IN THE SOIL ENVIRONMENT

  Soil contaminated by oil affects plant growth, the quality of surface and ground water, air environment, and other environmental media around the site. A clean Earth environment is so important to human beings and all living organisms that determining the biodegradability of B100 in a soil environment is another indispensable task.

The existing methods for determining biodegradation in soil can generally be classified into three categories (EPA, 1975). They are laboratory tests, greenhouse studies, and field studies. Laboratory tests, exemplified by bottle tests, are carried out under a controlled environment and are time-saving, convenient, and economical. The disadvantage is that they are not suitable for deciding whether the results can be used directly to predict full-scale treatment as the conditions under which they are tested do not closely mimic field conditions (Flathman et al., 1994). Field studies reflect the fate of the chemicals in conditions similar to the actual environment but they are high cost and need long time periods. Greenhouse studies such as pan studies are a combination of laboratory tests and field studies in terms of environmental simulation and complexity. In pan studies, experiments are typically set up in stainless steel, glass, insulated wood, or a plastic container which act as a small microcosm and provide biodegradation information more consistent with what can be expected in the field. Pan studies are very useful for estimating rates and extent of biodegradation of contaminants during full-scale treatment (Flathman et al., 1994). A lysimeter is an example of a pan study. Wang and Bartha (1990) used a 90 x 90 x 60 cm3 outdoor lysimeter to simulate spills of three hydrocarbon fuels (jet fuel, heating oil, and diesel) and found good correlation between hydrocarbon residue decline, microbial activity, and toxicity reduction. Finally, since re-vegetation of soils contaminated by fuel spills is often a desirable goal, plant bioassay such as seed germination is usually employed to evaluate toxicity of pollutants in a soil system (Wang and Bartha, 1989; Burnside et al., 1971; Sheets et al., 1968; MacRAE and Alexander, 1965).

Generally speaking, the test of biodegradability in soil is thought to be more difficult than in an aquatic system because of the high endogenous rates of soil respiration and its solid phase, which makes the pollutants difficult to access by microorganisms. Yet, most of the methods used in the aquatic phase are still available for assessing the potential of a material to be biodegraded in soil, including the CO2 evolution test, BOD (respiratory meter) method, GC method, and so on. The 14CO2 evolution test (biometer test), recommended by the EPA under TSCA (1979), is considered the best method, but is impractical when dealing with vegetable oils (Cornish et al., 1993) because no 14C-labeled rapeseed oil is available. Some methods, for example, oxygen consumption and CO2 production tests, are considered to be unsuitable for natural soils because high organic matter in the soil causes high respiration rates of soil organisms (Howard et al., 1975). Since the CO2 evolution method has some profound advantages such as being simple (therefore easily carried out), inexpensive (especially in running large experiments such as determination of degradation rates), and environmentally safe (no hazardous reagent involved such as hexane, methylene chloride, or chloroform as in the case of GC methods), an attempt was made to use the CO2 evolution method in the soil experiments while the GC method dominated the analysis.

The purpose of this study was to test the biodegradability of five different B100 fuels in the soil environment and compare them with 2-D reference diesel fuel. GC analysis was the major method employed in all the soil experiments and the CO2 evolution from all the substrates was also investigated as a comparison.

OBJECTIVES
The objectives of the experiments include:

  1. Determine the biodegradability of five B100 and diesel fuels in the soil environment in both flask and lysimeter tests by GC analysis.
  2. Examine the biodegradation rates for REE B100 and diesel in a bottle test by GC analysis and CO2 evolution method.
  3. Assess the toxicity of the B100 and diesel fuels on plant productivity by a seed germination test.
  4. Compare the biodegradability of B100 in two different kinds of soil.

MATERIALS
The soil
All of the soil used in the experiments was collected from the University of Idaho Plant Science Farm at the same time except for a few experiments in which forest soil was used as a comparison for biodegradation in different soil types. The compositions of the two kinds of soil are listed in Table 1. The soil was sifted through a 2.5 mm sieve and covered and stored in a refrigerator to maintain the moisture. The soil percent dry weight was determined by weighing approximately 20 grams of the original soil, oven drying 24 hours at 105-110°C, and calculated using Equation 1.

Table 1. The Soil Composition.

Equation 1

% of dry weight = g of dry soil / g of wet soil x 100
where,

g = grams, the unit of soil weight.


