| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Biosystems Engineering, Jordan Univ. of Science and Technology, P.O. Box 3030, Irbid, Jordan
b School of Engineering, Univ. of Guelph, ON, Canada N1G 2W1
c J.F. Sabourin and Associates, Inc., Ottawa, ON, Canada
d Dep. of Environmental Biology, Univ. of Guelph, Guelph, ON, Canada N1G 2W1
* Corresponding author (majed{at}just.edu.jo)
Received for publication February 3, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: PTE, phosphorus-trapping efficiency • VFS, vegetated filter strip
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Vegetated filter strips can be defined as bands of cropland adjacent to streams or drainage ditches that are set aside from crop production to be planted with permanent vegetation. When cropland runoff flows across the VFS, it undergoes a decrease in pollutant concentration and volume. These changes reduce the loading of pollutants in the receiving watercourse. In Canada, the use of VFS has been included in the Ontario Environmental Farm Plan and the Best Management Practices for Water Management and for Soil Management. In the United States, since 1988, filter strips are an approved USDA cost-share practice under the Conservation Reserve Program of the Food Security Act of 1985.
Phosphorus (P) exists in many forms in soil, water, and sediments. In runoff, P is generally divided into particulate and dissolved fractions by filtration through a 0.45-µm filter. Particulate forms (i.e., sediment-bound P) include sorbed P, organic P, and mineral P phases. Dissolved forms are normally considered to be orthophosphate, inorganic polyphosphates, and organic P compounds (McDowell and Sharpley, 2001a; Nelson and Logan, 1983). These P compounds exist in dynamic equilibrium between their dissolved and particulate forms. The desorption of soil P for individual runoff events has been related to the P content of surface soil, duration and volume of runoff event, and sediment load (McDowell and Sharpley, 2001b; Sharpley, 1985). Once in surface runoff, phosphorus can deposit along with sediments, adsorb to suspended solids, adsorp to surface soil and vegetation, be assimilated by microorganisms and plants, infiltrate down into soil profile, or move downslope with the runoff (Lee et al., 1989).
While sediment-removal studies are abundant, research studies that have dealt with P removal in VFS are very limited and the sparse results are somewhat contradictory. In a VFS field experiment, Dillaha et al. (1987) found that total P removal was closely related to sediment removal when runoff had high particulate P concentration. They found that P removal efficiency in 4.6-m-long filters varied from 49 to 73%, while corresponding sediment removal was slightly higher at 53 to 86%. Longer filters of 9.1 m were more efficient, with P removal ranging from 65 to 93% and sediment removal ranging from 70 to 98%. In this study more than 90% of the total phosphorus content was sediment bound. Another study (Magette et al., 1989) reported that VFS were less efficient in P removal compared with that of sediment removal. They found that the average total P removal for the 4.6- and 9.1-m-long filters was only 27 and 46%, respectively. The corresponding sediment removal efficiencies for the same study were 66 and 82%, respectively.
In a two-year VFS study under natural rainfall conditions, Daniels and Gilliam (1993) found that 6-m-long filters retained, on average, 60% of the total P load, and retained about 50% of the soluble P load. A similar study was conducted by Patty et al. (1997), who investigated the removal of soluble P load in VFS with 12 filters with lengths of 6, 12, and 18 m under natural rainfall conditions. They found that the average soluble P removal was 40, 52, and 87% for lengths of 6, 12, and 18 m, respectively. Corresponding average sediment load removal was 92, 98, and 99%, respectively. Infiltration was suggested as the main removal mechanism for P removal, especially in the case of longer filter strips.
Many other studies have suggested that infiltration is the primary mechanism of P removal, especially for runoff with high soluble P content such as runoff from land area receiving manure applications (Overcash et al., 1981; Chaubey et al., 1994; Srivastava et al., 1996). In Ontario, VFS for treatment of beef feedlot and dairy yard runoff are designed to allow for total infiltration of the design storm, as infiltration is more easily quantified than other VFS treatment effects (Toombs, 1997).
