Natural and Anthropogenic Influences on the Mount Hope Bay Ecosystem 2006 Northeastern Naturalist 13(Special Issue 4):173–198 Hypoxia in the Upper Half of Narragansett Bay, RI, During August 2001 and 2002 Christopher F. Deacutis1,*, David Murray2, Warren Prell2, Emily Saarman2, and Larissa Korhun2 Abstract - Narragansett Bay, RI, is considered to be a relatively well-mixed estuary not subject to extensive seasonal stratification and hypoxia. However, results of surveys of dissolved oxygen (DO) for the upper half of Narragansett Bay on August 15, 2001 and on August 6, 2002 have documented evidence of wide-area intermittent subpycnoclinal hypoxia (􀂔 3 mg l-1). For the August 2001 survey, severe hypoxic to near-anoxic levels were confined to the Providence River, the western side of Greenwich Bay, and a small area of Mount Hope Bay, but hypoxic levels below 2 mg l-1 were also experienced on the western side of the Upper Bay in an extensive, shallow oxygen minimum. Hypoxic bottom waters (􀂔 3 mg l-1) extended from the Upper Bay into the upper West Passage. Hypoxic waters covered approximately 66 km2 (36%) of the survey area for August 15, 2001. A more extensive and severe hypoxic event occurred during the August 2002 survey, when near-bottom waters of the entire Providence River and a large area of the Upper Bay and upper East Passage were severely hypoxic to near-anoxic, while other parts of the Upper Bay, upper East Passage and upper West Passage were hypoxic at depths greater than 5 m. Limited data for Mount Hope Bay in August 2002 documented small hypoxic areas of the southern end of that subembayment. The total hypoxic area for August 6, 2002 was approximately 93 km2 (65%) of the total area surveyed. Decreased estuarine circulation due to a severe drought may have contributed to the wider extent of hypoxic and near-anoxic waters in large areas of the upper half of Narragansett Bay recorded in the August 6, 2002 survey as compared with the August 15, 2001 survey. Results of the oxygen surveys affirm sediment profile camera work and limited benthic studies that previously suggested parts of the Mid Bay have become subject to increased organic loading impacts. These impacts can take place even under drought conditions, when only point source nutrients are the major contributors to nutrient loadings entering the upper half of Narragansett Bay. Introduction Narragansett Bay is a northeastern US, medium-sized (370 km2) temperate, semi-diurnal tidal estuary with a 4714-km2 watershed and three significant urbanized freshwater inflows: the Blackstone, Pawtuxet, and Taunton Rivers (Nixon 1995a; Nixon et al. 1995; Ries 1990; Robinson et al. 2003, 2004). Annual mean monthly river flow from these watersheds is only 104 m3 sec-1. Longitudinal salinity range is 􀂧 11–31 psu (depending on river flows), tidal range is 0.6 to 1.9 m, and average depth is 7.8 m (Bergondo et al. in press, 1Narragansett Bay Estuary Program, URI Graduate School of Oceanography, Narragansett, RI 02882. 2Brown University, Providence, RI 02912. *Corresponding author - deacutis@gso.uri.edu. 174 Northeastern Naturalist Vol. 13, Special Issue 4 Chinman and Nixon 1985, Ely 2002, Pilson 1985, Ries 1990). The tidal Providence River is actually part of the headwaters of the estuary, with a mesohaline mixing zone. The majority of runoff for the western part of the watershed enters the Bay through the Providence River (Ries 1990). Most of the major rivers (outside of the Taunton River) were dammed very close to their entry into the estuary over a century ago (Nixon 1995a, Ries 1990). South of the tidal Providence River area, the Bay is considered weakly stratified to well mixed (Nixon et al. 1995). Recent attempts to predict the sensitivity of various US estuaries to nutrient inputs have concluded that Narragansett Bay has “moderate” susceptibility to acknowledged high levels of nitrogen inputs, with low expression of impacts such as loss of submerged aquatic vegetation (SAV) and hypoxia (Bricker et al. 1999, Nixon et al. 1995). A significant influence on formulation of these conclusions comes from a lack of published hypoxic oxygen concentrations for Narragansett Bay outside of the tidal Providence River and the known physical characteristics of this estuary, which has been classified as “well mixed” to “partially mixed/weakly stratified” due to low freshwater inflow and significant tidal and wind response of the system (Kremer and Nixon 1978, Nixon et al. 1995, Pilson 1985, Ries 1990, Weisberg 1976, Weisberg and Sturges 1976). Survey results presented here, as well as other recent work, indicate that expression of symptoms of nutrient overload are greater than previously recognized. Narragansett Bay has experienced significant loss of SAV (Zostera marina L.) (Doherty 1997, Kopp et al. 1997) over the last 50–75 years. Reviews of available benthic-community data as well as a sedimentprofile camera study and reports have suggested that parts of the upper half of Narragansett Bay are experiencing impacts from excess organic loading, including increased incidence of opportunistic benthic species, degraded benthic habitat quality, significant growth of nuisance macroalgae, and bouts of low dissolved oxygen (which may reach hypoxic levels) since at least the mid- to late 1970s (Deacutis 1999, Frithsen 1990, Germano and Rhoads 1989, Granger et al. 2000, Valente et al. 1992). In addition, fish kills in some Mid-Bay areas have been attributed to probable severe hypoxia or anoxia by the state environmental agency over the last two decades (Deacutis 1999), and in August 2003, one of the largest fish kills recorded for Narragansett Bay was clearly associated with anoxia (Rhode Island Department of Environmental Management 2003). Hypoxia can have a wide range of negative impacts on the biological community. Severe hypoxia is associated with fish kills and mass mortality of benthic invertebrates (Baden et al. 1990, Breitburg 2002, Diaz 2001, Diaz and Rosenberg 1995, Lu and Wu 2000, Wu 2002) and can have a structuring influence on depth-specific zones for benthic communities (Gray et al. 2002, Rosenberg et al. 1992). Even moderate hypoxia can reduce growth rates of marine organisms, cause shifts in the benthic and pelagic community structure, and alter predator-prey interactions (Breitburg 2002, Breitburg et al. 2006 C.F. Deacutis, D. Murray, W. Prell, Emily Saarman, and L. Korhun 175 1994, Diaz and Rosenburg 1995, Pihl 1994, Pihl et al. 1992, Rosenberg et al. 1992). Where hypoxia is a recurrent problem, benthic and pelagic communities tend to shift dominance from large, long-lived species to more tolerant or opportunistic, short-lived species (Diaz 2001, Diaz and Rosenberg 2001, Pearson and Rosenberg 1978). This paper summarizes results from near-synoptic surveys of the upper (northern) half of Narragansett Bay for August 15, 2001 and August 6, 2002 and provides estimates of the geographic extent of hypoxic waters for those dates (here defined as waters with oxygen levels 􀂔 3.0 mg l-1). These surveys provide a contrast of hypoxia distribution across the upper half of the Bay for a typical average monthly summer river flow (July–August 2001), when nonpoint runoff as well as point-source nutrient loading might be expected to occur, and a severe drought period (July–August 2002), when nonpoint load was likely restricted but the significant point-source loading of nutrients from the major wastewater treatment plants continued. The July–August (2-month average) river flow for 2001 was typical for this summer period, ranking 48th out of 75 consecutive years in the flow record, while July–August 2002 flow had the 5th lowest flow out of the 75- year record (United States Geological Survey 2003). The results of these evening oxygen surveys provide evidence that wide areas of the upper half of Narragansett Bay are subjected to intermittent periods of hypoxia during summer months, with probable ecological consequences to benthic communities in these areas. Materials and Methods Approximately 75 stations were selected in the upper half of