Final Report: Agronomic Optimization for Phytoremediation of Polycyclic Aromatic Hydrocarbons

EPA Grant Number: R831072
Title: Agronomic Optimization for Phytoremediation of Polycyclic Aromatic Hydrocarbons
Investigators: Nedunuri, Krishnakumar , Lowell, Cadance , Okunade, Samuel
Institution: Central State University
EPA Project Officer: Hahn, Intaek
Project Period: October 1, 2003 through March 31, 2006 (Extended to June 30, 2006)
Project Amount: $336,649
RFA: Superfund Minority Institutions Program: Hazardous Substance Research (2002) RFA Text |  Recipients Lists
Research Category: Hazardous Waste/Remediation , Land and Waste Management , Safer Chemicals

Objective:

Phytoremediation is the use of green plants to remove, contain, or render harmless environmental contaminants (Cunningham and Berti, 1993). This project is an innovative effort to optimize agronomic practices critical to cleaning up polycyclic aromatic hydrocarbons (PAHs) from refinery waste disposal sites using native plant species in managed systems. Several researchers have adopted phytoremediation as an alternative to more expensive and labor-intensive engineering controls (soil washing, chemical addition) to treat refinery waste sites (declared by the U.S. Environmental Protection Agency (EPA) as Superfund sites). Phytoremediation essentially has progressed in two directions:

  1. Using non-native species that are shown to survive toxic environments and at the same time reduce contamination with a concomitant biomass growth (managed systems).
  2. Allowing native species to naturally revegetate the contaminated sites.

Non-native species have the advantage of aggressively degrading contaminants, and at the same time supporting large biomass. A 50% reduction in total petroleum hydrocarbons was observed over 21 months using rye grass on a refinery waste site in southeast Texas (Nedunuri, et al., 2000). However, these species must be carefully managed, requiring frequent irrigation, fertilizer application, and pesticide/herbicide additions. These may well be targeted species for phytoremediation; however, by themselves, non-native species cannot restore the natural ecosystem. The non-native species also may disturb natural plant succession in the vicinity of the contamination.

Allowing native species to naturally revegetate is an attractive proposition, since it saves on the cost of restoration and management. If natural revegetation can remediate a site similar to a managed system and establishes a plant community comparable to that existing in the vicinity, the outcome will be both site remediation and ecological restoration (Henry and Shann, 2004; Jodi Shann, personal communication). However, it is not certain whether native species can tolerate such harsh and abruptly introduced toxic environments. Comparisons of ecological recovery through plant invasion and succession on sites contaminated with organic pollutants have shown biodiversity differences of recovering plant communities (Olson and Fletcher, 2000). In situations where native species can encroach into contaminated areas, they may take longer to establish, since contamination in several of these Superfund sites exceeds plant tolerance levels.

Both native and non-native species have shown promise in degrading contaminants and at the same time have the limitations indicated above. Therefore, our rationale has been that the success with either approach depends not only on the plant’s ability to survive, adapt, and succeed in degrading contaminants at Superfund sites, but also on how one can best optimize conditions conducive to plant growth using agronomic practices. These agronomic practices can include soil tillage/using non-tilled soils, soil amendments, surfactants, fertilizer addition, irrigation, and pesticide addition.

The research focused on optimization of agronomic practices typically used to promote remediation efforts at organic contaminated sites. A holistic approach to these efforts should also include site restoration to the natural ecological community prevailing in the region surrounding the contaminated sites. Attempts are underway to remediate a PAH-contaminated site to a prairie grassland using species native to Ohio that may subsequently undergo natural succession to mesophytic forest. Many clean up efforts so far have only addressed site remediation and used species targeted for waste reduction. These species may not promote natural community succession quintessential to site restoration. For instance, poplar trees were recommended for degradation of monoaromatic petroleum compounds (BTX) since these trees were found to enhance microorganisms that can enhance biodegradation in the rhizosphere (Jordahl, et al., 1997). Introduced trees, however, would hamper the natural succession of herbaceous vegetation such as grasses.

The principal objective of this study is to optimize three agronomic practices on PAH-contaminated sites—the addition of soil amendments, the addition of nitrogen, phosphorus, and potassium (NPK) fertilizers, and the timing of irrigation. The goal of this study is to facilitate site remediation and ecological restoration to the natural community. The study is unique since it addresses site remediation using native species in Ohio, unlike other studies, which have employed non-native aggressive species. Using native species is the first and most important step in restoring the site to natural ecological community development within a disturbed site.

Research Hypotheses

  1. The extent of PAH degradation depends more on the optimization of agronomic practices than on the choice of native grass species.
  2. It is possible to restore the PAH-contaminated sites to native prairie grasslands, which will provide direction to natural succession in these disturbed sites.

Summary/Accomplishments (Outputs/Outcomes):

Approach

Task 1 (First Year): Design and conduct greenhouse studies at Central State University (CSU) to investigate the effect of agronomic practices—compost additions, NPK fertilizer and irrigation scheduling—on reduction of PAH-contaminated soil using native grass species.

Task 2 (Second Year): Design and conduct greenhouse studies at CSU to compare PAH reduction in soils collected from nearby refinery waste sites in agronomically managed and unmanaged containers where natural succession is mimicked using a heterogeneous planting of three native grass species chosen for their potential for PAH remediation from Task 1.

The research involved screening of six Ohio native grasses to determine their potential in surviving under toxic PAH wastes, study site selection, physical and chemical characterization of soil, and design of individual treatments. We set up greenhouse experiments involving six grasses and no plant control X four treatments X two irrigation regimes to determine the best survivors under optimal agronomic conditions. Three separate individual studies were designed for our students to conduct research on these activities. Four students worked in teams on each of these activities. These studies are described below.

Initial Screening of Native Ohio Grass Species to Determine Their Survival on a Petroleum-Contaminated Sludge

Rationale. The research team visited Phytoremediation laboratories at the EPA Environmental Research Laboratories in Cincinnati and met with the lead investigator, Mr. Steven Rock. In discussions that emerged out of the meeting and tour of these laboratories, the research team was provided with important guidelines to carry out this research using petroleum-contaminated sludge. The main issues the team discussed at this meeting were the following:

  1. Aging of the sludge and its non-availability to the plants.
  2. Porosity reduction as sludge was homogenized, resulting in poor drainage.
  3. Difficulty in degradation of aged petroleum hydrocarbons in the presence of species such as poplars.
  4. Lack of primary substrates for growth of bacteria.

Knowledge of these issues before we started our research work helped in processing the sludge material so as to ensure porosity and also to make the sludge bioavailable to the plants. We also felt that it was critical to find native plant species that can survive the harsher petroleum sludge environments in a smaller pot study before we embarked on a detailed greenhouse study on optimizing agronomic practices.

Materials and Methods. Sludge was collected from a harbor site near Gary, IN. Six grass species native to Ohio—Indian Grass (Sorghastrum nutans), Indian Wood Oats (Unolia latifolia), Switch Grass (Panicum virgatum), Canada Wild Rye (Elymus canadensis), Prairie Brome (Bromus kalmii), and Side-oats Grama (Bouteloua curtipendula). were evaluated for their survival and growth on the sludge. The rationale for choosing these species was that they are native to Ohio, vary in soil pH tolerance from 5.0–7.0, and provide good rooting depth from 0.25 to 0.3 m or 0.55 to 0.6 m for possible rhizosphere degradation. The pots were 4-inch standard plastic pots layered from bottom to top with gravel, a layer of topsoil, a layer of raw sludge broken into pieces, and a layer of topsoil. The topsoil provided a surface for seed germination. The pots were watered as needed. Addition of a bottom gravel layer ensured less than or nearly equal field capacity conditions to be maintained in the pot. This provided adequate aerobic conditions within the pot. Each species was planted in two pots and grown for 4 weeks as shown in Figure 1. Extent of root penetration and leaf and root biomass by dry weight were measured. The soil/sludge-root mass was separated from the pots, and soil was carefully removed with water to recover the entire root mass and above ground leaf for each grass species. Figure 2 shows one such separation for Switch Grass. Above ground and root biomass were calculated per plant.

