If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Petroleum hydrocarbon contamination resulting from fuel spills is a global issue that causes significant damage to the environment. This is particularly true in cold regions, where low temperatures reduce natural attenuation and degradation rates and make ecosystem recovery significantly slower than in temperate regions 1.
Healthy soils are important for maintaining overall ecosystem functionality and ecological processes, including plant productivity, degradation of organic matter, and nutrient cycling. These processes, as well as hydrology, aeration, and biogeochemistry in soils, are influenced heavily by biota living in the soil matrix. The abundance and diversity of microbial, fungal, plant, and invertebrate populations in soils are known to influence ecosystem processes and functions dramatically 2–4.
No universally accepted guidelines have been developed specifically for the assessment and remediation of petroleum hydrocarbon contamination in cold regions 5. Nations operating in polar and subpolar regions regulate the management of contaminated sites through domestic guidelines and legislation. To assess the impact of contamination on an ecosystem fully, a weight of evidence approach to environmental risk assessment, which includes chemical analysis, ecotoxicological testing, and ecological assessment is recommended 5, 6. Toxicity tests using a suite of species from different trophic levels are therefore an important component of risk assessments. Currently, little information is available on the toxicity of petroleum hydrocarbons and their impacts on ecosystem health within polar and subpolar regions. In addition, no standard toxicity tests with native terrestrial invertebrates have been developed for the subantarctic. In the absence of adequate polar and subpolar toxicity data, site managers in high latitudinal cold regions have relied on guidelines developed for temperate regions. The relevance of these data, which is based on the response of temperate species, is questionable when considering the unique characteristics of the polar environment and resident species. By gathering reliable toxicity data for native species, site managers will be able to incorporate ecologically relevant data into site-specific management guidelines more effectively.
The response of soil invertebrates to contaminated soils depends largely on the bioavailability of contaminants and the form in which they occur. Organic contaminants, including complex fuels, interact with the soil matrix and alter natural soil chemistry, which in turn influences the toxicity of the contaminant 7–9. Toxicity can be influenced by a range of soil properties, including pH and nutrient content; however, it is primarily soil organic carbon that influences the bioavailability of organic contaminants 9. Organic matter in the soil allows a larger percentage of organic contaminants to be sequestered into the soil matrix, making them less available to biota 10. In addition, a high content of organic matter decreases both the rate of release (vapor and/or aqueous phases) and the mobility of organic contaminants through the soil 11.
The composition of petroleum hydrocarbons in soil changes with residual time as a result of aging and weathering processes. Naturally occurring biological and physical processes, such as volatilization, leaching, and microbial degradation, selectively change the composition of petroleum hydrocarbon compounds, thereby altering the toxicity of contaminated soils to biota 12. Over time, as the chemistry of petroleum hydrocarbons in the soil changes, soils are left contaminated with what is referred to as an unresolved complex mixture of hydrocarbons (UCM; also referred to as recalcitrant petroleum hydrocarbon residuals) that are resilient to degradation 13, 14. In cold regions, these natural processes take extended periods of time, but can be enhanced through remediation practices. For example, adding nutrients and oxygen stimulate microbial populations, enabling them to degrade hydrocarbons faster and encouraging further volatilization 15, 16. Testing the toxicity of aged and fresh petroleum hydrocarbons in soils allows for realistic measures of toxicity, because the contaminants interact with the environment and change through time. This allows toxicity measures of the UCM to be incorporated into environmental risk assessments and guideline derivations.
Although several types of fuel have been used throughout the Antarctic and subantarctic, the most common fuel used at Australian Antarctic research stations is Special Antarctic Blend (SAB). This is a light diesel with alkanes in the range n-C9-14 peaking at around n-C12, with trace amounts of n-C15-23 also present. Special Antarctic Blend is also the most commonly spilled fuel at subantarctic Macquarie Island 17, where Australia operates a research station. Three sites associated with generating power for the station are known to have hydrocarbon contamination in the diesel range 17 and are currently undergoing bioremediation 15.
