Quantitative variation of HAP
The extraction of aromatic hydrocarbons by two experimental protocols was illustrated in Table 1.
The results showed a quantities of between 35 µgg-1 Ms and 910 µgg-1 Ms for the hexane-acetone extraction and between 62.5 µgg-1Ms and 1556 µgg-1 Ms for the dichloromethane-hexane extraction. These levels were compared with those form elsewhere in the world on similar studies (Table 2).
The highest levels were recorded in 1mS (1556 µgg-1), 3mN (1221.5 µgg-1) and 1mE sites (809.9 µgg-1). While the lowest levels were recorded in the 50m of Tunis road (35 µg / g) and the control site located at 500m from this road (49 µgg-1), the control site of Gremda road (51µgg-1) and the control site of Manzel Chaker road (62.5 µgg-1).
The differences observed between sites were linked to the intensity of road traffic and also the effect of meteorological factors. For the two extraction methods, the PAH levels decreased with the increase in the distance from the road. This trend was also reported in other studies (Kumar et al. 2011).
The Manzel Chaker soils recorded the highest levels; 1556 µg g-1 Ms in 1mS (with the dichloromethane-Hexane extraction) and 1221.5 µg g-1 Ms in the 1mE site (with the dichloromethane-hexane extraction). For hexane –acetone extraction, the contents recorded in 1mS represent 2 times higher than the content recorded at the same distance from Gremda road, and 4 times higher than the content of the site 1mW. While, for the dichloromethane-hexane extraction, the content recorded at 1m from Manzel Chaker road, is almost 3 times higher than that recorded at 1m from Gr2emda road, 2 times higher than that recorded at 1m in the east of Tunis and 2.5 times higher than that recorded at 1m in the west of Tunis road. For the two extraction solvent combination, the ratio of the contents recorded by the two extraction is illustrated in figure 2. The difference between the two protocols was more remarkable in the 3mW, 50mS, 10mW and 50m sites for which the dichloromethane – hexane extraction are 4.5, 3.3, 2.8 and 3.1 times higher than the hexane – acetone extraction respectively. For only 13% of sites, hexane-acetone extraction was more efficient than Dichloromethane-hexane. While for 8% of sites, extraction by both methods gives the same values (Fig.2).
This is apparently linked to the difference in density between dichloromethane (1.34) and hexane (0.78). Increasing the density of solvent results in the solubility of the compound to be extracted (Henrion, 1999). Several authors have used this extraction method (Mzoughi et al. 2002; Diago et al. 2010; Temilola et al. 2011; Obini et al. 2013).
We used the hexane for many reasons: low cost, high affinity with hydrocarbons especially when increase the contact time between them in room degree and the easy to evaporation and this point is very important according to use the rotary evaporator to maintain the evaporation point for hydrocarbons. Cabillic et al (2015) indicated the effectiveness of hexane in the extraction of hydrocarbons especially fluorenthene (Flu) and benzo (b) fluorenthene (BbF).
Qualitative analysis of HAP
Elemental analysis of polycyclic aromatic hydrocarbons showed the detection of 14 compounds (Figure 3). The ten most important compounds are Nap, Napthalene; Ace Nap, Acenapthene; Flu Fluorene; Flan, Fluoranthene; B (a) A, Benzo (a) Anthracene; Chry, Chrysene; B (a) P, Benzo (a) Pyrene; IP, Indeno (123cd) Pyrene; Di Benzo (ah) Anthracene; Benzo (ghi) Pyrelene.
The highest concentrations were recorded for penzo(a)pyrene (46 µg g-1 in the 1mS site), Chrysene (43 µg g-1 in the 1mS site), penzo(a)anthracene (30.2 µg g-1 in the 3mS site) and ndeno (123cd) pyrene (45 µg / g I the 1mSW site). The lowest levels were recorded for naphthalene (0.12 µg g-1 in the 500mMc site) and acenaphthene (0. µg g-1 in the 500mMc site). These less compounds correspond to light molecules. This situation is found in old soils with historical pollution where the lightest molecules have time to volatilize, to be degraded or carried away by groundwater or percolation (Loehr and Webster 1996).
