- Email: [email protected]
A modified ceramic sampler and lysimeter design for improved monitoring of soil leachates
Water Research 36 (2002) 799–804
Technical note
A modified ceramic sampler and lysimeter design for improved monitoring of soil leachates Mark S. Bowman*, Timothy S. Clune, Bruce G. Sutton Department of Crop Sciences (A20), University of Sydney, Sydney NSW, 2006 Australia Received 28 August 2000; accepted 9 May 2001
Abstract This paper describes the design and use of modified solution samplers and non-weighing lysimeters in a field experiment examining the response of a turf–soil system to landfill leachate irrigation over a 2-year period. The two designs were shown to offer significant performance advantages, were cost effective and overcame many of the reported problems commonly associated with each technique. The quantities of leached chloride detected in the 20%, 50% and undiluted leachate irrigated plots by the modified solution samplers (1677, 4319 and 8021 kg ha 1, respectively) and microlysimeters (1759, 4512 and 8160 kg ha 1, respectively) were significantly higher than the conventional sampler design (1407, 3767 and 7052 kg ha 1, respectively). Additionally, the microlysimeter design functioned reliably throughout the experiment, achieving solution flow rates analogous to the unconfined plots. Therefore, it was concluded that both the designs appear to be suitable for monitoring changes in soil solution composition associated with sub-surface wastewater irrigation. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Solution samplers; Lysimeters; Nitrate; Leachate; Chloride
1. Introduction Leaching of nitrogen (N) from agricultural and urban plant–soil systems is typically quantified using ceramic cup solution samplers. These samplers are primarily utilised because they are inexpensive and simple to install [1]. Although a range of sampler designs and installation procedures are widely used, their utility is limited by problems associated with preferential flow [2] and contamination due to reaction between the solution and ceramic cup [3]. Additionally, increased use of irrigation technologies such as sub-surface drip irrigation, have further decreased the sensitivity of these methodologies by increasing the potential for N plumes to pass the samplers undetected. As such, modifications to the conventional sampler designs may be required to obtain representative solution samples. *Corresponding author. Tel.: +61-2-6207-6350; fax: +61-26207-6341. E-mail address: [email protected] (M.S. Bowman).
An alternative technique to the use of ceramic cup samplers for determining N leaching losses from plant– soil systems is lysimetry. Lysimeters comprise a confined, intact soil column with a provision for solution sampling that allows an accurate measurement of nutrient source/sink relationships [1]. However, the utilisation of lysimeters is limited as they are expensive to install, may exhibit preferential flow around the edge of the monolith and failure rates approaching 35% have been reported [4–6]. The current study is part of a larger programme examining the feasibility of in situ bioremediation of Nrich leachate through the irrigation of turf–soil systems. As such, a number of leachate dilution treatments (20%, 50% and 100% leachate) were used to irrigate turf at the Newington landfill, Sydney. The primary aim of irrigation management is to minimise the movement of N into the surrounding environment whilst maximising N applications to meet the daily requirements of the landfill. The regulatory limit for the emission of nitrate (NO3 ) to waterways is 10 mg L 1 and ammonia (NH3) 0.1 mg L 1 [7]. Consequently, practical, inexpensive and
0043-1354/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 3 - 1 3 5 4 ( 0 1 ) 0 0 2 6 4 - 0
800
M.S. Bowman et al. / Water Research 36 (2002) 799–804
timely data sets are required to facilitate the planning of management systems that promote efficient bioremediation of the N-rich leachate. Ceramic cup soil solution samplers have been utilised to assess N leaching from subsurface irrigated experimental plots and lysimeters to provide a more reliable measure of N leaching losses from confined soil columns. To provide practical data for an evaluation of management practices and to overcome the design limitations outlined, modifications to the conventional ceramic cup sampler design and a new microlysimeter design utilising low-cost components and installation procedures are described.
