Earth and Planetary Science Letters 365 (2013) 25–37
Contents lists available at SciVerse ScienceDirect
Earth and Planetary Science Letters
journal homepage: www.elsevier.com/locate/epsl
The astronomical rhythm of Late-Devonian climate change
(Kowala section, Holy Cross Mountains, Poland)
David De Vleeschouwer a,n, Micha" Rakociński b, Grzegorz Racki b, David P.G. Bond c,
Katarzyna Sobień d, Philippe Claeys a
a
Earth System Sciences and Department of Geology, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussels, Belgium
Faculty of Earth Sciences, University of Silesia, B˛edzińska Str. 60, 41-200 Sosnowiec, Poland
c
Norwegian Polar Institute, Fram Centre, 9296 Tromsø, Norway
d
Polish Geological Institute—National Research Institute, Rakowiecka 4, 00-975 Warszawa, Poland
b
a r t i c l e i n f o
abstract
Article history:
Received 24 April 2012
Received in revised form
16 January 2013
Accepted 17 January 2013
Editor: G. Henderson
Rhythmical alternations between limestone and shales or marls characterize the famous Kowala
section, Holy Cross Mountains, Poland. Two intervals of this section were studied for evidence of orbital
cyclostratigraphy. The oldest interval spans the Frasnian–Famennian boundary, deposited under one of
the hottest greenhouse climates of the Phanerozoic. The youngest interval encompasses the Devonian–
Carboniferous (D–C) boundary, a pivotal moment in Earth’s climatic history that saw a transition from
greenhouse to icehouse. For the Frasnian–Famennian sequence, lithological variations are consistent
with 405-kyr and 100-kyr eccentricity forcing and a cyclostratigraphic floating time-scale is presented.
The interpretation of observed lithological rhythms as eccentricity cycles is confirmed by amplitude
modulation patterns in agreement with astronomical theory and by the recognition of precession cycles
in high-resolution stable isotope records. The resulting relative time-scale suggests that 800 kyr
separate the Lower and Upper Kellwasser Events (LKE and UKE, respectively), two periods of anoxia
that culminated in massive biodiversity loss at the end of the Frasnian. Th/U and pyrite framboid
analyses indicate that during the UKE, oxygen levels remained low for 400 kyr and d13Corg measurements demonstrate that more than 600 kyr elapsed before the carbon cycle reached a steady state
after a þ 3% UKE excursion. The Famennian–Tournaisian (D–C) interval also reveals eccentricity and
precession-related lithological variations. Precession-related alternations clearly demonstrate grouping
into 100-kyr bundles. The Famennian part of this interval is characterized by several distinctive anoxic
black shales, including the Annulata, Dasberg and Hangenberg shales. Our high-resolution cyclostratigraphic framework indicates that those shales were deposited at 2.2 and 2.4 Myr intervals respectively.
These durations strongly suggest a link between the long-period ( 2.4 Myr) eccentricity cycle and the
development of the Annulata, Dasberg and Hangenberg anoxic shales. It is assumed that these black
shales form under transgressive conditions, when extremely high eccentricity promoted the collapse of
small continental ice-sheets at the most austral latitudes of western Gondwana.
& 2013 Elsevier B.V. All rights reserved.
Keywords:
Late Devonian
Milanković forcing
limestone-shale rhythmites
Gondwanan glaciation
anoxic black shales
TR-cycles
1. Introduction
Throughout the Phanerozoic, the imprint of Milanković’s
astronomical cycles is recognized in lithological rhythms, as well
as in many geochemical and paleoecological proxies (e.g. Boulila
et al., 2011; Fischer and Bottjer, 1991; House and Gale, 1995;
Tyszka, 2009). The study of cyclical change in sedimentary
archives provides a better insight in the sensitivity of the Earth’s
climate system to cyclic variations in the orbital parameters of the
Earth (e.g. Lourens et al., 2010; Verschuren et al., 2009) and
n
Corresponding author. Tel.: þ32 486318938.
E-mail address: [email protected] (D. De Vleeschouwer).
0012-821X/$ - see front matter & 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.epsl.2013.01.016
demonstrates that climatic setting strongly influences the response
of climate to orbital forcing (Boulila et al., 2011; Giorgioni et al.,
2012). As a geo-chronometer, astronomical rhythms observed in a
sedimentary archive provide insight into absolute and relative
timing questions (Ghil et al., 2002; Kuiper et al., 2008; Muller and
MacDonald, 2000).
A relevant astronomical imprint into climatic proxies has been
successfully demonstrated in diverse cyclostratigraphic studies
of Devonian carbonate-shelf archives, exemplified by Chen
and Tucker (2003), Crick et al. (2002), De Vleeschouwer et al.
