Anthropogenically induced changes in sediment and biogenic silica fluxes in Chesapeake Bay

INTRODUCTION

Excess nutrients contributed to lakes and estuaries by human activities commonly lead to eutrophication and attendant environmental problems such as phytoplankton blooms, increased turbidity, and increased water-column oxygen demand (Conley et al., 1993; Boynton et al., 1995; National Research Council, 2000). Chesapeake Bay has been seriously affected and remains one of the ecosystems most at risk from nutrient overenrichment (National Research Council, 2000). The relatively well studied, although complex, processes of nutrient delivery, phytoplankton production, and water circulation in Chesapeake Bay lead to significant bottom-water anoxia in the middle part of the bay in the summer (Harding et al., 1992; Boynton et al., 1995; Malone et al., 1996). Anoxia in Chesapeake Bay has been noted since the early twentieth century (Sale and Skinner, 1917; Newcombe and Horne, 1938) and monitored since the 1980s (Harding et al., 1992, and references therein). Although the history and magnitude of changes in the bay since European settlement have been studied (e.g., Brush, 2001; Zimmerman and Canuel, 2002), they are less well known compared to twentieth century changes, and pre–twentieth century events are difficult to date, especially during the critical period of 100–300 yr ago.

Diatoms are major primary producers in Chesapeake Bay. The seasonal cycles of diatom blooms and biogenic silica uptake, recycling, and deposition in the bay are fairly well studied (Anderson, 1986; Fisher et al., 1988; Conley and Malone, 1992; Malone et al., 1996; Harding and Perry, 1997). During spring and fall diatom blooms, dissolved silica may briefly become a limiting nutrient (D'Elia et al., 1983; Anderson, 1986; Conley and Malone, 1992), and recycling of dissolved silica in the water column and regeneration from sediment pore water are extensive. In a typical year, 80%–90% of the primary productivity during the spring phytoplankton bloom is due to diatoms, and diatoms account for ∼50% of the total phytoplankton production for the year (D'Elia et al., 1983). Changes in diatom species assemblages, including planktonic:benthic ratios, and increases in the number of diatoms preserved in sediments in postsettlement time have been noted (Brush and Davis, 1984; Cooper and Brush, 1993; Cooper, 1995). Cooper (1995) also showed a major difference in biogenic silica flux between presettlement and postsettlement time.

As part of an effort to reconstruct past environmental changes and to determine the influence of past climates on Chesapeake Bay (Cronin et al., 1999), we obtained a series of sediment cores in the bay and performed a variety of geochronological and paleoenvironmental-proxy analyses. We focus here on the geochronology and biogenic silica results from the upper parts of these cores. In particular, we develop age models for sediment accumulation that are robust enough to allow calculation of changes in the flux of biogenic silica and total sediment to the bay floor.

METHODS

Chesapeake Bay, the largest estuary in North America, is a drowned river valley composed of broad, relatively shallow (<10 m) margins, surrounding a deeper (20–50 m) axial channel. This channel is the partly filled fluvial paleochannel of a former Susquehanna River drowned by Holocene sea-level rise (Colman et al., 1992). We obtained cores ranging from 1-m-long damped-entry cores to piston cores >20 m long taken with the Calypso system of the RV Marion-Dufresne at a variety of sites (Fig. 1 ). Sediment accumulation on the shallow margins of the bay is relatively slow and does not show consistent patterns related to European settlement (Colman et al., 2002). We therefore concentrated on two sites in the deep channel of the bay (Fig. 1), where sedimentation has been relatively rapid and continuous. We discuss results primarily from two gravity cores, PTMC-3 at site 1 and RD98-P1 from site 2. The ∼800 yr record from RD98-1 was extended with material from core MD99-2209.

We measured biogenic silica concentrations in Chesapeake Bay sediments by using a wet extraction method, for which the estimated analytical error is 10–20 mg·g⁻¹ (Carter and Colman, 1994). We also measured the water content of the samples and converted the measurements to densities and masses by using an assumed grain-specific density of 2.6 g·cm⁻³. We successfully radiocarbon dated the sediments, despite the problem of mixed carbon sources, by using accelerator mass spectrometer methods to analyze specific carbon fractions (mostly biogenic carbonate) of the sediments (Colman et al., 2002).

