Content uploaded by Mark B. Edlund
Author content
All content in this area was uploaded by Mark B. Edlund on Mar 28, 2015
Content may be subject to copyright.
ORIGINAL PAPER
Twentieth century eutrophication of the St. Croix
River (Minnesota–Wisconsin, USA) reconstructed
from the sediments of its natural impoundment
Mark B. Edlund ÆDaniel R. Engstrom Æ
Laura D. Triplett ÆBrenda Moraska Lafrancois Æ
Peter R. Leavitt
Received: 18 January 2008 / Accepted: 5 September 2008
!Springer Science+Business Media B.V. 2009
Abstract Evaluation of land-use effects on coastal
and marine ecosystems requires better understanding
of the role of rivers in regulating mass transport from
terrestrial to oceanic environments. Here we take
advantage of the presence of a riverine lake to use
paleoecological techniques to quantify impacts of
logging, European-style agriculture, urbanization and
continued terrestrial disturbance on mass transport
and water quality in the northern drainage of the
Mississippi River. Two 2-m sediment-cores recov-
ered in 1999 from Lake St. Croix, a natural
impoundment of the St. Croix River, were dated
using
210
Pb and
137
Cs, and analyzed for historical
changes (c. 1840–present) in sediment magnetic
susceptibility, inorganic and organic matter content,
biogenic silica, fossil pigments, and diatom micro-
fossils. Inorganic sediment accumulation increased
threefold between the mid-1800s and present,
whereas clear signs of eutrophication were only
evident after the mid-twentieth century when bio-
genic silica accumulation increased sixfold, diatom
accumulation rates increased 20- to 50- fold, and the
diatom community shifted from predominantly ben-
thic species to assemblages composed mainly of
planktonic taxa. Similarly, fossil pigment concentra-
tions increased during the 1960s, and diatom-inferred
total phosphorus (DI-TP) increased from *30 lg TP
l
-1
c. 1910 to *60 lgl
-1
since 1990, similar to
historical records since 1980. Together, these patterns
demonstrate that initial land clearance did not result
in substantive declines in water quality or nutrient
mass transport, instead, substantial degradation of
downstream environments was restricted to the latter
half of the twentieth century.
Keywords Biogenic silica !Diatoms !
Fossil pigments !Gulf of Mexico !
Hypoxia !Mississippi River !Nutrients !
Paleolimnology !Phosphorus
This is one of eight papers dedicated to the ‘‘Recent
Environmental History of the Upper Mississippi River’’
published in this special issue of the Journal of
Paleolimnology. D. R. Engstrom served as guest editor of the
special issue.
M. B. Edlund (&)!D. R. Engstrom !L. D. Triplett
St. Croix Watershed Research Station, Science Museum
of Minnesota, 16910 152nd St. N, Marine on St. Croix,
MN 55047, USA
e-mail: mbedlund@smm.org
B. M. Lafrancois
US Department of Interior, National Park Service, 16910
152nd St. N, Marine on St. Croix, MN 55047, USA
P. R. Leavitt
Department of Biology, University of Regina, Regina,
SK S4S 0A2, Canada
Present Address:
L. D. Triplett
Department of Geology, Gustavus Adolphus College,
St. Peter, MN 56082, USA
123
J Paleolimnol
DOI 10.1007/s10933-008-9296-1
Introduction
Landscape changes due to natural processes and
human modification have impacted receiving waters
worldwide. Elevated sediment and nutrient loads to
surface waters from point and non-point sources
characterize many anthropogenic impacts. The result
has been wide-spread eutrophication in North Amer-
ica, with *50% of impaired lakes and *60% of
impaired rivers suffering from excess nutrient loads
(Carpenter et al. 1998; Correll 1998), further leading
to coastal eutrophication and ‘‘dead zones’’ from
nutrient-driven plankton production and subsequent
oxygen depletion (Rabalais et al. 2002a,b,2007).
Establishing sound management policy for control-
ling or reducing nutrient inputs to surface waters may
mitigate these trends (Smith 2003), but requires two
specific pieces of information: assessment of current
loading trends and an understanding of historical
nutrient dynamics and mass transport. Historical
water quality records may provide this second factor,
but in most cases the establishment of water quality
monitoring came as a response to already-degraded
conditions. Furthermore, sampling and analytical
methods have changed, often making historical data
difficult to compare with modern values. Instead,
paleoecological approaches can provide historical
information on the timing, magnitude, and trajectory
of environmental change, although the approach is
rarely used in lotic ecosystems.
Here we evaluate the use of sediments in riverine
lakes as archives of land-use changes, water quality
decline, and mass transport of inorganic matter,
nutrients, and silica (Si) in an uppermost drainage of
the Mississippi River, a key region implicated in the
degradation of marine environments in the Gulf of
Mexico (Alexander et al. 2000). Mid-continental
land-use changes that drive the degradation of rivers
and coastal zones include population growth, dis-
charges from municipal treatment plants, broad-scale
upland (land clearance, conversion to European-style
agriculture, wetland drainage) and riparian (flood
control, channelization, riparian loss) change, and
application of chemical fertilizer (Rabalais et al.
2002a,b; Mulla and Sekely this issue). Recent
landscape change and environmental impacts are
readily monitored; however, an understanding of
historical land-use patterns and their impact on rivers
and coastal zones remains a critical component for
developing sound nutrient and sediment control
policy (Rabalais et al. 2002a,b,2007).
Paleolimnology offers a powerful approach for
assessing historical environmental change from lake-
sediment records, especially nutrient-driven trophic
change. But paleolimnology is often limited in its
utility when applied to rivers (but see Hall et al. 1999).
Erosion and sedimentation are usually too dynamic in
rivers to permit accumulation of stable, conformable,
and lengthy sediment sequences. In contrast, reser-
voirs or impoundments provide one sedimentary
environment where paleolimnological techniques
can be cautiously applied to rivers (Balogh et al.
1999; Engstrom et al. this issue). However, as most
reservoirs are end products of man-made dams, few
situations exist where river sediments have complete
pre- and post-Euro-American settlement records.
Lake St. Croix is a model system for determining
the impacts of European-style agriculture, urbaniza-
tion, and continued upland disturbance on water
quality and mass transport in a northern drainage of
the Mississippi River. The lake is natural impound-
ment of the St. Croix River, a major tributary to the
Upper Mississippi River, and preserves in its sediment
column a continuous record of environmental change
including the impact of Euro-American settlement,
land clearance, agriculture, and urbanization (Andersen
et al. 1996). This study reports the analysis of diatom
remains, fossil pigments, and sediment geochemistry
in two dated sediment-cores for quantitative and
qualitative measures of historical water chemistry,
sedimentation, ecological change, and productivity.
We demonstrate that water quality change was not
severe following initial land clearance, rather that
changes in mass export and water quality were likely
greatest after 1950, establishing a temporal linkage
between ecological impacts in the Upper Mississippi
drainage with degradation in the Gulf of Mexico.
Study site
Lake St. Croix encompasses the terminal 37 km of the
St. Croix River from Stillwater, Minnesota to Pres-
cott, Wisconsin along the interstate boundary (Fig. 1).
The St. Croix River (266 km long) drains a watershed
of approximately 22,196 km
2
in conjunction with 16
J Paleolimnol
123
secondary tributary systems. The river serves a
basinwide population of over 300,000 including more
than 50 permitted point source dischargers (Edlund
et al. this issue). The lower St. Croix, including Lake
St. Croix, was designated a National Scenic Riverway
in 1972. Lake St. Croix is a natural impoundment that
was formed approximately 9,500 years ago by down-
stream progradation of the delta produced by the
Mississippi River at the headwaters of Lake Pepin
(Eyster-Smith et al. 1991). Secondary deposition of
alluvial tributary fans within the lake divides Lake
St. Croix into four sub-basins (Fig. 1).
