314
Impact of Euro-American settlement on a
riparian landscape in northeast Iowa,
midwestem USA: an integrated approach
based on historical evidence, floodplain
sediments, fossil pollen, plant macrofossils
and insects*
R.G. Baker,† D.P. Schwert,‡ E.A. Bettis III,** and
C.A. Chumbley°
(†Department of Geology, University of Iowa, Iowa City, Iowa 52242-1379, USA;
‡Department of Geosciences, North Dakota State University, Fargo, North Dakota
58105-5517, USA; **Iowa Department of Natural Resources, Geological Survey
Bureau, 123 N Capitol, Iowa City, Iowa 52242, USA; °412 Highland Avenue, Iowa
City, Iowa 52242, USA)
Received 25 August 1992: revised manuscript accepted 21 January 1993
*
Iowa Quaternary Studies Contribution No. 53.
Abstract: European settlement and attendant forest clearance and agricultural activities in northeastern
Iowa caused changes in the landscape, vegetation, insect fauna and water quality unequalled in rate and
magnitude since the melting of Wisconsinan glaciers. Historical documents show that the upper part of
the Roberts Creek drainage basin was settled between AD 1840 and 1856, and the area was under
intensive cultivation by 1880. Extensive soil erosion beginning at this time resulted in increased runoff
and more frequent flooding; aggradation rates increased by one to two orders of magnitude over those in
presettlement times, and the entire floodplain was covered with up to 1 m of sediment. Channel
widening between about 1880 and 1930 allowed the stream to accommodate greater floods, overbank
deposition decreased, and further deposits were restricted mostly to the channel belt.
The presettlement vegetation was a stable mix of wet meadows and riparian shrubs on the floodplain,
a rich aquatic community in the stream, and oak savanna on the valley walls and upland. Disturbance
from soil erosion, floodplain erosion and floodplain deposition almost completely replaced both lowland
and upland communities with ruderal (disturbed ground) plants, many of them introduced weeds.
Regional insect communities were simultaneously affected by changes in land use. The presettlement
aquatic beetle fauna was dominated by species of dryopoid beetles that today inhabit only streams of
high water quality. Terrestrial beetle taxa included species of undisturbed grasslands and riparian forest.
The changes in the landscape resulted in a decrease in the diversity of terrestrial beetle taxa and caused
the near total elimination of dryopoid beetles in stream waters. Dominating the historic assemblages are
beetles associated with dung, polluted waters and cultivated plants, including host-specific immigrant
beetle species that are associated with immigrant plant species.
Key words: Euro-American settlement, human impact, riparian landscape, cultural landscape, historical
evidence, floodplain sediments, pollen, macrofossils, insects, Iowa, midwest.
315
Introduction
What impact have land settlement and modern agriculture
had on natural landscapes? This question is difficult to
answer, because (1) these changes in land use occurred over a
century ago in the US, and (2) those remaining natural
landscapes in the midwest cannot be tested because they are
already affected by such anthropogenic factors as increased
runoff and pesticide drift from surrounding areas. However,
some of the impacts of agriculture on natural landscapes can
be evaluated by comparing presettlement and postsettlement
fossil and lithologic records that include the last few hundred
years. These records reflect changes in land-use practices and
have been used to document some of the resulting impacts on
the environment. The purpose of this paper is to illustrate
how Euro-American settlement changed the character of the
landscape, stream behaviour, vegetation, beetle fauna and
water quality in a small northeastern Iowa drainage basin.
The record preserved in the Roberts Creek Basin in
northeastern Iowa (Figure 1) shows how significant and interrelated
the effects of historic land-use changes were on
vegetation, slope processes, hydrology, sedimentation and
water quality. These changes affected all parts of the landscape
to varying degrees, but nowhere was the impact greater
than on the floodplain. In order to evaluate the nature of
historic floodplain changes, we examined historical documents,
Holocene stratigraphy beneath the valley floor, average
floodplain aggradation rates, and aspects of the
palynology, plant macrofossils, beetles and chemistry of presettlement
and historical deposits. We used sites ranging in
age from about 3000 BP to the present. The historical deposits
are recognized by: (1) sediment stratigraphy, (2) the
abundance of plant macrofossils and beetles of introduced
species, (3) radiocarbon dates, and (4) the presence of recent
artifacts such as barbed wire, glass, bricks and other debris.
The historical data reveal the nature and timing of floodplain
changes, and the record from about 3000 BP to settlement
Figure 1 Location map. Area outlined in Clayton County is the
Roberts Creek drainage basin. Uncalibrated radiocarbon ages of
sediments at numbered sites are as follows: 1 =
modern; 2 =
150 BP;
3 =
1820 BP; 4 =
2850 BP; 5 =
380 BP; 6 =
70 BP; 7a =
modern;
7b =
2350 BP; 8 =
1220 BP; 9 =
2900 BP; 10 =
modern (see
Table 1).
time provides baseline information for assessing the changing
edaphic and ecologic conditions during this period.
The environmental impact of aboriginal agrarian activities
during the Neolithic was relatively slight, but it is recognizable
in pollen diagrams from Europe, the Middle East and
Japan (Tsukada, 1972; Birks et al., 1988). Even in North
America, similar small changes from native American agriculture
are noticeable (Delcourt et al., 1986; McAndrews,
1988). These effects are generally: (1) slight decreases in
pollen of forest taxa, suggesting that forests were cleared for
agriculture, (2) slight increases in pollen percentages of crop
species that were planted, and (3) increases in pollen of
disturbed ground plants, which became established on the
newly cleared areas and abandoned fields and gardens. The
changes in landscape and vegetation were apparently slight
because only small areas were affected and the changes
occurred slowly over a relatively long time.
In contrast, European settlement in North America caused
rapid and consequential changes in the environment over a
large region. A rise in Ambrosia (ragweed) pollen percentages
of 2-10 times presettlement values is the most widely
known marker for the settlement horizon in the eastern and
central United States (Webb et al., 1983), and increases in
sedimentation also occurred following cultivation (Knox,
1987). However, these studies only hint at the total changes in
the ecosystem and landscape as a consequence of settlement.