Figure 1. Canada Wild Rye and Side-Oats Grama Growing in a Small Pot Study After 1 Month of the Study

Figure 1. Canada Wild Rye and Side-Oats Grama Growing in a Small Pot Study After 1 Month of the Study

Figure 2. Leaf and Root Biomass for Switch Grass Separated From Soil and Sludge During the Initial Screening Study

Figure 2. Leaf and Root Biomass for Switch Grass Separated From Soil and Sludge During the Initial Screening Study

Results. The grasses were visually monitored for their growth over the period of time after the 2nd day, and 1st, 2nd, and 4th week of planting. Figure 3 shows the growth of two species—Side-oats Grama and Canada Wild Rye—after 2 days, the 1st week, and the 4th week in that order, from left to right. Both these grasses showed significant growth in 4 weeks. Indian Grass, Indian Wood Oats, and Prairie Brome did not show significant growth. Canada Wild Rye, Switch Grass, and Side-Oats Grama showed significant growth.

Figure 3. Growth of Side-Oats Grama and Canada Wild Rye After 2nd Day, 1st Week, and 4th Weeks of Planting
 
Figure 3. Growth of Side-Oats Grama and Canada Wild Rye After 2nd Day, 1st Week, and 4th Weeks of Planting
 
Figure 3. Growth of Side-Oats Grama and Canada Wild Rye After 2nd Day, 1st Week, and 4th Weeks of Planting

Figure 3. Growth of Side-Oats Grama and Canada Wild Rye After 2nd Day, 1st Week, and 4th Weeks of Planting

Canada Wild Rye visually had the largest root mass with extensive penetration into PAH-contaminated soil followed in order by Switch Grass, Side-Oats Grama, Indian Wood Oats, Prairie Brome, and Indian Grass. Canada Wild Rye also had the highest root biomass of 1.5 g/plant, followed by Indian Wood Oats with 0.5 g/plant, Side-Oats Grama with 0.3 g/plant, Switch Grass with 0.33 g/plant, Indian Grass with 0.12 g/plant, and Prairie Brome with 0.1 g/plant.

Conclusions. The preliminary screening study evaluated six grass species—Canada Wild Rye, Indian Grass, Indian Wood Oats, Prairie Brome, Side-Oats Grama, and Switch Grass—for their survival and growth on the sludge. The layered design of adding gravel, topsoil, sludge, and topsoil with amendments supported growth of these grasses. The major issues of inadequate low porosity of the sludge material, low aeration, and lack of nutrient enrichment typically encountered when these grasses were planted directly on the sludge have been resolved. Visual inspection and above and below ground biomass values suggested that Canada Wild Rye, Indian Wood Oats, and Side-Oats Grama may have the best survival potential when they are grown in the greenhouse study for optimizing the agronomic conditions to facilitate phytoremediation using selected native Ohio grasses.

Preparatory Work for Research

As a first step, selection of students was made based on good academic record, prior experience working on research projects, and an interest to pursue careers in the environmental fields. Two students from the Department of Biology (Mr. Williams and Ms. Burt) and two students from the Department of Water Resources Management (Mr. Meade and Ms. Smith) were selected. The principal investigators (PIs) spent the first month describing to the students (undergraduate research assistants, URAs) the nature of research, its importance to society, the hypothesis of the research, objectives, the toxicology of the contaminants, and their safe storage, handling, and disposal. URAs and PIs received 40-hour Hazardous Waste Operations and Emergency Response (HAZWOPER) training and obtained certification in the 1st week of December 2003. Efforts for selection of study site began with a visit to the phytoremediation laboratories of our project advisor, Mr. Steven Rock, at the EPA National Risk Management Research Laboratory (NRMRL) in Cincinnati, Ohio. Mr. Rock provided us with guidance in locating the PAH-contaminated sites that fall within EPA Region 5 jurisdiction. One of these sites was near the Gary, IN harbor and contained aged petroleum sludge. Initial screening efforts on finding potential native grass species as well as potential for biodegradation were made using a sample of the sludge collected from this location. Our efforts to collect the sludge in bulk quantities from Gary, IN did not meet with success due to extremely cold weather and difficulty in digging sludge hardened under prevailing frost conditions during the month of January. The PIs, with the assistance of the project advisor, were successful in identifying an industrial storage facility in Dayton, Ohio during the month of February. This facility experienced leaking of creosote onto the site and was identified as one of the high priority EPA clean up areas. The PIs obtained necessary clearance from EPA to collect the contaminated sludge to carry out phytoremediation in the CSU greenhouse using native grasses of Ohio. Both URAs and PIs went on a field visit to collect 300 gallons of contaminated sludge. The site contained high concentrations of creosote-based mixed waste containing PAHs and heavy metals. EPA has been investigating the use of phytoremediation for the clean up of this site. Students (URAs) spent one month of the summer at University of Cincinnati in Dr. Jodi Shann’s phytoremediation laboratory learning techniques for contaminant soil physical, chemical, and biological characterization. The training was carried out using the sludge obtained from the study site.

Soil-Contaminant Particle Size Distribution in Sludge Collected From Dayton, OH Undergoing Native Grass Phytoremediation

Abstract. This component of the overall study identified how characteristics of a typical silty loam soil could be altered due to its association with the creosote from the study site described in the previous section. Analysis of the waste conducted in the Department of Biology at the University of Cincinnati revealed excessive concentrations of PAHs and heavy metals. The site has two distinct regions: one hot spot (ditch), which contained the mixed waste having a mean PAH concentration of 4000 ppm, and a region towards the center of the site (center), which was mildly contaminated with a mean PAH concentration of 80 ppm. Site clean up efforts involved uniformly spreading contaminant, tillage, and application of vegetative treatments. Phytoremediation trials are underway using native grasses in a greenhouse study at CSU using layers of soil collected from this site. Remediation depends on bioavailability of these contaminants, and thus the water retention capacity of the contaminated soils. Soil characterization was performed on: (1) soil collected from the mildly contaminated region (mild); and (2) a 1:1 mixture collected from the ditch and mild region of the site (mixed). Soil gradation tests, bulk density, water holding capacity, plasticity index, and field capacity experiments were performed to determine differences in particle size and water retention at various size fractions. Soil gradation tests showed greater association of PAH to coarse soil particles passing through mesh number 10. Soil particle sizes finer than mesh number 100 were removed in the sample from the ditch due to their association with the PAH. The water holding capacity of the mixed soil associated with excessive waste was 30% lower than the corresponding value for the mildly contaminated soil.

Introduction. This study involved physical characterization of the sludge obtained from the study site in order to determine its particle size gradation, soil texture, soil pH, organic matter, and water holding capacity. The study is unique in addressing the effect of contaminant association on soil particle size distribution and hence on the water holding capacity of soils. The water holding capacity is key to determining the bioavailability of PAHs and other organics to the rhizosphere that will be supported by the six grasses used in this study.