(2012a, 2012b), Elrick and Hinnov (2007), Elrick et al. (2009),
Garcia-Alcalde et al. (2012) and House (1991, 1995). The
Devonian system is poorly constrained by radiometric dating
(see recent chronologies by Becker et al., 2012; Kaufmann,
26
D. De Vleeschouwer et al. / Earth and Planetary Science Letters 365 (2013) 25–37
2006), and the refined analysis of sedimentary rhythmicity in
stratigraphically extended and relatively continuous sequences
remains an overall unexploited field and an urgent research aim,
as stressed by House (2002) and Racki (2005). This study focuses
on the conodont-dated and widely studied Late Devonian shelfbasin successions at Kowala, Holy Cross Mountains (central
Poland; see regional geological setting in Chapter 2 and
Supplementary material) characterized by distinct limestoneshale rhythmites (Bond and Zatoń, 2003; Fig. 2 in Marynowski
and Filipiak, 2007; Fig. 2 in Marynowski et al., 2010; Fig. 3 in
Racka et al., 2010; Fig. 3 in Racki et al., 2002; Fig. 1A and B in
Racki, 2005; marly facies of Szulczewski, 1971; 1995). We test
whether or not this rhythmicity is the result of astronomical
forcing, within a climate on the eve of the transition from the
Devonian greenhouse to the Carboniferous icehouse (Racki, 2005,
p. 15; Streel et al., 2000). In addition, a comprehensive set of other
environmental proxies, such as stable isotopes, magnetic susceptibility and gamma-ray spectrometry, which record only minor
diagenetic/thermal overprint (Belka, 1990; Joachimski et al.,
2001; Marynowski et al., 2011), serve as indispensable tools for
the correct interpretation of the observed lithological variations
in terms of paleoenvironmental change.
2. Regional and stratigraphic setting
Extensive outcrops of Middle to Upper Devonian reefal and
basinal systems are exposed in the Holy Cross Mountains (see
Supplementary material), that formed a fragment of the once vast
equatorial carbonate shelves of south-eastern Laurussia (Racki,
1993; Racki et al., 2002; Szulczewski, 1995). The centrally located
Frasnian Dyminy Reef evolved into a Famennian pelagic ridge,
and this shoal was surrounded by drowned, oxygen-depleted
deeper-shelf areas (i.e. intrashelf basins): Ch˛eciny–Zbrza to the
south, and Łysogóry–Kostom"oty to the north (Racki, 1993;
Szulczewski, 1971, 1995).
The large, still active Kowala quarry near Kielce is well established
as a major reference site for multifaceted Late Devonian studies. This
more than 350 m thick, generally deepening-upward and continuous
succession from the Ch˛eciny–Zbrza basin is relatively well-dated
using conodonts (Fig. 1), and comprehensively investigated from
event-stratigraphical, paleontological and geochemical perspectives
(summarized in Racki, 2005). The intensely studied distinctive
Frasnian–Famennian and late Famennian bituminous horizons record
several recurrent anoxic pulses that are known to be of global extent
(Filipiak and Racki, 2010; Kazmierczak et al., 2012; Marynowski and
Filipiak, 2007; Marynowski et al., 2010; 2012; Racka et al., 2010).
Within the largely dark to black rhythmically bedded marly
series, late Frasnian to basal Carboniferous lithologic units H–N
(Fig. 1; Bond and Zatoń, 2003; Szulczewski, 1996) record the postreef phase of shelf evolution. In this study, two rhythmite-bearing
successions were analyzed for cyclostratigraphy (Fig. 1):
(1) The upper Frasnian to lower Famennian interval (sets H-1 to
H-4) encompasses the Frasnian–Famennian (F–F) boundary
beds and records one of the greatest biodiversity crises of the
Phanerozoic, the Late Devonian mass extinction (Bond et al.,
2004; Joachimski et al., 2002; Racki et al., 2002). Graded
bioclastic limestones (with redeposited reef-builders and
brachiopod coquinas) and/or thick (up to 1 m) slump layers
occur in the basal (H-1) unit only. An abrupt lithologic change
from wavy- and thin-bedded marly deposits, which are highly
fossiliferous in places (unit H-2), to cherty limestone strata
(H-3), is particularly significant for the F-F passage (Racki
et al., 2002). These relatively thick-layered calcareous sets,
6 to 8 m thick, with thin (up to 5 cm) shale breaks, constitute
the main lithologic peculiarity within the otherwise uniform
marly rhythmic succession of set H. In fact, the succeeding
monotonous fossil-impoverished unit H-4, which is more
than 100 m thick, includes regularly alternating marly limestones and shales, here analyzed in the lower part only (see
also Filipiak, 2009; Marynowski et al., 2011).
(2) The highest exposed interval contains the latest Devonian
thicker-layered dark platy limestone-shale couplets, interrupted by the Annulata and Dasberg black shales (set K;
details in Marynowski et al., 2010; Racka et al., 2010) and the
grey to brownish fossil-rich wavy-bedded to (sub)nodular
calcareous set with thin shale partings (Woclumeria limestone, set L, see Supplementary material; Marynowski and
Filipiak, 2007; Rakocinski, 2011). The Hangenberg bituminous shale (90 cm thick, total organic carbon (TOC) up to
22.5 wt%), that also records a major biotic crisis, albeit of less
significance than the F–F event, occurs above this interval. It
is succeeded by a 120 cm thick grey-greenish clayey/tuffite
succession with several interbedded limestone layers (the
newly proposed set M ¼sets B and C sensu Malec, 1995;
see also Marynowski and Filipiak, 2007, Fig. 2 therein). The
Devonian–Carboniferous (D–C) boundary occurs within the
upper part of this unit. The Lower Carboniferous comprises a
greenish clayey/tuffite interval with four thin black
shale intercalations (5 to 15 cm thick) and some nodular
carbonate horizons. This recently exposed succession is
named herein as set N ( ¼lower part set D sensu Malec,
1995, Fig. 8).