Earthworms are universally recognized as indicators of soil quality and are used extensively worldwide to assess the toxicity of contaminated soils 18–24. Standard toxicity tests have been developed by the Organisation for Economic Cooperation and Development (OECD) and the International Organization for Standardization (ISO) for the earthworms Eisenia fetida and Eisenia andrei25–27. Because the natural distribution of these species is limited to temperate and tropical regions, using these species for ecotoxicological testing in the subantarctic is inappropriate. Using the OECD and ISO standard procedures as a guide, tests were developed in the present study using a native Macquarie Island earthworm.
The present study examined the effect SAB had on the avoidance behavior, survival, and reproduction of Macquarie Island's most common native earthworm, Microscolex macquariensis. No single soil can represent fully the range of soil types that occur at Macquarie Island. To better understand the influence of organic matter content on toxicity, therefore, two soil types containing different organic carbon content were tested. The effects of fresh SAB contamination were evaluated, with additional efforts to examine the toxicity of aged SAB. By examining several endpoints, different soil types, and both fresh and aged SAB, a comprehensive evaluation of the risk of SAB to earthworms can be determined. This provides relevant data that can be used, in part, to derive site-specific petroleum hydrocarbon remediation targets for the subantarctic, which ensures contaminated sites no longer pose a significant environmental risk.
MATERIALS AND METHODS
Macquarie Island is a subantarctic island located approximately 1,500 km southeast of Tasmania in the Southern Ocean. It is part of the Macquarie Island Nature Reserve and is World Heritage listed. The island is under Tasmanian jurisdiction and is governed by the Tasmanian Parks and Wildlife Service, with a permanently occupied research station on the northernmost isthmus. Macquarie Island's climate is cool, wet, and windy with an average air temperature ranging from 3.3°C in the winter to 7.0°C in the summer. Mean soil temperature ranges from 2.8 to 8.5°C. Precipitation occurs approximately 312 d per year with an average annual rainfall of 920 mm and an average humidity of 89%. Winds are often gale force (>55.6 km/h) and are generally northwest to westerly. The average day length varies between 17 h in the summer to 7 h in the winter 28.
Three sites on Macquarie Island are contaminated with petroleum hydrocarbons (diesel-range), which are currently undergoing onsite remediation. These sites are located within station limits at the fuel farm and the main power house; structures that are associated with storing and using fuel for the station's power supply. The most recent fuel spill of approximately 500 L of SAB occurred in 2002 at the main power house. Other contaminated sites are historic with no accurate records of the time at which the spills occurred or the volume of fuel spilled. These contaminated sites have been described in detail previously 15, 17. Several options for remediation were assessed, and the limitations of natural hydrocarbon degradation were investigated 15. Based on a series of laboratory studies, in situ remediation, which enhanced microbial hydrocarbon degradation by aeration and nutrient addition, was determined to be the best option 16. Full-scale onsite remediation began in 2009.
The earthworm M. macquariensis is endemic to Macquarie Island, occurring in soils ranging from vegetated coarse sands at the shoreline to moist peat and mire soils on the plateau. They generally grow to a length of 24 to 40 mm but can reach 90 mm with a diameter of 2.5 to 4.6 mm 29. Little is known about the organism's life history and biology.
Earthworms used for these experiments were collected in May 2010 by hand sorting soil from an uncontaminated site on Macquarie Island (−54.4974°, 158.9395°). Worms were transported at 4°C to the Australian Antarctic Division, Kingston, Tasmania, where they were transferred to a temperature-controlled room held at 8 ± 2°C. They were maintained in peaty soil and were given sufficient water to simulate field moisture content of approximately 70% (on a soil dry wt basis), which mimics that found at Macquarie Island. Earthworms were fed oatmeal hydrated with deionized water every few days as required, and any food not consumed was removed to prevent the growth of mold. Although ISO and OECD protocols were used as a guide, the number of earthworms per vessel and number of replicates per treatment differed for the acute survival and reproduction tests. This was due to the fact that limited earthworms were available from the field collection.