Such changes, as documented in the biotic, sedimentary,
pedological and historical record, have rarely, if ever, been
studied simultaneously in a drainage basin.
Many previous studies in North America that documented
impacts of changes in land use have concentrated mainly on
lake sediments (Davis, 1973; 1989; Birks et al., 1976; Maher,
1977; Bradbury and Waddington, 1973; Brugham, 1978). For
example, Birks et al. (1976) compared the fossil and chemical
content of pre- and posthistoric lake sediment in lakes in
northwestern Minnesota. At Lake Shagawa, northern Minnesota,
Bradbury and Waddington (1973) related changes in
lake chemistry and biota to human activities, including mining.
Davis (1973), Maher (1977) and Brugham (1978) showed
increased sedimentation rates in lakes ranging in size from
small ponds to Lake Superior, caused by various human
activities. Some recent studies have shown the relationship
between human activities and stream behaviour. Knox (1987)
related changes in sedimentation in stream valleys in southwestern
Wisconsin to mining and other human activities, and
James (1991) showed changes in stream channel morphology
that were related to hydraulic mining in California.
The impact of Euro-American settlement on the insect
fauna of North America was as profound as that caused by
any event in the late Quaternary, including glaciation. Populations
of insect species representative of such habitats as
forest, native prairie and marshlands were replaced by ruderal
species as extensive deforestation, cultivation and drainage
of wetlands accompanied the westward expansion of
Euro-American settlement (McLeod, 1980; Turnbull, 1980).
Associated with settlement was the accidental or intentional
introduction of exotic insect species (Lindroth, 1957; Sailer,
1983). For example, of the 1683 immigrant arthropod species
reported by Sailer (1983) in the continental United States,
66% originated from the western palaearctic region.
The study of beetle remains from archaeological sites has
become a well established discipline in Europe (e.g., Buckland
and Kenward, 1973; Osborne, 1974; 1988; Kenward,
1976). In North America, however, such studies are only now
being initiated (e.g., Elias, 1986; 1990), and none until now
has examined anthropogenic changes of an insect fauna. The
discovery of a rich assemblage in northeastern Iowa permits a
316
palaeontological examination of the impact of human activities
on the species composition of the regional beetle fauna.
Roberts Creek is well suited for studying the effects of
agricultural land-use change. Modern vegetation near the
study area is mostly pasture on the floodplain and row crops
on the uplands, with a few scattered groves of oak. The low
permeability of the Ordovician shale that underlies the
floodplain has kept the water table at a relatively stable
position within the Holocene alluvial fill for the last 12500
years (Bettis, 1984). The creek has meandered across the
valley floor throughout the Holocene with little downcutting
and only modest aggradation, and numerous beaver-dam
ponds, small abandoned channel ponds, and point bars were
formed, rapidly buried with sediment, and later exposed by
lateral stream migration. These exposures in the creek banks
and cores collected from the valley floor are the sources of
fossil data. Fossil preservation is excellent, because deposition
was rapid and the water table was continuously high. The
area sampled is near the headwaters of the basin, guaranteeing
that the fossils are of local origin. Over 40 sites spanning
the entire Holocene were sampled along the creek. Sixty-one
radiocarbon dates, 53 of which were on wood, provide a
reliable chronology and indicate that each deposit represents
only about 100 to 200 years of accumulation (Chumbley,
1989; Chumbley et al., 1990). Of the 40 sections studied, ten
date from the last 3000 years and are the basis of this study.
Historical information
Conditions at the time of settlement can be approximated
from notes recorded during the General Land Office Survey
of Wisconsin Territory (1837-1849). At that time, large
prairie tracts occupied most upland and gently sloping areas,
while timber was restricted to a narrow belt along the Roberts
Creek channel. Human impacts were minimal; a military road
traversed the northeastern part of the basin, three small fields
were present in the basin’s northeastern uplands, and there
was a single field on the divide near the junction of Deer and
Roberts creeks, within the reach of the valley where our study
sites are located. The 1838 census of Wisconsin Territory
supports this image of few inhabitants; a population of 274 is
recorded for Clayton County, most of whom lived by the
Mississippi River. Soon thereafter Euro-American settlement
increased markedly. The population of Clayton County was
500 in 1840 and 1000 in 1850. By 1856 more than 1000 people,
mostly farmers, lived in Roberts Creek basin alone (Iowa
State Census, 1856), and by 1880 a large percentage of upland
and broad valley tracts in the basin was in wheat, maize and
hay (Benton and Gray, 1925).
Knox (1987) used historical records along with a variety of
sedimentological tracers derived from mining operations to
evaluate the timing of floodplain aggradation during the
historical period in southwest Wisconsin. He concluded that
tributary valleys aggraded primarily between about 1890 and
1930. After that time the meander belt expanded, and sedimentation
decreased sharply outside the meander belt.
Although we are not able to date precisely the period of
historical sedimentation because these age tracers were not
available in the Roberts Creek drainage basin, a similar
chronology is likely. One line of evidence that supports the
idea that historical sedimentation was nearly completed by
the 1920s in the Roberts Creek study area is that soils mapped
in this part of the valley in the 1920s were noted to be ’well
above overflow’ (Benton and Gray, 1925).
Methods
Deposits exposed in stream banks along Roberts Creek were
measured and described using standard procedures and nomenclature
(Soil Survey Staff, 1975). Deposits were also
examined in 7.6-cm-diameter cores taken with a drill rig.
Coring transects and bank exposures were used to determine
the thickness of historical alluvium across the valley floor.
Samples for radiocarbon dating were collected from stream
banks and cores. All samples were analysed by Beta-Analytic,
Incorporated, and the ages reported are uncorrected except
as noted (Table 1). Samples for pollen, macrofossils and
insects were dug by hand from cut banks along Roberts
Table 1 Radiocarbon ages of samples reported in this study. Site numbers refer to locations on Figure 1
Note:
*
p =
pollen, m =
macrofossils, b =
beetles.
