Cahokia’s emergence and decline coincided with shifts of flood frequency on the Mississippi River
Edited by James A. Brown, Northwestern University, Evanston, IL, and approved April 8, 2015 (received for review January 28, 2015)
Letter
June 16, 2015
Significance
Our paper evaluates the role that flooding played in the emergence and decline of Cahokia—the largest prehistoric settlement in the Americas north of Mexico that emerged in the floodplain of the Mississippi River around A.D. 1050. We use sediment cores to examine the timing of major Mississippi River floods over the last 1,800 y. These data show that Cahokia emerged during a period of reduced megaflood frequency associated with heightened aridity across midcontinental North America, and that its decline and abandonment followed the return of large floods. We conclude that shifts in flood frequency and magnitude facilitated both the formation and the breakdown of Cahokia and may be important factors in the declines of other early agricultural societies.
Abstract
Here we establish the timing of major flood events of the central Mississippi River over the last 1,800 y, using floodwater sediments deposited in two floodplain lakes. Shifts in the frequency of high-magnitude floods are mediated by moisture availability over midcontinental North America and correspond to the emergence and decline of Cahokia—a major late prehistoric settlement in the Mississippi River floodplain. The absence of large floods from A.D. 600 to A.D. 1200 facilitated agricultural intensification, population growth, and settlement expansion across the floodplain that are associated with the emergence of Cahokia as a regional center around A.D. 1050. The return of large floods after A.D. 1200, driven by waning midcontinental aridity, marks the onset of sociopolitical reorganization and depopulation that culminate in the abandonment of Cahokia and the surrounding region by A.D. 1350. Shifts in the frequency and magnitude of flooding may be an underappreciated but critical factor in the formation and dissolution of social complexity in early agricultural societies.
Sign up for PNAS alerts.
Get alerts for new articles, or get an alert when an article is cited.
The episodic breakdowns of early agricultural societies, many of which were situated in the floodplains of major rivers, are often attributed to episodes of severe drought (1–3) that correlate with cascading social and environmental feedbacks (4, 5). The causes of the decline and abandonment of Cahokia, a major late prehistoric population center that emerged in the floodplain of the central Mississippi River (6), remain unclear but have also been attributed to drought (7), as well as resource overexploitation (8), intergroup conflict, and sociopolitical factionalism and upheaval (6, 9, 10). Here, we present evidence that Cahokia emerged during a multicentennial period of enhanced midcontinental aridity that inhibited the occurrence of high-magnitude floods on the central Mississippi River, and that Cahokia’s decline and abandonment correspond to an increase in the frequency of large floods. These findings imply that the disintegration and dissolution of Cahokia may be, in part, societal responses to enhanced hydrological variability in the form of high-magnitude flooding.
Cahokia’s emergence as a regional center can be traced to the population growth and intensified cultivation of native domesticates that began around A.D. 400 in the floodplain of the central Mississippi River near modern-day Saint Louis, MO (6, 11, 12). By A.D. 1050, Cahokia emerged as a hierarchically organized cultural and political center in this region, which was inhabited by tens of thousands of individuals supported in part by the cultivation of native domesticates and maize (6, 13). Settlements affiliated with Cahokia were concentrated on higher-elevation ridges in the floodplain with access to a variety of resources (14), with no evidence that irrigation canals were constructed in this moist and temperate region. Populations in the Cahokia region continued to grow until ca. A.D. 1200, when the region’s population size and cultural prominence began to decline (6, 9, 15), and, by A.D. 1350, Cahokia and the surrounding region were almost completely abandoned (6).
Archaeologists have previously recognized the possibility that large floods could have affected the sociopolitical stability of Cahokia by disrupting food production and storage, damaging houses, and motivating individuals to relocate (6, 9, 16). Before the establishment of modern flood control infrastructure in the early 20th century (17), large floods like the A.D. 1844 event inundated extensive tracts of the central Mississippi River floodplain (Fig. S1), forcing residents to evacuate and causing widespread destruction (6) (historical accounts of flooding are available in SI Text). Past shifts in the locations of house basins and storage pits along elevational gradients in the Mississippi floodplain have been interpreted as indirect evidence for changing hydrological conditions (6, 9, 16), but direct evidence of flooding is rare in archaeological excavations from the Cahokia area (e.g., ref. 18).