The river basin has undergone significant land-use
changes since Euro-American settlers arrived in the
1840s. Settlement was initiated by logging interests
and the construction of the first sawmill in the lower
St. Croix watershed in 1839. Logging activity peaked
in 1889 and was done in conjunction with land
clearance and a shift to agriculture. A stable popula-
tion and agriculture dominated land use within the
Stillwater
Bayport
Hudson
Afton
Prescott
N
0 2 4 6 8
Kilometers
Lake St. Croix
Coring Sites
Basin1
KinnickinnicRiver
Valley Creek
Willow River
Mississippi River
6B,C
1A,B
X
Lakeland
X
Minnesota
Wisconsin
St.Croix
Watershed
Fig. 1 Map of Lake St.
Croix showing location
on Minnesota–Wisconsin
border, multiple sub-basins,
major towns and cities,
and location of coring sites
6B and 1B (X)
J Paleolimnol
123
watershed from 1880 to 1940; farming acreage peaked
c. 1935. From 1940 to present farming interests
declined with increased urbanization and parceling
associated with an expanding St. Paul-Minneapolis
metropolitan area in the lower St. Croix watershed
(Andersen et al. 1996; Edlund et al. this issue).
Water quality monitoring data for the St. Croix
River are available in minimal form since the 1950s;
regular monitoring only dates to the mid-1970s. A
general, although not significant, decrease in total
phosphorus (TP) from c. 80 lgl
-1
TP to 50 lgl
-1
TP
occurred between the time periods 1950–1975 and
1976–1992 (Troelstrup et al. 1993). A more recent
trend analysis used Lake St. Croix data from 1976 to
2001 and identified slight but statistically signifi-
cant TP declines of 0.6–1.0 lgl
-1
year
-1
for two
sites. Mean TP concentrations declined from 65 to
70 lgl
-1
in the mid-1970s to 46–49 lgl
-1
in the mid
1990s (Lafrancois et al. this issue).
Methods
Two sediment-cores were recovered from Lake St.
Croix in October 1999 using a drive-rod piston corer
equipped with a 2.4 m long, 7 cm diameter polycar-
bonate barrel (Wright 1991). Core ‘‘6B’’ was 2.03 m
long and recovered near Lakeland, Minnesota, from
14.86 m of water (44"56051.17800 N, 92"45019.34000 W;
Fig. 1). Core ‘‘1B’’ was 1.89 m long and recovered
upstream of Prescott, Wisconsin, from 11.9 m of
water (44"45027.58400 N, 92"48025.25000W; Fig. 1).
The top 20 cm of each core were extruded in the field
in 2-cm increments; the remaining core was sealed
and transported to 4"C laboratory storage.
Cores were cut into 1-m sections for magnetic
susceptibility logging on a Bartington MS2 core
logging sensor with an automated trackfeed. Suscep-
tibility measures were taken at 1-cm intervals which
integrated a signal over a 5–10 cm length of core.
Following susceptibility logging, cores were sec-
tioned in 2-cm intervals, and stored at 4"C.
Loss-on-ignition analysis used sediment subsam-
ples dried at 105"C for 24 h to determine dry density,
then heated at 550 and 1,000"C to determine organic,
carbonate, and inorganic content from post-ignition
weight loss (Dean 1974). Data are expressed
as percentage of dry sediment weight and mass
accumulation rates.
Sediment chronology was determined using standard
alpha (
210
Pb) and gamma (
137
Cs) mass spectrometric
techniques (Binford 1990; Engstrom et al. 2006). Core
dates and sedimentation rates were calculated using the
constant rate of supply model (Appleby and Oldfield
1978).
Secondary cores from transects 1 and 6 were
collected in July and October 2001 (core 1A, 1.76 m
recovery, 44"45028.96500N, 92"48037.74200W; core 6C,
2.00 m recovery, 44"56056.32700N, 92"44052.62200W)
for fossil pigment analysis and magnetic susceptibility
logging. Fifteen wet sediment samples from each of
the secondary cores were frozen, freeze-dried, and the
carotenoids, chlorophylls, and derivatives extracted
(18 h, 4"C, dark, under N
2
) using acetone, methanol
and water (80:15:5 by volume; Leavitt et al. 1989).
Extracts were separated into component pigments on a
Hewlett-Packard model 1050 high performance liquid
chromatography system following standard methods
(Hall et al. 1999) and reported relative to sedimentary
organic carbon contents.
Twenty freeze-dried subsamples (30 mg) from
each primary core were digested for biogenic silica
analysis using 40 ml of 1% (w/v) Na
2
CO
3
solution
heated at 85"C in a reciprocating water bath for 5 h
(DeMaster 1979; Conley and Schelske 2001). A 0.5 g
aliquot of supernatant was removed from each sample
at 3, 4, and 5 h. After cooling and neutralization with
4.5 g of 0.021N HCl solution, dissolved silica was
measured colorimetrically on a Lachat QuikChem
8000 flow injection autoanalyzer as molybdate reac-
tive silica (McKnight 1991).
Twenty-one sections from core 1B and 23 sections
from core 6B were analyzed for diatom remains.
Between 25 and 60 mg of freeze-dried material were
used for microfossil analysis. Samples were digested
following Stoermer et al. (1995) to remove organic
matter. Remaining material was dried onto coverslips
using settling chambers (Battarbee 1973) and the
coverslips mounted on microslides with Naphrax.
Siliceous remains (diatom valves and chrysophyte
cysts) along one or more random microslide transects
were counted on an Olympus BX50 microscope using
full oil immersion optics capable of N.A. 1.4 and
1250X until a total of 500 diatom microfossils was
reached. Counts were converted to absolute abun-
dance of cysts and whole diatom valves per g dry
sediment and combined with sedimentation rates to
calculate microfossil accumulation rates. Percent
J Paleolimnol
123
abundance of ecological groups reflects abundance
relative to all siliceous microfossils. Percent abun-
dances of diatom species are relative to total diatom
abundance.
Principal components analysis (PCA), a linear,
unconstrained ordination method, was used to
explore patterns in diatom species data from Lake
St. Croix sediments (CANOCO 4 software, ter Braak
and Smilauer 1998). Relatively short gradient
lengths, noted during preliminary detrended corre-
spondence analyzes (DCA), indicated that species
responses were not strongly unimodal and that a
linear ordination model was appropriate (ter Braak
and Smilauer 1998). Scaling for PCA ordinations was
focused on inter-species correlations, with species
scores divided by their standard deviations (ter Braak
and Smilauer 1998). Ordination biplots were created
for each sediment-core; sample identifiers and key
species were labeled.
Weighted averaging calibration and reconstruction
(Birks et al. 1990a) were used to infer historical water
column total phosphorus (TP) in Lake St. Croix. We
used the Ramstack et al. (2003) 55 Minnesota lakes
training set based on lakes with total phosphorus
values from 7.5 to 139.3 lgl
-1
. Training set species
and environmental data were analyzed using weighted
average regression software (C2; Juggins 2003) to
calculate TP optima for 108 diatom taxa in the
training set; the resulting transfer functions (boot-
strapped r
2
=0.68, P\0.05) were subsequently
applied with weighted averaging calibration to the
fossil diatom assemblages (Birks et al. 1990a; Juggins
2003). Within the Lake St. Croix cores, 68 diatom
taxa were present in [1% relative abundance in two
sediment samples or [5% relative abundance in one
sample (selection criteria from Ramstack et al. 2003).
