PESTICIDE ACCUMULATION RATES IN A MANAGED MARSH
ALONG LAKE ERIE
ALISON L. SPONGBERG∗ , JOHAN F. GOTTGENS and BARRY E. MULLER
Department of Earth, Ecological and Environmental Sciences, University of Toledo, Toledo, Ohio,
U.S.A.
(∗ author for correspondence, e-mail: [email protected], Fax: 419 530 4421)
(Received 20 December 2002; accepted 12 September 2003)
Abstract. Cores from two marshes along the southern shore of Lake Erie were 210 Pb age-dated and
analyzed for persistent pesticide pollutants using column chromatographic separation and analysis
with gas chromatography. Soils in both watersheds have low to very low permeability and are dominated by a very poorly drained silty clay. Land-use practices in the watersheds of either marsh changed
little since 1950; however, both watersheds are marked by decreased area dedicated to orchards and
concurrent increase in residential and road area. The increase in grain size in recent years may be
associated with a period of high water in Lake Erie since the early 1970s. The pesticide accumulation
rates were calculated and indicate an airborne source for HCHs and endrin and possibly current illegal
usage of DDT. The ratio of metabolite to parent molecule did not appear to change with sediment
age. Also, there appeared to be a delay between pesticide application and deposition in the marshes.
Keywords: accumulation rates, DDT, Lake Erie, 210 Pb age dating, persistent pesticides
1. Introduction
Wetlands may offer many environmental benefits, including acting as sinks for
nutrients, suspended solids, and pesticides (Ewel and Odum, 1978; Hey et al.,
1994) and therefore serving as filters for agricultural drainage. Mechanisms for
removal of chemicals include microbial processing, sedimentation and biomass
production (Kadlec and Knight, 1996). Short-term assimilation of constituents in
agricultural drainage by wetlands is reasonably well documented (Ewel and Odum,
1979; Klarer and Millie, 1989; Mitsch et al., 1989; Hammer, 1993; Kadlec and
Hey, 1994; Dortch, 1996). Recently, long-term trapping of nutrients and sediments
in farm runoff was quantified for a complex of Lake Erie marshes (Gottgens and
Liptak, 1998; Gottgens et al., 1999a). Unfortunately, the relative importance of
these marshes for the assimilation or transformation of pesticides in agricultural
runoff is poorly understood.
The ability of a wetland to remove chemicals from influent waters may vary
from year to year and diminish with time (Richardson, 1985). Factors affecting
such removal include flooding regime (Craft et al., 1988), composition and density
of the plant community, water-retention time, and geochemical processes (Kadlek
and Knight, 1996; Mitsch and Gosselink, 2000). Pesticide degradation, in particuWater, Air, and Soil Pollution 152: 387–404, 2004.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands.
388
A. L. SPONGBERG ET AL.
lar, depends on the environmental conditions during transport and burial. In streams
with pH of 6.5–8.5, some pesticides can remain in a non-ionic state and be dissolved in water rather than adsorbed to suspended solids (Frank et al., 1979). Persistence in soil depends on pH, temperature, and microbial activity (Sheets, 1970),
with decomposition and the behavior of pesticide residues being strongly dependent on the adsorption and desorption properties of the sediments (Kalousková,
1989). Dissolved organic matter within the soil and sediment can also affect adsorption or leaching of pesticides (Barriuso et al., 1992; Locke, 1992; Guo et
al., 1993). However, these were all short-term studies covering a maximum of a
few years. Programs of wetland acquisition and management designed to emphasize water-quality benefits of marshes, however, should not only consider shortterm data but also long-term information on contaminant assimilation by wetlands.
Consequently, programs that do not consider such long-term information lack a
complete scientific basis.
Although the assimilative capacity of marsh sediments depends on their physicochemical properties, and can vary from year to year, the long-term record of material transfer between water and sediment may be preserved in the sediment stratigraphy. Accumulation rates can be calculated for the past 100 years by dating
the sediment column using activities of radionuclides such as lead-210 (210 Pb)
and cesium-137 (137 Cs) (Goldberg, 1963; Gottgens et al., 1999b). The Winous
Point marshes in the Lake Erie watershed serve as a natural laboratory where
these long-term processes can be investigated. Building on the results of a pilot
study (Gottgens and Liptak, 1998), the long-term effects of agricultural runoff
on the accumulation capacity of two marshes at Winous Point were investigated.
