Use of the exotic plant oenanthe javanica in plant/rock filters for on-site waste disposal.

Introduction

Given the current emphasis on the environment, scientists are seeking to purify contaminated waters in ways that do not further contaminate air, water, or earth. The trend is toward recycling all natural resources, including water, by using as few harmful substances as possible and by using natural systems. Plants can provide an energy-efficient natural alternative to the chemical and mechanical wastewater treatment methods now in use (1). Domestic sewage can be treated by aquatic or emergent plants on a large scale; many smaller cities use aquatic plants as part of their wastewater treatment (2). On a smaller scale, individual residential wastewater can be treated on site with plants and natural processes. Almost one third of the homes in the United States are served by some form of on-site wastewater disposal system (3). Many rural residential sites are not suitable for the most common form of on-site wastewater disposal, and other means, including plant/rock filters, must be used (2,4).

Plant rock filters are a type of constructed wetland with a subsurface water flow. They use gravel as a growing medium for plants (2). In this type of wetland, the top few inches of gravel remain dry while fluid moves from one end to the other. Wolverton's studies using rooted vascular aquatic plants showed the positive advantage of plants in a rock filter system (5-7). The popularity of plant/rock filters as an alternative secondary treatment for domestic wastewater is increasing (8). For residential use they generally consist of a plastic-lined, gravel-filled pit connected to a septic tank at one end and to overland or underground discharge at the other end. Small operating size, simplicity of construction, and cost-effectiveness make plant/rock filters especially advantageous for residential use (8). Typical three-bedroom homes require a system size of 63 cubic meters ([m.sup.3]) to meet Mississippi state requirements. This system size also is recommended for meeting U.S. Environmental Protection Agency (U.S. EPA) guidelines for secondary effluent (five-day biochemical oxygen demand and total suspended solids of 30 milligrams per liter).

Currently, only eight different plants are [TABULAR DATA FOR TABLE 1 OMITTED] suggested for use in plant/rock filters, and five of those eight lose their foliage to frost. Accordingly, finding suitable plants for plant/rock filters is difficult and constitutes a topic of current research. In a technology assessment of subsurface-flow constructed wetlands in 1993, U.S. EPA listed the investigation of usable plant types (other than reeds, rushes, and cattails) as a high-priority research need (2). The need for suitable plants can be addressed by finding plants that exhibit growth during winter while planted in plant/rock filters.

Successful functioning of a plant/rock filter requires that it be stocked with plants whose traits are conducive to growth in the system. One characteristic needed in macrophytes planted in a plant/rock filter is an extensive root system capable of growing in gravel. Plant roots in the filter complement its functioning by providing filtration and aerobic microsites for nitrification. The pathway for nitrogen removal in plant/rock filter systems entails that plants with copious roots, which provide a greater area for this process, are preferable. When grown hydroponically, emergent plants form extensive root masses with excellent wastewater treatment potential (1).

Oenanthe javanica, commonly called "water dropwart," is an emergent plant with slender roots (9). The ratio of surface area to volume is high (10). These characteristics make it an ideal subject for investigation. The nitrogen pathway determines the advantage that plants that grow year round have over plants that are dormant in the winter - they maintain transpiration and denitrification. More oxygen is transposed to roots, providing aerobic microsites for bacteria, when plants remain in an actively growing state. Many factors, some of which are the amount of mature foliage and the amount of sunlight, can also affect oxygen transport capacity (11,12). Because Oenanthe javanica grows throughout the winter, it maintains the nitrogen pathway, transpiration, and bacterial activity year round.

Other characteristics needed for plants grown in plant/rock filters are rapid year-round growth rate, ability to survive freezing temperatures, and ability to grow in secondary effluent (13). In a plant/rock filter, plants and rocks contribute to the synergistic functioning of the system; the rock surfaces and plant roots provide excellent mediums for the filtration and adsorption of solids, as well as for the growth of bacteria (1). Plants that form dense masses of roots in response to wastewater as a nutrient source are preferable to plants that do not. Many plant species exhibit greatly expanded root systems when forced to grow hydroponically (14). Prolific plants are the most useful for pollution abatement, but the most prolific plants are strictly aquatic or are made dormant by winter temperatures (15). Since O. javanica had been previously observed by the author as an evergreen with an extensive root system, this study was designed to test this plant for suitability in plant/rock filters during winter.

