SAMPLE PREPARATION AND MEASUREMENT

The collected plant materials were kept frozen until chemical preparation. The samples were cut into small pieces, then homogenized before pretreatment. The standard acid-base-acid preparation protocol was used for the purification of the samples (Southon and Magana Reference Southon and Magana2010), subsequently, the samples were dried in a heating block at 50ºC. The dried samples were combusted in sealed glass tubes at 550ºC for 12 hr, using MnO₂ as an oxidant to convert the carbon of organic material to CO₂ (Janovics et al. Reference Janovics, Futó and Molnár2018). The liberated CO₂ gas was purified from H₂O and different contaminants in a dedicated vacuum line. The pure CO₂ gas was trapped in glass tubes (with Zn, TiH catalysts and iron powder) sealed by flame and converted to graphite following the sealed tube graphitization method (Rinyu et al. Reference Rinyu, Molnár, Major, Nagy, Veres, Kimák, Wacker and Synal2013). After graphitization, the graphite samples were pressed into aluminium targets for the accelerator mass spectrometry (AMS) measurements to determine the carbon isotopic composition of the plant samples (Molnár et al. Reference Molnár, Janovics, Major, Orsovszki, Gönczi, Veres, Leonard, Castle, Lange, Wacker, Hajdas and Jull2013). For data evaluation, “Bats” data reduction software was used (Wacker et al. Reference Wacker, Christl and Synal2010). For the reporting of ¹⁴C data, pMC and Δ¹⁴C units were used (Stuiver and Polach Reference Stuiver and Polach1977; Stenström et al. Reference Stenström, Skog, Georgiadou, Grenberg and Johansson2011).

For the environmental samples we use Δ¹⁴C units:

(1) $${\Delta ^{14}}C = \left( {{{{A_{SN}}} \over {{A_{ABS}}}} - 1} \right)\,\,\,\,1000\,{\raise0.5ex\hbox{$\scriptstyle 0$}\kern-0.1em/\kern-0.15em\lower0.25ex\hbox{$\scriptstyle {00}$}}\,\,\,\,\,\,\,\left( {{\rm{without \;age \;correction}}} \right)$$

Where A_SN is the stable isotope normalized specific radiocarbon activity of the samples, and A_ABS is the specific activity of the absolute radiocarbon standard (226 Bq/kg carbon) (Stenström et al. Reference Stenström, Skog, Georgiadou, Grenberg and Johansson2011).

Fossil Contribution

For all the measured samples, the fossil-fuel-derived CO₂ ratio (F) and fossil fuel CO₂ component (CO_2ff) was calculated by the following equations (Levin et al. Reference Levin, Kromer, Schmidt and Sartorius2003):

(2) $$ F = {{{\Delta ^{14}}{C_{bg}} - {\Delta ^{14}}{C_{meas}}} \over {{\Delta ^{14}}{C_{bg}} + 1000}\,{\raise0.5ex\hbox{$\scriptstyle 0$}\kern-0.1em/\kern-0.15em\lower0.25ex\hbox{$\scriptstyle {00}$}}} $$

Where Δ¹⁴C_bg is the radiocarbon content of the background sample, Δ¹⁴C_meas is the radiocarbon content of the measured sample (Baydoun et al. Reference Baydoun, Samad, Nsouli and Younes2015). In our case, the Δ¹⁴C_bg is represented by the sample with the highest measured radiocarbon content, due to the lack of proper local background atmospheric ¹⁴CO₂ measurements, the highest Δ¹⁴C value represents the lowest fossil carbon content.

(3) $$ C{O_{2ff}}\left( {ppmv} \right)\,\, = \,\,{{C{O_{2meas}}\left( {{\Delta ^{14}}{C_{bg}} - {\Delta ^{14}}{C_{meas}}} \right)} \over {{\Delta ^{14}}{C_{bg}} + 1000}\,{\raise0.5ex\hbox{$\scriptstyle 0$}\kern-0.1em/\kern-0.15em\lower0.25ex\hbox{$\scriptstyle {00}$}}} $$

Where Δ¹⁴C_bg is the radiocarbon content of the background sample, Δ¹⁴C_meas is the radiocarbon content of the measured sample, CO_2meas is the measured CO₂ content in the atmosphere at the time of sampling (Berhanu et al. Reference Berhanu, Szidat, Brunner, Satar, Schanda, Nyfeler, Battaglia, Steinbacher, Hammer and Leuenberger2017). In this study, the CO_2meas is represented by the annual CO₂ concentration (ppm) in 2018 at Mauna Loa, Hawaii, 408.5 ppm (NOAA 2019).