Both marshes are isolated by dikes from the open water of Lake Erie and have
drainage basins dominated by agricultural activity initiated by European settlers
during the middle of the 19th century. The two marshes have been managed by
the Winous Point Shooting Club (WPSC) in the same fashion except that the
upland dike of the west marsh was closed in 1978, thereby effectively separating that marsh from its watershed. Because the north marsh continued to receive
agricultural runoff from its drainage basin, these two marshes provided a system whereby a comparison between the accumulation rates of target analytes of
a marsh which received agricultural runoff could be made with a similar and adjacent marsh that did not. Such an experimental arrangement with two adjacent
marshes was not available in any other Ohio coastal marsh complex. We tested
the hypothesis that the marsh subjected to continued agricultural runoff has maintained consistently higher accumulation rates for target analytes since 1978 than
the reference marsh. By comparing recent accumulation rates with pre-1978 rates,
we also tested whether the impacted marsh continued to filter agricultural runoff or
whether assimilative saturation had occurred. Target analytes were those that have
a negative impact on water quality in the littoral and open water zones of Lake
Erie. The contaminants include sediment (turbidity), carbon, organic matter, nitrogen, phosphorus (total and bio-available), heavy metals, and selected persistent
PESTICIDE ACCUMULATION RATES IN A MANAGED MARSH
389
Figure 1. The study site is located at the Winous Point Shooting Club on Sandusky Bay along the
south shore of Lake Erie, Ohio.
pesticides and their breakdown products. Trapping of sediment, organic matter and
phosphorus was particularly pronounced in the impacted marsh, confirming earlier
findings from the pilot study (Gottgens and Liptak, 1998). Carbon and nitrogen
accumulation was less prominent in the core profiles, perhaps due to the significance of an atmospheric sink for these elements. The heavy metals showed mobility
throughout the sediment column, most likely due to uptake and redeposition by the
wetland plants. The present paper is a summary of the pesticide data.
390
A. L. SPONGBERG ET AL.
TABLE I
Land use (ha) in the North and West Marsh watersheds, Winous Point, Ohio.
Ten hectares of land were not covered by the 1939 aerial survey
Marsh
1939
1950
1957
1970
1988
Row crops
N
W
128
N.D.
149
152
155
154
153
154
149
144
Orchard
N
W
25
N.D.
16
5
8
3
8
2
9
2
Forest, marsh
or old field
N
W
46
N.D.
42
7
44
6
44
6
44
5
Roads and
residential
N
W
4
N.D.
6
7
6
8
6
9
11
21
N.D. = No Data.
2. Study Site
The project was carried out in the Winous Point marshes (latitude 41◦ 28 N, longitude 82◦ 59 W, Ohio, U.S.A.), situated between areas of agricultural land to the north
and Muddy Creek Bay to the south (Figure 1). The bay drains via Sandusky Bay
into the central basin of Lake Erie. These marshes are privately owned by WPSC,
established in 1856, and have been managed by professional wildlife biologists
since 1946. Original survey maps (Bourne 1820; WPSC, unpubl. surveys) indicate
that farms have defined the marshes’ upland boundary since their establishment.
Water flows into the wetland from the Sandusky River, several small creeks and
agricultural ditches, draining first into Sandusky Bay and then into Lake Erie. By
1920, ‘lakeward’ dikes were completed to protect some of these marshes from
wind and wave action from the open lake and bay (Gottgens et al., 1998). Growth
of emergent aquatic macrophytes is stimulated by periodic lowering of the water
level in the diked marshes from mid-May through early August (Sherman et al.,
1996). The marshes contain a combination of shallow water emergent vegetation,
and some woody plants on higher elevations.
The field sites for this project have been protected from physical disturbance
from Lake Erie wave energy with the use of dikes since before 1920. The property
includes a 260 ha marsh (North Marsh) subject to non-channelized runoff from
approximately 200 ha of farmland, primarily used for the production of corn and
soybeans. Runoff from similar farmland has been diverted from an adjacent 220 ha
marsh (West Marsh) since 1978. Soils underlying both marshes are dominated by
silty clays, of which the Toledo Silty Clay is the most common, with an average
clay content of 63% and an average silt content of 35%.
PESTICIDE ACCUMULATION RATES IN A MANAGED MARSH
391
Photocopies of aerial photos of the area (Ottawa County Natural Resource Conservation Service, Oak Harbor, Ohio) covering both watersheds were digitized and
analyzed for land use classification for the following years: 1939 (North marsh
only), 1950, 1957, 1970, 1988. Land-use in the watersheds of both the North Marsh
and West Marsh changed somewhat since 1950 (Table I). During that time period,
the area of land in either watershed that was dedicated to orchards was halved. The
amount of area occupied by ‘Roads and Residential’ in the watersheds doubled
since 1950. An average of 21% of the North Marsh watershed consisted of forest,
marshes, or old fields. However, the majority of both watersheds remained devoted
to row crops.