The middle segment of the Famennian succession remains
unexplored due to frequent diagenetic nodular lithologies that
obscure the depositional rhythmicity (Fig. 1; Marynowski et al.,
2007), as well as some fault disturbances and/or covered intervals
(Fig. 2 in Bond and Zatoń, 2003).
3. Materials and methods
3.1. Time-series analysis procedure
The frequency composition of lithological variations (shales,
marls, micrites, sparites, calcarenites and nodular limestones)
were analyzed via spectral analysis. To carry out time-series
analysis of lithological variations, each lithology was assigned a
different code and a quantified litholog was obtained. Spectral
analysis of proxy data, like the isotopic, magnetic susceptibility
and gamma-ray spectrometry records, was carried out after
detrending and interpolating the records to equally-spaced
intervals. Proxy data and quantified logs were analyzed using
the multitaper method (MTM; Thomson, 1982) as implemented
in the SSA-MTM Toolkit (Ghil et al., 2002).
The MTM-method was performed using 3 discrete prolate
spheroidal sequences (DPSS) as data tapers to compromise between
spectral resolution and sidelobe reduction. Small variations in
accumulation rate behave like phase modulations, and introduce
multiple spurious spectral peaks (Muller and MacDonald, 2000). We
chose the MTM because it averages these sidelobes into the main
peak and thereby gives a superior estimate of the true spectral
power. To assess whether or not the strongest spectral peaks are
statistically different from the red noise spectrum, the 95% confidence level (CL) is calculated (robust AR(1) estimation, median
smoothing window width¼(10Dt) 1, Log Fit, fNyquist ¼2/Dt).
The Continuous Wavelet Transform is used to decompose
the one-dimensional time-series into their two-dimensional
D. De Vleeschouwer et al. / Earth and Planetary Science Letters 365 (2013) 25–37
27
Fig. 1. Composite lithological section of the Upper Devonian strata at Kowala (after Szulczewski, 1996, Fig. 8, modified and supplemented in the topmost part by recently
exposed sets M and N). The reference succession shows environmental evolution typical of intermittently drowned shelf from reef (units A–C) to foreslope (D–G) to
intrashelf basin (H–N).
time–frequency representation. This method reveals the persistency
of (astronomical) periodicities along the studied sections and
provides a tool to detect changes in accumulation rate through
associated changes in cycle thickness. Wavelet software was provided by Torrence and Compo (1998), using the Morlet wavelet
function with wave number 6, zero-padding to the next higher
power of 2, spacing between discrete scales (Dj)¼0.25, smallest
scale (S0)¼2Dt, and the number of scales¼ logz ðN Dt=S0 Þ=Dj þ 1.
Frequency-selective filters or Gaussian band-pass filters isolate
and extract the components of signals associated with a specific
range of frequencies. We employ band-pass filters to assess the
behavior of a specific range of frequencies in a studied signal
using the Analyseries software (Paillard et al., 1996).
For a detailed description of the methodology of proxy
record compilation, the reader is referred to the Supplementary
material.
28
D. De Vleeschouwer et al. / Earth and Planetary Science Letters 365 (2013) 25–37
4. Results and discussion
4.1. Do limestone-marl rhythmites reflect primary environmental
change?
Rhythmical alternations between limestones, shales and marls
are abundant throughout the entire Phanerozoic. Before any
climatic interpretation can be made, or any astronomical cycle
can be recognized, the primary origin of the alternations between
different lithologies needs to be demonstrated (Westphal et al.,
2010). Several arguments strongly suggest that, with some
exceptions noted below, diagenesis can be excluded as the driver
for the decimeter-scale rhythmic interbedding that dominates the
Kowala section. (1) The immature character of the organic matter
indicates burial temperatures lower than 75 1C (Belka, 1990).
