Whilst the increase in ABA was not replicated in soil-based experiments, we did observe an increase in biological activity when both earthworms and plants were present. This could indicate a potential synergistic relationship between plants, soil microbes and earthworms, which could be further investigated using the developed methods. Differences between the results of the hydroponic and soilexperiments may in part be due to the use of different earthworm species. E. fetida are litter feeders and L. terrestris are an anecic species, and consequently they will interact with the soil differently. Differences between the two species in terms of e.g. sensitivity to toxicants and biochemistry in addition to behavioural differences are well established in the literature. The additional complexity of a soil matrix compared to hydroponic solutions will inevitably increase associated difficulties in the extraction. It is also possible that increased biological activity in soils compared to hydroponic experiments leads to degradation or conversion of phytohormones during extraction. There is therefore scope to improve the extraction method to achieve better recovery, allowing the observation of more subtle changes in phytohormone concentrations within soils.Perfluoroalkyl acids have been detected ubiquitously in water , biota and the atmosphere as well as in humanblood serum and breast milk . They have known and suspected toxic effects , and human exposure occurs via food . In response to concerns about these chemicals, the European Food Safety Authority established tolerable daily intakes for perfluorooctanoic acid and perfluorooctane sulfonic acid,and they have recently presented a proposal to add perfluorononanoic acid and perfluorohexane sulfonic acid while reducing the TDI for the sum of all four . To ensure that the TDIs are not exceeded, we must understand the sources of PFAAs in food. Crops are one possible vector for PFAAs into the food supply.This work aims to further our understanding of how PFAAs are transferred from soils into crops. Plant uptake of PFAAs via the roots has been studied using several experimental designs.
The first studies published were soil based experiments. Stahl et al. and Lechner et al. showed that the concentration of PFOA and PFOS in several crops was linearly proportional to the concentration in the soil in which they were grown. Since then there have been several reports of uptake of a broad spectrum of PFAAs in vegetation growing in biosolids-amended soils. They show that the length of the per- fluoroalkyl chain is the dominant variable influencing PFAA uptake in foliage. Foliage concentration factors are negatively correlated with chain length . For grasses, vertical grow tables an average decrease in FCF of 0.24 log units per CF2 group was observed , while for lettuce and tomato plants the average decrease was 0.3 log units per CF2 group . Regarding PFAA accumulation in root tissue, a much weaker influence of chain length has been observed. For instance, the variation in root concentration factors for C5eC10 perfluoroalkyl carboxylic acids was just 0.5 log units for radish, celery, tomato and pea . A similarly small variation was found between PFHxA, PFOA, PFBS, PFHxS and PFOS in wheat . In contrast, root concentration factors in chicory showed a pronounced dependence on the chain length, suggesting that root accumulation is influenced by species and soil type . Hydroponic experiments provide an opportunity to obtain a more systematic understanding of contaminant accumulation in plants. For instance, a hydroponic experiment was used to assess the influence of different metabolic inhibitors on the uptake of PFOA and PFOS in maize shoots . The influence of pH on PFAA uptake into maize roots was also elucidated in a hydroponic experiment, showing no effect in a pH range of 5e7 for nine of the ten PFAAs studied . A hydroponic study was used to explore the effect of temperature and salinity on PFAA uptake in wheat, identifying a positive effect for both, which was attributed to increased evapotranspiration . Hydroponic experiments have also been used to study how perfluoroalkyl chain length influences uptake in plants. PFAAs with perfluoroalkyl chain lengths ranging from 3 to 13 were all transferred via the roots to the plant foliage in lettuce, tomato, cabbage and zucchini . Transpiration stream concentration factors for C4eC10 PFAAs ranged over just a factor of two for three of the four species. Relatively high TSCFs of 0.05e0.8 showed that the PFAAs were clearly able to cross the Casparian strip and plasma membranes that prevent the passive entry of many polar molecules into the vascular tissue of the root . A weak influence of chain length on TSCF was also observed in grass . Hydroponic studies have also been used to study PFAA uptake into roots. In lettuce, the root-nutrient solution concentration factor decreased with chain length for C4eC6 PFCAs before increasing by almost 3 orders of magnitude from PFHxA to PFUnA.