Thirty-three taxa were common with the Ramstack
et al. (2003) TP training set. Initial TP estimates from
weighted averaging regression were corrected (Birks
et al. 1990a,b) using inverse deshrinking based on the
Ramstack et al. (2003) model. Bootstrapped error
estimates from the Ramstack model are based on the
initial log transformed data set; the log TP error is
-0.2488. Downcore reconstructions of environmental
variables are generally more robust if the fossil
assemblage is analogous to a modern sample in the
training set. To determine if core sections had suitable
modern analogs we used the ‘‘analog’’ function with a
chord dissimilarity measure in the analog package
(version 0.3–1; Gavin Simpson) of R (R Development
Core Team 2006). Dissimilarity scores between each
core level and its closest modern analog are compared
to the 1, 2, 5, and 10% quantile distributions of all
dissimilarity scores among the modern training set.
Downcore dissimilarity scores lower than the 1, 2, 5,
and 10% quantiles indicate excellent, very good,
good, and fair analogs, respectively, exist between the
core sample and the modern training set (Whitlock
et al. 1993).
Results
Chronology, magnetics, sedimentation rates
Both cores showed general monotonic downcore
declines in total
210
Pb activity from upper-level sedi-
ments to core depths with constant supported
210
Pb
levels (68 cm in 1B and 114 cm in 6B; Fig. 2a, b).
Unsupported
210
Pb activity in surface sediments was
relatively low and variable between cores, from 5 to
8 pCi g
-1
, suggesting some dilution of atmospheric
inputs by elevated sediment inputs (Fig. 2a, b). Cumu-
lative unsupported activity ranged from 47 to
68 pCi g
-1
, representing a mean unsupported
210
Pb
flux of between 1.5 pCi cm
-2
year
-1
(core 1B) and
2.2 pCi cm
-2
year
-1
(core 6B) in the lake. These
values range from 3 to 5 times mean regional
atmospheric deposition (0.45 pCi cm
-2
year
-1
; Urban
et al. 1990) indicating that a significant portion of the
210
Pb activity in the cores is derived from watershed
deposition and downstream export to Lake St. Croix.
Overall, correspondence between
210
Pb and
137
Cs was
good, with 1963–1964 deposition peaks of
137
Cs
correspond closely with the
210
Pb date models;
137
Cs
peaks are located at 24–26 cm in core 1B and 46–48 cm
in core 6B peaks (Fig. 2). Magnetic susceptibility
profiles and correlations between primary (1B, 6B) and
secondary cores (1A, 6C) are reported in Triplett et al.
(this issue).
Presettlement sediment accumulation rates (0.06–
0.08 g cm
-2
year
-1
) were similar for cores 6B and
1B (Fig. 3). Core 6B showed an earlier post-settle-
ment increase in sedimentation than core 1B.
Sedimentation rates began to increase in core 6B as
early as 1850 and continued with some variability to
the highest rate of 0.34 g cm
-2
year
-1
around 1970
(Fig. 3a). Since 1970, sedimentation rates in core 6B
J Paleolimnol
123
have decreased slightly to *0.25 g cm
-2
year
-1
.
Post-settlement increases in sedimentation were
not pronounced in core 1B until c. 1910 (Fig. 3b).
Sedimentation rates in this core increased to a
local maximum of 0.36 g cm
-2
year
-1
by 1959,
decreased to 0.23 g cm
-2
year
-1
by 1982, increased
slightly to present-day values of *0.30 g cm
-2
year
-1
(Fig. 3b).
180
160
140
120
100
80
60
40
20
0
1 10
Core Depth (cm)
Total 210Pb Activity (pCi g-1)
120
100
80
60
40
20
0
1 10
Total 210Pb Activity (pCi g-1)
6B 1B
120
110
100
90
80
70
60
50
40
30
20
10
0
0 50 100 150 200
Core Depth (cm)
210Pb Age (yr), before 1999 AD
80
70
60
50
40
30
20
10
0
0 50 100 150 200
210Pb Age (yr), before 1999 AD
6B 1B
(a) (b)
(c) (d)
Cesium-137 peak
(1963-64 AD)
Cesium-137 peak
(1963-64 AD)
Fig. 2 Lead-210 profiles
and chronology for core 6B
(a,c) and core 1B (b,d).
a,bTotal-
210
Pb activity
plots; supported
210
Pb
indicated by dashed line;
error bars represent ±1
SD. c,dCalculated age
(years before 1999 A.D.)
versus sediment depth (cm)
based on the constant rate of
supply (c. r. s.) model. The
1963–1964 peak deposition
of
137
Cs is indicated in both
cores
1840
1860
1880
1900
1920
1940
1960
1980
2000
0 0.1 0.2 0.3 0.4
210Pb Date (A.D.)
Sediment Accumulation (g cm-2 yr-1)
1840
1860
1880
1900
1920
1940
1960
1980
2000
0 0.1 0.2 0.3 0.4
Sediment Accumulation (g cm-2 yr-1)
6B 1B
(a) (b)
Fig. 3 Dry-mass
sedimentation accumulation
(g cm
-2
year
-1
) versus
210
Pb date for acore 6B
and bcore 1B
J Paleolimnol
123
Loss-on-ignition
Lake St. Croix sediments are dominated by inorganics
(Fig. 4). Percent organic, carbonate, and inorganic
content varied little during the last 200 years in cores
6B (14.40 ±1.40, 2.59 ±1.04, 81.60 ±2.27 mean ±
SD weight percent of each metric, respectively) and 1B
(9.22 ±1.23, 5.54 ±1.41, 85.23 ±1.08 mean ±SD
weight percent, respectively). In contrast, mass accu-
mulation rates of organic, inorganic and carbonate
deposition substantially increased starting c. 1850 in
core 6B and c. 1900 in core 1B. Inorganic accumulation
rates peaked in the 1960–1970s at about 0.3 g cm
-2
year
-1
before decreasing upcore. Carbonate and
organic accumulation rates have remained relatively
constant after the 1920s in both cores (Fig. 4).
Fossil pigments
A general increase in upcore concentrations of fossil
pigments, with especially dramatic increases after the
1960s, was noted in both cores 6C and 1A. In core 6C
low and variable concentrations of pigment deriva-
tives continued until c. 1925 (Fig. 5a). Following
1925, most pigment derivatives increased to an initial
concentration peak in the 1940s and then decreased
during the 1950s (Fig. 5a). Following the mid-1960s,
a sharp increase in concentration began which has
persisted through the last three decades (Fig. 5a). In
core 1A, low and variable concentrations of most
pigment derivatives characterized the sediments until
the mid-1930s (Fig. 5b). A period of low concentra-
tion followed, which lasted until the 1950s. Since that
time, fossil pigment concentrations increased signif-
icantly. For example, the blue-green indicators
canthaxanthin and echinenone show abrupt post-
1960s increases. Pigments have generally remained at
high concentrations in the sediment record since the
1970s (Fig. 5b).
Microfossil and biogenic silica analysis
Over 350 diatom taxa and chrysophyte cysts were
identified in sediment-cores from Lake St. Croix. Total
annual microfossil accumulation was higher in core
6B (6 910
6
–131 910
6
microfossils cm
-2
year
-1
)
than core 1B (0.5 910
6
–36 910
6
microfossils
cm
-2
year
-1
; Fig. 6a, b). Both cores had near-maxi-
mum abundance of microfossils in their surface
sediments (Fig. 6a, b) with major increases after
1950 (Fig. 6a, b). Accumulation of diatoms was low
and dominated by benthic species before c. 1950
(Fig. 6a, b). After 1950, flux of both benthic and
planktonic diatoms increased markedly, especially
planktonic species. By 1950, dominance shifted from
benthic to planktonic species in Lake St. Croix
(Fig. 6a, b).
Cores 1B and 6B have 1.2–14.6 weight percent
biogenic silica. Core 6B has markedly higher weight
percent and historical accumulation of biogenic
silica (Fig. 6c). Pre-Euro-American rates of biogenic
silica accumulation were *2.0–3.0 mg cm
-2
year
-1
(Fig. 6c, d). Biogenic silica accumulation began
increasing c. 1920 with more dramatic increases
after 1950 (Fig. 6c, d). In core 6B, biogenic silica
1800
1850
1900
1950
2000
0 0.1 0.2 0.3 0.4
Accumulation rate (
g
cm
-2
yr
-1
)
1800
1850
1900
1950
2000
0 10 20 80 90 100
% dry weight
1800
1850
1900
1950
2000
0 10 20 80 90 100
210Pb Date (A.D.)
