Journal of Great Lakes Research 38 (2012) 35–48

Contents lists available at ScienceDirect

Journal of Great Lakes Research j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j g l r

The eutrophication of Lake Champlain's northeastern arm: Insights from paleolimnological analyses Suzanne N. Levine a,⁎, Andrea Lini b, Milton L. Ostrofsky c, Lynda Bunting d, Heather Burgess b, Peter R. Leavitt d, Daun Reuter e, Andrea Lami f, Piero Guilizzoni f, Elizabeth Gilles b a

University of Vermont, Rubenstein School of Environment and Natural Resources, Aiken Center, Burlington, VT, 05405, USA University of Vermont, Department of Geology, Delahanty Hall, Burlington, VT, 05405, USA Department of Biology, Allegheny College, Meadville, PA, 16335, USA d Limnology Laboratory, Department of Biology, University of Regina, Regina, SK, Canada, S4S 0A2 e Science, Liberal Arts and Business Division, Paul Smith's College, Paul Smiths, NY 12970, USA f National Research Council, Institute of Ecosystem Study, 28922 Verbania Pallanza, Italy b c

a r t i c l e

i n f o

Article history: Received 9 October 2010 Accepted 19 April 2011 Available online 23 August 2011 Communicated by Eric Howe Keywords: Lake Champlain Sediment Paleolimnology Eutrophication Restoration

a b s t r a c t The trophic history of Lake Champlain's northeastern arm was assessed using a multi-proxy paleolimnological approach to provide sub-basin specific information for restoration planning. Sediment cores collected from Missisquoi Bay, St. Albans Bay, and the central Northeast Arm (Inland Sea) were analyzed for nutrients, organic carbon, carbon stable isotopes, biogenic silica, pigments, diatoms and soft algae microfossils. Results indicate that this arm of Lake Champlain was oligotrophic when Europeans arrived in 1609, and that clearance of N70% of catchment forest cover had minor impact on algal production. Instead, eutrophication of St. Albans Bay was concurrent with sewer installation and expansion in early 20th century, and again with urban development in the 1960–70s. In contrast, less urbanized Missisquoi Bay remained mesotrophic until agriculture intensified after 1970. Interpretation of central Northeast Arm trophic history is complicated because road and railroad causeways built in 19th century reduced sediment input to this basin for several decades. Nevertheless, high surfacesediment concentrations of nutrients, pigments and organic matter along with replacement of Cyclotella bodanica with more eutrophic Fragilaria crotonensis suggest substantial eutrophication in deep as well as shallow water after 1970. We conclude that effective restoration of the northeastern arm is possible, but will require stringent control of animal and human wastes and reduced use of crop fertilizers. © 2011 Published by Elsevier B.V. on behalf of International Association for Great Lakes Research.

Introduction Water quality degradation by eutrophication remains severe in many of the world's lakes despite decades of intense study. For example, the United Nations Educational, Scientific and Cultural Organization (UNESCO) reports high-biomass cyanobacterial blooms in all 65 countries surveyed, and in more than half of the lakes of lowland Europe, Africa, Australia, and China (Codd et al., 2005). In the United States, every biannual report of the Environmental Protection Agency (EPA) to Congress since the agency's founding in 1972 has cited excess nutrient or its consequences – algal blooms, macrophytes and hypoxia – as the principal reason why lakes fail to meet water quality standards (US EPA, 2009). Although many lakes with brief periods of eutrophication (e.g., the experimentally-treated ELA lakes, Levine and Schindler, 1989; Schindler, 1974) or rapid flushing (e.g., Lake Washington, Edmondson, 1996) recover quickly following reduction in nutrient input, other lakes exhibit multi-decadal lags in ⁎ Corresponding author. Tel.: + 1 802 656 2515. E-mail address: [email protected] (S.N. Levine).

biological response (e.g., Lago Maggiore, Salmaso et al., 2007) or even continued water quality degradation (e.g., Lough Neagh, Bunting et al., 2007). In fact, an estimated 40% of European lakes have shown no to minor response to nutrient reduction (Sas, 1989). To date, there is little agreement concerning the causes of such long-term delays. Persistent release of phosphorus from bottom sediments can delay lake recovery from nutrient diversion for years-to-decades (Schindler, 2006). In addition, limitation of algal growth by nitrogen supply seems to be more widespread than previously surmised (Bergström and Jansson, 2006; Elser et al., 1990), especially in agricultural regions where crop fertilizers and animal wastes have saturated soil P adsorption capacity (Carpenter et al., 1998). Furthermore, colonial cyanobacteria can be favored by fisheries harvests and species introductions that modify predator–prey regimes (Carpenter et al., 1995), as well as by altered littoral–pelagic zone linkages (Schindler et al., 1996), and climate warming (Schindler, 2001). Often, it is difficult to quantify the relative importance of diverse causes of water quality degradation because the time series of observations are too brief (years to decades) relative to the period of environmental change (decades to centuries). This issue is particularly pronounced for the world's large lakes where routine

0380-1330/$ – see front matter © 2011 Published by Elsevier B.V. on behalf of International Association for Great Lakes Research. doi:10.1016/j.jglr.2011.07.007

36

S.N. Levine et al. / Journal of Great Lakes Research 38 (2012) 35–48

limnological monitoring is often highly limited due to logistical constraints. Studies of the eutrophication of large lakes may be complicated further by high spatial variation in causal mechanisms (e.g., urbanization, forestry) and variation among locations within the lake in the magnitude of response to common and unique forcing mechanisms. For example, analysis of coastal marine ecosystems and the Laurentian Great lakes (e.g., Green Bay, Wisconsin) suggests that both the magnitude of water quality degradation and the precise mechanism leading to excess production (e.g., N vs. P) may vary among embayments and open water habitats (Savage et al., 2010). Similarly, sites within a lake may vary in the magnitude of response to common forcing factors (e.g., climate) as a result of habitat-specific variation in morphometry (stratification, internal nutrient loading), hydrology (water and nutrient residence times), and food-web structure (barriers to mobile predators, etc.). Fortunately, paleolimnological techniques can now be used in the major sub-basins of large lakes to better identify the common and site-specific causes of eutrophication (e.g., Bunting et al., 2007; Leavitt and Hodgson, 2001). In this paper, we describe the trophic history of the northeastern arm of Lake Champlain as determined from sediment records and explain how the information gained can be used to improve lake management. Lake Champlain provides an important and generalizable model to better understand the causes of persistent eutrophication in large North American lakes because of high spatial variation in land use practices, the degree of sub-basin eutrophication, and the precise management strategies employed to remediate water quality problems. For example, dense seasonal blooms of cyanobacteria typify both of the lake's large, shallow embayments (Missisquoi and St. Albans bays) while the deeper Northeast Arm is sometimes cyanobacteria-dominated but forms only occasional blooms of limited extent (Smeltzer et al., 2012). Eutrophication of Missisquoi Bay was manifested first as weed bed expansion in the 1970s (Myers and Gruendling, 1979), followed by intense summer cyanobacterial blooms by the early 1990s (Shambaugh, 2008; Smeltzer et al., 2012), and high levels of microcystin by 2000 (Boyer et al., 2004). By contrast, St. Albans Bay has suffered blooms and excessive weed growth since the 1960s, at least (Smeltzer, 2003). Similarly, sites have been subject to differential management, including use of copper sulfate (St. Albans Bay only) (1966–1970s) (Myers and Gruendling, 1979), banning of phosphate-based detergents in Vermont (1978-on), reduced nonpoint sources from farms (1980-on), P removal from sewage (1987-on), installation of water circulators (St. Albans Bay in 2007) and partial causeway removal to aid in flushing (Missisqoui Bay in 2007; Smeltzer, 2003; Smeltzer et al., 2008). Given this high temporal and spatial variability, it has proven difficult to develop effective lake management strategies. The specific objectives of this study include determination of “reference” (pre-settlement and pre-Industrial Age) conditions, quantification of the timing of the onset of cyanobacterial blooms, and assessment of the relative impacts of specific land uses and waste-water effluent on eutrophication in different parts of the lake. In general, we anticipated that the embayments would exhibit earlier and more severe eutrophication than the more open waters of the Northeast Arm, but that the magnitude of water quality change should vary with land use intensity, a factor that is well documented for the Lake Champlain catchment (see below). Specifically, we hypothesized that regional deforestation (1760–1870) should influence all sites, but the subsequent effects of animal husbandry (mostly Missisquoi Bay) and urban development (St. Albans Bay) should be more specific to individual sites. Methods Study area Lake Champlain lies in an ancient continental rift valley between the Adirondack Mountains of New York and the Green Mountains of