Here, we reconstruct the timing of major flood events in the Cahokia area using floodwater sediments deposited in floodplain lakes (19). During flood stages that hydrologically connect the main channel with floodplain lakes, floodplain lakes act as sediment traps that allow the suspended load of floodwaters to fall out of suspension (19, 20). The composition of floodwater sediments usually differs from locally sourced sediment deposited during nonflood conditions, particularly in grain size distribution (19). Across the wide floodplain of a low-gradient river like the Mississippi, overbank floods deposit well-sorted fine silt- and clay-sized sediments in distal floodplain lakes and depressions (21–23). Analysis of sediment records from multiple basins in similar geomorphic settings along the same river ensures that identified flood events are not due to localized erosion or flooding from tributaries (20).
Sediment cores were obtained from two floodplain lakes in the central Mississippi River valley that are abandoned channels of the Mississippi River: Horseshoe Lake (HORM12; Madison County, IL; 38.704767°N, 90.081279°W), situated 5 km northwest of Cahokia, and Grassy Lake (GRAS13; Union County, IL; 37.431701°N, 89.377466°W), located 190 km downstream from Cahokia (Fig. 1). Both lakes are young features of the high-sinuosity meander belt that formed before the main channel stabilized during the late Holocene into a narrow low-sinuosity meander belt (24). Before the 20th century, Horseshoe Lake attained surficial hydrological connectivity with the Mississippi River only during high-magnitude floods (>10-m stage at Saint Louis) that inundated the majority of the floodplain around Cahokia (Fig. S2). Modern flood control infrastructure now isolates Horseshoe Lake from floods (17, 25), including the A.D. 1993 flood (26).
Fig. 1.
Results and Discussion
The sediment cores collected from Horseshoe and Grassy Lakes are predominately dark brown massive silts interbedded with distinct light gray to pale brown layers of well-sorted silty clay that we interpret as floodwater sediments deposited by the Mississippi River (Fig. 2). These floodwater sediments are similar in texture to the sediments deposited in floodplain lakes and depressions by historical overbank Mississippi River floods (21–23), and their low organic content and low concentrations of microfossils (12) are consistent with floodwater sediments identified in previous paleohydrological studies (19). At both sites, the first principal component (PC1) of grain size distributions is closely correlated to median grain size and uniformity, and PC1 explains >55% of grain size variance. We define samples with scores in the upper quartile of PC1 as floodwater sediments. These samples have median grain sizes that are significantly lower (P < 0.005; paired t tests) and more uniform than nonfloodwater sediments. To establish the timing of flood events, Bayesian age–depth models were constructed for both cores (Fig. S3) based on 17 radiocarbon dates from terrestrial plant macrofossils (Table S1), core tops, and the timing of Anglo-American settlement based on the recent increase of Ambrosia-type pollen (12) using bacon v.2.2 (27).
Fig. 2.
