Holocene Biogeochemical and Pollen History of a Lake Erie , Ohio , Coastal Wetland

A five meter sediment core was taken from Old Woman Creek National Estuarine Research Reserve and State Natural Area and Preserve along the western basin of Lake Erie, U.S.A., to determine the historical biogeochemistry of the wetland. Analysis of pollen and sediment chemistry revealed that the area has remained a wetland since ca. 5,500 yr BP, despite changing lake levels. Pollen stratigraphy indicates a distinctive local succession, which has been divided into three zones. Sediments from sometime after glaciation to 5,500 yr BP were characterized by low concentrations of herb pollen; the next zone showed an establishment of hardwood forest vegetation. The modern local vegetation developed after deforestation (about 200 years ago), when sedimentation increased an order of magnitude, phosphorus deposition increased, and the ecosystem changed from a macrophyte dominated wetland to a plankton dominated marsh. After European settlement, the wetland retained its ability to act as a sink and biotic transformer of bioavailable phosphorus; however, abiotic processes seemed to be more important than the biotic transformations that dominated before deforestation. OHIO J. SCI. 94 (4): 87-93, 1994 INTRODUCTION Western Lake Erie is surrounded by a Lake-Plain (remnants of ancient lakes from previous glaciations and glacial retreat) whose gentle slopes are conducive to the formation of wetlands. The area was once known as "the Great Black Swamp"—90% of which was drained after European settlement, mostly for agriculture (Herdendorf 1987). Since wetlands are known for their ability to mitigate floods and retain and detoxify pollutants, it could be surmised that if some of these wetlands had remained intact, they might have reduced cultural eutrophication in the western basin of Lake Erie (Mitsch and Reeder 1992). The potential of a wetland to act as a sink or transformer of nutrients is tied to hydrology (Gosselink and Turner 1978, Kadlec et al. 1981), spatial and temporal landscape changes (Costanza et al. 1990), and productivity (Brown 1981, Heckman 1986, Whigham et al. 1988, Reeder 1994). Lake Erie wetlands are unique in that their hydrology and formation is dictated by both adjacent lake levels and watershed inputs. How the combined effects of these two hydrologic "forcing functions" would affect nutrient cycling in these ecosystems is not known. One theory is that the wetlands move back and forth with rising and falling lake levels (Herdendorf 1987, Mitsch et al. 1989). There is a paucity of information on the climatic, vegetational, and water level histories of the Lake Erie Lake-Plain and Till-Plain areas following the end of the Wisconsin Glaciation. Such information would help explain effects of changing climate, lake levels, and watershed deforestation. For example, there is disagreement on the interpretation of recent lake levels from the geologic record. Little data is available to directly determine the fluctuations of recent Lake Erie water levels and their 'Manuscript received 9 February 1994 and in revised form 6 June 1994 (#94-04). resultant effect on adjacent ecosystems. Coakley and Lewis (1985), and Herdendorf and Bailey (1989) are in agreement as to ancient lake levels but disagree regarding recent history. The present study explored the historic hydrologic and biogeochemical cycles in Old Woman Creek to assess what changes occurred to the ecosystem as the surrounding hydrology, climate, and landscape changed. Sediment records provide an opportunity to study long-term ecosystem processes such as how wetland landscapes respond to different hydrologic loadings and human impacts (Schoonmaker and Foster 1991). In order to understand the ecosystem from a long-term perspective, a sediment core of post-glacial stratigraphy has been used to approach these questions. The present study used three complementary analyses to reconstruct the ecosystem processes under different climates and human interferences. We combined chemical and pollen analysis with data on past productivity from fossil pigments to reconstruct part of the post-glacial history of Old Woman Creek wetland. This was then compared to recent data on b' ochemical processing of phosphorus in the wetlan Site Description and Previous Studies Old Woman Creek National Estuarine Research i 3rve and State Natural Area and Preserve (Old Woman Creek) is one of the last remnants of the pre-European settlement wetlands along western Lake Erie (Fig. 1). This 56 ha wetland lies on the edge of western Lake Erie near Huron, OH, U.S.A. Water depths in the wetland average about 50 cm or less, but this can change dramatically (up to 1 m) throughout the year, not only because of storm pulses from the watershed, but also because of adjacent lake level fluctuations and a barrier beach which may be opened or closed by hydrologic events. Currently, less than 30% of the wetland is covered by the dominant macrophyte, Nelumbo lutea (water-lotus). Therefore, unlike most wetland ecosystems, this system is dominated by open ss HOLOCENE LAKE ERIE WETLAND HISTORY VOL. 94 FIGURE 1. Old Woman Creek wetland and its watershed, showing coring sites. water primary producers rather than macrophytes (despite its shallow conditions). The geologic history of the Old Woman Creek area has been well documented; the most site specific study being