Soils used in toxicity tests were collected from an uncontaminated site on Macquarie Island (−54.4974°, 158.9395°). Soils were hand sorted to remove any invertebrates, stones, surface plant material, and fine roots and were then homogenized. Soils were maintained at field moisture content (based on dry wt) before and during testing (Table 1). Two soil types were collected and tested: a peaty soil with high carbon content and a coarse, sandy soil with low carbon content. High carbon soil was collected in two different batches from the same site; as a result, differences in soil physicochemical properties existed between the soils used in the tests (Table 1). The avoidance test was undertaken using a mixture of high- and low carbon soil to allow for a sufficient quantity of soil to perform the test. The physicochemical properties of the soils used in the tests were analyzed and presented in Table 1. Methods used to characterize soils and their associated extracts were based on standard methods 30, 31. Soil were spiked by mixing the required amount of SAB into 10 g of acid-washed sand and thoroughly mixing the SAB/sand mix into 1,080 g of soil until homogenized. All treatments and controls had the same amount of acid-washed sand added. The peaty, high carbon soils, on a dry weight basis, were extremely light. Therefore, to achieve to same target SAB concentrations as the sandy soil the quantity of SAB required was much less. Using such low volumes of fuel was problematic to achieving a homogeneous spike. To standardize spiking methods, the amount of SAB required was determined on a soil wet weight basis. Aged treatments were spiked four weeks before the test began, were separated into exposure jars, and kept under field conditions (8°C, 16:8 h light:dark). The jar lids remained on during this time and were removed once a week to mix the soil briefly and allow aeration. To allow for the SAB concentration to decrease with the aging process, aged treatments were spiked initially at higher concentrations than required. For the purpose of the present study, aging refers to a change in the chemical composition and concentration of SAB in the soil from its original form through volatilization, biodegradation, and sorption and sequestration 32. It is acknowledged that different degrees of aging exist, and the composition of the residuals will vary with time. Changes in SAB concentration at the initial spike, the beginning of the test (four weeks), and at the end of the test (16 weeks) are presented in Table 2.
Table 1. Physicochemical properties of soils used in each test
aData expressed as a mean (±standard deviation) of all replicates (n = 3).
Initial concentration of SAB at time 0.
Concentration of SAB at the beginning of the chronic reproduction test (after four weeks of aging).
Concentration of SAB at the end of the chronic reproduction test (after 16 weeks of aging).
TPH = total petroleum hydrocarbons (measured in mg SAB/kg soil, dry wt. The detection limit was 50 mg SAB/kg soil).
Two-chamber avoidance test
This sublethal test, derived from the ISO 17512-1 guidelines 25, assessed the earthworms' behavioral response to contamination. The test was carried out using one homogenized batch of field-collected soil in 1.7 L rectangular plastic containers, divided vertically into two equal sections. Half of the vessel contained freshly spiked SAB soil, and the other half contained uncontaminated control soil, both to a depth of 5 cm. To avoid directional bias the same side of the container always received the control soil. A control treatment with uncontaminated soil on both sides was also tested. Soils were spiked with SAB at three target nominal concentrations (125, 250, and 500 mg SAB/kg soil), and five replicate test containers were prepared per treatment. At the start of the tests, the divider was removed, and 10 adult earthworms were placed along the separating line in each test container. Containers were covered to prevent the earthworms from escaping and were incubated in temperature-controlled culture cabinets for 48 h under field-simulated conditions (8°C, 16:8 h light:dark). At the end of the test period, the control and test soils in each container were separated by adding the divider before the containers were removed from the culture cabinet. The number of earthworms on each side of the container was counted. Earthworms split by adding the divider were counted as one-half. Missing earthworms were counted as dead, and containers with more than one dead earthworm were considered invalid.
14-day survival test
The 14-d acute survival test was derived from OECD guidelines 207 and 222 26, 27. Soils were spiked with fresh SAB at five target nominal concentrations (100, 200, 300, 500, and 1,000 mg SAB/kg soil). Three replicate 500-ml glass jars, each containing 360 g of the SAB test soil or uncontaminated control soil (0 mg SAB/kg soil) were prepared per treatment. Two soil types were used (high carbon and low carbon content). Five sexually mature earthworms were added to each jar. Jars were incubated in temperature-controlled culture cabinets under field-simulated conditions (8°C, 16:8 h light:dark) for 14 d. Lids were kept on the jars for the duration of the experiment to prevent volatilization of the SAB. After 7 d, the lids were removed briefly to allow aeration. Earthworms were not fed during this experiment. At the end of the test, earthworm survival was measured by counting and recording the number of earthworms that survived in each test unit. Missing earthworms were assumed dead. The test was considered invalid if more than one adult per exposure jar in the controls died.
