Outcrop versus core and geophysical log interpretation of mid-Cretaceous paleosols from the Dakota Formation of Kansas

Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

Contents lists available at SciVerse ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology journal homepage: www.elsevier.com/locate/palaeo

Outcrop versus core and geophysical log interpretation of mid-Cretaceous paleosols from the Dakota Formation of Kansas Gregory J. Retallack a,⁎, David L. Dilcher b, c a b c

Department of Geological Sciences, University of Oregon, Eugene, OR 97403, U.S.A. Department of Biology, Indiana University, Bloomington, IN 47401, U.S.A. Department of Geology, Indiana University, Bloomington, IN 47401, U.S.A.

a r t i c l e

i n f o

Article history: Received 18 February 2011 Received in revised form 9 February 2012 Accepted 14 February 2012 Available online 23 February 2012 Keywords: Cretaceous Paleosol Kansas Paleoclimate Diagenesis Weathering

a b s t r a c t Paleosols of the mid-Cretaceous (late Albian to early Cenomanian) Dakota Formation in Kansas and Nebraska are well exposed in road, river and clay pit exposures, but show systematic overprinting by modern weathering. Examination of deep drill cores reveals that modern weathering effects include (1) oxidation of pyrite to jarosite, (2) reaction of jarosite and calcite to clear selenite, and (3) oxidation of sphaerosiderite and siderite nodules to hematite. Surface oxidation in this region extends no deeper than 2 m from the surface. Hematite nodules and coatings of leaves are found in core only at one stratigraphic horizon, which was not as appears in outcrop, a laterite. Transformed matrix density and photoelectric absorption logs of cores reveal that kaolinite clays are most abundant within deeply weathered paleosols (Sayela pedotype) at this stratigraphic level, near the Albian–Cenomanian boundary, when angiosperm forests were especially diverse and lived in a subtropical humid climate. Other paleosols of the early Cenomanian were less deeply weathered, and fossil plants of this age included temperate humid angiosperm mangal and swamp woodland, but conifer-dominated floodplain forests. This short-lived mid-Cretaceous warm–wet spike has also been recognized in other paleosols of Minnesota, Utah, Nevada, and New Mexico, and coincides with marine black shale of the Breitstroffer event, midcontinental marine transgression, a sequence-stratigraphic boundary, and a marked increase in early angiosperm diversity. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Weathering of paleosols in outcrop can compromise interpretation of their past environments. The best recourse when surface weathering is suspected is to examine the paleosols in deep drill core. Confusion of modern and ancient weathering was controversial for interpretation of redox and base status of Cambrian paleosols (Retallack, 2008) in South Australia and of Archean paleosols (Ohmoto et al., 2007) and banded iron formations (Hoashi et al., 2009) in Western Australia, but solved by examination of core and geophysical core logs. Furthermore, core logs come with useful geophysical logs such as gamma ray profiles (MacFarlane et al., 1989), and this paper attempts to convert such data to information critical for interpretation of paleosols (Retallack, 1997). This comparison of paleosols of the mid-Cretaceous Dakota Formation in Kansas within cores and outcrop (Fig. 1) thus itemizes common alterations of paleosols on exposure, and explores paleosol interpretation using wireline logs.

⁎ Corresponding author. E-mail address: [email protected] (G.J. Retallack). 0031-0182/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.palaeo.2012.02.017

This study of cores and geophysical logs has ramifications for interpretation of paleoclimates and paleoenvironments. For example, paleosols of the Dakota Formation identified as laterites in outcrop by Thorp and Reed (1949) and Reed (1954) are shown here to have been subtropical Ultisols, but not tropical laterites (Retallack, 2010). Furthermore, not all red paleosols in outcrops of the Dakota Formation were originally red or deeply weathered. One horizon of Ultisol paleosols records global warmth and humidity from a CO2 greenhouse spike, recorded elsewhere as the Breitstroffer marine anoxia event (Retallack, 2009), but other paleosols of the upper Dakota Formation now red in outcrop, were originally gray, sideritic paleosols of waterlogged coastal plains. Paleosols were first recognized in the mid-Cretaceous Dakota Formation by soil scientists (Thorp and Reed, 1949; Reed, 1954), but also have attracted geological attention (Retallack and Dilcher, 1981a,b; Joeckel, 1987, 1991; Porter et al., 1991). The Dakota Formation has long been known for its sandstone fossil floras, collected initially by George Miller Sternberg (Rogers, 1991) for Leo Lesquereux (1892), and more recently collected for shale floras with preserved cuticles (Dilcher et al., 1976; Dilcher, 1979; Basinger and Dilcher, 1984; Dilcher and Crane, 1984; Dilcher and Kovach, 1986; Skog and Dilcher, 1992, 1994; Wang, 2002; Wang and Dilcher, 2006; Dilcher and Wang, 2009; Wang et al., 2011). The Dakota Formation is also

48

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

Fig. 1. Map of locations examined within the Dakota Formation in central Kansas and Nebraska after Retallack and Dilcher, 1981a,b) and cross-section from borehole data after Hamilton (1994).

rich in fossil pollen (Farley and Dilcher, 1986; Kovach, 1988; Kovach and Dilcher, 1988; Ravn and Witzke, 1994, 1995; Hu et al., 2008a,b), but poor in fossils other than plants. Moths, weevils, leaf beetles, and midges are inferred mainly from damage to fossil leaves (Labandiera, 1998), and a few insect fossils (Retallack and Dilcher, 1981a). Fossil molluscs are rare molds and casts of freshwater to brackish forms (White, 1894; Siemers, 1976). A locality near Wilson (Fig. 1) has yielded 6 species of shark teeth, 6 different rays and 7 kinds of bony fish (Everhart et al., 2005), localities near Carneiro and Salina produced the giant crocodile Dakotasuchus kingi (Mehl, 1941; Vaughn, 1956), and another crocodile was found in Big Creek near Russell (D.E Hattin pers comm., 1978). A species of nodosaur (Silvisaurus condrayi Eaton, 1960) from near Minneapolis and also near Wilson (both in Kansas), and a large ornithopod from near Decatur, Nebraska (Barbour, 1931; Galton and Jensen, 1978), are the only dinosaur bones from the Dakota Formation, though dinosaur gastroliths and tridactyl footprints are also recorded (Joeckel et al., 2001; Liggett, 2005). This unbalanced pattern of preservation also may be related to paleosols (Retallack, 1998), and this paper also

attempts to detail taphonomic biases in reconstruction of past ecosystems from paleosol-rich sequences. 2. Geological background The age of the Dakota Formation in Kansas is bracketed by marine fossils in underlying and overlying rocks (Fig. 1). Fossil foraminifera are evidence for an Albian age of the underlying Kiowa Formation (Loeblich and Tappan, 1961) and an age of Cenomanian from foraminifera in the overlying Graneros Shale (Eicher, 1965) is supported by associated ammonites and clams (Hattin, 1967; Hattin et al., 1978). Palynofloras are evidence that the Lower Dakota Formation is late Albian, but the middle and upper Dakota Formation are lower to middle Cenomanian (Pierce, 1961; Farley and Dilcher, 1986; Ravn and Witzke, 1994, 1995; Hu et al., 2008a). The Albian–Cenomanian boundary has been recognized using palynology and carbon isotope chemostratigraphy within the Rose Creek clay pit south of Fairbury, Nebraska (Gröcke et al., 2006). This boundary is marked by exceptionally developed paleosols near the Rose

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

Creek pit and at other localities such as Carneiro and Delphos (Sayela pedotype of Fig. 2) at the marine regression forming the junction of J and D sequence stratigraphic units recognized in regional correlations

49

(Franks, 1975; Hamilton, 1994). This key sequence boundary is within the Terracotta Clay Member of the Dakota Formation, as defined by Plummer and Romary (1947) and Plummer et al. (1960), and not at

Fig. 2. Measured sections of paleosols in the Dakota Formation of Kansas (B–N) and Nebraska (A). Paleosols are classified into pedotypes using Lakota Sioux descriptors (Table 1), and the thickness is shown by boxes whose width corresponds to degree of development. Scales of development and calcareousness from field application of dilute acid follow Retallack (1997), and colors are from a Munsell chart.

50

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

the boundary of the Terracotta and overlying Janssen Clay Member (Hamilton, 1994). The deeply weathered Sayela paleosols can be regionally correlated through Nebraska (Porter et al., 1991) and Iowa (Witzke and Ludvigson, 1994), where they separate a lower Nishnabotna and upper Woodbury Member of the Dakota Formation (Karl, 1976). The deeply weathered paleosols developed in granite (Goldich, 1938), quartzite (Austin, 1970) and paleokarst (Sloan, 1964) beneath the base of the Dakota Formation in Minnesota are of comparable geological age and exceptional degree of weathering (Setterholm, 1994). Sandstones of the lower Dakota Formation (upper J) include large (up to 6 m diameter) spherical to ellipsoidal nodules of cross bedded sandstone cemented by calcite, notably at Mushroom Rock State Park (N38.726005° W98.030718°) south of Carneiro, at Rock City Park (N39.089852° W97.739759°) southwest of Minneapolis, and Quartzite Quarry Park (N39.028497° W97.162289°) southwest of Lincoln (McBride and Milliken, 2006). Sandstones of the upper Dakota Formation (unit D) lack such features, although they weather into hoodoos and show meandering paleochannels such as the Rocktown channel, near Russell (N38.966597° W98.857589°; Siemers, 1976; Retallack and Dilcher, 1981a). The Dakota Formation in Kansas is flat-lying and unmetamorphosed. Smectite and kaolinite are the principal clay minerals in the Dakota Formation of Kansas (Plummer and Romary, 1947; Plummer et al., 1960), Nebraska (Joeckel, 1987, 1991), Illinois (Frye et al., 1964) and Minnesota (Parham and Hogberg, 1964; Parham, 1970). These clays are unillitized by burial, which would not have exceeded 0.8 km in Kansas through Minnesota (Dyman et al., 1994; McBride and Milliken, 2006). Ptygmatic folding of clastic dikes to 38% of their former thickness has been observed in carbonaceous shales of the Dakota Formation in northwestern North Dakota (Shelton, 1962), where there was 1.6 km of overlying Cretaceous and 0.2 km of Cenozoic rocks (Dyman et al., 1994). Standard compaction formulae for paleosols (Sheldon and Retallack, 2001) give original thicknesses of Ultisols (Su in cm) and Histosols (Sh in cm) from thickness of comparable paleosol (Bp in cm) for a known burial depth (K in km), as follows. Su ¼ Sp

