3.2.3 Stocks in Soils
Due to their lipophilic properties, PAHs are mainly stored at the surface layer of soils
[52]; consequently, the contamination depth depends on land use. For undisturbed
soils such as forests, previous studies have shown that PAHs were mainly concentrated within the 8–10 cm of the topmost layer of soil [7, 53]. In agricultural areas,
regular ploughing usually resulted in homogenisation of the physico-chemical
properties of the surface layer [54], and the PAH content was assumed to be constant
through the ploughed layer. In the Seine River basin, a 25-cm-deep contaminated
layer was assumed. Therefore, the masses of contaminated soils were estimated
as the product of the surface considered by the contamination depth by an average
dry density of 1,350 kg m
À3 [7]. Concerning the stocks in the urban and industrial
areas, only the permeable fraction of the surface was expected to accumulate
PAHs and was therefore taken into account. Since the thickness of the contaminated
layer varied greatly depending on the site’s history, an average value of 10 cm was
used, as for undisturbed areas. Consequently, the estimation only constituted a lower
limit of the PAH stocks in urban areas. Contaminant stocks were estimated
by multiplying the masses of soils by random values of PAH contents drawn
among 130 samples from two studies [7, 55]. A significant difference was measured
between the soil PAH contents depending on their land use (Wilcoxon-MannWhitney test, p < 0.0001), especially between forest and agricultural samples and
urban and industrial samples. The samples collected in forested and agricultural
areas underwent an additional procedure to discriminate between the samples
depending on their distance to the road network with a 150-m threshold. Indeed,
previous studies reported specific traffic-related contamination within this range
[7, 56]. The stocks on both sides were quantified independently. Thus, 22%
of agricultural land and 13% of forest area were situated within the critical
distance to the road network. The PAH stocks in agricultural lands reached
S10 ¼ 24 Æ 17 Â 10
6 kg (n ¼ 61, Σ15), equally disseminated between the road
vicinity and the more remote areas. High relative uncertainty due to the heterogeneity of the PAH content database for agricultural lands was observed. In forested
areas, the PAH stock amounted to S11 ¼ 1.0 Æ 0.3 Â 10
6 kg (n ¼ 26) with a smaller
portion (12%) located close to roads. For urban and industrial areas, the PAH
amount reached S12 ¼ 1.8 Æ 0.6 Â 10
6 kg (n ¼ 45). Because further investigation
on the relationship between the road traffic and the additional PAH stocks in the
vicinity would be necessary to correctly quantify the pollutant amounts in the areas
concerned, the stock estimation was awarded a grade of 2. In addition, the flux of
PAHs spread with urban sludge was estimated based on the population, leading to
a total amount of F13 # ¼ 310 kg year
À1 (Σ13).
176
D. Gateuille et al.
Due to their lipophilic properties, PAHs are mainly stored at the surface layer of soils
[52]; consequently, the contamination depth depends on land use. For undisturbed
soils such as forests, previous studies have shown that PAHs were mainly concentrated within the 8–10 cm of the topmost layer of soil [7, 53]. In agricultural areas,
regular ploughing usually resulted in homogenisation of the physico-chemical
properties of the surface layer [54], and the PAH content was assumed to be constant
through the ploughed layer. In the Seine River basin, a 25-cm-deep contaminated
layer was assumed. Therefore, the masses of contaminated soils were estimated
as the product of the surface considered by the contamination depth by an average
dry density of 1,350 kg m
À3 [7]. Concerning the stocks in the urban and industrial
areas, only the permeable fraction of the surface was expected to accumulate
PAHs and was therefore taken into account. Since the thickness of the contaminated
layer varied greatly depending on the site’s history, an average value of 10 cm was
used, as for undisturbed areas. Consequently, the estimation only constituted a lower
limit of the PAH stocks in urban areas. Contaminant stocks were estimated
by multiplying the masses of soils by random values of PAH contents drawn
among 130 samples from two studies [7, 55]. A significant difference was measured
between the soil PAH contents depending on their land use (Wilcoxon-MannWhitney test, p < 0.0001), especially between forest and agricultural samples and
urban and industrial samples. The samples collected in forested and agricultural
areas underwent an additional procedure to discriminate between the samples
depending on their distance to the road network with a 150-m threshold. Indeed,
previous studies reported specific traffic-related contamination within this range
[7, 56]. The stocks on both sides were quantified independently. Thus, 22%
of agricultural land and 13% of forest area were situated within the critical
distance to the road network. The PAH stocks in agricultural lands reached
S10 ¼ 24 Æ 17 Â 10
6 kg (n ¼ 61, Σ15), equally disseminated between the road
vicinity and the more remote areas. High relative uncertainty due to the heterogeneity of the PAH content database for agricultural lands was observed. In forested
areas, the PAH stock amounted to S11 ¼ 1.0 Æ 0.3 Â 10
6 kg (n ¼ 26) with a smaller
portion (12%) located close to roads. For urban and industrial areas, the PAH
amount reached S12 ¼ 1.8 Æ 0.6 Â 10
6 kg (n ¼ 45). Because further investigation
on the relationship between the road traffic and the additional PAH stocks in the
vicinity would be necessary to correctly quantify the pollutant amounts in the areas
concerned, the stock estimation was awarded a grade of 2. In addition, the flux of
PAHs spread with urban sludge was estimated based on the population, leading to
a total amount of F13 # ¼ 310 kg year
À1 (Σ13).
176
D. Gateuille et al.