3. Methods
Five core locations picked from each of the studied marshes were selected in areas
with the greatest water depth, a high probability of inundation throughout the year,
and a history of minimal disturbance. At each site, two cores were taken. One core
from each location was used to determine sediment chronology, nutrient concentrations, grain size distribution, and pollen composition (Muller et al., in prep.). The
second core was used for metals and pesticide analyses. Clear polycarbonate core
barrels (7.5 cm i.d.) were used to collect sediment cores in both marshes. Cores
were taken from a small boat by inserting a barrel into the marsh sediments until
refusal. Because of the high water content and fine-grained sediment in the core,
the bottom of the core barrel was covered with a cap to prevent the sediments from
sliding out upon retrieval. Any headspace in the core barrel was carefully filled with
marsh water to limit disturbance of the sediment/water interface during transport.
The top of the core barrel was capped and secured with duct tape. Aluminum foil
was wrapped around the core barrel as insulation and to limit photoautotrophic
activity. Cores were transported back to the laboratory in a vertical position and
stored up to 14 days at 4 ◦ C, pending sectioning. Each core was sectioned at onecentimeter intervals. Because smearing may occur in the sediments in contact with
the core barrel, the outer 1cm of each section was discarded. The deepest 2 cm of
the core were discarded due to possible disturbance and contamination. Following
sectioning, samples were frozen pending further analyses.
Activities of total 210 Pb, supported 210 Pb and 137 Cs were measured independently and simultaneously by direct-assay using an intrinsic-germanium well-detector
following techniques described by Gäggeler et al. (1976) and Appleby et al. (1986).
Calculation of 210 Pb dates followed the Constant Rate of Supply (CRS) model
(Goldberg, 1963). To calculate age/depth relationships in the sediment cores, the
activity of unsupported 210 Pb was calculated by subtracting supported 210 Pb from
total 210 Pb in sediment samples. The activity of 137 Cs served as an independent
time marker in the sediment profile (Robbins and Edgington, 1975; Krishnaswami
and Lal, 1978), because maximum activities coincide with a period of widespread
392
A. L. SPONGBERG ET AL.
atmospheric nuclear testing in the early 1960s. Details on the radionuclide dating
protocol can be found in Gottgens et al. (1999b).
Preparation of sediment samples for pesticide analysis required column chromatographic separation of the organic extracts contained in the sediment. Initially, the
organic fraction was extracted from a known quantity of sediment using dichloromethane solvent in the Tecator Soxtec Extraction System HT 1043. Sulfur removal
was accomplished during this step using solvent cleaned copper strips. A twostep column chromatographic separation segregated co-eluting compounds within
these extracts. The first step used 0.5% diethyl ether in hexane and 40% diethyl
ether in hexane solvents in a florisil column. The second step used 0.5% toluene in
hexane and 25% diethyl ether in hexane in an acidized silica gel column. Duplicate
samples were extracted and analyzed for confirmation. Pesticide standards were
used to quantify the data and determine recovery efficiencies. Blanks and spiked
samples were also included in the protocol.
Gas chromatographic analysis of these extracts was performed on a Hewlett
Packard 5890 II gas chromatograph equipped with an HP-5 MN crosslinked 5% PH
ME Siloxane capillary column and an electron capture detector. Inlet and detector
temperatures were 205 and 250 ◦ C, respectively. Helium carrier gas was set to
6 mL min−1 flow. Splitless injection was used. The oven was set to 100 ◦ C for
2 min, ramped at 15 ◦ C min−1 to 160 ◦ C, ramped at 5 ◦ C min−1 to a final temperature of 250 ◦ C, that was held for 10 min. Analytical integration of the resulting
peaks was limited to a minimum area count of 1000 using an RTE integrator that
approximated the minimum detection limits for most compounds. Confirmation of
standard and many sample peaks was accomplished using a Hewlett Packard 6890
gas chromatograph equipped with an identical column and mass selective detector.
The gas chromatographic method was identical to the previously described method.
The pesticides chosen for analysis include those highly persistent chlorinated
compounds manufactured in the 1960–1980 era. These include aldrin and its metabolite dieldrin, endrin and its metabolite endrin aldehyde, DDT and metabolites
DDE and DDD, and three hexachlorohexane pesticides, α, β, and γ -HCH. These
pesticides were designed to be persistent in the environment and have since proven
to be highly detrimental to biota. All of these compounds have since been banned
from use in the United States, except γ -HCH which is currently highly restricted.
Accumulation rates (g m−2 yr−1 ) were calculated by determining the target analyte’s concentration per volume (g cm−3 ), integrating this concentration over depth
(cm), and dividing that by the 210 Pb-determined time (yr) that it took to deposit that
material.