317
Figure 2 Pollen diagram constructed from six sites showing ten selected pollen taxa.
Creek. Approximately 300 cm3 of matrix were collected for
pollen and macrofossil analysis and about 20 kg for insect
analysis at each site. Pollen samples were processed following
Flegri et al. (1989) using KOH, HCI, acetolysis and HF. In
addition, the sediment was sieved through a 7 Rm screen
(Cwynar et al., 1979), treated with 10% sodium hypochlorite
solution and floated using ZnCI (Chumbley, 1989). Pollen was
mounted in silicone fluid, and at least 300 pollen grains were
counted in each sample. Plant macrofossils were sieved using
0.5- and 0.105-mm screens, and wet residues were hand
picked. Samples were stored in glycerin with a pinch of
phenol as a preservative. Pollen and macrofossils were identified
using reference collections at the University of Iowa
Geology Department and are preserved in the repository
there. Plant nomenclature follows Gleason and Cronquist
(1963).
Not every sample was analysed for all biotic elements. Ten
sites have macrofossil analyses, whereas pollen and insects
were analysed from six sites. The uppermost sample on the
pollen diagram (Figure 2) is from historical sediments (site 6
on Figure 1) of about 1880 BP; the uppermost macrofossil
samples are from modern abandoned channel sediments.
Bulk samples of the sediment were processed for fossil
beetles according to procedures described in Ashworth
(1979). The beetle remains (mainly heads, pronota and
elytra) are disarticulated but otherwise well preserved. They
are reposited on microslides and in vials at the Quaternary
Entomology Laboratory, North Dakota State University,
Fargo, North Dakota. Grain size composition of the deposits
was determined using the pipette method (Walter et al.,
1978). Organic carbon, total nitrogen and total phosphorous
contents were determined at the Iowa State University Soil
Survey Laboratory using standard methods (Dick and Tabatabai,
1977; Yeomans and Bremner, 1991).
Results and discussion
Characteristics of historical alluvium
Historical alluvium in Roberts Creek basin is similar to that
found in other parts of the upper Midwest (Magilligan, 1985;
Bettis and Littke, 1987; W.C. Johnson, 1987; Knox, 1987;
Bettis, 1990). It consists primarily of very dark greyish brown
(10YR3/2) to dark brown (10YR4/3) planar bedded silt loam
and loam resting abruptly on the very dark grey (10YR3/1) to
black (10YR2/1), a horizon of the presettlement surface soil
(usually a mollisol). Bedding is usually absent in the upper
few decimetres of the postsettlement deposits because sedimentation
was slow or intermittent, allowing sufficient time
for bioturbation and other pedogenic processes to destroy
former bedding structures.
The thickness of historical alluvium varies across the valley
floor. Within the modern channel belt, the thickness of the
fine grained increment of the historical alluvium is extremely
variable, ranging from < 10 cm to 2 m. Outside the channel
belt its thickness averages 1 m where it buries Holocene
alluvium younger than about 3500 BP and 44 cm where it
buries older Holocene alluvium. The surface of these buried
deposits ranges from 0.5-1.5 and 1.7-2.4 m above the adjacent
channel bed, respectively. Low terraces underlain by alluvium
older than about 8000 BP are usually elevated more than
2.5 m above the adjacent channel bed and have little or nor
cap of historical alluvium. Knox (1987) found similar relationships
among thickness of historical alluvium, age of the
underlying presettlement deposits, and relative elevation of
the presettlement surface in the Driftless Area of Wisconsin.
Historical alluvium is also found at the base of valley wall
slopes where it varies from a few cm to 50 cm in thickness. In
a few places 1-2 m of historical alluvium were recorded
beneath small alluvial fans emanating from steep, healed
gullies on valley slopes.
Calculation of average sedimentation rates (from observations
at more than 50 sites) using an age of 150 years for the
base of the historical alluvium, and radiocarbon ages for late
Holocene and early to middle Holocene-age alluvium indicates
that aggradation increased from presettlement rates
of 0.028-0.13 cm year to an average of 0.7 cm year in the
historic period. The presettlement values are comparable to
those obtained by Knox (1987) in streams of similar size in
Wisconsin’s Driftless Area, but the historic values from
Roberts Creek are higher than those from the Driftless Area.
If most of the historical alluvium accumulated in the 40 years
between AD 1890 and 1930, as suggested by historical records,
then the historic sedimentation rate is even higher,
about 2.6 cm year-1.
The organic carbon content of historical alluvium is slightly
lower than that of the upper, humus-rich A horizon of the
318
buried presettlement soil, but significantly higher than that of
the presettlement alluvium beneath the A horizon. These
differences probably reflect the origin of a significant part of
the historical alluvium from erosion of the upper, organic-rich
part of soil profiles in the watershed. The contrast between
the organic carbon content of historical and presettlement
alluvium is not as great for deposits dating from 3500 BP or
later (the presettlement floodplain) as it is for older alluvial
deposits (presettlement low terraces). Total nitrogen relationships
parallel those of organic carbon. Total phosphorus
content is much lower in historical alluvium than in late
Holocene alluvium, but similar to that of early and middle
Holocene alluvium and the upper horizons of loess-derived
upland and valley slope soils. The chemical properties of
historical alluvium are consistent with an interpretation that
the unit was derived primarily from erosion of the upper
solum (A, E and upper B horizons) of prairie/forest transition
(Mollic Hapludalfs) and forest (Typic Hapludalfs) soils in the
Roberts Creek drainage basin.