12-week reproduction and survival test
The 12-week reproduction and survival test was derived from OECD guideline 222 27. Soils were spiked with SAB at six target nominal concentrations (250, 500, 1,000, 2,500, 5,000, 10,000 mg/kg). Three replicate 500-ml glass jars, each containing 360 g of the SAB test soil or control soil (0 mg SAB/kg soil) were prepared per treatment. Two soil types (high- and low carbon content) and two contamination treatments (fresh and aged SAB) were tested. To account for juvenile earthworms or cocoons remaining in the field-collected soil after thorough hand sorting, a second control soil (0 mg SAB/kg soil) was introduced that contained no additional earthworms. Earthworms or cocoons found in these controls were averaged and used as a correction factor for each soil type. Five sexually mature earthworms were added to each jar. Before the earthworms were added, they were rinsed, blotted dry, and weighed. During exposure, earthworms were fed oatmeal hydrated with deionized water. The moisture content of the soil was maintained at the set percentages reported in Table 1. All exposure jars were incubated in temperature-controlled culture cabinets under field-simulated conditions (8°C, 16:8 h light:dark) for six weeks. Jar lids were removed weekly to allow aeration and to remove unconsumed food, which was replaced with fresh food. At this time, jars were also weighed without their lids to account for water loss. Any losses were replenished with deionized water as required. Water content did not vary by more than 10% from the beginning of the test.
The exposure period for the adult survival test was six weeks, after which surviving adult earthworms were removed from the exposure jars and again rinsed, dried, and weighed. Jars were then returned to the culture cabinet for an additional six weeks to allow for juveniles to emerge from cocoons. Juveniles were extracted from the exposure jars by immersing the jars in a warm water bath (∼30°C), which encouraged earthworms to come to the surface. This continued until no more earthworms emerged for a period of 30 min 27. Cocoons were extracted by wet sieving the soil from each jar. Soil was washed through a series of sieves (one-half, 1, 2 mm), with the majority of cocoons found on the 1 mm sieve. Cocoons were then transferred to a beaker of water. Floating cocoons were scored as hatched, and cocoons that did not float were scored as unhatched 27. Test endpoints measured were six-week adult survival and the number of juveniles and cocoons produced after 12 weeks. The test was considered invalid if more than one adult per exposure jar in the controls died. Due to the lack of accurate life history knowledge on the test species, a validity criterion could not be specified for reproduction.
Analyzing soil properties
Total petroleum hydrocarbon
Approximately 10 g of soil was sampled from each exposure jar and added to a 40 ml glass vial with a Teflon septum in the lid. To each vial, 1 ml of hexane spiked with internal standard (containing 250 mg/L 1-bromoeicosane; 25.9 mg/L anthracene-d10; 49.8 mg/L ethylbenzene-d10; 62.4 mg/L 1-fluoroheptane; and 250 mg/L cyclooctane), 10 ml hexane, and 10 ml distilled water was added. Vials were weighed to four decimal places after each addition. The vials were tumbled overnight to allow for thorough mixing, and samples were then centrifuged for 10 min at 200 g. The hexane extract was removed using a glass pipette and stored in an 8 ml glass vial with a Teflon-lined lid until analyzed. Vials were oven dried at 105°C to obtain a soil dry mass.
Extracts were analyzed for TPH by gas chromatography using flame ionization detection (GC-FID; Agilent 6890N with a split/splitless injector) and an auto-sampler (Agilent 7683 ALS). Separation was achieved using an SGE BP1 column (35 m × 0.22 mm ID, 0.25 µm film thickness), with 1 µl of extract injected (1:15 pulsed split) at 310°C and 30 psi of helium carrier gas. After 1.3 min, the carrier gas pressure was adjusted to maintain a constant flow at 1.3 ml/min for the duration of the oven program. The oven temperature program was started at 50°C (held for 3 min) and increased to 330°C in increments of 18°C per min. Detector temperature was 340°C. Total petroleum hydrocarbon concentrations were determined using a calibration curve generated from standard solutions of SAB. Total petroleum hydrocarbons were measured using the ratio of the total detector response of all hydrocarbons to the internal standard peak response. The detection limit for SAB hydrocarbons was 50 mg SAB/kg soil. Reported TPH concentrations were the SAB hydrocarbons range (n-C9-18 and associated unresolved complex mixture). Note that this is outside the range of biogenic hydrocarbons; therefore, concentrations are not influenced by naturally occurring hydrocarbon signatures.
