Field and Laboratory Tests of Flow-Proportional Passive Samplers for

Feb 7, 2013 - Grab sampling for nutrient and other chemical/solid determinants in rivers is both coarse and temporally discrete. It is ineffective at ...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/est

Field and Laboratory Tests of Flow-Proportional Passive Samplers for Determining Average Phosphorus and Nitrogen Concentration in Rivers Philip Jordan,†,* Rachel Cassidy,†,‡ Katrina A. Macintosh,† and Joerg Arnscheidt† †

School of Environmental Sciences, University of Ulster, Coleraine, Northern Ireland BT52 1SA School of Planning Architecture and Civil Engineering, Queen’s University, Belfast, Northern Ireland BT9 5AG



S Supporting Information *

ABSTRACT: Flow responsive passive samplers offer considerable potential in nutrient monitoring in catchments; bridging the gap between the intermittency of grab sampling and the high cost of automated monitoring systems. A commercially available passive sampler was evaluated in a number of river systems encapsulating a gradient in storm response, combinations of diffuse and point source pressures, and levels of phosphorus and nitrogen concentrations. Phosphorus and nitrogen are sequestered to a resin matrix in a permeable cartridge positioned in line with streamflow. A salt tracer dissolves in proportion to advective flow through the cartridge. Multiple deployments of different cartridge types were undertaken and the recovery of P and N compared with the flow-weighted mean concentration (FWMC) from high-resolution bank-side analysers at each site. Results from the passive samplers were variable and largely underestimated the FWMC derived from the bank-side analysers. Laboratory tests using ambient river samples indicated good replication of advective throughflow using pumped water, although this appeared not to be a good analogue of river conditions where flow divergence was possible. Laboratory tests also showed good nutrient retention but not elution and these issues appeared to combine to limit the utility in ambient river systems at the small catchment scale.



INTRODUCTION Grab sampling for nutrient and other chemical/solid determinants in rivers is both coarse and temporally discrete. It is ineffective at delineating the precise dynamics of transfers during high- and low-flow events, and, as a result, it is difficult to accurately estimate parameter load and average concentrations from data collected using this sampling methodology.1 For eutrophication assessment and management, an implicit assumption is that if data collection continues over a long enough period then a spectrum of flow behaviors will eventually be captured in the data set, so enabling the assessment of both point and diffuse chemical transfer patterns. In terms of trajectories of change, however, low-flow patterns tend to be overrepresented and high-flow patterns tend to be underrepresented in monitoring data, which rely on grab samples. In such data sets there is, therefore, an inherent bias that can, in some catchments, be overwhelmed by the chronic presence of point source impact patterns.2,3 In recent years, in situ bank-side analysers have been used4,5 for the collection of continuous high-frequency river chemistry data and is a progression from the use of automatic samplers, which require subsequent collection and laboratory analysis.6−8 The collection of high-frequency data sets has, therefore, improved current estimates of periodic nutrient load, expanded © 2013 American Chemical Society

the potential of environmental modeling, and facilitated the capture of low- and particularly high-flow events.4,9−12 However, a major limitation to the widespread use of bankside analysers is the capital required to purchase, install and maintain the equipment.13 A move toward passive sampling techniques may provide a more parsimonious approach; covering flow ranges and acting as a surrogate for continuous time-series.14,15 Deployed in situ for a predefined period of time, and then collected for analysis, passive sampling methodologies have evolved through the development of diffusive gradient in thin-films (DGT) and diffusive equilibrium in thin-films (DET). This gel probe technology is both passive and, for DGT, time-integrated, and can be deployed in situ for the measurement of metals and, more recently, nutrients.16−18 Research to date has documented gel deployments in natural waters,16,19 soils,20,21 sediment pore waters,22−25 and marine and lacustrine sediments.26 Current application of gel passive samplers occurs in aquatic systems where solute flux rate is not Received: Revised: Accepted: Published: 2331

