −21.5‰ (corrected). The δ13C values in α-cellulose samples extracted from pine growing in these 3 forests show a significant correlation between them (rLA vs HK=0.71; rKK vs LA,=0.78, rHK vs KK=0.78 n=39, p<0.001), which confirms the similarity in tree response to changes in the ecosystem. The admixture of large amounts of fossil-fuel derived CO2, which is depleted in 13CO2 resulted in a dilution of the atmospheric 13CO2 (Suess effect) (Figure S4). An inverse trend between CO2 emission and δ13C was observed: the increase in CO2 emission decreased δ13C and decreased CO2 emission increased δ13C (Figure S4).
The relationships between climate and isotopic composition exhibit spatiotemporal diversity. The variation of the carbon isotope composition of tree ring cellulose caused by climate changes and anthropogenic emission of CO2, based on multiple regressions is illustrated in Figure S5.
The value of the correlation coefficient between the measured and modeled δ13C in α-cellulose are above 0.9 (n=39).
Different b-values (Table S1) for climatic parameters may confirm that not only individual climatic parameters but also other interrelated components of the climate system may affect the correlation coefficient for the isotopic series from these trees. Therefore, this model suggests that the Silesia region may not be a uniform region for isotopic composition in Scots pine due to these other sources of variability.
Our model (Sensuła and Pazdur 2011) attempts to assess the CO2 emission component based on multiple regression analyses of the δ13C records. The total fossil fuel emission of CO2 estimated from the multiple regression model followed a nationally reported pattern with the anthropogenic industrial emission and showed a similar trend as that evaluated for Poland earlier by Boden et al. (Reference Boden, Marland and Andres2016) (Figure S5).
Radiocarbon in Tree Ring Cellulose
The comparison of the 14C concentrations in tree ring cellulose of the examined populations and Δ14Co concentration reference values (“background”) is presented in Figure S6, and in Tables 1 and S2.
During the period investigated, the values of Δ14Co show a significant decrease: from approximately 400 to 40‰ in an exponential form. The high initial value of Δ14Co was due to the anthropogenic production of the 14C isotope during nuclear tests in the atmosphere, mainly in the 1960s. The Δ14Co decrease following this period is primarily due to a mixing of 14CO2 containing artificially generated 14C atoms with the global surface ocean in the Earth’s carbon cycle (Hua et al. Reference Hua, Barbetti and Rakowski2013).
A similar trend is evident in the results obtained for tree ring cellulose samples from Laziska, Kedzierzyn-Kozle and Dabrowa Gornicza (Table 2, Figure S6).
Mean value of Δ14Co in the atmosphere (from April to September of the given year) and Δ14C in tree ring cellulose and its uncertainty (u) in the given year. Δ14Co in the atmosphere for the period 1975–1986 was taken from the work of Hua et al. (Reference Hua, Barbetti and Rakowski2013), whereas for the period 1986–2014 from the work of Hammer et al. (Reference Hammer and Levin2017).
During the investigated period, not all Δ14C values in tree ring cellulose are characterized by lower values than Δ14Co. Differences between the values of Δ14Co in atmospheric CO2 and Δ14C in tree ring cellulose are observed. These differences (Tables 2 and S2) are not constant over time, but on average for the whole study period, the difference between Δ14Co and Δ14C in tree ring cellulose range from –16.6‰ to 6.1‰ in Dabrowa Gornicza, range from –3.5‰ to 20.6‰ in Laziska, range from –21.6‰ to 7.5‰ in Kedzierzyn-Kozle, respectively.
δ13C and 14C in Needle Samples
The patterns of the spatial and short-temporal variability of δ13C, Δ14C in pine needles of the three forests in the most industrialized part of Poland – in proximity to the heat and power plant in Laziska (LA), the nitrogen plant in Kedzierzyn-Kozle (KK) and the steelworks in Dabrowa Gornicza (HK) are summarized in Tables S3–S5.