% dry weight
1800
1850
1900
1950
2000
0 0.1 0.2 0.3 0.4
210Pb Date (A.D.)
Accumulation rate (g cm
-2
yr
-1
)
(c) (d)
(a) (b)
6B
6B
1B
1B
inorganic
organic
carbonate
Fig. 4 Sediment composition and accumulation as determined
by loss-on-ignition versus
210
Pb date. aPercent dry mass of
inorganics, organics. and carbonates. bAccumulation rates of
inorganics, organics and carbonates (g cm
-2
year
-1
)
J Paleolimnol
123
peaked in the 1960s and has fluctuated since around
34 mg cm
-2
year
-1
(Fig. 6c). In core 1B, biogenic
silica increased upcore to a modern maximum of
13.5 mg cm
-2
year
-1
(Fig. 6d).
There was a clear shift of dominance from benthic to
planktonic diatoms within Lake St. Croix during the
post-Euro-American settlement period (Figs. 7,8).
From presettlement to c. 1920, benthic diatoms
composed over 60% of the diatom assemblage pre-
served in both cores with planktonic species always
less than 30% relative abundance (Figs. 7,8). By the
1950s planktonic forms had become greater than
50% of the total assemblage. Planktonic abundance
has remained at *60–75% for the last 40 years
(Figs. 7,8).
In sediments deposited before 1920 in both cores,
the planktonic flora was dominated by Aulacoseira
granulata,A.ambigua and Stephanodiscus niagarae
(Figs. 7,8). The pre-1920 benthic flora was com-
posed of a diverse assemblage dominated by ‘‘small
fragilarioid’’ taxa and mono- and bi-raphid attached
forms (Figs. 7,8). After 1920, the plankton domi-
nants Aulacoseira granulata,A.ambigua, and
Stephanodiscus niagarae increased in relative abun-
dance (Figs. 7,8). New planktonic taxa appeared in
Lake St. Croix as conditions changed. Stephanodis-
cus parvus and Fragilaria capucina var. mesolepta
became established or increased in abundance shortly
after 1910 A.D.; A.subarctica became established for
the first 50 years of the twentieth century (Figs. 7,8).
In core 6B, Fragilaria capucina and A.alpigena
appeared during the 1930s and 1940s, and the
eutrophic indicators F.crotonensis,Cyclostephanos
invisitatus, and C.tholiformis became abundant after
1950 (Fig. 7). These same trends were seen but with
slightly lower abundances and slightly later appear-
ances of each taxon in core 1B (Fig. 8). Most benthic
taxa declined in relative abundance in both cores
during the last 200 years to minimum values in the
uppermost sediments (Figs. 7,8).
1800
1850
1900
1950
2000
0 25 50 75
alloxanthin
0 10 20 3 0 40
fucoxanthin
0 25 50
diatoxanthin
0 10 20 30
lutein-zea.
0 5 10 15
canthaxanthin
0246
echinenone
0 25 50 7 5 100
chlorophyll a
0 25 50
beta-carotene
Core 1A
Date (A.D.)
nmol g-1 organic matter
1800
1850
1900
1950
2000
0 50 100 150
alloxanthin
0 20 40 60
fucoxanthin
0 25 50 7 5 100
diatoxanthin
0 25 50 7 5 100
lutein-zea.
0 5 10 15 20
canthaxanthin
0 5 10 15 20
echinenone
0 50 100 150
chlorophyll a
0 25 50 7 5 100
beta-carotene
Core 6C
Date (A.D.)
nmol g-1 organic matter
(a)
(b)
Fig. 5 Downcore distribution of select fossil pigments and their derivatives for acore 6C and bcore 1A. Pigment concentrations are
expressed as nmol g
-1
organic matter as determined by loss-on ignition at 550"C. Note non-uniform scaling of abscissae
J Paleolimnol
123
Ordination of fossil diatom assemblages using PCA
revealed comparable temporal trends in species abun-
dances and an increasingly variable community
structure in both cores. The first two PCA axes
explained 78.8 and 7.4% of the variation in species
relative abundance in core 6B, and 83.0 and 7.6% of
the variation in core 1B. The first axis in both cores
describes the dominant shifts in species composition,
from benthic dominants (Pseudostaurosira brevistri-
ata, P. brevistriata v. inflata, Staurosirella pinnata,
and Staurosira construens) in pre-Euro-American to
early twentieth century samples, to planktonic domi-
nants (Fragilaria capucina v. mesolepta, F. crotonensis,
Stephanodiscus parvus, Aulacoseira ambigua, S. niag-
arae and A. granulata) in more recent samples (Fig. 9).
Although the timing of the shift from benthic-dominated
to planktonic dominated is lagged to the 1950s in core
1B, similar patterns of community change andincreased
upcore variability are recorded in both cores (Fig. 9).
Total phosphorus reconstructions
Between 55.1 and 87.4% of the diatom assemblage in
cores 1B and 6B were used in the reconstruction of
historical water column TP (Fig. 10). Upcore samples
had excellent to good modern analogs (Core 1B
1945–1999 and 6B 1957–1999). Except for a few
samples, older samples in both cores had fair modern
analogs. Reconstructed pre-settlement values were
between 26.3 and 31.4 lgl
-1
TP with downstream
core 1B showing slightly higher presettlement values
(Fig. 10). In both cores, historical water column TP
values remained at or near pre-settlement levels until
c. 1910–1920, when reconstructed TP values began to
increase in both sub-basins (Fig. 10). In core 6B, TP
values rose more rapidly to 50 lgl
-1
by the mid-
1930s and then continued in a general upcore
increase with some variability to modern recon-
structed values of 61–72 lgl
-1
TP (Fig. 10b). Core
1B showed a similar increase beginning c. 1910;
however, the rise has been more or less continuous to
modern reconstructed levels of 58–69 lgl
-1
(Fig. 10e). The DI-TP values for the past 30 year
(48.8–72.6 lgl
-1
) compared well with those derived
from monitoring data (46–73 lgl
-1
), except for
several anomalous higher values in the 1950s and
1960s, which were often based on only one or two
measures per year (Fig. 10b, e).
Discussion
Coastal eutrophication has been broadly linked to
inland landscape changes associated with increased
population, clearance, drainage and modification of
upland and riparian zones, conversion to agriculture,
and increased application of chemical fertilizer
(Rabalais et al. 2002a,b; Smith 2003). In the
Gulf of Mexico, the delivery of nutrients, particulary
N, has been identified as the primary factor in
development of hypoxic conditions; these modern
nutrient exports are significantly derived from the
1800
1850
1900
1950
2000
0 5 10 15 20 25
valves or cysts x 106 cm-2 yr-1
1800
1850
1900
1950
2000
0 20 40 60 80 100
210Pb Date (A.D.)
valves or cysts x 106 cm-2 yr-1
1B
6B
benthic
diatoms
chrysophyte cysts
planktonic
diatoms
1800
1850
1900
1950
2000
0 10 20 30 40
210Pb Date (A.D.)
bio
g
enic silica m
g
cm-2 yr-1
6B
1800
1850
1900
1950
2000
0 5 10 15
bio
g
enic silica m
g
cm-2 yr-1
1B
(a) (b)
(c) (d)
Fig. 6 Siliceous microfossil and biogenic silica accumulation
versus
210
Pb date. Microfossils are separated into three
ecological groups: chrysophyte cysts, planktonic diatoms, and
benthic/periphytic diatoms. aSiliceous microfossil accumula-
tion for core 6B. bSiliceous microfossil accumulation for core
1B. cBiogenic silica accumulation for core 6B. dBiogenic
silica accumulation for core 1B
J Paleolimnol
123
Upper Mississippi River (Alexander et al. 2000;
Rabalais et al. 2002a,b). Historical reconstructions
and coastal paleoecology in the Gulf suggest that the
development of large-scale hypoxia since the 1950s is
coincident with increased nitrogen loads and shifts in
nutrient ratios (Si:nitrate) from the Mississippi River
(Rabalais et al. 2002a,b,2007). However, little
evidence has been proffered on the historical role of
rivers, especially upstream tributaries, in regulating
land-use impacts and mass transport from the conti-
nental to coastal environments. Our study clearly
establishes the historical linkage between land-use,
increased nutrient and sediment loads, their ecolog-
ical impacts, and mass transport from the Upper
Mississippi drainage to coastal eutrophication in the
Gulf of Mexico.