 . .0:42 −0:58 −1 0:2K e

ð1Þ

Sh ¼ Sp

 . . 0:94 −0:06 −1 2:09K e

ð2Þ

Using these equations for kaolinitic paleosols (Ultisols) of the Dakota Formation gives reduction to 90% of former thickness at 0.8 km burial, and 82% at 1.8 km, and coals (Histosols) give reduction to 7% at 0.8 km and 6% at 1.8 km. Early diagenetic sphaerosiderite from methanogenic pedogenesis was in some locations followed by pyritization due to sulfate reduction (Brenner et al., 1981). Large (6 m) spherical nodules in sandstone paleochannels of the lower Dakota Formation in Kansas have oxygen isotope compositions interpreted as evidence of warm precipitation temperatures (30–50 °C)

and shallow (ca. 0.4 km) depths of burial (McBride and Milliken, 2006). Although called “concretions” by McBride and Milliken (2006), they lack the growth banding required of concretions in soil science terminology (Brewer, 1976), and are thus nodules. Most hematite in Dakota Formation paleosols is paleomagnetically reversed and so compatible with a Cretaceous age, but one sample is normal and oxidized well after burial (Faulds et al., 1996). 3. Materials and methods Stratigraphic sections were measured using tape and level on a Brunton compass, with degree of development and of dilute acid reaction estimated in the field using the scale of Retallack (1997) and color from a Munsell chart with tropical and gley pages (Fig. 2). Localities examined for this study in stratigraphic succession were: 1, Carneiro, Ellsworth Co. (Acme Brick Company Ramsay pit N38.708031° W98.090387°); 2, Fairbury, Jefferson Co. (Endicott Clay Company, Rose Creek pit N40.050094° W97.169206°); 3, Delphos, Cloud Co. (Braun's farm N39.336428° W97.604351°); 4, Kanapolis, Ellsworth Co. (Acme Brick Company, Vondra pit N38.714624° W98.115197°); 5, Hoisington, Barton Co. (3 localities in Kansas Brick and Tile company quarry, “north” N38.474784° W98.791444°, “southwest” N38.471302° W98.781502°, and “south” N38.469133° W98.780099°); 6, Ellsworth, Ellsworth Co. (gully at N38.816884° W98.137002°); 7, Bunker Hill, Russell Co. (2 sites on Linnenberger brothers ranch; “west” N38.908973° W98.663604° and “east” N38.910887° W98.656723°); 8, Concordia, Cloud Co. (Cloud Ceramic Company pit N39.533929° W97.597685°); 9, Black Wolf, Ellsworth Co. (gully at N38.739122° W98.369881°); 10, Russell, Russell Co (bluffs north of Saline River N38.966597° W98.857589°); 11, Wilson, Ellsworth Co. (gully at N38.782809° W98.493196°). Boreholes examined include KGS Braun #1 (N38.986373° W99.354256°) in Ellis Co., Jones #1 (N39.219036° W98.176458°) in Lincoln Co., and Kenyon #1 (N39.683807° W97.710725°) in Republic Co. The Dakota Formation from 146.9 m down to 239.3 m in Braun #1 (Macfarlane et al., 1990) is a useful reference section for relative stratigraphic level of these localities, estimated from local level relative to Sayela paleosols near D–J sandstone contact and Graneros Shale (Retallack and Dilcher, 1981a,b) as follows: locality 1 at 204–210 m, 2 at 197–204 m, 3 at 184–192 m, 4 at 169–176 m, 5 at 151–163 m, 6 at 157–175 m, 7 at 157–172 m, 8 at 147–164 m, 9 at 147–161 m, 10 at 147–168 m, and 11 at 135–169 m. Paleosols were classified into different field categories (pedotypes of Table 1) and named for non-genetic descriptions in the Lakota Sioux Native American language (Buechel, 1970). Pedotype terminology developed here can also be applied to other described paleosols, for example, Sayela paleosols near Brookville, Kansas (Thorp and Reed, 1949; Porter et al., 1991) and Endicott, Nebraska (Joeckel, 1991), and Casmu, Sagi and two Sapa paleosols along Little Blue River near Fairbury, Nebraska (Joeckel, 1987). The “Morton profile” of Sloan (1964) developed on granite beneath the Dakota Formation in Minnesota (Goldich, 1938) is an additional mid-Cretaceous pedotype. Comparable paleosols are found in the Cenomanian Dunvegan

Table 1 Diagnosis and classification of pedotypes of the Dakota Formation in Kansas. Pedotype

Lakota meaning

Diagnosis

Soil taxonomy

FAO soil map

Australian classification

Cahli

Coal

Fibrist

Dystric Histosol

Fibric Organosol

Casmu Kazela Makazi Sagi Sapa Sayela

Sandy Shallow Sulfur Auburn colored Dirty Reddest

Thick coal (O) over white clayey underclay with carbonaceous root traces (A) and bedded siltstone (C) Pyrite encrusted carbonaceous root traces (A) in bedded sandstone (C) Carbonaceous root traces (A) in gray shale (C) Carbonaceous claystone with pyrite encrusted root traces (A) over bedded shale (C) Ferruginized sandstone and root traces (A) in bedded sandstone and shale (C) Carbonaceous claystone with root traces (A) over sphaerosideritic claystones (By) Light gray clayey siltstone (A) over red mottled claystone (Bt)

Sulfipsamment Fluvent Sulfaquept Psamment Kandihumult Kandiudult

Thionic Fluvisol Dystric Fluvisol Thionic Fluvisol Ferralic Arenosol Humic Gleysol Ferric Acrisol

Intertidal Hydrosol Stratic Rudosol Intertidal Hydrosol Arenic Rudosol Oxyaquic Hydrosol Red Chromosol

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

Formation of Alberta and British Columbia, Canada (McCarthy and Plint, 1999, 2003; McCarthy et al., 1999; McCarthy, 2002), especially their “type 1” (like Casmu here),” type 2” (like Sapa), “type 3” (like Cahli), “type 4” (Cahli–Sapa–Casmu pedocomplex) and “type 5” (Sapa–Cahli pedocomplex). Type profiles of the pedotypes (Fig. 3) were sampled for field laboratory study. These and other data were the basis for pedotype-specific paleoenvironmental interpretations (Table 2), and taphonomic biases and outcrop overprints discussed in this paper. Samples of paleosols were collected in the field and petrographic thin sections prepared using kerosene as a non-polar coolant and epoxy resin as a specimen stabilizer of these friable specimens (Tate and Retallack, 1995). Samples were analyzed for major oxides using atomic absorption by Christine McBirney (University of Oregon), who also determined ferrous iron using the Pratt titration. Weight percent organic carbon was analyzed by the soil analytical laboratory at Oregon State University using the Walkley-Black titration. Bulk

51

density was determined weighing paraffin-coated clods some 2 cm 2 in and out of water, in order to calculate mass transfer and strain of elemental composition. Thin sections were point-counted using a Swift Automated stage and collator and an eyepiece micrometer to determine proportions of sand–silt–clay and mineral components of the paleosols (Fig. 3). Molar ratios were also calculated from bulk chemical analyses to give products over reactants of common soil forming chemical processes (Retallack, 1997), such as salinization, calcification, clayeyness, base loss and gleization (Fig. 2). A more detailed accounting of geochemical change following Brimhall et al. (1991) is mass transfer of elements in a soil at a given horizon (τj,w in mol) calculated from the bulk density of the soil (ρw in g cm − 3) and parent material (ρp in g cm − 3) and from the chemical concentration of the element in soils (Cj,w in wt.%) and parent material (Cp,w in wt.%). Changes in volume of soil during weathering are called strain by Brimhall et al. (1991), and estimated from an immobile element in soil (such as Ti

Fig. 3. Grain size and mineral (from point-counting) and chemical composition (from Walkley-Black titration and Atomic Absorption) of paleosols in the Dakota Formation of Kansas. Molar weathering ratios are designed to reveal degree of common soil forming reactions, such as base loss (Retallack, 1997).

52

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

Table 2 Interpretation of pedotypes of the Dakota Formation in Kansas. Pedotype

Paleoclimate

Ecosystems

Paleotopography

Parent material

Time for formation

Cahli

Not diagnostic for climate

Swampy lakes and oxbows

Smectite clay

3–13 ka

Casmu

Not diagnostic for climate

Supratidal estuarine flats

Not diagnostic for climate

Smectite clay and quartz silt Smectite clay

0.01 ka

Kazela Makizi

Not diagnostic for climate Not diagnostic for climate

Stream levee tops

Smectite clay and quartz silt Quartz sand

2–5 ka

Sagi Sapa

Warm temperate (14.2 ± 4.4 °C mean annual temperature) humid (1482 ± 182 mm mean annual precipitation), short dry season Subtropical (17 ± 4.4 °C mean annual temperature) humid (1527 ± 182 mm mean annual precipitation)

Angiosperm (Sapindopsis bagleyae, Liriophyllum kansense) swamp woodland Angiosperm (Araliopsoides cretacea) colonizing woodland Angiosperm (Aquatifolia fluitans) aquatic vegetation Angiosperm (Pabiania variloba) mangal with mussels (Brachidontes filiscuptus) Angiosperm (Araliopsoides cretacea) colonizing woodland Conifer (Sequoia condita) swamp woodland

Seasonally waterlogged floodplain

Smectite clay and quartz silt

112–640 ka

Angiosperm (Rogersia parlatoriii) wet forest

Well drained floodplain

Kaolinite clay and quartz silt

25–107 ka

Sayela

used here) compared with parent material (εi,w as a fraction). The relevant Eqs. (1) and (2) (below) are the basis for calculating divergence from parent material composition (origin in various panels of Fig. 4). "

τj;w

# i ρw ⋅ C j;w h ¼ εi;w þ 1 −1 ρp ⋅ C j;p "

ε i;w ¼

# ρp ⋅ C j;p −1 ρw ⋅ C j;w

ð3Þ

Streamside clayey swales Intertidal estuarine flats

0.01 ka

0.01 ka

some of the pedotypes as weakly developed (Entisols), and coal marks the Cahli pedotype as peaty (Histosols). Kaolinitic clays, low base status and thick clay-enriched subsurface horizons mark others as Ultisols, with some showing hematite mottles as evidence of good drainage (Sayela pedotype) and others carbonaceous with sphaerosiderite as evidence of poor drainage (Sapa pedotype). 4. Features of the paleosols

ð4Þ

Clear weathering trends were found in Sayela paleosols (Fig. 4), but other paleosols show variations reflecting sedimentary parent heterogeneity. These data also enable interpretation of the pedotypes in terms of modern soil classifications, such as those of the Soil Survey Staff (2000) of the United States Natural Resources Conservation Service, the Food and Agriculture Organization (1974) of the United Nations Educational, Scientific and Cultural Organization based in Paris, and the Australian soil classification (Isbell, 1996). Within the US system, for example, prominent relict bedding (Casmu, Sagi and Makizi) mark

Large woody root traces are a conspicuous feature of many paleosols in the Dakota Formation, and include carbonaceous remains of roots heavily encrusted with pyrite (Fig. 5D). Other root traces are filled with clay and sphaerosiderite, but retain the downward tapering, bracing and irregular distribution of plant roots. Most paleosol tops can be recognized in the field from the level at which root traces are truncated by overlying beds. The Dakota Formation has well preserved sedimentary structures, such as paleochannels and flaser and linsel bedding (Siemers, 1976), but also thick horizons in which primary bedding has been destroyed. These unbedded rock units have two particular characteristics of soil horizons: sharply truncated tops and diffuse alteration downward

Fig. 4. Mass balance geochemistry of paleosols in the Dakota Formation of Kansas, including estimates of strain from changes in an element assumed stable (Ti) and elemental mass transfer with respect to an element assumed stable (Ti, following Brimhall et al., 1991). Zero strain and mass transfer is the parent material lower in the profile: higher horizons deviate from that point due to pedogenesis.