Carbon content, water content, pH, electrical conductivity, and grain size
Carbon content was determined by loss on ignition. Approximately 2 g of soil were oven dried at 105°C to remove the soil water, then combusted at 550°C to remove the soil organic carbon. The percentage of water content was determined by drying 10 g of field-collected soil at 105°C. Soil pH and electrical conductivity were measured using a Radiometer PHM 210 pH meter and WTW 197i conductivity meter, respectively, with a 1:5 (by volume) soil:water suspension. Soil grain size percentages were determined using laser diffraction using a Malvern laser particle size analyzer. Before analyzing the grain size, organics were removed by immersing samples in 10 to 20 ml of 10 to 15% hydrogen peroxide. Grain size was divided into three fractions (gravel >2,000 µm; sand 63 to 2,000 µm; and mud <63 µm, where mud was a combination of the silt and clay fractions).
Analyzing the anions in water extracts of soil was carried out by ion chromatography using a Metrohm 761 Compact IC connected to a 766 IC Sample Processor. Additionally, a Bischoff Lambda 1010 UV-Vis detector was employed to confirm the identification of in samples. Separation of anions ( and ) was achieved within 18 min by injecting a 20 µl sample onto a Metrosep A Supp 5-150 column flushed with 3.2 mM Na2CO3/1.0 mM NaHCO3 eluent solution at 0.75 ml/min. Calibration of the conductivity detector was achieved using standard solutions in the range 0 to 50 mg/L; where necessary, more concentrated samples were diluted to allow for quantification within this range.
Ammonium () and phosphate () in soils were measured from KCl and extracts, respectively. Using the Tecan M200 Infinite microplate reader, concentrations of and were measured using colorimetry absorbance following a procedure based on Laskov et al.'s 33 microphotometric method. A linear calibration was constructed for standard solutions to calculate the concentration of and in the extracts.
Avoidance behavior was expressed as NR = (C–T)/n × 100, where NR = net response, C = number of earthworms in the control soil, T = earthworms in the test soil, and n = total number of earthworms. A positive net response indicates the earthworms were avoiding the test soil. The habitat function of the soil was considered impaired if on average ≥80% of earthworms were found in the control soil. For each concentration, the mean number of earthworms in the test soil was compared to the mean number of earthworms in the control soil using a one-tailed Student's t test. Results showing significantly more earthworms in the control soil compared to the test soil indicated an avoidance response 25. Point estimates, including effective concentrations (EC) and lethal concentrations (LC), were determined using the statistical software package ToxCalc Version 5.0.32. Data was tested for normality using Shapiro-Wilk's test and log (x + 1) transformed. Dunnett's test was performed to determine no observed effective concentration (NOEC) and lowest observable effective concentration (LOEC) values. Point estimates were determined using the maximum-likelihood probit test for normal data or the Trimmed Spearman-Kärber test for non-normal data. Linear interpolation was used to determine point estimates for reproduction data. Before analysis, juvenile and cocoon numbers were amended by subtracting the average number of juveniles and cocoons found in the respective controls without earthworms. All TPH concentrations were measured SAB range (n-C9-18 and associated unresolved complex mixture) at commencement of the test and were reported on a soil dry weight basis.
Soil total petroleum hydrocarbon concentrations
Measured total petroleum hydrocarbon (TPH) concentrations deviated by a maximum of 50% from target nominal concentrations in low carbon, fresh spiked soils (Table 3). However, measured TPH concentrations were approximately 300% higher than target nominal concentrations in high carbon, fresh spiked soils (Table 3). Homogeneity was achieved with the majority of treatments having approximately 15% variation in TPH. The greatest variation of approximately 30% was observed in the high carbon, aged soils (HCA; Table 2). For the aged treatments, TPH concentrations reduced substantially during the four-week aging period and reduced even further by the end of the experiment (Table 2).
Table 3. Measured Special Antarctic Blend (SAB) diesel concentrations used to determine point estimates for freshly amended SAB treatments.
High carbon soil
Low carbon soil
Data expressed as a mean (±standard deviation) of all replicates (n = 3).
TPH = Total petroleum hydrocarbons (measured in mg SAB/kg soil [dry wt]. The detection limit was 50 mg SAB/kg soil); NT = not tested.