Vermont (Fig. 1). Long (194 km), narrow (19 km maximum), and deep (up to 122 m), with an area of 1127 km 2, and a volume of 26 km 3, the lake is one of the largest in the United States. The lake is morphologically complex, with numerous islands and sub-basins further separated by artificial road and railroad causeways. The basin drains north to the St. Lawrence River through the Richelieu River, the two joining near Montreal, Quebec. The Champlain Canal, dug in 1823, connects the southern lake to the Hudson River and Erie Canal, creating a pathway for exotic species invasions (Marsden and Hauser, 2009). Regional climate is cool and humid with average annual minimum and maximum temperatures of − 13.9 and 26.0 °C, and mean annual rainfall and snowfall of 84.5 and 155 cm, respectively. This entire northeastern arm of Lake Champlain is generally iced over from mid-December through April. The Northeast Arm, which includes St. Albans Bay, lies entirely in Vermont, USA, whereas Missisquoi Bay is shared roughly equally by this state and Quebec, Canada (Fig. 1). Major characteristics of the study areas are summarized in Table 1. The central Northeast Arm (excluding St. Albans Bay) makes up about one-quarter (269 km 2) of Lake Champlain's surface area, but because it is shallower (mean and maximum depths, 13 and 49 m) than the Main Lake to the south, holds only 13% of lake volume (Myers and Gruendling, 1979). Net water flow is southward into Malletts Bay, but slow due to road and railroad causeways to the west and south that have largely isolated the water body. Circulation in summer is primarily through a simple high amplitude internal seiche (T. Manley, Middlebury College, pers. com.). St. Albans Bay occupies two sub-basins (mean depths, 13 and 4 m) separated by a rise and island. The inner bay is unstratified in summer, receives the inflow of the Mills River and Black and Jewett Creeks, and exhibits complex water exchange with the bay's dimictic outer basin and the central Northeast Arm beyond (Eglite, 2009). Missisquoi Bay, on the other hand, lies in its own shallow (mean depth 2.8 m) basin, which is connected to the Northeast Arm by a narrow channel at its southern end. It receives water from both eastern and western catchments, in particular via the Missisquoi, Pike and Rock rivers. Overall, the ratio of catchment to lake area (Ac:AL) for Missisquoi Bay (40) is double that of Lake Champlain as a whole (19), while its catchment area: basin volume ratio (Ac:VL) is close to 20 times greater (14,114 versus 827) than that of the entire lake. These characteristics are less extreme for St. Albans Bay (Table 1), whereas the central Northeast Arm (minus St. Albans Bay) is unusual in the small amount of direct runoff it receives. Current land use in the northern Lake Champlain Basin is predominantly agricultural, with intensive dairy operations, and feed corn as the principal cultivated crop. In 2001, the percent distribution of forested:agricultural:urban land in the St. Albans Bay catchment was 24:56:12 compared with 62:25:5 in the Missisquoi Bay catchment, and 38:41:11 for land draining directly to the central Northeast Arm (Troy et al., 2007). Missisquoi Bay is eutrophic, with mean total phosphorus (TP) and chlorophyll a concentrations of 47 and 15 μg/L, and mean Secchi transparency of 1.8 m for the period 1992–2007. The inner subbasin of St. Albans Bay also is eutrophic based on limited spatial surveys and remote sensing (Smeltzer et al., 2008; Wheeler, 2006), but the outer bay, where routine monitoring occurs, is meso-eutrophic (Table 1). By contrast, the Northeast Arm at the VT DEC monitoring site near Savage Island (Fig. 1) is mesotrophic (~16 μg TP/L, ~5 μg chlorophyll a/L, Secchi depth of ~5.0 m). Mean mass ratios of TN:TP average ~15: 1 in both bays and suggest, along with physiological indicators and microcosm enrichment experiments (US EPA, 1974; Levine et al., 1997), that nitrogen affects algal growth in combination with phosphorus influx. Overall, TN:TP is slightly greater (22) in the central Northeast Arm than in the two bays. Monitoring of phytoplankton in the northeastern lake reveals that summer cyanobacterial blooms consist of variable combinations of the N2 fixing genera Aphanizomenon, Anabaena, and Gloeotrichia, and nonheterocystous Microcystis and Woronichinia. Since 2006

S.N. Levine et al. / Journal of Great Lakes Research 38 (2012) 35–48

37

Fig. 1. Location of the study regions and core collection sites in Lake Champlain, with causeways depicted as thin lines. The main line down the lake is the boundary between New York and Vermont.

Aphanizomenon has been most frequently dominant at all VT DEC sampling sites in the region,Microcystis, Anabaena and Woronichinia are most often co-dominants at their sites in Missisquoi Bay, outer St. Albans Bay and near Savage Island in the Northeast Arm, respectively (Smeltzer et al., 2012). Interestingly, earlier studies reported Microcystis as the predominant bloomer in Missisquoi Bay (Shambaugh, 2008; Watzin et al., 2006). As well, diatoms are abundant during spring and in the summers of cool years (Aulacoseira at all sites, Fragilaria in St. Albans Bay), while the dinoflagellate Ceratium is also common in St. Albans Bay and at the Savage Island site in late summer and autumn. In general, summer zooplankton communities are composed mainly of cladocerans such as Bosmina longirostris, Daphnia retrocurva, Daphnia galeata mendotae, and Eubosmina coregoni, whereas winter assemblages are dominated by cyclopoid copepods (e.g., Diacyclops thomasi, Mesocyclops edax) (Shambaugh et al., 1999). Among fishes, Osmerus mordax (rainbow smelt) is the most common species in the Northeast Arm, while a number of taxa are common in the bays, including Perca flavescens (yellow perch), Lepomis macrochirus (bluegill), Lepomis gibbosus (pumpkinseed), Micropterus salmoides (largemouth bass), Micropterus dolomieu (smallmouth bass), Notropis atherinoides (emerald shiner), Notropis hudsonius (spottail shiner), Fundulus diaphanus (banded

killifish), and invasive Morone americana (white perch) (Bilodeau et al., 2004). Alosa pseudoharengus (alewife) invaded and spread through the northeast arm from 2003 to 2005, but was still relatively scarce when our cores were collected (Marsden and Hauser, 2009). Macrophytes are prolific in shallow regions of the northeastern arm in early summer, including Vallisneria americana (water celery), invasive Myriophyllum spicatum (European watermilfoil), Elodea canadensis (Canadian waterweed), Potamogeton perfoliatus (clasping leaf pondweed), and Najas flexilis (slender water nymph) (Bouchard and Arsenault, 2003). History of human activity Lake Champlain has a vibrant history due to its provision of a navigable waterway between the St. Lawrence and Hudson Rivers, particularly after construction of the Champlain and Erie canals in 1823 and the Chambly canal in 1844. Aboriginal land use included limited agriculture and forest burning to drive animals towards hunters; however, native human populations were small, and their impact on the lake is assumed to be minimal (Albers, 2000). European settlers subsequently engaged in subsistence farming, sheep pasturing for wool, commercial potash production, and logging, with Burlington,

38

S.N. Levine et al. / Journal of Great Lakes Research 38 (2012) 35–48

Table 1 Modern (1992–2007) characteristics of the three regions of Lake Champlain studied. VT DEC study sites in the embayments are nearer openings to the central Northeast Arm than our core sites, and thus may be less eutrophic. The Central Northern Arm site is our Savage Island site.

Morphometry Surface area (km2), As Mean depth (m), zave Max depth (m), zmax As:zave Volume (km3) Catchment1 Area (km2), AC AC:AS AC:zL Land Use2 % forest % agriculture % urban P load (kg/km2/y)3 Geochemistry4 Total P (mg/L), TP Total N (mg/L), TN TN:TP Fe (mg/L) Ca (mg/L) Dissolved Si (mg/L) Dissolved inorganic C (mg/L) pH Secchi depth (m) Chlorophyll a (μ/L) Chlorophyll a: total P 1 2 3 4

Missisquoi Bay

St. Albans Bay

77.5 2.8 4 27.7 0.220

7.2 8 12 0.9 0.023

Central Northeast Arm 269 13 49 20.7 3.50

3105 40 14,114

130 19 5652

234 0.9 18

62 25 5 1931

24 56 14 940

38 41 11 29

0.048 0.70 15 0.21 13.5 2.07 9.3 8.0 1.8 14.7 0.31

0.027 0.44 16 0.073 17.4 0.65 10.1 8.3 2.7 5.4 0.37

0.016 0.35 22 0.038 16.4 0.84 9.9 7.9 5.0 5.4 0.34

Catchment and morphometric data here exclude St. Albans Bay and its catchment. Gaddis (2007), International Missisquoi Bay Task Force (2004). Troy et al. (2007). VT DEC (http://www.anr.state.vt.us/dec/waterq).

Vermont, a major North American sawmill center in 19th century. As a result, Vermont forests were depleted rapidly, with ~70% removal by ca. 1870 (Albers, 2000). Furthermore, streams were dammed to power mills and straightened to permit rapid drainage of farmland. Recent geomorphic assessment of Lake Champlain tributaries indicates that many continue to undergo channel adjustments such as down-cutting, widening, and sediment dumping in response to historical changes in sediment and water loads (Kline and Cahoon, 2008). Similarly, a commercial fishery in the northeastern arm severely reduced stocks of lake whitefish (Coregonus clupeaformis), resulting in closing of the US portion in 1912 (Halnon, 1963). Canadian exploitation continued until 2004 when fish could no longer be caught. Over-exploitation of resources and opening of the western US by railroads led to collapse of the regional economy, such that abandonment of hillside farms allowed forest re-growth after 1870. Because regional cities experienced epidemics of cholera and other waterborne diseases (Leighton, 1905), many centers installed sewer systems in late 19th and early 20th centuries. Consequently, raw sewage was discharged into St. Albans Bay from 1900 to 1936, with primary treatment (solids removal) after 1936, and secondary treatment (dissolved organic matter removal) after 1955. Tertiary treatment to remove P was initiated in 1987. In contrast, Missisquoi Bay received very little urban waste water but its flushing may have been modestly reduced by an automobile causeway and bridge built across its southern mouth in 1937. The Northeast Arm was more seriously impacted by causeways: a wagon road was built between South Hero Island and the mainland at its southern end in 1850, then in 1899–1900, the railroad built causeways and bridges between the lake's large islands cutting off water exchange with the northwestern lake (Fig. 1). Major land use changes occurred in the region after 1960. First regional population growth prompted a proliferation of pavement