Based on the criteria described above, we identify eight flood events at Horseshoe Lake and five at Grassy Lake (Fig. 2). The five most recent floodwater deposits (flood events I–V) in Horseshoe Lake at A.D. 1810 (A.D. 1780–1880), A.D. 1630 (A.D. 1570–1720), A.D. 1520 (A.D. 1480–1640), A.D. 1400 (A.D. 1340–1510), and A.D. 1200 (A.D. 1080–1250) (ages reported as the mode and 95% confidence interval of probability density functions generated by bacon) temporally coincide with the five flood events recorded in Grassy Lake at A.D. 1800 (A.D. 1760–1880), A.D. 1570 (A.D. 1530–1730), A.D. 1500 (A.D. 1460–1670), A.D. 1390 (AD 1300–1570), and A.D. 1200 (A.D. 1000–1310); assuming that these five deposits represent the same floods, the joint age estimates for these five floods narrow to A.D. 1800 (A.D. 1780–1870), A.D. 1590 (A.D. 1550–1730), A.D. 1510 (A.D. 1470–1590), A.D. 1400 (A.D. 1340–1460), and A.D. 1200 (A.D. 1100–1260) (Fig. S4 and Table S2). The topmost event in both lakes corresponds with the large A.D. 1844 flood that inundated most of the central Mississippi River’s floodplain (6, 17) (see also SI Text), including the area around Cahokia (Fig. S1), but the other seven identified flood events predate the instrumental record. The three older floodwater deposits (flood events VI–VIII) in Horseshoe Lake at A.D. 580 (A.D. 510–630), A.D. 490 (A.D. 440–560), and A.D. 280 (A.D. 160–320) predate the formation of Grassy Lake. The synchronicity of the five most recent flood events at these two distant sites, and the correspondence of flood event I to a historically documented flood, strongly indicates that these floodwater sediments originated from the Mississippi River during high-magnitude floods.
The frequency of large floods in the central Mississippi River has shifted over the last two millennia (Fig. 3) as a result of atmospheric circulation patterns that promote or suppress precipitation over midcontinental North America (28–30). At Cahokia, large floods are concentrated in two intervals from ca. A.D. 300 to A.D. 600 and ca. A.D. 1200 to A.D. 1850. These intervals correspond to large flood events on the upper Mississippi River (29) and occur during periods of greater moisture availability over the midcontinent (30, 31). In contrast, no large floods occur between ca. A.D. 600 and A.D. 1200, when more arid conditions are documented across central North America (30–34). The close correspondence between midcontinental aridity and Mississippi River flood history implies that decreased moisture availability over the Missouri and Upper Mississippi River basins inhibited the intensive runoff from rainfall and/or snowmelt required to generate large floods (28, 29) that would have inundated Cahokia, neighboring settlements, and croplands in the floodplain of the central Mississippi River.
Fig. 3.
Shifts in the frequency and magnitude of flood events in the central Mississippi River valley also correspond with changes in settlement patterns, population size, and land use that define Cahokia’s emergence and decline (Fig. 3). Intensified cultivation of native domesticates in the central Mississippi River valley began during the early Late Woodland period, around A.D. 400–650 (11, 12, 35), when known settlements were concentrated on higher-elevation alluvial fans and terraces along the edge of the floodplain (16, 36). The absence of large floods from A.D. 600 to A.D. 1200 corresponds to the expansion of settlements to lower-elevation floodplain ridges that are separated by swales, sloughs, and old meander scars (14, 24, 36, 37), an increase in settlement numbers (6, 36), and continued intensification in the cultivation of native domesticates, and, after ca. A.D. 900, maize (11, 12). At A.D. 1050, toward the end of this multicentennial period of midcontinental aridity and infrequent large floods, Cahokia emerged as a hierarchically organized regional center that drew thousands of people from across the midcontinent (6, 38).
At the height of Cahokia’s size and cultural prominence, flood event V (ca. A.D. 1200)—the first large flood event in over 500 y—was of a magnitude sufficient to inundate croplands, food caches, and settlements across most of the floodplain, and would have forced residents to temporarily relocate to the higher elevations available along the edge of the floodplain and adjacent uplands. Floods in the central Mississippi River valley typically occur during the growing season (17, 29); unexpectedly high water levels at this time would have made most of the floodplain uninhabitable, and created serious and persistent agricultural shortfalls for Cahokia’s residents by destroying both crops and the agricultural surpluses from previous years stored in underground pits. After floodwaters receded, considerable effort would be required by the region’s residents to rebuild in place, clear fields of the sediment and debris that overbank floods deposit unevenly across the floodplain (21–23), and restore food production to preflood levels. Neighboring communities at higher elevations were not directly affected by flooding, but they likely played an important role during the flood by absorbing refugees from Cahokia and other settlements inundated by floodwaters, and providing labor, materials, and food in the flood’s aftermath. Maintaining political authority over this dispersed and fragmented population would have posed a significant challenge to a complex nonstate society like Cahokia, as similar societies around the world are typically limited in their ability to exert a strong and persistent hegemony over areas >40 km in diameter (39, 40). Through its direct and indirect impacts on the stability of the Cahokia region’s economy and the welfare of its inhabitants, this large and unprecedented flood held the potential to reshape the region’s sociopolitical dynamics long after the floodwaters receded.