Methods

To validate the emergent plant, Oenanthe javanica, for cold weather use in plant/rock filters, a field study was conducted from December 1993 through March 1994 in Adams County, Mississippi [ILLUSTRATION FOR FIGURE 1 OMITTED]. The study site climate is considered mild, although temperatures do occasionally drop below freezing. The chance of frost after December 2 and before April 7 is 90 percent. The previous five years, ending in 1993, had an average of 29 freezing days each winter (16).

The plant/rock filter was situated on a 5-acre single-family home site. The family consisted of two adults and three children, three, 10, and 12 years of age. Mississippi recommendations for system size are based on U.S. EPA guidelines and Tennessee Valley Authority (TVA) information with respect to flow rate, hydraulic retention time, and effluent characteristics. Effluent was routed from an existing septic tank to the filter. Because the septic tank does provide minimal treatment in the form of settling, it is considered a method of treatment; the term secondary does not refer to quality as defined by U.S. EPA, but to the number of different types of treatment.

Construction of the plant/rock filter began in October 1993 according to Mississippi State Department of Health Regulations (17). The construction criteria in these regulations are based on the current research and recommendations of U.S. EPA and TVA (2,4). A pit 30 by 90 m (10 by 30 feet), with vertical earthen walls and a maximum depth of 46 cm (18 inches) was dug. A construction grade transit ensured that the bottom of the pit was level. The pit was lined with two layers of 6-millimeter (mm) black polyethylene. Inlet and outlet structures were constructed with sewer grade 4-inch polyvinyl chloride (PVC) pipe. A center "T" of perforated pipe, held off the bottom of the pit with two common bricks at each end, formed the inlet. The outlet structure was a 3 m (10-foot) section of perforated sewer pipe held 30 cm (1 foot) above the bottom of the pit by steel stands. It was located in the center and parallel to the long sides of the pit as shown in Figure 2. A carpenter's level ensured that the pipe was level. The sides and ends of the filter were lined with two tiers of landscape timbers with a berm formed around them to keep runoff from entering. Approximately 13 [m.sup.3] (17 cubic yards) of Number 57 aggregate, washed gravel 2.54 to 3.16 cm (0.25 to 1.25 inches), filled the plant/rock filter. Sampling ports of perforated pipe were placed at approximately 3 m and 6 m (10 and 20 feet), and at the effluent end of the plant/rock filter. The 3 m and 6 m ports were located in the center of the filter. The end port was located against the side. An unglued elbow at the effluent end formed a water level adjustment device to be used until the plants became well established.

A member of the family Umbelliferae, Oenanthe javanica was selected for its root system and growth rate, described as extensive and rapid, respectively (18). This plant characteristically grows in low, swampy areas, in the same environment as reeds, rushes, and cattails (19). While the literature refers to it as an emergent plant, it will grow in water for an indefinite time, though not as vigorously as in damp soil. This leafy perennial, whose distribution ranges from Asia to South Africa and tropical Australia, is terete and ramose, with stems from 10 to 100 cm long, erect or ascending from a creeping base (19). Leaves are four-to-five pinnate, but variable, as is the rest of the plant in the dimensions of all its parts, the compoundness of its leaves, the length of the peduncles, the number of pedicels in the umbellules, and the dimensions of the fruits. According to Edwin G. Dyess, Deputy Director of the Bureau of Plant Industry of Mississippi, this plant has naturalized from California to British Columbia. Oenanthe javanica can be obtained from aquatic plant dealers. Because of its reliance on moist growing conditions, this plant would pose little risk to native plant systems outside of constructed wetlands, unless those systems are in unusually moist surroundings.