2. Materials and methods 2.1. Site description The study was conducted at the Newington Landfill Homebush Bay, 10 km west of Sydney, Australia. Subsurface drip irrigation (16 mm Geoflow Rootguard) supplied four treatment plots with town water (control), 20%, 50% and undiluted leachate derived from a subsurface drain at the nearby landfill (Table 1). The predominant turf species were kikuyu (Pennisetum clandestinum) and couch (Cynodon dactylon) with the soil profile consisting of a well-structured (bulk density 1.3 g cm 3) sandy loam topsoil (depth 200 mm) overlying a compacted clay subsoil layer to a depth of 600 mm. The topsoil was acidic (pH 5.8), non-saline (ECe 0.9 dS m 1) moderately sodic (71 mg Na+ kg 1), contained adequate organic carbon (2.09%) and had an effective cation exchange capacity (ECEC) of 10.3 cmol kg 1. 2.2. Solution samplers Soil solution chemistry was monitored in the leachate treatment plots using two porous ceramic cup sampler designs. The conventional design (Fig. 1) utilised ceramic cups (Cooinda Ceramics, Melbourne, size 7)
Table 1 Mean analysis of leachate and town water (7SD) Analyte
Leachate
Town water
pH EC (dS m 1) 1 NH+ 4 (mg L ) Cl (mg L 1) Na+ (mg L 1) SARa
6.871.2 17.674.6 285731.8 81727563 40127830 33
6.970.4 0.170.04 0.0570.01 1572.2 871.9 1
a
Sodium adsorption ratio [8].
Fig. 1. Schematic design for conventional ceramic cup samplers.
installed at two depths (300 and 600 mm, respectively) as described by Hansen and Harris [3] with six replicates in each of the four treatment plots. To complement the conventional solution sampler data, a revised solution sampler design was trialed during the field experiment. The modified samplers were constructed of 40 mm electrical conduit, installed to sit on the surface, sealed with a rubber stopper drilled to incorporate two sampling tubes (Fig. 2). The modified samplers differed significantly from the conventional sampler below the soil surface, with the 40 mm conduit split by a 3-way (25 mm) PVC fitting. The horizontal arm of the fitting was attached to a length of conduit (295 mm) onto which a ceramic cup was fitted. The bottom of the vertical arm of the PVC fitting was sealed with a pushon end cap (25 mm) forming a reservoir from which the solution was collected. The modified samplers were installed similar to the conventional design. However, a shovel was used to excavate the primary access hole for the main body and reservoir of the sampler. To accommodate the horizontal arm of the sampler, a small tunnel (length 305 mm)
M.S. Bowman et al. / Water Research 36 (2002) 799–804
801
Fig. 2. Schematic design schematic for the modified ceramic cup samplers. The horizontal arm (51 gradient) incorporated into the modified sampler facilitated the collection of solution samples from undisturbed soil beneath point-source emitters.
was excavated horizontally into the soil face from the primary access tunnel. Care was taken to prevent disturbance of the overlying soil and the end of the excavated tunnel was backfilled with sand (approximate diameter 30 mm). The arm of the sampler fitted with the ceramic cup was gently inserted into the sand. Each sampler was positioned at a slight angle (51) so that solution collecting in the horizontal arm drained into the reservoir. The tunnel surrounding the arm was backfilled with a soil/water slurry and a layer of bentonite was positioned around the reservoir and horizontal arm to minimise preferential flow around the body of the sampler. 2.3. Microlysimeters Nine microlysimeters were constructed (Fig. 3) at the Newington site to collect soil solution from confined, intact soil cores (diameter 300 mm depth 600 mm). Intact soil cores were collected using a soil corer mounted on a drill rig and were placed in a PVC casing which formed the body of the microlysimeter (height 1010 mm diameter 300 mm). A recessed PVC sheet (Fig. 3) sealed the base of each cylinder. A second PVC sheet, installed above the first (400 mm) formed a shelf to support the soil core and to enclose the collection vessel. The second shelf was supported by a steel ring attached to the inside of the PVC cylinder and was predrilled with five holes (19 mm diameter) to facilitate the
Fig. 3. Microlysimeter cut-away design schematics. The solution was collected by applying vacuum ( 200 mmH2O) to approximate field suction.
drainage of leachate into the collection vessel. Two pressure tubing lines were installed for the removal of the solution from within the collection vessel to the surface. Both the lines exited the top of the second PVC shelf and were attached to the inside of the PVC cylinder, providing surface sampling access to the buried collection vessels. Prior to the commencement of irrigation, the microlysimeters were subject to saturation and drying phases during a two-month equilibration period after which the space between the intact column and the PVC cylinder was filled with warm (501C) liquefied petroleum jelly to surround the soil monolith [5]. Irrigation was applied by pressure compensating emitters (4 L h 1) and was scheduled to achieve a 20% leachate dilution equivalent to the irrigation applied to the unconfined leachate plots.
