Water Quality Data

Table of Contents:
Prior Studies (before 2003)
2004-2005 Study
2006-2007 Study
2006-2007 Study Sites
2008 Study

Prior Studies (before 2003), Summary


The Yonkers stretch of the Saw Mill River was found in 1983 to contain the highest levels of heavy metals, PCBs, orthophosphate and PAHs in the river (Rogers 1984). Urban runoff was determined to be the source of these contaminants. In 1993-1994, the USGS and RPI conducted an assessment of the distribution of chlorinated organic compounds in sediments of the Hudson River Basin (Phillips et al. 1997). Although quantitative information was not given in the study, DDT and chlordanes (including trans-chlordane, cis-chlordane, cis-nonachlor and oxychlordane) were reported in Saw Mill River sediments. Due to the high levels of these chlorinated organics in the Saw Mill River and the other tributaries included in the survey, it was concluded that urban watersheds may constitute the major source of these compounds to large rivers such as the Hudson. In 1998, the USGS reported on the water quality of the Hudson River basin over the years spanning from 1992-1995 (Wall et al. 1998). They noted that the stream-bottom sediments of the Saw Mill contained the highest concentrations of metals among all sites in the USGS National Water-Quality Assessment (NAWQA) Program (Table 1).

Table 1. Concentrations of metals in Saw Mill River sediments (Wall et al. 1998)

Metal Concentration (mg/g)
Cadmium 6.9
Copper 410
Mercury 1.4
Nickel 72
Zinc 980


An extensive one-day sampling was performed on August 26, 1992 by the NYS DEC Stream Biomonitoring Unit. The NYS DEC concluded that a significant decline in the water quality of the Saw Mill River had occurred, specifically within the reach stretching between Nepera Park and downtown Yonkers (Bode et al. 1993). The severest degradation in water quality was observed in the downtown Yonkers area near the USGS station (Station 5).

A 1998 study by the Stream Biomonitoring unit found that the Yonkers stretch of the Saw Mill was receiving contaminants through “multiple urban sources in a defined stream segment,” and was considered medium priority for remediation (Bode et al. 2001). The Nepara Park area in North Yonkers was controlled by nonpoint source nutrient enrichment and the invertebrate community was dominated by filter-feeding caddisflies. Downstream, municipal and/or industrial inputs were deemed the cause of impared water quality and the invertebrate community was dominated by tolerant midges, snails and worms. The also noted that tissue analysis of crayfish revealed elevated levels of metals, PAHs and PCBs.

A document from the Army Core of Engineers reported on the degradation of the Saw Mill River over the past 20 years as a result of flood mitigation projects (Pearce 1999). They concluded that the river ecosystem was severely impacted due to improvements in flood control. Additionally, they concluded that channelization of the River in downtown Yonkers prevents the natural movement of aquatic life from the Hudson River into the Saw Mill. In particular, the stream bed design and the high flows in the channelized areas prevent the river from supporting aquatic life.

Table of Contents

2004-2005 Study


Managing partner:
Saw Mill River Coalition

Manhattan College, Saunders Trade and Technical High School

Key Findings:

The monitoring activity led to some intriguing data that provides an initial picture of the water quality conditions and the impact on the river of stormwater events. In general, the water quality index of the river is very poor (6 out of 100), with high phosphate levels, hard water, high turbidity, and poor or invasive plant life.

More specifically, the project obtained the following results:

Dissolved Oxygen (DO) levels were disturbingly low, probably as a result of limited rain during the summer of the study, but mostly due to the extreme lack of organisms, poor sediment, and lack of plant life in the stream. Stormwater events appeared to actually improve DO, as the stream became swift and agitated.
Nitrogen levels were very high in the fall, perhaps related to falling leaves
Salt impacts on the river before and after stormwater events were not unusually high, even though the nearby Saw Mill River Parkway receives tons of salt during the winter. It did appear, however, that salt concentrations were affecting the diversity of plant life adjacent to the river. Salt levels did fluctuate, especially in the winter, but never reached levels of concern for aquatic life.
Ammonia and nitrate levels were high at one site close to a residential area, suggesting the potential that lawn fertilizers are getting into the stream. Ammonia levels clearly rise during stormwater events.
Of all the metal tested, iron, manganese, and aluminum were somewhat high in both baseline and stormwater samples. However, iron and manganese levels are not out of the ordinary and could be a natural phenomenon related to native soil conditions. Aluminum would be a concern, but was found in the insoluble form, which is less problematic.
Conductivity decreased consistent with the stormwater events and stream dilution, and clearly can act as a general measure to demark the occurrence and extent of a stormwater event. In addition, the stream, as expected, is very flashy. The depth of the river doubled from two to four meters during stormwater events.

