A comparison of rain-related phosphorus and nitrogen loading from urban, wetland, and agricultural sources

Comprehensive watershed studies have been conducted for two lakes located in the Lake Michigan drainage system. Studies were conducted from March through October of 1979. During that interval, large differences in storm-related nutrient loading were measured from urban, wetland, and agricultural sources. Eliminating runoff due to melt of the snow pack, it was found that rain-related discharge from the urban area studied was 0.578 kg total-P and 3.688 kg total-N ha⁻¹ of watershed. Rain induced runoff from marshes in the same drainage basin transported 0.023 kg total-P and 0.585 kg total-N ha⁻¹ of catchment. Rainfall of approximately the same amount caused runoff from agricultural land of 0.180 kg total-P and 5.965 kg total-N ha⁻¹. Algae of both lakes were phosphorus limited; nitrogen was present in excess. Using constants from Nichols-Dillon relationships in the literature regarding phosphorus, phytoplankton biomass, and secchi disc transparencies, the urban input of phosphorus ha⁻¹ of drainage was sufficient to bring 0.96 ha-m of lake water to undesirable algal bloom status. Similarly, march input ha⁻¹ would bring an estimated 0.04 ha-m into bloom. By the same calculation, storm-related agricultural runoff would result in 0.30 ha-m of lake water becoming undesirably rich in algae. Knowing the number of hectares in these types of catchment and the volume available in a particular lake for phytoplankton production, decisions regarding cost-effective treatment of rain-related discharge can be made.     

Factors affecting nutrient loads in some Iowa streams

The export and concentration of inorganic nitrogen and total phosphorus from 34 watersheds in a northwestern Iowa lake district were measured during March 1971–August 1973. Annual nutrient losses were approximately 0.35 kg ha⁻¹ P. 6.7 kg ha⁻¹ NO₃-N, and 1.0 kg ha⁻¹ NH₃-N. A statistical analysis of the relationship between land-use and plant nutrients was used to determine differences among streams. Animal units in feedlots were significantly correlated with phosphorus and ammonia nitrogen (mg l⁻¹ and kg ha⁻¹ yr⁻¹). Nitrate nitrogen was negatively correlated with the percentage of watershed in marshland. Tile drainage and surface runoff from grasslands, feedlots, cornfields, and soybean fields were analyzed for nitrogen and phosphorus in spring 1974: mean values are given.     

The behaviour of phosphate, nitrate, chloride and hardness in twelve welsh rivers

Selected water quality data from 12 rivers in the area administered by the Welsh Water Authority were analysed for the period 1974–1981. Mean nitrate-nitrogen concentrations varied from 0.4 to 3.7 mg l⁻¹ and were significantly related to the intensity of average catchment run-off; mean orthophosphate-phosphorus concentrations ranged from the limit of analytical detection to 0.730 mg l⁻²; chloride from 11 to 42 mg l⁻¹ and total hardness (as CaCO₃) from 13 to 173 mg l⁻¹. Seasonal patterns of change in concentration were established, generally for all determinands at most sites, but no long-term trends were detected. Relationships between concentration and flow were established for most determinands at many sites, increasing flow generally resulting in decreased concentration. However, positive relationships between nitrate concentration and flow were established at seven sites. Mass flows (kg ha⁻¹ yr⁻¹) were calculated at nine sites only: nitrate-N 4.8–24.6; orthophosphate-P 0.16–3.81; chloride 79–334; total hardness (as CaCO₃) 196–1629. Orthophosphate flows were related to sewered population density, estimates of per capita and land drainage contributions being 1.9 g day⁻¹ and 0.112 kg ha⁻¹ yr⁻¹ respectively.     