Extensive studies have been conducted to determine soil particle size distribution (PSD) and its effect on water content determination in soils. Gee and Bauder (1986) suggested standard hydrometer methods to determine PSD in soils. Arya, et al. (1999) used a variable scaling parameter based on the number of particles within a given size distribution and used this parameter in a logistic growth equation to describe complete PSD in soils. Yang, et al. (2001) used fractal models to delineate PSD. Their model was validated with classic sieve analysis to determine the PSD. Single fractal dimension may not be adequate to characterize the particle sizes, which are shown to have a lognormal distribution. Whenever experimentally available PSDs are limited, complete distribution can be obtained using a logarithmic size distribution model. The model was validated for 125 soils using both wet sieving as well as hydrometer methods (Skaggs, et al., 2001). Hence, Posadas, et al. (2001) used multi-fractal dimensions to characterize contrasting particle sizes in soils. Dur, et al. (2004) used laser granulometry, and image analysis using transmission electron microscopy was used to determine the PSD of soil clay fractions. Their technique, however, was restricted to clay particle sizes finer than 0.2 micrometers. A lognormal PSD has been used to construct soil water retention curves, assuming that soil particles are random spheres distributed randomly in space (Chan and Govindaraju, 2004).

This experimental study focuses on PSD in disturbed soils contaminated with petroleum sludge. The influence of PSD on water holding capacity in sludge-affected soils is important to determine how roots can access water in soils associated with petroleum contaminants. Earlier studies (Gardner, 1986) suggested gravimetric methods to determine the water content in soils.

Materials and Methods. Contaminated soil was obtained from an industrial storage facility in Dayton, OH. The facility experienced creosote leaks and EPA had declared the site as a clean up priority under the Resource Conservation and Recovery Act (RCRA). The site has two distinct regions: a hot spot ditch, which contained the mixed waste (mixed) having a mean total PAH concentration of 4000 ppm in the sludge extract, and a region towards the center of the site, which was mildly contaminated (mild) with a mean total PAH concentration of 80 ppm. Samples were collected from both sites, and brought to the soils laboratory at CSU’s Department of Water Resources Management (WRM). The mild soil was oven dried for 24 hours, and the mixed soil was air dried for 48 hours, ground, and sieved for PSD. Figure 4 shows images of soil fractions of different sizes obtained after sieving samples from both center and ditch sites. Initial water content and the water content at saturation were determined for mild and mixed soils (three replicates) using gravimetric methods at Dr. Shann’s phytoremediation laboratories. The organic carbon content also was determined for soil samples (six replicates) collected from mild and mixed locations. Atterberg plasticity index was determined for samples from mild and mixed using standard procedures. Table 1 shows physical properties of these soil samples. All properties are determined using standard methods as outlined in Klute (1986).

Figure 4. Soil Fractions Sieved Through Meshes of Different Sizes for Mixed (Left Half) and Mild (Right Half) Soils

Figure 4. Soil Fractions Sieved Through Meshes of Different Sizes for Mixed (Left Half) and Mild (Right Half) Soils

Table 1. Physical Characteristics of Mild and Mixed Soils Collected From Mildly Contaminated and Highly Contaminated Regions of the Study Site

Location

Initial Water Content (gm/dry weight)

Water Content at Saturation

Atterberg Plasticity Index (strokes)

Total Organic Matter (gm/gm soil dry weight)

Soil pH

 

Mean

Std.

Mean

Std.

50 mesh

Mean

Std.

Mean

Std.

Mild

0.04

0.011

0.33

0.01

35

0.215

0.002

7.1

0.17

Mixed

0.04

0.006

0.14

0.01

38

0.28

0.003

6.67

0.02

Results and Discussion. Both samples have the same initial water content; however, mild soil has greater saturated moisture content than mixed soil. The mixed soil has a slightly higher organic matter. The pH of both samples remained close to neutral.

A simple sieve analysis has been utilized to compare the mild and mixed soils. Figure 4 shows sieve analysis showing soil particles of different size ranges retained in mesh numbers (corresponding to a given mesh size in mm) ranging from 1 (9.525 mm), 10 (2 mm), 20 (0.8 mm), 50 (0.3 mm), 100 (0.15 mm), 140 (0.1 mm), 200 (0.075 mm), and pan (less than 0.05 mm). The left half portion of Figure 4 shows soil PSD for mixed soil and the right half portion shows corresponding PSD for mild soil. It is evident that negligible amounts are retained on meshes 140, 200, and pan in samples collected from the mixed site. Most sludge adhered to soil particles in the size ranges corresponding to mesh numbers 10, 20, 50, and 100. The size gradation curve shown in Figure 5 is a lognormal plot of grain particle sizes against percentage of particles passing through a particular grain size. The gradation curve shows that center soil has 32.5% gravel, 60% sand, and 7.5% clay, whereas the ditch soil has 36% gravel and 64% sand. Figure 5 demonstrates that the soil associated with the sludge (mixed) retained greater mass in heavier size fractions when compared to that of uncontaminated soil (mild). It is clear that size fractions finer than 0.15 mm (mesh number 100) do not exist in mixed soil, which essentially means that sludge particles associate themselves with coarse to fine sand particles. Thus, small particle sizes corresponding to clay and silt fractions in the ditch soil adhered to coarse soil particle sizes due to their association with the petroleum sludge. Contamination of soil with sludge therefore tends to shift PSD from fine clay size particles to coarse gravel to fine sand particles, thereby increasing the porosity of the contaminated sludge.

Figure 5. PSD From Sieve Analysis for Soil Samples From Ditch and Center Locations

Figure 5. PSD From Sieve Analysis for Soil Samples From Ditch and Center Locations

The effect of this phenomenon on water retention by the soil particles corresponding to grain sizes 0.3 mm and 0.15 mm was determined by looking at the plasticity index. The higher the number of strokes it takes to break the soil, the greater is its ability to hold water. When the plasticity index was determined for soil fractions retained on mesh number 50, corresponding to particle size 0.3 mm, the mild soil took 35 strokes to break compared to the mixed soil, which took 25 strokes. Thus, mild soil has more water holding capacity than mixed soil. On mesh number 100, corresponding to a grain size of 0.15 mm, mild soil required 38 strokes and mixed soil required 33 strokes, yielding the same result, that even at particle sizes of 0.15 mm, water holding capacity is greater for mild soil than for the mixed soil. One can also notice that it is easier to hold water in mixed soil, as the particle size becomes finer, than in mild soil. The plasticity index was not determined for size fractions larger than 0.3 mm since these fractions cannot hold any water. Also, the plasticity index for fractions finer than 0.15 mm was not determined since mixed soil did not retain any weight corresponding to particle sizes finer than 0.15 mm.

Summary and Conclusions. This study uniquely described the effect of sludge association with soil on its PSD. In soils contaminated with sludge, finer particle sizes in the range of fine sand, silt, and clays tend to disappear as such and there is a shift in PSD from fine sand, silt, and clays to coarse sized particles such as gravel and coarse sand. This shift increases the porosity of the soil, in the size range greater than 0.3 mm, or mesh number 50. Thus, soils contaminated with sludge may tend to drain faster than uncontaminated soils. In the size ranges corresponding to mesh numbers 100 (0.3 mm) and 50 (0.15 mm), soils contaminated with sludge show lower water holding capacity than uncontaminated soils. The effect of sludge on soil used in this study is to reduce its water holding capacity. This observation is critical to the overall research where clean up of sludge-contaminated soils using native grasses is being investigated under optimal agronomic practices such as fertilizer and compost addition and irrigation. This component of the study suggests that sludge-affected soils require more frequent irrigation.

The findings of this research suggest that PSD models for soils need to be modified to take into account the effect of sludge on shifting the PSD from fine silt and clay size fractions to more coarse grained fractions such as coarse sand and gravel. The corresponding pedotransfer function models, which correlate the water retention to PSD, also need modification to incorporate the sludge content in soils. This study only qualitatively described the relative amounts of sludge adhering at different particle sizes; however, an exact quantity of sludge at different fractions is unknown. This information would be helpful when incorporating sludge content into PSD models and pedotransfer function models. Further study is needed to find the effect of sludge association with soils on the water retention curves of these soils.