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

53

Fig. 5. Field appearance of Cretaceous paleosols of the Dakota Formation Kansas; A, Sayela pedotype (below siltstone bench near top) in Acme clay pit east of Kanapolis; B, Sagi paleosol (in sun immediately below hammer) in levee facies, Kansas Brick and Tile Company pit south of Hoisington; C, sequence of paleosols along Saline River north of Russell (see Fig. 2B for interpretation); D, pyritized root traces (shining and golden) in gray Makizi pedotype of Cloud Ceramic Company pit south of Concordia; E, Cahli pedotype paleosol from Linnenberger's Ranch west, near Bunker Hill; F, Gypsum crystals littering gray shales along Saline River, north of Russell; G, red mottles in from oxidized sphaerosiderite from Sapa pedotype near Ellsworth.

(Fig. 5C,E). Carbonaceous surface horizons (A horizons) and coals (O horizons) are common in these paleosols, as are subsurface zones of red mottling (Bw horizon), clay enrichment (Bt horizon) and abundant replacive sphaerosiderite (Bg horizon). Paleosols of the Dakota Formation also have a variety of soil structures including subangular clayey units (blocky peds) surrounded by slickensided clay skins (argillans), and vertical structures (columnar peds) outlined by ferruginized planes (ferrans). Some of the intensely ferruginized units have been interpreted as laterite (Thorp and Reed, 1949) or plinthite (Joeckel, 1991), but chemical (Fig. 4) and petrographic analysis (Fig. 3) of likely rocks analyzed in this study falls short of the degree of iron enrichment found in plinthite, murram or laterite (Retallack, 2010). In thin section also, these paleosols show sepic plasmic fabrics of highly oriented and birefringent clay within fields of less oriented clay (Fig. 6C–D), which develop within

soils due to the complex and highly deviatoric stress fields imposed by desiccation, rooting and burrowing (Brewer, 1976). A distinctive microstructure revealed by examination under crossed nicols of underclays of Cahli pedotype paleosols is inundulic plasmic fabric (Fig. 6A,B, see also color figure of Retallack, 1997, color photos 55, 56). This nearly isotropic fabric may result from clay flocculation in wetlands (Brewer, 1976). 5. Paleoenvironmental interpretations Paleoenvironmental interpretations of fossil plants and soils are detailed in the following paragraphs and illustrations (Figs. 6 and 7), before considering the degree to which these interpretations may be compromised by surface weathering and paleoenvironmental biases in fossil preservation.

54

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

Fig. 6. Thin section photomicrographs of Cretaceous paleosols of Kansas: A–B, inundulic plasmic fabric from A horizon of Cahli pedotype on Linnenberger's Ranch east of Bunker Hill, under crossed nicols and plane light, respectively; C, climobimasepic plasmic fabric from Bw horizon of Makizi pedotype from Fairbury under crossed nicols; D, mosepic plasmic fabric from A horizon of Sapa paleosol along the Saline River, north of Russell. Specimens in collections of University of Oregon are R108 (A–B), R110 (C) and R101 (D).

5.1. Paleoclimate Paleosols of the Dakota Formation in Kansas, Nebraska, and Minnesota are non-calcareous (pedalfers), and such soils indicate a humid climate in which precipitation exceeds evaporation (Retallack and Dilcher, 1981a, b; Joeckel, 1991). A more precise estimate of paleoprecipitation can be gained from the paleohyetometer of Sheldon et al. (2002), using chemical index of alteration without potash (R = 100mAl2O3/(mAl2O3 + mCaO + mNa2O), in mol), which increases with mean annual precipitation (P in mm) in modern soils (R 2 = 0.72; S.E. = ± 182 mm), as follows. 0:0197R

P ¼ 221e

ð5Þ

This formulation is based on the hydrolysis equation of weathering, which enriches alumna at the expense of lime, magnesia, potash and soda. Magnesia is ignored because it is not significant for most sedimentary rocks, and potash is excluded because it can be enriched during deep burial alteration of sediments (Maynard, 1992). Unlike profile specific mass balance (Fig. 4), this method conflates weathering that contributed to the parent material as well as within the paleosols. The estimated mean annual precipitation for a Sayela paleosols of the Dakota Formation is 1527 ± 182 mm, and two Sapa paleosol received 1482 ± 182 and 1507 ± 182 mm (Fig. 9D), much more humid than modern Concordia, Kansas (740 mm average from 1964 to 1993: Wood, 1996). Deep weathering and kaolinitic composition of Dakota Formation paleosols are evidence of warm climates (Retallack and Dilcher, 1981a,b; Joeckel, 1991), and of paleoclimatic cooling in the smectitic upper Dakota Formation and overlying formations (Austin, 1970). A useful paleotemperature proxy for paleosols devised by Sheldon et al. (2002) uses alkali index (N = (K2O + Na2O)/Al2O3 as a molar ratio), which is related to mean annual temperature (T in °C) in modern soils by Eq. (6) (R 2 = 0.37; S.E. = ± 4.4 °C). T ¼ −18:5 N þ 17:3

ð6Þ

Estimated mean annual temperature using Eq. (6) for a Sayela paleosol of the Dakota Formation is 17.0 ± 4.4 °C, and for two Sapa paleosols 14.2 ± 4.4 and 14.6 ± 4.4 °C (Fig. 9C), thus quantifying the temperature decline suspected on the basis of clay mineralogy by Austin (1970). All estimates of paleotemperature are warmer than modern Concordia, Kansas (11.8 °C average from 1964 to 1993: Wood, 1996). The basal Dakota Formation thus enjoyed a subtropical humid climate, and the upper Dakota Formation a warm temperate humid climate in the Köppen system (Trewartha, 1982). There is no evidence from paleosols of frost heave or other periglacial structures to support inferences of glaciation from paleochannel morphology in the Dakota Formation (Koch et al., 2007). Paleosols of the Dakota Formation contain abundant charcoal, even in wetland Makizi and Cahli pedotypes (Fig. 2), as an indication of a dry season of wildfires. Wetland wildfires of the Everglades of Florida are generally in the driest months of December to February (Smith et al., 2001; Winkler et al., 2001). Dakota Formation paleosols do have some pedogenic slickensides and some cracks, but they are not organized into lentil peds or mukkara structure like those of the Vertisols in the modern Gulf coastal plain of Texas, where evaporation exceeds precipitation for 5 months of the year (Nordt et al., 2004). These paleoclimatic estimates from paleosols are in broad agreement with those from isotopic composition of pedogenic sphaerosiderite. Two analyzed sets of sphaerosiderite have δ 18O of −4.7‰ ± 0.2‰ (SCB4B-119.6) and −2.9‰ ± 0.5‰ (SCB4A-139) from the middle Dakota Formation in cores at Sergeant Bluff, Iowa (Ludvigson et al., 1998) and of −5.1‰ ± 0.3‰ (8.6 m) and − 4.1‰ ± 0.1‰ (7.6 m) at the Rose Creek locality near Fairbury Nebraska, and modeled to indicate mean annual precipitation of 1800 ± 1000 mm (Ufnar et al., 2008a). Fossil plants of the Dakota Formation are also evidence of warm, seasonally wet paleoclimate. Angiosperm dominance of swamp paleosols (Cahli pedotype) and intertidal paleosols (Makizi: Retallack and Dilcher, 1981a; Upchurch and Dilcher, 1990) is now only found in central America south of the United States, where intertidal vegetation is salt marsh rather than mangal (Avicennia germinans), and

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

55

Fig. 7. Reconstructed sedimentary setting, marine communities and pedotypes of the upper Dakota Formation in Kansas.

swamp woodlands are dominated by cypress (Taxodium distichum) rather than angiosperms (Hartshorn, 1983; Barbour and Billings, 1988). This indication of warm paleoclimate using modern analogs of swamp and mangal angiosperms must be viewed with caution considering the early stage of angiosperm evolution and global migrations represented by the Dakota flora (Retallack and Dilcher, 1986; Field et al., 2011). The unusually diverse Fort Harker fossil leaf assemblage of the middle Dakota Formation has been taken as physiognomic evidence of mean annual temperature of 24 °C, because of its common large, entire margined leaves (Upchurch and Wolfe, 1987). Mean annual temperature estimated from leaf margins and mean annual precipitation from leaf size for successive localities (after Sloan, 1964; Setterholm, 1994; Gröcke et al., 2006) in the upper Dakota Formation have been estimated as 21.1 ± 3 °C and 990 mm for Courtland in southeastern Minnesota, 18.6 ± 3 °C and 840 mm for Fairbury in Nebraska, 14.9 ± 3 °C and 790 mm for Delphos, and 18.3 °C and 1270 mm for Hoisington (Dilcher et al., 2005; errors from Peppe et al., 2010).

These lines of evidence demonstrate an unusual spike of warm– wet conditions in the lower Dakota Sandstone, which is also evident from the percentage of base-poor clay inferred by Macfarlane et al. (1989) from photoelectric absorption and density logs of Braun #1 core (Fig. 9B). Base-poor clay (kaolinite rather than smectite) is inferred from a cross plot of apparent matrix density (RHOMAA in g cm − 3) and apparent matrix photoelectric absorption (UMAA in barns cm − 3), because other minerals of comparable physical character such as muscovite and biotite are rare in point counts (Fig. 3). A thick interval of dominantly kaolinitic clay (Fig. 9B) is at the same level as Sayela paleosols in the field (Fig. 2). Thick kaolinitic paleosols are widespread in the “Dakota Formation” (see discussion of Witzke and Ludvigson, 1994) in Utah, where a recent paleoclimatic study of calcareous paleosols (Retallack, 2009) had to go west as far as the Baseline Sandstone and Willow Tank Formation of Nevada to obtain a pedocal record through that time interval (Fig. 9A). Comparable latest Albian–Cenomanian anomalously deep weathering also has been recorded in paleosols of New Mexico (Leopold, 1943; Mack, 1991), and Minnesota (Goldich, 1938; Sloan, 1964; Austin, 1970).

56

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

Fig. 8. Reconstructed biota of stream margins (Casmu pedotype) of the Dakota Formation in Kansas.

Triplehorn's (1971) objection that the change from deep weathering inferred by Austin (1970) was due to comparison of paleosol with shale is unjust, because paleosols like the Sapa pedotype are common in the upper Dakota Formation of Minnesota as well. This Albian– Cenomanian boundary (99 Ma) warm–wet paleoclimatic spike corresponds with the Breitstroffer oceanic anoxic event (Gradstein et al., 2004) and with a marked perturbation in isotopic balance of the carbon cycle (Gröcke et al., 2006). The basal Turonian Bonarelli oceanic anoxic event (93 Ma), was a similar warm–wet climatic spike, associated with a spike in atmospheric CO2 spike estimated from ginkgo (Retallack, 2009) and laurel leaf stomatal index (Barclay et al., 2010), but at a stratigraphic level above the Dakota Formation and Graneros Shale in Kansas (Austin, 1970; Hattin et al., 1978).