October 9, 2012 February 6, 2013 February 7, 2013 February 7, 2013 dx.doi.org/10.1021/es304108e | Environ. Sci. Technol. 2013, 47, 2331−2338

Environmental Science & Technology

Article

Phosphax (0.01 − 5.0 mg l−1) by digestion of a 10 mL unfiltered sample using the sulphuric acid persulphate method with subsequent addition of molybdate antimony and ascorbic acid and photometric determination of concentration.32,33 Total Reactive P is determined using the same approach, but excluding the digestion step, and is operationally the same as MRP. The Hach Lange Nitratax probe determines nitrate (NO3−N) and nitrite (NO2−N) in a water sample using UV excitation at 210 nm and detection (0.1−25 mg l−1 at sites 1 and 2; 0.1−50.0 mg l−1 at sites 3, 4, and 5). A reference beam at 350 nm compensates for turbidity and color changes. Both instruments were validated with laboratory measurements and, for TON, the data showed NO2−N to be consistently 0.05). With linear least-squares fits through the origin, the TP comparisons indicated that the passive samplers recovered 2333

dx.doi.org/10.1021/es304108e | Environ. Sci. Technol. 2013, 47, 2331−2338

Environmental Science & Technology

Article

Figure 3. (a−i) Scatterplots and regression relationships (solid line) of flow-weighted mean concentrations (FWMC) of TP and NO3−N compared with passive sampler cartridge recoveries and using different entrance frit sizes. The comparisons are made against unity (dashed line) and indicate between 29% and 44% of TP recovered and a satisfactory recovery for NO3−N up to approximately 2 mg l−1.

Figure 4. Comparison of river discharge during deployments with estimated sampler throughflow by tracer loss. The sampler throughflow volumes are estimated as proportions of river discharge that, when integrated with proportions of P and N flux (as eluted from the samplers), are estimates of FWM concentration. The comparisons do not show a consistent recovery of throughflow between and within sites.

2334

dx.doi.org/10.1021/es304108e | Environ. Sci. Technol. 2013, 47, 2331−2338

Environmental Science & Technology

Article

Table 2. Laboratory Pump Test Results of Passive Sampler Cartridge Recoveries (Means of Triplicates) Compared with Ambient River Sample Concentrations at Five Sites Passive Sampler Recovery Concentration

Laboratory Concentration

river

throughflow (l)

throughflow range (l)

P (mg l−1)

P range (mg l−1)

NO3−N (mg l−1)

NO3−N range (mg l−1)

throughflow (l)

TP (mg l−1)

TSP (mg l−1)

TRP (mg l−1)

SRP (mg l−1)

NO3−N (mg l−1)