Three phases of change in historical valley floor hydrology
and sedimentation are evident. The first occurred during the
early settlement period and involved land clearing and initial
cultivation. The amount of land in cultivation was small at the
beginning of this period, but increased dramatically in the
1850s. Initially the deep, organic-rich prairie and prairieforest
transitional soils of the upper part of Roberts Creek
basin continued to provide relatively good infiltration, so that
runoff remained relatively low. Experimental studies indicate
that the original vegetation and porous, humus-rich presettlement
top soil were extremely effective in retarding surface
runoff and sediment yield (Sartz, 1975; 1976). However, after
several years of cultivation with poor conservation practices,
the infiltration capacity of the soils in the basin decreased, and
the increased runoff began to reach the channel network. In
the last decade or two of the nineteenth century the second
phase began, as flood peaks and sedimentation increased
significantly because poor conservation practices intensified
soil erosion and augmented runoff. Frequent flooding and
relatively rapid floodplain aggradation continued until sometime
in the 1920s or 1930s, then began to decrease as the
channel adjusted to the changed hydrologic conditions by
widening. At that time the meander belt expanded, less of the
valley floor was flooded, and floodplain sedimentation decreased.
During the final stage, runoff and sediment delivery
to the valley decreased as a result of the adoption of better
soil and water conservation practices. This final phase of
valley floor changes has been characterized by decreased
valley floor sedimentation, lower peak flood height, and
maintenance of a relatively wide channel.
Plants
Pollen analysis of the cut bank and core sediments shows that
the presettlement vegetation pattern was established in the
area about 3000 BP (Figure 2; see also Chumbley et at., 1990).
Land survey records from the 1830s and 1840s (Chumbley,
1988; unpublished data) indicate that oak savanna occupied
much of the Roberts Creek basin. Prior to settlement the
major pollen types present were Quercus (15-20% ), Salix
(2-20% ), Poaceae (20-35% ), Asteraceae excluding Ambrosia
(10-15%) and Cyperaceae (20-25%). Pollen of aquatic
plants, especially Myriophyllum, was also relatively common.
This pollen rain supports the contention that the upland area
was an oak savanna and the floodplain supported willows
along with other wetland and aquatic plants.
Plant macrofossil analyses of these deposits (Figures 3 and
4) support the idea that a rich mix of aquatic, wetland, wet
prairie and riparian elements was present on the floodplain.
Figure 3 Summary of changes in plant macrofossils arranged in four
ecological groups: (1) aquatic, (2) prairie, (3) riparian and (4) ruderal
plants. 3A =
plotted by number of taxa; 3B =
plotted by number of
specimens per litre. Prehistoric bar is mean of seven sites (2, 3, 4, 5,
7a, 8 and 9 on Figure 1); historic bar represents one site (6 on
Figure 1), and modern bar is mean of two sites (1 and 7 on
Figure 1).
The major tree during presettlement time is Salix, with single
occurrences of Ulmus americana, Tilia americana and Quercus.
Forest herbs are restricted to Ranunculus septentrionalis,
which can grow in riparian forest environments. Shrubs are
represented by Cornus amomum subspecies obliqua, Cornus
sp. and Rubus, all floodplain or edge species. Disturbed
ground plants are sparingly present, but most cannot be
identified to species. These include Amaranthus, Chenopodium
and Oxalis. A variety of prairie indicator species occurs
in presettlement sites; among the more regularly occurring
are Amorpha canescens, Hypoxis hirsuta, Monarda fistulosa,
Pycnanthemum virginianum, Ratibida pinnata, Rudbeckia
hirta and Zizia aurea. Nine species of aquatic plants also grew
in the stream or in floodplain ponds, but their occurrence is
sporadic.
Macrofossils of the historical period (Figure 3) were found
in site 6 (Figure 1). This site is dated by a piece of barbed
wire, which was of the type called ’Glidden’s Barb, Onestrand
Variation’ (written communication from A.L. Fisk to
D.P. Schwert, 12 August 1989). The patent for this make was
issued in 1874, and the site is estimated to date from about
1880.
319
Figure 4 Plant macrofossil diagram of ruderal taxa at Roberts Creek during the last
3000 BP, showing the increase in abundance of most taxa following settlement.
Pollen percentages of Ambrosia and Poaceae rise sharply
in the historical sample to about 25% and 40% respectively,
whereas those of Quercus, total arboreal pollen, Asteraceae
and Cyperaceae decline to 10%, 20%, <5% and <5%
respectively. The rise in Ambrosia and the decline in Quercus
and other trees are widespread in the eastern and midwestern
United States and represent land clearance, but effects on
other taxa were not known.
The diversity and abundance of plant macrofossils of
prairie, aquatic and wetland taxa decreased after settlement
occurred (Figures 3 and 4). The number both of taxa and of
specimens of aquatic plants decreased from prehistoric (mean
of seven samples dated between 3000 BP ant - 150 BP) to
historic time (one sample from -AD 1880) and apparently
recovered slightly in modern time (two samples). Only Potamogeton
has been seen in the modern stream, though Chara
oospores are present in the modern samples. Prairie and
riparian taxa and specimens decrease markedly between
prehistoric and historic time, and the decrease continues to
the modern samples. Only rare Salix trees are present along
the modern stream, and prairie elements are completely
absent from cultivated areas, and rare to absent in pastures on
the floodplain. Ruderal annual plants are present in low
numbers and diversity prior to settlement, but their diversity
triples by -AD 1880, and the number of specimens increases
by about two orders of magnitude. The diversity remains high
in the modern samples, but the number of specimens drops to
about twice presettlement values.
These ’weeds’ included the introduced species, Pastinaca
sativa, Polygonum aviculare, Polygonum hydropiper, Polygonum
convolvulus, Setaria glauca, Setaria viridis and Taraxacum
officinale. Native ruderal genera present in small
numbers earlier in the Holocene are Chenopodium, Echinochloa
and Amaranthus sp. Euphorbia maculata appeared only
in historical sediment, and Polygonum pensylvanicum occurred
sparsely in mid-Holocene time, and increased following
cultivation.
The plant record indicates that major changes in vegetation
occurred following forest clearance and cultivation. The
details of the timing of these changes have yet to be established,
but by -AD 1880 most of the changes had occurred.