Large (6 m) nodules in paleochannel sandstones (Fig. 10) at the same stratigraphic level as deeply weathered Sayela paleosols and inferred warm–wet paleoclimatic spike is additional support for the conclusion of Swineford (1947) based on facies distribution that these exceptionally large nodules formed shortly after deposition. Unusually warm–wet climate could explain both the mobilization of large amounts of Ca and its precipitation within a deep weathering zone of a soil, as an alternative to the deep diagenesis model of McBride and Milliken (2006). Their chemical data is strongly suggestive of a role for meteoric water and soil formation: displacive carbonate (requiring low confining pressure), calcite replacement of feldspar (open system alteration), calcite 87/86Sr values of 0.707739 to 0.708700 (meteoric), calcite δ 13C of − 36‰ to −4‰ (methanogenic to meteoric), calcite δ 18 O of −13‰ to −6‰ (meteoric), and pyrite δ 34S of +8.5‰ and 12‰ (microbial sulfate reduction). Furthermore a cross plot of 60 analyses of calcite δ 13C and δ 18O (by McBride and Milliken, 2006) shows a significant linear trend (R 2 = 0.61, t test p b 0.001) with a high slope (m = 0.4), as found in pedogenic carbonate in a highly evaporative warm climate, as opposed to marine and deep diagenetic carbonate (Ufnar et al., 2008b). This unusual giant nodule horizon may be another consequence of unusual paleoclimatic perturbation. 5.2. Past vegetation Paleosols of the Dakota Formation are themselves evidence of plant formations. Pyritic carbonaceous claystones with abundant pyrite nodules (Makizi pedotype) and brackish–marine bivalves in growth position (Fig. 11A) are distinctive paleosols of mangal

Fig. 9. Paleoclimatic time series from calcareous paleosols of Utah and Nevada (A, from Retallack, 2009), compared with proportion of base poor clay inferred a cross plot of apparent matrix density (RHOMAA in g cm− 3) and apparent matrix photoelectric absorption (UMAA in barns cm− 3) values (B: after Macfarlane et al., 1989); and paleotemperature and paleoprecipitation (C–D) inferred from paleosols (solid squares, herein) and fossil plants (open diamonds, Upchurch and Wolfe, 1987; Dilcher et al., 2005).

Fig. 10. Giant subspherical nodules in lower Dakota Formation sandstones of Kansas, at Mushroom State Park, near Carneiro (A) and Rock City Park, near Minneapolis (B).

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

57

Fig. 11. Fossil preservation in paleosols, including (A) mussel (Brachidontes filiscuptus) in life position within a clayey mangal paleosol (Makizi pedotype) from Fairbury, and (B–E) fine preservation of fossil plants within ferruginized sandstones (Sagi pedotype) near Ellsworth, including inflorescences (B) of Eoplatanus serrata, seed capsule in side (D) and superior (C) view of Nordenskioldia borealis, and leaves of Araliopsoides cretacea and Brasenites kansense (E). Depositories and specimen numbers are Florida Museum of Natural History 15713–3313 (A), Smithsonian Institution Museum of Natural History 5005A (B), University of Kansas Natural History Museum 11043 (C–D) and 2951 (E).

communities, the pyrite derived from microbial reduction of marine sulfate (Altschuler et al., 1983). Sparsely pyritic coals with large woody roots in their underclay (Cahli pedotype) represent freshwater swamp woodland communities, as are well known in coal measures (Greb et al., 2006). Thick paleosols with clay enriched subsurface horizons (Ultisols) form over long periods of time under old-growth forest communities (Markewich et al., 1990). These well developed paleosols would have been well drained in the case of deeply rooted and oxidized Sayela pedotype, which are not so highly oxidized or ferruginized to qualify as laterites (Retallack, 2010) with which they have been compared (Thorp and Reed, 1949). Pedogenic sphaerosiderite (Ludvigson et al., 1998; Ufnar et al., 2008a) of carbonaceous Sapa pedotype paleosols is known to form in waterlogged soils of coastal marshes (Pye, 1981). Weakly developed paleosols in which rooting and burrowing have not proceeded far enough to destroy primary bedding features are comparable with soils supporting vegetation early in ecological succession after disturbance by flooding (McKenzie et al., 2004), and also come in a variety of redox conditions like comparable wetland soils today (Vepraskas and Sprecher, 1997). Sandy paleosols ferruginized in outcrop and in drill cores (Sagi pedotype) are comparable with modern freely drained soils, as indicated also by their deep patterns of rooting. Carbonaceous sandy and clayey Casmu pedotypes have siderite nodules of seasonally waterlogged soils (Pye, 1981), whereas gray shaley Kazela pedotype has lost little organic matter to aerobic oxidation. These three pedotypes can thus be interpreted to represent well drained levee top riparian woodland (Sagi), swale woodland (Casmu) and pond vegetation (Kazela). These different kinds of paleosols preserve floristically distinct fossil plant assemblages, and thus provide a guide to autochthonous plant communities. Some fossil plant localities (Hoisington, Courtland and Delphos: Wang, 2002; Wang and Dilcher, 2006; Wang et al., 2011) are laminated lake deposits in which plant remains from a

variety of distinct communities have been mixed. Other localities include subautochthonous plant remains close to growth position, such as mangal vegetation dominated by Pabiania variloba and Pandemophyllum kvacekii in Makizi pedotype paleosols at Fairbury (Upchurch and Dilcher, 1990), and swamp woodland vegetation of Liriophyllum kansense and Sapindopsis bagleyae in underclay of the Cahli pedotype at Bunker Hill east (Retallack and Dilcher, 1981a,b). These two localities also had a prominent understory component of ferns (Skog and Dilcher, 1994). Sandstone floras of the Dakota Formation come in two varieties: those in ferruginized siderite nodules (Casmu pedotype at Sternberg localities of Fort Harker illustrated by Rogers, 1991, and Hoisington specimens F111088–11091 in Condon Collection, University of Oregon) and those in ferruginized sandstone (Sagi pedotype: F111083–111087 from Hoisington in Condon Collection, and other historic museum specimens of Fig. 11). Both these streamside early successional vegetation types were dominated by Araliopsoides cretacea. A Kazela pedotype paleosol at Carneiro also yielded a small collection including redwood (Sequoia condita), epiphytic lycopsid (Onychiopsis goepperti), and sycamore (Eoplatanus serrata: F111092–F111098 Condon Collection). Conifers such as redwood are found in lake deposits of the Dakota Formation as well, and dominate its pollen (Pierce, 1961; Ravn and Witzke, 1994, 1995). This may be in part because the conifers were wind pollinated, whereas most angiosperm pollen types were insect pollinated (Hu et al., 2008b), but it is also likely that pedotypes for which fossil floras are unknown (Sayela and Sapa pedoptypes) supported coniferous forests (Retallack and Dilcher, 1981a,b). These inferences from paleosols and their distinct plant communities are compatible with what has already been inferred from fossil plants of the Dakota Formation. Large leaves, drip tips and vine like leaf shapes of the Fort Harker leaf assemblage have been interpreted as closed canopy multistratal rain forest (Upchurch and Wolfe, 1987),

58

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

and these authors were prescient in noting that this sandstone flora (of Casmu pedotype at stratigraphic level of Sayela pedotype) may have been of short duration, anticipating discovery here of the basal Cenomanian warm–wet paleoclimatic spike (Fig. 9). Upper Dakota Formation floras are slightly less diverse (87 species, rather than 93 for the Fort Harker sandstone floras), and have variety of features of temperate deciduous woodlands: moderate size leaves with limited insect and fungal damage, and stipules or other protection for leaf buds (Wang, 2002). These floras were a marked advance in diversity and modernization of early angiosperms associated with extensive Early Cretaceous coastal migrations and paleoclimatic changes (Retallack and Dilcher, 1981a, 1986; Wang et al., 2009; Field et al., 2011; Friis et al., 2011). 5.3. Past animal life The crocodile (Dakotasuchus kingi) and nodosaur (Silvisaurus condrayi) were both found in large ellipsoidal siderite nodules (Mehl, 1941; Vaughn, 1956; Eaton, 1960), characteristic of Casmu pedotype paleosols, so were a part of the early successional streamside community of plants noted above from nodules within that pedotype (Fig. 8). The alkaline geochemical environment required by siderite nodules may have been important for preservation of bone, otherwise dissolved by acidic deep weathering that has stripped most of the Dakota Formation of calcite, including the shells of bivalves (Fig. 11A). A variety of marine trace fossils have been described from the uppermost parts of the Dakota Formation (Siemers, 1976; Hattin et al., 1978), but few trace fossils were found in the intermittently waterlogged paleosols of the lower parts of the formation. 5.4. Paleotopography

Fig. 12. Log of Braun #1 well through the Dakota, Kiowa and Cheyenne Formations, showing variation in lithological composition (A), and proxies for paleosol development (B), calcareousness (C) and hue (D). Data is from Macfarlane et al. (1989, 1990).

Mangal to freshwater swamp paleosols of the Dakota Formation noted above are additional evidence for the fluvial to linear clastic shoreline sedimentary setting inferred for the formation from evidence of sedimentary facies and structures (Siemers, 1976; Hattin et al., 1978). Even the best drained pedotypes have oxidation (Fig. 5A), sepic plasmic fabrics (Fig. 6C–D), clay skins and root traces reaching down no more than 2 m, to levels of pyrite, sphaerosiderite, coal or other indications of a stagnant water table. Subdued paleotopography of alluvial valley landscapes is preserved at the base of both D and J units of the Dakota Formation preserved in boreholes in Kansas (Fig. 1: Hamilton, 1994) and Iowa (Witzke and Ludvigson, 1994). In Illinois and Minnesota the low hills of Precambrian granite and quartzite and Paleozoic limestone emerge from beneath cover of Dakota Formation (Frye et al., 1964; Sloan, 1964), although softened by thick paleosols from intense chemical weathering (Goldich, 1938; Austin, 1970; Triplehorn, 1971).

The best developed paleosols in the Dakota Formation are thick kaolinitic Ultisols comparable with those of the modern coastal plain of the southeastern United States (Markewich et al., 1990), where strongly developed soils of increasing age (A in ka) show the following increases in depth of solum (Ds in cm) and depth of argillic horizon (Dt in cm) with R 2 = 0.79 and 0.80, and standard errors on age of 140 and 137 ka, respectively.

