# LINJ 1998 Surface Water

## Science Center Objects

In addition to the text that follows, there has been significant analysis of the water-quality network data and biologic data from LI and NJ. The major literature sources are as follows.

New Jersey

Concern with the availability of adequate water resources and potential degradation of these resources as a result of population growth in New Jersey dates back to the turn of the century (Vermeule, 1894), Although a network of stations for collection of streamflow data was established by 1921, comparatively little information on the quality of streams is available before 1945 (Parker et al., 1964). A summary of streamflow data collected 1921-1950 and existing water quality data is provided in Parker et al. 1964. In this report, water budgets, and rainfall runoff-ratios, and flood and drought frequencies were computed for 232 streams in the Delaware River basin and New Jersey for the years 1921-1950. Flow duration curves were compared among streams located in different New Jersey/Pennsylvania regions. For example, the Q90/Qm (the ratio of the 90th percentile discharge in cfs to the median discharge) was reportedly lower for coastal plain streams than for mountain streams. Natural water quality at selected sites is discussed in terms of dissolved solids and electrical conductivity, composition of dissolved solids, salinity, temperature, and suspended sediment. Pollution is measured as dissolved oxygen and biochemical oxygen demand. Relations between dissolved solid concentrations and discharge were compared among coastal plain and mountain streams. In Coastal Plain streams dissolved solids increase with discharge, whereas in mountain streams, dissolved solids decrease with discharge.

In 1962, the USGS and the State of New Jersey agreed to cooperate in a long-term program to assess the water quality of streams throughout the State. One of the first products was a report summarizing data collected by the State Department of Health and others since 1950 and a statewide water-quality reconnaissance conducted in 1962-63. The report describes the water-quality characteristics of streams in the five physiographic provinces of NJ based on specific conductance, dominant major ions, relations of dissolved solids to discharge, and sediment loads (Anderson and George, 1966).

Results show how specific conductance increases with discharge in coastal plain streams and decreases with discharge in mountain streams. In the Valley and Ridge physiographic province, calcium, magnesium, and bicarbonate are the dominant ions; in the Piedmont physiographic province calcium, magnesium, and sulfate are the dominant ions, whereas in the Coastal Plain, sodium, potassium, sulfate and chloride are the dominant ions. The dominant ions in each physiographic province reflect the reactivity of the underlying bedrock, and the proximity of the streams to the coast and industry (Figure from Anderson and George, 1966.)

Seasonal variability in concentrations of dissolved solids was observed. High concentrations of dissolved solids were measured July through September, typically the low flow period of the year; low concentrations of dissolved solids were measured March through May, typically the high flow period of the year.

The lowest sediment yields in New Jersey are measured in streams in the New England physiographic province, north of the extent of Wisconsin glaciation and in the Valley and Ridge province. Highest sediment yields are measured in streams in the New England province, north of the extent of Wisconsin glaciation and in the Piedmont province.

In the 1960s through 1976, water quality samples were collected at a variable number of stations each year, ranging from 33 in 1967 to a high of 238 in 1976 (figure of current and discontinued water quality stations in New Jersey). Several reports were published summarizing water quality in the Passaic River and Raritan River Basins, the largest river basins in New Jersey, before major point source pollution control legislation was passed (Anderson and George, 1973, Anderson and Faust, 1974). A fixed-site USGS/NJDEP Ambient Water-Quality Monitoring Network (QW Network) was finally established in 1976.

Data collected as part of the cooperative network are published every year in Water Resources Data of New Jersey Surface water (Bauersfield and others, 19)and every other year since 1984 in The New Jersey State water quality inventory reports (NJDEP 1986, 1988, 1990, 1992). These reports characterize all major river basins and tributaries in New Jersey, list drainage areas, populations centers, qualitative land use, major impoundments and point sources, and compare DO, bacteria, nutrients, ammonia, and metals at the QW Network stations using a Water Quality Index (WQI). the stations sampled remain consistent from year to year, however, many have been discontinued based upon statistical evaluation (Robinson, written communication); currently, the QW Network consists of 79 sites sampled five times a year for dissolved oxygen, bacteria, nutrients, ammonia, dissolved solids and metals (or figure of sampling sites here?). A history of the QW Network since 1976 summarizing the changes in the sites in the network, sampling frequencies, sampling procedures, constituents sampled, and laboratories is in progress (Pustay, written communication).

Data collected as part of the cooperative network have been analyzed more quantitatively for several major river basins in New Jersey including the Passaic River (NJDEP, 1987), the Great Egg Harbor River (Watt and Johnson, 1992), Tom's River (Watt et al., 1994) and Mullica River (Johnson and Watt, 1995). Each of these reports include time series plots, summary statistics, box plots and stiff diagrams, and an evaluation of the water resources in the basin.

