Scots pine is a very useful archive of changes in ecosystems, because it has a specific sensitivity to local environmental conditions, including climate. (Schweingruber, 1985; Richter
Year-on-year variations in tree-ring widths are mainly caused by meteorological conditions. Industrial pollution impairs the sensitivity of trees to climatic factors. However, the trees retain their year-on-year tree-ring width changes, even during periods of high-pollution emissions.
Experiments and analyses of the stable isotope composition of annual tree rings conducted by various scientists (e.g., Craig, 1954; Burk and Stuiver, 1981; Leavitt and Long, 1982; Edwards and Fritz, 1986; Ehrelinger and Vogel, 1993; Farquhar and Lloyd, 1993; Yoder
A combination of several independent indicators constitutes a useful tool in environmental analysis, such as in dendrochronological and climatological research. Scientists attribute the variations in the stable isotope composition to climate changes, and also to anthropogenic effects (Gagen
According to Saurer and Siegwolf (2007), the stronger response of photosynthetic capacity (A) indicates that some species could enhance biomass accumulation due to the increasing CO2 during the last decade, whereas other species responded more strongly with reduced stomatal conductance and less transpiration and water loss. According to scientists (Farquhar and Lloyd, 1993; Ehlelinger and Vogel, 1993; Scheidegger
The observed anthropogenic impact on the carbon cycle is mainly related to various global industrial activities (Martin
The climate-radial growth relationships were analysed for the period 1951–2012, whereas the climate-stable isotopic relationships were analysed for the period 1975–2012. The software programs Statistica 12 (Statsoft Inc. 2014, Kraków, Polska) and DendroClim2002 (Biondi, 1997) were used for the statistical analyses.
The dendroclimatological sampling reported in this paper included 16 pine sites in three regions – Dąbrowa Górnicza near Huta Katowice (HK), Kędzierzyn-Koźle (KK), and Łaziska (LA). The sampling sites were located at varying distances from industrial factories (distance range: 1–20 km from the factories,
The localisation of the sampling sites.Stand (region) Lab code Geographical coordinates Kędzierzyn-Koźle (KK) KKT 50° 18′N; 18°20’E KKS 50° 18′N; 18°17′E KKB 50°21′N; 18°18’E KKA 50° 19′N; 18°15’E KKC 50°20′N; 18°19’E KKN 50°22′N; 18°23′E Laziska (LA) LAW 50°8′N; 18°53′E LAM 50°9′N; 18°56′E LAP 50°10′N; 18°58′E Dabrowa Gornicza (HK)HKB2 HKE 50°20′N; 19°23′E HKBB 50°24′N; 19°22′E HKB2 50°24′N; 19°21′E HKFF 50°21′N; 19°19′E HKF2 50°24′N; 19°28′E HKC2 50°26′N; 19°29′E HKD1 50°31′N; 19°31′E
At each site, 20 pine trees were sampled by taking one increment core per tree at a height of 1.3 m above ground. All 320 tree samples were dominant and co-dominant individuals without damage. The examined pine stands were located in similar habitat conditions around the industrial factories. All the pine stands were classified as fresh mixed broadleaved forest and aged between 80 and 100 years.
Meteorological data were provided by the Polish Institute of Meteorology and Water Management (IMGW-PIB). The temperature, humidity, sunshine duration and precipitation data were obtained from the meteorological stations in Katowice and Opole.