Test substrates and controls
The test substrates included:

    1. Rapeseed Ethyl Ester (REE)
    2. Rapeseed Methyl Ester (RME)
    3. Soybean Methyl Ester (SME)
    4. Neat rapeseed oil (NR)
    5. Neat soybean oil (NS)
    6. Phillips 2-D diesel (2-D)

Controls were:

  1. Soil with no test substrate but all other reagents and solvents were the same as in the test substrate system.
  2. Sterilized soil made by autoclaving the soil 30 minutes for three consecutive days with test substrates.

The germination test seed was:

  • Legency alfalfa

The composition and specific properties of the test substrates are the same as those in (Biodegradation in an aquatic environment).

The stock solution for the soil experiments was made by dissolving NH(4)NO(3) and K(2)HPO(4) in the required volume of DIW (distilled and deionized water) to ensure the ratios of C:N = 60:1 and C:P = 80:1 (Dibble and Barthha, 1979).

A total of three test systems in the soil were conducted. The test substrates and the concentrations are given in Table 2. System I was to compare the biodegradability of five different B100 and diesel fuels in soil and REE in two different types of soil. In System II various concentrations of REE and diesel from low to high were tested to obtain the biodegradation rates. System III was the lysimeter and seed germination tests.

Test system design
Table 2. Test system design in the soil

Test Methods
The flask test

The shake flask test method is the same as that used in the "Biodegradation In The Aquatic Environment" notes.

Lysimeter test
Each of four 32 cm x 6 cm (diameter x height) lysimeters (Fig. 1) holding 2.0 kg soil (dry weight) was spread with one of four different test substrates including B100 REE, RME, and NR, and diesel at an approximately average concentration of 50,000 ppm. A lysimeter with no substrate was used as the control. Each lysimeter was covered by a thin plastic film (with small holes). The lysimeters were watered periodically to maintain required moisture. All the lysimeters were kept in a green house to maintain a favorable temperature for microorganisms and plants. At the required time intervals, the lysimeter soils were sub-sampled using the following procedure. A small spoon was used to collect a small pit of soil, a total of 10 times from different parts of a lysimeter to a depth of 3 cm. The sub-sampled soil was then thoroughly mixed to prepare three 2 g composite samples. The hydrocarbon residues in the sample were then monitored by GC analysis.

Figure 1. A lysimeter system.

At time zero, 30 g of substrate contaminated soil were removed from each lysimeter and transferred into a 500 ml Erlenmeyer flask to conduct a bottle test for comparison. The bottles were covered by aluminum film to avoid water from evaporating but to allow air in. At the same sampling times as previously described, 2 g soil (triplicate) were removed from each bottle for GC analysis. The results were compared with those from lysimeters.

Seed germination
In the same lysimeter system as above, 100 seeds of Legency Alfalfa were seeded in the fuel-contaminated soil in each lysimeter (Fig. 2) at day 1, week 1, 3, and 6 for testing seed germination. The seed germination rate equals the ratio of seed germinated in the substrate plate and control, as shown in Equation 2.

Figure 2. Distribution of 100 Legency Alfalfa seeds in a lysimeter.

Equation 2
% Seed germination = Seed ger. in substrate plate / Seed ger. in control x 100

Gas chromatograph analysis
Soil extraction

The 2 g soil sample obtained previously was placed into a 24 ml vial with a Teflon cap. The soil was mixed with the same volume of anhydrous sodium sulfate to absorb moisture in the sample. One milliliter of internal standard (with accurate concentration known) was added into the vial to determine the extraction efficiency and serve as a quantitative standard. Immediately after adding the internal standard, 9 ml of extraction solvent were added. The extraction then was conducted in a bath sonicator (Model: Branson 3200) for 30 minutes with less than 10 vials at one time. The extract was transferred into a vial, sealed, and kept in a refrigerator (4°C) prior to GC analysis.

The parameters of GC analysis
The parameters of GC analysis for the B100 and diesel were the same as that in the aquatic experiments

Percentage of substrate disappearance (primary degradation) determined by GC
The formula for calculating the percentage of the substrate disappearance determined by GC analysis was the same as that in the aquatic experiments.

Measurement of CO2 evolution
Periodically, the 10 ml of NaOH plus 10 ml rinsing water (DIW) was removed for CO2 measurement by titration with 0.1 N HCL to the phenolphthalein end point. The reservoir was refilled with fresh NaOH. The samples were analyzed at sufficient time intervals in the test period to allow for a smooth biodegradation plot or the slopes of the CO2 evolution curves to be determined. The calculation of percent CO2 evolution was the same as that in the aquatic experiments.