The effects of geology and land use on the export of phosphorus from watersheds

The export of total phosphorus from 34 watersheds in Southern Ontario was measured over a 20-month period. The annual average export for igneous watersheds (i.e. those of the Canadian Shield) that were forested was 4.8 mg m⁻² yr⁻¹, significantly different from the average (11.0 mg m⁻² yr⁻¹) for watersheds that included pasture as well as forest. Similarly, on sedimentary rock, the mean export from forested watersheds (10.7 mg m⁻² yr⁻¹) differed significantly from those with forest and pasture (28.8 mg m⁻² yr⁻¹). The differences between watersheds of different geology but similar land use were also highly significant.Additional data from the literature supported our conclusions. Other forested igneous watersheds of plutonic origin averaged 4.2 mg m⁻² yr⁻¹ of total phosphorus exported: forested igneous watersheds of volcanic origin, however, averaged 72 mg m⁻² yr⁻¹. The overall average export from each type of watershed as classified by geology and land use was very similar to that for the same classification found in our study.The effects of agriculture and urbanization were to greatly increase the total phosphorus exported. Wide ranges of values probably reflect the intensity of land use.     

The role of woody vegetation in preventing ground water pollution by nitrogen from septic tank leachate

The ability of woody vegetation to remove nitrogen from septic tank leachate was studied in pine upland, oak upland and hardwood wetland habitats of the New Jersey Pinelands. The study was stimulated by the incorporation of a term for plant uptake in nutrient dilution models used for Pinelands land-use management decisions. Plant response was studied at sites involving septic tank leach fields and matched control sites for each habitat type. At each site, total biomass, net production, and tissue N concentrations for the dominant species was determined. The hardwood wetland habitat had a total biomass of 15.9 MT ha⁻¹, a net primary production of 5.4 MT ha⁻¹yr⁻¹ and a net N uptake 75–80 kg N ha⁻¹yr⁻¹. Tissue N values for wetland trees and shrubs did not show significant differences between control and experimental sites. The pine upland communities had a biomass of 55 MT ha⁻¹ and a net production of 5.7 MT ha⁻¹yr⁻¹; net N uptake ranged from 45 kg N ha⁻¹yr⁻¹ (control sites) to 56 kg N ha⁻¹yr⁻¹, (experimental sites). The oak upland communities had a biomass of 59 MT ha⁻¹ and a net primary production of 5.0 MT ha⁻¹yr⁻¹; net uptake ranged from 55 kg N ha⁻¹yr⁻¹ in the control sites to 69.3 kg N ha⁻¹yr⁻¹. Tissue N concentrations showed significant increases for tree but not shrub species in both upland habitats. The capacity of the upland woody plants to increase N uptake and storage appears to be related to rooting depth and to the proximity of the plants to the drain tile: only plants with deep taproots, growing close to (within 1 m) the trench showed significant increases in uptake. It is suggested that only by augmenting appropriate natural vegetation with supplementary tree plantings can upland vegetation be utilized to reduce nitrogen movement to groundwater from septic systems in sandy soil; vegetation at naturally-occurring densities will not have a significant effect on water quality.     

The contribution of mercury and chlorinated organics from urban runoff

The concentration of mercury and chlorinated organics in the combined sewer overflows of Rochester, New York (U.S.A.) have been measured. The resulting annual loadings of 0.034 kg ha⁻¹ (0.034 pounds acre⁻¹) and 0.011 kg ha⁻¹ (0.01 pounds acre⁻¹) of mercury and chlorinated organics measured in urban runoff are significant. A high degree of correlation exists between the measured concentration of the toxicants when compared with rainfall intensity, rainfall duration, antecedent number of dry days and the total suspended solids concentration characteristic of the combined wastwater.     

The quality of run-off from an irrigated market garden area in Brazil

This paper presents the results of a study carried out in Brazil to determine the effects of irrigation and fertiliser use on the quality characteristics of run-off from an agricultural area used for intensive cultivation of market garden produce.The results show that during dry weather the discharge of nutrients from the cultivated area when irrigation is taking place is equivalent to 7 kg ha⁻¹ yr⁻¹ N and 1 kg ha⁻¹ yr⁻¹ P. Considerable accumulation of nutrients took place in the soil during the growing season, over the 129-day survey period the nutrient content of the top soil increased by approximately 590 g m⁻²N and P and 780 g m⁻²K. This accumulation was derived from fertilisers which were applied at rates of 1136 kg ha⁻¹N, 978 P and 1367 K. During wet weather the nutrients which had accumulated in the soil were washed out.     