0
2000
1950
1900
1850
1800
Lake St. Croix Core 6B, percent relative abundance
planktonic diatoms
Aulacoseira granulata
Aulacoseira ambigua
Stephanodiscus niagarae
Fragilaria crotonensis
Cyclostephanos invisitatus
Fragilaria capucina
Aulacoseira alpigena
Cyclostephanos tholiformis
Stephanodiscus parvus
Frag. cap. v. mesolepta
Aulacoseira subarctica
Date (A.D.) Date (A.D.)
0
2000
1950
1900
1850
1800
benthic diatoms
Staurosirella pinnata
Fragilaria vaucheriae
Pseudost. brev. v. inflata
Martyana martyii
Staurosira construens
Amphora pediculus
Cocconeis neodiminuta
Plano. frequentissimum
St. construens v. venter
St. leptostauron v. dubia
Cocc. plac. v. euglypta
Pseudost. brevistriata
Cavinula scutelloides
80604020 0 40
302010 0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 1020
10080
6040
20 100 100 100 100 100 100 100 100 100 100 100 100
20 10020
Fig. 7 Relative abundance of planktonic and benthic diatom taxa present at [3% total diatom counts in core 6B sediments. Upper
panels are planktonic taxa and lower panels are benthic/periphytic taxa
J Paleolimnol
123
The St. Croix River, with its natural impoundment,
Lake St. Croix, has shown to be a model system for
determining the role of large rivers in recording
impacts of historical land-use change, the ecological
response to those impacts, and the resulting storage or
downstream transport of sediment and nutrient
loads. The sediment record from Lake St. Croix
accurately documented all known historical changes
in St. Croix basin land-use (Andersen et al. 1996).
Increased nutrients, increased diatom and biogenic
silica accumulation, shifts toward plankton-dominated
production, establishment of eutrophic indicators,
increased sedimentation rates, increased pigment
concentrations, and increased accumulation of
carbonates and organic matter provide clear and multi-
ple upcore signals of post-Euro-American settlement
Lake St. Croix Core 1B, percent relative abundance
benthic diatoms
Staurosirella pinnata
Fragilaria vaucheriae
Pseudost. brev. v. inflata
Martyana martyii
Staurosira construens
Amphora pediculus
Cocconeis neodiminuta
Plano. frequentissimum
St. construens v. venter
St. leptostauron v. dubia
Cocc. plac. v. euglypta
Pseudost. brevistriata
Cavinula scutelloides
0
2000
1950
1900
1850
1800
Date (A.D.)
planktonic diatoms
Aulacoseira granulata
Aulacoseira ambigua
Stephanodiscus niagarae
Fragilaria crotonensis
Cyclostephanos invisitatus
Fragilaria capucina
Aulacoseira alpigena
Cyclostephanos tholiformis
Stephanodiscus parvus
Frag. cap. v. mesolepta
Aulacoseira subarctica
Date (A.D.)
0
0
2000
1950
1900
1850
1800
80604020
0 80604020
40302010 0 40302010 0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 10
20
10 010 010 010 010 010 010 010 010 010 010 010 010
Fig. 8 Relative abundance of planktonic and benthic diatom taxa present at [3% total diatom counts in core 1B sediments. Upper
panels are planktonic taxa and lower panels are benthic/periphytic taxa
J Paleolimnol
123
eutrophication of Lake St. Croix. Most of these changes
did not occur immediately upon settlement and land
clearance. Instead, the changes occurred in the early to
middle twentieth century. The timing of many changes
recorded in the sediments was later in downstream core
1B compared to core 6B. Differences in upstream urban
centers and land-use, hence loads, and/or sufficient
residence time in the northern sub-basin to assimilate
sediment and TP loads through production and sedi-
mentation likely explains these lags.
1974
1997
1985
1963
1994
1936
1999
1968
1918
1951
1957
1892
1980
1812
1875
1859
1990
1699
Core 6B
A. ambigua
+1.0
-1.0
-1.0
A. granulata
S. niagarae
S. parvus
F. capucina
v. mesolepta
F. crotonensis
P. brevistriata
S. construens
S. pinnata
P. brevistriata
v. inflata
PCA axis 1 (78.8%)
1928
1907
1944
1966
1972
1959
1982
1992
1988
1999
1996
1946
1952
1939
1892
1910
1931
1826
1868
1922
1627
Core 1B
A. ambigua
+1.0
-1.0
A. granulata
S. niagarae
S. parvus
F. capucina
v. mesolepta
F. crotonensis
P. brevistriata
S. construens
S. pinnata
P. brevistriata
v. inflata
PCA axis 1 (83.0%)
(a) (b)
+1.0
PCA axis 2 (7.4%)
-1.0
+1.0
PCA axis 2 (7.6%)
Fig. 9 Principal Components Analysis ordination plots of
downcore samples (filled circles) and selected species scores
(filled squares) with percentage variance explained by each
axis indicated for acore 6B and bcore 1B. Dated (A.D.)
samples are connected in a temporal sequence. Diatom species
plotted include these subdominants: Aulacoseira ambigua,A.
granulata,Fragilaria capucina v. mesolepta,F.crotonensis,
Pseudostaurosira brevistriata,P.brevistriata v. inflata,Stau-
rosira construens,Staurosirella pinnata,Stephanodiscus
niagarae, and S. parvus
Core 6B
% diatoms used
1008060
Core 6B
-0.8-1.2-1.6-2.0
log TP
Core 1B
Diatom-inferred TP
(g/L)
12010080604020
**
***
*
% diatoms used
Core 1B
1008060
Date (A.D.)
-0.8-1.2-1.6-2.0
log TP
1800
1850
1900
1950
2000
Core 6B
(a) (b) (c) (d) (e) (f)
Diatom-inferred TP
(g/L)
*
Core 1B
12010080604020
**
*
*
*
*
**
(186 g/L)
Fig. 10 Diatom reconstructed water column total phosphorus
versus
210
Pb date. aDiatom reconstructed logTP (mg l
-1
), core
6B. Error bars represent sample specific errors for the
Ramstack et al. (2003) model. bDiatom-inferred TP (lg/l),
core 6B. Asterisks represent 5 year means of surface water TP
concentrations sampled at Hudson, Wisconsin (STORET and
MCES data). cPercent of fossil diatom flora used in TP
reconstruction, core 6B. dDiatom reconstructed logTP
(mg l
-1
), core 1B. Error bars represent sample specific errors
for the Ramstack et al. (2003) model. eDiatom-inferred TP
(lgl
-1
), core 1B. Asterisks represent 5 year means of surface
water TP concentrations sampled at Prescott, Wisconsin
(STORET and MCES data). fPercent of fossil diatom flora
used in TP reconstruction, core 1B
J Paleolimnol
123
Logging and land clearance for agriculture in the
St. Croix basin were initiated in the 1840s; logging
peaked c. 1890 (Andersen et al. 1996). In upper Lake
St. Croix, these land-use changes caused immediate
increases in sedimentation rates of primarily inor-
ganic clastics, a trend that continues upcore.