The period from 1951 to 2012 was characterised in the regional climate records by the following: an annual mean temperature of approximately 9°C (annual range of 6.5 to 10.5°C/yr), a mean annual precipitation of around 610 mm (annual range of 360 to 870 mm/yr), mean annual number of sunshine hours of approximately 1530 h (annual range of 1108 to 1978 h), and relative humidity of around 80% (annual range of 75% to 86%). The lowest precipitation was observed between the mid-1980s and the mid-1990s (
The tree-ring widths were measured to the nearest 0.01 mm, and were absolute-dated and rechecked using the COFECHA computer program (Holmes, 1983). The standardisation process of the tree-ring widths eliminates their medium- and long-term variability, and emphasises short-term variations. The tree-ring width series standardisation removes the non-climatic (long-term) trends in growth associated with increasing tree age – the major source of non-stationarity in the time series. Accordingly, the series of the annual sensitivity indices (
where
Then, the site chronologies were determined on the basis of the tree series. The indexed site chronology revealed the short-term variance due to the variation in climatic factors. For each pine stand, the following indicators were calculated:
To estimate the dendrochronologically uniform regions for pine, a cluster analysis of the indexed site chronologies (Ward’s method and 1-
In this study, we assessed the stability of the relationship between tree-ring cellulose δ13C and δ18O and the climate over the period 1975–2012 for Scots pine. The chronologies were based on a pooled-ring approach, with 10 trees per 3 sites (LAP in the LA region, ZKC in the KK region, and HKF2 in the HK region). The absolute-dated annual tree rings were manually separated as thin slivers, and then pooled (equally weighted per tree) and homogenised. The α-cellulose samples were extracted using Green’s method (1963), with further modifications (Pazdur
The carbon and oxygen stable isotope compositions were measured by IRMS (Isoprime, GV Instruments, Manchester, UK) in the Institute of Physics, the Silesian University of Technology, Poland. We reported the isotope values in the delta notation as follows (
for carbon (δ13C) and oxygen (δ18O) in respect of the international standard, which was Vienna Pee Dee Belemnite (VPDB) for carbon and Vienna Standard Mean Ocean Water (VSMOW) for oxygen.
To test the stability of the climate-isotope relationship over the period 1975–2012, we performed correlation analyses. The analysis of the temporal stability of the climate proxy was based on forward evolutionary intervals (base length:
where
The indexed site chronologies showed a high similarity of year-on-year processes and indicated the strong dynamics of the size of the radial growth of pines in various sites in certain years (
The rbt indices ranged from 0.399 to 0.521 (
No. of trees and cores HKE 20 0.279 0.886 7.7 0.255 37 HKBB 20 0.322 0.905 9.5 0.219 47 HKB 20 0.445 0.941 16.0 0.236 60 HKFF 20 0.300 0.896 8.6 0.283 43 HKF 20 0.334 0.909 10.0 0.286 45 HKC 20 0.545 0.960 23.9 0.226 66 HKD 20 0.227 0.855 5.9 0.265 36 LAW 20 0.518 0.953 21.5 0.284 64 LAP 20 0.399 0.930 13.3 0.285 51 LAM 20 0.514 0.954 21.2 0.308 63 KKA 20 0.361 0.918 11.3 0.204 49 KKB 20 0.421 0.935 14.5 0.197 54 KKC 20 0.355 0.916 11.0 0.201 48 KKN 20 0.319 0.903 9.4 0.158 46 KKS 20 0.424 0.936 14.7 0.191 55 KKT 20 0.371 0.922 11.8 0.210 50
The indexed site chronologies were correlated with each other within a common period from 1951 to 2012. The similarity between site chronologies decreased with distance, and was high within each of the considered regions; further, it was often high between sites from various regions (
To identify the reaches of the dendrochronologically uniform regions more clearly, a PCA was used. The first component (PC1) accounted for 65% of the variance among the site chronologies. All the indexed site chronologies correlated highly with PC1 (
We assumed that any characteristic incremental rhythm of pines in each region would be shaped by climatic factors. To test this, we correlated the indexed site chronologies with the climatic parameters. Thus, we obtained a series of correlation coefficients for each pine population (
The variability of climate conditions from year to year mainly determined the short-term variance of the radial increment of the trees. The first two components described the climate elements which had a significant impact (77%) on the radial-growth variability of the pines. The climate factors described by PC1 and PC2 were identified by a correlation-function analysis (
The first component also correlated with the February and March temperatures and the May and July precipitation values of the year of the ring formation. The climatic factors described by PC1 had a similar and significant impact on the radial growth of all Scots-pine populations. PC2 differentiated the indexed site chronologies and correlated significantly with the monthly precipitation of June and the July temperatures (
The pattern of the spatial and temporal variability of δ13C and δ 18O in the samples extracted from the annual tree rings of pines growing in the three regions, namely LA, HK, and KK, is illustrated in