Naphthalene and dibenzo (ah) anthracene were not detected at some sites on the Gremda and Manzel Chaker roads for both extraction systems. This seems to be linked to the variation in road intensity use between the three roads as well as the variation in type of vehicles circulated there. The low levels recorded for dibenzo (ah) anthracene is linked to its high boiling point. This compound is among the compounds assayed, the one which has the highest boiling point (524 ° C), which can modify the levels detected (Laurence et al. 1998). The results obtained showed an important variation between sites of the selected roads. Generally, a decrease in the concentrations of Manzel Chaker sites for most compounds as the sampling distance drifted further from the roads was observed. The first two sites, 1m and 3m from the three roads, recorded the highest concentrations followed by a sharp drop. As a result, the oil contamination appears to be concentrated to a short distance from the road of around 3m. These results were consistent with those of Kumar et al (2011). The low concentrations recorded for fluorene was speculated to be related to the mechanisms of its degradation by microorganisms (Manilal and Alexander 1991).
Soil characteristics
The pH of Manzel Chaker soil varies between 4.07 and 7. The low values; 4.07 and 4.7 were recorded at 3m and 1m in north of Manzel Chaker road respectively. For the two other sites (Tunis and Gremda), the pH values were between 7.1 and 8.97. The alkaline pH which is close to 9 for 1m and 3m near Gremda road and between 7.7 and 8.5 for the sites of Tunis road (Table 3).
The acidic pH of 1mS and 3mS sites of Manzel Chaker road increases microbial activity and therefore may increase the availability of hydrocarbons leading to increased toxicity (Leahy and Colwell, 1990). On the other hand, Yang et al (2001) indicated that the decrease in pH tends to decrease the availability of PAHs (Subramaniam et al. 2004).
In addition to pH, the richness of major elements (Ca2+, k+, Na+) can influence the availability of PAHs. The soil of Tunis were found to be poor in K+ (between 0.006% and 0.05%), Na+ (0.009% and 0.026%) and Ca2 + (less than 1%), the same for the Gremda soils where low levels of K+, Na+ and Ca2+ were recorded (Table 3). On the other hand, for Manzel Chaker soil, the results obtained showed high concentrations of Ca2+ in 1mS (2.57%) and 3mN (2.91%) sites.
The organic component of soil plays an important role in the availability of polycyclic aromatic hydrocarbons. Polluted soils contain little organic matter, thus preventing bacterial activity (Simeon et al. 2008). Our results showed that the OM content in Manzel Chaker soil was higher than 2% at the first site north of road and slightly over 1% up to a distance of 10m. While in the two sites 1mS and 3mS, the content of OM exceeded 2%. Therefore, the increased OM content in the 1mS and 3mS sites promotes microbial activity and hence the solubility of organic pollutants. In the immediate vicinity of Gremda road (1m), the MO content (1.52%) is almost 5 times higher than the value recorded at 500 m from the road (0.3%). On the other hand, we note a depletion in OM for the soil of Tunis where the contents do not exceed 1% for all the sites. On the other hand, the affinity of humic material for hydrophobic molecules such as PAHs is more important with lower polarity (Schlautman and Morgan 1993; Ping et al. 2006). These results seem to be related to an enrichment of polycyclic aromatic hydrocarbons for the sites rich in OM. The rise in OM content at the Manzel Chaker sites can decrease the biodegradation of these organic pollutants. Generally, studies carried out on the biodegradability of hydrophobic pollutants in soils show that their mineralization was much slower when the pollutant is in adsorbed form (Manilal and Alexander, 1991; Al Bashir et al., 1990). Polluted soils contain little organic matter, thus preventing bacterial activity (Simeon et al. 2008). The degrading effect by bacteria has been demonstrated by several studies (Wilson and Jones 1993; Breedveld and Sparrevik 2000; Canet et al. 2001; Antizar-Ladislao et al. 2005).
Germination
The monitoring of the direction of the winds during the study period showed that Gremda road sites were influenced by winds from N, NNE, NNW, SE, NE, ENE, E and ESE. So, five directions brought emissions from the road to sampling area. Tunis road faces NNE, NE, ENE, E and ESE winds. Thus, three wind directions were able to bring road traffic emissions to sampling sites. Manzel Chaker road was under the dominance of ESE, E, ENE, NE, NNE, N and NNW winds and 4 of them enhanced spreading of pollutants toward sampling sites. The distance was calculated from the shoulder of the road. In order to ensure homogeneity and similarity between the different samples, the 500m sample must be in the same area.