Optimization of Agronomic Practices for Enhancing Growth and Remediation Potential of Six Ohio Grass Species in PAH-Contaminated Soils

Abstract. PAHs are a class of very stable organic molecules consisting of carbon and hydrogen that are very persistent in nature. PAHs are highly carcinogenic and found in oil refinery wastes. Phytoremediation is a low-cost method of using plants to degrade these wastes to clean up and reuse field sites that have been polluted with these harmful hydrocarbons. Certain grasses not native to Ohio were found to be PAH degraders; however, these plant species could not restore the natural ecosystem on Ohio industrial waste sites identified by EPA. The purpose of this research was to evaluate the remediation potential of grass species native to Ohio that can help reestablish the prairie ecosystem destroyed by PAH contamination. A greenhouse investigation is under way to measure the growth of six different grass species native to Ohio grown on PAH-contaminated soil. Seeds of Canada Wild Rye, Prairie Brome, Side-oats Grama, Indian Grass, Indian Wood Oats, and Switch Grass were germinated in 5-gallon buckets layered with topsoil, PAH-contaminated soil and topsoil and grown for 5 months. Each layer is 76 mm thick. Four replicate buckets per grass species were treated with soil amendments such as N-Viro Soil (processed municipal biosolid), N-Viro Soil and fertilizer (NPK), or NPK and only topsoil as a control. Half of the replicates were treated with either of two watering schemes, watering every fourth day (frequent) or watering every sixth day (infrequent). Plant height, density, and chlorophyll were measured and significant differences determined using analysis of variance (ANOVA). Environmental parameters within the greenhouse such as temperature, humidity, light, soil moisture, and soil pH also were measured. Canada Wild Rye and Side-oat Grama treated with N-Viro Soil under infrequent irrigation showed the best growth performance. Canada Wild Rye had a plant count of 64 per bucket, a plant height of 24 cm, and a chlorophyll content of 42.4 SPAD units. Side-oats Grama had a plant count of 128 per bucket, a plant height of 42 cm, but a lower chlorophyll content of 23.8 SPAD units. Data showed that plants performed better under more frequent irrigation. Further study examined root morphology on this disturbed soil using a root bed scanner. Soil cores (30 mm diameter and depth to the gravel layer) were obtained for each treatment after 5 months. PAH concentration in the sludge and leachate are currently being determined after treatment with grasses for this period in the presence of soil amendments and irrigation.

Introduction. Optimization of plant-based remediation of refinery waste disposal sites—in terms of arriving at an optimal combination of soil amendments, fertilizer additions, soil conservation practices, and irrigation scheduling—is critical in managing plant-based site remediation efforts. Few researchers in the past have investigated these agronomic practices. Villamouz and Wilke (2001) found that the addition of compost assisted ryegrass in degrading total petroleum hydrocarbons. Ebbs, et al. (1997) found enhanced accumulation of zinc in Brassica species when a commercial grade soil amendment (GRO-MOR) was added. Sub-surface irrigation and daily additions of soluble N and P in greenhouse studies have been shown to enhance plant root growth with a concomitant stimulation of petroleum-degrading microorganisms, resulting in total PAH degradation (Hutchinson, et al., 2001). A poor management of one or more of these factors will result in longer time for site clean up and associated excessive costs.

The objective of this greenhouse study component of the research was to investigate best agronomic practices under which native Ohio grass species could survive, grow, and degrade PAHs in contaminated soils.

Materials and Methods.

Design of Individual Treatments. Each treatment involved using a particular grass, native to Ohio, growing on top of the sludge treated with one of the soil amendments such as compost (N-Viro Soil), inorganic fertilizer (NPK), a combination, or topsoil without amendments. The treatment was carried out using 5-galllon buckets, as they have been found to be a suitable size to mimic outdoor rhizosphere conditions (personal communications with Mr. Steven Rock, Environmental Engineer, NRMRL, EPA). Figure 6 shows individual layers in any typical bucket. The topmost layer was seeded from the surface with one of the six native grasses chosen for this study. The efficiency of the treatments in degradation of PAH was tested on native Ohio grasses (and a control with no grass species). The rationale for the selection of these species among several native species of Ohio stems from a variety of attributes including grass height (1–2 m), soil pH tolerance (5.0 to at least 7.0), rooting depth (0.25–0.3 m or 0.55–0.6 m), and medium to high seedling vigor (United States Department of Agriculture-Natural Resources Conservation Service [USDA-NRCS], 2003). No species chosen is associated with nitrogen-fixing bacteria. The species selected were Indian Grass, Indian Wood Oats, Switch Grass, Canada Wild Rye, Prairie Brome, and Side-oats Grama. Switch Grass has been shown to degrade PAH (Hutchinson, et al., 2001).

Figure 6. Design of Individual Treatment in 5-Gallon Bucket With a Bottom Tray for Leachate Collection

Figure 6. Design of Individual Treatment in 5-Gallon Bucket With a Bottom Tray for Leachate Collection

The topmost layer was 7.62 cm topsoil amended with one of the soil amendments (treatments) as part of the proposed design. The next layer beneath the soil amendment layer was the 7.62-cm sludge layer, which was comprised of uniform layers of ditch and center soils obtained from the study site. Below the sludge layer was a 7.62-cm topsoil layer that was introduced to provide nourishment to the roots penetrating the sludge layer as well as ensure aerobic conditions. The bottommost layer was a 7.62-cm gravel layer for free gravity drainage of excess water. This unique design allowed for optimal water retention, nutrient availability, and aerobic conditions within the rhizosphere. Consequently, the design provided an opportunity for the roots to come into contact with the sludge layer. One uniqueness in our design is the restoration of the porous nature of the sludge, since we have not ground the sludge. Past experience showed that grinding the sludge to a uniform size reduces variability; however, this process leads to hardening of the sludge, thereby making it difficult for root penetration.

Initial Sludge Preparation and Chemical Characterization. The sludge was transported to the greenhouse from the study site and air dried. Soil samples at the contaminated site were analyzed (Spectrum Analytic Inc, Washington Courthouse, Ohio). The soil composition in the sludge is given in Table 2.

Table 2.

pH

P

K

Mg

Ca

CEC*

Zn

Fe

Cu

Mn

OM (%)**

7.9

51

117

313

5174

17.5

42

245

28.8

144

3

Soil nutrients are extracted by Mehlich-3 (ICP) and are reported in ppm.
*CEC is cation exchange capacity.
**OM (%) is percentage organic matter.

URAs conducted initial soil characterization at Dr. Shann’s phytoremediation laboratories in the Department of Biology at the University of Cincinnati in the months of July and August as part of their summer training. PAH was extracted using an accelerated solvent recovery system using acetone, filtered using a cellulose filter, and then analyzed using gas chromatography (GC) with a flame ionization detector (FID). The total PAH concentration averaged over eight replicated samples in the sludge from the ditch site was 4000 ppm in the extract, and the corresponding concentration in the sludge from the center sites was 80 ppm. The sludge was analyzed for total heavy metal content using atomic absorption (AA) with flame detection. Sludge collected from the ditch site had 838 μg Pb/g, 52.2 μg Cd/g, and 490 μg Zn/g averaged over three replicates on a dry weight basis. Corresponding concentrations in the center site were 324 μg Pb/g, 10.4 μg Cd/g, and 208 μg Zn/g averaged over three replicates on a dry weight basis.

