• Johan C. Varekamp, Ellen Thomas, (E&ES, Wesleyan University, Middletown CT 06459-0139; jvarekamp@wesleyan.edu; ethomas@wesleyan.edu),
• Marilyn R. Buchholtz ten Brink, Ellen L. Mecray, (Coastal and Marine Geology, U.S. Geological Survey, 384 Woods Hole Road, Woods Hole, MA 02543, mtenbrink@usgs.gov)
• Sherri R. Cooper, Biology Department (Bryn Athyn College, Box 707, Benade Hall
• Bryn Athyn, PA 19009, slcooper@duke.edu ).
• Mark A. Altabet (School for Marine Science and Technology, University of Massachusetts, 706 Rodney French Blvd., New Bedford, MA 02744-1221, maltabet@umassd.edu)

Our research aims at using the shells of unicellular, bottom-dwelling organisms (foraminifera) as well as the remains of unicellular algae in sediments which represent the last few hundreds of years to the last few decades in order to find out evidence of changes in the Long Island Sound ecosystem. We argue that major changes in the ecosystem (affecting small organisms low in the food chain) may have played an important role in the lobster decline. We will date the cores to find out the timing of these ecosystem changes, and use geochemical parameters to reconstruct environmental changes that may have occurred at the time of ecosystem changes. The environmental changes that we can reconstruct include water temperature, salinity, eutrophication, oxygenation, and metal pollution. We are specifially interested in the development of hypoxia over time.

Our approach to understanding the underlying mechanisms for the occurrence of hypoxia in LIS is based on the following basic observation: Hypoxia or anoxia in bottom waters occurs when the rate of oxygen supply is less than the rate of oxygen consumption. These conditions are found when the LIS waters are:

1. stratified (salinity or thermal stratification, cutting off oxygen supply) and/or when
2. the supply of labile organic matter to the bottom waters increases, creating a larger oxygen sink (e.g., increased nutrient flux, more rapid nutrient cycling).

There are important linkages: primary productivity of organic matter increases with the increased sunlight intensity in spring, and higher water temperatures increase the rate of productivity (until a nutrient becomes limiting) and the rate of organic carbon mineralisation.

With higher temperatures, a condition is set up where both 1 and 2 may increase, promoting hypoxia. We use the recent geological past (last 1000, last 400 and last 50 years) as our guidance for the present and future, and determine under what conditions and at what times hypoxia/anoxia has occurred.

We developed parameters (proxies) that indicate bottom water quality: temperature, salinity and degree of oxygenation, and combine these data with faunal (foraminifera) and floral (diatoms, dinoflagellates) proxies for temperature and water level oxygenation. These primary data are then combined with observations on the presence and concentration of inorganic contaminants (mainly heavy metals such as Pb and Hg), sewage input (through counts of Clostridium perfringens spores), paleo-productivity (abundance of diatom frustules and the species composition of the flora; abundance of dinoflagellates and the species composition of the flora; abundance of benthic foraminifera and species composition of the fauna), and age information (dating of the core layers). In addition, we monitor the amount of stored organic carbon (productivity tracer) and the isotopic composition of Nitrogen in the organic matter (indicator of the source of the nutrient).

The proxies for bottom water temperature and salinity are derived from the isotopic composition and Ca/Mg ratio of calcite foraminiferal shells (d¹⁸O). The level of oxygenation is derived from the d¹³C values in the foraminiferal calcite, after correction for the salinity effect (d¹³C*). The paleo-productivity is derived from the C_org, N and biogenic silica (B_si) analyses in the sediment layers, in combination with the biological proxies.

The faunal proxies are tested in studies of other coastal environments. The ratio of benthic to planktic diatoms, for instance, has been used as a productivity indicator in Chesapeake Bay. The ratio of the abundance of Ammonia beccarii and Elphidium excavatum (A/E ratio), has been shown to be sensitive to oxygen levels in the Gulf of Mexico. We are testing whether at the higher latitudes of LIS there may be additional effects of water temperature, and influences from the food source for these organisms (e.g., composition of the phytoplankton).

Modern LIS and the last 50 years.