5.5. Parent material

A ¼ 4915:2Ds −343466

ð7Þ

Source terranes for the Dakota Formation were low hills of Precambrian quartzites, schists and granite, with a cover of Paleozoic quartz sandstone and limestone. Very few rock fragments or limestone clasts, and little feldspar were delivered to the quartz-rich sands (Witzke and Ludvigson, 1994) that formed parent materials to Dakota Formation sandy paleosols (Fig. 3). This itself is evidence of profound weathering of the source terrane, which is also apparent in the dominantly kaolinitic composition of shales and claystones of the lower and middle Dakota Formation ((Plummer and Romary, 1947; Plummer et al., 1960; Frye et al., 1964; Parham and Hogberg, 1964; Parham, 1970; Joeckel, 1987, 1991). Smectite is abundant near the top of the Dakota Formation at the transition to the marine Graneros Shale (Fig. 12), but the clay is not entirely of marine origin. A bentonite seen at Black Wolf (Fig. 2) and others within overlying marine rocks (Hattin et al., 1978) are

A ¼ 4844:4Dt −196090

ð8Þ

evidence that smectite clays were derived from alteration of volcanic ash derived from the western Cordillera (Dyman et al., 1994). Within thick (Sapa and Sayela) profiles, smectite and illite were weathered to kaolinite, as reflected by profound base depletion (Figs. 3 and 4). 5.6. Time for formation

For application to paleosols, observed solum and argillic depths of Sayela (60 and 40 cm) and Sapa pedotypes (110–160 cm and 90–100 cm respectively) need to be corrected for compaction using Eq. (1), above. These calculations reveal that these Cretaceous paleosols compare in age with mid-Pleistocene soils of the Georgia coastal plain (Markewich et al., 1990): 25 ± 140 ka from solum and 107 ± 137 ka from argillic horizons of a Sayela paleosol, and 332–640 ± 140 for solum and 112–138 ± 137 ka from argillic of several Sapa paleosols. The shorter time of development of Sayela paleosols, yet greater intensity of chemical weathering (Fig. 4) is another indication of the basal Cenonamian paleoclimatic spike. Coaly paleosols (Histosols like Cahli) are also indicators of durations because growth of woody plants in precursor peats is curtailed at subsidence rates of

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

more than 1 mm per year, and peat cannot survive aerobic decay at subsidence rates of less than 0.5 mm per year (Retallack and Jahren, 2008). Using known compaction of peat to coal (Eq. (2)), gives times of formation of 2.8 to 12.8 ka for Cahli paleosols with organic horizons now 20–45 cm thick. In coastal plain and alluvial settings paleosols with the degree of development seen in Casmu, Sagi and Kazela pedotypes have historical artifacts indicating durations of formation from decades to centuries (McKenzie et al., 2004), but those with large (5–10 cm) siderite or pyrite nodules (Casmu and Makizi) may have taken millennia to form, like modern coastal soils (Pye, 1981). While these weakly developed paleosols are common in the formation, the abundance of moderately to strongly developed paleosols (Fig. 2) is an indication of relatively low sediment accumulation rates on a tectonically stable low relief coast that was well vegetated. 6. Core-outcrop comparisons of Dakota Formation paleosols Comparison of outcrops with cores of paleosols can be useful for understanding unusual modes of fossil preservation, overprints of modern upon ancient weathering, and additional sources of geophysical data on paleosols. 6.1. Leaves preserved in red sandstone? A paradox presented by the 19th century fossil leaf floras of the Dakota Formation collected by the Sternberg family is their superb preservation in coarse ferruginized sandstone or red siltstone nodules (Rogers, 1991). Many leaves are wrinkled and partly decayed as if dried and curled on a leaf litter (Fig. 11E). Thin leaves of aquatic plants show less relief, but superbly detailed venation (Fig. 11E upper). Delicate three dimensional structures such as fruit capsules (Fig. 11C–D) and seed heads (Fig. 11B) appear uncompressed by burial, whereas formerly calcareous shells of bivalves preserved in growth position within clayey mangal paleosols (Makizi pedotype) are crushed flat in life position (Fig. 11A). Observations of such fossils from a Casmu and Sagi paleosol in Hoisington brick pit, and from Jones #1 core, show that the red color of the nodules is a surficial oxidation product of dark gray siderite nodules seen in core. In contrast, the red color of leaf-bearing sandstones is found also in core, and was red and oxidized before burial in the Cretaceous. In the fossil preservation terminology of Schopf (1975), the nodules are a form of chemically reducing authigenic cementation and the red sandstone fossils are oxidized impressions. Authigenic cementation is best known in the Francis Creek Shale (Middle Pennsylvanian) in that region of Illinois strip mines best known as Mazon Creek (Shabica and Hay, 1997). The mechanism of preservation of plants, fish, crabs and other fossils is decay, especially deamination, of the fossil itself, creating in surrounding mud an alkaline and reducing geochemical halo that promoted early diagenetic precipitation of siderite. Unlike hard durable siderite nodules with leaves found at Hoisington, the Dakota Formation leaves collected from Fort Harker by the Sternbergs (Rogers, 1991) are preserved in porous red silty nodules, which have been oxidized in outcrop, by oxidative dissolution of siderite to hematite. Ferruginized siderite nodules do not explain the high fidelity, red, leaf impressions in Sagi paleosols, which may be “microbial death masks” in the sense of Gehling (1999), but not of pyrite. In the Dakota Formation, pyrite forms massive nodules and root encrustations in mangal paleosols (Fig. 4D), but does not form surficial “death masks”. Pyrite framboids are ubiquitous within pyritized fossils, but not on their surfaces which retain original ribbing and striae (Grimes et al., 2002). Pyrite framboids leave characteristic clustered cavities when oxidized (Altschuler et al., 1983), not seen on close examination of Dakota Formation red leaves (Spicer, 1977). A predepositional death mask of goethite (ferric hydroxide) was proposed for such leaves by Spicer (1977), based on observations of rusty coatings

59

of modern leaves created by Sphaerotilus natans. These filamentous proteobacteria settle on surfaces by entanglement and by an adhesive base (Pellegrin et al., 1999). Sphaerotilus, and the similar Leptothrix are heterotrophic sludge thickeners, and thus fueled by partial leaf decay. Although widely considered as iron oxidizing bacteria, the necessity of iron for metabolism of Sphaerotilus and Leptothrix is controversial (van Veen et al., 1978). Ferric hydroxide encrustations on modern leaves observed by Spicer (1977) were only 2–3 μm thick, so were not responsible for observed differences in relief of aquatic leaves (Fig. 11E upper), densely veined leaves (Fig. 11E lower), and of woody reproductive structures (Fig. 11B-D). These observations are evidence that resistance to burial compaction was conferred by remnant refractory lignin rather than iron-hydroxide death masks which preserve surficial detail. Other fine-grained impressions in red sandstones for which this preservational mechanism is appropriate include Evolsonia from the Permian of Texas (Mamay, 1989), Sanmiguelia from the Triassic of Colorado (Tidwell et al., 1977), and Dickinsonia from the Ediacaran of South Australia (Retallack, 2007). 6.2. Jarosite and gypsum in coaly paleosols? Many Dakota Formation coals and carbonaceous shales in outcrop (especially Makizi pedotype) are coated with powdery masses of the yellow, rotten-egg-smelling mineral jarosite [KFe3+ 3(OH)6(SO4)2] and littered with clear crystals of gypsum (CaSO4), but neither mineral was seen in cores. Both are common soil minerals, jarosite in “cat clay” supratidal soils and mine tailings (Doner and Lynn, 1977) and gypsum in desert soils, where it forms replacive nodules and sand crystals including much pre-existing soil matrix (Retallack and Huang, 2010). These crystal forms are quite distinct from the clear selenite crystals found in Kansas (Fig. 5F), where paleosols of the Dakota Formation are evidence of much more humid climate than for modern gypsic soils (Table 2). Jarosite is found in close association with pyrite and carbonaceous root traces and claystone (Fig. 5D), but both jarosite and gypsum crystals were found littering marine shales of the Graneros Formation as well. The acidophilic and thermophilic archaebacterium Acidianus brierleyi and proteobacterium Thiobacillus ferrooxidans are known to oxidize pyrite to sulphuric acid or jarosite (Larsson et al., 1990; Olsson et al., 1995). Neutralization of this acid by calcite from limestone or mollusc shells (Fig. 11A) produces gypsum in modern coastal soils (Doner and Lynn, 1977). Both jarosite and gypsum are assumed to be modifications of these paleosols in outcrop until they can be demonstrated in core. 6.3. Hematite in carbonaceous paleosols? A surprising discovery from examination of Dakota Formation cores was that the Sapa paleosols with scattered red buckshot coloration in outcrop (Fig. 5C,G) were entirely gray in core. Sayela, like Sagi paleosols, were red in both core and outcrop, as noted for comparable stratigraphic levels in cores from Iowa (Witzke and Ludvigson, 1994). In thin section the red coloration of Sapa paleosols can be seen to be thin weathering rinds on individual sand-size spheroidal aggregates of sphaerosiderite, formed by oxidative dissolution of siderite to hematite in outcrop. The formation of sphaerosiderite during microbial respiration of these paleosols was thus remote from oxidizing surface conditions, as in modern coastal swamp soils (Pye, 1981). This result is compatible with methanogenic carbon isotopic compositions found in some of the Dakota Formation sphaerosiderites (Ludvigson et al., 1998). 6.4. Geophysical signals of paleosols in core? Projects to characterize the Dakota Formation aquifer in Kansas have generated a large amount of geophysical borehole data (Macfarlane et al., 1990), some of which is helpful in interpreting these mid-

60

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

Cretaceous paleosols. For example, degree of pedogenic development, calcareousness and hue are key variables in paleosol interpretation (Fig. 2), and can potentially be inferred from geophysical logs (Fig. 12). A general lithological log of the formation can be obtained from matrix density (RHOMAA in g cm − 3) and matrix z-number inferred from photoelectric absorption (UMAA in barns cm − 3), which in cross-plot arrays base-poor clays (kaolinite), base-rich clays (illite, smectite), and light minerals (quartz, feldspar) in different fields (Macfarlane et al., 1989). These have been converted to a proportional log of the three components using a matrix algebra computer algorithm by Doveton (1986). These data (Fig. 12A) show proportions of well developed clayey paleosols and of sandy sediment with weakly developed paleosols through the Braun #1 core. Spikes in base-poor clay are evidence of deeply-weathered kaolinitic paleosols, and elevated abundance of quartz and feldspar reveals local paleochannel and levee facies. A suitable proxy for paleosol development in the Dakota Formation is photoelectric index (Pe): a proxy for atomic number (z) of elements in the matrix, which in shales is effectively a proxy for iron abundance (Ellis, 1987). This proxy thus quantifies density of sphaerosiderite in Sapa paleosols and occasional thin hematitic Sagi paleosols in the upper Dakota Formation, and density of hematitic ferruginization in Sayela paleosols of the middle Dakota Formation. There is less evidence of well developed paleosols in the sandy lower Dakota Formation (Fig. 12B). A suitable proxy for paleosol pH or calcareousness, for which the field test is acid reaction (Fig. 2), is apparent matrix photoelectric absorption (UMAA), calculated from photoelectric index (Pe) and neutron density. The values of known local calcareous (Greenhorn Limestone) and sandy rock units (Longford Member of lower Kiowa Formation) divided limestones from clastics at a value of 9.6 barns cm − 3 (Macfarlane et al., 1990). These data reveal that the Braun #1 core (Fig. 11B) like the outcrop of the Dakota Formation is calcite poor, with exception of nodules in paleochannel sandstones (McBride and Milliken, 2006) and rare molluscan fossils (White, 1894; Siemers, 1976). A suitable proxy for redox status of paleosols, for which the field index is Munsell hue (Fig. 2), is Th/U ratio, derived from measurements of natural gamma ray radiation filtered to obtain ppm values for these different radioactive elements (Macfarlane et al., 1989). Values of this ratio of less than 2 reflect leaching of U under chemically reducing conditions, but values of 7 or more result from fixation of U under chemically oxidizing conditions (Adams and Weaver, 1958). These values for the Braun #1 core (Fig. 11D) show the expected highly oxidized level of Sayela paleosols atop sandstones of the middle Dakota Formation. Three other zones of oxidation in the upper Dakota Formation are non-calcareous, sandy, and iron poor, so reflect redeposited soil oxides in paleochannels. Critical contributions of these geophysical logs include demonstration that most of the clayey upper Dakota Formation is chemically reduced (Fig. 12D), as physical inspection of the cores confirms. The oxidation in outcrop of the upper Dakota Formation is thus a misleading indication of paleodrainage of these paleosols. In addition, the logs confirm that base-poor clay is most abundant at one stratigraphic level of Sayela paleosols in the middle Dakota Formation (Figs. 9B and 12A).