Immaturity to low-maturity is confirmed by an average vitrinite
reflectance value of 0.53% Rr (Marynowski et al., 2001), the
temperature of maximum release of hydrocarbons (Joachimski
et al., 2001; Marynowski and Filipiak, 2007; Marynowski et al.,
2007; 2010) and hydrogen, oxygen and conodont alteration
indexes (Joachimski et al., 2001). (2) Many primary sedimentary
structures are well preserved, including distinctly graded limestone layers, hummocky stratification, trilobite obrution deposits,
extensive lateral continuity, burrow fills, and web-like structures
typical of benthic cyanobacterial mats (Kazmierczak et al., 2012;
Marynowski et al., 2011; Racki et al., 2002; Radwanski et al.,
2009; Szulczewski, 1971). Moreover, delicate fossils, such as
Phyllocarida and other crustaceans, ostracod-shells and aptychi
of the goniatite Tornoceras (Dzik, 2006) as well as carbonized
thalli of large algae with filamentous internal structure (Fig. 3B in
Racki et al., 2002) are superbly preserved. The exceptional preservation of these fossils and sedimentary structures demonstrates minimal diagenetic alternation in this section and thus a
limited applicability of the early diagenesis unmixing model
(Westphal et al., 2010). (3) The character of d13C and d18O
variations in whole-rock samples appear to reflect a predominantly primary signature. Indeed, the isotopic data presented in
this study (Figs. 3E and F and 6G and H) fall entirely within the
range of d13C and d18O for Upper Devonian–Lower Carboniferous
brachiopod shell calcite (Giles, 2012; van Geldern et al., 2006;
Veizer et al., 1999). Furthermore, the recognized d13C trends in
calcites at Kowala correlate with organic carbon data (Fig. 3E and
G; Joachimski et al., 2001) and the d18O pattern shows the same
trends as diagenetically resistant conodont apatites (Fig. 5 in
Joachimski et al., 2009). Consequently, in this section, diagenesis
appears so moderate that post-depositional incorporation of light
isotopes into calcite is negligible. Above all, meteoric diagenesis,
paired with strongly distorted isotopic signatures due to the
influence of 12C and 16O enriched fluids, is essentially absent in
deeper-shelf facies. Indeed, the reported d13C and d18O series
exhibit frequent enrichments in heavy isotopes of 13C and 18O
respectively, which cannot be attributed to diagenetic processes
(Joachimski et al., 2001; van Geldern et al., 2006). Pressure
solution carbonate re-mobilization is attested by a crudely laminated ribbon-facies in sharp contact with neomorphosed crystalline carbonate and pressure-welded (stylolaminite) argillaceous
interbeds (especially in unit H-3; Figs. 12D and 16C in Racki et al.,
2002). This conspicuous diagenetic instability of the precursor
carbonate muds may be promoted by mass abundance of susceptible opaline skeletons (radiolarians, silicisponges) and chert
formation near the F–F boundary only (Vishnevskaya et al.,
2002). Nevertheless, the isotopic ratios still indicate a primary
origin of the alternating lithologies as manifested by the mostly
positive d13C values (Fig. 3E) in the broad F–F transition, concurrent with the global UKE isotopic pattern (Joachimski et al.,
2001; Marynowski et al., 2011; Racki et al., 2002). Shallow-burial
selective dissolution and cementation of the organic-rich carbonate muds during anaerobic oxidation of organic matter, as well
as secondary limestone-shale cyclicity caused by anaerobic
methane oxidation, are associated with exceedingly negative d13C
values, as low as 5% to 42% (Coniglio, 1989; Raiswell, 1988).
Therefore, early diagenetic processes in the mostly organic-enriched
carbonate-clayey deposits (TOC typically 40.5% in dark shales, but
420% in the black horizons) are precluded as the driver of the
rhythmic bedding.
In the F–F transition, a possible extreme condensation (or even
a cryptic hiatus?) embracing the basalmost triangularis Zone
should be taken into account, since the (late) Pa. triangularis
occurs right at the base of the Famennian (Racki et al., 2002). On
the other hand, syndepositional and shallow burial depth dissolution in starved sedimentation regimes is also commonly
documented in the Woclumeria limestone, especially by irregularly wavy inter-layer contacts and the abundance of dissolved
aragonite shells (encrustations on cephalopod internal mounds;
Rakocinski, 2011).
The rhythmical alternations between limestone and shale in
the studied intervals of the Kowala section are interpreted to be
primary depositional features, even if diagenetically enhanced
bedding (Bathurst, 1987) is a feature of some packages and event
beds in the succession (e.g. obrutionary trilobite rhythmites; Brett
et al., 2012). The associated facies changes reflect repetitive
variations in carbonate content, in response to changes in the
water column and in the geochemical conditions at the sediment/
water interface. These are most probably driven by climate.
Indeed, terrigenous clay and silt input is highly dependent on
prevailing wind and precipitation patterns, while carbonate
productivity is strongly affected by temperature, salinity, storm
frequency and wave energy (Elrick and Hinnov, 2007). The
observed lithological variations can thus be used as a continuous
high-resolution, 104 yr-scale, record of Late Devonian climate,
highly suitable for cyclostratigraphic purposes. For this reason,
lithological variations are used as the starting point for timeseries analysis and as the foundation for an astronomical cyclostratigraphy. Subsequently, geochemical proxies and their possible
cyclical character are placed in the established cyclostratigraphic
framework and interpreted in terms of climate change.