The A/E index increases from east to west in the modern Sound, which may reflect the lower dissolved oxygen values in the west. The greater intensity of algal blooms in west LIS and the change in composition of the "food" source from diatoms to dinoflagellates may also play a role.

The A-E index correlates with the concentration of the bacterial spore and sewage tracer in the sediment (Clostridium perfringens). The A/E index has increased in western LIS since the 1960s, conform observations of water oxygenation over that time period. The west LIS ecosystem has seen dramatic changes in composition over the last 5 to 6 years, with strong increases in the relative abundance of Ammonia beccarii but decreasing abundance of the total numbers of foraminifera.

Metal contaminants increase from east to west in the Sound, a result of westward transport of fine-grained sediment and the presence of significant contaminant sources in the west (waste water treatment plants, WWTP, the metal-rich Housatonic River sediment). Preliminary estimates suggest that 35% of Hg in the Sound comes from WWTPs and that 50 and 70% of the Cr and Cu, respectively, in western LIS stem from the Housatonic River. Part of the metal load is brought into the Sound during catastrophic flood events. For example, the floods of the 1950s created peaks in Hg, Cu and Cr in LIS sediments as a result of the injection of toxic sediment plumes from the Housatonic River. Episodic versus continuous contaminant input should be an important consideration in overall contaminant budget estimates.

Low salinity events in the Sound occurred during the 1950s, the1970s, 1990s and during the start of the 20^th century. These periods were very wet according to the Connecticut precipitation records, and had recorded major floods on land. Our water oxygenation proxy (d¹³C*) shows low values (low oxygen in bottom waters) during these events, suggesting that enhanced oxidation of organic matter in the bottom waters took place at these times.

The medium-term record shows several interesting features:  organic carbon levels in LIS sediment have almost doubled since the 1850s (from 1.5 to 2.8 wt. %), the nitrogen has become substantially isotopically heavier (enriched in ¹⁵N), which is usually seen as a feature of WWTP nitrogen, and the sewage tracer (C. perfringens) increased in abundance by a factor of ~100. The date of onset of this increase in abundance of the sewage tracer (an indicator of population density) and that of the metal contamination were the same (~ 1850), indicating an anthropogenic origin for the metal contamination. Most metal contamination records show a decrease over the last 30-40 years, probably a result of lower metal discharges in the watersheds. The biological productivity as indicated by the number of foraminifera per gram of sediment increased by almost a factor of 1000, but declined again over the last 30-40 years.

The long term record shows two peaks in the A/E index: one in modern times (last 50 years) and one around 900 years ago, during the so-called Mediaeval Warm Period. Clearly, the current period of modern global warming and the warm middle ages created conditions that were conducive to an increased abundance of Ammonia beccarii in Long Island Sound, either as a result of the warm bottom water temperatures or from a combination of hypoxia and warm temperatures. Our preliminary isotope data suggest that water oxygenation during this early warm period was indeed low.

Organic productivity in the Sound has increased since the industrialization and strong population increase in the mid-1800s. It was then that large scale disturbance of the watersheds started, such as land clearing, agriculture, biomass burning, metal pollution, manure effluents etc, which is seen to be reflected in changes in the foraminiferal population. The onset of bottom water hypoxia in the main Long Island Sound basins started much later, however, somewhere in the late1960s to early 1970s. Either a critical threshold was exceeded, or a combination of contamination and temperature effects led to the establishment of these chronic hypoxia conditions in summer. One could argue that high temperatures (especially if coupled with high precipitation) are a prime requirement for hypoxic bottom water conditions because our long term record shows only evidence for hypoxia during the two warmest intervals in the historic record (ealry Mediaeval times, recent period of warming). The period from 1850 to 2000 is characterized by a steady increase in C_org storage and increasing d¹⁵N, but the foraminiferal productivity show a more complex pattern, with strong declines over the last 40 years, probably the result of changes in phytoplankton composition with increasing eutrophication.

Ammonia-Elphidium index (blue, scale on right) and the abundance of the bacterial spore and sewage indicator Clostridium perfringens (red; scale on left), from samples collected in 1996/1997, plotted against longitude in Long Island Sound (west left, east right).