Science One Research Projects 2010-2011

Climatic and atmospheric influences on carbon isotope ratios stored in tree rings Fuglem, Naomi 2011

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


Fuglem_N_et_al_Sci_One_Prog_2010-2011.pdf [ 1.47MB ]
JSON: 1.0107226.json
JSON-LD: 1.0107226+ld.json
RDF/XML (Pretty): 1.0107226.xml
RDF/JSON: 1.0107226+rdf.json
Turtle: 1.0107226+rdf-turtle.txt
N-Triples: 1.0107226+rdf-ntriples.txt

Full Text

JUNE 2011 CLIMATIC AND ATMOSPHERIC INFLUENCES ON CARBON ISOTOPE RATIOS STORED IN TREE RINGS NAOMI FUGLEM & KRISTINA NELSON SCIENCE ONE PROGRAM THE UNIVERSITY OF BRITISH COLUMBIA VANCOUVER, CANADA 1 2NAOMI FUGLEM & KRISTINA NELSON SCIENCE ONE PROGRAM THE UNIVERSITY OF BRITISH COLUMBIA VANCOUVER, CANADA Abstract The δ13C values of 30 individual tree rings from a Douglas Fir (Pseudotsuga menziesii) were analyzed. These data were compared with local climatic data, total tree ring size and late- wood/earlywood ratio to determine the degree of correlation. The results provide indication of correlation between different aspects of tree growth and precipitation. A yellow cedar (Callitropsis nootkatensis) was also analyzed in ten year increments for long-term δ13C trends, which may be representative of atmospheric δ13C trends. Introduction Dendrochronology, the dating of tree-rings, has been used to examine historical climatic condi- tions and other environmental factors affecting tree growth. In the past, many studies have focused on ring width or growth rate. Growth rate has been correlated with precipitation, temperature, and snow-pack [1]. Researchers have been working to accurately model the effects these conditions have on ring width, so that tree rings can be used in paleoclimate reconstructions, revealing his- torical weather conditions. The process is complicated by the varying importance of each factor in different ecosystems; some ecosystems were found to be precipitation-limited, while others were more influenced by energy availability [2]. Recent research has been done to examine the carbon-13/carbon-12 ratio of tree rings and search for correlations with the historical climate. Relationships have been found between the carbon isotope ratio and temperature, humidity [3], precipitation, light, degree of canopy closure [4] and growth increment [5]. Furthermore, the carbon-13/carbon-12 ratio in each ring has been linked to the atmospheric CO2 level during ring formation [5]. This is because the burning of fossil fuels releases carbon originally fixed by C3 plants. These plants select for the lighter carbon isotope, carbon-12, and so fossil fuels contain predominately carbon-12. Burning them thus lowers the atmospheric carbon-13/carbon-12 ratio. This relationship between atmospheric carbon and tree ring carbon could allow us to reconstruct historical atmospheric CO2 levels. However, other factors also affect the carbon isotope ratio stored within tree rings. During photosynthesis the enzyme Rubisco converts atmospheric CO2 into carbohydrates, some of which are then stored in the tree as cellulose and lignin. The resulting carbon-13/carbon-12 ratio of the annual rings is affected by two main processes: CO2 entering the leaf and its subsequent selection by Rubisco. Under normal conditions, Rubisco selects strongly for the lighter, and thus more reactive carbon isotope, carbon-12. However, during periods of low precipitation the stomata, or pores, of the leaf close to reduce water transpiration. This limits the flow of CO2 into the leaves. With less CO2 available within the leaf, Rubisco does not select as strongly for carbon-12, and thus less discrimination against carbon-13 is found in the tree rings. In this study, the ring-sizes and carbon ratios of coastal Douglas fir (Pseudotsuga menziesii) and yellow cedar (Callitropsis nootkatensis) and were analyzed and compared with weather data and atmospheric CO2 levels. Since the atmospheric carbon-12 levels have been decreasing, rings from the yellow cedar covering 120 years were examined for a decaying trend. Regarding the Douglas fir, previous studies have found inconsistent results concerning the carbon ratio to temperature and carbon ratio to precipitation relationships [6]. This could indicate that such relationships are climate or species-specific. For these reasons our study simply aimed to further the understanding of the climates effect on carbon storage in trees. Methods A Douglas fir was sampled from Thetis Lake Regional Park at latitude 48.47181 N, longitude -123.48523 W, and altitude 66m. The tree was a dominant crown located on a >100% East-facing slope, with extremely shallow, rocky