and drainage systems that increased storm water and sediment delivery to the lake (Albers, 2000). Second, agriculture was intensified significantly, with a shift from small family to larger industrial facilities, importation of feed grains (and nutrients) from outside the catchment, selection for higher per capita production (e.g., milk per cow), and use of more artificial fertilizer to grow local crops (USDA, 2009). At present, the dairy industry is important throughout the northeastern arm catchment, but milking facilities are concentrated in the Missisquoi Bay catchment, while feed corn and hay production are important around St. Albans Bay (Watzin et al., 2005). Furthermore, manure is managed through waste redistribution to agricultural fields in ice-free seasons, a practice that commonly increases influx of nonpoint nutrients to surface waters (Carpenter et al., 1998). Restoration efforts began circa 1966, but varied substantially among sub-basins. For example, copper sulfate was applied mid-1960s–1970s to reduce cyanobacterial biomass in St. Albans Bay, but was abandoned due to perceived stimulation of nuisance macrophytes. Phosphorus influx was reduced in 1978 with a state-wide ban on P-rich detergents, continued with control of nonpoint P runoff from farms ca. 1980, and included P removal from municipal wastewaters after 1987, while water circulators were installed in 2007 to reduce internal P loading. Unfortunately, these strategies have been largely ineffective at reducing algal biomass and water-column TP concentrations (Smeltzer, 2003; Smeltzer et al., 2008). Initially, Missisquoi Bay was not managed intensively, as this site was relatively free of cyanobacteria during the 1970s (Myers and Gruendling, 1979), and exhibited mainly littoral-zone siltation and weed-bed expansion. However, since the start of lake monitoring in 1992, Missisquoi Bay has shown mean TP and chlorophyll a concentrations higher than anywhere else in Lake Champlain, and since cyanotoxin monitoring began in 2003, has hosted the most blooms with microcystin production (Watzin et al., 2006). Consequently, remediation of the bay is now a major management priority. Core acquisition and sampling Short cores (b1 m, ~400 years) were collected from the deepest points of inner St. Albans Bay and Missisquoi Bay (Fig. 1) in March 2006 using a gravity corer with a 2.5″ OD PVC liner tube. Piston cores collected at that time have been analyzed elsewhere (Burgess, 2007) and do not include most sediment parameters presented herein. Additional gravity cores were retrieved at two sites along the main axis of the Northeast Arm, near Savage Island in September 2005, and near Cheney Point in March 2008. All cores were wrapped in black plastic and kept on ice during transport to the laboratory, where they were stored in a walk-in refrigerator at 4 °C until they were logged for stratigraphy and extruded at a 1-cm interval (usually within 48 h) under low light and temperature (Leavitt and Hodgson, 2001). Cores were sampled for pigments and soft algae by completely filling vials or Whirl-Pac bags to exclude air and refrigerating or freezing samples until analysis. The remainder of each sediment section was collected into pre-weighed vials for drying, powdering with a mortar and pestle, and use in dry weight and geochemical analyses. Chronology and sediment accumulation rate determination Sediment chronology and mass accumulation rates were estimated by quantifying 210Pb radioactivity in sediments using the ultra-low background gamma spectrometry protocol of Engstrom et al. (2006). Sediment accumulation rates (SAR) were calculated from the mass of material present between dated layers. Estimates of dates for sediment layers older than the limit of detection of the 210Pb method were obtained in the Missisquoi Bay and St. Albans Bay cores by lining up their organic carbon profiles with those of radiocarbon dated piston cores collected by Burgess (2007) and interpolating between 210Pb and 14C dates. For the Cheney Pt. and Savage Is. cores we simply assumed that

S.N. Levine et al. / Journal of Great Lakes Research 38 (2012) 35–48

the mean SAR estimates for mid-19th century also applied lower down in the cores. Geochemistry Organic carbon (OC) and total nitrogen (TN) were analyzed at 1-cm resolution by combusting sediment in sealed tin capsules and analyzing the gas released in a CE Instruments NC 2500 elemental analyzer calibrated with OAS B-2152 (1.65% ± 0.02 C, 0.14% ± 0.01 N) and OAS B2150 (6.72%± 0.17 C, 0.50% ± 0.01 N) standards and using Eager 200 data handling software. The precision of the analyzer is ~1% of the quantity measured for OC, and ~0.5% for TN. Total and exchangeable (bioavailable) phosphorus (TP and EP) samples were digested in hot HCl:HNO3 or 10:10:4 citrate–bicarbonateascorbic acid solution (Anschutz et al., 1998), respectively. Solutions were analyzed on an Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES) following Druschel et al. (2005). Biogenic silica was measured using the procedure of Demasters (1981) in which hot (85 °C) 1% sodium carbonate is added to powered sediment and the appearance of silica in solution measured over a time course. Silica concentration is regressed against time to determine ambient concentration as the y-intercept of the resulting equation.

39

standard for cell quantification. The sediment is then washed, and mixed with glycerol and Fuschin B for mounting. Sediment for diatom analysis was clarified through treatment with hot H202 (30%) followed by soaking in CH3COOH (95%) (Bates et al., 1978; Morley et al., 2004), and the cleaned frustules mounted on slides with Naphrax. At least 300 valves or stomatocysts were identified and counted using the keys of Patrick and Reimer (1966). Results Stratigraphy All four cores contained light gray clay at the bottom that transitioned to coarser brown-gray clay at sediment depths of 60, 20, 9 and 4 cm in the Missisquoi, St. Albans, Savage Is. and Cheney Pt. cores, respectively. Except in the Savage Is. core, the gray-brown layer included dark brown or black plates, mottles, and streaks in those intervals with elevated OC, some of which contained identifiable macrophyte fragments (especially in the 2–4 cm interval of the St. Albans core). The St. Albans Bay core was primarily silt above 11 cm depth (deposition date ~1980), while the Cheney Pt. core contained both sand and silt above 2 cm (after ~1970). The latter core also had a well-developed rust-colored flocculation above the sediment surface.

Stable isotopes Chronology and sediment accumulation rate For measurement of carbon stable isotope ratios, 25–50 mg of each sample were combusted in quartz tubes at 900 °C and the resulting CO2 gas cryogenically purified on a vacuum line (Boutton, 1991). A VG SIRA II isotope ratio mass spectrometer (Environmental Stable Isotope facility, Geol. Dept., Univ. Vermont) was used to analyze the. CO2. Results are reported using standard delta (δ) notation, in units of ‰ relative to the inorganic standard V-PDB. Analytical precision was ±0.05‰ (based on replicate standards). Interpretation of results was based on Lake Champlain source material analyses conducted during 1995–1997: for terrestrial detritus, seston, and submerged macrophytes, δ13C ranges were −28 to −31, −25 to −30, and −14 to −23‰, respectively. Pigments Sediments for pigment analysis were stored in the dark at 4 °C until analysis using standard high performance liquid chromatography (HPLC) at the University of Regina (Leavitt et al., 1989) or the CNR Institute in Verbania Pallanza, Italy (Lami et al., 2000). In both cases, pigments were extracted into degassed mixtures of organic solvents (acetone, methanol) and water under an inert N2 atmosphere, filtered through 0.45-μm pore membrane filters, and injected into HPLC systems for separation and detection of chlorophylls, carotenoids and their derivatives using a C18 column and reversed-phase procedures. HPLC systems were calibrated and peaks identified using authentic pigment standards from the U.S. Environmental Protection Agency and unialgal cultures. Compounds isolated included parent and derivative chlorophylls a, b and c, and taxonomically diagnostic carotenoids from cryptophytes (alloxanthin), siliceous algae (mainly diatoms, diatoxanthin); chlorophytes and higher plants (chlorophyll b, pheophytin b), colonial cyanobacteria (myxoxanthophyll), Nostocales cyanobacteria (canthaxanthin), and total cyanobacteria (echinenone), as well as ubiquitous β-carotene (total algae). We report pigment concentrations both specific to dry sediment weights (μmol pigment/g dry mass) and as annual accumulation rates (μmol pigment/m2/y).

Radioisotope analysis indicated excess 210Pb activity in the upper 25 cm of cores that allowed layer dating back to ~1850 in all cores. Dating accuracy was good (SD 1–3 years) for sediment deposited since 1950, but increased down core to 10–20 years around 1900 and 27– 64 years ca. 1850 as expected (Fig. 2). Sediment accumulation rate based on 210Pb chronology followed a relatively smooth downward trend, suggesting slow rather than rapid transitions during land use change (Fig. 3). To avoid over-interpretation of temporal trends in sediment, nutrient and fossil accumulation rates and concentrations, decadal averages were calculated and used for the plotting of most sediment parameters (averages are shown at the start of each decade). Long-term sediment means were also estimated for comparison with the main intervals of land use change (e.g., deforestation, urbanization, etc.). Our earliest 19th century estimates indicate a sediment accumulation rate of ~0.2 kg/m 2/y throughout the Northeast Arm including St. Albans Bay, compared with a rate of ~0.6 kg/m 2/y in Missisquoi Bay, which has an especially large catchment (Fig. 3). During late 19th century deforestation, minimal increase in SAR was recorded in the two

Algal microfossils Slides of cyanobacteria (blue green algae) and chlorophyte (green algae) microfossils were prepared using the method of Bunting et al. (2007) in which fresh sediment is deflocculated in boiling 10% hydrogen peroxide with Lycopodium clavatum spores added as a

Fig. 2. Mean year of deposition of sediment layers in cores determined from excess 210 Pb content, and plotted at the top of 1-cm intervals: a) Missisquoi Bay, b) St. Albans Bay, c) Cheney Point, and d) Savage Island. Error bars indicate standard deviation around the means.