Narragansett Bay to examine the vertical water column structure of the upper Bay across wide areas. Stations were all located north of Conanicut Island, and placement was decided based on bathymetry, with a mix of deep (> 7.5 m) and shallow (< 7.5 m) stations, and with a mean minimum distance between stations of 0.87 km ± 0.44 km (Fig. 1). Monthly surveys were conducted by a multi-institutional group of volunteer scientists using 6–7 small (5.8–8.5 m) boats between midnight and 7:00 AM in July, August, and September in 2001 and June, July, August, and September in 2002. Survey dates were always chosen to coincide with projected weak neap tides when physical conditions were most conducive to onset of hypoxia (warm water, stratified water column, evening hours). Results presented here are for August 15, 2001 and August 6, 2002. Each boat was preassigned between 10 and 13 stations, and a global positioning system was used to navigate to each station. Vertical water column profiles of salinity, temperature, and dissolved oxygen were taken at pre-specified depths at each station using a hand-deployed, calibrated multiparameter water quality sonde attached to a weighted electronic data transfer cable and digital logger. 176 Northeastern Naturalist Vol. 13, Special Issue 4 High sensitivity teflon membranes were used for oxygen sensors, and all sondes included non-vented depth (pressure) sensors. Most oxygen sensors (6 of 7 used) were of the YSI rapid-pulse design. One sensor was a Hydrolab, with the standard membrane and standard Clark-type oxygen sensor design. All sensors were calibrated according to manufacturers’ recommended operating procedures prior to surveys, and calibration checks were recorded during surveys. Oxygen was calibrated to partial pressure of oxygen in 100% water vapor saturated air. Figure 1. Upper half of Narragansett Bay, RI. Stations in Narragansett Bay used in the 2001–2002 dissolved oxygen evening volunteer surveys are shown, along with geographic points of reference, the Providence River ship channel, and areas > 10.7 m depth. 2006 C.F. Deacutis, D. Murray, W. Prell, Emily Saarman, and L. Korhun 177 Data sheets were transcribed to electronic spreadsheets, and a geographic information system was used to produce maps of the oxygen-concentration distribution to illustrate spatial patterns of hypoxic waters across the Bay. Contours of DO distribution for August 15, 2001 and August 6, 2002 were drawn in ArcMap, using 2nd-order inverse-distance weighting for interpolation between nearest 5 neighbors within a 65º angled ellipse 1524 m wide × 2743 m long, with no minimum of neighboring data points. Separate distribution maps were developed for bottom DO concentrations and oxygen-minima values for each survey date. The areal extent (km2) of hypoxic bottom waters (􀂔 3.0 mg l-1) and oxygen minimum layer < 3 mg l-1 for each area surveyed was estimated from the kriged surfaces shown in the oxygen-distribution maps. Percent of area hypoxic was based on the area surveyed. Vertical density stratification strength was depicted by the proxy of 􀀶􀁭t (bottom 􀁭t–surface 􀁭t ) to examine that parameter in relation to hypoxia occurrence. Densities (􀁭t) based on the equation of state (UNESCO 1981) were calculated for each depth using salinity and temperature readings. Sigma-T units are defined as the difference in g l-1 between the density of the water in question and the density of fresh water at 4 ºC (= 1000 g l-1). Results were mapped by interpolation between stations using the same inverse-distance weighting method as above for oxygen. Results General patterns of hypoxia The geographic distribution of low-DO waters exhibited a consistent pattern for most summer neap-tide surveys conducted since 1999 (C.F. Deacutis, unpubl. data, Narragansett Bay Volunteer DO Monitoring Program). Surface water for the northern half of the tidal Providence River was generally saturated with respect to oxygen, while in the lower Providence River and Upper Bay, higher salinity surface water usually remained between 4–6 mg l-1 DO, although these areas were sometimes supersaturated due to heavy algae blooms (Tables 1–3). The Upper Bay, upper West Passage to Quonset Point, upper East Passage to Popasquash Point, mid- and western portion of Greenwich Bay, and northeastern areas of Mount Hope Bay and the vicinity of the Fall River sewage treatment plant on the southeastern shore of Mount Hope Bay all experienced hypoxic (􀂔 3 mg l-1) to severely hypoxic (􀂔 1 mg l-1) dissolved-oxygen levels during our surveys. Hypoxic water, when it occurred, was found just below the pycnocline in specific areas described below (Table 3). In the August 15, 2001 survey, three regions of subpycnoclinal severe hypoxia (􀂔 1.0 mg l-1) were apparent: the northern Providence River, western Greenwich Bay, and the vicinity of the mouth of the Lee and Cole Rivers in Mount Hope Bay. Most of the Providence River was hypoxic, with the most severe hypoxia (􀂔 1.0 mg l-1) in a shallow (3–5 m) oxygen minimum just below a shallow pycnocline (Table 1, Fig. 2). This oxygen minimum layer remained hypoxic, and extended 5 km south of Conimicut Point, from the 178 Northeastern Naturalist Vol. 13, Special Issue 4 Table 1. Summary of physical data for Narragansett Bay regions for the August 15, 2001 survey. Mean depth and mean dissolved oxygen (DO) concentration and range are indicated for the oxygen minima. DO Station Temperature (°C) Salinity (psu) Density (s) (mg l-1) Oxygen minima depth Surface Bottom Surface Bottom Surface Bottom Surface Bottom Depth (m) mg l-1 (m) Providence River Average 22.1 20.0 19.9 30.2 12.7 21.1 5.0 2.3 3.5 1.5 8.4 n = 13 Range 21.3–23.0 18.5–22.6 5.5–26.7 28.6–31.0 2.1–17.9 19.4–22.1 4.0–6.2 0.4–4.0 2.9–5.0 0.1–3.2 2.9–14.2 Upper Bay Average 22.5 19.2 26.9 30.9 17.9 21.8 6.9 3.3 6.0 2.6 9.4 n = 13 Range 22.0–23.3 17.6–21.3 24.3–29.0 30.3–31.2 16.0–19.6 21.0–22.4 5.8–9.0 1.6–5.1 2.5–10.5 1.6–5.0 3.2–14.0 East Passage Average 22.8 18.8 29.0 30.9 19.4 21.9 7.6 4.6 8.6 4.4 11.7 n = 6 Range 22.4–23.2 17.5–21.2 28.8–29.1 30.1–31.3 19.3–19.7 20.7–22.5 6.5–8.8 3.0–5.5 5.3–12.3 3.0–5.2 5.3–24.1 West Passage Average 22.5 19.5 29.0 30.9 19.5 21.7 8.6 4.8 8.1 4.4 10.2 n = 15 Range 21.7–23.0 17.7–22.5 28.0–29.9 28.9–31.5 18.7–20.4 19.4–22.7 6.3–10.9 2.2–8.2 3.5–18.1 2.2–8.2 3.5–20.0 Greenwich Bay Average 23.3 22.7 26.5 29.4 17.4 19.8 7.7 2.3 3.8 2.3 4.4 n = 11 Range 22.5–24.0 21.0–24.2 23.3–27.4 27.5–30.4 15.2–18.1 18.0–21.0 4.4–10.1 0.1–7.7 0.1–9.1 0.1–7.5 2.5–11.2 Mount Hope Bay Average 23.4 20.5 28.3 30.3 18.8 21.0 6.3 4.0 6.9 3.9 8.6 n = 18 Range 22.2–24.6 17.9–24.3 26.4–29.5 28.7–31.3 17.6–19.9 18.8–22.5 4.2–7.5 0.7–5.5 2.7–11.7 0.7–5.5 2.7–22.7 2006 C.F. Deacutis, D. Murray, W. Prell, Emily Saarman, and L. Korhun 179 Table 2. Summary of physical data for Narragansett Bay regions for the August 6, 2002 survey. Mean depth and mean dissolved oxygen (DO) concentration and range are indicated for the oxygen minima. DO Station Temperature (°C) Salinity (psu) Density (s) (mg l-1) Oxygen minima depth Surface Bottom Surface Bottom Surface Bottom Surface Bottom Depth (m) mg l-1 (m) Providence River Average 26.3 23.2 26.7 30.7 16.7 20.6 6.1 1.1 7.0 1.1 8.6 n = 15 Range 25.5–26.9 21.3–26.7 17.8–28.6 28.5–31.9 9.9–18.2 17.9–22.1 3.6–8.3 0.4–2.8 1.5–10.6 0.4–2.8 1.5–13.9 Upper Bay Average 26.1 23.3 30.4 31.2 19.5 20.9 6.0 2.3 6.2 1.9 9.2 n = 13 Range 25.1–27.6 20.1–27.3 29.0–31.5 30.2–31.8 18.4–20.6 19.0–22.2 3.9–7.0 0.2–6.7 0.3–12.1 0.1–6.4 3.0–13.8 East Passage Average 25.2 22.3 31.8 32.2 20.9 22.0 6.9 2.7 8.4 2.7 10.0 n = 5 Range 24.3–26.0 20.1–24.1 31.6–31.9 32.0–32.6 20.6–21.2 21.4–22.9 6.5–7.2 0.2–4.4 4.7–15.5 0.2–4.4 4.7–23.5 West Passage Average 24.9 23.0 30.8 31.3 20.2 21.1 5.6 2.4 7.6 2.4 7.8 n = 13 Range 23.6–25.7 21.1–25.0 30.4–31.6 30.7–31.8 19.7–20.8 20.5–21.9 3.5–7.8 1.0–5.5 1.5–15.3 1.0–5.5 1.5–15.3 Greenwich Bay Average 26.0 25.1 31.3 31.4 20.3 20.6 5.1 3.5 9.5 3.4 10.9 n = 2 Range 25.9–26.0 25.1–25.2 31.3–31.3 31.4–31.4 20.2–20.3 20.6–20.6 4.8–5.5 3.3–3.7 7.5–11.5 3.1–3.7 10.2–11.5 Mount Hope Bay Average 26.0 23.8 31.5 