Survival, growth, and reproduction in controls
In all tests, 100% of the adults survived in the controls. The average weight gain of adult earthworms in the controls after six weeks was 0.032 ± 0.002 g/earthworm, equivalent to approximately 7%. After six weeks, five earthworms produced an average of 7 ± 2.5 cocoons with 9 ± 4 earthworms per cocoon.
Earthworms showed significant avoidance at all treatment concentrations (all Student's t tests, p < 0.01). The earthworms' avoidance response indicated limited habitat functionality (≥80% avoidance level) at the lowest test concentration of 181 mg SAB/kg soil (Fig. 1). Exposure to higher SAB concentrations did not indicate limited habitat functionality by percentage avoidance. However, at higher SAB concentrations, earthworms developed sores on their epidermis and had reduced mobility, suggesting an inability to avoid contamination. In addition, one earthworm died in one replicate at the highest exposure concentration.
Earthworms were approximately 10 times more sensitive to SAB in low carbon, fresh spiked soils (LCF) compared to high carbon, fresh spiked soils (HCF) (median lethal concentration [LC50] = 103 and 1,114 mg SAB/kg soil, respectively; Table 4). Significant mortality relative to the control was observed at 182 mg SAB/kg soil and 1,425 mg SAB/kg soil for the LCF and HCF treatments, respectively (LOEC values; Table 4).
Table 4. Lethal concentrations for acute (14-d) and chronic (42-d) survival tests with the earthworm Microscolex macquariensis.a
Points estimates were calculated from measured Special Antarctic Blend diesel (SAB) range (n-C9-18 and associated unresolved complex mixture). Total Petroleum Hydrocarbons concentrations were calculated on a soil dry-weight basis and reported in mg SAB/kg soil.
CL = Confidence limit; HCF = high carbon soil spiked with fresh SAB; LCF = low carbon soil spiked with fresh SAB; HCA = high carbon soil spiked with SAB and aged for four weeks; LCA = low carbon soil spiked with SAB and aged for four weeks; NC = unable to calculate value.
Chronic reproduction and survival
Aged SAB was more toxic than fresh SAB to adult earthworm survival (Table 4). Fresh SAB, however, was more toxic than aged SAB to earthworm reproduction (Table 5). Earthworms were more sensitive to SAB in low carbon soils (LCA) than in high carbon soils (HCA and HCF), both in terms of survival (Table 4) and reproduction (Table 5). Due to low survival at low SAB concentrations in the acute test and the limited number of M. macquariensis individuals available to use in tests, fresh SAB contamination in a low carbon soil (LCF) was not tested in the chronic reproduction test. Cocoon production was less sensitive than juvenile production, as indicated by higher median effective concentration (EC50) values across all treatments (Table 5). For freshly spiked, high carbon soil, juvenile production was the most sensitive endpoint, with an EC50 of 317 mg SAB/kg soil (HCF). For aged SAB treatments, juvenile production was more sensitive to SAB in the low carbon soil than in high carbon soil, with EC50 values of 130 mg SAB/kg soil and 1,753 mg SAB/kg soil, respectively. No differences were observed in the ratio of hatched to unhatched cocoons or changes in earthworm biomass between treatments. Where possible, EC20s were reported for reproductive endpoints to provide more relevant endpoints to develop guidelines (Table 5).
Table 5. Effect concentrations for chronic reproduction tests with the earthworm Microscolex macquariensisa
Points estimates were calculated from measured Special Antarctic Blend diesel (SAB) range (n-C9-18 and associated unresolved complex mixture) Total petroleum hydrocarbons concentrations were calculated on a soil dry-weight basis and reported in mg SAB/kg soil.
CL = Confidence limit; HCF = high carbon soil spiked with fresh SAB; HCA = high carbon soil spiked with SAB and aged for four weeks; LCA = low carbon soil spiked with SAB and aged for four weeks; NC = unable to calculate value.
Special Antarctic Blend diesel in the low carbon soil was generally more toxic to adult survival than in the high carbon soil, regardless of aging (based on LOECs, because the LC50 for low carbon aged soil could not be determined). The same trend is true for reproductive endpoints (juvenile and cocoon production); however, the effect of SAB in low carbon, freshly spiked soils was not tested. For high carbon soils, fresh SAB was more toxic than aged SAB for reproductive endpoints; however, the opposite was observed for adult survival. Of the reproductive test endpoints, juvenile production was more sensitive than cocoon production to SAB, regardless of the soil type and aging.