Environmental perturbations often trigger a societal reorganization that either initiates societal breakdown or fosters resilience, with the outcome mediated by the preparedness and response of established social, political, and ideological institutions (4, 5, 41). Extensive inundation of the floodplain was unprecedented for the sociopolitical system established at Cahokia, and the return of large floods at ca. A.D. 1200 at the onset of regional depopulation (6), agricultural contraction (12), political decentralization (15), the construction of defensive palisades (42), destruction of outlying population centers (9, 43), and decline of monumental construction at Cahokia (9) implicate flooding as a factor in the reorganization of Cahokia’s sociopolitical structure that initiated its decline. In contrast to the large Mississippi River floods of the 19th century that fostered resilience and motivated legislation aimed at preventing damage from flooding (17), Cahokia’s leaders appear to have been unable to maintain the impression of security and stability following the economic upheaval created by the return of large floods. Sociopolitical disintegration progressed over the following century as the residents of the Cahokia area continued to relocate to other regions; by A.D. 1350, the sociopolitical system centered on Cahokia had completely dissolved (6, 10).
The declines of many early agricultural societies in the tropics and subtropics, including those of the ancient Puebloans, Classic Maya, Akkadians, and Harappans, are often attributed to drought and water limitation (1–3). In contrast, our work indicates that Cahokia, a sociopolitical system established in a moist and temperate region, emerged and flourished during a period of heightened midcontinental aridity and was instead vulnerable to flooding mediated by subcontinental-scale shifts in moisture availability. These findings do not preclude the role of additional factors in Cahokia’s decline, including more localized high-frequency hydroclimatic variability recorded by dendroclimatological data (7). Instead, our work emphasizes the sensitivity of fluvial systems to climatic variability (28, 29, 44) and shows that variation in flood frequency and magnitude may be an underappreciated but key factor in the development and disintegration of early agricultural societies, even in temperate regions (e.g., refs. 45 and 46). Floodwater deposits may be absent or obscured by postdepositional processes in archaeological contexts (18, 46), but, as demonstrated by this study, can be well preserved in lacustrine sedimentary records. Hydrological variability—both droughts and floods—appears to have profoundly shaped late Holocene ecosystems and societies (1, 3, 7, 33, 47), and, given the increase in the frequency of extreme hydroclimatic events projected for the 21st century (48), more work is needed to understand the coupled responses of sociocultural and hydrological systems to present and past climatic variability.
Methods
Sediment Core Extraction and Sampling.
Sediment cores with overlapping 0.5-m offsets were extracted from the in-filled thalwegs of Horseshoe Lake and Grassy Lake in May 2012 and June 2013, respectively, using a modified Livingstone piston corer with a Bolivia adapter for surface sediments. All cores were described, wrapped, and labeled in the field, then taken to the National Lacustrine Core Facility at the University of Minnesota, where they were longitudinally split, scanned for magnetic susceptibility, and photographed at high resolution. Primary core sections, overlapping core sections, and surface sediment sections were used to create continuous composite cores based on stratigraphy and magnetic susceptibility. These composite cores measured 5.5 m and 2.2 m for Horseshoe Lake and Grassy Lake, respectively. Only the top 4.4 m from Horseshoe Lake were used in the present study, because the bottom section of core is characterized by interbedded sands and clays that were deposited before the main channel of the Mississippi River migrated to the low-sinuosity meander belt along the western edge of the floodplain (12, 24). Sediment cores were sampled at 1-cm intervals and refrigerated in labeled Whirl-pak bags for future subsampling.