In the uplands, oaks in the savanna apparently had been
reduced substantially, and prairies had largely been ploughed
under. On the floodplain, native wetland, riparian and prairie
communities had been almost completely replaced by crops.
Ruderal elements were dominant in the fossil assemblage. By
modern time, the prairie and riparian communities had
further declined, and aquatic communities in the stream had
showed little recovery. Ruderal annuals were considerably
less abundant (though diversity remained high).
The correlation of a maximum in ruderal plants with the
time of maximum flooding is probably not coincidental. The
most abundant niche along the stream for disturbed ground
plants would be the bare soil exposed by both erosion and
deposition during floods. Ruderal species were at their maximum
during this time. The decrease in numbers of ruderal
species in modern samples may be explained by: (1) the
greater stability and lesser flooding that accompanied enlargement
of the meander belt, decreasing the area available
for colonization by these plants, and; (2) the use of herbicides
in the area in recent decades.
Insects
Twenty one families of beetles are represented in the pre- and
post-settlement assemblages (Schwert, in prep.). In addition
to beetles, the remains of molluscs, ostracods, spiders, orbatid
mites, and other orders of insects are present.
Pre-settlement species
Samples from the presettlement sites contain remarkably
diverse assemblages of ground beetles (Carabidae). Most of
the species represented are associated with water-marginal or
marshland environments. Several, however, including Cyclotrachelus
s. sodalis LeC., C. seximpressus LeC., Poecilus
lucublandus (Say), and Amara angustata (Say), are inhabitants
of open grasslands (Lindroth, 1966; 1968; Freitag,
1969). Others, including Scaphinotus elevatus (Fab.) and
Pterostichus stygicus (Say), are woodland species (Lindroth,
320
1961; 1966). Additional upland taxa include Hyperaspis brunnescens
Dobz., an apparently rare ladybird (Coccinellidae)
species known today only from Iowa and Illinois (Gordon,
1985), and numerous remains of the scarabaeid provisionally
identified as Macrodactylus subspinosus (Fab.), the common
’rose chafer’.
Dryopoid beetles, particularly elmids, are the numerically
largest faunal component of the presettlement assemblages,
representing approximately two-thirds of all beetle remains.
Five species of elmids (Dubiraphia cf. vittata [Melsh.], Macronychus
glabratus Say, Microcylloepus pusillus [LeC.], Optioservus
fastiditus [LeC.] and Stenelmis cf. crenata [Say]) and
one species of dryopid (Helichus striatus LeC.) are represented.
Dryopoid faunas of this diversity and richness exist
today only in association with the bedrock- or gravel-bedded
riffle zones of streams of exceptionally high water quality
(Brown, 1976; Hilsenoff, 1977).
Numerous water-marginal and marsh-associated taxa are
present in the remaining component of the presettlement
beetle assemblages. Saprophagous species, however, are
poorly represented; the occurrences of the dung-associated
scarabaeid Onthophagus hecate (Panz.) might have been in
association with occasional visits to the stream by large
herbivores such as bison.
Together, the insect assemblages show that a mixture of
open grasslands and riparian woodlands characterized the
drainage basin of presettlement Roberts Creek. The stream
itself was of trout-stream water quality, with gravel substrate
and with alternating riffles and pools bordered by muddy
banks and zones of marsh-associated plants.
Postsettlement species
Of the 12 species of carabids collectively represented in the
postsettlement samples, all but one (Cyclotrachelus sodalis )
are associated with marsh or water-marginal habitats. No
woodland-associated carabids are present, although two generalist
taxa of bark beetles (Scolytidae) occur.
Upland elements of the postsettlement assemblages are
dominated instead by scarabaeids (Scarabaeidae) and weevils
(Curculionidae). Of the six species of the scarabaeid dung
beetle Aphodius definitively or provisionally identified, all
are today regularly associated with cattle dung. Two, A
distinctus (Miiller) and A. granarius (L.), represent introductions
from Europe (Table 2), the latter apparently during the
seventeenth century aboard ships carrying cattle from England
(Sailer, 1983; Ratcliffe, 1991).
At least six species of the large postsettlement weevil
assemblage represent immigrant species that themselves are
associated with immigrant host plants (Table 2). Nearctic
weevil species (and their associated host plants) include
Hypera cf. compta Say (Polygonum coccineum), Nedyus
flavicaudis (Boh.) (Urtica sp.), Listronotus cf. appendiculatus
(Boh.) (Sagittaria spp.), and the billbug weevils Sphenophorus
parvulus Gyll. and S. minimus Hart. (serious pests of turfforming
grasses) (Puttler et al., 1973; Vaurie, 1983; JohnsonCicalese
et al., 1990; R.S. Anderson, pers. comm., 1990).
Other anthrophilic taxa represented in the postsettlement
assemblages include the palaearctic chrysomelid Phyllotreta
striolata (Fab.) (a widespread pest, particularly of cruciferous
vegetables) and the nearctic nitidulid Glischrochilus quadrisignatus
Say (a pest of fruit and maize) (Smith, 1985;
Bousquet, 1990). Fossils of the cryptophagid Atomaria ephippiata
Zimm. and a cucujid provisionally identified as Cryptolestes
ferrugineus (Steph.) are present; these species
commonly occur today in association with stored grains, the
latter being a serious pest, particularly of wheat (Bousquet,
1990).
Whereas the remains of dryopoid water beetles over~
whelmingly dominated the presettlement assemblages, they
are nearly absent from postsettlement samples (Figure 5).
Replacing elmids as the most abundant water beetle taxon is
the haliplid Peltodytes edentulus LeC., one of the few species
of aquatic beetles that appears to thrive in shallow, eutrophic
and polluted pools and streams.