Cooperative network data have also been used to answer other water quality questions of state-wide interest. Hay and Campbell (1990) analyzed trends in water quality data from 1980-96 for 86 sites and from 1976-86 for 69 of these sites using the Seasonal Kendall Tau and Censored Data Regression techniques. In general, downtrends were observed in trace elements and uptrends were observed in dissolved oxygen, fecal streptococci, specific conductivity, calcium, magnesium, sodium and chloride. Downtrends were observed in total organic nitrogen for the seven year study period only. Results of the analysis for other nutrients was equivocal. For example, downtrends were observed in total nitrogen at more sites in the seven year study period and uptrends were observed in total nitrogen at more sites in the eleven year study period.

Robinson et al. (199x) related trends observed by Hay and Campbell (1990) to basin characteristics using contingency tables. This report provides land use, population, point source, road deicing material use, fertilizer application and soil erosion rates for the 60 basins located entirely in New Jersey. In general results showed uptrends in pH, sodium, magnesium and chloride were associated with urban land use. Uptrends in pH, sodium, and chloride were also associated with basins having the highest water yields. Uptrends in fecal streptococci, ammonia, downtrends in pH, were associated with agricultural land use. Uptrends in sodium and chloride were associated with road salt application rate in the basin. The authors note a lack of association between nutrient trends and basin characteristics.

In addition the trend analyses, cooperative network data have also been modeled based on basin characteristics. Smith et al. (1993) developed a database of basin characteristics associated with 7,123 stream reaches in New Jersey using geographic information systems (GIS). They used digital elevation data to define stream reaches throughout the State and intersected this drainage network with population, landuse, and point source coverages so that this information was available for each of the 7,123 basins. Regression relations were developed relating basin characteristics, incorporating distance weighted decay, to total phosphorous concentrations at the cooperative network sites. The same data base and a similar approach was used to analyze the presence and distribution of organic contaminants and trace elements in two of the LINJ NAWQA retrospective reports accepted for publication in the Water Resources Bulletin. In addition to computing summary statistics and analysis of variance (ANOVA) for major drainage areas and physiographic provinces in New Jersey, the presence or absence of organic contaminants and trace element was related to basins characteristics.

This approach is used as part of an ongoing effort to develop a simple method for evaluating the relative importance of point source and non-point sources on stream water quality, Work is currently underway (State/Fed Coop study - Water-quality characteristics of NJ streams) to determine relations between concentration (DO, fecal coliform, hardness, Na, alkalinity, Cl, nutrients, ammonia, TOC, suspended sediment, boron and lead) and time, concentration and discharge, and loading and instantaneous discharge for all QW Network stations in NJ. The relation between load and discharge and concentration and flow for 15 parameters is defined statistically. For each site trends are examined in high flow and low flow samples separately. Increasing trends in high flow samples indicate an increase in contribution from non point sources, whereas an increasing trend in the low flow samples indicate an increase in contribution from point sources. In addition, diagrams are presented that show the strength of these relations (relative slopes of total nitrogen load in streamflow) in an entire basin.

To further evaluate the importance of point and non-point sources, a LINJ NAWQA retrospective study is in progress to analyze nutrient mass balance in the major basins of the study unit. An accounting spreadsheet was compiled including point source discharges, withdrawals and gaging station flow in downstream order for the Passaic, Raritan and Hackensack River basins, 1986-88. Point source nitrogen and phosphorous totals were calculated using the NOAA data base of discharge concentrations. Average annual nutrient loads at all water quality stations in the study unit were calculated using ESTIMATOR and MOVE1. The GIS data base of basin characteristics will be used to better evaluate and quantify the relative importance of point and nonpoint sources.

The work described above addresses the occurrence, distribution, loads and trends in nutrient concentrations (Table SW1) with the exception of the analysis of the distribution of trace elements and organic contaminants in bed sediments which was a major part of our analysis of existing data. Otherwise, very little information regarding the distribution and occurrence of toxics, such as pesticides and VOCs in surface water, is available. Ivahnenko and Buxton (1994) conducted a reconnaissance study of six drainage basins in NJ to determine the presence of pesticides from agricultural runoff in surface water. Drainage basins potentially affected by pesticide application and used for public supply were identified using GIS. Six basins were selected as most susceptible to pesticide contamination including the Lower Mine Hill Reservoir, South Branch of the Raritan River, Main Branch of the Raritan River, Millstone River, Manasquan River and Matchaponix Brook. All but the Lower Mine Hill Reservoir basin lie within the LINJ study unit. Twenty-eight surface-water samples were collected as part of the reconnaissance sampling, including 6 samples from water-treatment facilities and quality control samples. Although atrazine and metolachlor were detected in 86% of the samples, alachlor in 55% of the samples, and diazinon in 45% of the samples, all but one concentration measured during the study were less than the USEPA's recommended Lifetime Health Advisory Limit.