The first objective of the study reported here was to examine the efficiency of P removal under Ontario, Canada conditions for VFS that varied in length, slope, type of vegetation, and density of vegetative cover. A second objective was to identify P removal mechanisms in vegetated strips and record their relative importance.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
A fully replicated range of strip lengths was tested for red fescue mix (Type B), but only 5-m strips of the control (Type C), native (Type D), and perennial rye grass (Type A) treatments were studied. The perennial rye grass strip was unreplicated and the strips of native vegetation were twice as steep as the other strips. The A and B types were recently established and had lower density of coverage than the well-established Type D. The vegetation density, reported as percentage of vegetation cover, was measured for each filter by visual estimation of vegetation cover and by counting and measuring the diameter of grass punches within a 25-cm square frame. Table 1 summarizes the physical characteristics of each filter tested in this study.
|
The experimental pollutant load was selected to represent an expected edge-of-field water quality. A difficulty arose in defining typical cropland runoff because this type of runoff has highly variable characteristics. The literature shows that there is a wide range of sediment and phosphorus concentrations found in cropland runoff. Total suspended solids of farm runoff ranged from 90 mg L-1 (Shaeffer, 1982) to as high as 7000 mg L-1 (Hayes and Hairston, 1983). Sharpley and Smith (1989) found an average phosphorus concentration of 0.24 mg L-1 with values ranging from 0.03 to 2.67 mg L-1. For highly disturbed areas, however, runoff sediment concentrations up to 50 000 mg L-1 have been reported (Robinson et al., 1996). In their VFS experiments, Mickelson and Baker (1993) used an artificial runoff with 10 000 mg L-1 of sediment. A concentration of 4000 mg L-1 was arbitrarily selected for this study that, in turn, governed the P content of the artificial runoff. The resultant average concentration of total P in the artificial runoff was 2.37 mg L-1. Soluble P, estimated by measuring dissolved orthophosphate that had filtrated through a 0.45-µm opening filter, varied between 0.10 and 0.30 mg L-1.
A typical test run was divided into five different phases: a wetting phase, an unsaturated phase (Q1A), and three consecutive saturated phases (Q1B, Q.65, and Q.3) with flow rates of 1.0, 1.0, 1.0, 0.65, and 0.3 L s-1, respectively. During the wetting phase, clear water was applied at a rate of 1.0 L s-1 onto the filter strip. Most of this water infiltrated into the soil, as the strip was initially dry. As soon as runoff started at the filter's outlet, the inflow was switched from clear water to artificial runoff at 1.0 L s-1, thus starting Phase Q1A. Soil water conditions in the surface layer during this phase moved from unsaturated to near saturation by the end. The unsaturated phase lasted until the flow rate at the outlet became steady, which indicated that near-saturated soil conditions were obtained. Following this soil saturation, application of artificial runoff continued in three consecutive phases of 15-min duration: Q1B, Q.65, and Q.3 with flow rates of 1.0, 0.65, and 0.3 L s-1, respectively.
Sediment and phosphorus load was introduced into the influent water from a separate premixed sediment tank, which was made of a 1.2-m-high PVC pipe with a 0.15-m diameter. The soil and water mixture was kept homogenized by an air jet placed at the bottom of the tank operating at 400 kPa. The sediment–water mixture was supplied to the filter influent by a peristaltic pump and mixed with clear water before it entered the filter inlet as artificial runoff. The average inlet sediment and P concentrations were kept constant at 2700 and 2.37 mg L-1, respectively. A schematic diagram of the experimental setup is shown in Fig. 1 .
|
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Performance in Phosphorus Removal
Performance of VFS for phosphorus removal was assessed from total P load calculated from influent concentrations and outflow rate compared with influent runoff and P concentration. Excluding results from filter strips B2-2 and B10-1, for which flow-rate measurements were inaccurate, the phosphorus-trapping efficiency (PTE) varied widely from 31% (for Filter B2-1) to 89% (for Filter B15-2) with an average of 61% for all filters. This PTE is 28% lower and had higher variation (coefficient of variation [CV] = 26.6% and ranged from 31 to 89%) than that of the mean sediment-trapping efficiency (CV = 9.5% and ranged from 68 to 98%) reported by Abu-Zreig et al. (2002) for the same plots. Higher CV and range in P removal was also observed in previous studies with similar experimental conditions (Dillaha et al., 1987; Magette et al., 1989; Daniels and Gilliam, 1993). Phosphorus trapping efficiencies reported here are comparable with those found in the literature. In a two-year study under natural rainfall conditions, Daniels and Gilliam (1993) found that the average PTE for 3- and 6-m-long filter strips vegetated with fescue was 55 and 70%, respectively. Their results are within the range obtained in the present study for the D-type filters with well-established natural vegetation but higher than that obtained with B-type filters with recently established legumes and fescue. In another study, Dillaha et al. (1989) reported trapping efficiencies of 73 and 93% for 4.6- and 9.1-m-long strips with a slope of 11%. With a 16% slope, the corresponding trapping efficiencies were 49 and 65% for the 4.6- and 9.1-m-long strips. Despite differences on the experimental conditions of the work presented here, such as the use of artificial runoff and the absence of rainfall, phosphorous trapping efficiencies with vegetative filters were quite comparable with other studies.