The frequency of exposure of sites to road pollutants during the study period showed that the sites south of Manzel Chaker road are the most exposed to road pollutants (35% of wind directions are subject to bring the road emissions to the south side of Manzel Chaker road) followed by the EST sites of Tunis road (30% of wind directions can bring the emissions to the EAST side of the road) (Figure 4).
For this reason, we took the soil in the southern side of Manzel Chaker road, in the eastern side Tunis road and in the vicinity of Gremda road, to do the germination test. The germination capacity and rate of two species on roadsoil samples are shown in Figure 5.
For Gremda road soil samples, the germination percentage on the soil taken at a distance of 1 m was 40% and 54% for tomato and cucumber respectively. These values were not significantly different from those measured for the soil sampled at a distance of 3 m. Contamination of the soil collected near Gremda road affected the germination capacity of the species tested, up to 3m. Considering that the PAH contents in the soils of Gremda was low compared with those of Tunis and Manzel Chaker road, this disturbance in the germination of the species seems to be linked to metallic contamination as observed by Mbadra et al (2018).
The synergism between Zn and Pb affected the percentage of germination of tomato and cucumber.
At 1m from Tunis road, the germination percentages of tomato and cucumber were 80% and 72% respectively. The germination percentages in the control site were 82% and 72% respectively. Therefore, the low soil contamination of Tunis does not affect the germination percentages of two selected species. On the soil collected at 1m from Manzel Chaker road, the seed germination percentages were 44% and 72% for tomato and cucumber respectively, which were not significantly different from those collected at 500m from the road. Manzel Chaker's soil was contaminated with PAHs and showed no metallic contamination (Mbadra et al. 2018). Thus, contamination by PAHs does not seem to affect the germination of seeds of the studied species.
The germination speed of tomato seeds on soils 1m and 3m from Gremda road was zero since the germination capacity were less than 50%. On the other hand, 144 h and 120 h were necessary for the germination of 50% of the seeds on the control soils (500m) of Gremda and Tunis roads respectively. For Manzel Chaker road, 248 h were needed to germinate 50% of the tomato seeds on the soil 1m away from the road, 232 h on soil 3m and 152 h for the germination of half of the seeds on the control soil (500m). For cucumber, the germination speed was 86 h on soil T1 and 80 h on control soil of Tunis road, 26 h on MC1 soil, 24 h on control soil of Manzel Chaker road, 119 h on G1 soil and 124 h on control soil of Gremda road.
These results showed that soil near Manzel Chaker and Gremda roads reduced the germination rate of the tomato while no effect was observed on the germination of the cucumber. The tomato germination rate was most affected in the vicinity of the three roads. The PAH richness in the soil of Manzel Chaker road (1556 μg / g) affected the germination rate of Solanum esculentum with an increase of 98 h compared to the control site (500m) and did not affect the germination capacity of cucumber. According to Baker (1971) and Adam et al (2002), petroleum hydrocarbons can form a film on the seed, preventing the entry of oxygen and water. In addition, Sharifi et al (2007) showed the negative effect of hydrocarbons on the germination of Medicago truncatular (a species of the Fabaceae family).
At 1m from Tunis road, the germination percentages of tomato and cucumber were 80% and 72% respectively. The germination percentages in the control site were 82% and 72% respectively. Therefore, the low soil contamination of Tunis does not affect the germination percentages of two selected species.
On the soil collected at 1m from Manzel Chaker road, the seed germination percentages were 44% and 72% for tomato and cucumber respectively, which were not significantly different from those measured 500m from the road. These results showed that soil near Manzel Chaker and Gremda roads reduced the germination rate of the tomato while no effect was observed on the germination of the cucumber. The tomato germination rate was mostly affected in the vicinity of the three roads. The PAH richness in the soil of Manzel Chaker road (1556μg/g) affected the germination rate of Solanum esculentum, with an increase of 98 h compared to the control site (500m) which did not affect the germination capacity of cucumber.
The correlation analysis between the most abundant hydrocarbons and the germination parameters of the two species (Fig 6) showed a negative relationship between chrysene, penzo (a) pyrene and ndeno (123cd) pyrene and the speed and germination capacity of two species, exception the penzo (a) pyrene, which has a positive correlation with the speed germination of tomato. On the other hand, we noted many positive correlations between the less abundant elements and the germination parameters (Table 4). Napthalene, acenaphthalene, fluorene, fluorenthene and Di benzo (ah) anthracene showed a positive correlations with the germination capacity of cucumber. Likewise, these same components, with the exception of di benzo (ach) anthracene, showed a positive correlation with speed germination of tomato and also, we noted, a positive correlation between naphthalene, acenaphthalene, fluorene, fluorenthene. and the speed germination of tomato. So, the impact of higher molecular weight was more distinguished and more pronounced than that of low molecular weight. According to Baker (1971) and Adam et al (2002), petroleum hydrocarbons can form a film on the seed, preventing the entry of oxygen and water. In addition, Sharifi et al (2007) showed the negative effect of hydrocarbons on the germination of Medicago truncatular (a species of the Fabaceae family).