In a follow up study of the Millstone and Shark River basins (Buxton and Dunne, 1993), a pesticide vulnerability index based on percentage of agricultural land in the basin, pounds of pesticides applied per square mile and soil surface loss potential was compared with water-quality data. In eight baseflow and 24 stormflow samples of the Millstone River, low concentrations of alachlor, atrazine, metolachlor and simazine were measured. Higher concentrations were measured in stormflow samples. No pesticides were detected in 8 baseflow and 13 stormflow samples of Shark River.

VOC data for surface waters in NJ is limited to a 14 site synoptic survey in spring of 1994 on the Hackensack River and 11 samples collected during field reconnaissance as part of the LINJ NAWQA. These results are summarized in a fact sheet in editorial review.

New York

In New Jersey, many people depend on water supplied from the surface waters in the basins in which they reside. Residents of Long Island in Brooklyn and Queens counties obtain their drinking water from surface water in upstate New York, whereas people in Nassau and Suffolk counties depend entirely upon the underlying groundwater reservoir for their fresh-water supply. Early surface water studies on Long Island investigated the effects of urbanization, particularly increased impermeable area, on stream runoff and recharge to the groundwater reservoir (Sawyer, 1963, Seaburn, 1969). Based on a hydrologic comparison of East Meadow Brook, located in a watershed affected by urbanization and Mill Neck Creek, located in an unaffected watershed, for the period 1952-60, Sawyer (1963) calculated 2% or 63,000 gallon per day of recharge to the groundwater reservoir was lost because of the change in land surface. Seaburn (1969) related indices of urban development to increases in the volume of annual direct runoff to East Meadow Brook during the period 1937-66. In addition, Seaburn calculated that the average peak discharge of a 1-hour unit hydrograph increased from 313 cubic feet per second for storms 1937-43, to 776 cfs for storms in 1960-62. Rainfall-runoff analyses indicated direct runoff during 1960-66 was from 1.1 to 4.6 times greater than the corresponding runoff during 1937-43.

A 13 site water-quality network was established on Long Island through NASQAN and cooperative efforts in the mid 70's. In 1987, Suffolk County augmented the regular nutrient/inorganic sampling to include quarterly VOC and pesticides sampling at the 13 sites and annual sampling at about 75 other sites. VOC data for Long Island streams are summarized in a fact sheet in editorial review. Concern with surface water quality on Long Island relates not to supply but to the degradation of tourism and fishing industries and shellfishing habitat resulting from increased bacteria loads delivered to the Sound.

Several studies compared the quality of streams (dissolved solids, nitrate, detergent, temperature) in urbanized basins and less developed basins (Koch, 1970, Pluhowski, 1970, Koppelman et al., 1990), or streams in sewered areas versus non-sewered areas (Ragone et al., 1981, Reilly, et al., 1983). Koch (1970) compared the quality of streams in the sparsely populated Suffolk County to those in the densely populated Nassau County. Dissolved solids were higher, average nitrate concentrations 14 times higher and detergent concentrations 9-18 times higher in the Nassau County streams than the Suffolk County streams.

Pluhowski (1970) investigated the effects of urbanization on the temperature of streams on Long Island. Pond construction, clearcutting of vegetation from stream banks, increased storms runoff, and reduction in the amount of groundwater inflow were the factors that determined the significant differences in temperature among Long Island streams in basins of different degrees of urbanization. Streams draining urbanized basins were found to be 5-8 times warmer in summer than unimpacted streams.

Ragone (1981) compared nitrogen concentrations in groundwater and surface water from sewered and unsewered areas in Nassau County. Although no significant differences were detected in groundwater quality between sewered and non-sewered areas, nitrogen total concentrations in streams draining the sewered area are significantly lower than in those draining the unsewered area where 95% of the streamflow is thought to be derived from the shallow part of the upper glacial aquifer.

Prince et al. (1988) presented a conceptual model of streamflow generation on Long Island. Long Island streams are either relict glacial Outwash channels or recent erosional features that act as drains to the water table ad are fed by local shallow groundwater systems that flow above the regional groundwater system.