Effect of Filter Length
The effect of flowpath length on PTE was determined by comparing trapping efficiencies for the Type B strips, excluding filter strips B2-2 and B10-1. The results are shown in Fig. 2
. Phosphorus-trapping efficiency increased steadily with length of filter strip and the increase is highly significant (p = 0.003) using multiple range variance analysis at a 95% probability level. However, the increase in performance with length decreased rapidly as filter length increased beyond 10 m. A power regression line seemed to fit the PTE data versus filter length very well (PTE = 24.78 [length]0.437; R2 = 0.88), and the results are shown in Fig. 2. The average phosphorus trapping efficiencies of the 2-, 5-, 10-, and 15-m-long strips were 32, 54, 67, and 79%, respectively. These are lower than the sediment-trapping efficiencies reported by Abu-Zreig et al. (2002). The difference between sediment and phosphorus trapping appears to be large for short strips and small for longer strips. Hence, while shorter strips offer good sediment control, they offer much less control over phosphorus. Increasing the filter length from 10 to 15 m increased the PTE by 12% (Table 2), but had negligible influence on sediment trapping as reported by Abu-Zreig et al. (2002). This is probably due to the dilution of phosphorus in filter strips being higher in long filters compared with short ones. Other researchers have also observed a uniform decrease in P concentration in overland flow with flowpath length without vegetation (McDowell and Sharpley, 2002). Dilution was hypothesized as the major factor for this decrease. In the presence of vegetation and enhanced sediment deposition the decrease in P concentration will be even more profound.
|
|
|
|
Effect of Vegetation Type
The influence of vegetation type is shown in Fig. 4
, where the average and range of PTE values of the 5-m-long filters were plotted against vegetation type. As shown in Fig. 4, the PTE of Type D filters showed a higher PTE, 68% on average and ranging from 65 to 72%, compared with that of Type A (65% on average) and B (54% on average) filters, which ranged from 47 to 60%. The Tukey HSD multiple comparisons test was performed between filters of similar length assuming that the single PTE value of the A filter is an average of 3 with a standard deviation similar to that of D5 filters. The test has shown that PTE values of D5 filters were significantly higher than those of B5 filters (P = 0.04) and also higher than those of A5 filters, with no significant difference. The PTEs of A5 filters were somewhat higher than those of B5 filters (p = 0.12). The advantages of native vegetation (Type D strips) under low-flow conditions, Phases Q.65 and Q.3, were even more profound. As shown in Table 3, the PTEs of D5 filters in phases Q.65 and Q.3 were 72 and 92%, respectively, which were equal or even higher than PTEs of the 10- and 15-m-long B filters. Considering that the slope of Type D filters was double that of other filters and a decrease in PTE would be expected, as reported by Dillaha et al. (1989), the actual PTE of native vegetation filters (Type D) could be greater than reported.
|
|
Mechanisms other than sedimentation and infiltration can enhance P removal in VFS. Possible additional mechanisms include adsorption to plant and soil surfaces and absorption of soluble P by plants. Improved P performance in Type D filter strips can be linked to these mechanisms since Type D vegetation had higher vegetation coverage, a rougher flow surface, and the presence of a decaying thatch near the ground surface. Generally, Type D filter strips had higher water retention (39–58%) than Type B strips (37–44%), and also had average flow velocities of 0.04 m s-1, which were half of the Type B strips, with water velocities of 0.09 m s-1. Lower flow velocities and greater water retention resulted in more contact time with the vegetation and soil, less erosive power, and less transport capacity, especially for fine particles with high P content (McDowell and Sharpley, 2002).