Project Summary:

Overall the project was highly successful in engaging volunteers and generating sound water quality data focusing on storm water concerns in the Saw Mill River. Over the course of 15 months, more than 30 student volunteers from a local high school and two from a local engineering college (plus two professors and a high school science teacher) were involved in monitoring activities. The group obtained both baseline and stormwater samples during the project period. Sampling occurred at seven sites along the Saw Mill River, however the Hearst Street location served as the main stormwater assessment area.

In addition to the above monitoring efforts, one college student summarized previous water quality research on the Saw Mill River to place the new data in a historical context. In addition, a special project was completed during the year focusing on the impact of road salt on river conditions. Finally, the project was successful in communicating results to a broader audience. High school students created a project web site and stories about the initiative ran in The New York Times and the Riverdale Press, an award winning local publication.

In terms of monitoring, project activities included the following:

A Hydrolab Minisonde probe was deployed at Hearst Street to collect data on a continuous 24/7 basis focusing on stormwater issues. It was deployed three times during the project period, with each interval lasting 2-3 weeks. The probe was removed each time by volunteers and data transferred to a laptop computer for analysis in the classroom.
Baseline water sampling was conducted in the spring and summer of 2004, and in the spring of 2005. Seven sites were used looking at 12 different parameters, including shoreline plant diversity, macroinvertebrates and sediment texture.
One volunteer, a high school senior, monitored the river on a weekly basis during most of the project period: August 2004 to May 2005. She focused on several concerns: salinity, conductivity, pH, and temperature. Stormwater impacts on these issues were evaluated (see attached graphs).
Additional stormwater data was collected by the college students using ISCO Field Samplers. This field work was conducted on the Saw Mill twice during the project period: once in the summer of 2004 and once in the spring of 2005. They evaluated stormwater samples over an 18-hour period including before and during significant rain events.

Statutory Background:

The Federal Clean Water Act Section 305(b) requires all states to assess and report on the quality of all water bodies in their state. In 2002, New York State listed the Saw Mill River as a “Waterbody affected by toxic pollutants” with pesticides cited as the cause. Section 303(d) of the Federal Clean Water Act further requires states to identify from this assessment a list of “Impaired Waters” where the water quality is not sufficient to meet specific designated uses and where restoration and protection efforts beyond conventional technology-based controls are necessary to address water quality issues. For these waters, the states must consider the development of a Total Maximum Daily Load (TMDL) approach to reduce the input of the specific pollutants that restrict water uses. As of 2004, the Saw Mill River is cited on the List of Impaired Waters Requiring a TMDL for floatables and sediments contaminated with chlordane.

Data Collection:

Seven sites along the Saw Mill River were sampled.

Site # Site Location River Mile
1 Farragut Ave. 4.73
2 Hearst St. 4.28
3 Executive Blvd. 3.97
4 Nepperhan/Roberts Pl. 2.76
5 Torre Place 2.46
6 USGS Site (Centre St) 1.25
7 Walsh Rd. 0.94

These sites were selected as a result of their accessibility to the river. Note that Site 6 corresponds to sampling station 5 of the 1992 DEC Biological Assessment. A USGS map with the sampling locations marked is shown in Figure 1.

Samples were taken on two non-rainy days in 2004 (July 21 and October 15). The analysis performed on each sample included: pH, conductivity, total phosphate, orthophosphate, nitrate (as NO3-N), ammonia (as NH3-N), chemical oxygen demand (COD), total Coliform bacteria, and metals (Fe, Cu, Cd, Pb, Zn, Mn, As, Cr).

The results from baseline sampling for pH, conductivity, nitrate, ammonia, orthophosphate and COD are shown in Figure 2. In general, pH and conductivity were relatively constant at 8.0 and 800 mS respectively for both sampling days. Nitrate concentrations were fairly constant with distance across the river, however, the levels of nitrate were 2.5-times higher during the October sampling. This may be due to falling of leaves from surrounding trees in the Fall followed by decomposition. Ammonia levels did not change a great deal from July to October, however the data does suggest that levels were highest near river mile 4.0. This area represents a highly residential area and ammonia inputs may be coming from fertilization of lawns. Levels of orthophospahte and total phosphate (data not shown) increased with decreasing river mile and showed some variability across the two samplings. COD readings were highly variable, with no clear trends across the length of the river.

Of the metals analyzed (iron, zinc, copper, cadmium, manganese, nickel, arsenic, chromium), only iron and manganese were above our instrument detection limits. Iron concentrations increased with decreasing river mile, while manganese did not show ant obvious trends. Average concentrations of iron and manganese were 0.22 and 0.070 mg/L respectively.