Interpretation of the insect evidence
Records compiled from museum and literature sources indicate
that all of the beetle species identified from pre- and
postsettlement samples could potentially occur within the
region of Roberts Creek today. The postsettlement transformation
of the landscape by cultivation appears to have
profoundly affected the species diversity of the carabid and
scarabaeid faunas (Figure 6). In this analysis, the pre- and
postsettlement representations of these families are standardized
to species per kilogram of sample sediment. Carabid
diversity markedly decreases, with upland species in particular
being no longer well represented in postsettlement
Table 2 Exotic species of Coleoptera represented as fossils in samples of postsettlement age from the Roberts Creek basin, with notes on their
habitats, host plants, or food sources
Note: *Host plants of palearctic origin (Gleason and Cronquist, 1963).
321
Figure 5 Changes in average numbers of carabid and dung-associated
scarabaeid species represented per kilogram of sediment in pre- and
postsettlement deposits of Roberts Creek.
samples; the detrimental impact on carabid populations by
agricultural manipulation of modern landscapes has been well
documented (e.g., Luff and Rushton, 1989; Weiss et at., 1990).
The diversity of dung-associated scarabaeids, however,
sharply increases upon settlement, as cattle and other large
herbivores are introduced to graze proximal to Roberts
Creek.
Sedimentation and eutrophication associated with agricultural
transformation of the landscape likewise made the
waters of Roberts Creek inhospitable to most species of
aquatic Coleoptera. Although we commonly encountered live
elmids in abundance clinging to stones in riffle areas of the
nearby Turkey River, dryopoid beetles today rarely occur in
the waters of Roberts Creek (Kennedy and Miller, 1990). A
thick layer of sediment of postsettlement age now blankets
what was once the gravel or bedrock bottom of Roberts
Creek, effectively eliminating drypoid habitat. As might be
predicted, water-marginal beetles, inhabiting zones that are
directly affected by neither cultivation nor eutrophication,
represent the group least affected by Euro-American
settlement.
The faunal record of changes marking the transformation
of the Roberts Creek basin to an agricultural landscape bear
remarkable similarities to fossil insect records of the English
Midlands as observed by Osborne (1974; 1988). Forests that
had once characterized the region proximal to the River
Avon in Warwickshire were by late Bronze Age (ca. 2900 BP)
largely replaced by grasslands occupied by large grazing
herbivores such as cattle and sheep. As forest clearing and
ploughing of the landscape progressed, soil erosion led to
increased sedimentation and to the elimination of the rich
elmid populations of the River Avon.
Conclusions
The spread of agriculture in northeastern Iowa had several
impacts on the local ecosystem and landscape in the time
following settlement.
(1) During the land clearance phase, stable vegetation
types were replaced by disturbed ground plants, and soils
were degraded by cultivation and poor conservation pracFigure
6 Changes in total numbers of elmid elytra represented per
kilogram of sediment in pre- and postsettlement deposits of Roberts
Creek. Age of deposits is given in italics.
tices. This disturbance apparently was at its peak between
-AD 1880 and 1930.
(2) The loss of stable vegetation cover and infiltration
capacity caused by soil erosion resulted in increased runoff,
thereby increasing flood frequency and magnitude for several
decades after land clearance.
(3) These changes greatly accelerated hill slope erosion,
removing the A horizons of upland soils and depleting their
tilth and fertility.
(4) Increased sediment delivery to the valley network
resulted in the accumulation of up to 2 m of sediment derived
from uplands and slopes in the channel belt and between 0.4
and 1.0 m on the floodplain between -AD 1890 and 1920.
This new alluvium was fertile, but the rapid sedimentation
rate (approximately 5-25 times the presettlement rate) across
much of the valley floor resulted in edaphic conditions that
favoured colonization by disturbed ground plants.
(5) Since -AD 1920, channel widening has decreased
flooding and floodplain aggradation rates.
(6) Diversity of native plants has decreased, and native
plant communities have been lost to aggressive introduced
species that are temporary colonizers of disturbed ground.
(7) Water quality has been severely degraded from a clear,
cold-water trout stream to a polluted, silt-laden creek.
(8) Changes in beetle populations show that upland carabids
declined and taxa associated with crop plants and cattle
dung increased.
Acknowledgements
This research was supported by NSF Grants ATM-88-06482
to R.G. Baker and ATM-88-0571 to D.P. Schwert. For their
assistance in identification of beetle remains, we are grateful
to the staff of the Coleoptera Division of the Biosystematics
Research Centre, Agriculture Canada, to H.P. Brown of the
University of Oklahoma (Dryopidae and Elmidae), to R.S.
Anderson of the National Museum of Canada (Curculionidae)
and to R.D. Gordon of the Systematic Entomology
Laboratory, USDA (Scarabaeidae). For their help in both
field and laboratory analyses, we thank Joseph Krieg, Kevin
Krogstad, Sally Domke and Cynthia Jensen.
References
Ashworth, A.C. 1979: A method of recovering fossil insect remains
from clays, peats, and silts. In Erwin, T.L., Ball, G.E. and Whitehead,
D.R. editors, Carabid beetles, their evolution, natural history, and
classification, The Hague: W. Junk, 406.
322
Benton, T.H. and Gray, A.L. 1925: Soil survey of Clayton County,
Iowa. Washington, D.C.: U.S. Dept. of Agriculture, Bureau of
Chemistry and Soils.
Bettis III, E.A. 1984: Preliminary investigations of the stratigraphy
and chronology of northeastern Iowa alluvium and its archaeological
significance with special reference to the Turkey River Basin. Report
submitted to the Iowa State Historical Department, Division of
Historic Preservation, Des Moines.
&mdash;
1990: Holocene Alluvial Stratigraphy and Selected Aspects of
the Quaternary History of Western Iowa. Guidebook for the 37th
Midwest Friends of the Pleistocene Field Conference.
Bettis III, E.A. and Littke, J.P. 1987: Holocene alluvial stratigraphy
and landscape development in Soap Creek Watershed Appanoose,
Davis, Monroe, and Wapello counties, Iowa. Iowa Department of
Natural Resources, Geological Survey Bureau. Open File Report 87-2.
Iowa City.
Birks, H.H., Birks, H.J.B., Kaland, P.E. and Moe, D. 1988: The
cultural landscape - past, present and future. Cambridge: Cambridge
University Press.