40

S.N. Levine et al. / Journal of Great Lakes Research 38 (2012) 35–48

Fig. 3. On the left, decadal means for variables measured over four centuries: Fe:P ratio; total phosphorus (TP), organic carbon (OC), and total nitrogen (TN) concentrations. On the right, decadal averages for sediment (SAR), phosphorus (PAR), organic carbon (OCAR) and nitrogen (NAR) accumulation rates, with values plotted at the beginning of decades.

embayments (period means were 0.3 and 0.7 kg/m2/y, in St. Albans and Missisquoi respectively), while SAR declined at the central Northeast Arm sites, probably due to 1850–1900 causeway installations that isolated the region from major sediment sources. In addition, dams on its few tributaries may have retained sediment. With 1960s urban development, SAR increased modestly at the Cheney Pt. site, but not at the more southern Savage Is. site. In the embayments, SAR rose gradually between late 19th century and the 1960s then surged upward in St. Albans Bay with major urban development around St. Albans. By 1980, SAR stabilized at ~1.2 kg/m2/y at the St. Albans Bay site, whereas a steep upward trend began at the Missisquoi Bay site in the 1990s, bringing its modern average up to ~1.4 kg/m 2/y. Phosphorus TP concentration exhibited less variability than most sediment parameters (Fig. 3). With the exception of one particularly low value in the Cheney Pt. core, all sediments deposited before 1850 had TP values of 0.5–1.0 mg/g dwt. In general, TP trended upward in more recently

deposited sediments, with substantial increases after ca. 1980, although the P content of St. Albans Bay sediment has been relatively stable since ca. 1940. Modern mean (1990–2005) TP concentrations were similar in St. Albans and Missisquoi bay sediments (1.3 and 1.7 mg P/g dwt, respectively), about 70% greater than pre-settlement TP, but less than the concentrations observed at Cheney Pt. and Savage Is. (2.1–2.3 mg P/g dwt). The fraction of sedimentary TP present as exchangeable (bioavailable) P increased up core, from 15 to 21% in sediment deposited before European settlement, to 45–67% in near-surface sediment (Table 2). In general, annual accumulation of P (PAR) was greater at the Missisquoi Bay site than at other core locations, except for the period 1960–1990 when the St. Albans Bay site had higher rates. However, PAR stabilized in St. Albans Bay during the 1970s (at ~ 1.6 g/m 2/y), and was surpassed eventually by rapid PAR rise in Missisquoi Bay after 1980. The 1990–2005 mean for Missisquoi Bay was 2.5 g/m 2/y. PAR in the central Northeast Arm sites was very low (0.1 g/m 2/y) in mid20th century, but after 1970 increased to 0.6 g/m 2/y at the Cheney Pt. site, while staying minimal at Savage Is.

S.N. Levine et al. / Journal of Great Lakes Research 38 (2012) 35–48

41

Table 2 Means (and standard deviations) of geochemical measures in sediment deposited: pre-settlement (PS, ~ 1600–1750), at maximum deforestation (DF, ~ 1850–1900), from 1940 to 1960 (pre-modern, PM), and from 1990 to 2007 (modern, M). SAR, PAR, NAR, OCAR and BSiAR denote rates of accumulation of sediment, phosphorus, nitrogen, organic carbon and biogenic silica, respectively; TP and EP are total and exchangeable phosphorus; TN, total nitrogen; OC, organic carbon; and BSi, biogenic silica. SAR for PS is from Burgess (2007). Site

Period

SAR (kg/m2/y)

TP (mg/g dwt)

EP (% of TP)

PAR (g/m2/y)

Fe (mg/g dwt)

Mn (mg/g dwt)

Fe:P

N:P

Missisquoi Bay

PS DF PM M PS DF PM M PS DF PM M PS DF PM M

0.17 0.68 0.85 1.44 0.13 0.28 0.55 1.18 – 0.19 0.11 0.27 – 0.17 0.09 0.10

1.04 1.00 1.29 1.69 0.80 1.06 1.14 1.33 0.59 0.79 0.97 2.31 0.68 0.83 0.91 2.06

– – – – 15 (1) 19 (2) 32 (9) 45 (15) – – – – 21 27 24 67

– 0.67 (0.05) 1.09 (0.06) 2.46 (0.48) – 0.30 (0.05) 0.62 (0.14) 1.57 (0.24) – 0.15 (0.01) 0.10 0.63 – 0.14 0.08 0.21

35 (0) 41(2) 49 (2) 48 (5) 24 (0) 29 (2) 31 (1) 28 (3) – – – – 26 28 32 42

0.82 (0.01) 0.94 (0.06) 1.08 (0.06) 1.64 (0.30) 0.28 (0.00) 0.37 (0.02) 0.49 (0.02) 0.66 (0.17) – – – – 0.52 0.69 1.31 6.28

41 41 38 28 24 29 31 28 – – – – 33 34 36 20

2.2 (0.1) 2.0 (0.2) 1.8 (0.2) 2.4 (0.3) 3.4 (0.3) 3.3 (0.2) 4.2 (0.1) 4.5 (0.7) 3.4 (1.3) 2.7 (.2) 6.5 3.9 2.4 2.9 8.9 18.6*

Site

Period

% TN

NAR (g/m2/y)

% OC

OCAR (g/m2/y)

C:N

δ13C (‰)

BSi (mg/g dwt)

BSiAR (g/m2/y)

Missisquoi Bay

PS DF PM M PS DF PM M PS DF PM M PS DF PM M

0.19 (0.01) 0.20 (0.01) 0.24 (0.01) 0.40 (0.04) 0.27 (0.01) 0.33 (0.03) 0.41 (0.01) 0.59 (0.03) 0.12 (0.01) 0.20 (0.04) 0.49 1.8 – 0.24 0.81 2.00⁎

– 1.38 (0.03) 2.01 (0.33) 5.78 (1.26) – 0.94 (0.22) 5.63 (0.50) 6.86 (0.37) – 0.32 (0.01) 0.66 2.42 – 0.24 0.34 0.49⁎

2.24 (0.06) 2.40 (0.02) 2.60 (0.08) 3.67 (0.20) 2.88 (0.04) 3.36 (0.21) 4.36 (0.05) 5.00 (0.17) 1.63 (0.08) 1.68 (0.24) 4.65 6.57 0.90 1.10 3.00 3.70⁎

2.9 16 (2) 22 (3) 53 (13) 3.7 9.3 (2.0) 24 (6) 59 (3) – 3.1 (1.2) 4.9 18 – 1.8 2.6 3.6⁎

11.6 (0.5) 11.8 (0.3) 11.0 (0.2) 9.2 (0.5) 10.6 (0.2) 9.9 (0.3) 9.0 (0.1) 8.6 (0.2) 8.5 (0.0) 8.1 (0.3) 7.4 7.4 8.3 7.6 7.8 7.4⁎

− 26.7 − 26.8 − 26.6 − 26.6 − 24.6 − 24.1 − 24.5 − 24.7 − 26.0 − 25.4 − 24.7 − 26 − 25.9 − 25.6 − 25.3 − 25.1

9.4 8.1 – 32 (6) 8.1 9.0 – 29 (3) 6.5 (1.4) 8.1 (2.0) 25 46 9.6 16.1 28 62

– 1.8 (3.2) – 45 (12) – 2.6 27 (6) 34 (3) – 1.4 (0.4) 2.7 12.4 – 2.6 2.5 6.3

St. Albans Bay

Cheney Point

Savage Island

St. Albans Bay

Cheney Point

Savage Island

(0.09) (0.10) (0.35) (0.05) (0.14) (0.02) (0.10)

(0.29) (0.07) (0.11) (0.18) (0.03) (0.02) (0.04) (0.19) (0.20) (0.00)

(0.1) (0.1) (0.2) (0.1) (0.2) (0.1) (0.2) (0.1) (0.5)

(0) (1) (2) (2) (0) (2) (1) (3)

⁎ 0–1 cm sample lost for %OC and %N. Data here are for 1–2 cm, which includes sediment as old as 1975.

Sedimentary concentrations of potentially P-binding iron and especially manganese increased at all sites after settlement (Table 2). Initially Fe:P ratios were uniform, but declined by 15–20% after ca. 1900 to yield modern weight ratios of 20–28 (Fig. 3). This range is well above those observed in P-saturated sediments (ca. 8; Jensen and Thamdrup, 1993). Consequently, P retention should be good under the oxic conditions normally present in Lake Champlain, although short term releases may occur when exceptionally high algal biomass drives up pH (Jensen and Andersen, 1992). Organic matter Most of the sediment buried at the bottom of Lake Champlain was clastic, exhibiting low mean OC content prior to European settlement (0.9–2.9% of dry mass) (Table 2, Fig. 3). OC content varied little at any site during 19th century, but in St. Albans Bay increased over the first four decades of 20th century, and again after 1980, to attain a modern mean of 5.0%. OC also increased in central Northeast Arm sediments beginning in early 20th century to reach modern means similar to those in St. Albans Bay (3.7% at Savage Island, 6.6% at Cheney Point). In contrast, OC remained near background levels in Missisquoi Bay until mid-20th century then rose modestly to 3.7%. Accumulation rates of OC (OCAR) at the two central Northeast Arm sites varied little over 19th and early 20th centuries (1.8–3.2 g/m 2/y), but in the second half of the 20th century rose to 18 g/m 2/y at Cheney Pt. In contrast, OCAR began a slow increase in the embayments in late 19th century that became more abrupt in St. Albans Bay during the 1960s, and in Missisquoi Bay after 1980. Modern means for Missisquoi and St. Albans bays are similar, 53 and 59 g/m 2/y, respectively.

Nitrogen Pre-settlement concentrations of TN were similar in sediments from Missisquoi Bay and the central Northeast Arm (0.12–0.19% dwt), but slightly higher in St. Albans Bay sediment (0.27%). Temporal trends were like those of OC, except that TN increase in Savage Is. sediment during late 20th century was particularly dramatic (Fig. 3). As was true for TP, central Northeast Arm sediment had greater surface concentrations of TN than was observed in embayment sediments, with Savage Is. sediment especially N rich at 2%, vs. 0.9% at the Cheney Pt. site, and 0.4–0.6% in the bays. Despite elevated modern concentrations of sedimentary N, accumulation of N (NAR) was consistently low (b0.5 g/m 2/y) in the core from near Savage Is., which collected little inorganic sediment to dilute organic debris. NAR trends at the other sites were similar to those for OCAR, reflecting the largely organic nature of the N retained. Major rate increases occurred in St. Albans Bay and at the Cheney Pt. site during the 1960s and '70s, but were delayed until after 1980 in Missisquoi Bay. Modern (1990–2005) NAR means were similar in St. Albans Bay (6.9 g/m 2/y) and Missisquoi Bay (5.8 g/m 2/y), half as great at the Cheney Pt. site (2.4 g/m 2/y), and barely elevated at Savage Is. (0.5 g/m 2/y). Most sedimentary TN:TP ratios were b5 by weight, indicating much better retention of P than of N in the lake. C:N ratio C:N ratios were uniformly low (8.3–8.5 by weight) in pre-settlement sediments of the central Northeast Arm, and decreased further in 20th century to average 7.4 in recently deposited sediment (Fig. 4).