done by Herdendorf (1963). The glacial history has been described by Campbell (1955), and the depositional history by Frizado et al. (1986) and Buchanan (1982) who noted a rapid sedimentation rate following deforestation. The Old Woman Creek study site lies in the deepest portion of the preglacial Huron River Valley (Buchanan 1982). Herdendorf (1992) described Lake Erie coastal wetland formation in terms of the drowning of river mouths. After the Niagara outlet rebounded, the lake rose and stream velocities decreased. This formed the current deposits of alluvium in Old Woman Creek. A regional history of vegetational changes along climatic gradients for the Lake Erie region is provided in Braun (1961) and Spear and Miller (1976). Buchanan (1982) estimated sediment deposition rates using Ambrosia (ragweed) and C dates for Old Woman Creek. He found an increase in Ambrosia pollen (an indicator of European settlement) with an influx of till-plain and lake-plain sediment. Major studies of pollen and macrofossils in nearby areas include those done at Sunbeam Prairie (Kapp and Gooding 1964), Silver Lake (Ogden 1966), Refugee Road (Garrison 1967), Frains Lake (Kerfoot 1974), Nichols Brook (Calkin and McAndrews 1980), Cony Bog (Karrow et al. 1984), and Bucyrus Bog (Shane 1989). A late glacial and early to mid-Holocene pollen zone stratigraphy for the area south of the Great Lakes was devised by Shane (1987) who analyzed information from previous studies, as well as data from six new sediment cores. Shane noted that there was a warming and drying period after 10,300 yr BP. She also noted, between 8,000 and 4,000 yr BP, a period of further warming and drying characterized by lower lake levels and the establishment of hardwood forest. MATERIALS AND METHODS The wetland was cored from the ice using a modified Livingstone piston sampler (Livingstone 1955, Colinvaux 1964). Parallel cores were taken at the site near where a long core was drawn, analyzed, and dated by Buchanan (1982). The first core yielded approximately 2.5 m of sediment, and the second core yielded 5.3 m of sediment. X-radiographs permitted comparison of the depositional patterns of both cores—which seemed to be identical. The longer core (brought up in seven 1-m sections) was the subject of all analyses. Cores were stored wrapped in a least two layers of plastic cling wrap at 4° C. Pollen subsamples of 0.5 cm were taken at 20 cm intervals along the core (adjacent to sites of chemical analysis), and at 5 cm intervals within the ragweed peak zone. Pollen was extracted using standard acetolysis techniques (Faegri and Iversen 1975) with bromoform separation (Frey 1955). Pollen concentrations were determined by adding a known volume of Eucalyptus pollen to each sample (Stockmarr 1971). Identification of pollen was done at 400X and 630X magnification with a light microscope. The pollen sum equaled approximately 300 pollen grains per sample. In those depths with extremely low pollen concentrations, counts were made to less than 100 grains (although up to five times as many Eucalyptus grains may have been counted). The percentage diagram is based on the sum of total identified and unknown terrestrial pollen including trees, shrubs and terrestrial herbs. Spores and non-terrestrial plants were excluded from the sum, and Lycopodium (clubmoss) spores, Typha (cattails), and Nupar (water lilly) have been expressed as percentages of the pollen sum. Pollen was counted and diagrams were drawn using a computer program (Eisner and Sprague 1988). Samples were taken for chemical analysis at 10 cm intervals. Within 72 hours of the cores being opened the sediments were analyzed for sedimentary chlorophyllous degradation products (SCDP). Analysis of SCDP followed the procedure of Sanger and Crowl (1979): SCDPs were extracted from 1 cm samples with 100 ml of 90% acetone, and measuring spectrophotomic absorbances at 660-670 nm peaks in a 1 cm cell. The values are expressed per gram organic matter in SCDP units (Swain 1985). Percent organic matter was determined by the loss on ignition of an oven-dried 1 cm sample at 550" C (Dean 1974). Measurements of bioavailable phosphorus in Great Lakes tributaries suggest NaOH extractable P is a reasonable estimate of bioavailable P (Logan et al. 1979). Aliquots of dried sediment (0.8 g) were analyzed for bioavailable phosphorus by placing them in 50 ml centrifuge tubes with 40 ml of 0.1 M sodium hydroxide (Chang and Jackson 1957), then analyzing for orthophosphate with the OI IK) JOURNAL OF SCIENCI, U. C. RKKDKR AND W. R. HISNHR 89 ascorbic acid method (Murphy and Riley 1962). Total phosphorus (TP) was determined after digestion with perchloric acid (Sommers and Nelson 1972). Metal concentrations were determined via atomic absorption spectrophotometry. An aqua regia/hydrofluoric acid mixture was used to release metals in Teflon bombs as described by Burnas (1967). Bulk sediment samples were taken for radiocarbon dating (Table 1). Radiocarbon dating was performed by Beta Analytic Inc. (Coral Gables, FL). Recent dating of the core was obtained from pollen analysis: the Ambrosia peak was considered to be a result of deforestation with European settlement. Forest clearance for ship building and agriculture occurred about 180 years ago (Sears 1938), which coincides with "C dates for a Lake Ontario coastal wetland Ambrosia increase dated at 150 + 50 yr BP by McCarthy and McAndrews (1988).