New York Times Article

Testing a River After Storms, New York Times, Sunday, October 17, 2004

Table of Contents

2006-2007 Study

New York State Water Resources Institute

Study name:
“Effect of urban runoff on seasonal and spatial trends in the water quality of the Saw Mill River”

Managing partner:
Manhattan College, Assistant Professor, Richard Carbonaro

Saunders Trade and Technical High School, Saw Mill River Coalition

Current status:
Data began being collected in June 2006 at 12 sites on a weekly or bi-weekly basis. This will continue through at least May 2007. Outcomes from this study will be posted to this site by the end of February 2007.

2006-2007 Study Sites

Water Quality Monitoring Sites, June 2006 through 2007
Manhattan College

Site # Site Location River Mile
1 Chappaqua by train station
2 Pleasantville Rd & Manville Rd
3 Rt 9A at Rosedale Nursery
4 Where river crosses Old Saw Mill River Rd, Eastview
5 Elmsford, at Warehouse Lane
6 Woodlands Lake outflow
7 Macy Park, Ardsley
8 Hearst St., Yonkers 4.28
9 Torre Place, Yonkers 2.46
10 Walsh Rd, War Memorial Field 0.94

Table of Contents

2007: Effect of urban runoff on seasonal spatial trends in the water quality of the Saw Mill River

Project Overview

The Saw Mill River is a tributary to the Hudson River in the Lower Hudson River Drainage Basin. The headwaters of the river are located in the Village of Chappaqua in the town of New Castle. The River has several designated uses over its entire stretch, indicating diverse surroundings as it travels from Chappaqua to Yonkers, NY where it eventually discharges into the Hudson River. Additionally, along the Yonkers stretch, the river has been severely altered over the past 30 years by flood control projects and re-routing (Pearce, 1999).

Urban Runoff is one of the leading sources of water quality impairment in surface waters. A high percentage of urban areas makes the Saw Mill susceptible to contamination from urban runoff.
The primary objective of this work was to conduct a year-long continuous monitoring program for the entire stretch of the Saw Mill River. The following water quality parameters related to urban runoff were monitored: surface water nutrients (ammonia, nitrate, and total phosphorus) and fecal coliform bacteria. Temperature, conductivity, pH, total suspended solids, and turbidity were also measured.

This work included partnerships with Groundwork Yonkers and Saunders Trade and Technical High School in Yonkers. Groundwork Yonkers is a non-profit organization developed in 1999 dedicated to revitalizing, greening, and connecting people to the urban environment in lower Westchester county. It is the coordinator of the Saw Mill River Coalition, a partnership of non-profit groups, government agencies, and businesses, aimed to revitalize and protect the Saw Mill watershed.

Table 1: Personnel on this project

Name Level Field of Study Institution Role
Jason Lumish Undergraduate (Junior) Civil Engineering (major: Environmental) Manhattan College, Riverdale, NY Field sampling, sample analysis
Erica Hanley Undergraduate (Sophomore) Environmental Engineering Manhattan College, Riverdale, NY Sample analysis
Christopher Fanelli Undergraduate (Sophomore) Environmental Engineering Manhattan College, Riverdale, NY Database management
Eric Spargamino Graduate Environmental Engineering Manhattan College, Riverdale, NY Field sampling, sample analysis
Michael Lynch Graduate Environmental Engineering Manhattan College, Riverdale, NY Field sampling, sample analysis
Nicole Kerrison High School (Junior) Environmental Technology Saunders Trades & Technical High School, Yonkers, NY Sample analysis
Leslie Guadron High School (Sophomore) Environmental Technology Saunders Trades & Technical High School, Yonkers, NY Sample analysis


Sample Site Selection

Sites were selected along the Saw Mill River so that a representative spatial profile of water quality could be obtained (Table 2). Distances are reported as “Distance from Elm St.,” which is the location at which the Saw Mill River travels underground in Yonkers.

Sampling Frequency
Samples were collected approximately every one to two weeks over the period of June 26, 2006 to October 23, 2006.