42

S.N. Levine et al. / Journal of Great Lakes Research 38 (2012) 35–48

Presettlement ratios were slightly greater in St. Albans Bay (10.6) but also dropped to a mean of 8.6 in modern sediments. By contrast, sedimentary C:N patterns in Missisquoi Bay were complex with early values of 12 near 1600, declining to 11 by 1700 before rising to a peak of ~12.2 in the mid 1800 s. This ratio subsequently declined, at first slowly, but more abruptly after 1940, to attain a modern mean of 9.2. The last phase is consistent with a growing role for algae as primary producers. Carbon stable isotopes Sedimentary δ 13C varied little among sites during the presettlement period, with mean isotopic ratios of −26.7‰, − 25.9‰, and −24.6‰ recorded for Missisquoi Bay, the central Northeast Arm (both sites), and St. Albans Bay, respectively (Fig. 4, Table 2). Values of δ 13C increased further in St. Albans Bay with land clearing, attaining a mean value of − 23.9‰ during 1850–1900, before declining to the modern mean of −25.1‰. Similarly, cores from the central Northeast Arm showed rising δ 13C values beginning in late 19th, then a reversal to lower values in late 20th century. These trends are consistent with the hypothesis that submerged macrophytes first benefited from nutrient enrichment, but ultimately declined due to algal shading. Isotopic patterns were less pronounced for the Missisquoi Bay core, with decadal means ranging from −26.8‰ to − 26.6‰ over the four centuries assessed. Biogenic silica Biogenic silica (BSi) concentrations were low and varied little amongst the four sites from 1600 to 1850 (Fig. 5, Table 2). However, over the following century, BSi content increased at the St. Albans Bay and central Northeast Arm sites, while remaining low in Missisquoi Bay. Subsequent increase in BSi in Missisquoi Bay took place in the 1960s and 1970s. Modern concentrations are generally greater at the central Northeast Arm sites (N45 mg Si/g dwt) than in the embayments (b30 mg Si/g dwt), but this is largely due to differences in inorganic sediment dilution. Analysis of biogenic silica accumulation rates (BSiAR) indicated a 3 to 5-fold increase at the Cheney Pt. and Savage Is. sites between 1850 and 2005, contrasted with 13 and 25 fold increases in St. Albans Bay and Missisquoi Bay, respectively. Augmentation of BSiAR was complete in St. Albans Bay by 1970, but occurred in Missisquoi Bay mostly after this date.

Pigments Patterns of historical change in algal abundance were similar when calculated using either pigment concentration or accumulation rates (Figs. 5–7). As a general rule, total algal abundance (as β-carotene or Chl a + Pheo a) varied little from 1600 to 1900 in Missisquoi Bay and the central Northeast Arm, whereas primary producers increased as early as 1800 in St. Albans Bay (Fig. 5). Exponential increase in total algal pigment concentration was noted after 1930 in St. Albans Bay and the central Northeast Arm sites, but was restricted to after 1970 in semi-isolated Missisquoi Bay. Although pigments levels were similar in the embayments and the central Northeast Arm during the final quarter of the 20th century, accumulation rates were widely different due to the low sediment accumulation rates of the latter. Analysis of individual biomarker pigments revealed that historical patterns of abundance for the major algal groups (Figs. 6–7) were broadly similar to those recorded for the pigment indicators of total algal abundance (Fig. 5). In all cases, concentrations of individual biomarkers increased in St. Albans Bay in late 19th century, in the central Northeast Arm in early 20th century, and finally in Missisquoi Bay beginning in the 1960s (chlorophytes and cryptophytes) or 1970s (cyanobacteria and diatoms). Overall, initial eutrophication resulted in only modest changes in the relative abundance of the major algal groups (Fig. 8), and at three of four sites, the relative importance of chlorophytes (Chl b + Pheo b) declined in favor of increases in cryptophytes, diatoms and cyanobacteria. In contrast, assemblages at the Cheney Pt. site exhibited a higher initial proportion of cyanobacteria, albeit unicellular rather than colonial (low ratio of myxoxanthophyll to echinenone) (Fig. 7, Table 3), while the proportion of chlorophytes increased at the expense of cyanobacteria over time. Although most sediment samples contained cyanobacterial pigments characteristic of all forms (echinenone), colonies (myxoxanthophyll), and the Nostocales (canthaxanthin), high rates of cyanobacterial pigment accumulation were limited to the embayments in 20th century (Fig. 7). In general, these nuisance species were well established in St. Albans Bay by mid-20th century, and were present at current levels by the 1970s. In contrast, accumulation rates indicative of blooms developed in Missisquoi Bay only after 1980. Similarly cyanobacterial pigments rose to high concentrations in central Northeast Arm sediments by 1950, but accumulation rates remained modest throughout the century here.

Non-siliceous algal microfossils

Fig. 4. Decadal means over four centuries for the carbon source indicators C:N ratio and δ13C. C:N b 10 indicates that most carbon is of algal origin, and δ13C N − 25‰, that submerged macrophytes contribute substantially to C production.

Analysis of microfossils derived from chlorophytes and cyanobacteria revealed that the colonial cyanobacterium Gloeotrichia was common in the two embayments throughout most of the 19th and 20th centuries, often leaving more remnants behind than all other non-siliceous taxa combined (Fig. 9). This taxon is widely distributed in shallow New England lakes of all trophic categories (Carey et al., 2008). Filamentous chlorophytes such as Spirogyra, the colonial green Botryococcus, and desmids (Cosmarium, Closterium) also were common. Apart from the desmids, all species found in abundance were meroplanktonic, dividing life into benthic and planktonic stages, and capable of positive buoyancy (i.e., bloom formation) (Reynolds, 1984). Gloeotrichia responded strongly to early lake eutrophication in St. Albans Bay, exhibiting rising rates of fossil accumulation from about 1800–1970. Interestingly, this taxon then gave way to another N2-fixing genus, Anabaena, which was replaced in turn by the green alga Pediastum between 1990 and 2005. In contrast to in St. Albans Bay, accumulation of Gloeotrichia exhibited no pronounced trend in Missisquoi Bay until the 1970s, when its deposition rate briefly tripled before diminishing near the top of the Missisquoi Bay core. Annual rates of heterocyst (specialized cell for N2 fixation) accumulation peaked in Missisquoi Bay in the 1980s, and in St. Albans Bay in the 1990s.

S.N. Levine et al. / Journal of Great Lakes Research 38 (2012) 35–48

43

Fig. 5. Left column: decadal means over four centuries for sediment concentrations of the diatom indicator, biogenic silica (BSi), and the total algae indicators, chlorophyll a + pheophytin a, and B-carotene. Right column: annual rates of accumulation of these materials averaged at a decadal scale and encompassing 150–200 years.

Fig. 6. Left column: decadal means over four centuries for the sediment concentrations of the indicator pigments for diatoms (diatoxanthin), chlorophytes (green algae and macrophytes; chlorophyll b + pheophytin b) and cryptophytes (alloxanthin). Right column: annual rates of accumulation of these pigments averaged by decade and spanning 150– 200 years.

44

S.N. Levine et al. / Journal of Great Lakes Research 38 (2012) 35–48

Fig. 7. Decadal means for cyanobacterial pigment concentrations (on left) and accumulation rates (right) in sediments at the four sites. Echineone is present in all cyanobacteria, while myxoxanthophyll is associated with certain vesiculated (bloom forming) colonial species, and canthaxanthin is present in the order Nostocales.

Diatoms Twenty diatom taxa were identified in sediments, including five indicators of trophic state. In shallow Missisquoi Bay, minimal change in diatom assemblage was observed over the 400-year period assessed, meroplanktonic Aulacoseira ambigua being consistently dominant and Stephanodiscus niagarae co-dominant (Fig. 10). By contrast, Cyclotella bodanica, an indicator of oligotrophy, was a codominant with A. ambigua in St. Albans Bay and at the Savage Is. site

up until early 20th century, when its population collapsed. More mesotrophic Fragilaria crotonensis and S. niagarae subsequently increased in importance at these sites, the former becoming their diatom dominant after 1980. As was the case for pigments, the diatom assemblage at Cheney Pt. included fewer littoral representatives than elsewhere; Stephanodiscus niaga was consistently more abundant than A. ambigua, and C. bodanica always scarce. Fragilaria crotonensis also increased here during catchment settlement, and was the most abundant indicator species present after 1880.

Fig. 8. The relative abundance of algal pigments in sediments averaged by decade. Site labels indicate the beginning of each of four data sets. Pigments exist in live algae at different concentrations, thus pigment ratios do not indicate biomass distribution.