32.0 20.4 21.4 6.3 3.9 10.2 3.8 10.2 n = 9 Range 24.4–27.2 22.0–25.8 31.1–31.9 31.4–32.2 19.7–21.2 20.4–22.1 4.6–6.9 1.8–5.1 4.8–15.9 1.8–5.1 5.2–15.9 180 Northeastern Naturalist Vol. 13, Special Issue 4 Table 3. Area (km2 and %) of hypoxia in bottom waters and oxygen minimum layer for each Bay segment for each survey, and total area surveyed for each survey date by Bay segment. All areas based on kriged surface area from oxygen-distribution maps. Classification scheme: the upper value of each category is included in the category; thus 1–2 mg l-1 is >1 and 􀂔 2 mg l-1. Area (km2) Area (km2) Area (km2) Total area (km2) % of area Total area (km2) = 1 mg l-1 1–2 mg l-1 2–3 mg l-1 hypoxic* hypoxic* surveyed 8-15-01 8-6-02 8-15-01 8-6-02 8-15-01 8-6-02 8-15-01 8-6-02 8-15-01 8-6-02 8-15-01 8-6-02 Providence River Minimum 4.3 13.2 5.4 5.9 8.7 0.7 18.4 19.8 94 100 19.6 19.8 Bottom 0.5 12.1 2.7 6.9 7.7 0.8 10.9 19.8 56 100 19.6 19.8 Upper Bay Minimum - 28.4 10.9 2.6 18.6 3.0 29.5 34.0 72 87 40.8 39.1 Bottom - 19.8 1.4 9.5 18.7 3.6 20.1 32.9 50 84 40.8 39.1 East Passage Minimum - 6.4 - 1.1 - 3.6 - 11.1 0 61 23.2 18.2 Bottom - 6.0 - 1.4 - 3.7 - 11.1 0 61 23.2 18.2 West Passage Minimum - - - 15.3 6.4 10.5 6.4 25.8 12 58 54.8 44.4 Bottom - - - 15.3 5.7 10.2 5.7 25.5 10 58 54.8 44.1 Greenwich Bay Minimum 2.7 - 1.8 - 2.4 - 6.9 - 66 0 10.4 2.7† Bottom 2.7 - 1.8 - 2.3 - 6.8 - 65 0 10.4 2.6† Mount Hope Bay Minimum 0.3 - 2.6 0.1 2.2 2.6 5.1 2.7 15 14 34.7 19.3‡ Bottom 0.3 - 2.6 0.1 2.0 2.5 4.9 2.6 14 13 34.7 19.4‡ Total Minimum 7.3 48.0 20.7 25.0 38.3 20.4 66.3 93.4 36 65 183.5 143.6 Bottom 3.5 37.9 8.5 33.2 36.4 20.8 48.4 91.9 26 64 183.5 143.6 - Indicates 0. * Definition of hypoxia = 3 mg l-1. † Limited coverage of Greenwich Bay on Aug. 6, 2002—only two stations (at the mouth of the Bay) surveyed. ‡ Limited coverage of Mount Hope Bay on Aug. 6, 2002 —only the lower half of the Bay surveyed. 2006 C.F. Deacutis, D. Murray, W. Prell, Emily Saarman, and L. Korhun 181 Figure 2. Distribution of oxygen concentrations across the upper half of Narragansett Bay on August 15, 2001: left (a) shows minimum oxygen-layer concentration (mg l-1); right (b) shows bottom oxygen concentrations (mg l-1). a b 182 Northeastern Naturalist Vol. 13, Special Issue 4 Providence River into the Upper Bay. Hypoxia was asymmetrically distributed between the east and west sides of the Upper Bay, with a severely hypoxic (􀂔 2 mg l-1) shallow (2.5–6 m) oxygen minimum concentrated on the western side, and a more extensive shallow (5–6 m) hypoxic oxygen minimum extending to the east, together covering 72% of the Upper Bay (Table 3, Fig. 2). Hypoxic bottom water covered approximately half of the Upper Bay, and extended 3 km south into the West Passage (Table 3, Fig. 2a). Hypoxia did not reach the East Passage on this survey date. Over 65% of Greenwich Bay experienced near-bottom hypoxia, with the most severe levels (􀂔 1.0 mg l-1) occurring in the western half. Mount Hope Bay had only 15% of the surveyed area at hypoxic levels, all centered on the mouths of the Lee and Cole Rivers (Table 3, Fig. 2). The August 6, 2002 survey was characterized by higher salinities and warmer temperatures for the entire Bay compared with the August 2001 survey (Tables 1 and 2). July and August 2002 were extremely hot and dry, with only 9.14 mm rainfall recorded in Providence 4 days prior to the survey, and no significant rainfall for the previous month (NOAA 2002). River flows were at record lows, and estuarine circulation was likely minimal. During this particular weak neap tide, the area experienced extremely warm air temperatures (> 32 ºC for previous 􀂧 7 days). For the August 6, 2002 survey, waters below 5 m were severely hypoxic (< 1.0 mg l-1), covering the entire length of the Providence River, and a zone of severe hypoxia below 5 m, with a near-anoxic oxygen minimum between 6 and 9 m depth, extended across the Upper Bay in a northwest to southeast direction all the way to the southernmost station off Popasquash Point in the upper East Passage (Fig. 3). Almost 87% of the Upper Bay experienced hypoxia, most of this < 1.0 mg l-1 (Table 3). Salinity and temperature values for the two stations sampled at the mouth of Greenwich Bay showed them to be well mixed, and only near-hypoxic at the bottom. These stations were located near a strong vertical mixing zone. Available data from continuous oxygen-monitoring sondes deployed at several sites within Greenwich Bay for a benthic study all indicated that the inner area of this subembayment was experiencing severe bottom hypoxia at the time of this survey (Cichetti et al., in press). Data from the limited coverage of Mount Hope Bay in August 2002 exhibited some patches of hypoxic waters, especially along a zone near the eastern shore, but most of the lower half of Mount Hope Bay was not hypoxic (Fig. 3). No August 2002 data were available for the northeastern end for this survey. Because of the limited survey data (2 dates) and spatial resolution, discussion of differences in areal extent of hypoxia from a statistical approach is not possible. However, these limited data sets suggest that the total area in the upper half of Narragansett Bay impacted by hypoxia is highly variable. The August 6, 2002 hypoxic event for the Providence River and Upper Bay had several unique aspects compared with previous surveys. The areal extent and 2006 C.F. Deacutis, D. Murray, W. Prell, Emily Saarman, and L. Korhun 183 a b Figure 3. Distribution of oxygen concentrations across the upper half of Narragansett Bay on August 6, 2002: left (a) shows minimum oxygen layer concentration (mg l-1); right (b) shows bottom oxygen concentrations (mg l-1). 184 Northeastern Naturalist Vol. 13, Special Issue 4 amplitude (severity) (Table 3) were greater than any previously recorded since 1999 (C. Deacutis, unpubl. data). Hypoxic bottom water and a hypoxic oxygen-minimum zone encompassed an estimated 92 km2 (64%) and 93 km2 (65%) of the upper half of Narragansett Bay, respectively, for the August 6, 2002 survey compared with hypoxic zones of 48 km2 (26%) of bottom waters and 66 km2 (36%) of oxygen minima in the August 15, 2001 survey. Severely hypoxic waters extended from the pycnocline all the way to the bottom for the August 6, 2002 survey, with oxygen minima 􀂔 2 mg l-1 covering ca. 96% of the Providence River and 79% of the Upper Bay, and a similar impacted nearbottom area 􀂔 2 mg l-1 of 96% of the Providence River and 75% of the Upper Bay near-bottom waters. In contrast, the majority of severely hypoxic waters (􀂔 2 mg l-1) on August 15, 2001 were found in a shallow (3–6 m) oxygen minimum (50% of the Providence River and 27% of the Upper Bay), with lesser impact on the bottom (16% of the Providence River and 3% of the Upper Bay). Most bottom hypoxic waters were between 2 and 3 mg l-1 for these areas in the August 15, 2001 survey (Table 3). To further explore the relationship of stratification strength to severity of hypoxia, we used the difference between bottom and surface density (􀀶􀁭t ) to provide a measure of stratification resistance to mixing. Differences in salinity between surface and bottom waters have a strong influence on density-driven stratification, especially in the Providence River. Kriged results were mapped for the August 2001 and 2002 surveys. Stratification strength followed the surface salinity gradient in the Providence River and Upper Bay (Fig. 4a). Density differences were much weaker for the August 6, 2002 survey across the Bay due to minimal freshwater flows during that drought period. Area specific conditions Providence River. On August 15, 2001, Bay waters were quite warm at both top and bottom (Table 1). The Providence River, especially the northern end, exhibited a shallow (2–3 m), well-defined pycnocline and strong halocline associated with a recent (< 48 h prior) significant rainfall event (> 2.5 cm) in the urban Providence area. The region of maximum DO depression was located in the upper Providence River, from the hurricane barrier to about 1 km south of Fields Point, with values below 1.5 mg l-1 (Table 1, Fig. 2). An oxygen minimum occurred just below the shallow pycnocline, spreading across shallow (2–5 m) flats just south of Fields Point (Fig. 2a), with bottom waters reaching minimum DO concentrations (􀂔 0.8 mg l –1) near the sediment/water interface (4–5 m). Oxygen concentrations increased slightly with depth below the oxygen minima layer, but bottom waters remained hypoxic (< 3 mg l-1) for almost three-quarters of the Providence River’s length and 56% of its area (Table 3, Fig. 2b). In contrast to the shallow oxygen minima of August 15, 2001, on August 6, 2002, DO decreased gradually starting at near-hypoxic levels even at the surface (3.5–3.8 mg l-1), reaching hypoxic levels by 􀂧 2–3 m depth and dropping slowly and steadily to severe hypoxic concentrations (< 0.5–1.3 2006 C.F. Deacutis, D. Murray, W. Prell, Emily Saarman, and L. Korhun 185 mg l-1) in near-bottom waters (Table 2, Figs. 3a,b). A pycnocline began to form north of Field’s Point at 􀂧 3.5–4.5 m, increasing in depth down the Bay, being driven mainly by temperature differences between surface and bottom. Oxygen levels at the surface just south of Field’s Point were super-saturated, then returned to normoxic levels at the southern reach of the river. The depth where hypoxic oxygen levels were first encountered slowly increased from 2–3 m at the hurricane barrier to 􀂧 4 m mid-river, reaching 􀂧 5 m near the mouth of the river. Oxygen levels immediately below this depth rapidly reached severe hypoxic levels of < 1 mg l-1 by 4.5 m depth for the northern reach, increasing to 7–7.5 m depth for the rest of the river, so that severely hypoxic to near-anoxic water filled the ship channel for the entire span of the river (Tables 2 and 3, Fig. 3). The thickness of this hypoxic layer varied, ranging from 5–8 m thick in the northern reach of the river to 9–10 m down at Gaspee Point and 5–6 m at the southernmost end, with severely hypoxic water (< 1 mg l-1 ) encompassing most of this volume. The shallow flats below Field’s Point and Bullock’s Cove were both hypoxic, but in contrast to August 15, 2001, oxygen concentrations were not severely hypoxic (2–3 mg l-1) compared with the deeper waters in the ship channel. Upper Bay and East and West Passage. On August 15, 2001, a mid-depth (4.5–6 m) hypoxic (1.5–3 mg l-1oxygen) water mass extended 5 km south beyond Conimicut Point on the eastern side of the Upper Bay, underlying a shallow (3 m), warm, low-salinity lens exiting the lower Providence River. Bottom waters in the ship channel below this were between 3.1 and 4 mg l-1 (Table 1, Fig. 2). Vertical density stratification was weaker along the western side of the Upper Bay (Fig. 4a), with a less clearly defined pycnocline gradually becoming deeper (5–7 m) as one moved south. The western Upper Bay had a severely hypoxic (< 2 mg l-1) water mass extending from 􀂧 4.6 m to the bottom. Part of this water mass entered a deep hole at Warwick Neck, slightly increased in DO concentration (hypoxic to near-hypoxic) from 􀂧 4 m to the bottom, and continued into the upper West Passage for 􀂧 3 km. The East Passage did not experience hypoxia for the August 15, 2001 survey (Tables 1 and 3, Fig. 2). On August 6, 2002, weak density differences across most of the Bay were due mainly to a thermocline, with minimal salinity differences between surface and bottom waters (Table 2, Fig. 4b). Hypoxia encompassed almost the entire Upper Bay and the upper East Passage to the southernmost station at Popasquash Point (Table 3, Fig. 3). Hypoxic waters began at 􀂧 7 m depth in the ship channel at Conimicut Point, with severe hypoxic waters (< 2 mg l-1) below 7.5–8 m depth, and extended south 􀂧 2.6 km along the ship channel. At this point, slightly cooler, more saline hypoxic (2–3 mg l-1) bottom water in the lower East Passage ship channel was intercepted, forcing the pycnocline to rise slightly, raising the entire severely hypoxic oxygen zone up to 􀂧 5.5 m (seen as the < 1 mg l-1 hypoxic 186 Northeastern Naturalist Vol. 13, Special Issue 4 Figure 4. Density differences between surface and bottom (Delta Sigma T = bottom density - surface density) for: left (a) August 15, 2001; and right (b) August 6, 2002. a b 2006 C.F. Deacutis, D. Murray, W. Prell, Emily Saarman, and L. Korhun 187 zone above the ship channel area in Fig. 3a). The hypoxic layer in the ship channel was 6–8 m thick, extending from just below the weak pycnocline to the bottom of the channel throughout the eastern Upper Bay and upper East Passage, with most of this volume < 2 mg l-1 DO. An unusual mid-water near-anoxic layer (0.6–0.2 mg l-1) occurred in the Upper Bay at 􀂧 6–8 m depth. Waters below 8 m exhibited a minor increase in DO, although this water was still severely hypoxic (< 2 mg l-1). This nearanoxic layer was 1–2 m thick and extended northwest to southeast across the Upper Bay for 􀂧 5 km, covering most of the Mid Bay and the upper East Passage between 6–8 m depth (Fig 3a), intercepting the sediment-water interface where it coincided with the depth of this layer. Stratification and associated hypoxia disappeared at several stations located near the confluence between upper West Passage, Greenwich Bay, and the Upper Bay (Fig. 3). A deep (12 m) station immediately east of Warwick Neck and the two deep (21 m and 12 m) stations directly south of Warwick Neck (entering the West Passage) were all vertically well mixed, with no discernable pycnocline, having uniform temperature and salinity. These stations were near-hypoxic at the bottom, and exhibited surface oxygen concentrations depressed by 1–2 mg l-1 compared with other Upper Bay stations. Hypoxia in the West Passage began 􀂧 2 km south of the well-mixed area at Warwick Neck, occurring in a layer starting at 􀂧 4.5 m depth, and extending to the bottom, with the vertical extent depending upon bottom depth, ranging from 0.5–3.5 m thick. The stations just north and east of Quonset Point had the lowest DO readings, with severe hypoxia between 1–2 mg l-1 extending from 6–7 m depth to the bottom (Fig. 3b). Greenwich Bay. Greenwich Bay, a small (11.5 km2), shallow (2.6-m mean depth) subembayment on the west side of Narragansett Bay, receives little freshwater input. During the August 15, 2001 survey, stratification across Greenwich Bay was very weak, with minor differences in vertical density (Table 1, Fig. 4a), driven mainly by a small salinity differential of 1–2 psu. Bottom waters across the western side of Greenwich Bay were hypoxic, with bottom DO concentrations in Apponaug Cove and out to the center of Greenwich Bay reaching near-anoxic levels (􀂧 0.75 mg l-1) at the sediment surface. Bottom waters in inner Greenwich Cove to the southwest were not hypoxic due to supersaturated surface water levels from an algae bloom, but dropped rapidly to < 0.75 mg l-1 at the mouth of this cove (Table 3, Fig. 2). Surface waters of the mid- and outer portions of Greenwich Bay were also supersaturated, likely due to algae blooms common in this subembayment. Despite minimal vertical density structure, DO concentrations fell off sharply near the bottom sediments. Due to equipment failure, we were unable to take any readings far inside Greenwich Bay for the August 6, 2002 survey. Water column profiles for two stations at the mouth of Greenwich Bay showed only minor differences in salinity and temperature between surface and bottom, with little vertical 188 Northeastern Naturalist Vol. 13, Special Issue 4 density structure (Fig. 4b). These stations are proximate to the strong mixing zone just south of Warwick Neck mentioned above. The Greenwich Bay station closest to Warwick Neck was normoxic (3.7 mg l-1) from 8 m to the bottom (11 m), while the station just inside Greenwich Bay was near-hypoxic (3.1–3.3 mg l-1) from 7 m to the bottom (10 m). Although we did not have data for the inner areas of Greenwich Bay for August 6, 2002, oxygen data was recorded during this period for a separate benthic study using continuous oxygen-sampling instruments deployed at the bottom at 3 stations: one near the western shore of Greenwich Bay, one at mid-Greenwich Bay, and