Design of Treatments for the Greenhouse Study. The petroleum sludge was packed to the same bulk density and initial moisture content of the field soils. Six different combinations of compost, fertilizer, and irrigation were applied (see Table 3 below for Factorial Design of Treatments used in this study), with two irrigation-only controls. There were 24 treatments with a minimum of four replicates/treatment for each grass species as shown in Table 3. The topsoil control had only three replicates. There were three replicates serving as no-plant controls, however, undergoing irrigation. The number of replicates also was dictated by the space constraints in the greenhouse and the variability in soils collected from the sites.

Table 3. Factorial Design of Greenhouse Experiments to Evaluate Soil Amendments and Irrigation

Treatment

N-Viro Soil

NPK Fertilizer

Optimal Irrigation Schedules

Every 4th Day

Every 6th Day

1

X

 

X

 

2

X

 

 

X

3

 

X

X

 

4

 

X

 

X

5

X

X

X

 

6

X

X

 

X

7 (control—topsoil)

 

 

X

 

8 (control—topsoil)

 

 

 

X

Grasses were seeded into the 5-gallon buckets on July 20, 2004, in a random block design. N-Viro Soil was added to buckets 1, 2, 5, and 6. N-Viro soil is a pasteurized product manufactured from processed municipal biosolids (Fairborn, OH), which contains 0.5% N, 0.1% P, 0.1% K. NPK fertilizer (Scotts Starter Fertilizer) with the following composition was applied to the treatments labeled 3, 4, 5, and 6 shown in Table 3: 20% total N, 27% P2O5, and 5% K2O. The plants were watered from the surface to field capacity of the topsoil for one month so that all grasses germinated. Then, half of the treatments were irrigated once every 4th day (frequent) and the remaining half were irrigated every 6th day (infrequent). These schedules were derived by monitoring the soil moisture content in one of the control pots both in the topsoil and sludge layers. The layers were close to field capacity on the 4th day and completely dry after the 6th day. The same amount of water (0.2 gallons) was applied to each treatment. Figure 7 shows these treatments in the greenhouse. The 5-gallon buckets with orange bands wrapped around their top circumferences were the ones undergoing frequent irrigation. The greenhouse climate was monitored for minimum and maximum daily temperatures and humidity. Over the study period, plant height, plant density, soil moisture, and chlorophyll (using a chlorophyll meter [Minolta SPAD 502]) were measured once a month from each treatment.

Figure 7. Greenhouse (Department of Natural Sciences) Used for Task 1 Study Where the Individual Treatments Were Arranged in a Random Block Design

Figure 7. Greenhouse (Department of Natural Sciences) Used for Task 1 Study Where the Individual Treatments Were Arranged in a Random Block Design

After 5 months from the initial seeding of grasses, soil cores were removed and stored in a walk-in freezer at 4°C until analysis. Root characteristics such as length, surface area, and root density were determined using a protected flat bed root scanner, DELTA T Scan, and image processing software (Dynamax, Inc). Roots were extracted from the soil by hand, rinsed with water using a 2-mm sieve, blotted dry, and stored in the refrigerator at 32° F until further processing using the root bed scanner. Root biomass, both fresh- and dry-weight basis (above and below ground), root density, and root length were determined as described in Hutchinson, et al. (2001). Plant biomasses (above and below ground) were collected at the end of the study and measured as total dry weight.

The original sludge (sludge before treatment) from the site contained a mixture of petroleum hydrocarbons typically found in diesel fuel. The second year of our research therefore involved analysis of total petroleum hydrocarbons from the treatments undergoing phytoremediation in the presence of native Ohio grasses under different combinations of compost addition, NPK fertilizer, and irrigation scheduling. Earlier studies did not show significant uptake of these hydrocarbons into roots and leaves due to their high octanol-water partitioning coefficients. This study therefore evaluated only changes in total PAHs in the soil before and after the treatment with grasses. Three-inch cores were obtained from each treatment using a soil core sampler. These cores contained layers of soil amendment, sludge, and topsoil. The cores were stored at 4°C in a walk-in freezer until further analysis. At the time of analysis, cores were removed from the freezer, the individual layers were carefully separated from each core, and 10 g of sludge was collected from the sludge layer. The sludge was mixed with 10 g of anhydrous MgSO4 and placed in a soxhlet. The mixture was then extracted using 300 mL of 1:1 methylene chloride and acetone for 48 hours. The extract was then analyzed using gas chromatography/mass spectrometry (GC/MS). The mass under the chromatogram was quantitated using 100 ppm diesel fuel as an internal standard. The GC/MS was also calibrated using the 100-ppm PAH standard (EPA priority list). The extract from each core obtained after soxhlet extraction was also quantitated based upon our calibration and the amount of PAH in each treatment was estimated.

Results and Discussion. A two-factor ANOVA with replication was performed on chlorophyll measurements taken on December 2, 2004, across the six grass species and across the agronomic practices used in this study as treatments. There was no interaction between grasses and treatments, both at 95% and 99% confidence intervals. There were significant differences in mean chlorophyll measurements, both under frequent and infrequent irrigation regimes, across the six grass species used in this study. The result was the same, both at 99% and 95% confidence intervals. There were no significant differences among the treatments under frequent regime and infrequent regime at a 99% confidence interval. The Indian Grass did not show any improvement in terms of plant height as well as plant counts over the time. There were significant differences at a 95% confidence interval among treatments under the infrequent regime when Indian Grass was excluded from the list of the grass species.

Figure 8 compares chlorophyll measurements among all six grasses under the frequent irrigation regime in the presence of three soil amendments and topsoil as control. Indian Wood Oats had the highest mean chlorophyll under these conditions, followed by Prairie Brome, Canada Wild Rye, Side-Oats Grama, and Switch grass in the presence of N-Viro Soil applied as a soil amendment. Indian Grass did not perform well in the presence of N-Viro Soil. The trends were similar across the grasses in the presence of NPK fertilizer; however, Indian Grass performed better than Canada Wild Rye. When grasses were subjected to a combination of N-Viro Soil and NPK fertilizer, Indian Wood Oats, Canada Wild Rye, and Side-Oats Grama increased their chlorophyll contents, whereas Switch Grass, Indian Grass, and Prairie Brome decreased their chlorophyll contents. Except Indian Grass, all grasses showed lower chlorophyll content in the presence of a topsoil-only control.

Figure 8. Comparison of Chlorophyll Measurements Across Six Ohio Grasses That Had Undergone Frequent Irrigation and Were Grown on the PAH-Contaminated Sludge in the Presence of Three Soil Amendments and Top Soil as Control

Figure 8. Comparison of Chlorophyll Measurements Across Six Ohio Grasses That Had Undergone Frequent Irrigation and Were Grown on the PAH-Contaminated Sludge in the Presence of Three Soil Amendments and Top Soil as Control

Figure 9 compares chlorophyll measurements among all five grasses under an infrequent irrigation regime in the presence of three soil amendments and topsoil as a control. Indian Grass was eliminated from the comparison since it showed the greatest variability across the treatments (soil amendments). Side-Oats Grama showed the lowest chlorophyll among the five grasses in the presence of any soil amendment or control. Canada Wild Rye had the highest chlorophyll content in the presence of N-viro Soil, followed closely by the Indian Wood Oats, and then by Prairie Brome, Switch Grass, and Side-Oats Grama, in that order. Prairie Brome had the highest chlorophyll content in the presence of NPK fertilizer, followed by Side-Oats Grama, Indian Wood Oats, Canada Wild Rye, and Switch Grass, in that order. Grasses performed better in the presence of N-Viro Soil than NPK except in the case of Prairie Brome. The combination of N-Viro Soil and NPK fertilizer favored Canada Wild Rye, Indian Wood Oats, and Switch Grass over using NPK fertilizer alone. All grasses showed similar chlorophyll contents in the presence of topsoil as the combination of N-Viro Soil and NPK fertilizer.