S.N. Levine et al. / Journal of Great Lakes Research 38 (2012) 35–48

45

Table 3 Means (and standard deviations) of pigment concentrations and accumulation rates during periods of distinct land use and effluent discharge: PS (pre-settlement, ~ 1600–1750); DF (deforestation maximum, 1850–1900); PM (pre-modern, 1940–1960); and M (modern, 1990–2005). Chl and pheo indicate chlorophyll and pheophytin, respectively. Chl a + pheo a

B Carotene

Chl b + pheo b

Diatoxanthin

Site

Period

mmol/g dwt

mmol/m2/y

mmol/g dwt

mmol/m2/y

mmol/g dwt

mmol/m2/y

mmol/g dwt

mmol/m2/y

Missisquoi Bay

PS DF PM M PS DF PM M PS DF PM M PS DF PM M

1.8 (0.8) 2.3 (2.0) 5.7 (0.9) 48 (20) 5.5 13 35 (3) 48 (10) 0.7 (0.0) 2.2 (2.4) 16 66 1.1 1.6 16 38

– 2.2 (1.1) 4.8 (0.7) 69 (38) – 3.1 19 (4) 56 (15) – 0.3 (0.2) 1.7 18 – 0.3 1.4 3.8

0.08 (0.02) 0.08 (0.07) 0.30 (0.11) 5.6 (1.9) 0.37 1.3 2.9 (0.3) 4.1 (1.2) 0.28 (0.13) 0.15 (0.11) 0.55 – 0.10 0.21 1.7 3.1

– 0.08 (.01) 0.26 (0.12) 8.3 (3.3) – 0.31 1.59 (0.45) 4.85 (1.33) – 0.02 (0.01) 0.06 – – 0.03 0.15 0.31

0.8 (0.4) 1.7 (1.5) 2.9 (0.5) 15 (5) 1.3 3.8 7.7 (0.8) 9.1 (1.9) 0.04 (0.00) 0.14 (0.15) 1.6 4.0 0.5 0.8 9.3 13.8

– 1.8 (0.6) 2.5 (0.6) 21 (12) – 1.82 (0.6) 11 (2) 23 (6) – 0.02 (0.01) 0.17 1.1 – 0.14 0.81 1.4

0.14 (0.01) 0.13 (0.11) 0.33 (0.09) 3.2 (0.9) 0.31 1.1 3.5 (0.5) 4.7 (1.7) 0.19 (0.07) 0.22 (0.15) 1.0 2.5 0.08 0.12 1.2 1.2 2.6

– 0.13 (0.02) 0.28 (0.07) 4.3 (2.2) 0.37 0.87 1.9 (0.2) 5.6 (1.9) – 0.03 (0.01) 0.11 0.69 – 0.02 0.11 0.26

St. Albans Bay

Cheney Point

Savage Island

Alloxanthin

Echinenone

Myxoxanthophyll

Canthaxanthin

Site

Period

mmol/g dwt

mmol/m2/y

mmol/g dwt

mmol/m2/y

mmol/g dwt

mmol/m2/y

mmol/g dwt

mmol/m2/y

Missisquoi Bay

PS DF PM M PS DF PM M PS DF PM M PS DF PM M

0.17 (.0.00) 0.21 (0.19) 0.54 (0.20) 3.5 (0.8) 4.3 9.2 19 (2) 17 (5) 0.33 (0.16) 0.70 (0.33) 2.8 7 0.11 0.11 2.6 6.3

– 0.21 (0.01) 0.46 (0.17) 4.7 (2.1) – 0.3 1.8 (0.3) 5.0 (0.5) – 0.12 (.01) 0.30 1.9 – 0.02 0.23 0.63

0.08 (0.01) 0.11 (0.09) 0.28 (0.05) 3.4 (1.9) 0.31 0.87 2.4 (1.4) 2.8 (1.4) 0.46 (0.01) 0.35 (0.04) 1.9 3.2 0.05 0.26 0.73 1.95

– 0.11 (0.02) 0.11 (0.02) 4.7 (3.0) – 0.20 1.3 (0.4) 3.6 (1.0) – 0.06 (0.00) 0.20 0.88 – 0.006 0.06 0.20

0.16 (0.01) 0.27 0.33 (0.05) 1.8 (0.5) 0.18 0.44 1.3 (0.3) 3.0 (1.3) 0.03 (0.02) 0.19 (0.21) 1.4 – 0.05 0.12 0.36 4.4

– 0.14 (0.03) 0.27 (0.07) 2.0 (1.3) – 0.10 0.71 (0.15) 2.7 (1.1) – 0.05 (0.01) 0.14 – – 0.02 0.03 0.44

0.13 (0.04) 0.26 (0.05) 0.38 (0.05) 1.8 (0.5) 0.35 0.99 2.3 (0.1) 2.8 (0.4) 0.31 (0.07) 0.33 (0.22) 1.8 2.5 0.09 0.11 1.4 1.5

– 0.17 (0.00) 0.32 (0.02) 2.4 (1.2) – 0.23 1.3 (0.3) 3.3 (0.3) – 0.05 (0.01) 0.19 0.68 – 0.02 0.12 0.15

St. Albans Bay

Cheney Point

Savage Island

Discussion Analysis of a comprehensive suite of sedimentary parameters demonstrated that the northeastern arm of Lake Champlain has undergone severe eutrophication since European colonization. In particular, cores from Missisquoi and St. Albans bays show severalfold increases in annual N, P and OC storage, and up to twenty-fold

increases in the burial rates of BSi, microfossils, and pigments. Concomitant decline in sediment C:N ratios at these sites is consistent as well with interpretations of increased algal productivity. Although data from the two sites in the central Northeast Arm are more challenging to interpret due to the potential effects of causeway construction on water circulation and sediment deposition, greatly elevated TP, TN, and OC concentrations at both sites, and a modest

Fig. 9. Rates of accumulation of soft algae (cyanobacteria and chlorophyte) microfossils in the sediments of Missisquoi and St. Albans bays averaged by decade.

46

S.N. Levine et al. / Journal of Great Lakes Research 38 (2012) 35–48

Fig. 10. Relative abundance of diatom species in selected core layers of the four cores (dates give average age).

increase in pigment accumulation rates at Cheney Pt., clearly imply eutrophication. Therefore, despite broadly similar environmental conditions at each site prior to the onset of European land-use practices, the timing and extent of water quality degradation has proven to be site-specific, likely reflecting high spatial variability in morphology, hydrology and land use. Reference conditions Several lines of evidence suggest that this region of Lake Champlain was oligotrophic to meso-oligotrophic in the absence of farming and urbanization. First, pre-settlement assemblages of nonsiliceous algae in the two bays included a predominance of benthic and meroplanktonic taxa, such as the colonial chlorophyte Botryococcus, the filamentous green alga Spirogyra, and Gloeotrichia. These taxa thrive in shallow, clear lakes where they have simultaneous access to light and sediment nutrient sources (Barbiero and Welch, 1992; Batten and Grenfell, 1996). For example Gloeotrichia can grow in water nearly devoid of P and N by storing sediment-derived P during its benthic stage and fixing atmospheric N2 while planktonic (Barbiero and Welch, 1992). In addition, oligotrophic taxa of diatoms (e.g., Cyclotella bodanica) were common in pre-settlement St. Albans Bay and the southern Northeast Arm. Second, calculation of the product of modern limnological conditions (TP, Chl a) and the ratio of pre-settlement: modern mass accumulation rates suggests that both the nutrient and total algal content of surface waters were well below accepted thresholds for oligotrophy (Vollenweider and Kerekes, 1982). For example, multiplication of modern TP concentrations with ratios of change in P accumulation suggest that TP was ~4, 3, and 1.5 μg/L at the Missisquoi, St. Albans and Savage Island sites, respectively, assuming that conditions regulating the relationship between water column concentrations, deposition, and retention in sediment have been stable (Meyers and Ishiwatari, 1993). Using a similar approach, historical concentrations of chlorophyll a are estimated to range from 0.4 to 0.7 μg/L among coring sites. Overall, these values are much lower than the remediation goals currently in place for Lake Champlain (14–25 μg TP/L), which are based on perceived feasibility given present nutrient levels and land use practices (LCBP, 2003), suggesting that substantial additional improvements of water quality can be achieved by better identification of the sources of nutrient influx.

Effects of forest removal and regrowth Before initiating this study, we hypothesized that removal of N70% of forest cover between 1760 and 1870 should result in large-scale increases in soil erosion, influx of soil P, and eutrophication of the lake. Similarly, we anticipated that subsequent afforestation of the Missisquoi Bay catchment to N60% forest cover by mid-20th century should initiate a period of recovery before the renewal of eutrophication associated with intensified agriculture. Contrary to our expectations, we found little evidence of pronounced effects of either forest removal or recovery on water quality in the northeastern arm of Lake Champlain, beyond possible modest effects on algal production in St. Albans Bay during 19th century (Fig. 5). In fact, our estimates of sediment accumulation rate in the central Northeast Arm and St. Albans Bay during minimal forest cover (0.2–0.3 kg/m 2/y) were similar to the mean rates estimated in piston cores covering several millennia of pre-settlement sediment deposition (0.1–0.2 kg/m 2/y; Burgess, 2007). Similarly, we found little evidence of major augmentation of rates of accumulation of P, N, or organic matter prior to 20th century. Not surprisingly then, sedimentary pigments revealed evidence of only modest increases in total algal biomass (Fig. 5) and cyanobacteria (Fig. 7) during this period. By contrast, in 20th century, when forest re-growth was expected to reduce soil erosion and nutrient export to the lake, sediment accumulation rate actually increased several fold in the two embayments. Taken together, these patterns demonstrate that Lake Champlain has been resilient to catchment-scale modification of forest cover, at least in terms of the biogeochemical and biological parameters analyzed in this study. Other lakes in catchments logged in 19th century have recorded similarly slight algal responses in sediments, to the extent that changes in production are indistinguishable from natural variability (Räsänen et al., 2007; Scully et al., 2000). Specifically, modern studies of the effects of logging, fire, and experimental removal of trees document substantial, but only short-term (2–10 year), increases in inputs of nutrients, dissolved organic matter, and inorganic sediment to lakes (reviewed in Carignan et al., 2000). Limited effects of 19th century tree clearance may be related to the lower impact of non-mechanized tree removal on soil integrity, and to dampening of sedimentary indicators at low temporal resolution. Instead sustained eutrophication may require persistent ecosystem conversion (e.g., agricultural development, urbanization) or development of permanent roads that increase runoff into lakes