The physicochemical properties of the soils play a significant role in the relative toxicity of SAB to M. macquariensis. The acute survival test was run using two different Macquarie Island soils (Table 1): a low- and a high-carbon soil. There was a 10-fold increase in fresh SAB toxicity with decreasing carbon content between these two soil types (high carbon LC50 of 1,114 mg SAB/kg soil; low carbon LC50 of 103 mg SAB/kg soil). This difference in earthworm sensitivity to SAB exposure between the two soil types observed in the present study was as expected. The biological availability of chemicals to soil biota may be altered by the binding of contaminants to soil particles. In particular, the level of organic carbon in soil is known to affect the bioavailability significantly of a range of contaminants, including hydrocarbons 10. It can be expected, therefore, that soils with higher organic carbon content will have reduced availability of a contaminant and therefore reduced toxicity.
Aging occurs through the loss of the volatile and degradable hydrocarbon components, as well as sorption and sequestration of hydrocarbon components into the soil matrix. As a fuel spill ages, it changes the physicochemical and biological properties of soil, thereby altering the bioavailability and toxicity of contaminants 32. Evaporation is responsible for the loss of the volatile components, typically the n-alkanes (n-C9-14) and isoprenoids; leaching removes water soluble components (although with a limited effect on the bulk concentration of diesel due to the low solubility of bulk diesel compounds); and microbial degradation selectively transforms compounds based on their bioavailability. Microbes degrade n-alkanes first, typically lightest to heaviest, causing isoprenoids to become more dominant compared to the n-alkanes in aged soils 12. This process leaves the soil contaminated with organic compounds referred to as an unresolved complex mixture (UCM). The UCM is comprised of highly branched compounds, such as alkylbenzenes and tetralins, which resist natural weathering processes. Therefore, even after extensive biodegradation, UCMs often persist in the environment 14, 34. To estimate the toxicity of petroleum hydrocarbons in the environment more realistically, it is essential to understand the changes in toxicity that occur through biodegradation, binding, and sequestration processes and factor these into risk assessments. This study has started that process by including treatments aged for four weeks into the chronic reproduction test. Aged SAB contamination in low carbon soil was significantly more toxic than in high carbon soil, with EC50s of 130 mg SAB/kg soil and 1,753 mg SAB/kg soil for juvenile production, respectively. Comparing endpoints for fresh and aged SAB contamination in high carbon soil for juvenile production (EC50s of 317 and 1,753 mg SAB/kg soil, respectively) shows the significant extent to which aging can reduce the toxicity of hydrocarbon contaminated soil. Unexpectedly, however, aged SAB was more toxic to adult M. macquariensis survival than fresh SAB for high carbon soils. A possible reason is the aging process used. For the aged treatments, initial test concentrations were achieved by spiking the soils with more SAB than required. This allowed the processes of volatilization, biodegradation, and sequestration to occur to reduce the concentration to the initial test concentration. At the start of the experiment, aged treatments were more stable with less volatile components than the fresh treatments. This was due to the fact that low molecular substances disappear more rapidly than higher molecular substances during aging 35. The SAB concentration in fresh treatments, therefore, decreased more than in the aged treatments during the exposure period. In addition, the longer the soils are left to age, the more hydrocarbons are sequestered into the soil matrix, reducing their bioavailability. Given the foregoing, adult earthworms in the aged treatments were potentially exposed to higher concentrations of SAB for a longer period of time, resulting in perceived greater toxicity of the aged treatments compared to the fresh treatments. However, as the soil aged further and the contaminants became less bioavailable, the aged contaminated soil was less toxic than the fresh contaminated soil, possibly accounting for the difference in toxicity between the aged and fresh contamination in high carbon soil for adult survival and juvenile production. Further research into the effects of aged SAB contamination, focusing on the toxicity and effects of UCMs on the survival and reproduction of M. macquariensis, is required before definitive conclusions can be drawn on how this aging process affects toxicity.