soil where precipitation has an increased influence on tree growth. The surrounding vegetation was dominated by arbutus and Douglas fir. Using 12 and 16 JUNE 2011 CLIMATIC AND ATMOSPHERIC INFLUENCES ON CARBON ISOTOPE RATIOS STORED IN TREE RINGS3 inch increment borers, cores were extracted from chest height (1.3m) on the North, West, East and South sides of the tree. The cores were dried for an hour on a wood-stove, and then sanded lightly. All samples were photographed individually beside a tape measure with a Canon 5D camera. A double strobe lighting set up was used to minimize any shadowing effects A further sample was taken from an 870 year old yellow cedar stump cut roughly 25 years ago from Northern Vancouver Island. The sample was extracted using a carpenter Skilsaw. The saw blades were run along either side of the stump centre to remove a narrow (approx. 1.5 cm) triangular prism. The cores from the Douglas fir were compared to determine the tree’s chronology and then separated into individual rings using an exacto knife. These rings were combined with rings from the other three directions for each year. The yellow cedar sample was similarly separated into ten year intervals. Each sample was ground into a fine powder by one of two methods: drilling through the center of the ring with a dremel drill, or using a fine wood file. To eliminate contaminants, the Douglas fir samples underwent an AAA pre-treatment. The samples were placed into test tubes and with 2-3 mL of 0.5M HCl and heated to 95◦C for 1 hour to remove carbonates absorbed from the groundwater. The samples were transferred into Falcon tubes to be centrifuged at 4500 rmp for 20 min. The HCl was poured off of the wood pellets and the samples were transferred into test tubes where the process was repeated with 0.5M NaOH. This was done to remove any humic acid contaminants from the soil. Then the process was repeated a final time using 0.5M HCl to remove any CO2 absorbed from the air during the NaOH procedure. The samples were then rinsed and centrifuged to remove remaining acid. Finally, the samples were freeze-dried in a Labconco freeze dryer for 2-3 hours. However, as measurements indicated that the samples still contained moisture, a microcentrifuge tube stand was constructed to hold the samples for further freeze drying. Approximately 0.600 mg of each sample was weighed out on an analytical balance. The samples were put into small tin boats, folded into pellets and analyzed with an Isoprime Mass Spectrometer. In the mass spectrometer, the samples were combusted to vaporize the carbon stored inside the tree- rings. The vaporized carbon was ionized and then accelerated through a magnetic field. All the carbon ions experience an equal force from this magnetic field as they have equal charges. However, carbon isotopes have differing masses and the lighter isotope, carbon-12, was deflected to a greater degree. Thus, the mass spectrometer could measure the ratio of particles by the number hitting at calibrated detection positions. The data from the mass spectrometer was in the form of δ13C values. These represent a percent deviation in carbon ratio from a standard, as calculated with the following equation: (1) δ13C(0/00) = Rsample−Rreference Rreference × 1000 0/00 where Rsample is the carbon-13 to carbon-12 ratio found in our samples, and Rreference is the ratio of the standard used for comparison. Due to the extensive pre-treatments required and the time constraints of the project, only 30 annual tree rings from the Douglas fir and the 15 most recent sections from the yellow cedar were analyzed. With the Douglas fir, to ensure there was enough sample for replicates, the annual rings that produced only a small amount of sample were removed. Although this introduced a potential source of bias, the mass spectrometer is unable to produce reliable results for samples below a certain carbon content. Of the samples with sufficient mass, every second sample was analyzed. From each of the 30 samples, two replicates were measured by the mass spectrometer (except for three samples where there was not enough sample). For the yellow cedar, first the 12 most recent samples were analyzed for their δ13C values. Then a further 3 samples were analyzed with the Douglas fir samples. The two δ13C values produced for each Douglas fir and yellow cedar sample were averaged using a weighted mean based on the corresponding uncertainties with the equation: (2) x = ΣwixiΣwi 4NAOMI FUGLEM & KRISTINA NELSON SCIENCE ONE PROGRAM THE UNIVERSITY OF BRITISH COLUMBIA VANCOUVER, CANADA where wi is the uncertainty in each measurement xi. The uncertainty