4.2. Interval I: the Frasnian–Famennian transition
The time–frequency representation (Fig. 2A) and the MTMspectrum (Fig. 2B) are combined to give an overview of the
dominant periodicities in the lithological variations between
shale (coded 0), micrite (coded 2), sparite (coded 3) and calcarenite (coded 4; lithologic log by Bond and Zatoń, 2003, and
D. Bond unpublished data). This ranking scheme is chosen to
emphasize the alternations between shales and limestones
(micrite, sparite and calcarenite). A very strong low-frequency
cycle, characterized by 700–1250 cm thick cycles is highlighted
(Fig. 2A and B). The wavelet-analysis’ time–frequency representation demonstrates that this cycle is particularly strong in the
lowermost Famennian, between 30 and 120 m above the base of
the section. Secondly, a strong and distinct spectral peak occurs
between a 160 and 256 cm period (Fig. 2B). Wavelet analysis
indicates strong spectral power in this frequency range, although
the time–frequency representation is characterized by multiple
bifurcations (splitting or braiding). The interpretation of these
two important cyclical components as the result of an astronomical forcing parameter requires an independent time-control. In
this case, time-control comes from conodont zonation based on
Szulczewski (1995, 1996), Racki and Baliński (1998) and Racki
et al. (2002). The rhenana, linguiformis and triangularis zone
span 81.9 m of section. Consensus on the total duration of these
D. De Vleeschouwer et al. / Earth and Planetary Science Letters 365 (2013) 25–37
29
Fig. 2. Periodicities in the lithological variations of the upper Frasnian–Lower Famennian part of the Kowala section. (A) Time–frequency representation resulting from
wavelet (Morlet) analysis. Red colors indicate the stratigraphic levels (x-axis) and periodicities (y-axis) that are characterized by high spectral power. The whitish bands
and dashed lines indicate the periods associated with 405-kyr (E1) and 100-kyr (E2) eccentricity throughout the section. The bifurcation pattern at the 100-kyr period,
indicative for amplitude modulation by the 405-kyr cycles, is shown by rounded rectangles. (B) MTM-power spectra showing distribution of spectral power with 95%
confidence level (CL). (C) The E1-filtered lithological signal (purple line). (D) The E2-filtered lithological signal (olive colored line) and its amplitude envelope (purple
dashed line). (E) Lithological variations (black solid line). BP¼ Bandpass.
biozones has yet to be reached in the literature. The recalibrated
Devonian timescale of Kaufmann (2006) suggested a combined
duration of 5.8 Myr for these zones, implying an average accumulation rate of 14.12 m/Myr. De Vleeschouwer et al. (2012b)
found that Kaufmann’s (2006) Frasnian chronology overestimated
to duration of the entire stage by 17%, and so it is likely that true
duration is less than 5.8 Myr. In contrast, Bai (1995) indicated that
the rhenana to triangularis zone record only 3.2 Myr of time,
which would imply a much higher average accumulation rate at
Kowala of 25.6 m/Myr. Most likely, the true combined duration
of the rhenana, linguiformis, and triangularis zone lies between
these two end-members. Therefore, it is reasonable to assume an
average accumulation rate of around 20 m/Myr. In this scenario,
the observed 700–1250 cm cycles can be interpreted as the result
of 405-kyr eccentricity forcing and the 160–256 cm cycles as the
result of 100-kyr eccentricity forcing (Table 1). This interpretation
is reinforced by the observed bifurcations in the 2 m cycles,
every 405 kyr, because of the long-eccentricity amplitude modulation of the 100-kyr cycles (indicated by grey circles on Fig. 2A;
Meyers and Hinnov, 2010). However, two problems arise with
this interpretation: [1] at 40 m and 100 m bifurcations in the
2 m cycles are missing (indicated by ‘‘?’’ in Fig. 2A) and [2] in
the lower part of the interval, the low-frequency cycles lack the
strong spectral power that would be expected from a phasecoherent and stable 405-kyr eccentricity forcing. Different interpretations can explain these observations in a valid manner.
However, the astronomical interpretation is maintained as the
listed problems most likely reflect a non-response to orbital
forcing (see Supplementary material for a more detailed discussion). Indeed, the 40 m level coincides exactly to the interval
30
D. De Vleeschouwer et al. / Earth and Planetary Science Letters 365 (2013) 25–37
Fig. 3. Composite plot of the Frasnian–Famennian interval of the Kowala section showing: (A) Lithological variations; (B) The E2-filtered lithological signal (olive colored
line) and its amplitude envelope (purple dashed line); (C) Th/U ratio and its E1-filtered signal (purple dashed line); (D) Weight percentage of potassium and its E-1 filtered
signal (purple dashed line); (E) Carbon isotope record of calcite; (F) Oxygen isotope record of calcite and (G) carbon isotope record of organic matter. (H–K) MTM spectral
plots of Th/U, K, d13C and d18O records respectively (black lines) and their 95% CL (dashed lines).
between the LKE and the UKE recorded by monotonous, clay-rich
and fossil-impoverished micrites. Such a long interval ( 800 cm)
of unchanged lithology is exceptional at Kowala. Consequently,
the loss of low-frequency spectral power in the lower part of this
interval is interpreted as the expression of a climate system that
was disturbed in such a way that the response to orbital forcing
was overpowered. At the 100 m level, an exceptionally long
interval of unchanged lithology suggests weak orbital forcing, or
a climate system that was insensitive to orbital forcing, or a
combination of both. A possible hiatus is considered unlikely
because conodont biostratigraphy and microfacies do not reveal
any large-scale sedimentary discontinuity (Racki et al., 2002) and
the estimated duration between the LKE and UKE ( 800 kyr) is in
good agreement with other estimates (Chen and Tucker, 2003;
Chen et al., 2005).