(Retallack, 2009) and fossil plants of the Dakota Formation (Upchurch and Wolfe, 1987; Dilcher et al., 2005). A variety of mangal, coastal swamp, and riparian communities dominated by angiosperms and upland communities of conifers are confirmed by different kinds of paleosols in the upper Dakota Formation (Retallack and Dilcher, 1981a,b). Strong development of many paleosols in the Dakota Formation may reflect formation on a tectonically stable and well vegetated coastal plain, comparable with modern lowland Georgia, U.S.A. (Markewich et al., 1990). Examination of paleosols of the Dakota Formation in drill core reveals that some features of the paleosols that might otherwise be considered significant to paleoenviromental interpretation have been compromised by alteration in outcrop. Oxidative weathering of paleosols in outcrop has created red color and ferruginous weathering rinds of siderite nodules and spherulites (sphaerosiderite) not seen in the cores. Oxidation of pyrite to jarosite and formation of gypsum by reaction of sulfuric acid and calcite also are restricted to outcrops, and not found in the cores. Nevertheless, siderite and pyrite are original features of the paleosols, as in comparable modern freshwater swamp and mangal (Pye, 1981; Ludvigson et al., 1998). Fossil leaves in red nodules and sandstones from the Dakota Formation widely distributed in fossil collections around the world from the 19th century activities of the Sternbergs (Rogers, 1991) are of two kinds. Some were oxidized in modern outcrop from siderite nodules seen in cores and deep clay pits: a style of preservation comparable with that better known from Pennsylvanian fossils of Mazon Creek, Illinois (Shabica and Hay, 1997). Other fine impressions of leaves and fruits in red sandstones were preserved by microbial death masks of ferric hydroxide, not pyrite (Spicer, 1977). Drill cores are not only valuable because they provide convenient vertical sections of paleosols, but proxies for many pedogenically significant variables can be obtained from geophysical logs of boreholes. Iron content as an index of paleosol development for example can be quantified by photoelectric index. Carbonate content as an index of paleosol pH can be quantified by apparent matrix photoelectric absorption (UMAA), calculated from photoelectric index (Pe) and neutron density. Finally, redox status of paleosols can be quantified by Th/U ratio, derived from measurements of natural gamma ray radiation (Macfarlane et al., 1989). Such remote sensing of paleosols can confirm field observations, but does not supplant the need for ground truth provided by drill core and outcrops. Core and geophysical logs thus demonstrate that not all paleosols of the Dakota Formation were oxidized or deeply weathered. A shortlived deep-weathering spike is represented by Sayela paleosols in the middle Dakota Formation of Kansas, and was continent-wide judging from comparable kaolinitic paleosols in Utah (Witzke and Ludvigson, 1994), Nevada (Retallack, 2009), New Mexico (Leopold, 1943; Mack, 1991), and Minnesota (Goldich, 1938; Sloan, 1964; Austin, 1970). This Albian–Cenomanian boundary (99 Ma) warm–wet paleoclimatic spike corresponds with the Breitstroffer oceanic anoxic event (Gradstein et al., 2004), with a marked perturbation in isotopic balance of the carbon cycle (Gröcke et al., 2006), and a significant adaptive radiation and modernization of angiosperms (Wang et al., 2009; Field et al., 2011).

7. Conclusions

This project was initiated and partly funded by David Dilcher with funds provided by NSF (BMS75-02268, DEB77-04846, DEB7910720, BSR83-00476, BSR 84-01148, BSR86-16657) and Kansas Geological Survey grant 41-246-02. Roger Pabian, Bill Crepet, Howard Reynolds, Stephen Manchester, Mike Zavada, Carolyn Bagley, Tom Halter, Terry Lott, Judith Skog, Peter Dilcher, Charles Beeker, Karl Longstreth, Bill Macklin, Bob Schwartzwalder, Chuck Huang, Warren Kovach, Hongshan Wang, Xin Wang, Shusheng Hu, Ben Sanders and Frank Potter aided with fieldwork. John Doveton helped interpret

The Dakota Formation of Kansas includes common paleosols with well preserved root traces, soil horizons and soil structures (Retallack, 1997). The variety of paleosol types can be classified into pedotypes which reveal both spatial and temporal variations in paleoenvironments. Evidence from paleosols confirms that the Albian–Cenomanian boundary spike of paleoclimate was warmer and wetter than later in the Dakota Formation, as also inferred from regional paleosol studies

Acknowledgments

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

geophysical logs. Alan Wade, P. Allen Macfarlane, and Janet Coleman allowed access to cores of the Kansas Geological Survey. References Adams, J.A.S., Weaver, C.E., 1958. Thorium to uranium ratios as indications of sedimentary processes: example of concept of geochemical facies. American Association of Petroleum Geologists Bulletin 42, 387–430. Altschuler, Z.S., Schnepfe, M.M., Silber, C.C., Simon, E.O., 1983. Sulfur diagenesis in Everglades peat and the origin of pyrite in coal. Science 221, 221–227. Austin, G.S., 1970. Weathering of the Sioux Quartzite near New Ulm, Minnesota, as related to Cretaceous climates. Journal of Sedimentary Petrology 40, 184–193. Barbour, E., 1931. Evidence of dinosaurs in Nebraska. Nebraska State Museum Bulletin 1, 187–190. Barbour, M.G., Billings, W.D., 1988. North American Terrestrial Vegetation. Cambridge University Press, New York. 432 pp. Barclay, R.S., McElwain, J.C., Sageman, B.B., 2010. Carbon sequestration activated by volcanic CO2 pulse during Ocean Anoxic Event 2. Nature Geoscience 3, 205–208. Basinger, J.F., Dilcher, D.L., 1984. Ancient bisexual flowers. Science 224, 511–513. Brenner, R.L., Bretz, R.F., Bunker, B.J., Iles, D.L., Ludvigson, G.A., McKay, R.M., Whitley, D.L., Witzke, B.J., 1981. Road log and stop descriptions. In: Brenner, R.L. (Ed.), Cretaceous Stratigraphy and Sedimentation in Northwest Iowa and Northeast Nebraska, and Southeast South Dakota: Iowa Geological Survey Guidebook, 4, pp. 149–172. Brewer, R., 1976. Fabric and Mineral Analysis of Soils. Krieger, New York. 482 pp. Brimhall, G.H., Chadwick, O.A., Lewis, G.J., Comptson, W., Williams, I.-S., Danti, K.J., Dietrich, W.E., Power, M.E., Hendricks, D., Bratt, J., 1991. Deformational mass transport and invasive processes in soil evolution. Science 255, 695–702. Buechel, E., 1970. A Dictionary of the Teton Dakota Sioux Language. Red Cloud Indian School, Pine Ridge. 852 pp. Dilcher, D.L., 1979. Early angiosperm reproduction: an introductory report. Review of Palaeobotany and Palynology 27, 291–328. Dilcher, D.L., Crane, P.R., 1984. Archaeanthus; an early angiosperm from the Cenomanian of the Western Interior of North America. Missouri Botanical Garden Annals 71, 351–383. Dilcher, D.L., Kovach, W.L., 1986. Early angiosperm reproduction; Caloda delevoryana gen. et sp. nov., a new fructification from the Dakota Formation (Cenomanian) of Kansas. American Journal of Botany 73, 1230–1237. Dilcher, D.L., Wang, H., 2009. An Early Cretaceous fruit with affinities to Ceratophyllaceae. American Journal of Botany 96, 2256–2269. Dilcher, D.L., Crepet, W.L., Beeker, C.D., Reynolds, H.C., 1976. Reproductive and vegetative morphology of a Cretaceous angiosperm. Science 191, 854–856. Dilcher, D.L., Wang, H., Kowalski, E., 2005. Dakota Formation paleoclimate estimates from late Albian angiosperm leaves. Geological Society of America Abstracts 37 (5), 85–86. Doner, H.E., Lynn, W.C., 1977. Carbonate, halide, sulfate, and sulfide minerals. In: Dixon, J.B., Weed, S.B. (Eds.), Minerals in Soil Environments. Soil Science Society of America, Madison, pp. 75–98. Doveton, J.H., 1986. Log Analysis of Subsurface Geology—Concepts and Computer Methods. Wiley Interscience, New York. 273 pp. Dyman, T.S., Cobban, W.A., Fox, J.E., Hammond, R.H., Nichols, D.J., Perry, W.J., Porter, K.W., Rice, D.D., Setterholm, D.R., Shurr, G.W., Tysdal, R.G., Haley, J.C., Campen, E.B., 1994. Cretaceous rocks from southwestern Montana to southwestern Minnesota, northern Rocky Mountains, and great plains region. In: Shurr, G.W., Ludvigson, G.A., Hammond, R.H. (Eds.), Perspectives on the Eastern Margin of the Cretaceous Western Interior Basin: Geological Society of America Special Paper, 287, pp. 5–26. Eaton, T.H., 1960. A new armored dinosaur from the Cretaceous of Kansas. University of Kansas Paleontological Contributions 8, 1–24. Eicher, D.L., 1965. Foraminifera and biostratigraphy of the Graneros Shale. Journal of Paleontology 39, 875–909. Ellis, D.V., 1987. Well Logging for Earth scientists. Elsevier, New York. 532 pp. Everhart, M.J., Everhart, P.A., Ewell, K., 2005. A marine ichthyofauna from the upper Dakota Sandstone (Late Cretaceous). Kansas Academy of Sciences Transactions 108, 71. Farley, M.B., Dilcher, D.L., 1986. Correlation between miospores and depositional environments of the Dakota Formation (mid-Cretaceous) of north-central Kansas and adjacent Nebraska. Palynology 10, 117–133. Faulds, J.E., Ludvigson, G.A., Brenner, R.L., 1996. Preliminary paleomagnetic study of plinthic paleosols of the Dakota Formation. In: Witzke, B.J., Ludvigson, G.A. (Eds.), Mid-Cretaceous Fluvial Deposits of the Eastern Margin, Western Interior Sea: Nishnabotna Member, Dakota Formation: Iowa Geological Survey Guidebook, 77, pp. 39–42. Field, T.S., Brodribb, T.J., Iglesias, A., Chatelet, D.S., Baresch, A., Upchurch, G.A., Gomez, B., Mohr, B.A.R., Coiffard, C., Kvacek, J., Jaramillo, C., 2011. Fossil evidence for Cretaceous escalation in angiosperm leaf vein evolution. U.S. National Academy of Sciences Proceedings 108, 8363–8366. Food and Agriculture Organization, 1974. Soil Map of the World: v. 1, Legend. United Nations Educational, Scientific and Cultural Organization, Paris. 17 pp. Franks, P.C., 1975. The transgressive-regressive sequence of the Cretaceous Cheyenne, Kiowa, and Dakota formations of Kansas. In: Caldwell, W.G.E. (Ed.), Cretaceous System in the Western Interior of North America: Geological Association of Canada Special Paper, 13, pp. 469–521. Friis, E.M., Crane, P.R., Pedersen, K.J., 2011. Early Flowers and Angiosperm Evolution. Cambridge Univ. Press, Cambridge. 585 pp.