in each replicate was determined using standard deviations calculated for stan- dards run alongside our samples in the mass spectrometer. Since multiple standards were used, the average of the standard deviations was taken to represent the uncertainty for each set of replicates. Then, using the following equation: (3) σ22 = 1 Σ 1 σ21 σ2, the new standard deviation of the weighted mean, was found. σ1 represents the standard deviation of the original measurements. The photos taken prior to sample analysis were used to measure the growth increments of the tree. Using Adobe Photoshop, the contrast of each photo was increased to better distinguish each annual ring. With the ruler function, the pixel to mm ratio of each photo was determined from the tape measure photographed alongside each core. Using the same tool, the width of each ring was selected and converted into millimeters. The distances between annual rings were generally clearly marked by the transition from the latewood of one year to the earlywood of the next. A similar process was used to find the size of solely the latewood. However, this measurement proved more difficult due to the lack of a clear division between the earlywood and the latewood within an annual ring. Therefore this division was better defined using the Adobe photoshop color replacement tool, which highlighted the difference between the earlywood and latewood sections of the core (figure 1). Figure 1: Example of photos used in visual analysis of tree rings. The colour of the earlywood has been replaced here with Adobe Photoshop’s colour replace tool in order to make the earlywood to latewood division more distinct. Also visible, the ruler within the photo used to convert to actual tree ring size. Precipitation and temperature data were collected from the Environment Canada weather archives (available online). After comparison between numerous Southern Vancouver Island weather stations, the Victoria International Airport station data was chosen for this analysis. This was due to the continuity of the weather data from 1941 to 2006 and the similarity of the precipitation and tem- perature measures to the other weather stations in proximity to the sample location. The daily precipitation was summed by month, and into May/June and June/July combinations for analysis. These combinations were chosen because past research shows significant correlation between growth increments and early summer weather conditions [1]. Temperature data was converted into growing degree days (GDD), a measure of the growing season length. The following equation was used for the conversion: JUNE 2011 CLIMATIC AND ATMOSPHERIC INFLUENCES ON CARBON ISOTOPE RATIOS STORED IN TREE RINGS5 (4) GDD = ΣTmax−Tmin2 − Tbase where Tmax and Tmin are the maximum and minimum temperatures for each day and Tbase is the baseline temperature. To find the GDD, the average temperature above the baseline temperature is calculated for each day, and then these values are summed for the entire year. The baseline temperature used in calculating the GDD in this study was 10◦C [8]. The historic atmospheric δ13C data was found in a previous study that measured the δ13C in Antarctic ice cores [9]. These values were taken to represent the atmospheric δ13C for comparison and detrending of the δ13C data in this report. Since the δ13C values determined for the Douglas fir showed significant atmospheric trends (p<0.01), the data was detrended before climatic and growth increment correlation analysis. This was accomplished by creating a model of the atmospheric δ13C values from the Antarctic ice cores plotted against time. A third degree polynomial equation was created to fit the data as follows: (5) δ13C atmos = −1.7524 × 10−6(t)3 + 1.0075× 10−2(t)2 − 1.9312× 101(t) + 1.2335× 104 where δ13C atmos is the atmospheric δ13C level, and t is time in years since 0 A.D. The model values were then subtracted from the δ13C data for each annual ring of the Douglas fir. Thus, the detrended values of δ13C for the Douglas fir should represent the discrimination of carbon-13 by the tree alone (as influenced by climatic stress) (figure 2). Figure 2: Mean δ13C values found in Douglas fir rings plotted by age (blue). Detrended δ13C values (green) are plotted over the same time scale. There was also a decreasing trend in the growth increments measured for the Douglas fir. This age trend is commonly found in dendrochronology. As the tree ages the growth of the tree must be spread over a larger circumference, and thus the annual ring sizes decrease. To remove this trend, the growth increments were converted into tree ring indices. To do this a model was found of the original data by minimizing the chi-square value. Then this model was subtracted from the growth increment measurements. The resulting values (termed ring indices) are a measure of how far each growth increment lies from the trend. 