2.4. Solution sampling and analysis Solution samples were collected from the conventional and modified ceramic cup samplers on a monthly
802
M.S. Bowman et al. / Water Research 36 (2002) 799–804
basis. The vacuum system (Fig. 4) applied constant suction ( 200 mmH2O) to the microlysimeters to facilitate the movement of water through the intact columns. The solution samples were collected under suction ( 3000 mmH2O) on a monthly basis. All solution samples were assayed for pH, EC (dS m 1), chloride (Cl ), ammonium (NH+ 4 ), nitrate (NO3 ) and nitrite (NO2 ). The mass flow of species in solution was calculated based on the flux (L m 2 d 1) and concentration of solution passing the ceramic cups, derived from irrigation and weather data, and for the microlysimeters, based on the volume of solution collected in the reservoirs each month. The relative efficiency of each solution sampling technique was assessed using Cl in the leachate as a tracer to monitor the movement of the solution. Significant differences (Pp0:05) between means were determined by ANOVA.
3. Results and discussion Both the ceramic cup sampler designs appeared to offer a reasonable description of turf–soil response to leachate irrigation. In particular, the Cl tracer data provided a useful indication of the relative performance and descriptive capability of each soil solution sampling technique. The microlysimeters recovered more Cl than was applied by irrigation (102.8%, 101.7% and 102.3% for the 20%, 50% and undiluted leachate treated plots), suggesting that either some leaching of pre-existing soil Cl and/or airborne deposition salt on the treatment plots occurred. The conventional solution samplers recovered significantly less Cl (1407, 3767 and
7052 kg ha 1 yr 1 for the 20%, 50% and undiluted leachate treatments, respectively) than the modified design (1677, 4219 and 8021 kg ha 1 yr 1 for the 20%, 50% and undiluted leachate treatments, respectively) (Table 2). The rate of conventional sampler Cl recovery in the 20% leachate treated plot (79.9%) was particularly low compared to the modified samplers (95.3%) and microlysimeters (102.8%). These results suggest that the modified sampler design adequately described soil solution chemistry and could be confidently used to monitor N leaching losses. The modified soil solution sampler most effectively described the movement of applied Cl because it facilitated the continuous sampling of solution passing the cup between monthly sampling events. In contrast, the conventional samplers provided a discontinuous record of changes in solution chemistry because the design only allowed intermittent sample collection, limiting their capacity to detect solution fluctuations associated with high rainfall or irrigation events. Furthermore, the modified samplers were installed directly below the irrigation emitters, minimising the potential error associated with preferential flow. In comparison, the conventional samplers were installed adjacent to the emitter, increasing their susceptibility to preferential flow. Consequently, the modified sampler design offers an enhanced estimation of soil solution chemistry, particularly for monitoring sub-surface irrigation systems, where an appropriate placement of the sampler in relation to the emitter is necessary so as to obtain a representative soil solution. However, the modified soil solution sampler design reported significantly (Pp0:05) higher N leaching rates
Fig. 4. Solution sampling apparatus for the microlysimeters.
803
M.S. Bowman et al. / Water Research 36 (2002) 799–804
Table 2 Comparison of mean Cl (kg ha 1 yr 1) leaching losses determined by three solution sampling techniques. Data are presented for 20%, 50% and undiluted leachate irrigated plotsa Sampler Cl (kg ha 1) Microlysimeters Conventional Modified Recovery efficiency (%Cl )c Microlysimeters Conventional Modified
20%
50%
Undiluted
1759735 14077303 16777148b
45127104 37677526 42197292b
81607215 70527799 80217312b
102.8 79.9 95.3
101.7 83.4 93.5
102.3 86.4 98.3
a
Data are the mean (7SD) total of 24 monthly observations. Significant differences (Pp0:05) reported for conventional and modified samplers relative to the microlysimeter data. c Recovery efficiencies (%) equivalent to the percentage of Cl measured in solution relative to the difference between applied and accumulated soil Cl levels. b
than the conventional design (Table 2). The quantity of leached NO3 in the 20% leachate treated plot detected by the modified samplers was 61 kg ha 1 more than the conventional samplers. This discrepancy was attributed to the installation of the ceramic cups attached to the modified samplers directly below the irrigation emitters, minimising the potential sources of error associated with preferential flow. The design also facilitated a continuous sampling of the solution passing the cup between monthly sampling events. The conventional samplers only allowed intermittent sample collection, limiting their capacity to detect solution fluctuations associated with high rainfall or irrigation events. Solution flow rates through the soil profile (0– 600 mm) in the microlysimeters (5.270.5 mm d 1) were similar to the unconfined plots (5.770.4 mm d 1). This indicates that preferential flow was effectively controlled by the injection of petroleum jelly around the soil monolith and that the vacuum system overcame the disruption of the soil matrix following the installation of the solution sampling apparatus. Furthermore, in contrast to other studies [3] that have reported high failure rates, all microlysimeters were operational at the completion of the experiment. Therefore, from a solute transport perspective, the capacity of the Newington microlysimeters to maintain flow rates approaching those of the unconfined plots and their continued functionality suggests that the design is suitable for the assessment of turf–soil response to leachate irrigation management and in other soil solution monitoring applications. An analysis of solution samples indicated that the microlysimeters provided unambiguous quantification of leachate composition changes following exposure to the turf-soil system. During the experiment, changes in sample NO3 and dissolved salt were highly correlated (Pp0:05) with high intensity rainfall events (Fig. 6) and
Fig. 5. Comparison of mean conventional ceramic sampler, modified solution sampler and microlysimeter NO3 leaching data. The data presented are the means of triplicate samples.