RESULTS AND DISCUSSION

Age models for mass accumulation at the two deep-water sites (Fig. 2 ) show progressive increases in mass-accumulation rates since ca. A.D. 1750–1800. Two straight-line segments fit the data for each site nearly as well as the continuous curves (Colman et al., 2002). The average post–A.D. 1800 accumulation rates for these sites (0.7 and 1.4 g·cm⁻²·yr⁻¹, respectively) are similar to the range for the middle to northern part of the bay (∼0.2–1.2 g·cm⁻²·yr⁻¹) determined by previous studies (Officer et al., 1984; Adelson et al., 2001; Nie et al., 2001). The increase in mass-accumulation rate (sediment flux) implied by the two-line-segment model amounts to a factor of 3.6 for site 1 and 4.2 for site 2 (Colman et al., 2002). Total sediment flux calculated from the curves in Figure 2 has increased continuously to the present (Fig. 3A ), except for a short reversal in the early twentieth century at site 2. Accumulation-rate changes have been ascribed to changes in land use, especially clearance of forests for agriculture (Brush and Davis, 1984; Brush, 1989, 2001; Cooper and Brush, 1991; Cooper, 1995), although shoreline erosion is also a significant source of sediment. In addition, the continued increase in accumulation rates despite twentieth century reforestation requires other explanations, such as urbanization or mobilization from streamside storage (Colman et al., 2002).

Previous studies of biogenic silica (BSi) concentrations in Chesapeake Bay sediments (Cooper, 1995; Cornwell et al., 1996; Zimmerman and Canuel, 2000) show inconsistent changes in the upper parts of cores. BSi concentrations in our cores (Fig. 3B) show relatively modest changes. At site 1, the concentration remained constant until the early twentieth century, after which it has shown a small increase. At site 2, BSi concentration began to decrease in the 1600s and did not increase again until the early twentieth century. However, the concentration data are deceptive because the flux of total sediment has been changing dramatically through this time interval.

For paleoenvironmental purposes, where total sediment fluxes are changing with time, BSi data need to be calculated as fluxes (BSi concentration times total sediment flux) (Fig. 3C), rather than as concentrations. The preserved flux of BSi has increased substantially at each site (Fig. 3C), beginning in each case ca. 200–250 yr B.P. (A.D. 1750–1800). For both sites, the increase is by a factor of 4–5 compared to the long-term presettlement (between 1500 and 400 yr B.P.) average (Fig. 4 ). Thus, despite differences in location and in overall rates of sediment accumulation, both sites show a fourfold to fivefold increase in preserved biogenic silica flux. Previous studies of both biogenic silica (e.g., Cooper and Brush, 1993; Cooper, 1995; Cornwell et al., 1996) and other proxies of productivity (e.g., Cooper and Brush, 1991; Zimmerman and Canuel, 2000) have related such changes to a nutrient increase due to human activities, and we concur. However, our data are the first measurements of biogenic silica fluxes in the bay in well-dated cores that extend to presettlement time.

The intensity of human activities affecting the observed fluxes of sediment and biogenic silica (Fig. 4) to the floor of Chesapeake Bay (e.g., land clearance for agriculture, road and urban construction, fertilizer application to agricultural fields, and wastewater disposal) has varied through time. Examples include the recent construction of sewage treatment plants, which have decreased the amount of phosphorus delivered to the bay, and land clearance for agriculture, which peaked in the mid–nineteenth century and was followed by twentieth century reforestation (Brush and Davis, 1984; Brush, 1989, 2001; Cooper and Brush, 1991). Our data do not distinguish among these activities, but the steady increases in the fluxes of sediment and biogenic silica to the present indicate that the combined sources have not yet abated, despite changes in human activity. The temporal pattern is consistent with scenarios in which land clearing has been surpassed by other sources of terrigenous sediments, such as construction, urbanization, and streamside storage. It is also consistent with scenarios in which nutrients first derived from eroded forest soils have been surpassed by those contributed by twentieth century applied fertilizers. A similar sequence has been suggested for the Mississippi Delta, where biogenic silica increased at the time of land clearance in the basin, then decreased before a major increase in the twentieth century (Turner and Rabalais, 1994).