Much of what is known about surface waters in Long Island resulted from the Long Island segment of the Nationwide Urban Runoff Program (NURP, Long Island Regional Planning Board, 1982) and the Flow Augmentation Needs Study (FANS, Suffolk County Department of Public Works, 1993). The Long Island Regional Planning Board report on the NURP study includes an explanation of the history behind recharge basins constructed throughout LI. Results of the study indicate detention an recharge of runoff in dry streambeds may remove indicator bacteria. The FANS report published in 1993 indicates that the decline in water table levels and decrease in stream length expected to result from sewering are not yet observed--no significant differences were measured between groundwater levels in sewered and non-sewered areas. Work continues as part of the FANS program to document natural vegetation variation in quadrats in four stream basins representing different types of wetlands: Sampawams Creek, Carll's River, Carman's River, and Santapogue Creek.

A series of investigations on the effects of urbanization on East Meadow Brook (e.g. Brown et al., in review) and urban stormwater runoff (Ku and Simmons, 1986) have quantified sources of pollution. Ku and Simmons monitored 46 storms in five recharge basins in representative landuse areas for trace elements, nutrients, composite organics and bacteria. The metal and bacteria load were found to decrease through the unsaturated zone. As in New Jersey, most water quality studies focus on nutrients and bacteria and little is published on the occurrence of toxics such as pesticides, VOCs and trace elements in bed sediments.

Biological Monitoring

Biomonitoring programs were also initiated by the NJDEP and the NYDEC in the mid-70's in cooperation with the USEPA. In NJ, 31 existing stations in the USGS/NJ Ambient Water-Quality Monitoring Network were adopted as routine biomonitoring sites. These sites were used to establish baseline biological information and were supplemented with localized intensive surveys wherever NJDEP priorities dictated. Because of the effects of rapid development (point and non-point pollution proliferation), many of the original 31 stations became either inaccessible or highly degraded. The NJDEP needed an updated reference site database that was robust enough to use with either site-specific or watershed-based surveys. From 1989 through 1991, most of NJ was surveyed to identify 43 sites in 8 ecoregions that qualified as biological reference sites as part of the Ambient Biomonitoring Network (AMNET). In addition, an entire state watershed-based inventory was initiated in 1991 and by 1995 has resulted in a biomonitoring network of more than 600 sites ranging in size and impact (Figure ECO3). Plans are to sample these sites on a 5-year rotation. The AMNET now serves as a solid foundation for the statewide water-quality inventory (305B) and planning/management decisions involving surface water-quality standards and biocriteria (NJDEP, 1994). Similar to NJDEP's AMNET network, the NYDEC's Stream Biomonitoring Unit has been monitoring water quality in NY since 1972. This biomonitoring effort, was a follow-up to sampling conducted by the Conservation Department (1926-1939), that documented many cases of severe pollution in NY's rivers and streams. During 1972-1992 over 721 sites on 170 streams were sampled. Of these, 373 are continuous statewide monitoring sites that include 16 sites on Long Island and 7 sites in the Ramapo River that fall within the boundaries of the LINJ SU. NYDEC's Stream Biomonitoring Unit documents the occurrence of trends in water quality of lotic systems in NY with more than twenty years of data (Bode et al., 1993).

The LINJ liaison committee has identified a need to develop a better understanding of the effects of and processes associated with (1) inputs of toxic materials such as, trace elements, VOCs, pesticides, and other synthetic organic compounds, (2) nutrient enrichment, (3) sediment, particularly as related to the fate and transport of toxic materials and nutrients, (4) stormwater quality, and (5) interbasin transfers of water. They suggested the NAWQA study focus on relations between sources and loads of toxics, sediment, nutrients, land use, accumulation in bed sediment, bioaccumulation in tissues, effects on aquatic communities, and other factors.

In summary, experience in the district and liaison committee discussion indicate that effects of land use, especially urban, on water quality are the primary issues in the LINJ SU. Consequently, our surface water activities will focus on pesticides, VOCs, nutrients, and (to some extent) trace elements in the urban environment and how these toxics affect biological communities.

### Surface-Water Activities

The LINJ liaison committee identified a need to develop a better understanding of effects of and processes associated with surface-waters including (1) inputs of toxic materials such as, trace elements, VOCs, pesticides, and other synthetic organic compounds, (2) nutrient enrichment, (3) sediment loading, particularly as related to the fate and transport of toxic materials and nutrients, (4) stormwater quality, and (5) interbasin transfers of water. They suggested the NAWQA study focus on relations between sources and loads of toxics, sediment, nutrients, land use, accumulation in bed sediment, bioaccumulation in tissues, effects on aquatic communities, and other factors.