Birks, H.H., Whiteside, M.C., Stark, D.M, and Bright, R.C. 1976:
Recent paleolimnology of three lakes in northwestern Minnesota.
Quaternary Research 6, 249-72.
Blatchley, W.S. and Leng, C.W. 1916: Rhynchophora or weevils of
north eastern America. Indianapolis: Nature Publications.
Bousquet ,Y. 1990: Beetles associated with stored products in Canada:
an identification guide. Ottawa, Canada: Agriculture Canada Research
Branch Publication 1837.
Bradbury, J.P. and Waddington, J.C.B. 1973: The impact of European
settlement on Shagawa Lake, northeastern Minnesota, U.S.A.
In Birks, H.J.B. and West, R.G., editors, Quaternary Plant Ecology,
Oxford: Blackwell, 289-317.
Brown, H.P. 1976: Aquatic dryopoid beetles (Coleoptera) of the
United States. Cincinnati, Ohio, USA: US Environmental Protection
Agency Water Pollution Control Research Series 18050-
ELD04/72.
Brugham, R.B. 1978: Human disturbance and the historical development
of Linsley Pond. Ecology 59, 19-36.
Buckland, P.C. and Kenward, H.K. 1973: Thorne Moor: a palaeoecological
study of a Bronze Age site. Nature 241, 405-406.
Chumbley, C.A. 1988: Late glacial and Holocene vegetation. In
Green, W., editor, Archaeological and paleoenvironmental studies in
the Turkey River Valley, northeastern Iowa, Report to the Iowa State
Preserves Board, 74-109.
&mdash;
1989: Late-glacial and Holocene vegetation of the Roberts
Creek Basin, northeast Iowa. University of Iowa Ph.D. Dissertation,
154 pp.
Chumbley, C.A., Baker, R.G. and Bettis III, E.A. 1990: Midwestern
Holocene paleoenvironments revealed by floodplain deposits in
northeastern Iowa. Science 249, 272-74.
Cwynar, L.C., Burden, E. and McAndrews, J.H. 1979: An inexpensive
method for concentrating pollen and spores from fine grained
sediments. Canadian Journal ofEarth Sciences 16, 1115-20.
Davis, M.B. 1973: Pollen evidence of changing land use around the
shores of Lake Washington. Northwest Science 47, 133-48.
&mdash;
1989: Retrospective studies. In Likens, G.E., editor, Long-term
studies in ecology, approaches and alternatives, New York: SpringerVerlag,
71-89.
Delcourt, P.A., Delcourt, H.R., Cridlebaugh, P.A. and Chapman, J.
1986: Holocene ethnobotanical and paleoecological record of human
impact on vegetation in the Little Tennessee River Valley, Tennessee.
Quaternary Research 25, 330-49.
Dick, W. and Tabatabai, M.A. 1977: An alkaline oxidation method
for the determination of total phosphorous in soils. Soil Science
Society ofAmerica Proceedings 41, 511-14.
Elias, S.A. 1986: Fossil insect evidence for late Pleistocene paleoenvironments
of the Lamb Spring site, Colorado. Geoarchaeology 1,
381-87.
&mdash;
1990: The timing and intensity of environmental changes during
the Paleoindian period in western North America: evidence from the
insect fossil record: In Agenbroad, L., Mead, J. and Nelson, L.,
editors, Megafauna and man, Flagstaff: Northern Arizona University
Press, 11-14.
Faegri, K., Kaland, P.E. and Krzywinski, K. 1989: Textbook ofPollen
Analysis. New York: John Wiley and Sons.
Freitag, R. 1969: A revision of the species of the genus Evarthrus
LeConte (Coleoptera: Carabidae). Quaestiones Entomologicae 5,
89-212.
General Land Office Survey of Wisconsin Territory 1837-1849.
Gleason, H.A. and Cronquist, A. 1963: Manual of vascular plants of
northeastern United States and adjacent Canada. New York: D. Van
Nostrand.
Godfrey, L.D. and Yeargan, K.V. 1989: Effects of clover root
curculio, alfalfa weevil (Coleoptera: Curculionidae), and soil-borne
fungi on alfalfa stand density and longevity in Kentucky. Journal of
Economic Entomology 82, 1749-56.
Gordon, R.D. 1985: The Coccinellidae (Coleoptera) of America
north of Mexico. Journal ofthe New York Entomological Society 93,
1-912.
Harde, K.W.1981: A fieldguide in colour to beetles. London: Octopus
Books.
Hisenoff, W.L. 1977: Use of arthropods to evaluate water quality of
streams. Wisconsin Department of Natural Resources (Madison)
Technical Bulletin No. 100, 15 pp.
Hoffman, A. 1954: Faune de France. 59. Col&eacute;opt&egrave;res curculiondes
(Deuxieme Partie). Paris: Federation Francaise des Soci&eacute;t&eacute;s de
Sciences Naturelles.
Iowa State Census 1856.
James, L.A. 1991: Incision and morphologic evolution of an alluvial
channel recovering from hydraulic mining sediment. Geological
Society ofAmerica Bulletin 103, 723-36.
Johnson, J.P. 1944: The imported long-horned weevil Calymycterus
setarius Roelofs. Bulletin of the Connecticut Agricultural Experiment
Station 19, 121-42.
Johnson, W.C., editor, 1987: Quaternary environments of Kansas.
Kansas Geological Survey, Guidebook Series 5, Lawrence.
Johnson-Cicalese, J.M., Wolfe, G.W. and Funk, C.R. 1990: Biology,
distribution, and taxonomy of billbug turf pests (Coleoptera: Curculionidae).
Environmental Entomology 19, 1037-46.
Kennedy, J.O. and Miller III, J.G. 1990: A survey of the benthic
macroinvertebrates of the Big Spring Basin, Iowa. Journal of the
Iowa Academy ofSciences 97, 46-54.
Kenward, H.K. 1976: Reconstructing ancient ecological conditions
from insect remains; some problems and an experimental approach.