4

0.987

0.902−1.119

0.014

4.83

4.20−5.30

0.843

0.032

0.029

0.027

0.023

6.436

5

0.754

0.723−0.775

0.025

4.93

4.60−5.20

0.843

0.125

0.101

0.095

0.086

5.449

6

0.884

0.781−1.006

0.014

0.33

0.30−0.40

0.843

0.030

0.024

0.013

0.011

0.534

7

1.190

0.950−1.603

0.011

0.87

0.50−1.10

0.843

0.081

0.070

0.060

0.056

1.682

8

0.830

0.735−0.947

0.015

0.013− 0.015 0.024− 0.026 0.012− 0.016 0.008− 0.013 0.013− 0.017

2.00

1.80−2.10

0.843

0.084

0.064

0.054

0.046

3.736

between 30% and 50% of TP FWMC as calculated from the in situ analysers. The larger frit size group (100 to 160 μm) appeared to give the least variation returning an r2 of 0.89 on the comparisons of FWMC but with a recovery of 40% (part c of Figure 3). Comparison with TRP FWMC from the bankside analysers (parts d−f of Figure 3), the passive samplers indicated a more complete recovery (69% for the 40−100 μm frit sizes) although the scatter in the relationship was too great to be useful (part e of Figure 3). With NO3−N, the relationship between FWMC indicated underestimation of the FWMC across the three frit size ranges at concentrations up to approximately 2.5 mg l−1. Above this concentration, recoveries were much higher than the FWMC and highly variable. The overestimated concentrations correspond to three deployments, two in Site 4 and one in Site 5 (Figure 1). One of these passive sampler overestimates was known to be due to laboratory error. However, the others are inexplicable and are well beyond the TON/NO3−N concentration ranges observed before or after these deployments. The 20 μm frit size range (part g of Figure 3) was most consistent, with 148% recovery over the full range of concentrations and least underestimation up to 2.5 mg l−1 (43% recovery, r2 = 0.5). Figure 4 shows the relationship between the total discharge (Q) and estimated sampler throughflow based on salt loss for catchments with multiple deployments. Estimated cartridge throughflow is approximately 9 orders of magnitude lower than the river discharge over the deployments, which in part reflects the small fraction of the channel cross-sectional area occupied by the sampler. The variation in throughflow estimates for the deployments in each catchment is considerable and particularly poor in the river at site 2 (Figure 1) with an almost inverse relationship between Q and cartridge throughflow and a high degree of estimated throughflow for individual deployments. Laboratory pump test data, with mean recoveries of triplicate cartridges and ambient river water P fractions, NO3−N, and throughflow volumes, are shown in Table 3. At the time of sampling, the five rivers ranged in TP concentration from 0.030 mg l−1 to 0.125 mg l−1 and in NO3−N concentration from 0.534 mg l−1 to 6.436 mg l−1. During the laboratory pumping test, a total volume of 0.843 l was passed through the cartridges at a constant flow rate (35.125 mL hr−1) during a 24 h period. ANOVA (Table 3) indicated that there was a statistically significant difference between passive sampler P recoveries and river water concentrations for TP, total soluble phosphorus (TSP) and TRP; differences between soluble reactive phosphorus (SRP), NO3−N and throughflow recovery were

Table 3. One-Way ANOVA Results Testing for Differences between Cartridge Recovery and Laboratory Concentrations for Ambient River Water Samplesa passive sampler recovery

laboratory concentration

df

MS

F

P

P P P P NO3−N throughflow

TP TSP TRP SRP NO3−N throughflow

1,8 1,8 1,8 1,8 1,8 1,8

0.902 0.671 0.452 0.339 0.087 0.003

19.293 15.208 6.878 4.816 0.386 1.190

0.002c 0.005c 0.031b 0.060 0.552 0.307

a

df = degrees of freedom; MS = mean square; F = F statistic; P = probability. bP < 0.05. cP < 0.01.

not significant. Of note is the relationship between cartridge P recovery and SRP (P = 0.06), which was only just not significant at the 95% level. This is supported by measurements on river water expelled from the cartridges that were subsequently found to have SRP concentrations lower than the limits of detection, yet were not fully recovered in the elution step.



DISCUSSION Flow-proportional passive sampling of nutrients in rivers remains an attractive option for routine river monitoring campaigns; optimizing the under-representation of grab sampling and avoiding the technological constraints and expense of high-resolution sampling.13,30 Results from this study, which used high-resolution data sets to compare with passive sampler deployments were, however, variable with regard to nutrient recovery. The approach to providing a FWMC for nutrients was based on two key assumptions. These were, first, that water would enter and exit cartridges unimpeded and deposit nutrient ions at a rate proportional to streamflow, and, second, that the throughflow of water was accurately approximated by the loss of tracer salt. In terms of the first assumption, hydraulic fluid simulations (not part of the original experimental design but undertaken as part of quality control checking and included as Figure E of the Supporting Information) demonstrated that flow would tend to diverge round the cartridges. There are few published studies relating to open channel flow around porous media due to the complexity of both simulating flow numerically with a free surface and the difficulties in measuring in experimental set ups. The most relevant investigated the influence of porosity and Reynolds number on flow within and around a permeable 2335