Figure 9. Comparison of Chlorophyll Measurements Across Five Ohio Grasses That Had Undergone Infrequent Irrigation and Were Grown on the PAH-Contaminated Sludge in the Presence of Three Soil Amendments and Topsoil as Control

Figure 9. Comparison of Chlorophyll Measurements Across Five Ohio Grasses That Had Undergone Infrequent Irrigation and Were Grown on the PAH-Contaminated Sludge in the Presence of Three Soil Amendments and Topsoil as Control

A two-factor ANOVA with replication was performed on grass height measurements taken on December 3, 2004, across the six grass species and across the agronomic practices used in this study as treatments. There was no interaction between grasses and treatments at 99% confidence intervals. There were significant differences in mean grass height measurements both under frequent (every 4th day) and infrequent (every 6th day) regimes across the six grass species used in this study at a 99% confidence interval. There were significant differences among the treatments under the frequent irrigation regime at a 99% confidence interval, and not under the infrequent irrigation regime. However, there were significant differences at 95% confidence intervals among treatments under the infrequent regime when Indian Grass was excluded from the list of the grass species.

Figure 10 compares grass height across the six grass species used in this study and across the three soil amendments and topsoil control treatments under the frequent irrigation regime. Switch Grass grew tallest, followed by Side-Oats Grama and Canada Wild Rye. Prairie Brome did not show significant growth. Except for Indian Grass, all grasses grew taller in the presence of N-Viro Soil than in the presence of NPK fertilizer. There were no significant differences in grass heights either in the presence of N-Viro Soil or a combination of N-Viro Soil and NPK fertilizer, though the contrary was true in cases of Switch Grass and Indian Wood Oats. Grasses performed better in the presence of each treatment than in the topsoil control.

Figure 10. Comparison of Grass Heights Across Six Ohio Grasses That Had Undergone Frequent Irrigation and Grown on the PAH-Contaminated Sludge in the Presence of Three Soil Amendments and Topsoil as Control

Figure 10. Comparison of Grass Heights Across Six Ohio Grasses That Had Undergone Frequent Irrigation and Grown on the PAH-Contaminated Sludge in the Presence of Three Soil Amendments and Topsoil as Control

Figure 11 shows a comparison across five grass species and across the treatment under the infrequent irrigation regime. Switch Grass and Side-Oats Grama attained the largest grass heights, with N-Viro Soil being the most favored soil amendment. Both Prairie Brome and Indian Wood Oats did not perform well under any treatment. In fact, these grasses performed better in the presence of the combination of N-Viro Soil and NPK fertilizer than any individual treatment. Generally, grasses performed better when grown in the presence of soil amendments than in the topsoil control.

Figure 11. Comparison of Grass Heights Across Six Ohio Grasses That Had Undergone Infrequent Irrigation and Grown on the PAH Contaminated Sludge in the Presence of Three Soil Amendments and Topsoil as Control

Figure 11. Comparison of Grass Heights Across Six Ohio Grasses That Had Undergone Infrequent Irrigation and Grown on the PAH Contaminated Sludge in the Presence of Three Soil Amendments and Topsoil as Control

Individual grass blades were counted in each treatment and for every type of grass. These numbers were monitored over the time as plant counts. One of the main challenges in comparing grasses based on their plant counts after a period of time was that initially these were not the same across all treatments. In order to determine how significant were changes in individual treatments after a period of time compared to changes across the treatments initially, changes in plant counts in each treatment over a period of time were computed. These changes were significant when compared to initial changes in plant counts across all treatments. Thus, plant counts taken after a period of time (in this case, after 5 months) from the start of planting can still be used to compare the grasses subjected to the different treatments. A two-factor ANOVA with replication showed significant differences across grasses and treatments only at a 90% confidence interval. Figure 12 shows plant counts for every grass subjected to one of the four treatments under the frequent irrigation regime, measured on December 4, 2004. Side-Oats Grama showed the greatest plant counts compared to those of any other grass. Canada Wild Rye and Prairie Brome followed next in terms of large plant counts. Indian Wood Oats and Switch Grass had smaller plant counts among all grasses with very few plant counts for Indian Grass. Side-Oats Grama showed the greatest plant counts in N-Viro Soil, followed by NPK fertilizer, combination, and topsoil, in that order. Figure 13 shows a similar comparison under the infrequent irrigation regime. Side-Oats Grama showed the largest plant counts under all treatments when compared to those of any other grass. This grass grew best under N-Viro Soil conditions. Other grasses did not show significant growth under any treatment. Among other grasses, Canada Wild Rye showed significant growth in terms of plant counts.

Figure 12. Comparison of Plant Counts (Number of Individual Grass Blades) Across Six Ohio Grasses That Had Undergone Frequent Irrigation and Grown on the PAH-Contaminated Sludge in the Presence of Three Soil Amendments and Topsoil as Control

Figure 12. Comparison of Plant Counts (Number of Individual Grass Blades) Across Six Ohio Grasses That Had Undergone Frequent Irrigation and Grown on the PAH-Contaminated Sludge in the Presence of Three Soil Amendments and Topsoil as Control

Figure 13. Comparison of Plant Counts (Number of Individual Grass Blades) Across Six Ohio Grasses That Had Undergone Infrequent Irrigation and Grown on the PAH-Contaminated Sludge in the Presence of Three Soil Amendments and Topsoil as Control

Figure 13. Comparison of Plant Counts (Number of Individual Grass Blades) Across Six Ohio Grasses That Had Undergone Infrequent Irrigation and Grown on the PAH-Contaminated Sludge in the Presence of Three Soil Amendments and Topsoil as Control

Figure 14 shows the penetration of roots through one of the soil cores across the layers. Penetration of roots through the sludge layer was found across all treatments. PAHs appear to have no effect on root growth. The high octanol-water partition coefficients for these compounds would make them essentially unavailable in soil pore water. This could be the reason for lack of toxicity due to PAHs on root biomass. Figure 15 shows the contact surface area of roots in each grass per gram of soil cored from the treatments under frequent irrigation. Switch grass showed significant root growth followed by Canada Wild Rye and Side-Oats Grama. Indian Wood Oats did not support any growth while Indian Grass supported the lowest growth among other grasses with the greatest variability across treatments. Except for Switch Grass, all grasses showed the greatest root biomass in the presence of the combination of N-Viro Soil and NPK fertilizer. All grasses undergoing a combination of treatments performed better than the topsoil control. Individual amendments performed similar to the topsoil control. Figure 16 is a corresponding plot of root surface area for each grass undergoing infrequent irrigation. Switch Grass in the presence of a combination of N-Viro Soil and NPK showed the greatest root growth. Both Side-Oats Grama and Switch Grass showed greater root surface area in the presence of amendments than the topsoil control. Canada Wild Rye supported lower root surface in the presence of treatments than the topsoil control. Prairie Brome did not show significant differences across treatments. Indian Wood Oats supported the lowest root surface area with significant variability in treatments.