In Fig. 2C and D respectively, the low- and high-frequency
cycles were filtered out and the amplitude envelope of the highfrequency cycles is calculated using the Hilbert transform. The
amplitude envelope of the 100-kyr eccentricity cycles seems
to mimic the pattern of the 405-kyr cycles, as is expected from
astronomical theory (Laskar et al., 2004). The similarity between
these signals confirms the veracity of the 405-kyr cycles in the
lower part of the section, despite the moderate spectral power. At
the 40 m and 100 m levels, the non-response to orbital forcing
removed the 100-kyr cycles, while the 405-kyr cycle is still
observable. Indeed, a 0.5 Myr non-response is not apparent
in the 405-kyr bandpass filtered record due to the smoothing
inherent in the filtering procedure. The correspondence between
the filtered 405-kyr cycles and the 405-kyr pattern in the
amplitude envelope of the 100-kyr cycles enables the delineation
of 13 complete 405-kyr eccentricity cycles, indicated by purple
bands on Fig. 2. Since these thirteen 405-kyr cycles span 105 m,
a cyclostratigraphically derived average accumulation rate of
19.9 m/Myr can be calculated. Based on this cyclostratigraphic
interpretation, we estimate the time difference between the
Lower and the Upper Kellwasser Event (LKE and UKE respectively)
to include two 405-kyr eccentricity cycles, i.e. 0.8 Myr. This
duration compares well the 10 cycle-sets (100 kyr each) presented by Chen and Tucker (2003) and Chen et al. (2005) between
the 4rd-order sea-level transgressions associated with the LKE of
the early Late rhenana Zone and the UKE of the Late linguiformis
Zone in Guangxi, southern China. Moreover, it agrees well with
the time-difference between the sequence boundary dates of
the two latest Frasnian TR-cycles (i.e. respectively 376.0 and
D. De Vleeschouwer et al. / Earth and Planetary Science Letters 365 (2013) 25–37
31
Table 1
Astronomical interpretation in the upper Frasnian–lower Famennian part of the Kowala section.
Half biostratigr.
acc. rate
Biostratigr. derived
acc. rate
Cyclostratigr. derived
acc. rate
Double biostratigr.
acc. rate
Sed. rate
7.11 m/Myr
14.21 m/Myr
19.9 m/Myr
28.42 m/Myr
Frequency range
700–1250 cm
160–256 cm
1.0–1.8 Myr
225–360 kyr
493–880 kyr
113–180 kyr
352–628 kyr
80–129 kyr
246–440 kyr
56–90 kyr
375.2 Ma; Haq and Schutter, 2008), of which the respective
transgressions are associated with LKE and UKE anoxic deposition
(Bond and Wignall, 2008).
Geochemical proxy data is plotted next to the cyclostratigraphically delineated eccentricity cycles in Fig. 3. The Th/U ratio, a
proxy for the intensity of anoxia, shows a gradual increase
towards more oxic values throughout the section, with important
meter-scale variations superimposed on the long-term trend. The
MTM spectral plot of the Th/U ratio in the lowermost 115 m of the
section (Fig. 3H) demonstrates a moderate peak of spectral power
around 0.00125 cycles/cm, i.e. the frequency at which one would
expect the influence of 405-kyr eccentricity forcing. A more
important spectral peak (495% CL) is observed at frequency
0.0025, indicating 200-kyr climatic fluctuations. Interestingly,
this 200-kyr rhythm is also observed in the carbon and
strontium records of F–F sections from South China (Chen et al.,
2005). The source of this signal might be the 174 kyr amplitude
and frequency modulation of obliquity (Hinnov, 2000). However,
more high-resolution records are needed to confirm this link. The
spectral curve also clearly rises around frequency 0.005 cycles/
cm, i.e. the frequency of 100-kyr cycles, but the resolution of the
Th/U record is too low to firmly distinguish these cycles. A similar
pattern is recorded by the concentration of potassium (K wt%;
Fig. 3I), which is mainly a proxy for the detrital clay fraction. The
dashed lines in Fig. 3C and D represent the 405-kyr variations of
the Th/U ratio and K concentration respectively. In particular, the
K concentration follows the 405-kyr cycles that were delineated
based on lithology. Consequently, it is suggested that stronger
lithological variations occur during periods of intense input of
detrital clays. The Th/U data in Fig. 3C indicates that during the
UKE, oxygen levels remained depleted for 400 kyr. This is
in agreement with previous estimates based on pyrite framboids
(Bond et al., 2004).
Carbon and oxygen isotopes were measured at decimeter
resolution, through a 16 m interval that straddles the F–F boundary. Joachimski et al. (2001) lower-resolution carbon isotope
record from another Kowala section is plotted for comparison.