Biogenic silica preserved in sediments is an index of diatom productivity and hence overall phytoplankton productivity, but the relationships are complex. Factors that complicate the relationship between biogenic silica and diatom productivity include (1) the productivity of other (nondiatom) phytoplankton and changes in the composition of the phytoplankton community, (2) recycling of silica in the water column, (3) grazing of diatoms by higher trophic-level organisms, (4) dissolution of diatoms on or in the sediments, and (5) other sources of biogenic silica, such as phytoliths. In addition, species assemblages of diatoms have shifted with time (Cooper, 1995), and the importance of other groups, such as filter feeders (especially oysters), has changed during historic time. The relative contribution of diatom productivity to total phytoplankton productivity is likely not constant, especially during nutrient overenrichment (Malone et al., 1996). Despite the complications, we consider biogenic silica flux to be a useful index of productivity changes, as it is in many other environments.

Few measures of long-term changes in the productivity of the bay are available for comparison with our estimated fourfold to fivefold increase in biogenic silica flux. On the basis of modern measurements and nutrient loading rates for mature forested lands, Boynton et al. (1995) estimated that total nitrogen and total phosphorus loading rates have increased by factors of 6–8 and 13–24, respectively, since precolonial time. Since the late nineteenth century, total organic carbon (TOC) accumulation has increased by a factor of 5, and lipid biomarker compounds from algal and bacterial sources show twofold to tenfold enrichments (relative to TOC) in the middle part of the bay (Zimmerman and Canuel, 2002). The concentration of chlorophyll-a, related to phytoplankton biomass, in surface waters of the mesohaline part of the bay doubled between 1950 and 1994 (Harding and Perry, 1997). Biogenic silica fluxes calculated by Cooper (1995) suggest a change of more than an order of magnitude between presettlement and postsettlement time. However, these presettlement flux estimates are probably too low, because they are based on single, base-of-core radiocarbon ages on total organic carbon, which are likely much too old (Colman et al., 2002).

Productivity limitation by N, P, or Si occurs on a complex set of temporal and spatial scales in the bay (Malone et al., 1996), but the continued increase in BSi flux until the present in our cores suggests that dissolved silica has not become an overall limitation on diatom productivity. Dissolved-silica concentrations may briefly approach limiting levels during spring diatom blooms and may control the size of the bloom and the timing of its termination (D'Elia et al., 1983; Anderson, 1986; Conley and Malone, 1992). In addition, dissolved-silica concentrations in many river systems have decreased in recent time, mostly owing to eutrophication of upstream lakes and reservoirs (Conley et al., 1993; Rabalais et al., 1996; National Research Council, 2000). Despite this potential for limitation, the flux of biogenic silica continues to increase in Chesapeake Bay.

Climate changes, such as those indicated for the past 1 k.y. by tree-ring studies (Stahle et al., 1998) and various fossil and geochemical proxies of past salinity and temperature in the bay (Cronin et al., 2000, 2002; Brush, 2001; Willard et al., 2002), have the potential to control the overall productivity and water-column stratification of the bay, primarily through the influence of climate on streamflow. However, the decade- to century-scale climate changes appear to be modest, and they seem to have had no effect on the diatom and macrophyte communities in the bay (Brush, 2001). These climate changes contrast markedly with the large, nearly monotonic, secular trends in biogenic silica flux (Fig. 4).

A short reversal in the nearly monotonic increase in biogenic silica is observed in the early twentieth century sediment at site 2 (Fig. 4). A similar reversal, apparently slightly earlier, occurred at site 1. It is tempting to correlate these reversals with dam construction on the Susquehanna River in the early twentieth century, because those dams trap both sediments and nutrients (Langland and Hainly, 1997) and because phytoplankton productivity in upstream reservoirs may lead to decreases in dissolved silica in the bay (Conley et al., 1993). Because of dating uncertainties and the limited number of core sites, however, this correlation awaits further study for confirmation.

CONCLUSIONS

Biogenic silica is a useful index of water-column productivity, and our results allow comparisons of sedimentation and productivity conditions in the middle part of the Chesapeake Bay before and after European settlement. These comparisons address the full range of changes that have occurred since settlement and should be useful in the design of nutrient and suspended-sediment targets for restoration of the bay. This approach should also be useful for other areas, including lakes, estuaries, or coastal seas in which large changes in sediment accumulation, eutrophication, and bottom-water anoxia occur as a result of human activity. Chesapeake Bay waters were substantially affected by human activities, beginning shortly after European settlement. Our results emphasize the fundamental interplay between sediment loads derived from the drainage basin, nutrients transported in solution or sorbed to those sediments, and phytoplankton productivity in the bay.