one deployed just south of Greenwich Bay. All these stations experienced severe hypoxic DO levels (< 1 mg l-1) between August 1 and 6, 2002, simultaneous with our August 6, 2002 survey date (Cichetti et al., in press). This suggests that inner Greenwich Bay was likely experiencing severe hypoxic levels during our August 6, 2002 survey, at least for bottom waters. Mount Hope Bay. Overall density differences for Mount Hope Bay were less than those in the Providence River and much of the Upper Bay for the August 15, 2001 survey (Table 1, Fig. 4a). Severely hypoxic to near-anoxic waters (0.7–1.2 mg l-1) in August 2001 were limited to the mouths of the Lee and Cole Rivers, with hypoxic waters (< 3 mg l-1) extending slightly beyond their mouths, and near-hypoxic waters (3.3–4.0 mg l-1) associated with the mouth of the Taunton River (Fig. 2a,b). For August 6, 2002, only the southern half of Mount Hope Bay was surveyed, and patterns of hypoxia were harder to discern. Surface temperatures were very warm for the southwestern stations sampled. Salinities were uniformly high for both surface and bottom, with minor vertical density differences (Table 2, Fig. 4b). Oxygen in the bottom waters at all stations near the mouth of Mount Hope Bay and just outside Mount Hope Bay in the East Passage were slightly depressed, but did not reach hypoxic levels. The only “hot spots” for low oxygen were seen at three stations: a 12-m deep hole just south of Mount Hope, with a bottom layer 􀂧 0.5 m thick between 2–3 mg l-1; and two deep (􀂧 10-m) stations in the ship channel on the southeastern shore (Fig. 3a), just south of the Fall River sewage treatment plant outfall. These latter stations had a hypoxic layer that extended from 􀂧 5 m depth to the bottom (10 m), with bottom oxygen at 1.8–2.2 mg l-1(Fig. 3). Discussion Near-bottom hypoxia occurs when some level of stratification develops, and bottom waters are cut off from atmospheric oxygen, decreasing oxygen concentrations due to water column and benthic respiration, including bacterial respiration during decomposition of organic matter (Diaz 2001). This rate of oxygen loss is accelerated by eutrophication (excess organic production due to excess nutrient loadings [Nixon 1995b]). Concentration of particulate organic matter, duration of stratification, and 2006 C.F. Deacutis, D. Murray, W. Prell, Emily Saarman, and L. Korhun 189 water temperature all play significant roles in the intensity of hypoxia for an area (Breitburg 2002, Diaz 2001). Residence time is another factor (Breitburg 2002), and local (cove/area specific) residence time may be one of the most critical aspects of an area’s risk of hypoxia formation. The near-synoptic surveys conducted on August 15, 2001 and August 6, 2002 corroborate previous reports which had indicated that impacts related to hypoxia and high organic loading were possibly occurring in Upper Narragansett Bay (Deacutis 1999, Frithsen 1990, Germano and Rhoads 1989, Granger et al. 2000, Valente et al. 1992). Our results showed that hypoxic and severely hypoxic zones extend significantly into Upper–Mid Bay areas beyond the previously known hypoxia-impacted area of the tidal Providence River. These areas have previously been characterized as well mixed, having little risk of hypoxia formation. The surveys provide a near-synoptic view of hypoxia in the Bay, but data from a continuous water-quality monitoring buoy system in Narragansett Bay helps to put this information into a temporal context. Based on 15 minute continuous time-series data for temperature, salinity, and dissolved oxygen for a buoy monitoring site off the northwest corner of Prudence Island and a second site in the southern reach of the tidal Providence River north of Conimicut Point, the August 15, 2001 survey occurred early in an extended hypoxic event in the Providence River from August 9 through August 31, 2001 (Bergondo et al., in press). The oxygen levels continued to decrease at the buoy sites following our August 15, 2001 survey, so the 2001 data here probably does not fully portray the final maximum extent of that hypoxic event. In contrast, the August 6, 2002 survey occurred just prior to a complete mixing event due to significant winds later in the day (Bergondo et al., in press), fully depicting the maximum extent of this extremely severe 6- day hypoxic event just prior to its breakup. Given the temporal context provided by time-series data from the same period, hypoxic conditions captured in these snapshot surveys are likely not rare, isolated events, but probably occur with regular frequencies associated with cyclic decreases in tidal energies (neap tides) and warm water temperatures. Analyses of the time-series data (Bergondo et al., in press) is likely to provide a much better understanding of frequency and duration, and the role tides and estuarine circulation play in setting up adequate physical conditions for the generation of these hypoxic events. For the Providence River and Upper Bay, areas of hypoxia match the well-known north–south nutrient and chlorophyll gradients (Oviatt et al. 2002). The Providence River is known to experience seasonal low dissolved- oxygen conditions due to its mesohaline estuarine circulation, entry of significant nitrogen loads from seven upstream urban sewage treatment plants, and the resultant algal growth and die-off (Turner 1997). These waters pass into the Upper Bay through estuarine circulation and tidal advection, providing a clear path for the high organic load to reach beyond the tidal Providence River. 190 Northeastern Naturalist Vol. 13, Special Issue 4 The asymmetrical distribution of hypoxic waters on August 15, 2001 along the western side of Narragansett Bay (Fig. 2a) may be linked to this estuarine circulation. The oxygen minimum at Conimicut Point and the western side of the Upper Bay was shallow to mid-water. If these intermediate- density hypoxic waters are advected south, Coriolis force would deflect this flow to the west, while inflowing cooler, more saline oceanic waters at the bottom would follow the ship channel, producing the observed distribution of hypoxia (Fig. 2). The northwest–southeast distribution of the more extensive and severely hypoxic to near-anoxic layer in the Upper Bay and upper East Passage in the dry August 6, 2002 survey (Fig. 3) seems to confirm this hypothesis since the increased residence time due to lack of estuarine flow would cause stratified, hypoxic waters to advect more slowly out of the Bay, allowing oxygen to continue decreasing and sweeping the water mass slowly down the Bay. A more perplexing area is that of western Greenwich Bay, which experienced severe hypoxia in August 2001 (and other unpublished surveys). Although some hypoxia has previously been documented in parts of Greenwich Bay (Granger et al. 2000), and Valente et al. (1992) found evidence of poor benthic habitat quality, this phenomenon was believed to be limited in its frequency of occurrence, especially considering the minor stratification potential for this shallow subembayment with very limited freshwater input (Granger et al. 2000). The lack of a mid-water oxygen minima in our Greenwich Bay data, and clear association of the most severe hypoxia with the sediment–water interface suggests that hypoxia in this area may be driven by sediment oxygen demand and benthic respiration rates. Another possible factor in hypoxia generation in Greenwich Bay may also involve compass orientation of the major axis of a subembayment in relation to the prevailing winds of the area. Greenwich Bay is an east–west oriented body, with the major exchange point at the eastern end with the West Passage of Narragansett Bay. Prevailing summer winds are out of the southwest for Narragansett Bay. The western side of Greenwich Bay may be particularly susceptible to severe hypoxia development. This area may have poor local flushing (Abdelrhman 2005), possibly related to local topography, including bulkheading and dredging of the small shallow western coves for extensive marina development. Local nutrient loadings are fairly high due to a sewage treatment plant effluent as well as significant nonpoint source loadings from a high density of septic systems and urbanized land use (Granger et al. 2000). Several multi-species fish kills on the western side of Greenwich Bay have occurred following westerly winds during neap tidal periods in 2001 and 2002, and a recent welldocumented large fish kill was directly linked to an anoxic event (Rhode Island Department of Environmental Management 2003). Breitburg (2002) noted that wind-driven upwelling of hypoxic waters and subsequent “jubilees” of organisms trying to escape the hypoxia have been associated with specific wind directions. 