Carbon Isotope Signature of Polycyclic Aromatic Hydrocarbons (PAHs): Evidence for Different Sources in Tropical and Temperate Environments? WOLFGANG WILCKE,* MARTIN KRAUSS, AND WULF AMELUNG Institute of Soil Science and Soil Geography, University of Bayreuth, D-95440 Bayreuth, Germany In tropical soils, naphthalene and, partly also, perylene occur at elevated concentrations while pyrolytic higher molecular weight PAHs are almost absent. We hypothesize that there are recent biological PAH sources in the tropical environment related with woody plants and termites. We used the C isotope signature of individual PAHs in temperate and tropical soils and in tropical wood and termite nests to distinguish different PAH sources. The mean δ 13 Cvalues of the benzo[ b+j +k]fluoranthenes and of benzo- [ a+e]pyrenes in temperate soils ranged between -24.6‰ and -25.3‰, being similar to values reported in the literature for PAHs with pyrolitic origin. The mean δ 13 C values of perylene decreased in the order temperate soils (-27.0‰) > termite nests (-31.4‰) > tropical soil (-32.4‰), while those of naphthalene (-24.6‰ to -26.2‰) were similar among the tropical and temperate soils, tropical wood, and termite nests. Our results support the assumption that perylene in the tropical environment is recently biologically produced, as indicated by the depletion in 13 C. The C isotope composition of naphthalene, however, cannot be used to distinguish different sources. Introduction Polycyclic aromatic hydrocarbons (PAHs) are produced during pyrolysis offossil fuels or modern plant material and ubiquitously distributed in the environment (1, 2). In the temperate zone, the largest part of the PAHs results from anthropogenic activity and is mainly stored in soils (3). Th e little published work on PAHs in tropical soils indicates that the concentrations of most PAHs are lower than in the temperate zone and that the composition ofthe PAH mixture is markedly different. While in the temperate zone benzo- fluoranthenes, chrysene, and fluoranthene often dominate the PAH mixture,in the tropics,naphthalene,phenanthrene, and perylene are most abundant in many soils, particularly in rural areas (2). The works of Chen et al. (4, 5) and Wilcke et al. (6) have shown that the sources of naphthalene, phenanthrene, and perylene in tropical soils may be related to the activity of termites and woody plants because termite nests and wood from tropical regions contain high naphthalene and, partly also,high phenanthrene and perylene concentrations,while the concentrations ofPAHs that are typicallyassigned to the pyrolysis of fossil fuels or modern organic matter and that dominate the PAH pattern in temperate soils are low (2). Considerable evidence has accumulated that perylene is produced biologically under anaerobic conditions, because of its presence in groundwater-influenced soils and in sediments where combustion-derived PAHs are almost absent (7-10). While Venkatesan (8) stated that perylene is microbially produced in anaerobic environments by incom- plete degradation ofpigmentsfrom insects,fungi,and marine organisms containing perylene quinones, other authors assume a microbial production independent of specific precursors(11).Asin termite gutsanaerobicconditionsoccur (12), it seems possible that perylene is produced there by intestinal microorganisms. In the past decade, the C isotope signature of individual PAHs has been used to determine the contribution from different sources to the PAH burden of soils and sediments (13-16). Enzymatically catalyzed biological processes fre- quentlydiscriminate 13 C, resulting in a shift ofthe δ 13 Cvalue (i.e., the 13 C/ 12 C ratio relative to a standard) from that of the atmospheric CO2 to more negative 13 C-depleted values. For plantsfollowingthe C3 metabolicpathway,the main sources of fossil fuels, δ 13 C values of -32‰ to -22‰ are common in bulk biomass with a mean of -27‰ (17). The C isotope signature ofpyrolysis-derived PAHs resembles the δ 13 Csignal of the fuel, depending on combustion conditions. For coal combustion and gasification processes, high combustion temperatures lead to PAHs depleted in 13 C as compared to parent materials, while for low temperatures the δ 13 C values of PAHs are close to those of the parent coals (18). Furthermore, it may not be ruled out that transport and postdepositional processes change the δ 13 C values of py- rolysis-derived PAHs after their emission to the environment. The objective of our study was to test the hypothesis that the compound-specific δ 13 C value allows for perylene and naphthalene produced by recent biological processes to be distinguished from those that were produced by pyrolysis of fossil fuels. For this purpose, we determined δ 13 C values of combustion-derived PAHs in anthropogenically contami- nated temperate soils and compared these results with the C isotope signature of presumably biological naphthalene and perylene in tropicalwood,termite nest,and soilsamples. Materials and Methods Samples. As temperate reference samples in which the PAHs were predominantlyderived from the pyrolysis offossilfuels, we chose three urban topsoils (0-5 cm) comprising a house garden,a former gasworksite,and a roadside soil.The house garden and gas work site were in the city of Bayreuth and the roadside soilnear a heavilyfrequented road in the village of Stephanskirchen, Germany. The soils were anthropogeni- cally strongly modified Cambisols (house garden and gas work site) and a Leptosol (roadside, 19). In the Brazilian Amazon region,we collected in each ofthe Terra firme (never flooded), Va ´rzea (seasonally flooded by white water), and Igapo ´ (seasonally flooded by black water) regions near Manaus one composite sample ofeach oftopsoil(0-10 cm ), dead wood, and the central part of the nests of the wood- feeding termite genus Nasutitermes. The soil types in Brazil were Ferralsols (Terra firme)and Fluvisols (Va ´rzea and Igapo ´ ), and organic C and PAH concentrations are shown in Table 1. All samples were air-dried at ambient conditions near the location where theywere collected and sieved to <2 m m . Air *Corresponding author phone: ++49 30 314 73538; fax: ++49 3031473548;e-mail: wolfgang.wilcke@tu-berlin.de.Present address of all authors: Department of Soil Science, Institute of Ecology, Technical University of Berlin, Salzufer 11-12, D-10587 Berlin, Germany. Environ. Sci. Technol. 2002, 36, 3530-3535