6NAOMI FUGLEM & KRISTINA NELSON SCIENCE ONE PROGRAM THE UNIVERSITY OF BRITISH COLUMBIA VANCOUVER, CANADA Results 15 samples from the yellow cedar were analyzed, each containing approximately 10 annual rings of the tree (roughly covering years 1835 to 1985). These values show a decreasing trend similar to that of the atmospheric δ13C data. They peak, however, in the 1960s interval, where atmospheric δ13C values are at a minimum. The correlation coefficient between these samples and atmospheric ice core data was determined to be -0.15 (figure 3), when only considering the 12 most recent 10-year intervals from the yellow cedar. The later 3 data points cannot be compared along with the first 12 data points. The standards run along with both sets of data showed slightly different δ13C values, thus the yellow cedar values may also be altered. However, if the last 3 intervals are included, the correlation was stronger (r = 0.37). Of the first 12 samples, the standard deviation was 0.1 for the first replicates and 0.3 for the second replicates. The last 3 samples were run with the Douglas fir samples, and had standard deviations of 0.03 for the first replicates, and 0.05 for the second replicates. The Pearson correlations coefficients were determined for numerous pairs of data sets. The precipitation was analyzed against the average growth increments, latewood increments, earlywood increments, latewood to earlywood ratio, ring indices, average δ13C values and detrended δ13C values of the Douglas fir (figure 4). In these comparisons, the precipitation data for each month was analyzed separately and in combinations of May/June and June/July to find the highest correlation. The δ13C values were also compared with growing degree days (GDD), the average ring increments, ring indices, and latewood to earlywood ratio to determine the level of correlation. The standard deviation of the Douglas fir δ13C values was found to be 0.03 for the first replicates, and 0.05 for the second replicates. Figure 3: Yellow cedar δ13C values plotted over atmospheric δ13C from Antarctic ice cores in corresponding ten year intervals. The data covers 150 years. A non-significant correlation coefficient was found to be 0.37 (p>0.05). Light green triangles represent the samples run on a separate day. JUNE 2011 CLIMATIC AND ATMOSPHERIC INFLUENCES ON CARBON ISOTOPE RATIOS STORED IN TREE RINGS7 Figure 4: The correlation coefficients between monthly precipitation and various tree ring measurements of the Douglas fir are shown. The measure of correlation, r, was determined using the Pearson correlation coefficient. For the growth measurements (left 5 data sets), when r>0.248, p<0.05 and when r>0.322, p<0.01. For the δ13C values (right 2 data sets), when r>0.361, p<0.05, and when r>0.463, p<0.01. The average latewood increments showed the greatest correlation to precipitation. March, May/ June, June, June/July, and July precipitation all correlated with p<0.01. The highest level of correlation was found for June/July at r=0.68. The earlywood increments showed no significant correlation to precipitation. The earlywood to latewood ratios were correlated less to monthly pre- cipitation than the latewood increments though the same months showing correlation with p<0.01. March precipitation was an exception, being much more strongly correlated with the latewood to earlywood ratio. The average growth increments and the detrended growth increments showed sim- ilar peak correlations, however, in general, the detrended data had higher correlation coefficients. The mean δ13C values were strongly correlated with the precipitation from May (p<0.01), and with precipitation from March and November (p<0.05). The detrended δ13C values showed signif- icant correlation in the same months, with the November correlation even more negative (p<0.01). Additionally, the May precipitation had r=-0.58 in the detrended δ13C values, indicating a higher negative correlation than with the mean δ13C values (r=-0.55). Of the monthly precipitation analyzed, May precipitation was found to correlate the most strongly with the δ13 in the tree rings (figures 5 and 6). Since the detrended δ13 showed more correlation than the original average, it was graphed against the May precipitation using a basic scatter plot. A chi-square test was used to create a linear model best representing the data