dx.doi.org/10.1021/es304108e | Environ. Sci. Technol. 2013, 47, 2331−2338

Environmental Science & Technology

Article

part of the channel during low flows. At high flows, however, the transverse velocity profile may alter and the zone of highest velocity may be at a different point in the channel cross-section. This may account for the poor relationship between Q and cartridge throughflow for Monaghan (site 2). In terms of the second assumption, loss of cartridge throughflow potential could also occur due to hydraulic resistance and/or the accumulation of algae (biofouling) or river particulate material at certain times of year (Figures F and G of the Supporting Information). High-flow storm events also tend to transport stream debris, and this was particularly prevalent during autumn due to the transportation of allochthonous organic matter. Variable responses were observed during the 12 deployments with some influenced by low flows, where flow velocities were at the threshold (0.2 m s−1) for operation of the cartridge system and others by storm events. Nevertheless, throughflow estimates using the calcium citrate dissolution method were found to be a satisfactory proxy during laboratory pump tests with known volumes recovered in blind trials. Again, the ambiguity with field trials perhaps points to conditions in ambient river deployments impacting cartridge throughflow as highlighted above. This is particularly true where large flow gradients exist, despite some reasonable recoveries reported in deployments of porous media and in other flow environments.28−30 In this case, however, P cartridge recoveries in general showed greater similarity to the time-weighted mean concentrations (comparisons not shown), particularly for TRP across the frit sizes trialled (79% recovery for 20 μm; 127% for 40−100 μm and 103% for 100−160 μm) with the 100−160 μm frit sizes showing least variation (r2 = 0.85). The precise nature of the P fractions recovered by the cartridges also remains unclear. Whereas the comparisons were made on TP and TRP, the most consistent cartridge recoveries were, on average, only approximately 40% of the TP measured during the 12 deployments. It is reasonable to assume that the porous resin media would act as a collection point for P in the soluble phase as well as desorbed P from the particulate phase. However, as might be expected from resin based media, the TRP data available in this study indicated that the main fraction recovered was likely to be unrelated to a largely particulate fraction. The laboratory pump trials sought to establish cartridge P and N recoveries and throughflow efficiencies under controlled conditions. Results found lower than expected P recoveries in the trials with ambient water despite accurate throughflow. These tests measured very low level SRP concentrations exiting the cartridges, reported as below the limit of detection, thus indicating the efficient adsorption of inorganic soluble P in the resin media. Despite this, recovery of eluted P was only 16− 53% of the actual P pumped through the cartridges. It remains unclear as to why, under controlled laboratory conditions, pumping test P recoveries were low (i.e., there were reasonable NO3−N recoveries) but suggests that other reactive species (in the total reactive P pool) were not eluted. Importantly, these trials suggested that efficient retention of (a proportion) of the total P mass had occurred and was not restricted to the soluble P mass, part of which also seemed to be retained following elution in the limited laboratory trials using a standard PO4 solution (Figure D of the Supporting Information). Implications. In this study, multiple factors have been highlighted regarding the performance of the passive sampler cartridges that appear to have operational issues in ambient

circular cylinder.36 As simulated velocities increased (higher Reynolds numbers) the velocities in the porous cylinder decreased, due to the effect of recirculating wakes, which developed downstream of the cylinder. The effect was most pronounced for lower porosities. Although it would be highly speculative to extrapolate these results to the sampler cartridges without extensive experimental validation (which was beyond the scope of this field assessment of the technology), it does suggest that it is risky to assume that the throughflow in the cartridge increases linearly with stream velocity. Analysis of discharge percentiles and the Q5:Q95 flashiness metric during deployment periods did appear to indicate some control on the recovery of nutrients (passive sampler FWMC expressed as a percentage of bankside analyzer FWMC) with less recovery in flashier deployments (Figure 5). Also, from these data, periods of very low dynamic discharge (i.e., Q5:Q95 < 10) appeared to show variable recoveries but tending to overestimates.

Figure 5. Comparison of high 5th percentile and low 95th percentile discharges (Q5:Q95 ratios) during the deployment periods with the percentage difference in nutrient recovery. The discharge ratios, a metric of runoff flashiness, appear to show some control on how much nutrient is recovered compared to FWMC from bankside analysis.