TABLE 2

Winter Growth of Oenanthe javanica

                               January 1994

Plant #        Weight(*)        Mean Weight       SD       t-Test

26              155.52
78              194.39
99              155.52
174             217.72
200             295.48

Total          1,018.62           203.72        57.77      4.23(**)

                              (p = .01) 3.75

                               February 1994

Plant #        Weight(*)        Mean Weight       SD       t-Test

64              334.36
68              419.89
75               15.55
108             132.19
157             520.98

Total          1,422.96           284.59       207.53      2.05

                              (p = .01) 3.75

March 1994

Plant #        Weight(*)        Mean Weight       SD       t-Test

119             152
152             785.35
230             660.94
231             699.82
80              559.85

Total          3,459.90           691.98        88.09     15.17(**)

                              (p = .01) 3.75

* in grams.

** Highly significant.

Beginning in April 1993, Oenanthe javanica was propagated by the author. All propagation and growing was done under natural conditions. On November 1 and 2, 1993, 270 plantings of Oenanthe javanica were made into the plant/rock filter (see photo on page 20). Rooted cuttings, some consisting of more than one individual plant and grown in typical nursery six-packs, were used. As much soil as possible was washed from the roots, and the plants were allowed to drain before being weighed. Plants were planted immediately after being weighed, and each one was numbered. Effluent from the septic tank was routed to the plant/rock filter two weeks later.

The composition of water hyacinths and other aquatic macrophytes removed from a treatment system can provide an estimate of their nutrient removal potential (20,21). Therefore, to estimate the nutrient removal potential of Oenanthe javanica, the composition of stems and leaves of sample plants removed from the plant/rock filter during the study period was determined. Two analyses were conducted on parts of five randomly selected Oenanthe javanica plants, and one analysis was taken entirely from Plant 157. The average of these three analyses is given in Table 1.

For this field study, two components were monitored: winter growth of Oenanthe javanica and performance of the plant/rock filter. The plant was evaluated by weight gain with Student's t-test, and the plant/rock filter was observed for three important parameters of wastewater: five-day biochemical oxygen demand (BOD5), pH, and total suspended solids (TSS). Random samples of five plants were made on January 31,1994, February 28, 1994, and March 31, 1994, according to the method described by Langley (22). Samples were obtained by loosening with a shovel, if necessary, and gently pulling the plant out of the gravel. Gravel was removed from the roots, and the plants were rinsed and allowed to drain. Samples were placed in plastic bags for transport to the weighing site.

The [BOD.sub.5] was evaluated twice in January 1994 and February 1994, and three times in March 1994, according to the Standard Methods of the American Public Health Association (23). Influent and effluent were taken by grab samples and placed in clean, unused, plastic milk containers. Samples were transported immediately to the testing site. U.S. EPA requirements for secondary effluent of 30 milligrams per liter (mg/L) were met on two of the samples. TSS were evaluated twice in January 1994 and February 1994, and three times in March 1994, according to Standard Methods (23). The same samples taken for [BOD.sub.5] also served for the TSS tests. The pH of the system was measured with an Oakton(R) glass electrode pH meter. During December, measurements were recorded almost daily. After that, measurements were recorded three times a week. The meter was calibrated before each measurement. System fluid temperature readings were recorded simultaneously with pH by means of an Oakton temperature probe. All measurements were taken directly in the sampling ports, except the influent, which was brought up with a plastic sampling bottle. After sampling, the influent was returned via the sampling port. Sampling ports were capped when not in use.

Results and Discussion

Oenanthe javanica was shown to grow in the winter in Adams County, Mississippi, at a level that was highly significant (p = .01) for the first and third sample dates, as measured by weight gain of the plant using Student's t-test [ILLUSTRATION FOR FIGURE 3 OMITTED]. Typical evergreen plant/rock filter plants enter a dormant state in winter and cannot function at optimal rates as they do in warm weather. Advantages of a plant/rock filter system with plants growing through the winter are increased nitrification-denitrification, greater filtration from root growth, and continued loss of liquid by transpiration (24). Vigorous growth of Oenanthe javanica in winter should allow quick establishment of complete plant cover and full-depth root penetration in plant/rock filters, even when the plants must be planted late in the year. Oenanthe javanica endured below freezing temperatures less than one week after being planted in the filter and continued to grow with loss of only a single plant. Final sample plants increased in weight an average of 597.43 grams (g) from an average beginning weight of 94.55g - an average increase of 591 g per plant, or over six times their weight at inception of the study (Table 2). Since the conclusion of the study, the plants have thrived, thinning some in September and October after blooming all summer; with cooler fall weather, they have grown vigorously from the crown.