S.N. Levine et al. / Journal of Great Lakes Research 38 (2012) 35–48

(Carignan et al., 2000). In the absence of these changes, stream channel adjustments that affect sediment and nutrient export to lakes, (widening, down cutting) develop slowly over decades (Kline and Cahoon, 2008), as does bedload transport downstream, creating the paradox of rising sediment deposition rates during forest recovery. Additionally, when sediment arrives at the river mouth, it may be buried rapidly without substantial interaction with the water column (Gregory, 2006). Nutrient runoff from farms and cities Patterns of geochemical and biological change in the sediments of the northeastern arm of Lake Champlain and Missisquoi Bay are most consistent with induction of eutrophication by nutrients from urban and agricultural sources. As noted above, raw (1900–1936), primary treated (1936–1955) and secondary treated sewage (1955–1987) was discharged into St. Albans Bay but not Missisquoi Bay for much of the 20th century. Consistent with the known importance of such point sources, accumulation rates of OC, TN, TP, total algae and cyanobacteria increased substantially after 1900 in St. Albans Bay, whereas similar rises in Missisquoi Bay occurred only after agricultural intensification began in the 1970s. Similarly, Gloeotrichia spp., a taxon which often responds quickly to initial increases in nutrient influx in New England lakes (Carey and Rengefors, 2010; Carey et al., 2008), showed mounting biomass in St. Albans Bay nearly 100 years before it was stimulated in Missisquoi Bay. In contrast to that of St. Albans Bay, the Missisquoi Bay catchment has undergone only limited population increase (~35%) since 1950. Instead, this embayment has been subject to substantial intensification of dairy activities. In fact, the Missisquoi Bay catchment was the source of nearly 25% of all P entering Lake Champlain during 2001(Troy et al., 2007), 70% of which was derived from agricultural practices that emphasize high density feedlots, importation of nutrient-fortified grains, use of fertilizers in the production of silage, extensive drainage tiling, and application of manure directly to crop fields (LCBP, 2003; Sims et al., 1998). At present, ~ 65% of nonpoint P influx to Missisquoi Bay is associated with livestock and manure application (Michaud and Laverdière, 2004). The St. Albans Bay catchment has experienced similar changes in agricultural practices, but has evolved to specialize more in regional hay and corn production than in direct milk production. Impact of causeways Little is known about the impacts of causeways built across Lake Champlain to support road (1850) and railroad (1899–1900) traffic. Our study suggests for the first time that the central Northeast Arm may have been disproportionately influenced by hydrological isolation, due to its extremely low catchment: surface area ratio (0.9) relative to the rest of the lake (19, for the whole lake). For example, sediment and nutrient accumulation rates declined until ca. 1960 when lakeside development provided new sources of sediment (Fig. 3). On the other hand, Missisquoi Bay showed no alteration of eutrophication rate following bridge and causeway construction near its outlet in 1937. Instead algal blooms developed half a century later, concomitant with the intensification of regional agriculture. Given this bay's history of good water quality prior to 1970, we believe that nutrient source reduction alone will reduce its symptoms of eutrophication, while causeway removal (partially completed in 2007) will not (see also Mendelsohn et al., 1997). However, we recognize that additional channel dredging to lower the elevation of Missisquoi Bay's outlet would speed up P flushing from the embayment. Implications for lake management Our study suggests that the precise causes of eutrophication vary with sub-basin in Lake Champlain and that this pattern may be

47

common in other large freshwater and coastal marine ecosystems. Specifically, we noted that shallow embayments exhibited the most substantial ecosystem degradation, but only in response to the development of major urban (St. Albans Bay) or agricultural (Missisquoi Bay) nutrient sources locally. Consequently we suggest that management focus should be primarily on reduction of nutrient inputs associated with human and animal wastes rather than on land use practices that contribute to natural soil erosion, mainly because the lake has exhibited substantial resilience to the loss and re-growth of forests, but has been degraded severely in parallel to documented rises in wastewater influx. As in other lake catchments (Bunting et al., 2007), it may be particularly important to reduce P importation in soil fertilizers, feed grains, and fodder to avoid a net accumulation of nutrients in soil and sediment. In addition, we recommend that the traditional practice of manure spreading on fields is inadvisable because such nutrients are both easily transported in runoff and groundwater, and highly bioavailable to algae once in lakes (Michaud and Laverdière, 2004). After 110 years of sewage inputs to St. Albans Bay, managers should anticipate only gradual recovery of water quality. Especially challenging is this bay's pre-settlement community of meroplanktonic species (e.g., Gloeotrichia), which may mediate substantial internal nutrient loading as water column nutrient sources diminish. Also worrisome is the high P content of sediments in the central Northeast Arm, as this factor may lead to future increases in internal P loading should deepwater hypoxia intensify. Although engineering solutions are attractive for this lake region (e.g., sediment sealing, artificial circulation, causeway removal), we conclude that reduction of external nutrient influx will be most effective at improving water quality, particularly in the shallow embayments. This finding is likely to be relevant to other large aquatic ecosystems currently experiencing pronounced eutrophication (e.g., Lake Winnipeg, the Baltic Sea). Acknowledgments We thank Zoraida Quinones-Rivera (U. Regina) and Stefano Gerli (CNR) for assistance with pigment analysis, Dan Engstrom (U. Minn.) for 210Pb dating of the sediment cores, and Helen Carr for help with Fig. 1. The USGS Water Centers Program and NOAA SeaGrant (Emerging Threats to Lake Champlain) funded this project. References Albers, J., 2000. Hands on the Land: A History of the Vermont Landscape. MIT Press, Cambridge, MA0-262-01175-1. Anschutz, P., Zhong, S.J., Sundby, B., Mucci, A., Gobeil, C., 1998. Burial efficiency of phosphorus and the geochemistry of iron in continental margin sediments. Limnol. Oceanogr. 43, 53–64. Barbiero, R.P., Welch, E.B., 1992. Contribution of benthic blue-green-algal recruitment to lake populations and phosphorus translocation. Freshw. Biol. 27, 249–260. Bates, C.D., Coxon, P., Gibbard, P.L., 1978. A new method for the preparation of clay rich sediment samples for palynological investigation. New Phytol. 81, 459–463. Batten, D.J., Grenfell, H.R., 1996. Chapter 7D. Botryococcus. In: Jansonius, J., McGregor, D.C. (Eds.), Palynology: Principles and Applications. Amer. Assoc. Stratigr. Palynol. Found, pp. 205–214. Bergström, A., Jansson, M., 2006. Atmospheric nitrogen deposition has caused nitrogen enrichment and eutrophication of lakes in the northern hemisphere. Global Change Biol. 12, 1–9. Bilodeau, P., Dumas, B., Masse, H., 2004. Compositon et état de santé de la communaut des poissons de la baie Missisquoi, lac Champlain, été 2003. Rapport technique Misistère des Ressouces naturelles, de la faune et des parcs, Direction de l'aménagement de la faune de Montréal de Laval et de la Montérégie, Longueuil, p. xii (+ 43 + annexes). Bouchard, A., Arsenault, S., 2003. Herbiers aquatiques de la Baie Missisquoi, Lac Champlain. EXXEP Environment, p. 16 (plus 13 appendices). Boutton, T.W., 1991. Stable carbon isotope ratios of natural materials: I. Sample preparation and mass spectrometric analysis. In: Coleman, D.C., Fry, B. (Eds.), Isotope Techniques. Academic Press, San Diego, pp. 155–172. Boyer, G., Watzin, M.C., Shambaugh, A.D., Satchwell, M.F., Rosen, B.R., Mihuc, T., 2004. The occurrence of cyanobacterial toxins in Lake Champlain. In: Manley, T. (Ed.), Lake Champlain: Partnerships and Research in the New Millennium: Proceedings of the Lake Champlain Research Consortium, May 20th 2002, Saint-Jean-sur-Richelieu, Quebec, pp. 241–257.