A detailed analysis of the carbon isotope composition of needles in each investigated site confirms that environmental records are not homogeneous within a single research region. There are significant differences in the stable isotopic and 14C composition of the needles between pine populations and between needles from year to year.
An analysis of the isotopic composition of pine growing in Laziska region (Tables S3–S5, Figure S1 and Figure 1) shows the lowest δ13C value in the LA_10 population (located near a road, near detached houses, about 7 km from a factory), while LA_11 and LA_9 show similar values.
Spatial and short-time variation of the stable carbon isotope and 14C composition of needles from 15 sampling sites in 3 regions (see Table 1).
For pine growing in Dabrowa Gornicza region (Tables S3–S5, Figure 1) the lowest δ13C values were noted in HK_15 (20 km from factories, very near a road).
In one population (HK_14), there were no differences in carbon isotopic composition between samples collected at different times, or between the winter or in the summer. Also, there was no significant difference between δ13C values in samples collected in 2013 and 2014. For pines growing in Kedzierzyn-Kozle region the lowest δ13C values were noted in KK_1 and KK_4 (near factories and roads). Populations growing at a distance of 6 km from factories and far from the road (KK_3 and KK_5) show no differences in carbon isotopic composition either in the winter or in the summer, also there was no significant difference between δ13C values in samples collected in 2013 and 2014.
Comparing δ13C values in the pine needles grown in 2012 (collected in winter 2013) with δ13C values in the pine needles grown in 2013 (collected in summer 2013), the values varied from −2.3‰ to 1.5‰. Depletion in δ13C was observed in half of the investigated sites. Usually, but not always, an increase in δ13C corresponds to an increase in Δ14C was observed. Comparing δ13C values in the pine needles created in 2013 (collected in summer 2013) with δ13C values in the pine needles created in 2014 (collected in summer 2014), δ13C in summer 2014 was mostly higher than in summer 2013 (in 11 of the 15 sampling sites).
In 2012, δ13C in tree ring cellulose in KK_3 was equal to –24.1‰, whereas δ13C in tree ring cellulose in LA_11 was equal to –23.5‰. The difference between δ13C in the needles (Table S3) and tree ring cellulose was about 5‰ in KK_3 and LA_11, and the δ13C difference between needles and tree ring cellulose (the ratio of δ13C in cellulose to δ13C in needles created in 2012) was a factor of 1.2 in both investigated sites (KK_3 and LA_11). However, the ratio of 14C concentrations in tree rings and foliage varied, in one case >1, while in another sampling site it was<1 as shown in Table 1.
Transect studies δ13C and 14C gradients have shown inhomogeneous changes in carbon isotopes composition in the needles (Figure 1).
There is no linear response between δ13C, Δ14C and space localization of the distance of sampling sites from factories. The analysis of 14C in trees shows regional variability and differences between each of the sites investigated. On the one hand, the variability in carbon isotope composition of the biosphere may be related to CO2 emissions from fossil fuel combustion, which increases the atmospheric CO2 concentration and alters its isotopic composition (e.g., Craig Reference Craig1954). In terms of 14C, the global CO2 becomes depleted in 14C, because the CO2 from fossil fuels is devoid of any 14C (Suess Reference Suess1955). Local depletion of the radioactivity concentration (local Suess effect) can be observed in areas with higher CO2 emissions from fossil fuel combustion (Rakowski et al. Reference Rakowski, Pawełczyk and Pazdur2000; Molnar et al. Reference Molnár, Bujtás, Svingor, Futó and Světlík2007; Pazdur et al. Reference Pazdur, Nakamura, Pawełczyk, Pawlyta, Piotrowska, Rakowski, Sensuła and Szczepanek2007; Svetlik et al. Reference Svetlik, Povinec, Molnár, Meinhardt, Michálek, Simon and Svingor2010; Baydoun et al. Reference Baydoun, Samad, Nsouli and Youness2015).