Nutrient budget for a hypertrophic reservoir

A nutrient budget for the shallow, hypertrophic Ardleigh Reservoir, a pumped storage scheme in eastern England, is described for the period 1979–1982. Algal succession in the reservoir was typical of eutrophic waters, with maximum chlorophyll-a of 98 mg m⁻³. Although the reservoir did not stratify thermally, the concentrations of SRP, Mn and Fe increased in bottom waters during summer. The weight ratio of inorganic N to inorganic P ranged from 720 to 5. On average, SRP represented 72% of the total P content of the reservoir.Some 44% of water input was of pumped river water, 48% being of direct catchment flow. The specific loading of SRP was 5.014 g m⁻² yr⁻¹.Ninety per cent of the annual SRP load was derived from pumped water and 60% of the SRP load was retained in the reservoir. Nitrate input was more diffuse, with approx. 33% from pumped water and 66% from catchment flow. A net release of P from the sediment of 23 mg P m⁻² day⁻¹ was recorded in summer, equivalent to 33% of annual mean external SRP loading. Strategies of P control are discussed in relation to loading models.     

Coliform bacteria from diffuse sources as a factor in estuarine pollution

The contribution of 849 ha of rural watershed to the fecal coliform (FC) pollution of the Rhode River, a subestuary of Chesapeake Bay with a surface area of 485 ha. was estimated. The watershed with an animal population of 0.6 animal unit ha⁻¹ discharged between 7.5 × 10⁶ and 669 × 10⁶ FC ha-day⁻¹. The FC discharge rate was seasonal and largely dependent upon the water flow. Total coliform (TC) discharge was influenced by the same factors as the FC discharge. It was calculated that on the average less than 1% of FC produced by the animals on the land was washed down by water runoff. Occasionally the FC discharge rate in the runoff reached 4–6% of the FC produced by the animals. Fecal coliforms persisted in the water. The FC numbers were high in the Rhode River close to the discharge points and further away were diluted by the river volume. It was estimated, using the maximum number of FC in the runoff, that 2600 m³ of well mixed receiving water was needed for every ha of watershed area not to exceed the safe water standards (14 FC MPN 100 ml⁻¹ for shellfish harvesting. Three factors having a role in FC pollution of an estuarine ecosystem were emphasized as a result of this study: (1) the rural watershed contributed substantial quantities of FC to the estuary; (2) the season of the year may determine the level of pollution entering from rural sources into the estuary; (3) the persistence of bacteria in the estuary may increase the pollution level contributed by the watershed especially at low water temperatures.     

The feasibility of using disused water treatment filter beds for rearing carp (Cyprinus Carpio L.) in the U.K.

The feasibility of rearing carp (Cyprinus carpio) at a disused water treatment works in East London (U.K.) was investigated. The 10 ha of filter beds represented an atypical site for carp culture; in order to assess whether in-house aquaculture would enable a Water Authority to gain advantage in cost, availability and disease-free status of stock, a series of trials was undertaken. These enabled comparison of the weight of carp produced in relation to: extensive cultivation, artificial feeding, a range of stock densities and a range of ages and sizes. The filter beds were unsuitable for natural spawning, early fry production and the rearing of first year fish (0⁺ age group). However, growing-on of subsequent stages (i.e. 0⁺ − 1⁺ or 1⁺ − 2⁺) gave good survival (60–80%) and growth rates. Fish stocked at 20 g in April/May reached 200–400 g by September and 1 kg plus, 1 yr later. Production levels of 300–350 kg ha⁻¹ yr⁻¹ without artificial foods (i.e. extensive cultivation) and up to 2.2 tonnes ha⁻¹ yr⁻¹ with supplementary feeding were comparable with mainland European fish farms. These high values reflect the eutrophic nature of this site and successful exploitation of the available secondary production. It was possible to harvest 15,000–20,000 carp yr⁻¹ for restocking. Carp rearing integrated with existing operations and activities on site (e.g. pipe and equipment stores) and represents a novel, alternative use of a semi-redundant site. Cost benefit assessment showed that fish production was economically viable returning nominal profit margins of between £1.10 and £2.06 kg⁻¹ of carp harvested. The resultant cost/benefit ratios per kilogram harvested were 1:1.9 for extensive production (without feeding) and 1:1.34–1.43 when intensive cultivation employing artificial foods was used.     