In addition to a remarkable growth rate, these plants also showed great vitality, with only two plants expressing a lack of vigor. Plant 21 died of what may have been Pythium root rot on February 20, 1994. Plant 75, a sample plant, showed similar symptoms and, after being weighed, was sent to Mississippi State University for a pathology report. Pythium root rot was diagnosed by the plant disease laboratory While this study ended with 153 plant clumps (many had grown together), the number could increase in time as recumbent stems root or seeds germinate. Further decline in number of plants is not anticipated as Oenanthe javanica is a perennial and the plants were well established by the end of the study

Some systems incorporate a layer of top soil or mulch over the gravel in which plants are grown (25). Over time, this layer may work down into the gravel and contribute to clogging and failure of the plant/rock filter. Oenanthe javanica was planted directly into the gravel and grew well, proving that this potentially detrimental layer is unnecessary with this plant. It is important for plants used in wastewater treatment to accumulate large quantities of mineral nutrients, especially nitrogen and phosphorus (26). The use of soil may discourage mineral pickup from the effluent, as plants use soil minerals instead of those in liquid effluent. Nitrogen and phosphorus levels of Oenanthe javanica grown in the plant/rock filter suggest harvesting may be a path for their removal. The nitrogen content of Oenanthe javanica is 45 percent more than that of water hyacinth and five percent more than that of Lemna minor - two plants used in wastewater treatment and harvested for nutrient removal (26,27). The phosphorus level is 23 percent more than that of water hyacinth and 30 percent less than that of Lemna minor.

Results of [BOD.sub.5] tests done for this study are presented in Figure 4. The plant/rock filter performance proved it can meet U.S. EPA secondary effluent requirements for [BOD.sub.5] and TSS within four months, although it was not able to maintain those levels throughout the study period. (This is not unusual for plant/rock filters.) During March the best three tests showed TSS lowered from 474 to 16 mg/L, 763 to 28 mg/L, and 498 to 10 mg/L for reductions of 96.62 percent, 96.33 percent, and 97.99 percent, respectively, proving that plant/rock filters can surpass U.S. EPA minimum requirements for secondary effluent of 30 mg/L [ILLUSTRATION FOR FIGURE 5 OMITTED]. The plant/rock filter TSS effluent mean was 50.14 mg/L, reflecting a mean reduction of 72.85 percent from the influent. On January 6, 1994, testing showed that the [BOD.sub.5] was not 30 mg/L or less. Water temperatures were 9.5 C, which may have been low enough to lower the BOD removal rate, as low temperatures have been shown to affect this rate (1). Plant/rock filters, even under highly controlled conditions, do not achieve steady-state situations (28). This field study did not have controlled influent, and fluctuations may account for the lack of maintenance of the BOD removal rate. The pH of the plant/rock filter system remained within U.S. EPA requirements for secondary effluent (6.0 to 9.0) during this study [ILLUSTRATION FOR FIGURE 6 OMITTED]. The filter adjusted the pH of the influent within the first 10 feet of the system to a maximum of 7.6 from influent pH as high as 8.1. Average pH was 7.6 for the influent and 7.2 for the effluent and ranged from 6.9 for effluent to 8.1 for influent.