48

S.N. Levine et al. / Journal of Great Lakes Research 38 (2012) 35–48

Bunting, L., Leavitt, P.R., Gibson, C.E., McGee, E.J., Hall, V.A., 2007. Degradation of water quality in Lough Neagh, Northern Ireland, by diffuse nitrogen flux from a phosphorus-rich catchment. Limnol. Oceanogr. 52, 354–369. Burgess, H.D. 2007. Geochemical indicators of productivity change in Lake Champlain, USA-Canada. MSc. Thesis, Univ. Vermont, Burlington, VT, 180 pp. Carey, C.C., Rengefors, K., 2010. The cyanobacterium Gloeotrichia echinulata stimulates the growth of other phytoplankton. J. Plankton Res. 32, 1349–1354. Carey, C.C., Weathers, K.C., Cottingham, K.L., 2008. Gloeotrichia echinulata in an oligotrophic lake: helpful insights from eutrophic lakes. J. Plankton Res. 30, 893–904. Carignan, R., D'Arcy, P., Lamontagne, S., 2000. Comparative impacts of fire and forest harvesting on water quality in boreal shield lakes. Can. J. Fish. Aquat. Sci. 57 (Suppl. 2), 105–117. Carpenter, S.R., et al., 1995. Biological control of eutrophication in lakes. Environ. Sci. Technol. 29, 784–786. Carpenter, S.R., Caraco, N.F., Correll, D.L., Howarth, R.W., Sharpley, A.N., Smith, V.H., 1998. Nonpoint pollution of surface waters with phosphorus and nitrogen. Issues Ecol. 3, 1–12. Codd, G.A., Azevedo, S.M.F.O., Bagnchi, S.N., Burch, M.D., Carmichael, W.W., Harding, W.R., Kaya, K., Utkilen, H.C., 2005. CYANONET. A Global Network for Cyanobacterial Bloom and Toxin Risk Management. Initial situation assessment and recommendations. HP-VI. Tech. Doc. Hydrol., 76. Internat. Hydrol. Progr, Paris. Demasters, D.J., 1981. The supply and accumulation of silica in the marine environment. Geochim. Cosmochim. Acta 45, 1715–1732. Druschel, G.K., Hartmann, A., Lomonaco, R., Oldrid, K., 2005. Determination of sediment phosphorus concentrations in St. Albans Bay, Lake Champlain: assessment of internal loading and seasonal variations of phosphorus sediment-water column cycling. Report to the Vermont Agency of Natural Resources (71 pp.) http://www. vermont.gov/cleanandclear/stalbansbay.htm. Edmondson, W.T., 1996. The Uses of Ecology: Lake Washington and Beyond. Univ. Wash. Press, Seattle, WA. (352 pp.). Eglite, E., 2009. Circulation dynamics of St. Albans Bay. Sr. Thesis, Middlebury College, Middlebury, VT, 125 pp. Elser, J.J., Marzolf, E.R., Goldman, C.R., 1990. Phosphorus and nitrogen limitation of phytoplankton growth in the freshwaters of North America: a review and critique of experimental enrichments. Can. J. Fish. Aquat. Sci. 47, 1468–1477. Engstrom, D.R., Schottler, S.P., Leavitt, P.R., Havens, K.E., 2006. Reevaluation of the cultural eutrophication of Lake Okeechobee using mulitproxy sediment records. Ecol. Appl. 16, 1194–1206. Gaddis, E.J.B., 2007. Landscape modeling and spatial optimization of watershed interventions to reduce phosphorus load to surface waters using a processoriented and participatory research approach: a case study in the St. Albans Bay watershed, Vermont. Ph.D. thesis, Univ. Vermont, Burlington, VT, 306 pp. Gregory, K.J., 2006. The human role in changing river channels. Geomorphologie 79, 172–191. Halnon, L., 1963. Historical survey of Lake Champlain's fishery. Vermont Fish and Game Job Completion Report. F-1-R-10 Job 6. Essex Junction, VT (96 pp.). International Missisquoi Bay Task Force, 2004. Final report to the International Joint Commission (60 pp.). Jensen, H.S., Andersen, F.O., 1992. Importance of temperature, nitrate, and pH for phosphate release from aerobic sediments of four shallow, eutrophic lakes. Limnol. Oceanogr. 37, 577–589. Jensen, H.S., Thamdrup, B., 1993. Iron-bound phosphorus in marine sediments as measured by bicarbonate-dithionite extraction. Hydrobiology 253, 47–59. Kline, M., Cahoon, B., 2008. Restoring ecological processes in protected river corridors. Proc. Amer. Water Res. Assoc. Summer Specialty Conference on Riparian Ecosystems and Buffers. Virginia Beach, Virginia. Lake Champlain Basin Program, 2003. Opportunities for Action: An Evolving Plan for the Lake Champlain Basin. LCBP, Grand Isle, VT. Lami, A., Guilizzoni, P., Marchetto, A., 2000. High resolution analysis of fossil pigments, carbon, nitrogen, and sulfur in the sediments of eight European Alpine lakes: the MOLAR project. J. Paleolimnol. 59, 15–28. Leavitt, P.R., Hodgson, D.A., 2001. Sedimentary pigments. In: Smol, J.P., Birks, H.J.B., Last, W.M. (Eds.), Tracking Environmental Change Using Lake Sediments, volume 3: Terrestrial, Algal and Siliceous Indicators. Kluwer Acad, Dordrecht, The Netherlands, pp. 295–325. Leavitt, P.R., Carpenter, S.R., Kitchell, J.F., 1989. Whole-lake experiments: the annual record of fossil pigments and zooplankton. Limnol. Oceanogr. 34, 700–717. Leighton, M.O., 1905. Preliminary report on the pollution of Lake Champlain. Water Supply and Irrigation Paper 121, Series I, Quality of Water, 9. USGS, Washington, D.C. Levine, S.N., Schindler, D.W., 1989. Phosphorus, nitrogen and carbon dynamics during the recovery of a lake from eutrophication. Can. J. Fish. Aquat. Sci. 46, 2–9. Levine, S.N., Shambaugh, A.d., Pomeroy, S.E., Braner, M., 1997. Phosphorus, nitrogen and silica as controls on phytoplankton biomass and species composition in Lake Champlain (USA-Canada). J. Great Lakes Res. 23, 131–148. Marsden, J., Hauser, M., 2009. Exotic species in Lake Champlain. J. Great Lakes Res. 35, 250–265. Mendelsohn, D.L., Rines, H., Isaji, T., 1997. Missisquoi Bay Field Study and Hydrodynamic Model Verification. Vermont Geological Survey, Vermont Agency of Natural Resources, Waterbury, VT.

Meyers, P.A., Ishiwatari, R., 1993. Lacustrine organic geochemistry—an overview of indicators of organic matter sources and diagenesis in lake sediment. Org. Geochem. 7, 867–900. Michaud, A.R., Laverdière, M.R., 2004. Cropping, soil type and manure application effects on phosphorus export and bioavailability. Can. J. Soil Sci. 84, 284–295. Morley, D.W., Leng, M.J., Mackay, A.W., Sloane, H.J., Rioual, P., Battarbee, R.W., 2004. Cleaning of lake sediment samples for diatom oxygen isotope analysis. J. Paleolimnol. 31, 391–401. Myers, G.E., Gruendling, G.K., 1979. Limnology of Lake Champlain. Champlain Basin Study. New England River Basins Commission, Burlington, VT, p. 407. Patrick, R., Reimer, C.W., 1966. The Diatoms of the United States, vol. I. Acad. Nat. Sci., Philadelphia, PA. Räsänen, J., Kenttämies, K., Sandman, O., 2007. Paleolimnological assessment of the impact of logging on small boreal lakes. Limologica Ecol. Manage. Inland Wat. 37, 193–207. Reynolds, C.S., 1984. The Ecology of Freshwater Phytoplankton. Cambridge Univ. Press, Cambridge. Salmaso, N., Morabito, G., Garibaldi, L., Mosello, R., 2007. Trophic development of the deep lakes south of the Alps: a comparative analysis. Arch. Hydrobiol. 170, 177–196. Sas, H., 1989. Lake Restoration by Reduction of Nutrient Loading: Expectations, Experiences, Extrapolations. Acad. Verlag. Richarz., St. Augustin, Germany. (497 pp.). Savage, C., Leavitt, P.R., Elmgren, R., 2010. Effects of land use, urbanization, and climate variability on coastal eutrophication in the Baltic Sea. Limnol. Oceanogr. 55, 1033–1046. Schindler, D.W., 1974. Eutrophication and recovery in experimental lakes: implications for lake management. Science 184, 897–899. Schindler, D.W., 2001. The cumulative effects of climate warming and other human stresses on Canadian freshwaters in the new millennium. Can. J. Fish. Aquat. Sci. 58, 18–29. Schindler, D.W., 2006. Recent advances in the understanding and management of eutrophication. Limnol. Oceanogr. 51, 356–363. Schindler, D.E., Cottingham, K.L., He, X., Hodgson, J.R., Kitchell, J.F., Soranno, P.A., 1996. Food web structure and littoral zone coupling to pelagic trophic cascades. In: Polis, G.A., Winemiller, K.D. (Eds.), Foodwebs: Integration of Patterns and Dynamics. Chapman Hall, pp. 96–105. Scully, N.M., Leavitt, P.R., Carpenter, S.R., 2000. Century-long effects of forest harvest on the physical structure and autotrophic community of a small temperate lake. Can. J. Fish. Aquat. Sci. 57 (S2), 50–59. Shambaugh, A.d., 2008. Historical phytoplankton densities at Missisquoi Bay, Station 50. Lake Champlain Basin Program Report. VT ANR, Waterbury, VT. Shambaugh, A., Duchovnay, A., Mcintosh, A., 1999. Survey of Lake Champlain's plankton. In: Manley, T.O., Manley, P. (Eds.), Lake Champlain in Transition: From Research towards Restoration: AGU. Water Science and Application, 1, pp. 323–340 (Washington, DC). Sims, J.T., Simard, R.R., Joern, B.C., 1998. Phosphorus losses in agricultural drainage: historical perspective and current research. J. Environ. Qual. 27, 277–293. Smeltzer, E., 2003. The phosphorus problem in St. Albans Bay. A Summary of Research Findings. VT ANR, Waterbury, VT. Smeltzer, E., Telep, P., Shambaugh, A., Stangel, P., 2008. Evaluation of the Effectiveness of SolarBee® Water Circulation Devices in Reducing Algae Blooms in St. Albans Bay, Lake Champlain. VT ANR Report, Waterbury, VT, http://www.anr.state.vt.us/ cleanandclear/SolarBee%20Report%20Final.pdf. Smeltzer, E., Shambaugh, A d., Stangel, P., 2012. Environmental change in Lake Champlain revealed by long-term monitoring. J. Great Lakes Res. 38, 6–18. Troy, A., Wang, D., Capen, D., O'Neil-Dunne, J., MacFaden, S., 2007. Updating the Lake Champlain Basin land use data to improve prediction of phosphorus loading. Tech. Rep., 54. Lake Champlain Basin Program, Burlington, VT. US EPA, 1974. EPA National Eutrophication Survey. Report on Lake Champlain, New York and Vermont, EPA Regions I and II. Pacific Northwest Environ. Res. Lab., Corvallis, OR. US EPA, 2009. National Water Quality Inventory: Report to Congress. 2004 Reporting Cycle. Office of Water. EPA Publ. No. 841-R-08-001. USDA, 2009. The Census of Agriculture. National Agricultural Statistics Service, http:// www.agcensus.usda.gov/. Vollenweider, R.A., Kerekes, J., 1982. Eutrophication of waters. Monitoring, assessment and control. OECD Cooperative Programme on Monitoring of Inland Waters (Eutrophication Control), Environment Directorate. OECD, Paris (154 pp.). Watzin, M.C., Smyth, R.L., Cassell, E.A., Hession, W.C., Manning, R.E., Wang, D., 2005. Ecosystem indicators and an environmental score card for the Lake Champlain Basin Program. Lake Champlain Basin Program. Tech. Rep., 46. Watzin, M.C., Fuller, S., Kreider, M., Couture, S.M., Levine, M., Boyer, G.L., 2006. Monitoring and evaluation of cyanobacteria in Lake Champlain. Summer 2005. Lake Champlain Basin Program. Tech. Rep., 55. Wheeler, S.M., 2006. An Evaluation of the Utility of Remote Sensing for Monitoring Cyanobacteria in Lake Champlain (VT-NY-PQ). MSc. Thesis, Univ. VT, Burlington, VT.