points: (6) δ13 = −0.015× precip− 15.8 where precip represents the total May precipitation of a given year. 8NAOMI FUGLEM & KRISTINA NELSON SCIENCE ONE PROGRAM THE UNIVERSITY OF BRITISH COLUMBIA VANCOUVER, CANADA Figure 5: Scatter plot of the highest correlating month’s precipitation with detrended δ13C from the Douglas fir. The correlation coefficient was found to be -0.58, with p<0.01. A linear model was fit to the data by minimizing the chi-square value. The model found was δ13C = -0.015 ×precip−15.8, with a chi-square value of 7518.7. Error bars represent standard deviation (σ) of 0.0257, except for years 2000, 1997, and 1988 where σ = 0.03. Figure 6: The May precipitation from the Victoria International Airport station plotted over the detrended Douglas fir δ13C values of the same years. The correlation coefficient found was -0.58 (p<0.01). JUNE 2011 CLIMATIC AND ATMOSPHERIC INFLUENCES ON CARBON ISOTOPE RATIOS STORED IN TREE RINGS9 Figure 7: Growing degree days were plotted against the δ13C of Douglas fir. The correlation coefficient was found to be 0.34, thus showing the data sets do not correlate significantly (i.e. p>0.05). The model equation was δ13C = 0.0008(GDD)−17.0, with a minimized chi-square value of 10 573.6. Error bars represent standard deviation (σ) of 0.0257, except for years 2000, 1997, and 1988 where σ = 0.03. The correlation coefficient between the δ13C value of the Douglas fir and growing degree days was found for the original and the detrended δ13C values. The original δ13C values had r=-0.14, showing a non-significant negative correlation. The detrended δ13C values generated r=0.34 (figure 7), corresponding to a positive correlation (p>0.05). By minimizing chi-square, a model was developed: (7) δ13C = 0.0008(GDD)−17.0 where GDD represents the annual growing degree days. Correlation coefficients were found between average growth increments and original or detrended δ13C values; similarly, the coefficients were found between tree ring indices and the two sets of δ13C values. The highest correlation was found between tree ring indices and the detrended δ13C data, with r = -0.52 (figures 8 and 9). Thus, p<0.01 and the correlation is significant. By minimizing the chi-square value, a linear model equation was found relating the tree ring indices to the detrended δ13C values as follows: (8) δ13C = −0.39 × index− 16.25 where index refers to the tree ring growth increment corrected for age-trends. 10NAOMI FUGLEM & KRISTINA NELSON SCIENCE ONE PROGRAM THE UNIVERSITY OF BRITISH COLUMBIA VANCOUVER, CANADA Figure 8: The detrended Douglas fir δ13C values plotted over the ring indices of corresponding years. The correlation coefficient was found to be = -0.52 (p<0.01). Figure 9: Scatter plot of tree ring indices against the detrended δ13C values of Douglas fir. The data have a correlation coefficient of r = -0.52, thus p<0.01. The model found was δ13C = -0.39×index−16.25, with a chi-square value of 8364.0. Error bars represent standard deviation (σ) of 0.0257, except for years 2000, 1997, and 1988 where σ = 0.03. The correlations between the latewood to earlywood ratio and the original and detrended δ13C values from the Douglas fir annual rings were determined. The level of correlation was very similar with r= -0.21 for the original mean δ13C values and r = -0.20 for the detrended δ13C values (figure 10). These results do not show significant correlation, p>0.05. Nevertheless, a model was found by JUNE 2011 CLIMATIC AND ATMOSPHERIC INFLUENCES ON CARBON ISOTOPE RATIOS STORED IN TREE RINGS11 minimizing chi-square: (9) δ13C = −0.68(lw/ew) − 15.97 where (lw/ew) represents the latewood to earlywood ratio of the Douglas fir. Figure 10: Detrended δ13C data from individual rings of Douglas fir core are plotted against the latewood to earlywood ratio of corresponding years. The correlation coefficient is r = -0.20 (p>0.05). A model was found: δ13C = −0.68(lw/ew) − 15.97 by minimizing chi-square to 12018.5. Error bars represent standard deviation (σ) of 0.0257, except for years 2000, 1997, and 1988 where σ = 0.03. Discussion Yellow Cedar The yellow cedar δ13C data from the 15 most recent ten-year intervals was plotted in figure 3 with atmospheric data for the past 150 years. Because replicates were analyzed by the mass spectrometer on two separate days, where the averages of δ13C found in various standards differed slightly, it is difficult to determine the level of correlation between the two data sets. The correlation coefficient between atmospheric and yellow cedar δ13C is -.37 when all points are considered, however becomes -0.16 when including on the 12 most recent points (p>0.05). This is likely caused by a systematic error, since higher average δ13C values were found in the standards on the days where the oldest 3 points were analyzed. Though the true correlation value between the yellow cedar and atmospheric δ13C is uncertain, even the highest possible value is not significant. The yellow cedar samples were cut from a stump of uncertain background. Its exact location of origin and growing conditions are not known exactly; thus the data collected could be subject to additional influences. The wood sample may have also accumulated contaminants while being stored for approximately 25 years in a wood working shop. Although the analyzed samples were not taken from the exposed surface of the wood, they did not undergo pretreatment and so contamination is a likely source of error. The samples were not pretreated due in part to time constraints and also because past studies have found that whole wood (i.e. not pretreated) showed very similar trends in δ13C values when compared with samples from the same tree that were pretreated to remove extractives [4]. 12NAOMI FUGLEM & KRISTINA NELSON SCIENCE ONE PROGRAM THE UNIVERSITY OF BRITISH COLUMBIA VANCOUVER, CANADA Douglas Fir: Precipitation and Ring Size Correlations Comparing the precipitation and various ring increment data, the highest correlation was found between the precipitation data and the size of the annual latewood increments. The latewood forms in the later part of the growing season primarily as a result of the growing conditions during this time of year [10]. In Victoria, precipitation acts as a limiting factor only in the late spring and sum- mer months. The precipitation of months after July showed no significant correlation because the late summer months are generally very dry resulting in reduced tree growth [11]; and the growing season comes to an end in the fall/winter months. Thus the majority of tree growth depends on the precipitation conditions in early summer months. The latewood increments reflect the amount growth for the corresponding year, and thus the highest correlation was found in the late spring to early summer months May, June and July. The earlywood increments showed no significant correlation to the monthly precipitation data because the earlywood growth is generally influenced by conditions of the previous year [10]. Since the latewood was highly positively correlated to monthly precipitation and the earlywood showed no correlation, it follows that the correlation of the ratio between these data should be influenced mostly by the latewood increments. The correlation of the latewood to earlywood ratio and monthly precipitation was less than that of the latewood alone. This is likely due to earlywood increment variation based on previous climatic conditions. The total ring increments and the tree ring in- dices also showed correlation with similar monthly trends to that of the latewood. The total ring increment is dependent on the latewood growth, thus correlates to precipitation for the reasons mentioned previously. The tree ring indices showed higher correlation than the original mean in- crements because the influence from the tree age had been removed. Douglas Fir: Precipitation and δ13C Correlations The precipitation showed a significant negative correlations with δ13C values in the tree rings, as modelled by the above graph of May precipitation against δ13C of the Douglas fir. This negative correlation is due to the response of the tree to low levels of precipitation. When in a water stressed environment, the tree closes its stomata to prevent water loss, which also decreases the amount of CO2 that can enter the leaves. When this occurs, Rubisco cannot discriminate as strongly against the carbon-13 isotope and more of this isotope is stored in the tree ring. Thus, decreased levels of precipitation cause an increase in the ratio of carbon-13 to carbon-12, and an increase in the δ13C value. The model for the May precipitation and the detrended δ13C value, equation 6, shows that with no precipitation, the δ13C stored in the tree would be -15.8 0/00. With an atmospheric δ13C of around -60/00 to -80/00, this means that Rubisco will discriminate against carbon-13 even when extremely water stressed. The Douglas fir δ13C data showed significant negative correlation with the March, May, and No- vember precipitation, the highest of which was the May precipitation. This was a surprising result due to the lack of correlation in neighbouring months. For example, April precipitation showed a slight positive correlation with the mean δ13C data. There are several possible