2006 C.F. Deacutis, D. Murray, W. Prell, Emily Saarman, and L. Korhun 191 Mount Hope Bay had a limited extent of hypoxia in August 2001, concentrated mainly at the small river mouths (Table 3, Fig. 2b). A 1972– 1973 oxygen study of Mount Hope Bay (Brown University 1973) provides an interesting historical perspective on our results. Results of the Brown University oxygen surveys found the lowest DO readings (0.7–0.9 mg l-1) at the mouth of the Lee and Cole Rivers, the same locale where we observed severely hypoxic levels (0.7–1.2 mg l-1) on August 15, 2001. Both our August 6, 2002 survey and the Brown study found a large hypoxic zone along the southeastern shore by the sewage treatment plant. However, the 1972 Brown University study also documented more extensive low DO (< 3.5 mg l-1) across wide areas of mid-Mount Hope Bay, suggesting that in the early 1970s, Mount Hope Bay may have experienced more extensive low-oxygen events than we recorded. The sewage treatment plant has been significantly upgraded since that study, and dryweather flows from combined sewer illegal connections have since been eliminated, so this is not an unreasonable positive shift in oxygen conditions. However, the limited coverage for the August 2002 survey precludes any definitive statements about this subembayment. Stratification strength and hypoxia Results for both August 2001 and August 2002 strongly suggest that stratification intensity (degree of density difference) is not the most important factor in developing severe hypoxia. In August 2001, the upper Providence River was characterized by a strong vertical density gradient due to surface-to-bottom salinity differences, creating strong densitydriven stratification, and exhibiting a mid-water (subpycnocline) oxygen minimum (Figs. 2b and 4a). However, during this same survey, Greenwich Bay experienced weak density gradients (Fig. 4a), yet exhibited near-bottom severe hypoxic to near-anoxic levels as low as those found in the northern and mid-Providence River (Fig. 2a). In July and August 2002, the Narragansett Bay watershed experienced an extreme drought as well as an extended period of warm air temperatures, with minimal river flows. Such conditions limit typical estuarine circulation and increase residence time for the Bay (Pilson 1985). However, nutrient loadings from nonpoint sources should be lower, although significant point source nutrient loadings would continue relatively unabated. Point sources are estimated to contribute > 65% of the total nutrient loading of Narragansett Bay (RI Governor’s Narragansett Bay and Watershed Planning Commission 2004), most of it entering through the tidal Providence River. All areas of the Bay had a very weak stratification signature based on density differences for August 6, 2002 (Fig. 4b), yet a much larger area of the Upper Bay and upper East and West Passage experienced hypoxic, severely hypoxic, or near-anoxic conditions compared to August 15, 2001 (Table 3, Figs. 2 and 3). These results suggest that decreased flushing, warm water temperatures, and other factors, such as the present nutrient loading from point sources leading to excess organic 192 Northeastern Naturalist Vol. 13, Special Issue 4 production, are more important than the stratification strength for development of severe but intermittent hypoxia in upper Narragansett Bay, at least during weak tidal-energy periods (neap tides). The strength of vertical density gradients may play a role in extending the duration of hypoxia in the Providence River beyond the period when strong spring tides have normally destratified the water column in the Upper Bay and East and West Passages. The Providence River, therefore, has the potential for longer periods of hypoxia than areas further down the Bay. Some areas may be resistant to stratification due to physical mixing energies linked to local hydrographic features. An example would be the area just south of Warwick neck, where a deep (21-m) hole and deep (12- m) Y-shaped natural channel exists (Fig.1). On our August 6, 2002 survey, stations at the mouth of Greenwich Bay and two other stations located just South of that point showed vertical homogeneity in terms of temperature and salinity as well as higher bottom and lower surface oxygen concentrations compared with other nearby stations. Mixing gyres, likely due to such bottom topographic features and their interaction with tidal and wind-driven currents, may play a role in the strong vertical mixing structure seen in such areas. Previous attempts to assess hypoxia and other eutrophication impacts to Narragansett Bay have been based on inadequate monitoring data, both in frequency and timing (Bricker et al. 1999, Desbonnet and Lee 1991, Doering et al. 1990, Olsen and Lee 1979). Most oxygen measurements taken in the Bay have ignored the tidal state (neap vs. spring). Results from our surveys strongly suggest that tidal state is important in assessing the areal extent of hypoxia in the upper half of Narragansett Bay. Recently completed analyses of continuous time-series oxygen data for the Upper Bay (Bergondo et al., in press) suggest that the occurrence of hypoxia in parts of Narragansett Bay appears to follow a tidally periodic, seasonal frequency. If this holds true, then Upper Narragansett Bay exhibits an intermittent form of hypoxia linked tightly to the neap–spring tidal cycle, reported previously in the York and Rappahannock Rivers, VA. (Diaz 2001, Haas 1977). Such short-term cyclic hypoxic events are more difficult to record unless the monitoring system is either continuous or directed towards periods of maximal risk due to physical driving factors, as observed here in this study. Hypoxia is clearly occurring in this partially to well-mixed estuary well beyond the end of the tidal Providence River at Conimicut Point, reaching into the upper half of the West and East Passages, and western Greenwich Bay under certain tidal and meteorological conditions (late in neap summer tides, perhaps also affected by decreased estuarine flows during drought conditions as well as advection following significant rainstorm events in the urban head of the estuary). Future time-series analyses should provide a more definitive understanding of driving factors and absolute frequencies of hypoxia in Upper Narragansett Bay. 2006 C.F. Deacutis, D. Murray, W. Prell, Emily Saarman, and L. Korhun 193 Ecological impacts One important aspect to consider in relation to potential impacts from the observed hypoxic events is the possibility of near-shore, shallow subtidal habitat impacts from shallow subpycnoclinal waters exhibiting severe hypoxic to near anoxic oxygen levels, as seen on August 6, 2002. Such mid-water (6–9 m depth) oxygen minima have not been unusual in our surveys; only the severity was noteworthy in August 2002. Such layers may have impacts on those specific depth zones where the oxygen minimum layer intersects the subtidal habitat. This may produce unusual benthic community structuring as has been observed elsewhere (Rosenberg et al. 1992) when shallow subpycnoclinal benthic populations are altered by lethal hypoxic levels, but deeper communities may escape the lowest oxygen levels. In addition, greater potential exists to advect or upwell these shallower waters into near-shore habitats under the right wind conditions. It is likely that the intermittent but extensive hypoxia events observed in the