Determination of biodegradation rates
Determination of substrate disappearance rate or primary biodegradation rates (defined as volumetric substance removal rate, q(v), ppm day(-1)) from GC analysis was the same as that in the aquatic analysis (refer to section 1.8 on page 35).

Quality control and data analysis

  1. Duplicate experiments were applied to System IA, IIA, and IIIA
  2. Triplicate samples were used in each experiment(1) (therefore, total 6 samples for each substrate in System IA, ILA, and IIIA).
  3. The control was biologically inactive soil (soil was autoclaved three times in three continuous days) with the same substrates as other test samples.
  4. The data were analyzed statistically using the SAS approach.

RESULTS
Results of GC analysis (System IA)
In each of the duplicate experiments 1a and 1b, the three samples were averaged and the arithmetic means were calculated. The average substrate disappearance versus time for five different B100 fuels and diesel at the initial concentration of 10,000 ppm is summarized in Table 3. The substrate depletion cuves for two duplicate experiments 1a and 1b are schematically shown in Figures 3 and 4. In 28 days, the average percent substrate disappearance for five B100 fuels reached 76-92%, average of 81% and for diesel was 52%(11). It demonstrates that all B100 fuels are readily degradable in the soil environment.

As one can see from the data of Table 3 and Figure 5, in experiment lb the measured amount of residual hydrocarbon was depleted rapidly in 7 days. However, it increased slightly in 14 days, which is not reasonable. Possibly mistakes incurred in preparing the internal standard or GC parameters. Experiment 1b for diesel was no degradation observed at all. The reason for this was not clear.

The GC analysis of substrate disappearance at time 0, week 1, 2 and 4 for RME and SME are shown graphically in Figures 5 and 6, separately, where the peak diminishes from time 0 to week 4 can be seen.

Footnotes:
I. Duplicate was used in a few cases.

II. The duplicate experiment of the diesel failed -- no degradation observed.

Table 3. Substrate disappearance in 28 days by GC analysis for experiment 1a and 1b.

Figure 3. Substrate disappearance for five B100 fuels, and the control (sterilized soil) in experiment 1a (initial concentration = 10,000 ppm).

Figure 4. Substrate disappearance for five B100 fuels, diesel, and the control (sterilized soil) in experiment 1b (initial concentration = 10,000 ppm).

Figure 5. GC analysis: Biodegradation of RME in the soil at time 0, week 1, 2, and 4 (concentration = 10,000 ppm). The peak at retention time. (RT) = 2.8 - 2.9 is the internal standard Methyl 17:0.

Figure 6. GC analysis: Biodegradation of SME in the soil at time 0, week 1, 2, and 4 (concentration = 10,000 ppm). The peak at retention time (RT) = 2.8 - 2.9 is the internal standard Methyl 17:0.

Results of CO2 evolution (System IB)
The results from CO2 evolution for all the substrates are given in Table 4 and graphically presented in Figure 7. The rate of CO2 evolution was much slower than that of the substrate disappearance. The percentage of CO2 evolution for five B100 fuels was 15-20% with an average of 18% while that for 2-D was only 9% in 28 days.

Table 4. Percent C02 evolution in 28 days in the soil (concentration = 10,000 ppm).

Figure 7. C0(2) evolution vs. Time for five B100 fuels and diesel (concentration = 10,000 ppm).

Results of C02 evolution from two types of soil (system 1c)

The C02 evolution from REE in two different soil types at a concentration of 10,000 ppm (System lc) is summarized in Table 5 and shown graphically in Figure 8. These data show that the percentage of C02 evolution from the forest soil was slightly less than that from the farm soil. It suggests that soil type may affect the degragda ion of B100 fuels.

Table 5. CO2 evolution from two types of soils

Figure 8. CO2 evolution vs. time for REE from two types of soils: the farm and the forest soil (concentration = 10,000 ppm).

Results of System 11--Biodegradation rates for REE and diesel Results of GC analysis (System IIA)
Various REE concentrations were tested by GC method and triplicate samples averaged. The REE disappearance in 28 days and the rate for each concentration is given in Table 6 (the detailed data are summarized in Appendix 10) and the REE disappearance curves are graphically presented in Fig. 9. The result demonstrates that even at a concentration as high as 100,000 ppm, REE can be degraded over 50% in 28 days. Diesel was tested in the same manner as REE. The result for diesel is summarized in Table 8 and illustrated in Fig. 10. A comparison of the degradation of REE and diesel is suggested in Table 7). It is clear that the degradation of diesel is much slower than that of REE in all concentrations. The higher the substrate concentration, the larger the difference.