explanations for the decrease in carbon-13 stored in the tree rings with higher precipitation for these months. Since the average temperatures during these months are generally quite mild in Victoria, it is possible that Rubisco functions most efficiently with carbon-12 isotopes at these temperatures. Thus, when there are high levels of precipitation, carbon-13 is highly discriminated against. On another note, this temperature and precipitation combination may be optimal for water retention in the tree. Without the risk of excess transpiration, the stomata may open more, allowing carbon-12 to cycle into the leaf. JUNE 2011 CLIMATIC AND ATMOSPHERIC INFLUENCES ON CARBON ISOTOPE RATIOS STORED IN TREE RINGS13 Douglas Fir: Growing Degree Days and δ13C Correlations The growing degree days showed a positive correlation to the δ13C values of the Douglas fir that is not significant. This is likely due to the relatively mild temperatures that occur in the Victoria region and the location of the tree on an steep East-facing slope. These conditions would decrease the influence of sunlight and temperature on the growth of the tree, and increase that of precipi- tation. To explain the slight positive correlation, an increase in temperature causes the stomata of the leaves to close, thus as per above, Rubisco discriminates against carbon-13 to a lesser extent. Therefore, an increase in temperature causes an increase in the δ13C value. Considering the GDD as a measure of growing season length, a higher GDD could also result in a higher δ13C in the forest atmosphere. Plants discriminate against carbon-13, so as the growing season progresses, the air in the forest slowly becomes depleted of carbon dioxide composed of carbon-12 atoms. With a higher concentration of carbon-13 in the air, the discrimination by Rubisco has a lesser effect and a higher δ13C is stored in the tree ring. The model created for the trend between δ13C and GDD, shows that an increase in GDD results in a very slight increase in δ13C stored in the tree (equation 7). Douglas Fir: Ring Indices, lw/ew Ratio Correlated with δ13C by Precip The δ13C values showed a strong negative correlation with the tree ring indices. These two measurements are both strongly correlated with precipitation. Since water is required for the tree to undergo photosynthesis among other biological processes, higher levels of precipitation allow for more growth. On the other hand, high levels of precipitation cause increased discrimination against carbon-13, and thus lower δ13C is stored within the annual tree growth. Therefore, when comparing the relative values together, a highly negative correlation can be found. This is indicated by the negative slope found by the model representing the relationship between δ13C and the ring index (equation 8). The scatter plot of the latewood to earlywood ratio against the δ13C of the tree rings shows a slight negative correlation. Although this correlation is not significant, we saw previously that the latewood to earlywood ratio increased with higher levels of precipitation. Since δ13C values are lowered by increased precipitation, this results in a negative correlation between the latewood ratio and δ13C. Sources of Error and Further Studies The experimental nature of this project allowed for several areas of possible error. For example, the data for the tree core samples were run alongside some of the initial tests of the Isoprime mass spectrometer. The machine is still in a process of calibration and as such there is some uncertainty surrounding the values found. However, that the second replicates displayed similar trends and values as the first offers some assurance of the data accuracy. Contamination of samples during pretreatment may have also had an effect on the results obtained. Since the sample size analyzed was so small (approximately 0.600 mg) any contamination could have a large effect on the carbon isotope value determined. The sources of these possible impurities include air-borne contaminants, sample residues on equipment, and several instances of test tube mix-up during the multiple transfers that occurred during pre-treatment. 