Both carbon and oxygen isotope records show a decreasing trend
leading up to the UKE and exhibit intense positive excursions
associated with the event. The d13Corg record from Joachimski
et al. (2001) similarly shows positive excursions related with the
LKE and UKE. These authors interpreted the positive excursions in
d13C and d13Corg as the result of higher photoautotrophic production, triggered by enhanced continental nutrient-supply parallel
with short-term sea-level rise and upwelling of anoxic waters
that invaded formerly well-oxygenated environments (see
another interpretation in Kazmierczak et al. 2012). The consequent burial of large amounts of organic carbon is expected to
result in a drop of atmospheric pCO2. Joachimski et al. (2001)
failed to observe a change in the photosynthetic carbon isotope
fractionation (ep), as would be expected after a significant pCO2.drop. Indirect evidence that a pCO2 drop is actually associated
with the UKE comes from our d18O record (Fig. 3F) that indicates
rapid cooling during the UKE (1% positive excursionE4 to 5 1C
cooling), similar to the oxygen isotope shift documented in
conodont apatites from this same succession (see Fig. 5 in
Astronomical
Interpretation
405-kyr ecc.
100-kyr ecc.
Joachimski et al., 2009). Putting these isotopic records in our
cyclostratigraphic framework, it suggests that for the carbon cycle,
after the þ3% UKE excursion in d13Corg, more than 600 kyr were
needed before a more or less steady state was reached (Fig. 3G).
The MTM spectral curves of our high-resolution isotope data
show sharp and strong ( 495% CL) spectral peaks (Fig. 3J and K):
d13C and d18O display an important spectral peak around frequency 0.011 cycles/cm. Following our astronomical interpretation, these cycles have a duration of 47 kyr, and cannot be
directly attributed to an astronomical forcing factor. In contrast,
the elevated spectral power observed for both d13C and d18O
around frequency 0.027 cycles/cm occurs exactly at the frequency
at which one expects the influence of 18-kyr precession. The
recognition of this 18-kyr period in environmental proxies such
as d13C and d18O demonstrate the importance of precessional
forcing on the Late-Devonian tropical climate. Indeed, in modern
tropical environments precession is the most important driver of
104-yr scale climate change, most often by influencing monsoonal
patterns and intensity (Kutzbach, 1981; Kutzbach and Liu, 1997;
Tuenter et al., 2003). In comparison with the modern world,
Devonian tropical climate potentially exhibited an even more
intense monsoonal circulation (Streel et al., 2000), characterized
by seasonally wet-and-dry climates (Cecil, 1990), on which
precessional variations undoubtedly had a significant influence
(De Vleeschouwer et al., 2012a). Despite the fact that 20 Myr
separate the present study from the Belgian section analyzed by
De Vleeschouwer et al. (2012a), they share a similar paleogeographic setting on the southeastern margin of Laurussia, at a
tropical latitude. Thus, a comparable paleoclimatic model can be
applied to interpret the precessional cycles observed in both sections.
The presence of a precessional rhythm in Frasnian–Famennian
climate change at Kowala also explains why eccentricity-related
periods show up in the shale-limestone alternations. Eccentricity
only marginally influences insolation, but it appears as the
envelope of the precession parameter (Muller and MacDonald,
2000).
4.3. Interval II: Devonian–Carboniferous interval
The results of MTM and wavelet analyses on the upper
Famennian to lower Tournaisian lithological variations are presented in Fig. 4A. This part of the Kowala section is characterized
by shales (coded 0), marls (coded 1), nodular limestones (coded 2)
and micrite (coded 2.5; Marynowski and Filipiak, 2007;
Marynowski et al., 2010; Racka et al., 2010). This scheme is
thought to reflect relative changes in clay-content. In the upper
part of the studied interval, an important low-frequency component is suggested at periodicities between 170 and 390 cm. A very
strong decimeter-scale cycle is denoted, characterized by a
40–80 cm period. Wavelet analysis suggests that this cycle is persistent through the entire Famennian part of the section, though
characterized by multiple bifurcations, and that it disappears
upon reaching the Tournaisian. Between 1600 and 2900 cm,
the importance of a high-frequency cycle with a 8–12 cm periodicity is demonstrated by both MTM and wavelet analysis. The
interpretation of the observed cyclicity requires independent
32
D. De Vleeschouwer et al. / Earth and Planetary Science Letters 365 (2013) 25–37
Fig. 4. Periodicities in the lithological variations of the Upper Famennian–Lower Tournaisian part of the Kowala section. (A) Time–frequency representation resulting from
wavelet (Morlet) analysis. Red colors indicate the stratigraphic levels (x-axis) and periodicities (y-axis) that are characterized by high spectral power. The whitish bands
and dashed lines indicate the periods associated with 405-kyr (E1), 100-kyr (E2) eccentricity and precession throughout the section. The bifurcation pattern at the 100-kyr
period, indicative for amplitude modulation by the 405-kyr cycles, is shown by rounded rectangles. (B) MTM-power spectra showing distribution of spectral power with
95% confidence level (CL). (C) The E1-filtered lithological signal (purple line). (D) The E2-filtered lithological signal (olive colored line) and its amplitude envelope (purple
dashed line). (E) Lithological variations (black solid line). BP ¼Bandpass. (For interpretation of the references to color in this figure legend, the reader is referred to the web
version of this article.)
biostratigraphic time-control: the Annulata Event, recorded in the
lower part of this interval occurs within the uppermost trachytera
Zone. According to Kaufmann’s (2006) chronology, 4.7 Myr
separate the trachytera conodont zone and the D–C boundary.