Radiocarbon Dates.
Seventeen wood and charcoal samples from terrestrial plant macrofossils extracted from the Horseshoe and Grassy Lake cores were collected and submitted for Accelerator Mass Spectrometry (AMS) radiocarbon dating (Table S1). Plant macrofossils were rinsed with deionized water obtained from an Academic Milli-Q water purifier with filter for organic carbon, and then dried for 24 h at 60 °C before submission to the Center for Applied Isotope Studies at the University of Georgia, or DirectAMS, for AMS dating.
Particle Size Analysis.
Sediment subsamples of 0.5 cm3 from the Horseshoe Lake and Grassy Lake cores were pretreated with 1 M HCl, to remove carbonates (i.e., gastropod shells), and rinsed with deionized water. Laser diffraction particle size analysis cannot distinguish coarse organic particles (e.g., roots, wood fragments) from coarse mineral grains, so these organics were removed by ignition at 360 °C for 2 h in a muffle furnace. Pretreated samples were then homogenized with mortar and pestle before their particle size distributions were measured on a Malvern Mastersizer 2000MU laser diffraction particle size analyzer after being dispersed with 10 mL of dispersant (0.5% sodium hexametaphosphate) and sonication for 5–15 min (midrange power, 10-μ tip displacement). To ensure disaggregation and full dispersion, samples were repeatedly sonicated and remeasured until a reproducible grain size distribution was observed. A base sampling resolution of 5–6 cm was initially used, with a higher sampling resolution (1–2 cm) used around core depths with lighter sediment color and/or low organic content that represented potential floodwater deposits. In total, 156 and 87 samples were measured at Horseshoe Lake and Grassy Lake, respectively, for a combined mean sampling resolution of 2.7 cm. A principal components analysis was performed on the full particle size distribution for all samples in each lake in R v. 2.15.1 using the princomp() function; PC1 explains >55% of grain size variance at both sites.
Age–Depth Modeling.
Age models were produced for Horseshoe Lake and Grassy Lake (Fig. S3) using bacon v.2.2 (27), calibrating all radiocarbon dates using IntCal13 (26). Additional chronological controls included the core tops (A.D. 2012 and A.D. 2013 for Horseshoe Lake and Grassy Lake, respectively) and the Euro-American settlement horizon (A.D. 1800 ± 25 y) identified from the abrupt increase in Ambrosia pollen (12) at 44 cm in Horseshoe Lake and 36 cm at Grassy Lake. To model the variable sedimentation rate in these floodplain lakes caused by flood events, we set section thickness to 2 cm, then imposed a high sedimentation rate (0.5 y/cm) on floodwater deposits > 5 cm in thickness identified from the particle size analyses, and imposed slower sedimentation rates (10 y/cm) on nonfloodwater sediments. Imposing the Bacon default sedimentation rate of 20 y/cm on nonfloodwater sediments resulted in age models that failed to pass through many calibrated chronological controls, probably because this default sedimentation rate is based primarily on small upland lakes whose geomorphic setting differs substantially from the floodplain lakes used in this study (49). Probability density functions (PDF) were obtained for each floodwater deposit using the Bacon.Age.d() function in bacon, which outputs all ages for a given depth. To develop joint PDFs and more tightly constrained age estimates for flood events I–V, the individual PDFs from Horseshoe and Grassy Lakes were multiplied together and rescaled to sum to a probability of 1 (Table S2).
Data Availability
Data deposition: The data reported in this paper are available in Dataset S1.