Bulk density values were determined for each section of each core. These values
were generated as part of the 210 Pb assay sample preparation process. Percent
OM was determined using the loss-on-ignition method (Dean, 1974). Particlesize analysis was performed with a Malvern Mastersizer Laser Analyzer (MMLA)
(Malvern Instruments Inc.) at the Water Quality Laboratory at Heidelberg College,
PESTICIDE ACCUMULATION RATES IN A MANAGED MARSH
393
TABLE II
Summary of core chronology data from the North and West
marshes, Winous Point, Ohio. Values in parentheses are 90%
confidence intervals in years
Core
ID
ID
Cumulative
residual 210 Pb
(Bq cm−2 )
210 Pb fallout
137 Cs onset
(Bq cm−2 yr−1 )
(210 Pb year)
N1
N2
N6
N7
W4
W5
W9
W10
0.57
0.50
0.41
0.52
0.58
0.80
0.46
0.75
0.018
0.015
0.013
0.016
0.018
0.025
0.014
0.023
1969 (6)
1962 (9)
1957 (8)
1974 (4)
1958 (10)
1960 (6)
1958 (9)
1945 (7)
Tiffin, Ohio. We selected this method because it permits particle-size analysis in
small samples (e.g., less than 1 g) that are common in core studies.
4. Results and Discussion
The time intervals of relevance to the study of the long-term accumulation of
chemicals in the marsh sediments were 1920–1977, 1978–1987, and 1988–1997.
The years 1920 and 1978 were significant in the study because 1920 was the approximate date for the completion of the dikes surrounding the marshes and 1978
was the year that the West Marsh’s upland dike was closed. Consequently, after
1978 runoff from the West Marsh’s watershed was directed around the marsh and
released directly into Muddy Creek Bay. The most recent 10 yr interval was used
to evaluate sustained accumulation of target analytes.
4.1.
210
210
P B CHRONOLOGIES
Pb activities were measured for all sections of all ten cores taken in this study.
Core chronologies deemed reliable had unsupported 210Pb activities that decreased
with core depth, similar unsupported 210 Pb cumulative residuals within the accepted fallout range for this region, and 210 Pb-dates that matched well dates determined
from the profiles of 137 Cs fallout (Robbins and Edgington, 1975; Appleby and
Oldfield, 1986). Two cores, one from each marsh, did not meet these criteria and
were considered unreliable. Chronology data for the remaining four cores from
394
A. L. SPONGBERG ET AL.
TABLE III
Average accumulation rates (g m−2 yr−1 ) for organic matter, and sediments from
the North Marsh and West Marsh, Winous Point, Ohio. Values are pooled from
four cores per marsh. Values in parentheses are one standard deviation
Interval
1997–1988
1987–1978
1977–1920
Organic matter
Sediments
West
North
West
North
313 (92.3)
264 (100)
131 (44.5)
251 (78.9)
221 (128)
105 (47.3)
2370 (1220)
2590 (1400)
1300 (316)
1480 (199)
1350 (376)
811 (145)
each marsh are summarized in Table II. Counting errors for unsupported 210 Pb
activities averaged only ±0.016 Bq g−1 with a mean sample weight of 1.810 g
and count times between 19 and 46 hr per sample. Monte Carlo simulations were
used to estimate error associated with the calculation of age and sedimentation rate
(Gottgens et al., 1999b). Dating uncertainty increased with the age of the sediment.
Ninety per cent confidence intervals were ±2 yr in deposits laid down 10 yr before
present, ±5 yr at 25 yr, ±10 yr at 50 yr, and ±20 yr at 80 yr before present. We
restricted the interpretation of the chronologies to sediments deposited during the
last 80 yr and averaged accumulation rates for long intervals (e.g., 1920–1977,
1978–1987, and 1988–1997) to reduce error.
4.2. G RAIN - SIZE DISTRIBUTION OVER TIME
With the exception of the last 30 yr of deposits in the North Marsh, core profiles
for both marshes showed a minimum of 60% silt and clay-sized particles during
the last 80 yr of deposition. After 1970 grain-size distribution shows coarsening
upward to the present day. This time period coincided with the beginning of a
high water period in the lake and may reflect increased erosional influx of coarse
particles from the North Marsh’s watershed. The West Marsh, isolated from its watershed during that time interval, did not show such an increase. Particularly in the
North Marsh, low concentrations of sand-sized particles coincided with relatively
dry periods as reflected in the 1940 data point (1935–1944) and the 1960 data point
(1955–1964).
Average accumulation rates for bulk sediment and organic matter were calculated for each marsh for selected time intervals of relevance to this study (Table III).
Post-1978 accumulation rates were consistently higher than pre-diversion levels.