leachate application rates during the summer months. A strong correlation was identified between the modified solution sampler and microlysimeter NO3 breakthrough data (Fig. 5). The levels of NO3 reported at the base of the microlysimeters were not significantly (PX0:05) different to the modified solution sampler data and did not exceed the regulatory limit of 10 mg NO3 L 1 [7]. Therefore, a chemical analysis of the microlysimeter solution data demonstrated that the design utilised in the current programme has the capacity to provide direct feedback of changes in leachate N composition following exposure to the turf–soil system, and to be used as an effective real-time irrigation management tool. Additionally, the strong correlation observed between the modified ceramic cup sampler and microlysimeter data confirms that soil solution samples obtained using the modified design are better able to
804
M.S. Bowman et al. / Water Research 36 (2002) 799–804
modified ceramic cup sampler design. Chloride leached from the 20%, 50% and undiluted leachate irrigated plots by the modified solution samplers (1677, 4319 and 8021 kg ha 1, respectively) and microlysimeters (1759, 4512 and 8160 kg ha 1, respectively) were significantly higher than the conventional sampler design (1407, 3767 and 7052 kg ha 1, respectively). The potential of using the modified ceramic cup and new microlysimeter designs to monitor soil solution in other similar in situ wastewater treatment systems appears significant.
Acknowledgements We are pleased to acknowledge Waste Service NSW for funding and logistical support of this research. The technical support provided by Stephen Garland and Stuart Farrar is also gratefully acknowledged. Fig. 6. Newington field site rainfall, evapotranspiration and irrigation volumes (mm).
References describe solution movement in sub-surface irrigated soils than the conventional design. Although more time consuming to install than the conventional design, excavation and backfilling of the tunnel to accommodate the arm of the modified sampler was readily accomplished, allowing an easy installation of the modified samplers.
4. Conclusions The monitoring protocol adopted at the Newington landfill test site allowed an accurate quantification of the effect of the turf–soil system on N and dissolved salt composition of the leachate. The modified solution samplers provided a good compromise between the higher installation cost of the microlysimeters and poorer quality of data obtained from the conventional samplers. The new solution samplers offered an enhanced estimation of soil solution chemistry, particularly the
[1] Addiscott TM, Whitmore AP, Powlson DS. Farming, fertilizers and the nitrate problem. CAB International, Wallingford, 1991. p. 55–72. [2] Wood WW. Atechnique using porous cups at any depth in the unsaturated zone. Water Res 1973;9:486–8. [3] Hansen EA, Harris AR. Validity of soil water samples collected from porous ceramic cups. Soil Sci Soc Am Proc 1975;39:528–36. [4] Reeder JD. Anon-weighing lysimeter design for field studies using nitrogen-15. Soil Sci Soc Am J 1985;50:1224–7. [5] Cameron KC, Harrison DF, Smith NP. A method to prevent edge flow in undisturbed soil cores and lysimeters. Aus J Soil Res 1990;28:879–86. [6] Young MH, Wierenga PJ, Mancino CF. Large weighing lysimeters for water use and deep percolation studies. Soil Sci 1996;161:491–501. [7] Australian and New Zealand Australian water quality guidelines for fresh and marine waters, Canberra. [8] Rhoades JD, Kanidah A, Mashali AM. The use of saline waters for crop production. FAO Irrigation and Drainage Paper, 48, Food and Agriculture Organisation of the United Nations, Rome, 1992.