Liaison committee and other discussions indicate that the effects of land use (non-point), especially urban, on water quality are the primary issues in the LINJ SU. Consequently, our surface-water activities have focused on pesticides, VOCs, nutrients, and (to some extent) trace elements in the urban environment and how these toxics affect biological communities.

HIP plans as modified by conference calls

As a result of the conference call and other communications between the NLT and LINJ, our FY97 SW network included (Figure SW-1):

2 indicator IFS (Bound Brook at Middlesex; Upper Great Egg Harbor R at Sicklerville),
3 indicator BFS (Neshanic R at Reaville; Saddle R at Ridgewood; Stony Brook at Princeton), and
2 integrator BFS (Passaic R at Two Bridges; Raritan R at Queens Bridge).

Because of the need to define VOC and pesticide occurrence in the LINJ urban environment, however, all sites were sampled for VOCs and pesticides in FY96 per their BFS or IFS frequency. Raritan River at Queens Bridge was designated as an integrator IFS in FY97 and we continued the monthly sampling of pesticides and VOCs through FY97. Other BFS were not sampled for pesticides after September, 1996 and VOCs after December, 1996. LINJ stratification (Figure Stratification)is discussed in the Environmental Setting.

### CONTAMINANTS IN BED SEDIMENT AND TISSUE

Because of the wealth of existing bed sediment data, sampling for contaminants in bed sediment and tissue (BS&T) was not a priority in our study unit during FY96-97. Two papers analyzing the presence and distribution of organic contaminants and trace elements, however, were accepted for publication in Water Resources Bulletin of AWRA. The historical basis for these papers consisted of trace element and organic contaminant data from bed sediment samples collected periodically at 295 sites throughout NJ from 1974 to 1993. Samples were collected by the USGS in cooperation with the NJDEP and analyzed at the NWQL. One-third of the sites in the network were sampled each year on a rotating basis; the number of samples collected per site over the study period ranged from 1 to 13. Sample locations included small, low-order streams in addition to locations on major rivers. Bed sediments were collected at four sites in the LINJ fixed-site network 7-11 times since 1974. Because of the wealth of existing data and the lack of consensus among the biologists regarding the relation between bed sediment and tissue contaminant concentrations, we delayed bed sediment sampling until October, FY98.

Accomplishments in FY 1996-97

Two factsheets and two journal articles on the distribution and occurrence of trace elements and organic contaminants in NJ streambed sediments were published.

O'Brien, A.K., 1997, Presence and distribution of trace elements in New Jersey streambed sediments,
U.S. Geological Survey Fact Sheet-049-97.
Stackelberg, P.E., 1996, Presence and distribution of chlorinated organic compounds in New Jersey
streambed sediments, U.S. Geological Survey Fact Sheet-118-96.
O'Brien, A.K., 1997, Presence and distribution of trace elements in New Jersey streambed sediments,
Journal of the American Water Resources Association, 33(2), p. 387-403.
Stackelberg, P.E., 1997, Presence and distribution of chlorinated organic compounds in New Jersey
streambed sediments, Journal of the American Water Resources Association, 33(2), p. 271-284.

Proposed BS&T work in FY 1998
Contaminants in bed sediments and tissue were not sampled during FY96-97, however, 14 sites were sampled for bed sediment in FY98 (October 1997). Tissues were collected at 6 of the 7 LINJ fixed sites (not yet at Passaic R at Two Bridges) and at two synoptic sites. The 14 sites selected for bed sediment sampling include the 7 fixed sites and 7 sites sampled as part of the VOC/pesticide/algae/benthic invertebrate synoptic studies completed FY96-FY97. The 7 additional sites selected from the 32 synoptic sites meet the following criteria: 1) no bed sediment data were collected previously as part of the NJDEP/USGS network, 2) fish population was surveyed during FY96 or FY97 as part of biological activities, and 3) biological community is comparable to majority of sites in the synoptic network. Although some sites on LI and the Coastal Plain of NJ satisfied the first two criteria, they did not fit the third criteria and are therefore a lower priority. These sites will be sampled for bed sediment in September of FY98 if end-of-year funds are available. Further details on sites, protocols, and other issues are contained in the Bed Sediment and Tissue Sampling section of the Ecological Activities chapter of this workplan.

### BASIC FIXED SITES

Our basic fixed site network consists of three indicator sites and two integrator sites. Indicator BFS include Saddle River at Ridgewood, Neshanic at Reaville, and Stony Brook at Princeton. The two integrator BFS are Passaic River at Two Bridges and Raritan River at Queen's Bridge.