The sites selected for investigation were in the proximity of recognized CO2 emitters; therefore, the lack of a pronounced Suess effect in two of the sites may seem surprising. This phenomenon might be explained by the factors associated with the local carbon cycle and pine physiology because carbon compounds synthesized in the fall and stored during the winter are utilized when trees start radial growth in the spring. The pines used a material from previous year to create new tissues and the first layer of early-wood (Białobok et al. Reference Białobok, Boratynski and Bugała1993). Evergreen leaves may contain carbon fixed from the previous season or even previous years. The significant scatter of 14C concentration (up to 70‰) in the plants growing polluted area comparing to the 14C concentration in the plants growing in the clean area has been observed also by Baydoun et al. (Reference Baydoun, Samad, Nsouli and Youness2015). Also the results of the systematic Δ14C measurements of soil CO2 flux performed in the forest environment for 1998–2001, Gorczyca et al. (Reference Gorczyca, Kuc and Różański2013) from closed-system gas collection at the soil surface suggested that Δ14C values of the CO2 emitted from the forest soil were ~40‰ higher than current atmospheric background values. Moreover the influence of bio-components combusted together with fossil fuels cannot be excluded at the moment, as additional research is required. This effect has also been observed in Niepolomice Forest (Pazdur et al. Reference Pazdur, Kuc, Pawełczyk, Piotrowska, Sensuła and Różański2013). Relatively high 14C concentration may be also connected with the other effect connected with the emission of 14CH4 by factories, that may raise the 14C concentration in the atmosphere (Molnar et al. Reference Molnár, Bujtás, Svingor, Futó and Světlík2007).
The “masking effect” caused by the carbon cycle is undoubtedly present in all investigated sites, although it may have a different extent, depending of the local soil characteristics. In general, it may even cause the Δ14C in tree rings to reach values above Δ14Co, which was observed in a few cases. The values of Δ14C in tree rings and needles lower than Δ14Co, i.e. indicating a local Suess effect, were observed in some samples, especially in the Laziska site (LA). The emission of 14C-free CO2 from the nearby heat and power plant may be assumed to be higher than that from other industrial sites (chemical factory for KK or steel factory for HK), and this is consistent with the presented data.
The effect of the nearby Katowice agglomeration, possibly causing an enhanced Suess effect, was not observed in HK. In fact, a local Suess effect in cities has been previously observed for sites close to city centers (Rakowski et al. Reference Rakowski, Pawełczyk and Pazdur2000; Pazdur et al. Reference Pazdur, Kuc, Pawełczyk, Piotrowska, Sensuła and Różański2013). However, in our investigation, some sampling sites, Δ14C in trees (cellulose and needles) is higher than current atmospheric background values. A higher Δ14C concentration in tree-rings than in the “clean” air has also been observed in specific years (between 1970s and 1990s) in urban areas by Rakowski et al. (Reference Rakowski, Pawełczyk and Pazdur2000). The variation in δ13C and Δ14C has been noted in all investigated sites. Differences have been noted between the values of Δ14Co in atmospheric CO2 and Δ14C in tree ring cellulose; some samples of tree rings and needles have a more positive value than Δ14Co in atmospheric CO2. This phenomenon might be explained by the factors associated with the local carbon cycle and additional research is required. The influence of different sources of carbon isotopes connected with the effects of roads, housing energy, household heating sources, low and high stack emission including methane emission by factories, bio-components combustion cannot be excluded at the moment.
ACKNOWLEDGMENTS
The authors wish to express their sincere gratitude to everyone who contributed to making these investigations possible, particularly Magdalena Opała and the technical staff of Silesian University of Technology, who helped in the sample collection and technical work.
This work was supported by National Science Center, Poland [grant number DEC-2011/03/D/ST10/05251]; the Ministry of Science and Higher Education [grant number BKM-507/RIF/2013, BKM-509/RIF/2014 and BKM-513/RIF/2015]; the rector’s grant in the area of research and development, Silesian University of Technology [grant number RGJ-45/RIF/2017] and Silesian University of Technology grant [grant number 14/990/RGH17/0096]
SUPPLEMENTARY MATERIAL
To view supplementary material for this article, please visit https://doi.org/10.1017/RDC.2018.59