Furthermore, the approach assumes that the velocity at the location in which the cartridge is installed remains proportional to Q for all flow magnitudes. In the stream, as stage height increases, the position of the cartridge remains constant while the location of the maximum flow velocity in the vertical profile will change. In addition, the most hydraulically active part of a natural channel is likely to vary with increasing flow (depending on bed and channel morphology both at and upstream of the deployment site). Cartridges were installed to be in the fastest 2336

dx.doi.org/10.1021/es304108e | Environ. Sci. Technol. 2013, 47, 2331−2338

Environmental Science & Technology

Article

(4) Jordan, P.; Arnscheidt, J.; McGrogan, H.; McCormack, S. Characterising phosphorus transfers in rural catchments using a continuous bank-side analyser. Hydrol. Earth Syst. Sci. Discuss. 2007, 11, 372−381. (5) Palmer-Felgate, E. J.; Mortimer, R. J. G.; Krom, M. D.; Jarvie, H. P. Impact of point-source pollution on phosphorus and nitrogen cycling in stream-bed sediments. Environ. Sci. Technol. 2010, 44, 908− 914. (6) Lennox, S. D.; et al. Estimating the contribution from agriculture to the phosphorus load in surface water. In Phosphorus Loss from Soil to Water; Tunney, H., Carton, O. T., Bookes, P. C., Johnston, A. E., Eds.; CAB International; Wallingford, 1997; pp 55−75. (7) Facchi, A.; Gandolfi, C.; Whelan, M. J. A comparison of river water quality sampling methodologies under highly variable load conditions. Chemosphere 2007, 66, 746−756. (8) Douglas, R. W.; Menary, W.; Jordan, P. Phosphorus and sediment transfers in a grassland river catchment. Nutr. Cycl. Agroecosys. 2007, 77, 199−212. (9) Fealy, R. M.; Buckley, C.; Mechan, S.; Melland, A.; Mellander, P. E.; Shortle, G.; Wall, D.; Jordan, P. The Irish Agricultural Catchments Programme: catchment selection using spatial multi-criteria decision analysis. Soil Use Manage 2010, 26, 225−236. (10) Mellander, P.-E.; Jordan, P.; Wall, D. P.; Melland, A. R.; Meehan, R.; Kelly, C.; Shortle, G. Delivery and impact bypass in a karst aquifer with high phosphorus source and pathway potential. Water Res. 2012, 46, 2225−2236. (11) Melland, A. R.; Mellander, P. E.; Murphy, P. N. C.; Wall, D. P.; Mechan, S.; Shine, O.; Shortle, G.; Jordan, P. Stream water quality in intensive cereal cropping catchments with regulated nutrient management. Environ. Sci. Policy 2012, 24, 58−70. (12) Jordan, P.; Melland, A. R.; Mellander, P.-E.; Shortle, G.; Wall, D. The seasonality of phosphorus transfers from land to water: Implications for trophic impacts and policy evaluation. Sci. Total Environ. 2012, 434, 101−109. (13) Wade, A. J.; Palmer-Felgate, E. J.; Halliday, S. J.; Skeffington, R. A.; Loewenthal, M.; Jarvie, H. P.; Bowes, M. J.; Greenway, G. M.; Haswell, S. J.; Bell, I. M.; Joly, E.; Fallatah, A.; Neal, C.; Williams, R. J.; Gozzard, E.; Newman, J. R. Hydrochemical processes in lowland rivers: insights from in situ, high-resolution monitoring. Hydrol. Earth Syst. Sci. 2012, 16, 4323−4342. (14) Gorecki, T.; Namiesnik, J. Passive sampling. Trends Anal. Chem. 2002, 21, 276−291. (15) Namiesnik, J.; Zabiegala, B.; Kot-Wasik, A.; Partyka, M.; Wasik, A. Passive sampling and/or extraction techniques in environmental analysis: a review. Anal. Bioanal. Chem. 2005, 381, 279−301. (16) Zhang, H.; Davison, W.; Gadi, R.; Kobayashi, T. In situ measurement of dissolved phosphorus in natural waters using DGT. Anal. Chim. Acta 1998, 370, 29−38. (17) Pichette, C.; Zhang, H.; Davison, W.; Sauve, S. Preventing biofilm development on DGT devices using metals and antibiotics. Talanta 2007, 72, 716−722. (18) Pichette, C.; Zhang, H.; Sauve, S. Using diffusive gradients in thin-films for in situ monitoring of dissolved phosphate emissions from freshwater aquaculture. Aquaculture 2009, 286, 198−202. (19) Dahlqvist, R.; Zhang, H.; Ingri, J.; Davison, W. Performance of the diffusive gradients in thin films technique for measuring Ca and Mg in freshwater. Anal. Chim. Acta 2002, 460, 247−256. (20) Zhang, H.; Davison, W.; Knight, B.; McGrath, S. In situ measurements of solution concentrations and fluxes of trace sulphate in soils using DGT. Environ. Sci. Technol. 1998, 32, 704−10. (21) Harper, M. P.; Davison, W.; Zhang, H.; Tych, W. Kinetics of metal exchange between solids and solutions in sediments and soils interpreted from DGT measured fluxes. Geochim. Cosmochim. Acta 1998, 62, 2757−2770. (22) Zhang, H.; Davison, W.; Mortimer, R. J. G.; Krom, M. D.; Hayes, P. J.; Davies, I. M. Localised remobilization of metals in marine sediment. Sci. Total Environ. 2002, 296, 175−187.