Acknowledgments
We thank the Illinois Department of Natural Resources for lake access; the Illinois State Geological Survey for providing the lidar data; R. Brugam, R. Criss, S. Goring, J. Kelly, J. Knox, J. Mason, D. Mladenoff, and M. Ozdogan for discussions; J. A. Brown and two anonymous reviewers for comments on an earlier version of this manuscript; S. Atkinson, J. Blois, K. Burke, J. Gill, J. Kershner, S. Knapp, M. Martinez, D. McCay, G. Schellinger, B. Rongstad, and C. Zirbel for field and/or laboratory assistance; and K. Brady, A. Myrbo, A. Noren, and R. O’Grady of the National Lacustrine Core Facility at the University of Minnesota for help with initial core descriptions. This work was supported by the National Science Foundation (Grants BCS-1333070 and DGE-1144754), the National Geographic Society (Grant YEG-9008-11), the National Lacustrine Core Facility, the Geological Society of America, and a Packard Fellowship to (D.A.F.).
Supporting Information
Supporting Information (PDF)
Supporting Information
- Download
- 1.32 MB
pnas.1501904112.sd01.xlsx
- Download
- 30.34 KB
References
1
PB deMenocal, Cultural responses to climate change during the late Holocene. Science 292, 667–673 (2001).
2
Y Dixit, DA Hodell, CA Petrie, Abrupt weakening of the summer monsoon in northwest India ∼4100 yr ago. Geology 42, 339–342 (2014).
3
L Giosan, et al., Fluvial landscapes of the Harappan civilization. Proc Natl Acad Sci USA 109, E1688–E1694 (2012).
4
KW Butzer, Collapse, environment, and society. Proc Natl Acad Sci USA 109, 3632–3639 (2012).
5
GD Middleton, Nothing lasts forever: Environmental discourses on the collapse of past societies. J Archaeol Res 20, 257–307 (2012).
6
GR Milner The Cahokia Chiefdom: The Archaeology of a Mississippian Society (Smithsonian Inst. Press, Washington, DC, 1998).
7
LV Benson, TR Pauketat, ER Cook, Cahokia’s boom and bust in the context of climate change. Am Antiq 74, 467–483 (2009).
8
NH Lopinot, WI Woods, Wood overexploitation and the collapse of Cahokia. Foraging and Farming in the Eastern Woodlands, ed EM Scarry (Univ Florida Press, Gainesville), pp. 206–231 (1993).
9
JE Kelly, Contemplating Cahokia’s collapse. Global Perspectives on the Collapse of Complex Systems, eds JA Railey, RM Reycraft (Maxwell Mus Anthropol, Albuquerque, NM), pp. 147–168 (2009).
10
TR Pauketat The Ascent of chiefs: Cahokia and Mississippian Politics in Native North America (Univ Alabama Press, Tuscaloosa, 1994).
11
ML Simon, KE Parker, Prehistoric plant use in the American Bottom: New thoughts and interpretations. Southeast Archaeol. 25, 212–257 (2006).
12
SE Munoz, S Schroeder, DA Fike, JW Williams, A record of sustained prehistoric and historic land use from the Cahokia region, Illinois, USA. Geology 42, 499–502 (2014).
13
TR Pauketat, NH Lopinot, Cahokian population dynamics. Cahokia: Domination and Ideology in the Mississippian World, eds Pauketat TR, Emerson TE (Univ Nebraska Press, Lincoln), pp 103–123. (1997).
14
S Schroeder, Power and place: Agency, ecology, and history in the American Bottom, Illinois. Antiquity 78, 812–827 (2004).
15
MBD Trubitt, Mound building and prestige goods exchange: Changing strategies in the Cahokia chiefdom. Am Antiq 65, 669–690 (2000).
16
JE Kelly, FA Finney, DL McElrath, SJ Ozuk, Late Woodland period. American Bottom Archaeology, eds CJ Bareis, JW Porter (Univ Illinois Press, Urbana), pp. 104–127 (1984).
17
FJ Dobney River Engineers on the Middle Mississippi: A History of the St. Louis District, U.S. Army Corps of Engineers (US Gov Print Off, Washington, DC, 1978).
18
JE Kelly, Formative developments at Cahokia and the adjacent American Bottom: A Merrell tract perspective. PhD dissertation, University of Wisconsin–Madison, Madison, WI. (1980).