Technical note
A modified ceramic sampler and lysimeter design for improved monitoring of soil leachates Mark S. Bowman*, Timothy S. Clune, Bruce G. Sutton Department of Crop Sciences (A20), University of Sydney, Sydney NSW, 2006 Australia Received 28 August 2000; accepted 9 May 2001
Abstract This paper describes the design and use of modified solution samplers and non-weighing lysimeters in a field experiment examining the response of a turf–soil system to landfill leachate irrigation over a 2-year period. The two designs were shown to offer significant performance advantages, were cost effective and overcame many of the reported problems commonly associated with each technique. The quantities of leached chloride detected in the 20%, 50% and undiluted leachate irrigated plots by the modified solution samplers (1677, 4319 and 8021 kg ha 1, respectively) and microlysimeters (1759, 4512 and 8160 kg ha 1, respectively) were significantly higher than the conventional sampler design (1407, 3767 and 7052 kg ha 1, respectively). Additionally, the microlysimeter design functioned reliably throughout the experiment, achieving solution flow rates analogous to the unconfined plots. Therefore, it was concluded that both the designs appear to be suitable for monitoring changes in soil solution composition associated with sub-surface wastewater irrigation. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Solution samplers; Lysimeters; Nitrate; Leachate; Chloride
1. Introduction Leaching of nitrogen (N) from agricultural and urban plant–soil systems is typically quantified using ceramic cup solution samplers. These samplers are primarily utilised because they are inexpensive and simple to install [1]. Although a range of sampler designs and installation procedures are widely used, their utility is limited by problems associated with preferential flow [2] and contamination due to reaction between the solution and ceramic cup [3]. Additionally, increased use of irrigation technologies such as sub-surface drip irrigation, have further decreased the sensitivity of these methodologies by increasing the potential for N plumes to pass the samplers undetected. As such, modifications to the conventional sampler designs may be required to obtain representative solution samples. *Corresponding author. Tel.: +61-2-6207-6350; fax: +61-26207-6341. E-mail address: [email protected] (M.S. Bowman).
An alternative technique to the use of ceramic cup samplers for determining N leaching losses from plant– soil systems is lysimetry. Lysimeters comprise a confined, intact soil column with a provision for solution sampling that allows an accurate measurement of nutrient source/sink relationships [1]. However, the utilisation of lysimeters is limited as they are expensive to install, may exhibit preferential flow around the edge of the monolith and failure rates approaching 35% have been reported [4–6]. The current study is part of a larger programme examining the feasibility of in situ bioremediation of Nrich leachate through the irrigation of turf–soil systems. As such, a number of leachate dilution treatments (20%, 50% and 100% leachate) were used to irrigate turf at the Newington landfill, Sydney. The primary aim of irrigation management is to minimise the movement of N into the surrounding environment whilst maximising N applications to meet the daily requirements of the landfill. The regulatory limit for the emission of nitrate (NO3 ) to waterways is 10 mg L 1 and ammonia (NH3) 0.1 mg L 1 [7]. Consequently, practical, inexpensive and
0043-1354/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 3 - 1 3 5 4 ( 0 1 ) 0 0 2 6 4 - 0
800
M.S. Bowman et al. / Water Research 36 (2002) 799–804
timely data sets are required to facilitate the planning of management systems that promote efficient bioremediation of the N-rich leachate. Ceramic cup soil solution samplers have been utilised to assess N leaching from subsurface irrigated experimental plots and lysimeters to provide a more reliable measure of N leaching losses from confined soil columns. To provide practical data for an evaluation of management practices and to overcome the design limitations outlined, modifications to the conventional ceramic cup sampler design and a new microlysimeter design utilising low-cost components and installation procedures are described.