upper half of Narragansett Bay in the August 15, 2001 survey and especially the August 6, 2002 survey produced significant negative impacts on sensitive benthic species. Such regular intermittent seasonal stress factors are likely to have been influencing the benthic community in Narragansett Bay over the last three decades, based on a review of fish kills, anecdotal information, and sparse available oxygen data (Deacutis 1999), as well as a review of benthic community information for Narragansett Bay (Frithsen 1990) and a sediment-profile camera survey of the area (Valente et al. 1992). Frithsen (1990) noted that a change in the description of the dominant benthic species began to occur starting in the late 1970s for the Mid-Bay region, shifting from a Nepthys-Nucula community to a Mediomastus- Nucula community. Despite the fact that earlier studies had used sieves adequate to capture at least part of the Mediomastus ambiseta Hartman population, there was a complete absence of this species in benthic data lists collected in Mid Bay prior to 1975. However, by 1977, Mediomastus appeared as a dominant species in collections. Frithsen argued that this shift appears to be real, and suggested this may be indicative of greater organic enrichment to the Mid-Bay region. Grassle et al. (1985) also suggested that the Mid-Bay community had undergone change prior to their 1976 benthic collection efforts. Data from a recent Final Environmental Impact Statement Report for the Providence River channel dredging project confirms continued absence of Nepthys, while Mediomastus has become common in the Mid-Bay area (US Army Corps of Engineers 2001). Our studies confirm that hypoxia, one of the consequences of such organic enrichment, does indeed appear in the Upper Bay and even further south today. A sediment profile imaging study for Narragansett Bay (Valente et al. 1992) using organism-sediment index (OSI) values also provided evidence for organic enrichment impacts beyond the recognized degraded habitat in 194 Northeastern Naturalist Vol. 13, Special Issue 4 the Providence River. Their map of the shallow apparent redox potential discontinuity depth (RPD) for stations in the upper half of Narragansett Bay sampled in August 1988 showed a similar shape to our maps of hypoxia and near hypoxia for August 15, 2001. Conclusions Results presented here indicate that Narragansett Bay is subject to periodic summer hypoxic events that can extend over much of the northern half of Narragansett Bay and, in some areas, may reach severe hypoxic (< 1.0 mg O2 l-1 ) to near-anoxic conditions. Most of the severe impacts occur in the Providence River and Greenwich Bay, but the Upper Bay and the northern reaches of the West and East Passages and limited areas of Mount Hope Bay are experiencing significant exposure to hypoxia under certain tidal and meteorological conditions. Results for the oxygen surveys affirm sedimentprofile camera work and some benthic studies that previously suggested parts of the Mid Bay have become subject to increased organic loading impacts. These impacts can take place even under drought conditions, when only point-source nutrients are the major contributors to nutrient loadings entering the upper half of Narragansett Bay. Acknowledgments Funds for some of the equipment used for these studies were provided through a grant to the Rhode Island Department of Environmental Management (RIDEM) from the National Oceanic and Atmospheric Administration (NOAA), and funds for coordination were provided by the Narragansett Bay National Estuary Program. We would like to thank Don Pryor, Candace Oviatt, and the manuscript reviewers who provided thoughtful suggestions and comments. We are grateful for GIS help from Paul Jordan of RIDEM and Jeff Albert and Lynn Carlson of Brown University, and computation from Philip Howell of Brown University. Our surveys have been dedicated to two of our close colleagues, Dr. Mark Gould of Roger Williams University and Dr. Dana Kester of the University of Rhode Island, both of whom sadly passed away during these studies. We gratefully acknowledge donated labor and use of YSI equipment and/or boats from: Derek Koloski of NE3, Roger Race of Endico/YSI Inc., and the United States Environmental Protection Agency (USEPA) Atlantic Ecology Division and New England Regional Labs; and donated boats, equipment, and support from: RIDEM; Candace Oviatt, Stan Cobb, and RISeaGrant of URI; Warren Prell of Brown University; Save The Bay, Inc.; Tim Scott of Roger Williams University; and Christian Krahforst of Massachussetts Bays Program. We are indebted to all the members of “The Insomniacs:” the volunteers who donated time and use of boats, and who were willing to lose sleep to join us in these oxygen surveys. The following is an abridged list (excluding the authors) of the members of the Narragansett Bay Estuary Program volunteer evening dissolved oxygen monitoring efforts (1999–2003): Andrew Altieri, Steve Clemens, John LaRiviere, Joseph Orchado, Don Pryor, and many other colleagues and students at Brown University; James Shine and crew from the Harvard School of Public Health; Christian Krahforst and his crew from Massachussetts Bays Program; Ali Armstrong, Taylor Ellis, and others from the Bay Commission; Diane Ferland, Rich Ribb and 2006 C.F. Deacutis, D. Murray, W. Prell, Emily Saarman, and L. Korhun 195 others of the Narragansett Bay Estuary Program; Tom Kutcher, Ken Raposa, Patty Richardson and friends from the Narragansett Bay National Estuarine Research Reserve; Megan Higgins of Rhode Island Coastal Resources Management Council; Ed Everich, Najih Lazar, Joe Migliore, Scott Olszewski, and many others at RIDEM; Brad Bourque, Stephen O’Shea, Skip Pomeroy, Andrew Tate, Paul Webb, and their many colleagues and students at Roger Williams University; John Torgan, Topher Hamblett, and Andrew Lapisky of Save The Bay, Inc.; Chris Calabretta, Chris Melrose, Laura Reed, Kim Whitman, and many other students of URI; Donald Cobb, Michelle Kraczkowski, Charles Strobel, and their many colleagues at the USEPA AED lab; Tim Bridges, Tom Faber, Jerry Keefe, and colleagues at the USEPA New England Regional Lab; Margherita Pryor of the USEPA NE Region; and Tom Halavik of the US Fish and Wildlife Service. Another Box O’Joe coming up! Literature Cited Abdelrhman, M. 2005. Simplified modeling of flushing and residence times in 42 embayments in New England, USA with special attention to Greenwich Bay, Rhode Island. Estuarine and Coastal Shelf Science 62:339–351. Baden, S.P., L.O. Loo, L. Pihl, and R. Roseberg. 1990. Effects of eutrophication on benthic communities including fish: Swedish west coast. Ambio 19:113–122. Bergondo, D., D.R. Kester, H.E. Stoffel, and W L.Woods. In press. Time-series observations during the low sub-surface oxygen events in Narragansett Bay during summer 2001. Marine Chemistry. Bricker, S.B., C.G. Clement, D.E. Pihalla, S.P. Orlano, and D.R.G. Farrow. 1999. National Estuarine Eutrophication Assessment : Effects of nutrient enrichment in the nation’s estuaries. NOAA, National Ocean Service, Special Projects Office and the National Centers for Coastal Ocean Science, Silver Spring, MD. 71 pp. Breitburg, D.L. 2002. Effects of hypoxia, and the balance between hypoxia and enrichment, on coastal fishes and fisheries. Estuaries 25(4b):767–781 Breitburg, D.L., N. Steinburg, S. DuBeau, C. Cooksey, E.D. Houde. 1994. Effects of low dissolved oxygen on predation on estuarine fish larvae. Marine Ecology Progress Series 104:235–246. Brown University. 1973. Report on Mount Hope Bay 1972–1973. Unpublished report funded by the NSF Student Originated Studies Program. Brown University Library. 192 pp. Chinman, R.A., and S.W. Nixon. 1985. Depth area-volume relationships in Narragansett Bay. URI Graduate School of Oceanography, Narragansett, RI, Marine Technical Report. No. 87. 64 pp. Cicchetti, G., J. Latimer, S. Nelson, B. Burgen, S. Rego, and L. Coiro. In press. Links between measured near-bottom oxygen and sediment profile camera measurements in coastal Rhode Island, USA. Journal of Marine Systems. Deacutis, C. 1999. Nutrient impacts and signs of problems in Narragansett Bay. Pp. 7–23, In Kerr, M. (Ed.) . 