While the accumulation of the shorter chained compounds was explained by uptake with the transpiration stream, the uptake of the longer chained compounds was attributed to sorption to the surface tissue of the roots . Hydroponic experiments with tomato, cabbage and zucchini showed a strong positive relationship between root-hydroponic solution concentration factor and chain length for C4eC11 PFAAs, indicating that root-surface sorption was the dominant uptake mechanism for all of the PFAAs in these species . In detailed experiments with a hydroponic model plant system , Müller et al. also concluded that the root uptake of all but the shortest PFAAs was governed by sorption and observed that the dead roothydroponic solution concentration factor increased by almost 3 orders of magnitude from PFBA to PFOS. Comparing the results from hydroponic and soil experiments, there are clear differences in the chain length dependence of PFAA uptake. In foliage, the hydroponic studies show a weak dependence of uptake on chain length, while soil studies show a very strong dependence. The opposite is the case in roots; the hydroponic studies show a strong positive chain length dependence that is attributed to sorption to root surfaces, while the soil studies show a weak dependence. It is unclear what the reasons for these differences are, and how and to what extent findings from hydroponic studies can be transferred to natural soil systems. Sorption of PFAAs to soil solids is certainly an important factor, as this reduces the fraction of chemical available for uptake by the roots. To be able to sorb to the root surface or be taken up with the transpiration stream, the compounds first need to be present in pore water. Long chain compounds sorb strongly to the soil; hence, for a long chain PFAA much higher concentrations in soil are required to generate a given concentration in pore water than for short chain PFAAs . However, there may be other factors that affect the comparability of hydroponic and soil systems. For instance, some contaminants appear to be taken up through the action of root exudates , which would be highly diluted or not present under hydroponic conditions. Another possibility is that differences in the nature of root tissue when grown under hydroponic conditions influence PFAA uptake and translocation. The uptake of the PFAAs could also be influenced by other solutes present in the soil. To explore these questions, we conducted a lysimeter experiment in which lettuce was grown in soil containing PFAAs, and compared this with our previous hydroponic experiment conducted with the same plant species, chemicals, sample preparation and analysis. The lysimeter soil was spiked with 11 PFCAs and 2 perfluoroalkane sulfonates . Four lysimeters were used, each with a different spiking level. At maturity the lettuce was harvested and the roots and leaves were analyzed separately.
Additionally, the PFAA concentrations in soil and pore water were determined. The measurement of concentrations in pore water facilitated comparison of this experiment with our earlier hydroponic greenhouse study, and thereby identification of differences in the uptake into roots and leaves between soil and hydroponic growth environments.The field experiment was conducted at the Fraunhofer Institute for Molecular Biology and Applied Ecology IME in Schmallenberg, Germany. Lettuce plants were grown in 5 lysimeters, one containing soil with background concentrations of PFAAs , and 4 with intended concentrations of individual PFAAs in soil of 0.1 mg/kg, 1 mg/kg, 5 mg/kg and 10 mg/ kg . This compares with PFOA and PFOS concentrations of ~1 mg/kg measured in contaminated agricultural soil in Arnsberg, ~30 km from Schmallenberg . The results from the highest spiking level were not used because the lettuce plants were significantly smaller at the time of harvest than those growing in the lower exposure levels, indicating that PFAAs had phytotoxic effects . Phytotoxic effects of PFAAs have been reported elsewhere . Each lysimeter had a surface area of 1 m2 and a total depth of 60 cm. The lysimeters were each filled with ~450 kg sand and ~450 kg of loamy sand . This resembled a typical soil from northwestern Germany. The soil used for the upper layer is available as a reference soil from Fraunhofer IME . The spiking of the soil was done stepwise. First a stock solution was prepared containing all PFAAs in methanol. With this stock solution 2 kg of soil were spiked. Afterwards the 2 kg spiked soil was mixed with approximately 90 kg of soil in a concrete mixer to achieve the desired concentration. This was repeated 5 times for each layer in each lysimeter. Samples were taken from each batch and combined to determine the initial PFAA concentration in the soil of each lysimeter. The lettuce plants were pre-grown in a greenhouse for 2 weeks in non-spiked soil before they were transferred to the lysimeters. Within one week of preparing the spiked soil, 20 lettuce seedlings were put in each lysimeter . The seedlings were watered after planting, and kept humid by rain events until harvest with supplementary watering when needed . After 72 days the lettuce plants were harvested . The plants were divided into roots and foliage, packed in freezer bags and stored at 20 C until analysis. Soil samples were taken with a soil corer when the plants were harvested. The soil core, which was taken from the top to the bottom of the lysimeter,flower pot was divided between the upper and lower soil layers, and the soil was packed in freezer bags and stored at 20 C for later separation of pore water and analysis.Before homogenization with a household blender the roots were rinsed with demineralized water to wash off residual soil and then carefully dried superficially with paper towels. As no residual soil was visibly apparent on the leaf samples, no cleaning was performed. The extraction method used is based on the modification Vestergren et al. proposed for the method published by Hansen et al. . Briefly, 10 g of the homogenate were weighed into a 50 mL PP tube and spiked with mass-labeled surrogate standards.
After adding 5 mL of 0.4 M NaOH solution and vortexmixing, the samples were left in the refrigerator over night to allow the internal standards to distribute in the slurry. Next, 4 mL of 0.5 M tetrabutylammonium hydrogensulfate solution and 5 mL of a carbonate buffer were added to the samples and thoroughly mixed. After adding 10 mL MTBE and vortex-mixing for 1 min the samples were sonicated for 10 min. Phase separation was achieved by centrifuging for 10 min at 3000 rpm. The MTBE phase was transferred to a new 50 mL PP tube and the extraction repeated two times. The extracts were combined and concentrated to approximately 2 mL using a Rapidvap . After adding 1 g of sodium sulfate to Florisil SPE-cartridges to remove any remaining water in the extracts, the cartridges were conditioned with 10 mL MeOH and 10 mL MTBE before they were loaded with the extract. The elution of the non-polar matrix was done with 10 mL MTBE before the target compounds were washed off the cartridge with 10 mL MeOH/MTBE . This extract was again evaporated to 1 mL final volume. An additional clean-up step following the Powley method with ENVI-Carb was added when the final extract was still strongly colored.