2. Materials and methods 2.1. Site description The study was conducted at the Newington Landfill Homebush Bay, 10 km west of Sydney, Australia. Subsurface drip irrigation (16 mm Geoflow Rootguard) supplied four treatment plots with town water (control), 20%, 50% and undiluted leachate derived from a subsurface drain at the nearby landfill (Table 1). The predominant turf species were kikuyu (Pennisetum clandestinum) and couch (Cynodon dactylon) with the soil profile consisting of a well-structured (bulk density 1.3 g cm 3) sandy loam topsoil (depth 200 mm) overlying a compacted clay subsoil layer to a depth of 600 mm. The topsoil was acidic (pH 5.8), non-saline (ECe 0.9 dS m 1) moderately sodic (71 mg Na+ kg 1), contained adequate organic carbon (2.09%) and had an effective cation exchange capacity (ECEC) of 10.3 cmol kg 1. 2.2. Solution samplers Soil solution chemistry was monitored in the leachate treatment plots using two porous ceramic cup sampler designs. The conventional design (Fig. 1) utilised ceramic cups (Cooinda Ceramics, Melbourne, size 7)
Table 1 Mean analysis of leachate and town water (7SD) Analyte
Leachate
Town water
pH EC (dS m 1) 1 NH+ 4 (mg L ) Cl (mg L 1) Na+ (mg L 1) SARa
6.871.2 17.674.6 285731.8 81727563 40127830 33
6.970.4 0.170.04 0.0570.01 1572.2 871.9 1
a
Sodium adsorption ratio [8].
Fig. 1. Schematic design for conventional ceramic cup samplers.
installed at two depths (300 and 600 mm, respectively) as described by Hansen and Harris [3] with six replicates in each of the four treatment plots. To complement the conventional solution sampler data, a revised solution sampler design was trialed during the field experiment. The modified samplers were constructed of 40 mm electrical conduit, installed to sit on the surface, sealed with a rubber stopper drilled to incorporate two sampling tubes (Fig. 2). The modified samplers differed significantly from the conventional sampler below the soil surface, with the 40 mm conduit split by a 3-way (25 mm) PVC fitting. The horizontal arm of the fitting was attached to a length of conduit (295 mm) onto which a ceramic cup was fitted. The bottom of the vertical arm of the PVC fitting was sealed with a pushon end cap (25 mm) forming a reservoir from which the solution was collected. The modified samplers were installed similar to the conventional design. However, a shovel was used to excavate the primary access hole for the main body and reservoir of the sampler. To accommodate the horizontal arm of the sampler, a small tunnel (length 305 mm)
M.S. Bowman et al. / Water Research 36 (2002) 799–804
801
Fig. 2. Schematic design schematic for the modified ceramic cup samplers. The horizontal arm (51 gradient) incorporated into the modified sampler facilitated the collection of solution samples from undisturbed soil beneath point-source emitters.
was excavated horizontally into the soil face from the primary access tunnel. Care was taken to prevent disturbance of the overlying soil and the end of the excavated tunnel was backfilled with sand (approximate diameter 30 mm). The arm of the sampler fitted with the ceramic cup was gently inserted into the sand. Each sampler was positioned at a slight angle (51) so that solution collecting in the horizontal arm drained into the reservoir. The tunnel surrounding the arm was backfilled with a soil/water slurry and a layer of bentonite was positioned around the reservoir and horizontal arm to minimise preferential flow around the body of the sampler. 2.3. Microlysimeters Nine microlysimeters were constructed (Fig. 3) at the Newington site to collect soil solution from confined, intact soil cores (diameter 300 mm depth 600 mm). Intact soil cores were collected using a soil corer mounted on a drill rig and were placed in a PVC casing which formed the body of the microlysimeter (height 1010 mm diameter 300 mm). A recessed PVC sheet (Fig. 3) sealed the base of each cylinder. A second PVC sheet, installed above the first (400 mm) formed a shelf to support the soil core and to enclose the collection vessel. The second shelf was supported by a steel ring attached to the inside of the PVC cylinder and was predrilled with five holes (19 mm diameter) to facilitate the
Fig. 3. Microlysimeter cut-away design schematics. The solution was collected by applying vacuum ( 200 mmH2O) to approximate field suction.
drainage of leachate into the collection vessel. Two pressure tubing lines were installed for the removal of the solution from within the collection vessel to the surface. Both the lines exited the top of the second PVC shelf and were attached to the inside of the PVC cylinder, providing surface sampling access to the buried collection vessels. Prior to the commencement of irrigation, the microlysimeters were subject to saturation and drying phases during a two-month equilibration period after which the space between the intact column and the PVC cylinder was filled with warm (501C) liquefied petroleum jelly to surround the soil monolith [5]. Irrigation was applied by pressure compensating emitters (4 L h 1) and was scheduled to achieve a 20% leachate dilution equivalent to the irrigation applied to the unconfined leachate plots.