Nitrate behaviour in streamflow from a grassland catchment in Devon, U.K.

NO₃---N levels and loadings in the River Dart, Devon, U.K. have been intensively monitored during the study period February 1975–January 1983, and NO₃---N behaviour in streamflow is compared with that of Mg which has predominantly contrasting origins and controls in runoff. A strongly developed annual regime characterises daily mean NO₃---N concentrations which peak at 4.5 mg l⁻¹ in December and reach a minimum of 1.1 mg l⁻¹ in early autumn, and this regime is reflected in a positive relationship between NO₃---N concentration and discharge. Storm-period NO₃---N behaviour also exhibits a marked seasonal variation, and “dilution” responses are typical of the winter period whereas a “concentration” effect is more characteristic of summer months and is attributed to the influence of soil throughflow. Spring and autumn are associated with a transition in storm-period behaviour, and the reaction of NO₃---N concentrations during individual events or over a sequence of storms at any time of the year may also vary in response to a complex interplay of factors.NO₃---N transport is more strongly biased towards storm events than is Mg, and more than 50% of the total NO₃---N load was removed in only 10% of the study period. The mean annual NO₃---N load for the study basin (24 kg ha⁻¹) is higher than that recorded in many grassland catchments of the U.K. but does not reflect an unusually high rate of fertiliser application. Unlike several other arable and grassland drainage basins studied in Britain, there is no evidence in the River Dart of a significant upward trend in NO₃---N concentrations and loads since 1975.     

A comparison of nutrient sources and behaviour in two lowland subcatchments of the River Wye

Seasonal changes in concentrations and mass flow of dissolved silicon, orthophosphate and inorganic nitrogen were studied during 1978 in two lowland catchments of the R. Wye, the Frome and the Trothy, which are of similar drainage area but are subjected to different land use. In addition more intensive investigation of water quality changes were undertaken during storm events.Seasonal patterns of silicon concentrations were similar in the two rivers with major depletions in May and annual losses to diatoms were estimated to be equivalent to 5–12% of annual mass flow. Peak concentrations of orthophosphate were generally recorded during periods of low river flow in both rivers but concentrations were substantially greater in the Frome (mean, 0.42 mg l⁻¹) than the Trothy (mean, 0.19 mg l⁻¹), reflecting differences in population density. The principal form of inorganic nitrogen in both catchments was nitrate, but mean concentrations were considerably greater in the Frome (mean NO₃-N. 4.99 mg l⁻¹) than in the Trothy (mean NO₃-N. 2.93 mg l⁻¹) and seasonal variations were markedly different. Concentrations in the Frome increased during the summer months (maximum, 7 mg l⁻¹ whilst concentrations in the Trothy decreased during this period and this was reflected in positive and negative relationships between concentration and river flow respectively. Such differences in behaviour cannot be attributed to contributions from sewage and it is suggested that they may reflect differences in the proportion of groundwater contributing to river flow during the summer period and the nitrate content of that groundwater.Intensive sampling during storm events indicated that, overall, orthophosphate concentrations decreased and silicon and nitrate concentrations increased with increased flow, but there was considerable variation in solute behaviour during different events.     

Phosphorus adsorption to riverine suspended matter: Implications for the P-budget of Lake Constance

Dissolved phosphorus from Lake Constance water was adsorbed by particulate matter from the Alpenrhein whenever the equilibrium phosphorus concentration of 0.85 μmol l⁻¹ was exceeded. When shaken in phosphate-free lake water the solids liberated small amounts of P (1% of the total-P). The maximum adsorption capacity of the particles was estimated to be 2.8 μmol g⁻¹ (dry substance). Only part of the phosphorus once experimentally sorbed to the solids was shown to desorbe in P-free lake water.During 1981 1.47·10⁹ kg of particulate matter was discharged into Lake Constance from the Alpenrhein. By adsorption these sedimenting materials had the potential to remove 18–25·10³ kg of dissolved phosphorus from the water column, thus comprising almost 2% of the total P-load in 1981.     