river environments as compared with high-resolution data sets. These interacting factors were: P fraction speciation and proportions in the eluted liquid; inconsistent N recoveries in river environments; possible flow divergence leading to loss of flow/velocity-proportional nutrient adsorption and loss of throughflow; potential changing vertical and cross-sectional river velocity profiles; physical and biological interference (debris or algal growth); laboratory elution and recovery procedures for nutrients in ambient river (and possibly laboratory) samples. Discharge flashiness was highlighted as a hindrance to good recoveries and this could be avoided in less dynamic rivers at this scale, or where this is not as influential at larger river catchment scales. However, the data also indicated that rivers with very low dynamic discharges may also cause problems with recovery during deployments. Further investigation into these factors would allow the effectiveness of passive samplers to be quantified in the context of river systems and informed adaptations made to improve function, ensure unimpeded throughflow and limit biofouling. At present, national monitoring programmes fail to capture diffuse nutrient transfers during storm events and, as a result, the interpretation of these data for monitoring the recovery of (agricultural) nutrient pressures is open to question. Whereas successful passive sampling may increase data richness by integrating concentrations over deployment periods, the issues associated with the passive samplers in this study were found in catchments where diffuse storm-associated nutrient risk is high and where the need for monitoring is greatest.



ASSOCIATED CONTENT

S Supporting Information *

Sampler design, deployments, and deployment issues are provided. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank and acknowledge contributions from science and technical staff at the University of Ulster (Coleraine), Dundalk Institute of Technology (Dundalk), University of Leeds (Leeds), and the Agricultural Catchments Programme, Teagasc, Wexford. This research was funded by the Environmental Protection Agency (Wexford) under STRIVE (2007-2013) funding (2000-ET-MS-5-S2).