Presettlement sedimentation rates were similar to
those found in Lake Pepin, a natural impoundment on
the Mississippi River (Engstrom et al. this issue);
however, post-settlement increases in Lake St. Croix
were much lower than post-settlement increases in
Lake Pepin. From c. 1870 to 1940, agriculture was
the primary land-use in the lower St. Croix. Contin-
ued increases in sedimentation over this time period
reflect the increasingly agricultural character and
increasing intensity of agriculture in the landscape.
Andersen et al. (1996) reported that 50% of the lower
St. Croix basin was in agriculture by 1880, and that
by 1928 over 75% of SW Wisconsin farms, including
in the lower St. Croix, had serious erosion problems.
Sedimentation rates continued to increase in Lake
St. Croix into the 1960s. Analysis of additional cores
from Lake St. Croix showed that catastrophic
erosional events along several tributaries in mid-
century contributed major sources of sediment to the
river (Triplett et al. this issue). Post-1960 decreases in
sedimentation rates were confirmed by monitoring
(Lafrancois et al. this issue) and may be an indication
of less intensive agriculture in the region, improved
soil conservation and management practices, or
reestablishment of cover as agricultural lands were
converted from cropland to pastureland or to low- to
medium-density urban land-use.
Although agricultural acreage decreased in the St.
Croix basin after 1935, historical records document
the increase use and reliance on chemical-based
fertilizers (Mulla and Sekely this issue) and the
increased delivery of phosphorus from point sources
to the river (Edlund et al. this issue). Diatom-based
reconstructions of historical water column TP track
these land-use and loading trends to show that TP
concentrations in Lake St. Croix are currently
between two- and threefold greater than prior to land
clearance. Presettlement values of 25–30 lgl
-1
began to rise c. 1910–1920 to 50 lgl
-1
by 1930,
before rapidly rising in the 1940–1960s to maximum
TP concentrations during the 1990s (58–72 lgl
-1
).
Importantly, modern diatom-inferred TP closely
approximated values measured directly during the
last two decades (Kroenig and Stark 1997; Malischke
et al. 1994). The higher historical TP levels reported
by Troelstrup et al. (1993) between 1950 and 1972
(median 80 lgl
-1
) and from STORET and MCES
records (Lafrancois et al. this issue; Fig. 10) were not
clearly shown in our reconstructions; however,
these early records are based on very few data (often
one or two measures per year). Although the
Ramstack et al. (2003) diatom calibration set is
statistically robust, a lack of representative hypereu-
trophic lakes in the calibration lake set limits diatom-
inferred TP reconstructions to maximum levels of
approximately 85 lgl
-1
. Thus some of the differ-
ence between diatom-inferred TP and monitoring
records from the 1960s and early 1970s (Fig. 10) may
reflect truncated response curves for eutrophic taxa
(Anderson 1997).
Siliceous microfossil and biogenic silica accumu-
lation similarly tracked reconstructed TP levels;
increased diatom productivity is a common ecolog-
ical response to nutrient loads (Conley et al. 1993;
Rabalais et al. 2007). Both cores had increased flux of
microfossils and biogenic silica beginning in the
1910s that peaked in the 1960s–1990s. The largest
increases in siliceous microfossil and biogenic silica
accumulation occurred in the 1950s to early 1970s.
There was concordance in timing of changes between
microfossil and biogenic silica accumulation; how-
ever, the magnitude of change between these biotic
indicators was vastly different. An increased upcore
abundance of small-celled and more lightly silicified
forms associated with eutrophy accounts for the
difference between a 20- and 50-fold increase in
microfossil accumulation and a 5.5-fold increase in
biogenic silica from presettlement levels. That an
increase in TP of only two to three fold could produce
such increases in both microfossil and biogenic silica
accumulation reflects the differences in recycling
rates of P (and N) in comparison to silica cycling and
loading rates (Conley et al. 1993; Conley 2000).
Concerns have been voiced regarding increased
burial of silica in reservoirs and impoundments due
to greater diatom production (from eutrophication);
increased deposition of biogenic silica is ultimately
affecting delivery rates of dissolved silica to the
world’s oceans (Conley et al. 1993; Humborg et al.
2000). Although it is a natural impoundment, Lake
St. Croix fits this global pattern of altered silica
cycling as a result of increased diatom productivity
J Paleolimnol
123
and rates of biogenic silica deposition (this study;
Triplett et al. this issue) (Fig. 11).
Both cores had post-settlement shifts from benthic
to planktonic productivity, a typical trophic response to
increased nutrient loads (Schelske et al. 1999; Vade-
boncoeur et al. 2003). While benthic diatoms decreased
in relative abundance, they actually increased in
absolute abundance and accumulation. Thus, both
ecological components of the diatom community
responded positively to eutrophication with greater
productivity, especially planktonic productivity
(Conley et al. 1993). The shift to plankton-dominated
production saw the relative increase of the plankton
dominants (Aulacoseira granulata,A.ambigua and
Stephanodiscus niagarae) and introduction and estab-
lishment of many planktonic diatoms considered
ubiquitous indicators of eutrophy (e.g., Cyclostephanos
invisitatus,C.tholiformis,Fragilaria crotonensis,
F.capucina and var. mesolepta). PCA confirmed the
significance of the shift from benthic to planktonic
dominance, and demonstrated increased variability of
the diatom community after the 1930s in core 6B and
after the 1950s in core 1B. Furthermore, the initial
increase in planktonic relative abundance occurred
not at initial land clearance or the onset of agriculture,
but much later and in concert with increasing TP
levels c. 1910–1920; larger increases in planktonic
abundance occurred post-1950s in both cores.
Fossil pigment records corroborate the timing of
environmental changes associated with twentieth
century eutrophication of the St. Croix River. Core
6C had initial increases in most pigments and
derivatives in the 1920s followed by more dramatic
increases in the 1960s. Core 1A primarily showed
increases following 1960. These changes are gener-
ally consistent with the records of biogenic silica and
siliceous microfossil accumulation. The blue-green
indicators canthaxanthin and echinenone show abrupt
post-1960s increases and may be associated with the
growth of bloom-forming blue-green algae during the
summer months. Cyanobacterial blooms have been
noted since the 1960s in Lake St. Croix (Brook in
Troelstrup 1993). Aphanizomenon flos-aquae was
noted in the St. Croix River as early as 1928, but not
at bloom levels (Reinhard 1931), which is consistent
with lower downcore levels of blue-green pigments.
Lastly, increased upcore flux of carbonates is likely
a signal of greater within-system productivity associ-
ated with increased nutrient loads to the river
(Engstrom and Swain 1986; Hodell et al. 1998). The
increased upcore flux of organics may signal
increased within-system productivity and/or increased
inputs of allochthonous material accompanying ero-
sion given the early post-settlement increases; the
former likely plays a significant role given the
corroborative increases in diatom, biogenic silica
and pigment accumulation.
By most standards (TP, chlorophyll a, Secchi,
various trophic state indices), modern Lake St. Croix
is eutrophic (MPCA 2001). It exceeds nutrient
standards and has recently been declared impaired
by the State of Minnesota. We have demonstrated
with multiple paleolimnological indicators that con-
ditions in St. Croix River are far from its natural pre-
settlement mesotrophic condition (Fig. 11). The most
dramatic changes in water quality, nutrient load, and
0 50 100
% planktonic
0 5 10 15
BSi (m
g
/cm2 yr)
0 10 20 30 40 50
diatoxanthin
012345
echinenone
1800
1850
1900
1950
2000
0 0.2 0.4
Date (A.D.)
sed rt (
g
/cm2 yr)
20 40 60 80
DI-TP (
g
/l)
0 200 400
pop'n (1000s)
0 0.5 1 1.5
Glauconite (%)
0 25 50 75 100
Point source P
Gulf of
Mexico
St. Croix
Basin
Core 1B Core 1A
St. Croix
Basin
Core 1B Core 1BCore 1B Core 1A
0 25 50 75 100
% PEB
Gulf of
Mexico
Core MRD05-4Core G27
Fig. 11 Summary diagram of changes recorded in Lake St.