61

Frye, J.C., Willman, H.B., Glass, H.D., 1964. Cretaceous deposits and the Illinoian glacial boundary in western Illinois. Illinois State Geological Survey Circular 364, 1–25. Galton, P.M., Jensen, J.A., 1978. Remains of ornithopod dinosaurs from the Early Cretaceous of North America. Brigham Young University Studies in Geology 25, 1–10. Gehling, J.G., 1999. Microbial mats in terminal Proterozoic siliciclastics. Ediacaran death masks. Palaios 14, 40–57. Goldich, S.S., 1938. A study in rock weathering. Journal of Geology 46, 17–58. Gradstein, F.M., Ogg, J.G., Smith, A.G., 2004. A Geologic Time Scale 2004. Cambridge Univ. Press, Cambridge. 589 pp. Greb, S.F., DiMichele, W.A., Gastaldo, R.A., 2006. Evolution and importance of wetlands in earth history. In: Greb, A.F., DiMichele, W.A. (Eds.), Wetlands through Time: Geological Society of America Special Paper, 399, pp. 1–40. Grimes, S.T., Davies, K.L., Butler, I.B., Brock, F., Edwards, D., Rickard, D., Briggs, D.E.G., Parkes, R.J., 2002. Fossil plants from the Eocene London Clay; the use of pyrite textures to determine the mechanism of pyritization. Journal of the Geological Society of London 159, 493–501. Gröcke, D.R., Ludvigson, G.A., Witzke, B.L., Robinson, S.A., Joeckel, R.M., Ufnar, D.F., Ravn, R.L., 2006. Recognizing the Albian–Cenomanian (OAE-1d) sequence boundary using plant carbon isotopes, Dakota Formation, Western Interior Basin. Geology 34, 193–196. Hamilton, V., 1994. Sequence stratigraphy of Cretaceous Albian and Cenomanian strata in Kansas. In: Shurr, G.W., Ludvigson, G.A., Hammond, R.H. (Eds.), Perspectives on the Eastern Margin of the Cretaceous Western Interior Basin: Geological Society of America Special Paper, 287, pp. 79–96. Hartshorn, G.S., 1983. Plants. In: Janzen, D.H. (Ed.), Costa Rican Natural History. University of Chicago Press, Chicago, pp. 118–158. Hattin, D.E., 1967. Stratigraphic and paleoecologic significance of macroinvertebrate fossils in the Dakota Formation (upper Cretaceous) of Kansas. In: Teichert, C., Yochelson, E.L. (Eds.), Essays in Paleontology and Stratigraphy, Raymond C. Moore commemorative volume: Department of Geology University of Kansas Special Volume, 2, pp. 570–589. Hattin, D.E., Siemers, C.T., Stewart, G.F., 1978. Guidebook: Upper Cretaceous stratigraphy and depositional environments of western Kansas. Kansas Geological Survey Guidebook Series, 3. 102 pp. Hoashi, M., Bevacqua, D.C., Otake, T., Watanabe, Y., Hickman, A.H., Utsunomiya, S., Ohmoto, H., 2009. Primary haematite formation in an oxygenated sea 3.46 billion years ago. Nature Geoscience 2, 301–306. Hu, S., Dilcher, D.L., Jarzen, D.M., Taylor, D.W., 2008a. Early steps of angiosperm-pollinator coevolution. U.S. National Academy of Sciences Proceedings 105, 240–245. Hu, S., Jarzen, D.M., Dilcher, D.L., 2008b. New species of Cenomanian angiosperm pollen, Cenomanian of Minnesota, U.S.A. Palynology 32, 17–26. Isbell, R.F., 1996. The Australian Soil Classification: Revised Edition. CSIRO Publishing, Melbourne. 144 pp. Joeckel, R.M., 1987. Paleogeomorphic significance of two paleosols in the Dakota Formation (Cretaceous) southeastern Nebraska. Contributions to Geology University of Wyoming 25, 95–102. Joeckel, R.M., 1991. Plinthic paleosols in Lower Dakota Formation (Cretaceous(Albian– Cenomanian) at Endicott clay pit. In: Johnson, W.C. (Ed.), Second International Paleopedology Symposium Field trip: Kansas Geological Survey Open File Report, 93-30-16/17, pp. 1–3. Joeckel, R.M., Cunningham, J., Phillips, P.L., Ludvigson, G.A., Witzke, B.J., Corner, R.G., Brown, G.W., 2001. The first report of dinosaur tracks from Nebraska. Journal of Vertebrate Paleontology Abstracts 21, 66. Karl, H.A., 1976. Depositional history of Dakota Formation (Cretaceous) sandstone in southeastern Nebraska. Journal of Sedimentary Petrology 46, 124–131. Koch, J.T., Koch, A.M., Brenner, R.L., 2007. A large incised fluvial channel within the Mid-Cretaceous Dakota Formation, Washington County, KS; evidence of Cretaceous glaciation? Geological Society of America Abstracts 39 (3), 71. Kovach, W.L., 1988. Quantitative paleoecology of megaspores and other dispersed plant remains from the Cenomanian of Kansas, U.S.A. Cretaceous Research 9, 265–283. Kovach, W.L., Dilcher, D.L., 1988. Megaspores and other dispersed plant remains from the Dakota Formation (Cenomanian) of Kansas, U.S.A. Palynology 12, 89–119. Labandiera, C.C., 1998. The role of insects in Late Jurassic to middle Cretaceous ecosystems. In: Lucas, S.G., Kirkland, J.I., Estep (Eds.), Lower to Middle Cretaceous Terrestrial Ecosystems: New Mexico Museum of Natural History and Science Bulletin, 14, pp. 105–124. Larsson, L., Olsson, G., Holst, O., Karlsson, H.L., 1990. Pyrite oxidation by thermophilic archaebacteria. Applied and Environmental Microbiology 56, 697–701. Leopold, L.B., 1943. Climatic character of the interval between the Jurassic and Cretaceous in New Mexico and Arizona. Journal of Geology 51, 56–62. Lesquereux, L., 1892. The flora of the Dakota Group. U.S. Geological Survey Monograph 17, 1–400. Liggett, G.l., 2005. A review of the dinosaurs from Kansas. Kansas Academy of Sciences Transactions 108, 1–14. Loeblich, A.R., Tappan, H., 1961. Cretaceous planktic foraminifera. Part I: Cenomanian. Micropalaeontology 7, 257–305. Ludvigson, G.A., González, L.A., Metzger, R.A., Witzke, B.J., Brenner, R.L., Murillo, A.P., White, T.S., 1998. Meteoric sphaerosiderite lines and their use for paleohydrology and paleoclimatology. Geology 26, 1039–1042. Macfarlane, P.A., Doveton, J.H., Goble, G., 1989. Interpretation of lithologies and depositional environments of Cretaceous and Lower Permian rocks by using a diverse suite of logs from a borehole in central Kansas. Geology 17, 303–306. Macfarlane, P.A., Whittemore, D.O., Townsend, M.A., Doveton, J.H., Hamilton, V.J., Coyle, W.G., Wade, A., Macpherson, G.L., Black, R.D., 1990. The Dakota aquifer program: annual report, FY89. Kansas Geological Survey Open File Report 90-27. 302 pp.

62

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63

Mack, G.H., 1991. Paleosols as an indicator of climatic change at the Early–Late Cretaceous boundary, southwestern New Mexico. Journal of Sedimentary Petrology 62, 483–494. Mamay, S.H., 1989. Evolsonia, a new genus of Gigantopteridaceae from the Lower Permian Vale Formation, north-central Texas. American Journal of Botany 76, 1299–1311. Markewich, H.W., Pavich, M.J., Buell, G.R., 1990. Contrasting soils and landscapes of the Piedmont and Coastal Plain, eastern United States. Geomorphology 3, 417–447. Maynard, J.B., 1992. Chemistry of modern soils as a guide to Precambrian fossil soils. Journal of Geology 100, 279–289. McBride, E.F., Milliken, K.L., 2006. Giant calcite-cemented concretions, Dakota Formation, Kansas. Sedimentology 53, 1161–1173. McCarthy, P.J., 2002. Micromorphology and development of interfluve paleosols; a case study from the Cenomanian Dunvegan Formation, NE British Columbia, Canada. In: Griffith, L.A., McCabe, P.J., Williams, C.A. (Eds.), Fluvial Systems through Space and Time; from Source to Sea: Bulletins of Canadian Petroleum Geology, 50, pp. 158–177. McCarthy, P.J., Plint, A.G., 1999. Floodplain palaeosols of the Cenomanian Dunvegan Formation, Alberta and British Columbia, Canada; micromorphology, pedogenic processes and palaeoenvironmental implications. In: Marriott, S.B., Alexander, J. (Eds.), Floodplains; Interdisciplinary Approaches: Geological Society of London Special Publication, 163, pp. 289–310. McCarthy, P.J., Plint, A.G., 2003. Spatial variability of palaeosols across Cretaceous interfluves in the Dunvegan Formation, NE British Columbia, Canada; palaeohydrological, palaeogeomorphological and stratigraphic implications. Sedimentology 50, 1187–1220. McCarthy, P.J., Faccini, U.F., Plint, A.G., 1999. Evolution of an ancient coastal plain; palaeosols, interfluves and alluvial architecture in a sequence stratigraphic framework, Cenomanian Dunvegan Formation, NE British Columbia, Canada. Sedimentology 46, 861–891. McKenzie, N., Jacquier, D., Isbell, R., Brown, K., 2004. Australian Soils and Landscapes: An Illustrated Compendium. Commonwealth Scientific and Industrial Organization, Melbourne. 416 pp. Mehl, M.G., 1941. Dakotosuchus kingi, a crocodile from the Dakota of Kansas. Journal of the Scientific Laboratories, Denison University 36, 47–65. Nordt, L., Wilding, L.P., Lynn, W., Crawford, C.C., 2004. Vertisol genesis in a humid climate of the coastal plain of Texas. Geoderma 122, 83–102. Ohmoto, H., Watanabe, Y., Allwood, A., Burch, I.W., Knauth, L.P., Yamaguchi, K.E., Johnson, I., Altinok, E., 2007. Formation of probable lateritic soils approximately 3.43 Ga in the Pilbara Craton, Western Australia. Geochimica Cosmochimica Acta 71 (15S), A733. Olsson, G., Pott, B.-M., Larsson, L., Holst, O., Karlsson, H.T., 1995. Microbial desulfurization of coal and oxidation of pure pyrite by Thiobacillus ferrooxidans and Acidianus brierleyi. Journal of Industrial Microbiology and Biotechnology 14, 420–423. Parham, W.E., 1970. Clay mineralogy and geology of Minnesota's kaolin clays. Minnesota Geological Survey Special Publication 10, 1–142. Parham, W.E., Hogberg, R.K., 1964. Kaolin resources of the Minnesota River Valley, Redwood and Renville Counties, a preliminary report. Minnesota Geological Survey Report of Investigations 3, 1–43. Pellegrin, V., Juretschko, S., Wagner, M., Cottenceau, G., 1999. Morphological and biochemical properties of a Sphaerotilus sp. isolated from paper mill slimes. Applied and Environmental Microbiology 65, 156–162. Peppe, D.J., Royer, D.L., Wilf, P., Kowalski, E.A., 2010. Quantification of large uncertainties in fossil leaf paleoaltimetry. Tectonics 29, TC3015. doi:10.1029/ 2009TC002549. Pierce, R.L., 1961. Lower Upper Cretaceous plant microfossils from Minnesota. Minnesota Geological Survey Bulletin 42, 1–86. Plummer, N., Romary, J.F., 1947. Kansas clay, Dakota Formation. Geological Survey of Kansas Bulletin 67, 1–241. Plummer, N., Bauleke, M.P., Hladik, W.B., 1960. Dakota Formation refractory clays and silts in Kansas. Geological Survey of Kansas Bulletin 142, 1–52. Porter, D.A., McCahon, T.J., Ransom, M.D., 1991. A hydromorphic paleosol in the Dakota Formation (Cretaceous), northcentral Kansas. In: Johnson, W.C. (Ed.), Second International Paleopedology Symposium Field Trip: Kansas Geological Survey Open File Report, 93-30-7, pp. 1–5. Pye, K., 1981. Marshrock formed by iron sulphide and siderite cementation in salt marsh sediments. Nature 294, 650–652. Ravn, R.L., Witzke, B., 1994. The mid-Cretaceous boundary in the Western Interior Seaway, central United States: implications of palynostratigraphy from the type Dakota Formation. In: Shurr, G.W., Ludvigson, G.A., Hammond, R.H. (Eds.), Perspectives on the Eastern Margin of the Cretaceous Western Interior Basin: Geological Society of America Special Paper, 287, pp. 111–128. Ravn, R.L., Witzke, B., 1995. The palynostratigraphy of the Dakota Formation (?late Albian–Cenomanian) in its type area, northwestern Iowa and northeastern Nebraska, USA. Palaeontographica B234, 93–171. Reed, E.C., 1954. Evidence of ancient soils in Cretaceous and pre-Cambrian rocks, Kansas–Nebraska. Nebraska Academy of Sciences Proceedings 64, 15. Retallack, G.J., 1997. A Colour Guide to Paleosols. John Wiley and Sons, Chichester. 175 pp. Retallack, G.J., 1998. Fossil soils and completeness of the rock and fossil record. In: Donovan, S.K., Paul, C.R.C. (Eds.), The Adequacy of the Fossil Record. John Wiley and Sons, Chichester, pp. 131–162. Retallack, G.J., 2007. Growth, decay and burial compaction of Dickinsonia, an iconic Ediacaran fossil. Alcheringa 31, 215–240. Retallack, G.J., 2008. Cambrian palaeosols and landscapes of South Australia. Australian Journal of Earth Sciences 55, 1083–1106.