1999. Nutrients and Narragansett Bay: Proceedings of a Workshop on Nutrient Removal from Wastewater Treatment facilities. RI Sea Grant. Narragansett, RI. 64 pp. Desbonnet, A., and V. Lee. 1991 Historical trends in water quality and fisheries resources, Narragansett Bay, Rhode Island . Uniersity of Rhode Island Coastal Resources Center Contribution No. 100 and National Sea Grant Publication # RIUT91- 001 . Graduate School of Oceanography, URI, Narragansett, RI. 101 pp. 196 Northeastern Naturalist Vol. 13, Special Issue 4 Diaz, R.J. 2001. Overview of hypoxia around the world. Journal of Environmental Quality 30(2):275–281. Diaz, R.J., and R. Rosenberg. 1995. Marine benthic hypoxia: A review of its ecological effects and the behavioral responses of benthic macrofauna. Oceanography and Marine Biology Annual Review 33:245-303. Diaz, R.J., and R. Rosenberg. 2001. Overview of anthropogenically-induced hypoxic effects on marine benthic fauna. Pp.129–146, In N.N. Rabalais and R.E. Turner (Eds.). Coastal Hypoxia: Consequences for Living Resources and Ecosystems. Coastal and estuarine Studies No. 58. American Geophysical Union, Washington, DC. 463 pp. Doering, P., C. Oviatt, and B. Walsh. 1990. Characterizing late summer water quality in the Seekonk River, Providence River, and Upper Narragansett Bay. Narragansett Bay Project Report #NBP-90-49. 58 pp. Doherty, A. 1997. Historical distribution of eelgrass in Narragansett Bay, RI. P. 52, In Proceedings of 14th Biennial Estuarine Research Federation International Conference: “The State of Our Estuaries” Oct.12–16, 1997. Providence, RI. Ely, E. 2002. An overview of Narragansett Bay, Rhode Island. RI Sea Grant, Narragansett, RI. 8pp. Frithsen, J. 1990. The benthic communities within Narragansett Bay. NBP Report #NBP-90-28. 90pp. and Appendices A–C. Germano, J., and D. Rhoads. 1989. Narragansett Bay sediment quality survey, August 1988. SAIC final report to Narragansett Bay Project, Report # NBP-89- 23. Narragansett Bay Estuary Program, URI Coastal Institute, Narragansett, RI. 171 pp. Granger, S.L., M. Brush, B. Buckley, M. Traber, M. Richardson, and S.W. Nixon. 2000. An assessment of eutrophication in Greenwich Bay. Paper No. 1, In M. Schwartz (Ed.). Restoring Water Quality in Greenwich Bay: A Whitepaper Series. Rhode Island Sea Grant, Narragansett, RI. 20 pp. Grassle, J.F., J.P. Grassle, L.S. Brown-Leger, R.F. Petrecca, and N.J. Copley. 1985. Subtidal macrobenthos of Narragansett Bay. Field and mesocosm studies of the effects of eutrophication and organic input on benthic populations. Pp. 421–434, In J.S. Gray and M.E. Christiansen (Eds.). Marine Biology of Polar Regions and Effects of Stress on Marine Organisms. Wiley, New York, NY. Gray, J.S., R. Shiu-sun Wu, and Y.Y. Or. 2002. Effects of hypoxia and organic enrichment on the coastal marine environment. Marine Ecology Progress Series 238:249–279. Haas, L. 1977. The effect of the spring-neap tidal cycle on the vertical salinity structure of the James, York, and Rappahannock Rivers, Virginia, USA. Estuarine and Coastal Marine Science 5:485–496. Kopp, B., A. Doherty, and S. Nixon. 1997. A guide to the site selection for eelgrass restoration projects in Narragansett Bay, RI. Final Report to the RI Aqua Fund Council. Rhode Island Department of Environmental Management, Division of Water Resources, Providence, RI..22 pp. and appendix. Kremer, J.N., and S.W. Nixon. 1978. A Coastal Marine Ecosystem: Simulation and Analysis. Springer-Verlag, New York, NY. 271 pp. Lu, L., and R.S. Wu. 2000. An experimental study on recolonization and succession of marine macrobenthos in defaunated sediment. Marine Biology 136:291–302. Nixon, S.W. 1995a. Metal inputs to Narragansett Bay: A history and assessment of recent conditions. RI SeaGrant Publication P1388. 76 pp. 2006 C.F. Deacutis, D. Murray, W. Prell, Emily Saarman, and L. Korhun 197 Nixon, S.W. 1995b. Coastal marine eutrophication: A definition, social causes, and future concerns. Ophelia 41:199–219. Nixon, S.W., S.L. Granger, and B.L. Nowicki. 1995. An assessment of the annual mass balance of carbon, nitrogen, and phosphorus in Narragansett Bay. Biogeochemistry 31:15–61. NOAA. 2002. National Weather Service historical daily precipitation data for Providence, RI. Olsen, S., and V. Lee. 1979. A summary and preliminary evaluation of data pertaining to the water quality of upper Narragansett Bay. University of Rhode Island., Coastal Resources Center, Narragansett, RI. 189 pp. Oviatt, C., A. Keller, and L. Reed. 2002. Annual primary production in Narragansett Bay with no bay-wide winter–spring phytoplankton bloom. Estuarine, Coastal ,and Shelf Science 54(6):1013–1026. Pihl, L. 1994. Changes in the diet of demersal fish due to eutrophication-induced hypoxia in Kattegat, Sweden. Canadian Journal of Fisheries and Aquatic Sciences 51:321–336. Pihl, L., S.P. Baden, R.J. Diaz, and L.C. Schaffner. 1992. Hypoxia-induced structural changes in the diet of bottom-feeding fish and crustacea. Marine Biology 112:349–361. Pilson, M. 1985. On the residence time of water in Narragansett Bay. Estuaries 8:2–14. Ries III, K.G. 1990. Estimating surface-water runoff to Narragansett Bay, Rhode Island and Massachusetts. US Geological Survey, Water Resources Investigations Report 89-4164. 44 pp. Rhode Island Department of Environmental Management (RIDEM). 2003. The Greenwich Bay fish kill, August 2003: Causes, impacts, and responses. 32pp. Available at: http://www.state.ri.us/DEM/pubs/fishkill.pdf. Accessed July 20, 2004. Rhode Island Governor’s Narragansett Bay and Watershed Planning Commission. 2004. Nutrient and Bacteria Pollution Panel Report, March 2004. 33 pp. Available at: http://www.ci.uri.edu/GovComm/Documents/Phase1Rpt/Docs/Nutrient- Bacteria.pdf. Accessed July 20, 2004 Robinson, K., S. Flanagan, J. Ayotte, K. Campo, A. Chalmers, J. Coles, and T. Cuffney. 2004. Water quality in the New England coastal basins, Maine, New Hampshire, Massachusetts, and Rhode Island, 1999–2001. US Geological Survey Circular 1226. 38 pp. Robinson, K., J. Campbell, and N. Jaworski. 2003. Water quality trends in New England rivers during the 20th century: US Geological Survey Water Resources Investigations Report 03-4012. 20 pp. Rosenberg, R.B., L.O. Loo, and P. Moller. 1992. Hypoxia, salinity, and temperature as structuring factors for marine benthic communities in a eutrophic area. Netherlands Journal of Sea Research 30:121–129. Turner, C. 1997. Development of a Total Maximum Daily Loading for nutrients: Results in the Providence and Seekonk Rivers. RIDEM Report, Rhode Island Department of Environmental Management, Division of Water Resources, Providence, RI. Aug. 20, 1997. 51 pp. + appendix. UNESCO 1981. Unesco Tenth report of the joint panel on oceanographic panels and standards. UNESCO technical papers in marine sciences No. 36. 24 pp. 198 Northeastern Naturalist Vol. 13, Special Issue 4 United States Army Corps of Engineers. 2001. Providence River and Harbor Maintenance Dredging Project Final Environmental Impact Statement. Appendix C: Biological Resources. US Army Corps of Engineers New Endland District, Concord, MA. August 2001. United States Geological Survey. 2003. USGS Water Resources of Rhode Island, Surface Water Data for Rhode Island, Monthly Statistics for 1929–2003. Available at: http://waterdata.usgs.gov/ri/nwis/sw. Accessed July 1, 2004. Valente, R.M., D.C. Rhoads, J.D. Germano, and V.J. Cabelli. 1992. Mapping of benthic enrichment patterns in Narragansett Bay, Rhode Island. Estuaries 15:1–17. Weisberg, R.H. 1976. The nontidal flow in the Providence River of Narragansett Bay: A stochastic approach to estuarine circulation. Journal of Physical Oceanography 6(5):721–734. Weisberg, R.H., and W. Sturges. 1976. Velocity observations in the West Passage of Narragansett Bay: A partially mixed estuary. Journal of Physical Oceanography 6(3):345–354. Wu, R.S. 2002. Hypoxia: From molecular responses to ecosystem responses. Marine Pollution Bulletin Article 45:35–45.