E. Thomas, F. Lugolobi and J. C. Varekamp; Department of Earth & Environmental Sciences, Wesleyan University, Middletown CT, USA

INTRODUCTION

Since the early 1970s Long Island Sound (LIS) has experienced dysoxia in bottom waters during summer, when waters are stratified. Hypoxia is most severe close to NYC (western LIS). High population density leads to nutrient (N, P) pollution by waste water treatment plant effluents.This causes eutrophication and creates an oxygen deficit below the pycnocline (5-10 m depth).

Fiugre 1: Hypoxia in LIS, August 2002. Data from CT DEP, Water Quality Monitoring Program; Atmosphere and hypoxia.

 

SAMPLES

We studied core samples to trace the development of eutrophication and hypoxia in LIS over time. Cores were taken in the Eastern (G1C1), Central (D2C3, B1C2, B5C5, B7C1), Western (A1C1, A4C1) Basins and in the Narrows (WLIS68, WLIS75), above and below pycnocline depths. Age models are preliminary, and are based on ²¹⁰Pb and ¹³⁷Cs data as well as data on Hg pollution and correlation to dated marsh cores. In addition, we used a pollen record from core A4C1.

Figure 2: Location of Cores in LIS

WATER MIXING IN LIS:

Oxygen isotopes are conservative and LIS water samples plot on straight mixing lines for salinity and d¹⁸O between rivers and Atlantic sea water. Carbon isotope values of DIC in LIS waters are expected to plot on a mixing curve in a salinity-d¹³C diagram, but the observed values are much lighter as a result of the oxidation of organic matter. Flux averaged mean of CT River and Housatonic River water: d¹⁸O =~-9.5 ^o/_oo, d¹³C = ~-9.5 ^o/_oo.

 Figure 3: Mixing model and observed values, LIS.

VARIATIONS IN d¹³C UNRELATED TO MIXING OF SEA AND RIVER WATER:

The measured d¹³C in foraminiferal tests reflects the d¹³C of DIC in LIS waters, which depends on salinity, and the amount of oxidized C_org added to the waters. We separate these terms as follows:

• Measure d¹⁸O values in foraminiferal tests (Elphidium excavatum).
• Calculate d¹⁸O water, using off set in tests from isotopic equilibrium, ~ + 1.1 ^o/_oo
• Estimate salinity from LIS mixing model (blue line in figure above)
• Estimate predicted d¹³C -DIC using the pink mixing curve in figure above
• Subtract calculated from observed values: d¹³C * = d¹³C _carbonate-d¹³C _mix
• d¹³C* (Excess d¹³C ): indicates how much oxidized organic carbon has been added to the water.

Figure 5: Elphidium excavatum.

Mg/Ca DATA IN FORAMINIFERA

Part 1 (Figure 6).

Mg/Ca values in foraminifera tests that were collected alive (Rose-Bengal stained, SSL samples) do not correlate with bottom water temperatures (Figure 6) measured at the time of collection. Elphidium-species live up to 12 months and the foraminiferal calcite thus reflects a time-integrated Mg/Ca (and temperature ) signal. The observed Mg/Ca values in SSL probably reflect largely the age structure of the foraminifera and not the in situ measured water temperature during collection.

Figure 6: Lack of correlation between Mg/Ca values in 'living' foraminifera and measured temperature at time of collection.

Part 2 (Figure 7).

Mg/Ca data in SSL and core samples show no correlation with d¹⁸O data in the same samples. Calcite d¹⁸O values are influenced by both salinity and temperature. The range in Mg/Ca values in core samples (cores in Figure 7) is similar to that in SSL samples (grabs in Figure 7), indicating the time-integrated character of both sample sets.The core samples represent an average of 5-10 years. Seasonal bottom water temperatures in LIS vary by about 20 ^oC, and Mg/Ca values from core forminiferal calcite samples thus do not provide information on long-term bottom water temperature changes. The core calcite Mg/Ca values provide information on the mass of foraminiferal calcite produced per season, and as such carry paleo-environmental information (work in progress).

Figure 7.

ISOTOPE DATA IN FORAMINIFERA

Part 1 (Figure 8)

Paleo-salinities are obtained from calcite d¹⁸O values using a mean annual temperature of 12.5 °C, an in vivo isotopic effect of 1.1 ^o/_oo and the modern LIS mixing model. Calculated salinities in core WLIS 75 (Figure 8, below) show two lows that correlate with the wet periods of the early 1900s and mid-1950s. On this time scale, d¹³C* values are strongly correlated with the paleo salinities, with light values corresponding to periods of low salinity (enhanced freshwater input). Modern observations show that plankton blooms occur in LIS during and shortly after enhanced run-off.

Figure 8: Record from core WLIS 75.

Part 2 (Figure 9).

Since the mid 19th century, paleo-salinity has varied strongly (18-32 ^o/_oo) (Figure 9, below). Most cores show a salinity dip in the mid 1950s, a regionally wet period with catastrophic floods. The lowest paleo-salinities occur in cores close to the CT shore line and those in westernmost LIS.

Figure 9.

Part 3 (Figure 10).

Paleo-salinity in LIS shows no clear trend over the last 1000 years, but data for the 20^th century show more structure, possibly an artifact of the higher sampling density (higher sedimentation rates). Alternatively,LIS environments (and climate?) were more variable in that period. The d¹³C* values fluctuated between 0.5 and minus1.5 ^o/_oo from 1000 to 1900 AD, but were significantly lower and more variable (extreme low value at minus 5.5 ^o/_oo ) since ~ mid 19^th century. These data indicate that primary productivity in LIS increased dramatically in the 20^th century, and part of the generated C_org was oxidized, part buried in the sediment.

Figure 10.

CONCLUSIONS Paleo-salinity in LIS bottom waters shows no consistent, significant trend with time since ~1000 AD, but there may have been more variability since the mid 19th century. Excess d13C values became significantly lighter since the mid 19th century, indicating enhanced production and oxidation of organic matter. Hypoxia probably was rare over the last 1000 years, but became prevalent in the 20th century. Enhanced productivity and possibly modern global warming are to blame. The degree of hypoxia and salinity lows correlate only during the 20th century, possibly the result of enhanced productivity after flood events (nutrient inputs) with resulting hypoxic conditions. In pre-anthropogenic times, the floods did not deliver abundant nutrients and no associated hypoxia occurred.

THANKS TO: Funding by Connecticut Sea Grant, EPA, and the Connecticut Department of Environmental Protection. Cores and 210Pb and 137Cs data collected by M. R. Buchholtz ten Brink (U.S.G.S., Woods Hole, MA), pollen data by K. Beuning (University of Wisconsin, Eau Clair).