Table 2: Sampling Locations

Site ID Location Latitude Longitude Distance from Elm St. (miles)
S1 Chappaqua Metro North Station, Chappaqua, NY 41.1565 -73.7756 22.93
S2 Pleasantville Rd/Manville Rd., Pleasantville, NY 41.1332 -73.7962 19.48
S3 Saw Mill Road, Eastview, NY 41.0811 -73.8287 14.5
S4 Saw Mill River Road (Rosedale Nursery), Hawthorne, NY 41.0952 -73.8108 16.33
S5 Warehouse Lane, Elmsford, NY 41.0637 -73.817 12.61
S6 Executive Blvd., Elmsford, NY 41.0704 -73.8082 N/A
S7 Woodlands Lake (spillway), Dobbs Ferry, NY 41.024 -73.8452 8.62
S8 V.E. Macy Park, Ardsley, NY 41.0183 -73.8463 8.26
S9 Lawrence St., Ardsley, NY 41.0025 -73.8563 6.57
S10 Hearst St., Yonkers, NY 40.9744 -73.8688 4.13
S11 Torre Pl., Yonkers, NY 40.9517 -73.8779 2.17
S12 Walsh Rd., Yonkers, NY 40.9379 -73.8893 0.4


Basic Water Quality
A summary of basic water quality parameters (pH, conductivity TSS and turbidity) are reported in Table 3. Median, minimum, maximum, and standard deviations are listed.

Table 3: Summary of Basic Water Quality Parameters(a)

Site # Site pH Conductivity (μS/cm) TSS (mg/L) Turbidity (NTU)
S1 7.36 (7.10 -7.70) ; 0.19 632 (186 – 891); 190 2.9 (1.0 – 17) ; 11 3.84 (1.6 -10.9) ; 3.56
S2 7.62 (7.32 -7.90) ; 0.20 605 (190 – 793); 170 1.6 (0.5 -35) ; 18 1.87 (0.74 -18.6) ; 5.09
S3 7.60 (7.32 -7.90) ; 0.19 663 (249 – 829); 173 2.1 (1.0 -61) ; 23 1.46 (1.02 -38.9) ; 11.2
S4 7.53 (7.16 -7.80) ; 0.20 621 (222 – 795); 176 2.1 (0.9 -46) ; 21 2.19 (0.91 -27.2) ; 7.6
S5 7.46 (7.13 -7.85) ; 0.22 755 (190 – 1041); 245 2.1 (1.1 -14) ; 7.6 2.52 (2.08 -16.8) ; 4.48
S6 7.64 (7.22 -7.90) ; 0.22 781 (389 – 1005); 165 3.2 (1.6 -64) ; 23 2.39 (1.79 -42.1) ; 11.9
S7 7.60 (7.14 – 8.00) ; 0.25 674 (198 – 823); 182 4.8 (3.7 -60) ; 17 4.88 (2.66 -59.1) ; 17.3
S8 7.70 (7.32 – 8.00) ; 0.20 666 (185 – 824); 193 4.6 (1.9 -41) ; 11 3.11 (1.97 -46.9) ; 13.2
S9 7.57 (7.23 -7.90) ; 0.21 762 (184 -846) ; 203 3.9 (2.3 -61) ; 17 3.21 (1.00 -60.1) ; 17.2
S10 7.72 (7.17 – 8.00) ; 0.22 758 (192 -869) ; 203 4.4 (1.6 -64) ; 18 3.07 (1.50 – 53.0) ; 15.2
S11 7.81 (7.28 -8.17) ; 0.27 740 (205 -826) ; 187 6.9 (3.1 -58) ; 15 3.86 (2.73 -45.6) ; 12.7
S12 7.67 (7.31 -8.02) ; 0.23 751 (205 -862) ; 197 4.9 (2.5 -59) ; 16 3.89 (2.33 -46.1) ; 12.8

(a) Data reported as median (min – max); standard deviation

The pH of the Saw Mill River showed very little spatial and temporal variability and is generally within the acceptable levels for freshwater streams.
Electrical conductivity (EC) showed large temporal variability, in part due to very low conductivity readings recorded on 8/28/06 during the middle of a heavy rainfall event. In general, the EC values are on the high side of freshwater streams, and not uncommon for streams impacted by urban runoff.

Water high in TDS can be harmful to fish and plants and may limit biodiversity.  Assuming a conversion factor of .7, the approximate median and maximum TDS of all samples taken from the Saw Mill River are 500 and 700 mg/L respectively. While there is some uncertainty in our choice of conversion factor for TDS, the river appears to be impacted by dissolved solids.

Fecal Coliform Bacteria

Sample sites S11 (Torre Pl., Yonkers, NY) and S12 (Walsh Rd., Yonkers, NY) had the highest median fecal coliform counts. These two sites are the southernmost sites sampled and are in a highly urbanized area of downtown Yonkers.