14NAOMI FUGLEM & KRISTINA NELSON SCIENCE ONE PROGRAM THE UNIVERSITY OF BRITISH COLUMBIA VANCOUVER, CANADA In order to minimize possible sources of error during pretreatment, equipment was rinsed with deionized water and methanol between each sample. Aluminum foil was also used to catch and transfer wood particles for pretreating, and were disposed of after each use. Our study raised some interesting questions regarding the correlation between monthly precip- itation and δ13C data, such as the high correlation between certain monthly precipitation and contrasting results for nearby months. Because of the limited replicates and species analyzed in this study, there is need for further research in this area to verify the results found and to increase our understanding. The research in this report and future research in dendrochronology may help us to understand more about the conditions that affect tree growth. This could have important implications for looking at historical climatic trends and predicting the effect of future climate changes on trees. Research into specific carbon storage in trees may also help us better understand the carbon seques- tration process. Further studies could look into the efficiency of storing carbon-12 versus carbon-13, and possible methods of selecting more efficient forms of Rubisco using tree ring analysis. Conclusion The δ13C within the ten-year increments of a yellow cedar showed possible correlation to the atmospheric δ13C trends. However, some replicates were analyzed separately, and that may have influenced the correlation found. Although not significant, the Douglas fir showed a negative correlation between growing degree days and δ13C values within the annual growth rings. There was also no significant correlation between precipitation and earlywood widths in the Douglas fir. On the other hand, the latewood widths showed a high level of correlation to March and May to July precipitation. The ratio of latewood to earlywood widths was also positively correlated to the precipitation of these months. The δ13C values from the annual tree rings of the Douglas fir were also significantly correlated to March, May and November precipitation as well as ring indices (a measure of ring width). Acknowledgements We would like to acknowledge Brian Chisholm, Michael Richards and Liz Jarvis, who very gen- erously donated their time and lab equipment without which this project would not have been possible. We would also like to thank Rob Guy, Lori Daniels, and Sue Watts, who answered some of our initial tree questions or pointed us in the right direction. We would like to thank our mentor, Fok-Shuen Leung for acquiring access to trees (that we unfortunately did not get to sample) and for being so untroubled when we had no data. Lastly, thank you to Kai Fuglem for providing the camera set up and photography assistance, and to Peter Fuglem for assistance with the extensive weather data and excel spreadsheets. References [1] Littell, J., Peterson, D., Tjoelker, M. (2008). Douglas-fir growth in mountain ecosystems: water limits tree growth from stand to region. Ecological Monographs, 78(3), 349368. [2] Waring, R., Running, S. (2007). Forest ecosystems: Analysis at multiple scales (3rd ed.). Burlington, MA: Elsevier Academic Press. [3] Edwards, T., Graf, W., Trimborn, P., Stichler, W., Lipp, J., Payer, H. (2000). δ13C response surface resolves humidity and temperature signals in trees. Geochimica et Cosmochimica Acta, 64(2), 161-167. [4] Walia, A., Guy, R., White, B. (2010). Carbon isotope discrimination in western hemlock and its relation- ship to mineral nutrition and growth. Tree Physiology, 30(6), 728-740. JUNE 2011 CLIMATIC AND ATMOSPHERIC INFLUENCES ON CARBON ISOTOPE RATIOS STORED IN TREE RINGS15 [5] Tans, P., Mook, W. (1980). Past atmospheric CO2 levels and the 13C/ 12C ratios in tree rings. Tellus, 32(3), 268-283. [6] Mazany, T., Lerman, J., Long, A. (1980). Carbon-13 in tree-ring cellulose as an indicator of past climates. Nature, 287, 432-435. [7] El-Shaarawi, A., Piegorsch, W. (Ed.). (2002). Encyclopedia of Environmetrics (Vol. 1). West Sussex, UK: Wiley. [8] Gordon, R., Bootsma, A. (1993). Analyses of growing degree-days for agriculture in Atlantic Canada. Clim. Res., 3, 169-176. [9] Francey, R., Allison, C., Etheridge, D., Trudinger, C., Enting, I., Leuenberger, M., Langenfelds, R., Michel, E., and Steele, L. (1999). A 1000-year high precision record of δ13C in atmospheric CO2. Tellus. 51(2), 170-193. [10] Gagen, M., McCarrol, D., Edouard, J. (2004). Latewood Width, Maximum Density, and Stable Carbon Isotope Ratios of Pine as Climate Indicators in a Dry Subalpine Environment, French Alps. Arctic, Antarctic, and Alpine Research. 36(2), 166-171. [11] Sarris, D., Christodoulakis, D., Korner, C. (2007). Recent decline in precipitation and tree growth in the eastern Mediterranean. Global Change Biology. 13, 1187-1200.


Citation Scheme:


Usage Statistics

Country Views Downloads
Japan 6 0
Canada 2 0
City Views Downloads
Tokyo 6 0
Saint John 2 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}


Share to:


Related Items