Phosphorus Removal in Relation to Sediment Removal
Experimental results presented here have shown that the relative importance of mechanisms for P removal is different from that of suspended solids. The average P removal in VFS (61%) was less than that of suspended solids (84%) (Abu-Zreig et al., 2002). Furthermore, the removal of P was seen to increase more steadily with filter length compared with the removal of suspended solids. Sediment trapping efficiencies for the 2-, 5-, 10-, and 15-m-long Type B filters were 65, 81, 92, and 91%, respectively (Abu-Zreig et al., 2002). The corresponding phosphorus trapping efficiencies of B filters were 32, 54, 67, and 79%, respectively. Increasing the length of filter strips from 2 to 15 m would increase the PTE by 47% compared with only 25% in the case of sediment trapping.
The main reason for this difference is the change in relative importance of the removal mechanisms in relation to the uneven distribution of P in different size classes of aggregates or soil particles. In this study more than 90% of P concentration in the runoff was sediment bound. Phosphorus tends to be more present in the particles smaller than 100 microns, that is, silt and clay fractions (He et al., 1995). It is also known that sediment settling in VFS is directly related to particle size. In a simulation study, Abu-Zreig (2001) found that the sediment trapping efficiency in a 3-m-long filter for sand (d = 0.2 mm), silt (d = 0.01 mm), and clay (d = 0.002 mm) particles was about 90, 60, and 2%, respectively.
Hence, while filter strips 2 m long were efficient in trapping sand and similar-sized large aggregates (Abu-Zreig et al., 2002), they were not efficient in removing clay particles, which contain most of the P. The PTE of the 15-m-long filters (79%) was higher than that of short filters because longer filters retain smaller particles better than short filters via deposition and provide more infiltration opportunity of overland flow.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Related articles in JEQ:
This article has been cited by other articles:
|
|
 |
|
  K. E. Smith, R. A. Putnam, C. Phaneuf, G. R. Lanza, O. P. Dhankher, and J. M. Clark Selection of Plants for Optimization of Vegetative Filter Strips Treating Runoff from Turfgrass J. Environ. Qual., August 8, 2008; 37(5): 1855 - 1861. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Kumar, S.H. Anderson, L.G. Bricknell, R.P. Udawatta, and C.J. Gantzer Soil hydraulic properties influenced by agroforestry and grass buffers for grazed pasture systems Journal of Soil and Water Conservation, July 1, 2008; 63(4): 224 - 232. [Abstract] [PDF]
|
|
|
|
|
|
D. J. Wagner, H. A. Elliott, R. C. Brandt, and D. Jaiswal Managing Biosolids Runoff Phosphorus Using Buffer Strips Enhanced with Drinking Water Treatment Residuals J. Environ. Qual., June 23, 2008; 37(4): 1567 - 1574. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. R. Sistani, G. E. Brink, and J. L. Oldham Managing broiler litter application rate and grazing to decrease watershed runoff losses. J. Environ. Qual., March 1, 2008; 37(2): 718 - 724. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. P. Udawatta, C. J. Gantzer, S. H. Anderson, and H. E. Garrett Agroforestry and Grass Buffer Effects on Pore Characteristics Measured by High-Resolution X-ray Computed Tomography Soil Sci. Soc. Am. J., January 25, 2008; 72(2): 295 - 304. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Steinke, J. C. Stier, W. R. Kussow, and A. Thompson Prairie and Turf Buffer Strips for Controlling Runoff from Paved Surfaces J. Environ. Qual., January 25, 2007; 36(2): 426 - 439. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. P. Udawatta, S. H. Anderson, C. J. Gantzer, and H. E. Garrett Agroforestry and Grass Buffer Influence on Macropore Characteristics: A Computed Tomography Analysis Soil Sci. Soc. Am. J., August 22, 2006; 70(5): 1763 - 1773. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Blanco-Canqui, C. J. Gantzer, S. H. Anderson, and E. E. Alberts Grass Barriers for Reduced Concentrated Flow Induced Soil and Nutrient Loss Soil Sci. Soc. Am. J., November 1, 2004; 68(6): 1963 - 1972. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. P. Udawatta, P. P. Motavalli, and H. E. Garrett Phosphorus Loss and Runoff Characteristics in Three Adjacent Agricultural Watersheds with Claypan Soils J. Environ. Qual., September 1, 2004; 33(5): 1709 - 1719. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Soil Science Society of America Journal | Journal of Plant Registrations | The Plant Genome | |||