All median fecal coliform bacteria values were above 200 org./100 mL. This is significant because the NYS DEC criteria (1999) for fecal coliforms states that the monthly geometric mean shall not exceed 200 org./100ml. Bacterial contamination can originate from point or nonpoint sources. For the Saw Mill River watershed, stormwater runoff, and municipal wastewater discharges are likely causes of the observed high fecal coliform bacteria counts. While fecal coliform bacteria are subject to inactivation upon direct exposure to sunlight, portions of the Saw Mill River are under canopy and rapid die-off is unlikely.

Wastewater generated from much of the area of the Saw Mill River is treated by only the Yonkers Wastewater Treatment plant. While the recorded values at site S1 may seem extremely high, they are significantly less than levels commonly found in raw sewage and sewer overflows. Sewer overflows occurring during wet weather in the upper watershed may therefore be responsible for high fecal coliform bacteria counts at site S1.


In general, all ammonia concentrations are relatively low with most values falling below .4mg N/L, and the largest value recorded being 0.79 mg N/L. Median total ammonia levels are generally slightly higher during or after storm events as indicated by the daily precipitation record.

Total ammonia levels clearly increase during rainfall. This is evident by comparing the temporal profiles for ammonia and conductivity. A cross plot of ammonia versus conductivity yielded a correlation coefficient for ammonia and conductivity of .86 (2=24). Therefore, stormwater runoff is an important non-point source of ammonia to the river.

Municipal wastewater, agricultural runoff, fertilizers, and animal feedlots are potential sources of ammonia (and nitrate) to streams.

Nitrate and Phosphorus

Nitrate concentrations in the Saw Mill River did not show a large amount of spatial variability. All nitrate readings were less than 1.0 mg/L. At these levels, nitrate is not expected to be toxic to aquatic life. Many recorded values for nitrate in late summer/early fall exceeded the EPA reference condition for total nitrogen, even without considering contributions of ammonia and organic nitrogen.

The total phosphorous values observed here are higher than the USEPA Ecoregion XIV reference conditions, but are in line with levels found for rivers and streams flowing through urbanized areas. While freshwater lakes are usually P-limited, freshwater lakes can be either N or P-limited (Thomann and Mueller, 1987).

Table of Contents

2008 Study

Hudson River tributaries, such as the Saw Mill River, function as spawning areas for migratory fishes. This is one of the significant ecosystem roles that these tributaries play. Hudsonia, a non-profit institute for environmental research and education conducted a study of migratory fishes in the mouth of the Saw Mill River during the spring of 2008. The Hudsonia group made four collecting trips to the mouth of the Saw Mill River in 2008. During these trips they caught a total of 17 fishes of two species, white perch (Morone Americana) and white sucker (Catosomus comersonii). These fish were caught using 50 ft. long by 6 ft. deep gil nets with 1.25 inch bar mesh, which are ideal nets for catching most migratory species of fish. The fish were identified, measured, photographed, and then released alive. Both of the species that were caught are migratory fishes that spawn in tributaries. While white perch spawn in tributaries and in the main tidal Hudson River, white sucker appears to completely depend on tributary spawning.

The number of fish and variation in species collected in the Saw Mill River is small compared to all other Hudson River tributaries however it is possible that white perch numbers are generally low in the Hudson River this year. The Hudsonia group was able to make hypotheses as to why these numbers were so low. First, the Saw Mill River is covered by cement during part of its path and therefore very dark, which can discourage fish from spawning there. This potential roadblock could be remedied by the daylighting of the Saw Mill River, planned for a portion of the stream adjacent to the sample site. A second problem is the concrete “flume” which exists in the Saw Mill River mouth under the Yonkers train station and may be preventing access to the river. Many migratory fish have limited swimming capability when it comes to ascending obstacles. This “flume” therefore, could prevent advantages of the future daylighting from reaching the migratory fish and thus should be mitigated. Though the number of fish found during the 2008 tests was low, it is important that fish have access to the Saw Mill for spawning, especially the white suckers which depend entirely on tributary spawning to sustain their populations.

Table 1. Fishes collected in gil nets in the mouth of the Saw Mill River, 2008. # is the specific number of individuals collected, Mean TL is the average total length of the specimens, Range is the range of total lengths of the specimens, and CPUE is number collected per net-hour.

Date Species # of Fish Mean TL Range CPUE
5-Apr White Perch 1 23.2 n/a 0.33
22-Apr White Perch 1 19.7 n/a 0.4
22-Apr White Sucker 1 39.3 n/a 0.4
10-May White Sucker 3 32.5 31.6-33.4 1
27-May White Perch 4 20.6 19.0-21.2 1.33
27-May White Sucker 7 33.5 29.7-42.0 2.33

Table of Contents

Design and Hosting by Jackson Whelan and Ditto! Design! | Powered by WordPress