Accomplishments in FY 1996-97

A reconnaissance survey of VOCs in eight streams located in a variety of land-use settings across NJ was conducted in March/April, 1996. At a reporting level of 0.2 ug/l, MTBE was the most frequently detected VOC, occurring in seven of eight streams with concentrations ranging from 0.2 to 4.9 µg/l. Largest concentrations (> 2.5 µg/l) were measured in the three highly urbanized basins (urban land use > 60%). Concentrations of MTBE in samples from the five basins with smaller amounts of urban land were all less than 1.0 µg/l. BTEX compounds were detected only at Passaic R. at Two Bridges and Raritan at Queens Bridge, sites representing our two integrator basins. A fact sheet summarizing this reconnaissance data along with existing VOC data for LI and NJ streams was published in 1997 by Terracciano and O'Brien (FS-063-97).

Regular network sampling began April 22, 1996. To further investigate the occurrence of VOCs and pesticides in study unit streams, both basic and intensive fixed sites were sampled for VOCs and pesticides through the fiscal year. A total of 17 samples were collected at each of the 2 IFS and 7-8 samples were collected at the 5 augmented BFS. In summary, 14 VOCs, 19 herbicides, and 8 insecticides were detected among the 69 samples collected at the 7 BFS/IFS sites April 22 - September 30, 1996.

MTBE (58%), chloroform (53%), TCE (32%), TCA (30%), PCE (25%) and methylene chloride (25%) were the most frequently detected VOCs in samples collected at the seven fixed sites. The highest detection frequencies and concentrations of MTBE, TCE, TCA, and PCE were observed at Bound Brook at Middlesex, the urban IFS indicator basin with 68% urban (44% res. 24% ind.) land use. Detection frequencies of most detected VOCs were highest in April and, except for the high number of detects in the August samples, decreased through the summer.

Atrazine (100%), prometon (100%), metolachlor (96%), simazine (96%) and alachlor (67%), carbaryl (61%) and diazinon (52%) were the most frequently detected pesticides in samples collected at the seven fixed sites. In general, the highest concentration of a given pesticide was observed at the site with the highest detection frequency. Highest concentrations of atrazine (10.0 µg/l) and alachlor (4.7 µg/l) were observed at Stony Brook at Princeton, a site draining developing formerly agricultural land. Highest concentrations of prometon (0.099 µg/l), carbaryl (1.5 µg/l), and diazinon (0.3 µg/l) were observed at Bound Brook at Middlesex. Highest concentrations of metolachlor (5.2 µg/l) were observed at Raritan R at Queens Bridge, an integrator basin of mixed land use. Highest simazine (0.1 µg/l) concentration was observed at Great Egg Harbor River at Sicklerville, a Coastal Plain site draining rapidly developing land. Pesticide detection frequencies were generally higher in June and July for alachlor, chloropyriferos, and DCPA, however, patterns were not distinguishable for most of the other frequently detected pesticides. Detection of 2,4D was twice as high in April as in any of the other months. See Accomplishments in FY 1997 under IFS for reports in progress summarizing seasonal variability in pesticide and VOC concentrations at BFS and IFS, April, 1996 through April, 1997.

During FY97, basic fixed sites were sampled monthly for major ions, nutrients, DOC/SOC, and suspended sediment and for VOCs in November and December only. Several high flow samples were collected on regularly scheduled sampling days. High flow samples were substituted for regularly scheduled samples where possible because three of the four BFS are sampled an additional five times per year as part of the USGS/NJDEP cooperative network. BFS were included in the VOC and pesticide synoptic studies conducted January, 1997 and June, 1997 (see section Water Quality Synoptics). BFS sampling is summarized. The flow for each sample collected at BFS was plotted on the flow duration curve. Samples collected in FY97 cover a wider range of flow conditions than those collected in FY96. The exception is at Saddle River at which high flow samples did not exceed those collected in FY96.

Raritan River at Queens Bridge was designated as an integrator IFS in FY97 and we continued the monthly sampling of pesticides and VOCs through FY97. Other BFS were not sampled for pesticides after September, 1996 and VOCs after December, 1996.

Continuous discharge is available for Saddle River at Ridgewood, Neshanic at Reaville, and Stony Brook at Princeton. Discharge data is obtained by correlation with sites directly upstream for Passaic at Two Bridges and Raritan River at Queen's Bridge.

Temperature loggers were installed at all BFS in April, 1997. Temperature data are collected hourly using Optic Stowaway Temp data loggers. Conductivity is not measured continuously at basic fixed sites. Additional bed sediment and nutrient data are available for four of the BFS through the cooperative USGS/NJDEP QW network.