Based on these calculations, the fossil carbon contribution can be estimated in the air and indirectly in the vegetation.

RESULTS AND DISCUSSION

The differentiation of the Δ¹⁴C data of evergreen plant samples is not simple, because the growing season is not as well-defined, as the case of deciduous species. Due to this situation, the comparison to atmospheric ¹⁴CO₂ data is not the same for every sample, because some leaves may have created at different time and the evergreen leaves may store fixed carbon from previous years (Pataki et al. Reference Pataki, Randerson, Wang, Herzenach, Grulke, West, Bowen, Dawson and Tu2010). For this reason, we compared our results to each other and the local “background” to identify the most radiocarbon depleted areas. Table 2 lists the results of AMS measurements of plant samples. The highest Δ¹⁴C data, equal to or greater than 11.8‰ were measured at the sampling point no. 3, 6, and 11. These sampling points were at rice fields (no. 3 and 6) and an urban background area on the southern coast (no. 11). These areas were relatively distant from direct urban CO₂ emissions and industrial facilities, a minimum 100 m away from roads. Although the longevity of evergreen leaves can be more than one year (Ewers and Schmid Reference Ewers and Schmid1981), in our case, only fresh leaves were collected.

Table 2

Summary of ¹⁴C results taken at different sampling districts of Bali, Indonesia.

* The sample with the highest measured radiocarbon content (Δ¹⁴C_bg).

The average Δ¹⁴C value is 2.2‰ but the scatter is ± 19‰ in case of 13 sampling point. There are two higher values, which are samples taken from urban districts, close to busy crossroads with heavy traffic at sampling points no. 9 and 10. Furthermore, there are two relatively lower Δ¹⁴C data at sampling points no. 4 and 13 from different parking places of a tourist attraction and the airport of Denpasar. The most polluted area was a crossroad (no. 10) where the traffic is moderated by a traffic light with a long waiting time for cars and scooters. This has the result that vehicles’ idling engines emit radiocarbon-free CO₂ in high concentration continuously at this point source. In this approach, the crossroads can be identified as local point sources of fossil-fuel derived CO₂. The second most polluted area is a crossroad as well, at the end of the motorway, a six-lane road, where the conditions are similar to the no. 9 sampling point. These areas are well shielded by buildings and concrete fences, which can regulate the mixing of the air by local winds. The urban background samples and materials from rice fields show generally higher than 5‰ Δ¹⁴C data (from 7.5 ± 4.6‰ to 18.2 ± 4.6‰). The 13 Δ¹⁴C data points show that the radiocarbon concentration of air in most parts of Bali island is not so depleted.

The spatial distribution is given in Figure 2 and shows that the most depleted area is the densely populated southern region where local urban and tourist activity causes heavy traffic. The radiocarbon-free CO₂, mainly due to vehicular traffic, can dilute the local radiocarbon concentration in the frequented area. These data show that the Suess effect is observable on a local scale in a less industrialized region such as Bali as well. Close to Mount Batur, in a volcanic caldera (sampling point no. 4) there is no clearly observable radiocarbon depletion, the relatively lower Δ¹⁴C data (0.9 ± 4.5‰) is presumably a consequence of the effect of tourist traffic in the nearby parking places, but cannot be clearly attributed as volcanic or magmatic CO₂ emission. Our data shows that the Δ¹⁴C level is a little higher than zero ‰ in this equatorial, tropical region. It seems the declining radiocarbon bomb-peak is close to the preindustrial level. The fossil-fuel contribution to the local environment varies between 0.3 and 25.8 ppmv compared to the local background sample (Figure 3).

Figure 2

Map of the Δ¹⁴C results in Bali, Indonesia.

Figure 3

Fossil-fuel component (CO_2ff) using radiocarbon concentration in leaf samples. The transparent gray bar with dashed line indicates the error of the CO_2ff calculation.

As shown by the Δ¹⁴C data, sampling sites no. 9 and 10 are the most fossil-carbon loaded regions. These sites have dense traffic every day all year long, which causes high local fossil fuel originated CO₂ emission by fuel burning and exhaust. The other 10 samples show that the rest of the island is less affected by fossil carbon, with an average CO_2ff is 6.4 ± 7.5 ppmv.