Table 6. Disappearance of REE vs. time in 28 days in the soil at various concentrations

Table 7. A comparison of percent disappearance of REE and diesel in 28 days

Figure 9. Disappearance of REE vs. time at various concentrations determined by GC analysis.

Figure 10. Disappearance of diesel vs. time at various concentrations derived by GC analysis.

Table 8. Disappearance of diesel vs. time in 28 days in the soil at various concentrations

The primary biodegradation rates, qV for REE and diesel can be derived from the slopes of substrate depletion curves (Fig. 9 and 10). The dynamic responses of REE and diesel can then be obtained by plotting qv against the substrate concentrations.

Results of CO2 evolution (System IIB)
The CO2 evolution was conducted at the same time as GC analysis for a comparison. The percent CO2 evolution for various concentrations of REE and diesel are summarized in Table 9 and 10 and illustrated in Fig. 11 and 12, respectively. As one can see, the CO2 evolution curves for diesel has a longer acclimation period and less degradation.

Table 9. CO2 evolution at various REE concentration vs. time

Table 10. CO2 evolution at various diesel concentrations vs. time

Figure 11. CO2 evolution from various REE concentrations vs. time in 28 days.

Figure 12. CO2 evolution from various diesel concentrations vs. time in 35 days.

A comparison of CO2 evolution and GC analysis
A comparison of primary degradation and CO2 evolution for REE in 28 days is presented in Table 11. As it illustrates, the CO2 evolution is much slower than the substrate disappearance. The higher the REE concentration, the larger the difference. A comparison of GC calibration curves at concentrations of 10,000 and 60,000 ppm are shown in Fig. 13 and 14. The former (lower concentration) shows that the maximum degradation rates from both CO2 evolution and GC analysis happened in the first 14 days. The latter shows that while the rate of REE depletion (by GC) slows down tremendously after week 2, yet the CO2 evolution increases constantly. The results suggest that CO2 evolution be not a suitable method for determining substrate biodegradation rate in a soil environment, especially at high concentrations. However, if calibrated properly and if the substrate concentration is low, it can still be considered as a method for qualifying the biodegradability of a pollutant. Experiment IC, comparing the biodegradation in two types of soil was conducted by the CO2 evolution method and was based on this consideration.

Table 11. A comparison of percent degradation and CO2 evolution of REE at day 28

Figure 13. A comparison calibration of CO2, evolution and GC analysis at a REE concentration of 10,000 ppm.

Figure 14. A comparison calibration of CO2 evolution and GC analysis at a REE concentration of 60,000 ppm.

Results of Lysimeter and seed germination tests (System III)
Results of Lysimeter test (System IIIA)
The substrate disappearance with time in the lysimeter and flask test are given in Table 12 and 13 and the percentage of substrate disappearance vs. time for both tests are schematically shown in Fig. 15 and 16, respectively (all the triplicate samples were averaged). The result shows that there is no significant difference on the primary degradation of B100 in the lysimeter and flask tests. For example, the degradation of REE in the lysimeter and the flask tests were 75% and 67% and the degradation of neat rapeseed oil in the two methods were 84% and 87%, respectively. But the figures show that the disappearance of diesel in the lysimeter in 48 days is twice what was measured in the flask test, 48% vs. 25%, which implies that diesel evaporation might occur in the diesel-spread lysimeter.

Table 12. Substrate disappearance in lysimeter test

Table 13. Substrate disappearance in flask test days Concentration (ppm)

Figure 15. Percent disappearance of REE, NR, and diesel vs. time in the lysimeter test

Figure 16. Percent disappearance of REE, NR, and diesel vs. time in the flask test

Again, the fungus growing on the B100 and the diesel in the lysimeters followed the same pattern as that in the bottles.