Ahammed et al (2012) reported the effect of pyrene on the photosynthetic mechanisms of tomato. Similarly, Onyema (2013) showed the effect of pyrene and B [a] P on the early seedling growth of Lolium perenne. On the other hand, the studies of Alam al (2010) pointed in another direction by showing the role of plants in the degradation of phenanthrene and pyrene.
Plant growth
The shoot and root biomass results are shown in the table 5. On the soil sample at 1m from Gremda road, the tomato root biomass was 0.05 g, while it was 0.54 g on the control soil (500 m). Tomato root biomass increased further away from this road. In contrast, for cucumber, the root biomass was 0.08 g on the soil 1 m from Gremda, while it was 0.078 g on the control soil (500 m).
On the soils of Manzel Chaker road, no significant difference was detected for the root biomass for tomato and cucumber. Indeed, the results showed that the root biomass was 0.6 g for the tomato and 0.5 g for the cucumber. This seems to be related to the metallic contamination of the soil in the vicinity of Gremda demonstrated in a previous study (Mbadra et al. 2018). Shoot biomass decreases more than root biomass in hydrocarbon-contaminated soil (Xu and Johnson, 1997; Waltonet al. 1994). Likewise, for the soils of Tunis road, the results obtained do not showed any variation in the root biomass. This biomass was 0.7g and 0.5g for tomato and cucumber respectively. On the Manzel Chaker soil, the shoot biomass of tomato was 0.14g at 1m from the road, it was significantly lower than the control soil (0.62g). These results corroborate with those of Kordybach et al (2000) who showed the effect of chrysene on the aerial elongation of tomato. For cucumber, this shoot biomass was 0.7 g in the five selected sites.
The richness of the first Manzel Chaker site (1m) by PAHs (1556μg / g) affected the above-ground biomass of Solanum esculentum with a reduction of 81%. However, this contamination did not affect the shoot biomass of cucumber. Therefore, the high levels of PAHs affected the aerial elongation of tomato. On the other hand, Han et al (2018) who worked on another species of the solanaceae family (Solanum nigrum) can resist to the low contamination by penzo (a) pyrene.
The development of the tomato on the soil taken 1m from Manzel Chaker road showed a significant reduction of 60% in the elongation of the aerial part compared to that cultivated on a soil sampled at 500 m from the same road. In contrast, for cucumber, the elongation of the aerial part was 18 cm for all sites. Shoot elongation was not affected compared to that developed on uncontaminated soil Manzel Chaker's soil was rich in PAHs. Therefore, the decrease in tomato growth was attributed to the high PAH content. In fact, hydrocarbons can participate in the inhibition of cell division in the cells of the leaf (Henner, 2000). Tesar et al (2002) noticed a 96% reduction in ryegrass biomass after 30 days of growth on soil contaminated with 25 g / kg of petroleum hydrocarbons. In agreement with our results, Cuevas et al (2008) showed the inhibition of aerial growth of herbaceous plants after 40 days when they were sown on a medium containing 1 mg / g of PAH.
For Gremda road, at the end of the experiment, the growth of tomato was in the order of 22cm at all sites. The growth of the two species on the soil taken at 1m from Gremda road showed that the contamination of these soils did not affect the elongation of the aerial part of lycopersicum esculentum and cucumis sativus. The cultivation of these two species on the soil taken at 1m from Tunis road showed the absence of any contamination effect on the elongation of tomato and cucumber shoots.
On the other hand, the contamination of Manzel Chaker's soils by hydrocarbons affected the aerial elongation of Tomato with a reduction of 60% compared to the control site. However, this contamination did not affected the aerial elongation of cucumber.