Since 1920 the West Marsh has always had higher sediment and organic matter
accumulation rates than the North Marsh. This may be related to the higher percentage area in row crops in the West Marsh watershed or to the trapping of these
PESTICIDE ACCUMULATION RATES IN A MANAGED MARSH
395
materials in some remaining areas of old field and wet forests situated between
farmland and the North Marsh.
Consequently, the pattern of material accumulation in the West Marsh was an
unsatisfactory reference for the North Marsh and we focused our analysis on quantifying the change over time within each marsh, rather than evaluating the differences between both marshes. Normalizing accumulation rates to each marsh’s prediversion rates reduced the variability associated with differences in each marsh’s
respective watershed.
Sediment accumulation rates in the North Marsh from 1978–1987 were 1.7
times higher than those calculated for the 1920–1978 period. In the West Marsh,
this rate nearly doubled compared to pre-diversion levels (1920–1977). From 1988
to 1997, sediment accumulation in the West Marsh decreased. On the other hand,
the rate at which the North Marsh trapped sediment continued to increase during
the last two decades. Interestingly, the organic matter fraction of recent deposits
was higher in the North Marsh than in the West Marsh (17% vs. 13%) perhaps
reflecting continued inflow of nutrients from farmland.
4.3. P ESTICIDE ACCUMULATION RATES
The soil series represented in the watersheds of both marshes were characterized by
their high clay-content, low to very low permeabilities, resulting in low percolation
and high runoff rates. Both watersheds had large areas used for row crops that
received significant quantities of fertilizer and pesticides as described in the soil
survey of the region (Paschall et al., 1928). The combination of soils with low
permeabilities and the application of persistent chemicals created a situation where
significant quantities of chemicals may have been present in runoff from these
fields. Although both watersheds have approximately 150 ha of land in row crops,
row crops only comprise 71% (average) of the North Marsh watershed. Twentyone percent of the North Marsh watershed was forest, marsh or old field, mostly
situated between the farms and the North Marsh, which may have filtered the runoff
from cropland.
Farms became established along the Northern border of the Winous Point wetlands prior to founding of the Shooting Club in 1856. Common crops traditionally
were corn, wheat and soybeans with the latter crop dominating in recent times.
The Brough farm, which drains into both marshes was established as an orchard
pre-1950, followed by dairy farming until the mid-1970s, prior to crop farming.
Former orchard operations included extensive use of pesticides such as DDT. Accumulation rates for the pesticides in both marshes (Figures 2–5) may be related
to anthropogenic activities. Even though these pesticides are banned from current
use, their persistent presence in the core profiles indicated continued leaching from
watershed soils, accumulation from atmospheric fallout, cycling in marsh food
webs, or illegal use.
396
A. L. SPONGBERG ET AL.
Figure 2. Accumulation rates versus year for α-, β-, and γ -hexacholorhexane pesticides in the North
and West marshes, WPSC.
PESTICIDE ACCUMULATION RATES IN A MANAGED MARSH
397
Figure 3. Accumulation rates versus year for the pesticide endrin and its metabolite endrin aldehyde
in the North and West marshes, WPSC.
398
A. L. SPONGBERG ET AL.
Figure 4. Accumulation rates versus year for the pesticide DDT and its metabolites DDE and DDD
in the North and West marshes, WPSC.
PESTICIDE ACCUMULATION RATES IN A MANAGED MARSH
399
Figure 5. Accumulation rates versus year for the pesticide aldrin and its metabolite dieldrin in the
North and West marshes, WPSC.
400
A. L. SPONGBERG ET AL.
HCH compounds, which include the popular pesticide lindane, increased in
both marshes in the mid-1960s (Figure 2). The North Marsh, which is still impacted by agricultural runoff showed an abrupt increase around 1965, followed
by relatively stable accumulation rates until the past five years, when they seem
to decrease. The West Marsh showed a steady increasing trend continuing to the
present.
Endrin was present in very low accumulation rates in both marshes until the
1960s as well (Figure 3), however, the West Marsh sediments showed roughly
twice the concentration of endrin as those in the North Marsh. The proportion
of parent pesticide to its metabolite remained stable up through the most recent
samples despite its ban from usage in this country.
DDT and its metabolites show a very different pattern than the other pesticides
(Figure 4). Accumulation rates of DDT, although still being detected in the most
recent sediments, reached their maximum in the 1950s for both marshes. This
correlates well with the presence of orchards in the drainage basin. Decreasing
concentrations occurred in the 1970s correlating with its ban from usage in this
country. Rates were similar in both marshes. The major metabolite detected in both
basins was the anaerobically generated compound, DDD. Degradation of DDT was
therefore occurring in the reduced anoxic zone within the sediment, as opposed to
degradation in oxidized surface sediments or degradation prior to transport and
burial. Concentrations within the past five years have been very low to absent in
the West Marsh. Evidence of recent input, however, was seen in the North Marsh.