Figure 14. Figure Shows Root Penetration Through Soil Amendment and Sludge Layers Observed Across All Treatments
 
Figure 14. Figure Shows Root Penetration Through Soil Amendment and Sludge Layers Observed Across All Treatments

Figure 14. Figure Shows Root Penetration Through Soil Amendment and Sludge Layers Observed Across All Treatments

Figure 15. Comparison of Root Surface Area Across Six Ohio Grasses Undergoing Frequent Irrigation and Grown on PAH-Contaminated Sludge in the Presence of Soil Amendments and Topsoil as Control

Figure 15. Comparison of Root Surface Area Across Six Ohio Grasses Undergoing Frequent Irrigation and Grown on PAH-Contaminated Sludge in the Presence of Soil Amendments and Topsoil as Control

Figure 16. Comparison of Root Surface Area Across Six Ohio Grasses Undergoing Infrequent Irrigation and Grown on PAH-Contaminated Sludge in the Presence of Soil Amendments and Topsoil as Control

Figure 16. Comparison of Root Surface Area Across Six Ohio Grasses Undergoing Infrequent Irrigation and Grown on PAH-Contaminated Sludge in the Presence of Soil Amendments and Topsoil as Control

A chromatogram of the original sludge before treatment using grasses and soil amendments is shown in Figure 17 below. Soxhlet extraction followed by chemical analysis using GC/MS showed that the sludge contained essentially a host of petroleum hydrocarbons typically found in diesel fuel in addition to PAH quantified by the GC-MS after calibration with the 100-ppm standard. Figure 18 shows a chromatogram of one of the treatments in the factorial design involving the six grasses, four soil amendments, and two irrigation regimes discussed in the 1st year of the annual report. This figure corresponds to the treatment of the sludge using Canada Wild Rye in the presence of a combination of N-Viro Soil and NPK fertilizer under the infrequent irrigation regime. Such chromatograms were generated for each treatment. A total of 142 chromatograms were generated, though we had 180 total treatments under the factorial design. This is due to the fact that some of the treatments did not show any growth of grasses and because some cores did not catch the sludge as the soil was cored through the layers of soil amendment, sludge, and topsoil in the layered design as discussed in the 1st year of the report. The total petroleum hydrocarbons (TPH) from the sludge in each treatment were estimated using 100 ppm diesel fuel as the internal standard. As shown from the chromatograms of the sludge before and after treatment in Figures 17 and 18, TPH concentration was lowered 10-fold (Note: graphs were plotted on different scales). Composition of contaminant seems to have shifted from high to low molecular weight compounds.

Figure 17. The Figure Shows a Chromatogram of the Original Sludge Before Treatment Using Grasses and Soil Amendments

Figure 17. The Figure Shows a Chromatogram of the Original Sludge Before Treatment Using Grasses and Soil Amendments

Figure 18. Figure Shows Chromatogram of the Sludge After Treatment With Canada Wild Rye, N-Viro Soil Under Infrequent Irrigation.

Figure 18. Figure Shows Chromatogram of the Sludge After Treatment With Canada Wild Rye, N-Viro Soil Under Infrequent Irrigation. (Note: The scale on y-axis is one order of magnitude lower than that of Figure 17.)

With frequent irrigation (Figure 19), there was a 10-fold reduction in TPH concentrations across all treatments, including the no-grass control. Canada Wild Rye, Side-Oats Grama, and Indian Grass reduced TPH concentrations at least 90% in the presence of N-Viro+NPK. TPH reduction is greater in the presence of soil amendments than without amendments (topsoil).

Figure 19. Figure Shows TPH Concentration Across Treatments, Each Using One of Six Native Ohio Grasses and One of the Soil Amendments Such as Compost (N-Viro Soil), Inorganic Fertilizer (NPK), a Combination of N-Viro and NPK, and Only Topsoil as a Control Under Frequent Irrigation Regime

Figure 19. Figure Shows TPH Concentration Across Treatments, Each Using One of Six Native Ohio Grasses and One of the Soil Amendments Such as Compost (N-Viro Soil), Inorganic Fertilizer (NPK), a Combination of N-Viro and NPK, and Only Topsoil as a Control Under Frequent Irrigation Regime

With infrequent irrigation (Figure 20), there was again a 10-fold reduction in TPH concentrations across all treatments including the no-grass control. Canada Wild Rye and Side-Oats Grama with N-Viro Soil and N-Viro+NPK showed up to a 95% TPH reduction compared to control. A two-factor ANOVA with replication was performed on TPH measurements across the six grass species and across the agronomic practices used in this study as treatments. There were no significant differences in mean TPH concentrations both under frequent (every 4th day) and infrequent (every 6th day) regimes across the six grass species used in this study.

Figure 20. Figure Shows TPH Concentration Across Treatments, Each Using One of Six Native Ohio Grasses and One of the Soil Amendments Such as Compost (N-Viro Soil), Inorganic Fertilizer (NPK), a Combination of N-Viro and NPK, and Only Topsoil as a Control Under Infrequent Irrigation Regime

Figure 20. Figure Shows TPH Concentration Across Treatments, Each Using One of Six Native Ohio Grasses and One of the Soil Amendments Such as Compost (N-Viro Soil), Inorganic Fertilizer (NPK), a Combination of N-Viro and NPK, and Only Topsoil as a Control Under Infrequent Irrigation Regime

We found growth of bacteria in the contaminated soil, which can degrade diesel fuel, and phenanthrene, which is indicative of the presence of petroleum hydrocarbon degraders in the contaminated sludge. The second phase of the project was designed to see whether a combination of these grass species would work in an ecological setting; however, since the individual species by themselves have shown significant degradation, we have deliberated on the necessity for the second phase. The plan was to reuse the sludge from the first phase after characterization so that we would not need the fresh contaminated sludge; however, the original sludge was significantly cleaned up, as our results have indicated. Since significant reduction of TPH was observed in our treatments with soil amendments, we conducted a study to investigate the role of bacteria as a logical next step in evaluating microbial degradation in the presence of grasses.

Treatments carrying Canada Wild Rye and Side-Oats Grama were analyzed for their support of microbial growth using the classical approach of plate counts and also using recent molecular approaches using DNA. Our students, under the guidance of Dr. J.R. Shann, carried out these studies in the Biology department at the University of Cincinnati. Enrichment cultures were prepared by taking soil from the rhizosphere of these two grasses, minimal salts media, and an organic contaminant of choice to isolate organisms in the soil that can use these complex carbon sources as food. Once the bacterial strains were enriched, their ability to degrade different compounds (diesel fuel, phenanthrene, and 2-4-D) was tested. The isolated bacteria degraded these common organic contaminants, thus confirming the presence of hydrocarbon degraders in the contaminated soil. The same enrichment cultures were used to determine molecular diversity by using polymerase chain reaction (PCR) and denaturing gradient gel electrophoresis (DGGE). This procedure was based on the amplification of the 16-S rRNA sequence of a region known as V3. Bacterial rDNA genes at this location are different generally down to the genus and species level, making it a convenient way to look at diversity. Treatments supported significant microbial diversity of typical petroleum hydrocarbon degraders.

Summary and Conclusions. This study uniquely identified native grass species of Ohio that could survive and grow in a highly PAH-contaminated sludge. All grasses responded to the soil amendments and irrigation. The individual treatments were made in such a way so as to allow for adequate aerobic conditions within the sludge and soil, sufficient nourishment for roots through nutrients, optimal water content, and free gravity drainage. Parameters such as chlorophyll content, plant height, and plant count were used to evaluate the growth of these species over the time. The study investigated optimal agronomic practices, such as combinations of soil amendments and irrigation that would enable these native grass species to survive and grow on the PAH-contaminated sludge. Side-Oats Grama and Canada Wild Rye showed the best growth performance among the six grass species studied. These grasses performed better when treated with N-Viro Soil or a combination of N-Viro Soil and NPK fertilizer, and under the infrequent irrigation regime. Switch Grass grew tallest under the frequent irrigation regime; however, it had lower chlorophyll content and plant counts when compared to Side-Oats Grama or Canada Wild Rye. Prairie Brome had high chlorophyll content under the infrequent irrigation regime and when treated with N-Viro Soil, however, it had a low plant count and low grass height. Indian Grass and Indian Wood Oats can be eliminated from further consideration since these species had the lowest plant counts, grass heights, and chlorophyll contents. Indian Grass also showed the greatest variability across all treatments. Our study also showed that the native Ohio grasses significantly reduced TPH contamination in the presence of a combination of soil compost and fertilizer under infrequent irrigation. Significant differences were found amongst grasses in terms of grass height, chlorophyll, and remediation potential. Side-Oats Grama, Canada Wild Rye, and Switch grass appear to grow best, as well as reduce TPH contamination. N-Viro Soil and N-Viro+NPK supported greater grass growth and reduction in TPH than topsoil and NPK. Reduction in TPH was evidenced using chromatograms and also was supported by significant root penetration across all layers, including the sludge layer. Presence of TPH did not affect the growth of grasses, as measured by both above ground and below ground biomasses. Roots may have provided the necessary root exudates and substrates for microbial growth that reduced TPH in sludge over a period of 5 months using an optimal combination of irrigation and soil amendments. Microbial degradation was confirmed by plate counts and the presence of typical hydrocarbon degraders. Microbial diversity of these species was also observed through molecular techniques.