19
A Gilli, FS Anselmetti, L Glur, SB Wirth, Lake sediments as archives of recurrence rates and intensities of past flood events. Dating Torrential Processes on Fans and Cones—Methods and Their Application for Hazard and Risk Assessment, eds MS Schneuwly-Bollschweiler, M Stoffel, F Rudolf-Miklau (Springer, New York), pp. 225–242 (2013).
20
L Glur, et al., Frequent floods in the European Alps coincide with cooler periods of the past 2500 years. Sci Rep 3, 2770 (2013).
21
RH Kesel, KC Dunne, RC McDonald, KR Allison, BE Spicer, Lateral erosion and overbank deposition on the Mississippi River in Louisiana caused by 1973 flooding. Geology 2, 461–464 (1974).
22
B Gomez, JD Phillips, FJ Magilligan, LA James, Floodplain sedimentation and sensitivity: Summer 1993 flood, upper Mississippi River valley. Earth Surf Processes Landforms 22, 923–936 (1997).
23
NC Grover, Floods of Ohio and Mississippi Rivers January-February 1937 US Geol Surv Water Suppl Pap 838:1–746. (1938).
24
ER Hajic, Geomorphology of the northern American Bottom as context for archaeology. Ill Archaeol 5, 54–65 (1993).
25
I Karthic, RB Brugam, W Retzlaff, K Johnson, The impact of nitrogen contamination and river modification on a Mississippi River floodplain lake. Sci Total Environ 463-464, 734–742 (2013).
26
; NASA Landsat Program Landsat 5 scene LT50240331993231XXX02 (US Geol Surv, Sioux Falls, SD, 1993).
27
M Blaauw, JA Christen, Flexible paleoclimate age-depth models using an autoregressive gamma process. Bayesian Anal 6, 457–474 (2011).
28
JC Knox, Sensitivity of modern and Holocene floods to climate change. Quat Sci Rev 19, 439–457 (2000).
29
JC Knox, North American palaeofloods and future floods: Responses to climatic change. Palaeohydrology: Understanding Global Change, eds KJ Gregory, G Benito (Wiley, Hoboken, NJ), pp. 143–164 (2003).
30
JC Montero-Serrano, et al., Changes in precipitation regimes over North America during the Holocene as recorded by mineralogy and geochemistry of Gulf of Mexico sediments. Global Planet Change 74, 132–143 (2010).
31
JM Daniels, JC Knox, Alluvial stratigraphic evidence for channel incision during the Mediaeval warm period on the central Great Plains, USA. Holocene 15, 736–747 (2005).
32
CA Woodhouse, JT Overpeck, 2000 years of drought variability in the central United States. Bull Am Meteorol Soc 79, 2693–2714 (1998).
33
ER Cook, et al., Megadroughts in North America: Placing IPCC projections of hydroclimatic change in a long-term palaeoclimate context. J Quat Sci 25, 48–61 (2010).
34
; PAGES 2k Consortium, Continental-scale temperature variability during the past two millennia. Nat Geosci 6, 339–346 (2013).
35
AD Fortier, TE Emerson, DL McElrath, Calibrating and reassessing American Bottom culture history. Southeast Archaeol. 25, 170–211 (2006).
36
DL McElrath, AC Fortier, The early Late Woodland occupation of the American Bottom. Late Woodland Societies: Tradition and Transformation Across the Midcontinent, eds Emerson TE, McElrath DL, Fortier AC (Univ Nebraska Press, Lincoln), pp 97–121. (2000).
37
WP White, S Johannessen, PG Cross, LS Kelly, Environmental setting. American Bottom Archaeology, eds CJ Bareis, JW Porter (Univ Illinois Press, Urbana), pp. 15–33 (1984).
38
PA Slater, KM Hedman, TE Emerson, Immigrants at the Mississippian polity of Cahokia: Strontium isotopic evidence for population movement. J Archaeol Sci 44, 117–127 (2014).