Lake Sammamish response to wastewater diversion and increasing urban runoff

Lake Sammamish has shown a decrease in its mean annual concentration of phosphorus following diversion of about one third of the external loading in 1968. During 1971–1975 the P concentration averaged 27 μg l⁻¹, in contrast to the prediversion (1964–1966) concentration of 33 μg l⁻¹, and may be equilibrating near the predicted steady state concentration of 22 μg l⁻¹. Neither phytoplankton biomass or Secchi visibility has changed following diversion, however the blue green component of the phytoplankton decreased by nearly 50%. The failure of biomass and visibility to improve is probably a result of similar pre- and postdiversion winter spring epilimnetic P concentrations. The marked reduction in P since diversion occurred during and prior to fall overturn and may have represented a supply of P for later summer early fall blue green algal populations that declined after diversion.Runoff from a rapidly developing westside portion (18%) of the watershed is contributing substantially to P loading of the lake. Development to a density of about ten dwellings ha⁻¹ has increased loading possibly on the order of 14%. Future development of the eastside portion (26% of watershed) may increase loading by 20% and be equivalent to nearly one half of the P previously diverted in 1968.     

The effect of sewage sludge land disposal on the microbiological quality of groundwater

The long term effect of farmland disposal of anaerobically digested sewage sludges (alum, iron and lime) on the microbiological quality of groundwater was investigated using the lysimeter system. It was observed that after 4 years of heavy sludge application (5100 kg TKN ha⁻¹ 4 yr⁻¹), 92–98% of the sewage micoorganisms in the sludge had perished in the soils systems. Analyses of leachates and soils for coliforms and heterotrophic bacterial populations revealed that there was little possibility of microbiological contamination of groundwater by the practice of sludge farmland disposal, provided that the groundwater table was not too high and the soil was well-drained.     

Effects of water regime during rice-growing season on annual direct N2O emission in a paddy rice-winter wheat rotation system in southeast China

Annual paddy rice-winter wheat rotation constitutes one of the typical cropping systems in southeast China, in which various water regimes are currently practiced during the rice-growing season, including continuous flooding (F), flooding-midseason drainage-reflooding (F-D-F), and flooding-midseason drainage-reflooding and moisture but without waterlogging (F-D-F-M). We conducted a field experiment in a rice-winter wheat rotation system to gain an insight into the water regime-specific emission factors and background emissions of nitrous oxide (N₂O) over the whole annual cycle. While flooding led to an unpronounced N₂O emission during the rice-growing season, it incurred substantial N₂O emission during the following non-rice season. During the non-rice season, N₂O fluxes were, on average, 2.61 and 2.48 mg N₂O-N m⁻² day^− 1 for the 250 kg N ha^− 1 applied plots preceded by the F and F-D-F water regimes, which are 56% and 49% higher than those by the F-D-F-M water regime, respectively. For the annual rotation system experienced by continuous flooding during the rice-growing season, the relationship between N₂O emission and nitrogen input predicted the emission factor and background emission of N₂O to be 0.87% and 1.77 kg N₂O-N ha^− 1, respectively. For the plots experienced by the water regimes of F-D-F and F-D-F-M, the emission factors of N₂O averaged 0.97% and 0.85%, with background N₂O emissions of 2.00 kg N₂O-N ha^− 1 and 1.61 kg N₂O-N ha^− 1 for the annual rotation system, respectively. Annual direct N₂O-N emission was estimated to be 98.1 Gg yr^− 1 in Chinese rice-based cropping systems in the 1990s, consisting of 32.3 Gg during the rice-growing season and 65.8 Gg during the non-rice season, which accounts for 25-35% of the annual total emission from croplands in China.     