Source of this contamination is unknown. The continued, although decreasing,
input of these compounds into the marshes after its ban indicates persistent cycling
of this contaminant in marsh food webs, illegal usage, or delayed transport into the
wetland due to chemical or physical processes in the watershed.
Aldrin and its metabolite, dieldrin, followed a trend similar to that of endrin
(Figure 5). However, unlike the other pesticides, accumulation rates of aldrin in
the North Marsh were higher than in the West Marsh. Increases in the North Marsh
began around 1970, whereas the increase in the West Marsh was not apparent
until 1975. Furthermore, the presence of this banned compound in the most recent
sediments indicated either illegal usage or delayed transport into the wetland.
Table IV shows the pesticide accumulation rates averaged over the following
three time-intervals: 1920–1977, 1978–1987, and 1988–1997. The first interval
represents the time period prior to diking the West Marsh. Many of the trends
previously discussed are masked in these averaged data. However, data for all compounds except the DDT and its metabolites showed similar overall trends within
each marsh. The HCH, endrin and aldrin compounds showed similar averages in
the two most recent intervals in both marshes, with a substantial decrease in the
oldest interval. No major management changes occurred in the wetland marshes in
the late 1980s or 1990s, other than the ‘supposed’ reduced or discontinued usage
of the compounds in question.
PESTICIDE ACCUMULATION RATES IN A MANAGED MARSH
401
TABLE IV
Averaged pesticide accumulation rates (µg m−2 yr−1 ) for the Winous Point
Marsh cores
March
Time
interval
HCH
α
HCH
β
HCH
γ
Endrin
parent
Aldehyde
North
North
North
West
West
West
1997–1988
1987–1978
1977–1920
1997–1988
1987–1978
1977–1920
4.99
4.80
1.43
5.73
6.09
1.36
4.29
3.58
1.61
6.50
4.91
1.39
4.88
5.46
1.65
7.30
5.05
1.41
0.81
0.73
0.05
2.26
2.07
0.19
0.93
1.31
0.09
3.23
2.82
0.30
Marsh
Time
interval
DDT
parent
DDE
DDD
Aldrin
parent
Dieldrin
North
North
North
West
West
West
1997–1988
1987–1978
1977–1920
1997–1988
1987–1978
1977–1920
1.40
1.00
0.78
0.44
1.68
1.59
1.95
3.92
0.77
2.32
5.11
4.30
4.73
8.44
6.46
3.93
12.55
13.09
39.70
33.69
6.00
29.17
29.77
6.29
25.80
12.40
2.96
30.90
17.56
4.90
Interestingly, the North Marsh, which still receives agricultural runoff, does
not always show the higher pesticide accumulation rates, possibly reflecting an
airborne source for the pesticides. Furthermore, the West Marsh is surrounded by
a greater proportion of agricultural land, whereas the North Marsh is bordered, in
part, by old fields, forest and marsh lands. The higher proportion of silt- and claysized particles in the West Marsh profiles could also contribute to this difference.
The ratio of parent pesticide to its metabolite does not change greatly with time
for any of the pesticides (Table V). Recent sediments contain about the same percent or more of parent molecule, as do the sediments deposited during the assumed
timeframe of maximum pesticide usage, with the exception of Aldrin. Presence
of parent molecule is usually interpreted to mean recent usage. However, the persistent ratio from current sediments to sediments >50 years old suggests that the
compounds are of former applications. Isotopic data preclude the interpretation of
sediment mixing. Therefore, the transport of the pesticides into the sediment record
may have been inhibited by some chemical or physical process such as sorption.
Surprisingly, if this is indeed the case, the pesticides are still degrading at roughly
the same rates as the pesticide buried within the anoxic sediment column.
402
A. L. SPONGBERG ET AL.
TABLE V
Ratio of parent to metabolite compound accumulation rates
Endrin P/M
DDT/DDE
DDT/DDD
Aldrin P/M
0.87
0.55
0.55
0.72
0.25
1.01
0.30
0.12
0.12
1.54
2.72
2.03
0.70
0.73
0.65
0.19
0.33
0.37
0.11
0.13
0.12
0.94
1.69
1.28
North
1997–1988
1987–1978
1977–1920
West
1997–1988
1987–1978
1977–1920
5. Summary
Soils in both watersheds have low to very low permeabilities and are dominated
by the very poorly drained Toledo Silty Clay. Land-use practices in the watersheds
of either marsh changed little since 1950; however, both watersheds are marked
by decreased area dedicated to orchards and concurrent increase in residential
and road area. Comparing land-use between the two watersheds showed a higher
percentage of the West Marsh was used for row crops and that the North Marsh
contained substantial forested marsh and old-field areas that may have functioned
to filter farm runoff.