39
DJ Hally, MT Smith, Jr JB Langford The archaeological reality of de Soto’s Coosa. Columbian Consequences, ed DT Thomas (Smithsonian Inst Press, Washington, DC) Vol 2, 121–138 (1990).
40
DJ Hally, The territorial size of Mississippian chiefdoms. Archaeology of Eastern North America: Papers in Honor of Stephen Williams, ed JB Stoltman (Mississippi Dep Archives Hist, Jackson), pp. 143–168 (1993).
41
SC Meeks, DG Anderson, Drought, subsistence stress, and population dynamics. Soils, Climate and Society: Archaeological Investigations in Ancient America, eds DJ Wingard, SE Hayes (Univ Colorado Press, Boulder), pp. 61–83 (2013).
42
WR Iseminger The Archaeology of the Cahokia Palisade (Illinois Hist Preserv Agency, Springfield, 1990).
43
TR Pauketat, AC Fortier, SM Alt, TE Emerson, A Mississippian conflagration at East St. Louis and its political-historical implications. J Field Archaeol 38, 210–226 (2013).
44
TR Kidder, Climate change and the Archaic to Woodland transition (3000–2500 Cal B.P.) in the Mississippi River Basin. Am Antiq 71, 195–231 (2006).
45
TR Kidder, H Liu, M Li, Sanyangzhuang: Early farming and a Han settlement preserved beneath Yellow River flood deposits. Antiquity 86, 30–47 (2012).
46
KW Butzer, I Miralles, JF Mateu, Urban geo-archaeology in Medieval Alzira (Prov. Valencia, Spain). J Archaeol Sci 10(4), 333–349 (1983).
47
DG Anderson, DW Stahle, MK Cleaveland, Paleoclimate and the potential food reserves of Mississippian societies: A case study from the Savannah River valley. Am Antiq 60, 258–286 (1995).
48
C Prudhomme, et al., Hydrological droughts in the 21st century, hotspots and uncertainties from a global multimodel ensemble experiment. Proc Natl Acad Sci USA 111, 3262–3267 (2014).
49
S Goring, et al., Deposition times in the northeastern United States during the Holocene: Establishing valid priors for Bayesian age models. Quat Sci Rev 48, 54–60 (2012).
Information & Authors
Information
Published in
Classifications
Data Availability
Data deposition: The data reported in this paper are available in Dataset S1.
Submission history
Published online: May 4, 2015
Published in issue: May 19, 2015
Keywords
Acknowledgments
We thank the Illinois Department of Natural Resources for lake access; the Illinois State Geological Survey for providing the lidar data; R. Brugam, R. Criss, S. Goring, J. Kelly, J. Knox, J. Mason, D. Mladenoff, and M. Ozdogan for discussions; J. A. Brown and two anonymous reviewers for comments on an earlier version of this manuscript; S. Atkinson, J. Blois, K. Burke, J. Gill, J. Kershner, S. Knapp, M. Martinez, D. McCay, G. Schellinger, B. Rongstad, and C. Zirbel for field and/or laboratory assistance; and K. Brady, A. Myrbo, A. Noren, and R. O’Grady of the National Lacustrine Core Facility at the University of Minnesota for help with initial core descriptions. This work was supported by the National Science Foundation (Grants BCS-1333070 and DGE-1144754), the National Geographic Society (Grant YEG-9008-11), the National Lacustrine Core Facility, the Geological Society of America, and a Packard Fellowship to (D.A.F.).
Notes
This article is a PNAS Direct Submission.
Authors
Competing Interests
The authors declare no conflict of interest.
Metrics & Citations
Metrics
Citation statements
Altmetrics
Citations
Cite this article
112 (20) 6319-6324,
Export the article citation data by selecting a format from the list below and clicking Export.
Cited by
Loading...
View Options
View options
PDF format
Download this article as a PDF file
DOWNLOAD PDFLogin options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginRecommend to a librarian
Recommend PNAS to a LibrarianPurchase options
Purchase this article to access the full text.