The average value of pine δ13C in each sampling site was approximately 23.9‰, whereas the average value of pine δ18O in each sampling site was approximately 29.7‰. However, different trends in δ13C and δ18O were observed in each sampling site. The typical decline in δ13C due to the global atmospheric trend could be observed after 1990. It is possible that the typical decline in δ13C prior 1990s might have been masked by another factor, or factors. In KK, the pine δ13C series data ranged from –24.6‰ to –23.2‰ (uncorrected) and from –23.1‰ to –21.9‰ (corrected); in LA, the pine δ13C series data ranged from –24.4‰ to –23.1‰ (uncorrected) and from –23.4‰ to –21.6‰ (corrected); and in HK, the pine δ13C series data ranged from –25.3‰ to –22.9‰ (uncorrected) and from –23.6‰ to –21.5‰ (corrected). The mathematical correction for the atmospheric δ13C trend to remove the effect of the changes in δ13C was carried out using a dataset of high-frequency measurements (McCarroll and Loader, 2004; McCarroll
The correlations between the values of δ13C for three forests and the correlations between the values of δ18O for three forests: Dąbrowa Górnicza near Huta Katowice (HK), Kędzierzyn-Kożle (KK), and Łaziska (LA) (all values are significant at p < 0.05).HK - 0.61 0.56 KK 0.61 - 0.66 LA 0.56 0.66 - HK - 0.34 0.51 KK 0.34 - 0.52 LA 0.51 0.52 -
In certain periods, a significant positive correlation between δ13C and δ18O was found, whereas in other time periods, this correlation was not significant (
Across all the sites, elevated anthropogenic carbon-dioxide emission increased the 13C-derived water-use efficiency on average by 40% (
The resulting isotopic chronologies were correlated against meteorological parameters from the nearest meteorological station in Opole (for KK) and in Katowice (for LA and HK) (
In the case of δ13C, the chronologies from each site (
On the other hand, the pines growing in the KK region differed in their isotopic response to humidity. A relationship between humidity and the carbon isotopic composition revealed a high correlation in spring and summer, and the influence of the humidity of the previous year could be noted. Further, the response of the pines to precipitation was different in different sites; however, some similarities were observed for the pines growing in the HK and KK regions, whereas a relatively high insolation increases δ18O. The pines growing in LA showed the highest correlations and the strongest climate-oxygen isotopic composition of pine.
In the case of δ18O, the values in the chronologies from each site (
To test the stability of the climate-isotope relationship over the period 1975–2012, we performed correlation analyses, and
For the δ13C pines, correlations with the precipitation amount were negative for summer (July or August). For the pines growing in LA, the correlations with precipitation were temporally unstable. The δ18O correlations with precipitation were temporally unstable in KK and HK, whereas in LA the negative correlation with precipitation in August was strong and stable. Correlations between sunshine and both isotopes were mostly positive for the growing-season sunshine. In general, for δ13C, sunshine in July was significant for determining the δ13C signal for all the three relationships studied, and sunshine in summer (June-September) was important for determining the δ13C signal in LA and HK. For δ18O, sunshine in summer (June-September) was important for determining the δ18O signal in all the investigated sites. In contrast, for pines growing in LA and HK a strong significant correlation was observed with the sunshine spread over much of a given year. The significant correlations of δ13C with humidity were predominantly negative for the summer months (June-September) in all the investigated sites. The sign of the correlation in December changed from negative (in LA and KK) to positive (in HK). The most consistent significant correlations with humidity determined the δ13C signal in pines growing in LA and HK. For δ18O pines, the correlations with humidity were unstable in HK and KK, whereas in LA, the correlation was stronger and more stable. The significant correlations of δ18O with humidity were positive for the months of the previous year (November-December) and negative for the months of the current year (June-September). Relationships between climate and the isotopic composition exhibited spatiotemporal diversity. A combination of parameters can impact on the isotopic composition of pine (Leegood and Edwards, 1996; McCarroll and Loader 2004; Sensuła and Wilczyński, 2017).
The similarity of the incremental reactions (year-on-year) of pines from various sites was relatively high. Therefore, a supra-regional factor was determined. We assumed that this was a climate factor. The answer was provided to us by the PCA carried out later in this study.
The mean correlation coefficient, expressed population signal, signal-to-noise ratio, and variance in the first component calculated for each pine population, indicated a strong climatic signal in the indexed site chronologies. The mean sensitivity implied a strongly differentiated susceptibility to short-term environmental factors. One such factor that changed from year to year was climatic conditions. The variance explained by the first eigenvector indicated a high similarity of the radial growth reactions of the pines.