Proposed BFS work in FY 1998

During FY98 basic fixed sites will be sampled monthly for major ions, nutrients, DOC/SOC, and suspended sediment. We will make a concerted effort to obtain high flow samples at sites lacking adequate high flow data, particularly Saddle River at Ridgewood.

INTENSIVE FIXED SITES

Our intensive fixed sites (IFS) are Bound Brook at Middlesex (NENJ-Urban) and Great Egg Harbor River at Sicklerville (CP-Developing Urban), both indicator sites. Raritan River at Queens Bridge was designated an integrator IFS (or augmented BFS) in FY97 and we continued the monthly sampling of pesticides and VOCs through FY97.

Accomplishments in FY 1996-97

During FY96, IFS were sampled weekly for major ions, nutrients, DOC/SOC, suspended sediment, VOCs and both pesticide schedules April - July and sampled biweekly for all constituents in August and September. During FY97, indicator IFS were sampled biweekly in October and monthly in November and December for all constituents (VOCs were sampled twice at Bound Brook at Middlesex in November and December). Biweekly VOC sampling resumed January through March (when pesticide samples were collected monthly). Pesticides were collected weekly in April (when monthly VOC sampling resumed), biweekly May through August, and monthly in September. Major ions, nutrients, DOC/SOC, and suspended sediment samples also were collected.

Streamflow, specific conductance and temperature were monitored continuously at the two intensive fixed sites. A stream gage shelter was installed at Great Egg Harbor River at Sicklerville on March 27, 1996. An ADR records stage every 15 minutes. A Hydrolab was installed on April 11, 1996, to monitor specific conductance and temperature every hour. A Sierra-Misco model 5096 radio transmitter previously had been installed at Bound Brook at Middlesex to monitor flood stages. The data logger in the Sierra-Misco gage was reprogrammed to record stage every 15 minutes. The stage data is downloaded to the Prime computer and stored as continuous-record stage. A CR10 and minimonitor probe were installed at Bound Brook at Middlesex on July 24, 1996 to continuously monitor specific conductance and temperature every hour.

IFS sampling is summarized in Tables 3a, b, and c. The flow for each sample collected at each IFS was plotted on the flow duration curve. Samples were collected at a wide range of flow conditions at the IFS. Figure SW98-4 shows the mean daily discharge for the period of record through the end of FY97 at each IFS along with the instantaneous discharge for each sample collected.

Streamflow, specific conductance and temperature were monitored continuously at the two intensive indicator fixed sites. Mean daily discharge and specific conductance are shown for each IFS for the period of record in. A temperature logger was installed at Raritan River at Queens Bridge, April, 1997. Temperature data are logged hourly.

Factsheets summarizing seasonal variability in pesticide and VOC concentrations at BFS and IFS, April, 1997 through April, 1998 are in progress:

"Seasonal variability in VOC concentrations in New Jersey streams" R.G. Reiser and others.
The purpose of this factsheet is to summarize VOC data collected at the 7 fixed sites April, 1996 through April, 1997.

"Seasonal variability in pesticide concentrations in New Jersey streams" A.K. O'Brien and others.
The purpose of this factsheet is to summarize pesticide data collected at the 7 fixed sites April, 1996 through April, 1997.

An analysis of stormflow variability in nutrients, VOCs and pesticides at the seven fixed sites will be underway this fall (Pflaumer and others).

Proposed IFS work in FY 1998

During FY98 IFS will be sampled monthly for VOCs from October-April and monthly for pesticides from October-June. IFS will be operated as BFS from July-September, 1998. This would give us a minimum of two-full years of IFS type data at the two indicator sites, basically, two high-level' seasons for VOCs (December-March) and three high-level' seasons for pesticides and nutrients (April-June).

### WATER QUALITY SYNOPTICS

Accomplishments in FY 1996-97

No synoptic water-quality studies were funded for FY96. In FY97, the two reach synoptic studies of VOCs and the study unit VOC and pesticide synoptic studies were carried out as planned. All VOC synoptic samples were collected January 28-31, 1997. To determine within basin variability/sources, 9 samples were collected in the Bound Brook watershed, 4 samples were collected in the Santapogue Creek watershed, and 3 samples each were collected in the Sampawams Creek and Swan River watersheds as part of the reach studies of VOCs. A total 51 stream samples were collected on LI and in NJ (Figure SW98-6) with an additional 3 replicates, 2 blanks, and 2 spikes.

All pesticide synoptic samples were collected June 9-20, 1997. A total of 53 stream samples were collected on LI and in NJ with an additional 3 replicates, 3 blanks, and 2 spikes. No significant amount of rain fell during the two week sampling period.