REFERENCES

(1) Johnes, P. J. Uncertainties in annual riverine phosphorus load estimation: Impact of load estimation methodology, sampling frequency, baseflow index and catchment population density. J. Hydrol. 2007, 332, 241−258. (2) Foy, R. H. Variation in the reactive phosphorus concentrations in rivers of northwest Europe with respect to their potential to cause eutrophication. Soil Use Manage. 2007, 23, 195−204. (3) Cassidy, R.; Jordan, P. Limitations of instantaneous water quality sampling in surface-water catchments: Comparison with nearcontinuous phosphorus time-series data. J. Hydrol. 2011, 405, 182− 193. 2337

dx.doi.org/10.1021/es304108e | Environ. Sci. Technol. 2013, 47, 2331−2338

Environmental Science & Technology

Article

(23) Fones, G. R.; Davison, W.; Hamilton-Taylor, J. The fine-scale remobilization of metals in the surface sediment of the North-East Atlantic. Cont. Shelf Res. 2004, 24, 1485−504. (24) Leermakers, M.; Gao, Y.; Gabelle, C.; Lojen, S.; Ouddane, B.; Wartel, M. Determination of high resolution pore water profiles of trace metals in sediments of the Rupel River (Belgium) using DET and DGT techniques. Water Air Soil Poll. 2005, 166, 265−286. (25) Gao, Y.; Leermakers, M.; Gabelle, C.; Divis, P.; Billon, G.; Ouddane, B. High-resolution profiles of trace metals in the pore water of riverine sediment assessed by DET and DGT. Sci. Total Environ. 2006, 362, 266−277. (26) Mortimer, R. J. G.; Davey, J. T.; Krom, M. D.; Watson, P. G.; Frickers, P. E.; Clifton, R. C. The effect of macrofauna on pore-water profiles and nutrient fluxes in the intertidal zone of the Humber estuary. Estuar. Coastal Shelf Sci. 1999, 48, 683−699. (27) O’Brien, D.; Hawker, D.; Shaw, M.; Mueller, J. F. The performance of passive flow monitors and phosphate accumulating passive samplers when exposed to pulses in external water flow rate and/or external phosphate concentrations. Environ. Pollut. 2011, 159, 1435−1441. (28) de Jonge, H.; Rothenberg, G. New device and method for fluxproportional sampling of mobile solutes in soil and groundwater. Environ. Sci. Technol. 2005, 39, 274−282. (29) Rozemeijer, J.; Van Der Velde, Y. P. E.; de Jonge, H.; Van Geer, F.; Broers, H. P.; Bierkens, M. Application and evaluation of a new passive sampler for measuring average solute concentrations in a catchment scale water quality monitoring study. Environ. Sci. Technol. 2010, 44, 1353−1359. (30) de Jonge, H.; New sampling method for monitoring of N, P in surface water. In Diffuse Phosphorus Loss Risk Assessment, Mitigation Options and Ecological Effects in River Basins; Heckrath, G., Rubæk, G. H., Kronvang, B., Eds.; The 5th International Phosphorus Workshop (IPW5) − Proceedings; Aarhus University, 2007; pp 87−90. (31) Haygarth, P. M.; Sharpley, A. N. Terminology for Phosphorus Transfer. J. Environ. Qual. 2000, 29, 10−15. (32) Murphy, J.; Riley, J. P. A modified single solution method for the determination of phosphate in natural waters. Anal. Chim. Acta 1962, 27, 31−6. (33) Eisenreich, S. J.; Bannerman, R. T.; Armstrong, D. E. A simplified phosphorus analysis technique. Environ. Lett. 1962, 9, 43− 53. (34) Macintosh, K. A.; Cassidy, R.; Jordan, P.; Arnscheidt, J. A.; McCarthy, V.; Jordan, S.; Linnane, S. Testing a New Technology for Monitoring Nutrients in Rivers (STRIVE Report Series No. 81); Environmental Protection Agency,Wexford, Republic of Ireland, 2011. (35) Macintosh, K. A.; Jordan, P.; Cassidy, R.; Arnscheidt, J. A.; Ward, C. Low flow water quality in rivers; septic tank systems and high-resolution phosphorus signals. Sci. Total Environ. 2011, 412, 58− 65. (36) Yu, O.; Zeng, Y.; Lee, T. S.; Chen, X. B.; Low, H. T. Steady flow around and through a permeable circular cylinder. Comput. Fluids 2011, 42, 1−12.

2338

dx.doi.org/10.1021/es304108e | Environ. Sci. Technol. 2013, 47, 2331−2338