Ecological Entomology 1, 7-17.
Knox, J.C. 1987: Historical valley floor sedimentation in the Upper
Mississippi Valley. Annals of the Association of American Geographers
77, 224-44.
Lindroth, C.H. 1957: The faunal connections between Europe and
North America. New York: John Wiley and Sons.
&mdash; 1961: The ground-beetles (Carabidae, excl. Cicindelinae) of
Canada and Alaska (2). Opuscula Entomologica Supplementum 20,
1-200.
&mdash; 1966: The ground-beetles (Carabidae, excl. Cicindelinae) of
Canada and Alaska (4). Opuscula Entomologica Supplementum 29,
409-648.
&mdash; 1968: The ground-beetles (Carabidae, excl. Cicindelinae) of
Canada and Alaska (5). Opuscula Entomologica Supplementum 33,
649-944.
Litsinger, J.A. and Apple, J.W. 1984: Summer diapause of the clover
leaf weevil, Hypera punctata and lesser clover leaf weevil,
Hypera nigrirostris, in Wisconsin. Great Lakes Entomologist 17,
83-86.
Luff, M.L. and Rushton, S.P. 1989: The ground beetle and spider
fauna of managed and unimproved upland pasture. Agriculture
Ecosystems and Environments 25, 195-205.
Magilligan, F.J. 1985: Historical floodplain sedimentation in the
Galena River basin, Wisconsin and Illinois. Annals ofthe Association
of American Geographers 75, 583-94.
Maher, L.J., Jr. 1977: Palynological studies in the western arm of
Lake Superior. Quaternary Research 7, 14-44.
McAndrews, J.H. 1988: Human disturbance of North American
forests and grasslands: The fossil pollen record. In Huntley, B. and
Webb III, T., editors, Vegetation history, Dortdrecht: Kluwer,
673-97.
McAvoy, T.J., Kok, L.T, and Trimble, J.T. 1983: Biological studies of
Ceutorhynchus punctiger (Coleoptera: Curculionidae) on dandelion
in Virginia. Annals of the Entomological Society of America 76,
671-74.
323
McLeod, J.M. 1980: Forests, disturbances, and insects. The Canadian
Entomologist 112, 1185-92.
Moore, I. 1979: Platystethus cornutus Gravenhorst reported from the
United States is P. degener Mulsant and Rey (Coleoptera: Staphylinidae).
Coleopterists Bulletin 33, 32.
Moore, I. and Legner, E.F. 1971: A review of the nearctic species of
Platystethus (Coleoptera: Staphylinidae). Pan-Pacific Entomologist
47,260-64.
Osborne, P.J.1974 : An insect assemblage of early Flandrian age from
Lea Marston, Warwickshire, and its bearing on the contemporary
climate and ecology. Quaternary Research 4, 471-86.
&mdash;
1988: A late Bronze Age insect fauna from the River Avon,
Warwickshire, England: its implications for the terrestrial and fluvial
environment and for climate. Journal of Archaeological Science 15,
715-27.
Puttler, B., Thewke, S.E. and Warner, R.E. 1973: Bionomics of three
nearctic species, one new, of Hypera (Coleoptera: Curculionidae),
and their parasites. Annals of the Entomological Society of America
66, 1299-1306.
Ratcliffe, B.C. 1991: The scarab beetles of Nebraska. Bulletin Nebraska
State Museum 12, 1-333.
Sailer, R.I. 1983: History of insect introductions. In Wilson, C.L. and
Graham, C.L., editors, Exotic plant pest and North American agriculture,
New York: Academic Press, 15-38.
Sartz, R.S. 1975: Controlling runoff in the Driftless Area. Journal of
Soil and Water Conservation 30, 92-93.
&mdash;
1976: Sediment yield from steep lands in the Driftless Area. In
Proceedings of the Third Federal Inter-Agency Sedimentation Conference,
Washington, D.C.: U.S. Government Printing Office,
I-123-31.
Smith, E.H. 1985: Revision of the genus Phyllotera Chevrolat of
America north of Mexico. Part I. The maculate species (Coleoptera:
Chrysomelidae, Alticinae). Fieldiana (Zoology, New Series) 28,
1-168.
Soil Survey Staff 1975: Soil Taxonomy. Washington, D.C.: U.S.
Department of Agriculture Handbook 436.
Tsukada, M. 1972: The history of Lake Nojiri, Japan. Connecticut
Academy ofArts and Sciences Transactions 44, 339-65.
Turnbull, A.L. 1980: Man and insects: the influence of human
settlement on the insect fauna of Canada. The Canadian Entomologist
112, 1177-84.
Vaurie, P. 1983: A catalog of the Coleoptera of America north of
Mexico. Family: Curculionidae. Subfamily: Rhynchophorinae. Washington,
D.C.: U.S. Department of Agriculture Handbook 529-143a.
Walter, N.F., Hallberg, G.R. and Fenton, T.E. 1978: Particle-size
analysis by the Iowa State University Soil Survey Laboratory. In
Hallberg, G.R., editor, Standard Procedures for Evaluation of
Quaternary Materials in Iowa, Iowa City: Iowa Geological Survey,
Technical Information Series No. 8, 61-74.
Webb III, T., Cushing, E.J. and Wright, H.E., Jr. 1983: Holocene
changes in vegetation of the Midwest. In Wright, H.E., Jr., editor,
Late Quaternary environments of the United States, Volume 2, The
Holocene, Minneapolis: University of Minnesota Press, 142-65.
Weiss, M.J., Balsbaugh, E.U., Jr., French, E.W. and Hoag, B.K.1990:
Influence of tillage management and cropping system on ground
beetle (Coleoptera: Carabidae) fauna in the northern Great Plains.
Environmental Entomology 19, 1388-91.
Yeomans, J.C. and Bremner, J.C.1991: Carbon and nitrogen analysis
of soils by automated combustion techniques. Communications in
Soil Science and Plant Analysis 22, 843-850.