Endpoints calculated in the present study add to existing toxicity data for Macquarie Island soils and will be used to develop site-specific remediation guidelines for hydrocarbon contamination, both for Macquarie Island, and more generally, for other subpolar and polar regions. Rayner et al. 15 used a hydrocarbon distribution model to determine that nonaqueous phase liquid (NAPL) droplets start to form in Macquarie Island soils at hydrocarbon concentrations between 50 to 1,000 mg/kg, depending on soil type and carbon content. Soil biogeochemical endpoints, including nitrification, denitrification, carbohydrate utilization, and total soil respiration, were developed for Macquarie Island soils by Schafer et al. 36 to examine SAB contamination. The potential nitrification activity was the most sensitive endpoint to the native microbial community, with an IC20 value of 190 mg SAB/kg soil. These studies present sensitivities consistent with the present study and are well below the recommended investigation levels for diesel fuel (n-C9–40) contamination of 1,000 mg/kg set by the National Environment Protection Council 37. This suggests that the current investigation levels are not protective of sensitive ecological receptors and that lower and more conservative values are required to protect soil health in the subantarctic. In addition, the influence of soil carbon content needs to be considered in risk assessments. Canada has produced guidelines based on soil type (coarse vs fine-grained soils) to take into account that various soil types effect different toxicological properties 38. This system or a similar approach would be appropriate for developing remediation guidelines for sites in the subantarctic, such as Macquarie Island. Macquarie Island is governed by Tasmanian State Jurisdiction and hence falls under Australian legislation for hydrocarbon remediation guidelines. Investigating levels for diesel range petroleum hydrocarbon contamination currently used in the Arctic vary greatly, ranging from 100 mg/kg in Norway to 2,000 mg/kg in Canada 5. Such differences in standards among countries lead to questions regarding the most appropriate environmentally relevant remediation target. The most ecologically protective and relevant guidelines should incorporate thorough site-specific ecological risk assessments, encompassing chemical, toxicological, and ecological data, highlighting compounds that require remediation to pre-determined protective concentrations 5, 6. Using M. macquariensis as an indicator of soil quality, the endpoints developed in the present study suggests that soil SAB concentrations of between approximately 50 to 200 mg SAB/kg soil would be a sufficiently protective remediation target depending on soil type. Earthworm data will be incorporated into species sensitivity distribution curves that include a range of taxa and endpoints to develop remediation targets that sufficiently protect the subantarctic at the ecosystem level 39, 40.
Microscolex macquariensis was chosen as a test species because it is a reasonably large, native, common earthworm on Macquarie Island. However, limited life history information is available for this species, and the time required for reproduction was unknown. Therefore, a 6- to 12-week exposure period for producing cocoons and progeny was chosen. It is possible that less time is required for a test with this species. Life history information reported in this study will allow testing with this species to develop in the future.
The soil types used in the present study were vastly different, but are representative of Macquarie Island soils. The peaty, organic matter rich soils were particularly problematic to amend with SAB. High quantities of organic material absorb quickly the SAB, making homogenization difficult. Compared to the sandy soil, the peaty soil filled a large volume, on a dry weight basis, making it impractical to add the same quantity of SAB to differing volumes of soil to achieve the target concentration. As a result, the quantity of SAB required was calculated on a soil wet weight basis and standardized between soils. Because the peaty soil contained a greater percentage of water, calculations of SAB concentration, on a dry weight basis, were inflated compared to target concentrations and concentrations attained in the sandy soil. This had a limited effect on the overall results, because all point estimates were calculated with measured SAB concentrations.
In conclusion, for effective hydrocarbon remediation guidelines to be set in environmentally fragile and ecologically unique areas such as Macquarie Island, ecotoxicological studies are required using both native invertebrate fauna and soil processes. The present study has effectively determined survival and reproductive endpoints for both a sandy, course soil and a peaty, organic matter rich soil for the earthworm M. macquariensis. The results reported will contribute to the data set that will be used to derive remediation targets for petroleum hydrocarbon contaminated sites on Macquarie Island. The results can also be used to develop petroleum hydrocarbon soil standards for the subantarctic and other polar regions with similar soil types.
Funding for the present study was provided by the Australian Antarctic Division. The authors thank the Risk and Remediation team at the Australian Antarctic Division, especially G. Hince, A. Palmer, S. Stark, and T. Raymond.