Overall accumulation rates for sediment and OM were higher in the West Marsh
relative to North Marsh since 1920. North Marsh sediments were composed of
approximately 80% silt and clay until about 30 yr ago when a dramatic increase
in the sand-sized proportion began. This increase may be associated with a period
of high water in Lake Erie since the early 1970s. Conversely, low proportions of
sand-sized clasts coincided with periods of low water levels. West Marsh sediments
were composed of approximately 60% clay and silt since 1920.
Whereas nutrients are required by biota within a particular range of concentrations, pesticides and many other xenobiotic compounds may be toxic even at
very low concentrations. Many researchers have documented the occurrence of
persistent organic pollutants (POPs) including polychlorinated biphenyls, polycyclic aromatic hydrocarbons and older organochlorine pesticides in lake and estuarine sediments (Gschwend and Hites, 1981; Jeremiason et al., 1994). Loading
of POPs may occur through atmospheric deposition, transport in solution or as
adsorbed species. Many of these compounds are resistant to degradation. The hydrophobic character of POPs leads to preferential sorption on particulate matter
PESTICIDE ACCUMULATION RATES IN A MANAGED MARSH
403
where they can persist for decades in the sedimentary record. In the Great Lakes,
maximum levels of POPs in sediments correspond well to periods of peak use and
manufacture of these chemicals in the U.S.A. (Eisenreich et al., 1989).
The pesticide data from Winous Point showed variations of aldrin, endrin, HCHs,
and DDT with depth that can be attributed to agricultural use. High values of HCHs
and endrin in West Marsh sediments since mid-1960’s point to a possible airborne
source. The ratio of metabolite to parent molecule did not appear to change with
sediment age. Also, there appeared to be a delay between pesticide application and
deposition in the marshes.
Acknowledgements
This work would not have been possible without the cooperation from the Winous
Point Marsh Conservancy and its wetlands biologist, Roy Kroll. We thank Nick
Kusina, Shelly Spera, and Michael Benedict for laboratory assistance and the Water
Quality Laboratory at Heidelberg College for the use of their particle-size analyzer.
This research was supported by grants from the Lake Erie Protection Fund, Society
of Wetland Scientists Student Awards Program, and University of Toledo Lake Erie
Center.
References
Appleby, P. G., Nolan, P. J., Gifford, D. W., Godfrey, M. J., Oldfield, F., Anderson, N. J. and Battarbee, R. W.: 1986, ‘210 Pb dating by low-background gamma counting’, Hydrobiologia 143,
21–27.
Barriuso, E., Baer, U. and Calvert, R.: 1992, ‘Dissolved organic matter and adsorption-desorption of
dimefuron, atrazine, and carbetamide by soils’, J. Environ. Qual. 21, 359–367.
Craft, C. B., Broome, S. W. and Seneca, E. D.: 1988, ‘Nitrogen, phosphorus and organic carbon pools
in natural and transplanted marsh soils’, Estuaries 11(4), 272–280.
Dean, W. E.: 1974, ‘Determination of carbonate and organic matter in calcareous sediments and
sedimentary rock by loss on ignition: A comparison with other methods’, J. Sed. Petrol. 44,
249–253.
Dortch, M. S.: 1996, ‘Removal of solids, nitrogen, and phosphorus in the Cache River Wetland’,
Wetlands 16(3), 358–365.
Eisenreich, S. J., Capel, P. D., Robbins, J. A. and Bourbonniere, R.: 1989, ‘Accumulation and
diagenesis of chlorinated hydrocarbons in lacustrine sediments’, Environ. Sci. Technol. 23,
1116–1126.
Ewel, K. C. and Odum, H. T.: 1978, ‘Cypress Swamps for Nutrient Removal and Wastewater Recycling’, in M. B. Wanielista and W. W. Eckenfelder Jr. (eds), Advances in Water and Wastewater
Treatment Biological Nutrient Removal, Ann Arbor Sci. Pub., Inc., Ann Arbor, MI, pp. 181–198.
Frank, R., Sirons, G. J., Thomas, R. L. and McMillan, K.: 1979, ‘Triazine residues in suspended
solids (1974–1976) and water 1977 from the mouths of Canadian streams flowing into the Great
Lakes’, J. Great Lakes Res. 5, 131–138.
Gäggeler, H., Von Gunten, H. R. and Nyffeler, U.: 1976, ‘Determination of 210 Pb in lake sediments
and air samples by direct gamma ray measurement’, Earth Planet. Sci. Lett. 33, 119–121.