High-rate overland flow

Recommended loading rates for treating raw domestic wastewater by overland flow are 6.3–15 cm wk⁻¹. Information provided in the literature yields little insight regarding the upper range of hydraulic loading rates that could be effectively treated by overland flow. Therefore, field investigations were conducted to evaluate the performance of the overland flow system at overland flow rates from 0.95 m³ day⁻¹ m⁻¹ width of slope (13 cm wk⁻¹ to 4.15 m³ day⁻¹ m⁻¹ (57 cm wk⁻¹).Preliminary treated municipal wastewater was pumped to overland flow slopes, each approx. 3.7 m wide and 36.5 m long. The slope of each plot was 2.5%. The cover crop consisted of a mixture of ryegrass, bluegrass and fescue grass. The plots were operated for 2 years at six different hydraulic loading rates.Effluent BOD₅ concentration averages varied from 6 to 11 mg l⁻¹. The reduction of influent BOD₅ concentration ranged from 87 to 93%. Mean effluent suspended solids values were from 6 to 9 mg l⁻¹ with reductions of influent concentrations of 91–95%. Hydraulic application rate had little effect on percent BOD₅ or suspended solids removal.Total phosphorus reductions were minimal at all hydraulic application rates due to limited soil water contact.Ammonia concentration in the effluent ranged from 1 mg l⁻¹ NH₃-N at the 0.95 m³ day⁻¹ m⁻¹ (13 cm wk⁻¹) applied flow rate of 11.7 mg l⁻¹ NH₃-N at the 4.15 m³ day⁻¹ m⁻¹ (57 cm wk⁻¹) loading rate. Ammonia and nitrogen reductions decreased as the applied flow rate increased. Consequently, lower overland flow rates are necessary for nitrogen removal.The use of high-rate overland flow could potentially reduce the land necessary for this form of land application, if nutrient removal was not a local concern.     

Comparison of denitrification rate estimation techniques in a large, shallow lake

A comparison was made between various incubation techniques and a mass balance model for estimating annual denitrification rates in Lake Okeechobee, Florida, a large, shallow lake. Incubation techniques involving isolation of the sediment from nitrate and labile organic carbon sources tended to produce low rate estimates, while diagenetic models based on nitrate profiles tended to produce unrealistically high rate estimates. Each technique has inherent errors when used with frequently suspended sediments, but most incubation techniques were in general agreement with mass balance calculations. Denitrification rate estimates range from 0.5 to 1.3 g N m⁻² yr⁻¹, which represents 9–23% of the average annual nitrogen inputs to the lake.     

Analysis of long-term and seasonal river water quality changes in Latvia

An analysis of long-term historical records of the concentrations of major ions, TDS and nutrients for 20 river sites in Latvia is reported. Periods of water quality observations ranged from 15 to 43 years. A study of the quarterly adjusted time series showed that characteristic features of the data are non-normal distributions, seasonality, serial correlation and presence of significant trends that are mostly positive. The application of state-of-the-art software, based on non-parametric statistics such as the Seasonal Kendall slope estimator and the Seasonal Hodges-Lehmann estimator, made it possible to investigate these water quality records more accurately than other methods allow.Typical seasonal variations and concentration-discharge relationships were analyzed for different constituents. It was shown that fertilizer application and marsh land reclamation can cause widespread and intensive river water quality changes. Concentration increases of as much as 5–10 times that of background values were detected for NO₃⁻, Cl⁻, Na⁺ + K⁺ and SO₄²⁻. The main water quality changes took place in the 1960s and the early 1970s when fertilizer applications and reclamation works increased. After that, concentration increases for constituents other than NO₃⁻ and SO₄²⁻ were statistically insignficiant. The significant increases for NO₃⁻ and SO₄²⁻ were probably due to the additional impact of increased atmospheric deposition. The results of long-term changes of river loads entering the Baltic Sea and the Gulf of Riga from Latvian territory are examined.     

Upward trend in nitrate concentrations in rivers discharging into lough neagh for the period 1969–1979

The six major rivers discharging into Lough Neagh, Northern Ireland have been sampled at intervals ranging from 7 to 15 days during the study period 1969–1979. The mean concentration of nitrate has increased from 1.41 mg N 1⁻¹ in 1969–1970 to 2.41 mg N 1⁻¹ in 1978–1979. The maximum observed concentrations of nitrate have increased from 2.9 mg N 1⁻¹ in 1969–1970 to 7.5 mg N 1⁻¹ in 1979–1980. A number of possible causes of the increase in concentration have been considered and it was concluded that the greater intensity of grassland management as evidenced by an increase in fertiliser usage from 42 kg N ha⁻¹ in 1969 to 106 kg N ha⁻¹ in 1979 has made a significant contribution to the increase in nitrate. The increase in nitrate concentration may be the causal factor for the marked decline in nitrogen-fixing species of blue-green algae from the phytoplankton of Lough Neagh during the latter half of the study period.