2.4. Solution sampling and analysis Solution samples were collected from the conventional and modified ceramic cup samplers on a monthly
802
M.S. Bowman et al. / Water Research 36 (2002) 799–804
basis. The vacuum system (Fig. 4) applied constant suction ( 200 mmH2O) to the microlysimeters to facilitate the movement of water through the intact columns. The solution samples were collected under suction ( 3000 mmH2O) on a monthly basis. All solution samples were assayed for pH, EC (dS m 1), chloride (Cl ), ammonium (NH+ 4 ), nitrate (NO3 ) and nitrite (NO2 ). The mass flow of species in solution was calculated based on the flux (L m 2 d 1) and concentration of solution passing the ceramic cups, derived from irrigation and weather data, and for the microlysimeters, based on the volume of solution collected in the reservoirs each month. The relative efficiency of each solution sampling technique was assessed using Cl in the leachate as a tracer to monitor the movement of the solution. Significant differences (Pp0:05) between means were determined by ANOVA.
3. Results and discussion Both the ceramic cup sampler designs appeared to offer a reasonable description of turf–soil response to leachate irrigation. In particular, the Cl tracer data provided a useful indication of the relative performance and descriptive capability of each soil solution sampling technique. The microlysimeters recovered more Cl than was applied by irrigation (102.8%, 101.7% and 102.3% for the 20%, 50% and undiluted leachate treated plots), suggesting that either some leaching of pre-existing soil Cl and/or airborne deposition salt on the treatment plots occurred. The conventional solution samplers recovered significantly less Cl (1407, 3767 and
7052 kg ha 1 yr 1 for the 20%, 50% and undiluted leachate treatments, respectively) than the modified design (1677, 4219 and 8021 kg ha 1 yr 1 for the 20%, 50% and undiluted leachate treatments, respectively) (Table 2). The rate of conventional sampler Cl recovery in the 20% leachate treated plot (79.9%) was particularly low compared to the modified samplers (95.3%) and microlysimeters (102.8%). These results suggest that the modified sampler design adequately described soil solution chemistry and could be confidently used to monitor N leaching losses. The modified soil solution sampler most effectively described the movement of applied Cl because it facilitated the continuous sampling of solution passing the cup between monthly sampling events. In contrast, the conventional samplers provided a discontinuous record of changes in solution chemistry because the design only allowed intermittent sample collection, limiting their capacity to detect solution fluctuations associated with high rainfall or irrigation events. Furthermore, the modified samplers were installed directly below the irrigation emitters, minimising the potential error associated with preferential flow. In comparison, the conventional samplers were installed adjacent to the emitter, increasing their susceptibility to preferential flow. Consequently, the modified sampler design offers an enhanced estimation of soil solution chemistry, particularly for monitoring sub-surface irrigation systems, where an appropriate placement of the sampler in relation to the emitter is necessary so as to obtain a representative soil solution. However, the modified soil solution sampler design reported significantly (Pp0:05) higher N leaching rates
Fig. 4. Solution sampling apparatus for the microlysimeters.
803
M.S. Bowman et al. / Water Research 36 (2002) 799–804
Table 2 Comparison of mean Cl (kg ha 1 yr 1) leaching losses determined by three solution sampling techniques. Data are presented for 20%, 50% and undiluted leachate irrigated plotsa Sampler Cl (kg ha 1) Microlysimeters Conventional Modified Recovery efficiency (%Cl )c Microlysimeters Conventional Modified
20%
50%
Undiluted
1759735 14077303 16777148b
45127104 37677526 42197292b
81607215 70527799 80217312b
102.8 79.9 95.3
101.7 83.4 93.5
102.3 86.4 98.3
a
Data are the mean (7SD) total of 24 monthly observations. Significant differences (Pp0:05) reported for conventional and modified samplers relative to the microlysimeter data. c Recovery efficiencies (%) equivalent to the percentage of Cl measured in solution relative to the difference between applied and accumulated soil Cl levels. b
than the conventional design (Table 2). The quantity of leached NO3 in the 20% leachate treated plot detected by the modified samplers was 61 kg ha 1 more than the conventional samplers. This discrepancy was attributed to the installation of the ceramic cups attached to the modified samplers directly below the irrigation emitters, minimising the potential sources of error associated with preferential flow. The design also facilitated a continuous sampling of the solution passing the cup between monthly sampling events. The conventional samplers only allowed intermittent sample collection, limiting their capacity to detect solution fluctuations associated with high rainfall or irrigation events. Solution flow rates through the soil profile (0– 600 mm) in the microlysimeters (5.270.5 mm d 1) were similar to the unconfined plots (5.770.4 mm d 1). This indicates that preferential flow was effectively controlled by the injection of petroleum jelly around the soil monolith and that the vacuum system overcame the disruption of the soil matrix following the installation of the solution sampling apparatus. Furthermore, in contrast to other studies [3] that have reported high failure rates, all microlysimeters were operational at the completion of the experiment. Therefore, from a solute transport perspective, the capacity of the Newington microlysimeters to maintain flow rates approaching those of the unconfined plots and their continued functionality suggests that the design is suitable for the assessment of turf–soil response to leachate irrigation management and in other soil solution monitoring applications. An analysis of solution samples indicated that the microlysimeters provided unambiguous quantification of leachate composition changes following exposure to the turf-soil system. During the experiment, changes in sample NO3 and dissolved salt were highly correlated (Pp0:05) with high intensity rainfall events (Fig. 6) and
Fig. 5. Comparison of mean conventional ceramic sampler, modified solution sampler and microlysimeter NO3 leaching data. The data presented are the means of triplicate samples.