404
A. L. SPONGBERG ET AL.
Goldberg, E. D.: 1963, ‘Geochronology with 210 Pb’, in Radioactive Dating, Internat. Atom. Energy
Agency, Vienna, pp. 121–131.
Gottgens, J. F. and Liptak, M. A.: 1998, ‘Longterm assimilation of agricultural runoff in a Lake Erie
Marsh’, Verh. Internat. Verein. Limnol. 26, 1337–1342.
Gottgens, J. F., Swartz, B. P, Kroll, R. W. and Eboch, M.: 1998, ‘Long-term GIS-based records of
habitat changes in a Lake Erie coastal marsh’, Wetlands Ecol. and Manage. 6, 5–17.
Gottgens, J. F., Spongberg, A. L. and Muller, B. E.: 1999a, ‘Long-term Non-point Pollution Abatement by a Lake Erie Marsh and its Implications for Wetland Restoration Policies’, Final Report
to the Ohio Lake Erie Commission, Lake Erie Protection Fund 97-49, Toledo, Ohio, p. 79.
Gottgens, J. F., Rood, B. E., Delfino, J. J. and Simmers, B. S.: 1999b, ‘Uncertainty in paleoecological
studies of mercury in sediment cores’, Water, Air, and Soil Pollut. 110, 313–333.
Gschwend, P. M. and Hites, R. A.: 1981, ‘Fluxes of polycyclic aromatic hydrocarbons to marine and
lacustrine sediments in the northeast United States’, Geochim. Cosmochim. Acta. 45, 2359–2367.
Guo, L., Bicki, T. J., Felsot, A. S. and Hinesley, T. D.: 1993, ‘Sorption and movement of alachlor in
soil modified by carbon-rich wastes’, J. Environ. Qual. 22, 186–194.
Hey, D. L., Kenimer, A. L. and Barrett, K. R.: 1994, ‘Water quality improvement by four
experimental wetlands’, Ecol. Eng. 3, 318–197.
Jeremiason, J. D., Hornbuckle, K. C. and Eisenreich, S. J.: 1994, ‘PCBs in Lake Superior, 1978–
1992 – Decrease in water concentration reflect loss be volatilization’, Environ. Sci. Tech. 28(4),
903–914.
Kadlec, R. H. and Hey, D. L.: 1994, ‘Constructed wetlands for river water quality improvement’,
Wat. Sci. Tech. 29(4), 159–168.
Kadlec, R. H. and Knight, R. L.: 1996, Treatment Wetlands, CRC Press Inc., Boca Raton, FL, p. 893.
Kalousková, N.: 1989, ‘Adsorption of atrazine on humic acids’, J. Environ. Sci. Health B24, 599–617.
Klarer, D. M. and Mille, D. F.: 1989, ‘Amelioration of storm-water quality by a freshwater estuary’,
Archiv. Hydrobiol. 116, 375–389.
Krishnaswami, S. and Lal, D., 1978, ‘Radionuclide Limnochronology’, in A. Lerman (ed.), Lakes –
Chemistry, Geology, Physics, Springer-Verlag, New York, NY, pp. 153–177.
Locke, M. A.: 1992, ‘Sorption-desorption kinetics of alachlor in surface in surface soil from two
soybean tillage systems’, J. Environ. Qual. 21, 558–566.
Mitsch, W. J., Reede, B. C. and Klarer, D. M.: 1989, ‘The Role of Wetlands in the Control of Nutrients
with a Case Study of Western Lake Erie’, in W. M. Mitsch and S. E. Jørgensen (eds), Ecological
Engineering, Wiley-Intersci. Pub., New York, NY, pp. 129–158.
Mitsch, W. J. and Gosselink, J. G.: 2000, Wetlands, 3rd ed., John Wiley & Sons, New York, NY,
p. 920.
Paschall, A. H., Steele, J. G., Conrey, G. W. and Phillips, S. W.: 1928, Soil Survey of Ottawa County,
Ohio, USDA Bureau of Chemistry and Soils, Washington, D.C., p. 38.
Richardson, C. J.: 1985, ‘Mechanisms controlling phosphorus retention capacity in freshwater
wetlands’, J. Water Poll. Contr. Fed. 55, 495–505.
Robbins, J. A. and Edgington, D. N.: 1975, ‘Determination of recent sedimentation rates in Lake
Michigan using 210 Pb and 137 Cs’, Geochim. Cosmochim. Acta 39, 285–304.
Sheets, T. J.: 1970, ‘Persistence of triazine herbicides in soils’, Residue Reviews 32, 287–310.
Sherman, D. E., Kroll, R. W. and Engle, T. L.: 1996, ‘Flora of a diked and an undiked southwestern
Lake Erie wetland’, Ohio J. Sci. 96(1), 4–8.