The use of PCA (Figure 7) showed that the existence of fluorene, fluorenthene, naphthalene and chrysene in the same pool as the germination capacity of tomato and the germination rate of tomato and cucumber, indicated the absence of the impact of the four compounds on the germination of tomato and cucumber (PCA1). The existence of large olive trees in the areas studied participated in the biodegradation of these molecules which have low molecular masses. Plants can indirectly influence the dissipation of PAHs through the stimulation of microorganisms capable of degrading them (Shahsavari et al. 2015; Tejeda-Agredano et al. 2013). In contrast, PCA 2 showed the effect of fluorenthene, naphthalene, chrysene and benzo (a) pyrene on the germination capacity of two selected species, on the tomato germination rate and the elongation of the aerial part of tomato. Therefore, these compounds did not show an effect on the germination rate of cucumber and also the elongation of its aerial part which can indicate the resistance of cucumber to contamination by hydroacrbons. According to Guo et al (2015) a weak treatment of tomato seeds with naphthalene affected the dry mass of whole plant and vigorous seedling index. Similarly, abbasi et al (2013) showed that naptalene with 0.02% has an effect on growth, nutrient uptake and fruit yield of tomato. Abou El-Yazied et al (2011) showed that concentrations between 25 and 50ppm of naphthalene, affected growth, photosynthetic pigments and tomato fruits. Studies by Oguntimehina et al (2010) showed that fluorenthene affected tomato root elongation, as well as photosynthetic capacity and several physiological parameters.The studies of Somtrakoon et al (2013) have shown the impact of these toxic compounds on the aerial and root elongation of economic crop in Thailand. These compounds were considered the most carcinogenic (Lacoste et al. 2013). The existence of shoot elongation of cucumber in the same group with ndeno (123cd) and the benzo (a) anthracene indicated the absence of effect of these hydrocarbons on the aerial elongation of cucumber. Wang et al (2011) showed the impact of benzo (a) pyrene on germination and in the elongation of Wheat. Similarly, Wagner and Wagner-Hering (1971) showed that benzo [a] pyrene have a negative effect on rice, wheat and corn at concentrations more than 1.2 mg kg-I soil. In the same, Campbell et al (2006) who worked on Cannabis sativa, showed the negative effect of chrysene and benzo (a) pyrene on the germination and elongation of this species.
The use of PCA (Figure 7) showed that the existence of fluorene, fluorenthene, naphthalene and chrysene in the same pool as the germination capacity of tomato and the germination rate of tomato and cucumber, indicated the absence of the impact of the four compounds on the germination of tomato and cucumber (PCA1). The existence of large olive trees in the areas studied participates in the biodegradation of these molecules which have low molecular masses. Plants can indirectly influence the dissipation of PAHs through the stimulation of microorganisms capable of degrading them (Shahsavari et al. 2015; Tejeda-Agredano et al. 2013). In contrast, PCA 2 showed the effect of fluorenthene, naphthalene, chrysene and benzo (a) pyrene on the germination capacity of two selected species, on the tomato germination rate and the elongation of the aerial part of tomato. Therefore, these compounds did not show an effect on the germination rate of cucumber and also the elongation of its aerial part which can indicate the resistance of cucumber to contamination by hydroacrbons. According to Guo et al (2015) a weak treatment of tomato seeds with naphthalene affected the dry mass of whole plant and vigorous seedling index. Likewise, abbasi et al (2013) showed that naptalene with 0.02% has an effect on growth, nutrient uptake, incidence of blossom end rot, fruit yield of tomato. In the same direction, Abou El-Yazied et al (2011) showed that concentration between 25 and 50ppm of naphthalene, affected growth, photosynthetic pigments and tomato fruits. Studies by Oguntimehina et al (2010) showed that fluorenthene affected tomato root elongation, as well as photosynthetic capacity and several physiological parameters. The studies of Somtrakoon et al (2013) have shown the impact of these toxic compounds on the aerial and root elongation of economic crop in Thailand. These compounds were considered the most carcinogenic (Lacoste et al. 2013). The existence of shoot elongation of cucumber in the same group with ndeno (123cd) and the benzo (a) anthracene indicated the absence of effect of these hydrocarbons on the aerial elongation of cucumber. Wang et al (2011) showed the impact of benzo (a) pyrene on germination and in the elongation of Wheat. Similarly, Wagner and Wagner-Hering (1971) showed that benzo [a] pyrene have a negative effect on rice, wheat and corn at concentrations more than 1.2 mg kg-I soil. In the same, Campbell et al (2006) who worked on Cannabis sativa, showed the negative effect of chrysene and benzo (a) pyrene on the germination and elongation of this species.