A retrospective factsheet (FS-063-97) by Terracciano and O'Brien (1997) describing the presence and distribution of VOCs in LI streams based on an existing NWIS data and data from the Suffolk County Department of Health was compiled in FY96 and published in FY97.

Factsheets summarizing the spatial variability VOC and pesticide concentrations in NJ and LI streams from FY97 synoptic survey data are in progress:

"Volatile organic compounds in New Jersey and Long Island streams" A. K. O'Brien, R.G. Reiser and H. Gylling. This factsheet summarizing the results of the study unit VOC synoptic study is currently in editorial review.

"Pesticides in New Jersey and Long Island streams" H. Gylling and others. The purpose of this factsheet is to present the results of the study unit pesticide synoptic study.

Proposed Synoptic work in FY 1998

Four synoptic studies are proposed for FY98. They are designed to gain a better understanding on the spatial variability and likely sources of pesticides and/or VOCs in Bound Brook, 3 LI streams, and 12 NJ Coastal Plain streams.

VOC Reach Study of three streams on Long Island (January or February, 1998).
The purpose of FY97 synoptic VOC study of urbanized stream reaches was to determine the downstream variability in VOC concentrations and, hence, some indication on the variability in sources of VOCs in these streams. Consistent patterns were not evident in the data collect January, 1997 and higher VOCs seen in LI streams. We propose to conduct intensive field reconnaissance of the three LI streams sampled to explain the concentrations observed in FY97. We will then modify our sampling design and collect an additional sample at ten sites in the 3 systems. Streams to be sampled are Swan River, Sampawams, and Santapogue Creeks (see Proposed Work in FY 1997 VOC Reach Study of three streams on Long Island). Streams will be sampled for VOCs, field pH, and conductivity only. Stream and air temperature will also be recorded. Discharge measurements will be made at ungaged sites. This study will be conducted by study unit personnel in NJ and LI offices, in late January or February, 1998.

VOC storm sampling in Bound Brook (Winter 1997-98).
Contrasting patterns in downstream VOC concentrations were observed in the two major tributaries to our Bound Brook IFS sampled during the VOC reach study in January 1997. Along Bound Brook, the more industrialized reach, VOC concentrations decreased downstream from the most upstream site where relatively high VOC concentrations were observed and a point source is suspected. Along Green Brook, a more residential/commercial reach, VOC concentrations increased downstream from NDVs. VOC concentrations at Bound Brook at Middlesex, our IFS, were between those observed in the two tributary sites just upstream, both during the baseflow synoptic and on another occasion when those same three sites were sampled during high flow. To better define how VOC concentrations vary at our IFS during stormflow from non-point and point sources, we propose to collect about three samples at each of 3 sites (Bound Brook indicator IFS and two upstream sites on the Green Brook trib) over a winter storm hydrograph.

Coastal Plain Pesticide/Algae/Benthic Invertebrate Synoptic Study (late May, early June, 1998).
FY97 synoptic study of pesticides focused primarily on northern NJ to compliment the algae/benthic invertebrate synoptic study conducted September/October, 1996 and the pesticide synoptic study conducted at the same time by the NY District. A synoptic study of pesticides/algae/benthic invertebrates is proposed for FY98 to determine the spatial variability in pesticide concentrations and detections in the Coastal Plain. About 10 sites will be selected to compliment NJ District work currently ongoing in the Tom's River/Metedeconk basins and the Pineland Commission work currently ongoing in the Mullica River basin. Interest in Tom's River basin is primarily related to rapid development and local concern with recent cancer cluster research and water quality. Interest in the Mullica River basin is primarily related to cranberry farming practices. Streams will be sampled for pesticides, nutrients, major ions, DOC/SOC, and suspended sediment. Water quality samples and biological samples will be collected and processed in the field by regular sampling teams, 1-2 sites per day. Sampling should be completed in about two weeks or less.

Pesticide reach study of Bound Brook (June, 1998).
The highest number of non-agricultural pesticides in LINJ surface waters was observed at our Bound Brook IFS indicator site. To determine the sources of pesticides in this urban indicator stream, a synoptic study is planned to sample reaches draining sub-watersheds of different land use. One tributary to Bound Brook drains a watershed of primarily industrial/residential land use and the other tributary drains a watershed of primarily residential/commercial land use. We propose to sample 9 locations along the two main tributaries (4 on Green Brook and 5 on Bound Brook) for pesticides, field pH, DO, and conductivity only. Stream and air temperature will also be recorded. Discharge measurements will be made at ungaged sites. Sampling will be conducted in June, when pesticide concentrations are highest. The IFS will also be sampled as part of that station's regular schedule.