Acknowledgments

The investigators would like to sincerely thank Professor Subramania I. Sritharan, Chairperson of the Department of Water Resources Management for his encouragement, advice, and administrative support. Guidance and collaboration from Professor Jodi Shann and her co-workers in the department of Biology at University of Cincinnati, Ohio, were also very much appreciated.

References:

Amellall N, Portal JM, Vogel T, Berthelin J. Distribution and location of polycyclic aromatic hydrocarbons (PAHs) and PAH-degrading bacteria within polluted soil aggregates. Biodegradation 2001;12(1):49-57.

Arya LM, Leij FJ, Th van Genuchten M, Shouse PJ. Scaling parameter to predict the soil water characteristic from particle-size distribution data. Soil Science Society of America Journal 1999;63(3):510-519.

Binet Ph, Portal JM, Leyval C. Fate of polycyclic aromatic hydrocarbons (PAH) in the rhizosphere and mycorrhizosphere of ryegrass. Plant and Soil 2000;227:207-213.

Cahn MD, Hummel JW, Brouer BH. Spatial analysis of soil fertility for site-specific crop management. Soil Science Society of America Journal 1994;58:1240-1248.

Chan TP, Govindaraju RS. Estimating soil water retention curve from particle-size distribution data based on polydisperse sphere systems. Vadose Zone Journal 2004;3:1443-1454.

Dur JC, Elsass F, Chaplain V, Tessier D. The relationship between particle-size distribution by laser granulometry and image analysis by transmission electron microscopy in a soil clay fraction. European Journal of Soil Science 2004;55(2):265-271.

Cunningham SD, Berti WR. Remediation of contaminated soils with green plants: An overview. In Vitro Cellular and Developmental Biology-Plant 1993;29P:207-212.

Ebbs SD, Lasat MM, Brady DJ, Cornish J, Gordon R, Kochian LV. Phytoextraction of cadmium and zinc in contaminated soil. Journal of Environmental Quality 1997;26(5):1424-1430.

Gardner WH. Water content. In: Klute A, ed. Methods of Soil Analysis, Part 1. Madison, WI: American Society of Agronomy, 1986, pp. 493-544.

Gee GW, Bauder JW. Particle-size analysis. In: Klute, A., ed. Methods of soil analysis, Part 1. Madison, WI: American Society of Agronomy, 1986, pp. 383-412.

Guha S, Peters CA, Jaffe PR. Multisubstrate biodegradation kinetics of naphthalene, phenanthrene, and pyrene mixtures. 1999;65(5):491-499.

Henry HF, Shann JR. Stability of contaminants on an aged petroleum refinery land treatment unit. Soil and Sediment Contamination 2004;13(2):187.

Hutchinson SL, Schwab AP, Banks MK. Phytoremediation of aged petroleum sludge: effect of irrigation techniques and scheduling. Journal of Environmental Quality 2001;30(5):1516-1522.

Jordahl JL, Foster L, Schnoor JL, Pedro A. Effect of hybrid poplar trees on microbial populations important to hazardous waste bioremediation. Environmental Toxicology and Chemistry 1997;16:1318-1321.

Joshi MM, Sungyuu L. Effect of oxygen amendments and soil pH on bioremediation of industrially contaminated soils. Energy Sources 1996;18(3):233-242.

Klute A. Methods of soil analysis. Part 1 physical and mineralogical methods, Agronomy Series No. 9. American Society of Agronomy 1986;2nd edition, pp. 493-544.

Krishnan G, Horst GL, Darnell S, Powers WL. Growth and development of smooth bromegrass and tall fescue in TNT-contaminated soil. Environmental Pollution 2000;107:109-116.

Mannetze LT, Jones RM. Measuring biomass of grassland vegetation. In: Field and laboratorymethods for grassland and animal productionresearch. CAB International 2000;151-177.

Nedunuri KV. Modeling of heavy metal movement in rhizosphere soils. PhD Thesis. Purdue University, 1999.

Nedunuri KV, Govindaraju RS, Banks MK, Schwab AP, Chen Z. Evaluation of phytoremediation for field-scale degradation of total petroleum hydrocarbons. Journal of Environmental Engineering 2000;126(6):483-490.

Olson PE, Fletcher JS. Ecological recovery of vegetation at a former industrial sludge basin and its implications to phytoremediation. Environmental Science and Pollution Research 2000;7(4):195-204.

Posadas AND, Giménez D, Bittelli M, Vaz CMP, Flury M. Multifractal characterization of soil particle-size distributions. Soil Science Society of America Journal 2001;65:1361-1367.

Reilley KA, Banks MK, Schwab AP. Dissipation of polycyclic aromatic hydrocarbons in the rhizosphere. Journal of Environmental Quality. 1996;25(2):212-219.

Skaggs TH, Arya LM, Shouse PJ, Mohanty BP. Estimating particle-size distribution from limited soil texture data. Soil Science Society of America Journal 2001;65:1038-1044.

U.S. EPA. 1987. Gas chromatography/mass spectrometry for semivolatile organics: capillary column technique. Method 8270.

U.S. EPA. 1992. Framework for risk assessment. Washington DC: EPA/630/R-92/001.

U.S. EPA. 1997. Innovative uses of compost. Bioremediation and Pollution Prevention. Solid Waste and Emergency Response: EPA 530-F-97-042.

Villamouz J, Wilke MW. Effect of compost on phytoremediation of diesel contaminated soil. Water Science and Technology 2001;43(2):291-295.

Yang Z-Y, Juo J-L. Interpretation of sieve analysis data using the box-counting method for gravelly cobbles. Canadian Geotechnical Journal 2001;38(6):1201-1212.


Journal Articles on this Report : 1 Displayed | Download in RIS Format

Other project views: All 16 publications 1 publications in selected types All 1 journal articles
Type Citation Project Document Sources
Journal Article Nedunuri KV, Lowell C, Meade W, Vonderheide AP, Shann JR. Management practices and phytoremediation by native grasses. International Journal of Phytoremediation 2010;12(2):200-214. R831072 (Final)
  • Abstract from PubMed
  • Full-text: Miami University-PDF
    Exit
  • Abstract: Taylor & Francis-Abstract
    Exit
  • Supplemental Keywords:

    RFA, Scientific Discipline, INTERNATIONAL COOPERATION, Waste, Contaminated Sediments, Environmental Microbiology, Hazardous Waste, Bioremediation, Hazardous, microbiology, plant species, industrial waste, microbial degradation, bioavailability, biodegradation, contaminated sediment, PAH, contaminated soil, petrochemical waste, contaminants in soil, bioremediation of soils, biochemistry, refinery waste, phytoremediation

    Progress and Final Reports:

    Original Abstract
  • 2004 Progress Report
  • 2005 Progress Report