Result of seed germination
The result of seed germination test is summarized in Table 13 and percent seed germination for each substrate vs. time is schematically shown in Fig. 17. B100 fuels REE, RME, and NR-spread soils had lower seed germination rates at day 1 and week 1 seeding than week 3 and 4 since fungus grew rapidly and spread out everywhere in the lysimeter in two weeks after fuel-spreading. About 20% of the seeds germinated but died underground. However, after week 3 when most of the B100 was degraded and fungus began to disappear visually, the seed germination rates in B100-spread lysimeters increased. After week 6, all three B100 fuel-spread lysimeters reached the highest germination rates, 92-98%. Among the three B100 fuels, neat rapeseed has the highest average rate (from four seedings), about 87%. The seed germination from the diesel-spread lysimeter was at least 7 days later than that from B100-spread plates. The diesel-spread lysimeter did not see fungus growth until after week 4. This is probably the reason why the diesel-spread lysimeter had a rather high germination rate at week 3. But at the week 6 seeding, the diesel plate was full of fungus and the seed germination rate dropped tremendously (20%). Why the effect is so severe is not clear.

Table 14. Seed germination in the fuel-spread lysimeter soils seeded at day 1, week 1, 3, and 6 after the spreading.

Figure 17. Seed germination rates vs. time from all oil-spread lysimeters

CONCLUSION

  1. All the B100 fuels are readily biodegradable in the soil environment (both flask and lysimeter tests). At a concentration of 10,000 ppm, the average percentage of primary degradation for all the B100 fuels in 28 days was 81% and for diesel was only 54%.

  2. B100 REE has a rather high biodegradation rate in the soil environment. Even at the concentration as high as 100,000 ppm, its degradation rate reached 3900 ppm day-1. More than 50% of REE was degraded in 28 days. The degradation rate for diesel is much slower. The higher the concentrations, the larger the difference. At a concentration of 100,000 ppm, only 16% of the diesel was degraded in 28 days.

  3. The rates of C02 evolution (representing the mineralization rates) for both B100 and diesel are much slower than the primary degradation determined by GC. At concentration of 100,000 ppm, only 5% C02 evolution was detected from REE and 3% from diesel. At a concentration of 5000 ppm, the corresponding figures were 23% and 13%. It appears that attempts to use the C02 evolution method to determine the degradation of a pollutant in a soil environment may be limited. Soil type may have an effect on the biodegradation rate for a substrate. Therefore, the results in this presentation might only be suitable for the specific soil environments reported.

  4. There is no significant difference between lysimeter and flask tests in B100-spilled soil. The percent primary degradation of REE and neat rapeseed oil reached 75% and 84% in the lysimeters following 6 weeks of spraying at the initial concentration of 50,000 ppm. The percent primary degradation in the flask tests under the same conditions were 67% and 87% for REE and neat rapeseed. However, the percent primary degradation of diesel was 48% vs. 24% in the lysimeter and flask tests, respectively. Evaporation of diesel might occur in the diesel-spread lysimeter.

  5. Biodegradation can restore a B100 fuel spill contaminated soil in 4-6 weeks to a degree that it can support plant germination. However, the seed germination test showed that B100 contaminated soil did have an effect on plant growth for the first 3 weeks due to the rapid growth of microorganisms during the period of degrading the fuels. Four weeks after the spreading, the average seed germination rate for three B100 (REE, RME, and NR) was 85% and after 6 weeks from application of the oil reached 95% under the initial concentration of 50,000 ppm. The effect of diesel on plant growth needs additional testing.

REFERENCES
1. Burnside, O.C., Wicks, G.A., and Fenster, C.R., 1971. Dissipation of dicamba, picloram, and 2,3,6-TBA across Nebraska, Weed Science, 19: pp 323-325.
2. Cornish, Alex, Battersby, Nigel S. &Watkinson, Robert, J., 1993. Environmental fate of mineral, vegetable, and transesterified vegetable oils. Pestic. Sci., 37: 173-178.
3. EPA, 1975. Review and evaluation of available techniques for determining persistence and routes of degradation of chemical substances in the environment. US Department of Commerce National Technical Information Service, pp 224. PB-243 825.
4. Howard, P.H., Saxena, P.R. Durkin, and L.T. Ou, 1975. Review and evaluation of available techniques for determining persistence and routes of degradation of chemical substances in the environment, EPA 560/5-75-006.
5. MacRae, C. and Alexander, Martin, 1965. Microbial degradation of selected herbicides in soil, J. Agr. Food Chem., 13:72-75.
6. Sheets, T.J., Smith, J.W., and Kaufman, D.D., 1968. Persistence of benzoic and phenylacetic acids in soils, Weed Science, 19(2):217-222.
7. Wang, Xiaoping and Bartha, Richard, 1990. Effects of bioremediation on residues, activity and toxicity in soil contaminated by fuel spills, Soil biol. Biochem. 22(4):501505.

Revised 20041120