The greatest distance between the pine sites was 66 km. The conifer chronologies in North America correlated at the 99% level over a distance of 1000 km (Cropper and Fritts, 1982), and in Spain over a distance of 500 km (Richter
The variability of climate conditions from year to year determined the short-term variance of the radial increment of trees. The first two components described the climate elements which had a significant impact on the radial growth variability of pines. The climatic factors as described by PC1 had a similar and significant impact on the radial growth of all Scots-pine populations. PC2 differentiated pine sites. The pines in each region reacted differently to factors described by the second component.
Our results implied that high July precipitation and high winter temperatures were beneficial to the radial growth of the pines. Under drought stress in the vegetation season, the trees had relatively low cambial activity, thereby reducing the cambial division and cell-wall thickening (Irvine
However, low precipitation in May had a positive impact on the radial growth of the pines. Low precipitation indicates low cloudiness which does not limit solar radiation. Therefore, we concluded that solar radiation could stimulate cambium activity.
The temperature in winter was the main limiting factor for the activity of pines. In late winter and early spring, a relatively high air temperature initiated physiological and biochemical processes faster, and caused vascular cambium activation earlier (Gricar
We have observed that a wet and cloudy previous September had a positive impact on the radial increment of pines in all sites. Such weather conditions are favourable for the carbohydrate accumulation by trees used in the next year to build wood cells (Hoch
Different effects on the growth of the pine trees from individual regions (KK, LA, and HK) were observed in the cases of the precipitation in June and the July temperature. Scots pine from HK was more sensitive to a low temperature in July and a low precipitation in June in the year of the ring formation. This could be attributed to the cooler and drier climate in this region.
The correlation analysis yielded mostly expected results. Positive relationships with temperature and sunshine, and negative relationships with precipitation and humidity, were observed for δ13C. In contrast, a positive relationship with temperature and a negative relationship with humidity were observed for δ18O. However, it has been noted that the correlations showed certain temporal instability differently in each of the investigated sites. The current growing-season climate (from April to September) dominated the δ13C and δ18O signals, despite the significant relationships found for several months from the previous year. As far as δ18O is concerned, there might be a depth-dependent temporal offset resulting from an effective uptake of the soil water by the roots (Saurer
The isotopic composition of pines growing in LA has been the most sensitive to climate changes, and the chronologies for both elements reflect the largest number of climatic parameters.
The differences in the relationship between climate and the carbon and oxygen stable isotope composition of the trees growing in different areas can be attributed to the effect of the masking of the climatic signal by pollution (Leonelli
The climate-isotope composition of tree-ring cellulose relationships is not stable over time. The previous analysis shows that from 1950 to 2000, the correlation coefficients between δ13C and the climate factors in the Niepolomice Forest (Poland) were not stable over the entire studied period (Sensuła
This study revealed that the Scots-pine populations did not show radial-growth differences within the same climatic microregion, and could be considered dendrochronologically homogeneous. The delineation of three dendrochronologically homogeneous regions was important for the dendroclimatological work. The similarity of the short-term signal in the tree-ring width decreased with the distance between regions, but was often high between sites from various regions. Therefore, the geographical criterion did not sufficiently explain the grouping of the pine populations.
The long and frosty winters had a negative impact on the radial growth of the pines in all the regions. Pine trees were also sensitive to a shortage of rainfall during July, the period of intense vascular-cambium division. However, Scots pine also demonstrated characteristic sensitivity to the climatic conditions of the region in which it grew. The specific impact on the radial growth of the pines depended on the pluvial condition in June, and the thermal condition in July, in each of the three regions.
The present investigation also showed differences in the relationship between climate and the stable isotopic composition of pine populations growing in the same climatic microregion. The importance of the current growing-season climate was revealed, and the impact of the other months was detected. The signals of climate changes recorded in the stable isotope composition of tree-ring cellulose was not stable over time. Relationships between δ13C and temperature were positive and between δ13C and sunshine positive, whereas relationships between δ13C and precipitation were negative and between δ13C and humidity negative. The relationship between δ18O and temperature was positive, and between δ18O and humidity negative. These relationships were expected in the light of the theoretical backgrounds. However, the correlations showed some variations between the investigated sites. The effect of the climatic isotope signals in the tree rings can be masked by the biosphere contamination.