This field study showed that plant/rock filter systems can meet U.S. EPA requirements, even before plants have grown a full growing season without influent as recommended by the Tennessee Valley Authority (4). The roots of Oenanthe javanica reached the filter bottom by the end of the study, as shown in the top photo. Roots were approximately 30 cm long, and ends were darkened, apparently as a result of penetrating the sediment at the bottom of the filter. The increase in root mass, resulting in greater filtration, may account for increase in removal of TSS from the filter during the last month of this study. Root penetration to the filter bottom may be a factor in increasing TSS removal, as the removal rate increased over time with root growth. TSS removal was 98 percent for the last test, compared with 6.3 percent for the first test.

The prolific growth of Oenanthe javanica proved that it is well suited to plant/rock filters in areas of the country with weather conditions similar to those in Adams County, Mississippi. This prolific winter growth was not anticipated, and the phenomenal growth of this plant was an unexpected and pleasant surprise (The bottom photo shows Oenanthe javanica in the plant/rock filter on March 31, 1994; compare this picture with the photo on page 20.) As stated earlier, this plant has naturalized to British Columbia and may grow as well in plant rock/filters in those northern regions as it did in this study. Areas for further study involving Oenanthe javanica could include summer growth rate, rate at which oxygen molecules are transported to the roots, nutrient removal potential, transpiration rate, low-temperature tolerance, maximum root length, and comparison tests with other plants commonly used for sewage treatment.

In conclusion, the emergent plant Oenanthe javanica was established as possessing all the characteristics needed to qualify for plant/rock filters and wastewater treatment: prolific growth, extensive root system, resistance to pests and pathogens, and tolerance of freezing temperatures. Its use in plant/rock filters should be encouraged in both new and old systems. Some states use wetlands with open water to purify highway rest stop effluent. The use of plant/rock filters in this setting could alleviate the accompanying insect infestations. Near municipalities, this wetland alternative can be the least expensive method for processing treatment plant effluent (2). When other plants have lost their leaves to frost or have died for other reasons, Oenanthe javanica can be added to older systems to improve their functioning - not only during the cold weather but all year. Its rapid growth makes it ideal for new systems; it provided full coverage in four months, growing in the worst of winter, and should cover even more quickly at more favorable times of the year. Anywhere a high water table, unsuitable soil, lack of space, or other factors restrain the use of more common treatment systems, a plant/rock filter might be considered for wastewater treatment. Even in coastal areas of Mississippi and Louisiana, where plant/rock filters are common, frost severely limits their effectiveness in winter. Oenanthe javanica shows potential to become one of the most highly recommended plants for use in plant rock filters to produce a high-quality, environmentally acceptable effluent.

Editor's note: During the peer review process for this manuscript, the author was asked whether plant/rock filters were suitable for use with cluster systems. The authors response was that since a cluster system serves many homes, a plant/rock filter for those homes would have to be sized accordingly. If properly sized, however, it should provide excellent secondary treatment.

REFERENCES

1. Stowell, R., J. Colt, R. Ludwig, and G. Tchobanoglous (1981), "Concepts in Aquatic Treatment System Design," In Proceedings of the American Society of Civil Engineers, 107(EE5):919-940.

2. "Subsurface Flow Constructed Wetlands for Wastewater Treatment: A Technology Assessment" (1993), Washington, D.C.: U.S. Environmental Protection Agency.

3. Salvato, J.A. (1992), Environmental Engineering and Sanitation, New York: John Wiley and Sons, Inc., pp. 477-512.

4. Steiner, G., and J.T. Watson (1993), General Design, Construction, and Operation Guidelines: Constructed Wetlands Wastewater Treatment Systems for Small Users Including Individual Residences, 2nd ed., Chattanooga, Tenn.: Tennessee Valley Authority.

5. Wolverton, B.C. (1982), "Hybrid Wastewater Treatment System Using Anaerobic Microorganisms and Reed (Phragmites communis)," Economic Botany, 39(4):373-380.

6. Wolverton, B.C., W.R. Duffer, and R.C. McDonald (1983), "Microorganisms and Higher Plants for Waste Water Treatment," Journal of Environmental Quality, 12(2):236-242.