Croix (cores 1A, 1B) and Gulf of Mexico sediments and
historical changes in the St. Croix River basin. From left to
right,panels depict sedimentation rate (g cm
-2
year
-1
),
biogenic silica flux (mg cm
-2
year
-1
), diatom-inferred total
phosphorus (lgl
-1
), and percent planktonic diatom taxa in
core 1B, echinone (nmol g
-1
organic matter) and diatoxanthin
(nmol g
-1
organic matter) in core 1A, basin population
(1000s), point source phosphorus loadings (tons P year
-1
;
see Edlund et al. this issue), percent glauconite on coarse grains
from Gulf of Mexico core G27 (adapted from Rabalais et al.
2007), and PEB index scores (% low-oxygen-tolerant benthic
foraminifers) from box core MRD05-4 (adapted from
Osterman et al. 2008)
J Paleolimnol
123
ecological condition occurred in the mid-twentieth
century, coinciding with increased nutrient loads from
point sources and increased use of chemical fertilizers
(Edlund et al. this issue; Mulla and Sekely this issue).
The downstream transport of nutrients has followed a
similar trend; phosphorus exports from the St. Croix
River increased steadily from the 1940–1990s (Trip-
lett et al. this issue). Targeted reductions in point
source loads of P have somewhat reduced P loads and
export; unfortunately, nitrate concentrations in and
export from the St. Croix River have continued to rise
(Edlund et al. this issue; Lafrancois et al. this issue).
There remains little doubt that nutrient loads from
the Mississippi River are the primary cause of coastal
eutrophication and development of the hypoxic ‘‘dead
zone’’ in the Gulf of Mexico (Rabalais et al. 2003).
Furthermore, significant quantities of nutrients reach-
ing the Gulf are derived from the Upper Mississippi
River (Alexander et al. 2000). Whereas N exports to
the Gulf have been most strongly implicated, P and
Si, alone or in elemental ratio with N, have also been
identified as contributors to Gulf hypoxia, at least
seasonally (Dodds 2006; Scavia and Donnelly 2007).
Our analysis of sediment records from St. Croix
River unambiguously establishes a temporal link to
post-1950s coastal degradation in the Gulf (Fig. 11)
by documenting a similar timeline of ecological
change in the Upper Mississippi River including the
impacts of all major land-use changes, degradation of
water quality, increased nutrient export, and the mid-
twentieth century eutrophication of the major tribu-
tary to Mississippi River.
Acknowledgments This project was supported in part by the
Minnesota Pollution Control Agency, the Metropolitan Council
Environmental Services (MCES), and NSERC Canada. Thanks
are due Scott Schellhaass (MCES) and Mike Perneil for coring
assistance, Tara Bromenshenkel, Jill Coleman, Ellen Mallman,
Diana Roen and Kelly Thommes for geochemistry, biogenic
silica, and
210
Pb dating, Joy Ramstack for providing access to
her diatom training set, and Jasmine Saros and an anonymous
reviewer for improving earlier versions of the manuscript. Doug
Schnurrenberger performed magnetic susceptibility logging at
the Limnological Research Center, Department of Geology and
Geophysics, University of Minnesota-Twin Cities.
References
Alexander RB, Smith RA, Schwarz GE (2000) Effect of stream
channel size on the delivery of nitrogen to the Gulf of
Mexico. Nature 403:758–761. doi:10.1038/35001562
Andersen O, Crow TR, Lietz SM, Stearns F (1996) Transfor-
mation of a landscape in the upper mid-west, USA: the
history of the lower St. Croix River valley, 1830-present.
Landsc Urban Plan 35:247–267. doi:10.1016/S0169-
2046(96)00304-0
Anderson NJ (1997) Reconstructing historical phosphorus
concentrations in rural lakes using diatom models. In:
Tunney H, Carton OT, Brookes PC, Johnston AE (eds)
Phosphorus loss from soil to water. CAB International,
Oxford, pp 95–118
Appleby PG, Oldfield F (1978) The calculation of lead-210
dates assuming a constant rate of supply of the unsup-
ported lead-210 to the sediment. Catena 5:1–8. doi:
10.1016/S0341-8162(78)80002-2
Balogh SJ, Engstrom DR, Almendinger JE, Meyer ML, Johnson
DK (1999) History of mercury loading in the Upper Mis-
sissippi River reconstructed from the sediments of Lake
Pepin. Environ Sci Technol 33:3297–3302. doi:10.1021/
es9903328
Battarbee RW (1973) A new method for estimating absolute
microfossil numbers with special reference to diatoms.
Limnol Oceanogr 18:647–653
Binford MW (1990) Calculation and uncertainty analysis of
210-Pb dates for PIRLA project lake sediment-cores. J
Paleolimnol 3:253–267. doi:10.1007/BF00219461
Birks HJB, Juggins S, Line JM (1990a) Lake surface-water
chemistry reconstructions from palaeolimnological data.
In: Mason BJ (ed) The surface water acidification program.
Cambridge University Press, Cambridge, pp 301–313
Birks HJB, Line JM, Juggins S, Stevenson AC, ter Braak CJF
(1990b) Diatoms and pH reconstruction. Philos Trans R
Soc Lond B Biol Sci 327:263–278. doi:10.1098/rstb.
1990.0062
Carpenter SR, Caraco NF, Correll DL, Howarth RW, Sharpley
AN, Smith VH (1998) Nonpoint pollution of surface
waters with phosphorus and nitrogen. Ecol Appl 8:559–
568. doi:10.1890/1051-0761(1998)008[0559:NPOSWW]
2.0.CO;2
Conley DJ (2000) Biogeochemical cycles and nutrient manage-
ment strategies. Hydrobiologia 410:87–96. doi:10.1023/
A:1003784504005
Conley DJ, Schelske CL (2001) Biogenic silica. In: Smol JP,
Birks HJB, Last WM (eds) Tracking environmental change
using lake sediments, vol 3: terrestrial, algal, siliceous
indicators. Kluwer Academic Publishers, Dordrecht, pp
281–293
Conley DJ, Schelske CL, Stoermer EF (1993) Modification of the
biogeochemical cycle of silica with eutrophication. Mar
Ecol Prog Ser 101:179–192. doi:10.3354/meps101179
Correll DL (1998) The role of phosphorus in the eutrophication
of receiving waters: a review. J Environ Qual 27:261–266
Dean WE (1974) Determination of carbonate and organic
matter in calcareous sediments and sedimentary rocks by
loss-on ignition; comparison with other methods. J Sed
Res 44:242–248
DeMaster DJ (1979) The marine budgets of silica and 32Si.