leachate application rates during the summer months. A strong correlation was identified between the modified solution sampler and microlysimeter NO3 breakthrough data (Fig. 5). The levels of NO3 reported at the base of the microlysimeters were not significantly (PX0:05) different to the modified solution sampler data and did not exceed the regulatory limit of 10 mg NO3 L 1 [7]. Therefore, a chemical analysis of the microlysimeter solution data demonstrated that the design utilised in the current programme has the capacity to provide direct feedback of changes in leachate N composition following exposure to the turf–soil system, and to be used as an effective real-time irrigation management tool. Additionally, the strong correlation observed between the modified ceramic cup sampler and microlysimeter data confirms that soil solution samples obtained using the modified design are better able to
804
M.S. Bowman et al. / Water Research 36 (2002) 799–804
modified ceramic cup sampler design. Chloride leached from the 20%, 50% and undiluted leachate irrigated plots by the modified solution samplers (1677, 4319 and 8021 kg ha 1, respectively) and microlysimeters (1759, 4512 and 8160 kg ha 1, respectively) were significantly higher than the conventional sampler design (1407, 3767 and 7052 kg ha 1, respectively). The potential of using the modified ceramic cup and new microlysimeter designs to monitor soil solution in other similar in situ wastewater treatment systems appears significant.
Acknowledgements We are pleased to acknowledge Waste Service NSW for funding and logistical support of this research. The technical support provided by Stephen Garland and Stuart Farrar is also gratefully acknowledged. Fig. 6. Newington field site rainfall, evapotranspiration and irrigation volumes (mm).
References describe solution movement in sub-surface irrigated soils than the conventional design. Although more time consuming to install than the conventional design, excavation and backfilling of the tunnel to accommodate the arm of the modified sampler was readily accomplished, allowing an easy installation of the modified samplers.
4. Conclusions The monitoring protocol adopted at the Newington landfill test site allowed an accurate quantification of the effect of the turf–soil system on N and dissolved salt composition of the leachate. The modified solution samplers provided a good compromise between the higher installation cost of the microlysimeters and poorer quality of data obtained from the conventional samplers. The new solution samplers offered an enhanced estimation of soil solution chemistry, particularly the
[1] Addiscott TM, Whitmore AP, Powlson DS. Farming, fertilizers and the nitrate problem. CAB International, Wallingford, 1991. p. 55–72. [2] Wood WW. Atechnique using porous cups at any depth in the unsaturated zone. Water Res 1973;9:486–8. [3] Hansen EA, Harris AR. Validity of soil water samples collected from porous ceramic cups. Soil Sci Soc Am Proc 1975;39:528–36. [4] Reeder JD. Anon-weighing lysimeter design for field studies using nitrogen-15. Soil Sci Soc Am J 1985;50:1224–7. [5] Cameron KC, Harrison DF, Smith NP. A method to prevent edge flow in undisturbed soil cores and lysimeters. Aus J Soil Res 1990;28:879–86. [6] Young MH, Wierenga PJ, Mancino CF. Large weighing lysimeters for water use and deep percolation studies. Soil Sci 1996;161:491–501. [7] Australian and New Zealand Australian water quality guidelines for fresh and marine waters, Canberra. [8] Rhoades JD, Kanidah A, Mashali AM. The use of saline waters for crop production. FAO Irrigation and Drainage Paper, 48, Food and Agriculture Organisation of the United Nations, Rome, 1992.