7. Wolverton, B.C., L.K. Marble, and R.C. McDonald (1984), "Removal of Benzene and Its Derivatives from Polluted Water Using the Reed/Microbial Filter Technique," Journal of the Mississippi Academy of Sciences, 29:119-127.

8. Reed, S.C., and D.S. Brown (1992), "Constructed Wetland Design - The First Generation," Water Environment Research, 64(6):776-781.

9. Tanaka, T. (1976), "Oenanthe Javanica," In Cyclopedia of Edible Plants of the World, ed., S. Nakao, Tokyo, Japan: Keigaku Publishing Co., pp. 504.

10. Zoku, S. (1965), "Oenanthe L.," In Flora of Japan, ed., J. Ohwo, Washington, D.C.: Smithsonian Institution, p. 674.

11. Sculthorpe, C.D. (1967), "The Biology of Aquatic Vascular Plants," London: Edward Arnold Ltd., pp. 64-65.

12. Dacey, WK. (1980), "Internal Winds in Water Lilies: An Adaptation for Life in Anaerobic Sediments," Science, 210:1017-1019.

13. Reddy, K.R., and W.F. Debusk (1985), "Growth Characteristics of Aquatic Macrophytes Cultured in Nutrient-Enriched Water: II. Azolla, Duckweed, and Salvinia," Economic Botany, 39(2):200-208.

14. Dinges, R. (1982), In Natural Systems for Water Pollution Control, New York: Van Nostrand Reinhold Co., pp. 118-130.

15. Rogers, H.H., and D.E. Davis (1972), "Nutrient Removal by Water Hyacinth," Weed Science, 20(5):423-425.

16. Mid-South Agricultural Weather Service Center (1989-1993), "Daily Temperature Records - Natchez Station," Stoneville, Miss.

17. "Regulation Governing Onsite Wastewater Disposal Systems" (1992), Jackson, Miss.: Mississippi State Department of Health.

18. Flowering Plants of the World (1978), London: Oxford University Press, pp. 219-221.

19. Buwalda, B. (1949), "Oenanthe," In Flora Malesiana, Kluwer Academic, pp. 135-136.

20. O'Brien, W.J. (1981), "Use of Aquatic Macrophytes for Wastewater Treatment," Journal of Environmental Engineering, 107 (EE4):681-698.

21. Reddy, K.R., and W.F. Debusk (1985), "Nutrient Removal Potential of Selected Aquatic Macrophytes," Journal of Environmental Quality, 14(4):459-462.

22. Langley, R. (1973), Practical Statistics, New York: Drake Publishers.

23. Standard Methods for the Examination of Water and Wastewater (1992), 18th ed., Washington, D.C.: American Public Health Association.

24. Reddy, K.R., C.W. Lindau, and W.H. Patrick, Jr. (1989), "Nitrification-Denitrification at the Plant Root-Sediment Interface in Wetlands," Limnology and Oceanography, 34(6):1004-1013.

25. Steiner, G.R., and D.W. Combs (1991), "Small Constructed Wetlands Systems for Domestic Wastewater Treatment and Their Performance," Presented at "Constructed Wetlands for Water Quality Improvement: An International Symposium," Pensacola Fla., Oct. 21-24.

26. Boyd, C.E. (1969), "Vascular Aquatic Plants for Mineral Nutrient Removal from Polluted Waters," Economic Botany, 23:95-103.

27. Weber, S.A., and G. Tchobanoglous (1985), "Nitrification in Water Hyacinth Treatment Systems," Journal of Environmental Engineering, 111 (5):699-713.

28. Skipper, D. (1990), "A Rock-Plant Filter Bench-Scale Study and Computer Model," Ph.D. thesis, Louisiana State University Agricultural and Mechanical College, p. 34.

29. Harvey, R.M., and J.L. Fox (1973), "Nutrient Removal Using Lemna minor," Journal of the Water Pollution Control Federation, 45(9):1928-1938.

Corresponding Author: Florence Hargis, Environmental Health Program Specialist, Bureau of Environmental Health, Mississippi State Department of Health, 2423 North State Street, Jackson, MS 39215-1700.