Ph.D. Dissertation, Yale University
Dodds WK (2006) Nutrients and the ‘‘dead zone’’: the link
between nutrient ratios and dissolved oxygen in the
northern Gulf of Mexico. Front Ecol Environ 4:211–217.
doi:10.1890/1540-9295(2006)004[0211:NATDZT]2.0.CO;2
J Paleolimnol
123
Edlund MB, Triplett LD, Tomasek M, Bartilson K (this issue)
From paleo to policy: partitioning the historical point and
nonpoint phosphorus loads to the St. Croix River, USA. J
Paleolimnol. doi:10.1007/s10933-008-9288-1
Engstrom DR, Swain EB (1986) The chemistry of lake sedi-
ments in time and space. Hydrobiologia 143:37–44. doi:
10.1007/BF00026642
Engstrom DR, Schottler SP, Leavitt PR, Havens KE (2006) A
re-evaluation of the cultural eutrophication of Lake Oke-
echobee, Florida, using multiproxy sediment records. Ecol
Appl 16:1194–1206. doi:10.1890/1051-0761(2006)016[11
94:AROTCE]2.0.CO;2
Engstrom DR, Almendinger JE, Wolin JA (this issue) Histor-
ical changes in sediment and phosphorus loading to the
Upper Mississippi River: mass-balance reconstructions
from the sediments of Lake Pepin. J Paleolimnol. doi:
10.1007/s10933-008-9292-5
Eyster-Smith NM, Wright HE Jr, Cushing EJ (1991) Pollen
studies at Lake St. Croix, a river lake on the Minnesota/
Wisconsin border, USA. Holocene 1:102–111. doi:
10.1177/095968369100100202
Hall RI, Leavitt PR, Quinlan R, Dixit AS, Smol JP (1999) Effects
of agriculture, urbanization, climate on water quality in
the northern Great Plains. Limnol Oceanogr 44:739–756
Hodell DA, Schelske CL, Fahnenstiel GL, Robbins JL (1998)
Biologically induced calcite and its isotopic composition
in Lake Ontario. Limnol Oceanogr 43:187–199
Humborg C, Conley DJ, Rahm L, Wulff F, Cociasu A, Ittekkot
V (2000) Silicon retention in river basins: far-reaching
effects on biogeochemistry and aquatic food webs in
coastal marine environments. Ambio 29:45–50. doi:
10.1639/0044-7447(2000)029[0045:SRIRBF]2.0.CO;2
Juggins S (2003) C
2
User guide Version 1.3. Software for
ecological and palaeoecological data analysis and visual-
isation. Version 0.82. University of Newcastle, Newcastle
upon Tyne
Kroenig S, Stark J (1997) Variability of nutrients in streams in
part of the Upper Mississippi River Basin, Minnesota and
Wisconsin. US Department of the Interior, US Geological
Survey Fact Sheet FS-164-97
Lafrancois BM, Magdalene S, Johnson DK (this issue) A
comparison of recent water quality trends (1976–2004)
with sediment-core records for two riverine lakes of the
Upper Mississippi River basin. Lake St. Croix Lake Pepin.
J Paleolimnol. doi:10.1007/s10933-008-9294-3
Leavitt PR, Carpenter SR, Kitchell JF (1989) Whole-lake
experiments: the annual record of fossil pigments and
zooplankton. Limnol Oceanogr 34:400–717
Malischke J, Ryan D, Sorge B, Larson N, Hartman J (1994) St.
Croix Basin water quality management plan. Wisconsin
Department of Natural Resources Publ 270-94-REV,
Madison
McKnight R (1991) QuikChem method 10-114-27-1-A,
Determination of silica in waters by flow injection anal-
ysis, 0.2–20.0 mg SiO
2
/L. Lachat Instruments, 6645 West
Mill Road, Milwaukee, Wisconsin 53218-1239
Minnesota Pollution Control Agency (MPCA) (2001) Lake
Water Quality Database. http://data.pca.state.mn.us/cgi-
bin/lkwq95ReadFull.pl?lakeid=82-0001
Mulla DJ, Sekely AC (this issue) Historical trends affecting the
accumulation of sediment and phosphorus in Lake Pepin,
Upper Mississippi River, USA. J Paleolimnol. doi:
10.1007/s10933-008-9293-4
Osterman LE, Poore RZ, Swarzenski PW (2008) The last 1000
years of natural and anthropogenic low-oxygen bottom-
water on the Louisiana shelf, Gulf of Mexico. Mar
Micropaleontol 66:291–303
R Development Core Team (2006) R: a language and envi-
ronment for statistical computing, version 2.4.1. The R
foundation for statistical computing, ISBN 3-900051-07-
0. http://www.r-project.org
Rabalais NN, Turner RE, Scavia D (2002a) Beyond science
into policy: Gulf of Mexico hypoxia and the Mississippi
River. Bioscience 52:129–142. doi:10.1641/0006-3568
(2002)052[0129:BSIPGO]2.0.CO;2
Rabalais NN, Turner RE, Wiseman WJ Jr (2002b) Gulf of
Mexico hypoxia, a.k.a. ‘‘The Dead Zone’’. Annu Rev Ecol
Syst 33:235–263. doi:10.1146/annurev.ecolsys.33.010802.
150513
Rabalais NN, Turner RE, Sen Gupta BK, Platon E, Parsons ML
(2007) Sediments tell the history of eutrophication and
hypoxia in the norther Gulf of Mexico. Ecol Appl
17:S129–S143. doi:10.1890/06-0644.1
Ramstack JM, Fritz SC, Engstrom DR, Heiskary SA (2003)
The application of a diatom-based transfer function to
evaluate regional water-quality trends in Minnesota since
1970. J Paleolimnol 29:79–94. doi:10.1023/A:1022869
205291
Reinhard EG (1931) The plankton ecology of the Upper Mis-
sissippi, Minneapolis to Winona. Ecol Monogr 1:395–
464. doi:10.2307/1943079
Scavia D, Donnelly KA (2007) Reassessing hypoxia forecasts
for the Gulf of Mexico. Environ Sci Technol 41:8111–
8117. doi:10.1021/es0714235
Schelske CL, Donar CM, Stoermer EF (1999) A test of pa-
leolimnologic proxies for the planktonic/benthic ratio of
microfossil diatoms in Lake Apopka. In: Mayama S, Idei
M, Koizumi I (eds), Proceedings of the fourteenth inter-
national diatom symposium. Koeltz Scientific Books,
Koenigstein, pp 367–382
Smith VH (2003) Eutrophication of freshwater and coastal
marine ecosystems. A global problem. Environ Sci Pollut
Res 10:126–139. doi:10.1065/espr2002.12.142
Stoermer EF, Edlund MB, Pilskaln CH, Schelske CL (1995)
Siliceous microfossil distribution in the surficial sedi-
ments of Lake Baikal. J Paleolimnol 14:69–82. doi:
10.1007/BF00682594
ter Braak CJF, Smilauer P (1998) CANOCO reference manual
and user’s guide to Canoco for windows: software for
Canonical community ordination (Version 4). Micro-
computer Power, Ithaca
Triplett LD, Engstrom DR, Edlund MB (this issue) A whole-
basin stratigraphic record of sediment and phosphorus
loading to the St. Croix River, USA. J Paleolimnol. doi:
10.1007/s10933-008-9290-7
Troelstrup NH Jr, Foley JL, Perry JA (1993) Changing patterns
of water quality and biology within the lower St. Croix
National Scenic Riverway. LCMR work elements B.2 and
B.6. Legislative Commission on Minnesota Resources, St
Paul
Urban NR, Eisenreich SJ, Grigal DF, Schurr KT (1990)
Mobility and diagenesis of Pb and
210
Pb in peat. Geochim
J Paleolimnol
123
Cosmochim Acta 54:3329–3346. doi:10.1016/0016-7037
(90)90288-V
Vadeboncoeur Y, Jeppesen E, Vander Zanden MJ, Schierup H-H,
Christoffersen K, Lodge DM (2003) From Greenland to
green lakes: cultural eutrophication and the loss of benthic
pathways in lakes. Limnol Oceanogr 48:1408–1418
Whitlock C, Bartlein PJ, Watts WA (1993) Vegetation history
of Elk Lake. In: Bradbury JP, Dean WE (eds) Elk Lake,
Minnesota: evidence for rapid climate change in the
North-Central United States. Geol Soc Am, Boulder,
Colorado, USA, Special Paper 276, pp 251–274
Wright HE Jr (1991) Coring tips. J Paleolimnol 6:37–49. doi:
10.1007/BF00201298
J Paleolimnol
123