Retallack, G.J., 2009. Greenhouse crises of the past 300 million years. Geological Society of America Bulletin 121, 1441–1455. Retallack, G.J., 2010. Lateritization and bauxitization events. Economic Geology 105, 655–667. Retallack, G.J., Dilcher, D.L., 1981a. A coastal hypothesis for the dispersal and rise to dominance of flowering plants. In: Niklas, K.J. (Ed.), Paleobotany, Paleoecology, and Evolution, 1. Praeger, New York, pp. 22–77. Retallack, G.J., Dilcher, D.L., 1981b. Early angiosperm reproduction: Prisca reynoldsii, gen. et. sp. nov. from mid-Cretaceous coastal deposits in Kansas, U.S.A. Palaeontographica B179, 103–137. Retallack, G.J., Dilcher, D.L., 1986. Cretaceous angiosperm invasion of North America. Cretaceous Research 7, 227–252. Retallack, G.J., Huang, C.-M., 2010. Depth to gypsic horizon as a proxy for paleoprecipitation in paleosols of sedimentary environments. Geology 38, 403–406. Retallack, G.J., Jahren, A.H., 2008. Methane release from igneous intrusion of coal during Late Permian extinction events. Journal of Geology 116, 1–20. Rogers, K., 1991. The Sternberg Fossil Hunters: A Dinosaur Dynasty. Mountain Press, Missoula. 288 pp. Schopf, J.M., 1975. Modes of fossil preservation. Review of Palaeobotany and Palynology 20, 27–53. Setterholm, D.R., 1994. The Cretaceous rocks of southwestern Minnesota: reconstruction of marine to nonmarine transition from the eastern margin of the Western Interior Seaway. In: Shurr, G.W., Ludvigson, G.A., Hammond, R.H. (Eds.), Perspectives on the Eastern Margin of the Cretaceous Western Interior Basin: Geological Society of America Special Paper, 287, pp. 97–110. Shabica, C.W., Hay, A.A. (Eds.), 1997. Richardson's Guide to the Fossil Fauna of Mazon Creek. Northeastern Illinois University Press, Chicago. 308 pp. Sheldon, N.D., Retallack, G.J., 2001. Equation for compaction of paleosols due to burial. Geology 29, 247–250. Sheldon, N.D., Retallack, G.J., Tanaka, S., 2002. Geochemical climofunctions from North American soils and application to paleosols across the Eocene–Oligocene boundary in Oregon. Journal of Geology 110, 687–696. Shelton, J.W., 1962. Shale compaction in a section of Cretaceous Dakota Sandstone, northwestern North Dakota. Journal of Sedimentary Petrology 32, 873–877. Siemers, C.T., 1976. Sedimentology of the Rocktown channel sandstone, upper part of the Dakota Formation (Cretaceous) central Kansas. Journal of Sedimentary Petrology 46, 97–123. Skog, J.E., Dilcher, D.L., 1992. A new species of Marsilea from the Dakota Formation of central Kansas. American Journal of Botany 79, 982–988. Skog, J.E., Dilcher, D.L., 1994. Lower vascular plants of the Dakota Formation in Kansas and Nebraska, U.S.A. Reviews of Palaeobotany and Palynology 80, 1–18. Sloan, R.E., 1964. The Cretaceous system in Minnesota. Minnesota Geological Survey Report of Investigations 5, 1–64. Smith, S.M., Newman, S., Garrett, P.B., Leeds, J.A., 2001. Differential effects of surface and peat fire on soil constituents in a degraded wetland of the northern Florida Everglades. Journal of Environmental Quality 30, 1998–2005. Soil Survey Staff, 2000. Keys to Soil Taxonomy. Pocahontas Press, Blacksbrug. 600 pp. Spicer, R.A., 1977. The pre-depositional formation of some leaf impressions. Palaeontology 20, 907–912. Swineford, A., 1947. Cemented sandstones of the Dakota and Kiowa Formations in Kansas. Kansas Geological Survey Bulletin 79, 53–104. Tate, T.M., Retallack, G.J., 1995. Thin sections of paleosols. Journal of Sedimentary Research A65, 579–580. Thorp, J., Reed, E.C., 1949. Is there laterite in rocks of the Dakota Group? Science 209, 69. Tidwell, W.D., Simper, A.D., Thayn, G.F., 1977. Additional information concerning the controversial Triassic plant; Sanmiguelia. Palaeontographica B163, 143–151. Trewartha, G.T., 1982. Earth's Problem Climates. University of Wisconsin Press, Madison. 334 pp. Triplehorn, D.M., 1971. Weathering of the Sioux Quartzite near Ulm, Minnesota as related to Cretaceous climates: a discussion. Journal of Sedimentary Petrology 41, 603–604. Ufnar, D.F., Ludvigson, G.A., González, L.A., Gröcke, D.R., 2008a. Precipitation rates and atmospheric heat transport during the Cenomanian greenhouse warming in North America; estimates from a stable isotope mass-balance model. Palaeogeography Palaeoclimatology Palaeoecology 266, 28–38. Ufnar, D.F., Gröcke, D.R., Beddows, P.A., 2008b. Assessing pedogenic calcite stable isotope values; can positive linear covariant trends be used to quantify palaeoevaporation rates? Chemical Geology 256, 46–51. Upchurch, G.R., Dilcher, D.L., 1990. Cenomanian angiosperm leaf megafossils, Dakota Formation, Rose Creek locality, Jefferson County, southeastern Nebraska. U.S. Geological Survey Bulletin 1915 55 pp.. Upchurch, G.R., Wolfe, J.A., 1987. Mid-Cretaceous to early Tertiary vegetation and climate: evidence from fossil leaves and woods. In: Friis, E.M., Chaloner, W.G., Crane, P.R. (Eds.), The Origin of Angiosperms and their Biological Consequences. Cambridge University Press, Cambridge, pp. 75–104. van Veen, W.L., Mulder, E.G., Deinema, M.H., 1978. The Sphaerotilus-Leptothrix group of bacteria. Microbiological Review 42, 239–356. Vaughn, P.O., 1956. A second specimen of the Cretaceous crocodile Dakotasuchus. Kansas Academy of Sciences Transactions 59, 379–381. Vepraskas, M.J., Sprecher, S.W., 1997. Overview. In: Vepraskas, M.J., Sprecher, S.W. (Eds.), Aquic Conditions and Hydric Soils: The Problem Soils: Soil Science Society of America Special Publication, 50, pp. 1–22. Wang, H., 2002. Diversity of angiosperm leaf megafossils from the Dakota Formation (Cenomanian, Cretaceous), northwestern Interior, U.S.A. Unpublished PhD thesis, University of Florida, Gainesville, 437 p.

G.J. Retallack, D.L. Dilcher / Palaeogeography, Palaeoclimatology, Palaeoecology 329–330 (2012) 47–63 Wang, H., Dilcher, D.L., 2006. Aquatic angiosperms from the Dakota Formation (Albian, lower Cretaceous), Hoisington III locality, Kansas, U.S.A. International Journal of Plant Sciences 167, 385–401. Wang, H., Moore, M.J., Soltis, P.S., Bell, C.D., Brockington, S.F., Alexandre, R., Davis, C.C., Latvis, M., Manchester, S.R., Soltis, D.E., 2009. Rosid radiation and the rapid rise of angiosperm-dominated forests. U.S. National Academy of Sciences Proceedings 106, 3853–3858. Wang, H., Dilcher, D.L., Schwarzwalder, R., Kvacek, J., 2011. Vegetative and reproductive morphology of an extinct Early Cretaceous member of Platanaceae from the Braun's Ranch locality, Kansas, USA. International Journal of Plant Sciences 172, 139–157.

63

White, C.A., 1894. Notes on the invertebrate fauna of the Dakota Formation, with descriptions of new molluscan forms. U.S. National Museum Proceedings 17, 131–138. Winkler, M.G., Sanford, P.R., Kaplan, S.W., 2001. Hydrology, vegetation, and climate change in the southern Everglades during the Holocene. In: Wardlaw, B.R. (Ed.), Paleoecological Studies of South Florida: Bulletins of American Paleontology, 361, pp. 57–99. Witzke, B.J., Ludvigson, G.A., 1994. The Dakota Formation in Iowa and the type area. In: Shurr, G.W., Ludvigson, G.A., Hammond, R.H. (Eds.), Perspectives on the Eastern Margin of the Cretaceous Western Interior Basin: Geological Society of America Special Paper, 287, pp. 43–78. Wood, R.A., 1996. Weather of U.S. Cities. Gale Research, New York. 1075 pp.