According to Becker et al. (2012, Table 22.2), the same interval
spans 4.1 Myr. Bai (1995) again suggests a shorter duration for
the same interval (2.9 Myr), but in this case, the true combined
duration of the considered biozones is most probably much closer
to the estimates of Kaufmann (2006) and Becker et al. (2012),
given the multiplicity of U–Pb ID-TIMS absolute ages available for
the latest Famennian (Richards et al., 2002; Streel, 2000; Trapp
et al., 2004; Tucker et al., 1998). In the studied section, 25 m span
the interval between the Annulata Event and the Famennian/
Tournaisian boundary, implying an average accumulation rate of
5.7 m/Myr. This value is in stark contrast to the 4-times higher
accumulation rate found at the Frasnian–Famennian interval of
the studied section. Evidence for slower accumulation also comes
from diverse sclerobionts that utilized the shells of dead cephalopods as a hard substrate for their settlement (uppermost
Famennian; Rakocinski, 2011) and from corrosional surfaces
related to syndepositional shallow burial-depth dissolution
(Berkowski, 2002). Also, as the section becomes much more shale
dominated, a slower accumulation could be expected, possibly in
response to the shutdown in primary productivity following
various extinction events (i.e. a less productive carbonate factory)
and to starved sedimentation regimes during the Famennian
global climate cooling (Caputo et al., 2008; Isaacson et al., 2008;
Joachimski et al., 2009; Streel et al., 2000). A 5.7 m/Myr
accumulation rate gives rise to the interpretation of the lowfrequency cycles as the result of 405-kyr eccentricity forcing
(Table 2). The 48–75 cm cycles match 100-kyr eccentricity cycles
and the 8–12 cm cycles can be seen as the result of precessional
forcing (Table 2). The extremely low spectral power in the 405-kyr
frequency range in the first portion of the record questions this
interpretation, as the 405-kyr eccentricity cycle is remarkably
phase-coherent and stable. However, the strong bifurcating pattern in the 100-kyr frequency range appears to be a pristine shorteccentricity signal. The lack of spectral power between 0 and
1400 cm is seen as the result of the non-linear recording of
climate forcing. Fig. 4C and D shows the bandpass filters at the
frequencies of 405-kyr and 100-kyr eccentricity respectively. The
amplitude envelope of the 100-kyr cycles, calculated using the
Hilbert transform, appears to bundle the 100-kyr cycle in groups
of four, as expected from astronomical theory. However, at
several stratigraphic positions, a mismatch between the amplitude modulation of the 100-kyr cycles and the 405-kyr cycles is
observed. This phase-lagged amplitude modulation does not
undermine the astronomical interpretation, as phase and amplitude distortions can be induced by the non-linear translation of
sea-level change to offshore sediment flux (Laurin et al., 2005).
Mainly based on the amplitude envelope of the 100-kyr cycles,
12 complete 405-kyr eccentricity cycles were delineated and are
D. De Vleeschouwer et al. / Earth and Planetary Science Letters 365 (2013) 25–37
33
Table 2
Astronomical interpretation in the upper Famennian–lower Trounaisian part of the Kowala section.
Half biostratigr.
acc. rate
Biostratigr. derived
acc. rate
Double biostratigr.
acc. rate
Sed. rate
2.85 m/Myr
5.7 m/Myr
11.4 m/Myr
Frequency range
170–390 cm
40–80 cm
8–12 cm
0.6–1.4 Myr
140–280 kyr
28–42 kyr
298–684 kyr
70–140 kyr
14–21 kyr
150–342 kyr
35–70 kyr
7–10 kyr
Astronomical
interpretation
405-kyr ecc.
100-kyr ecc.
18-kyr prec.
Fig. 5. Periodicities in the lithological variations of the upper Famennian part of the Kowala section, between 1500 and 3000 cm. This portion is analyzed in detail because
it is characterized by high-frequency lithological variations. (A) Time–frequency representation resulting from wavelet (Morlet) analysis. Red colors indicate the
stratigraphic levels (x-axis) and periodicities (y-axis) that are characterized by high spectral power. The whitish bands and dashed lines indicate the periods associated
with 100-kyr (E2) eccentricity and precession throughout the studied interval. The strong bifurcation pattern at the precessional period, indicative for amplitude
modulation by the 100-kyr cycles is shown by rounded rectangles. (B) MTM-power spectra showing distribution of spectral power and 95% confidence level (CL). (C) The E2-filtered
lithological signal (olive colored line). (D) The precession-filtered lithological signal (green line) and its amplitude envelope (olive colored dashed line).
(E) Lithological variations (black solid line). BP¼ Bandpass. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
indicated by purple bands in Fig. 4. Together, these 12 cycles
(4.86 Myr) span 2650 cm, resulting in a refined accumulation rate
of 5.45 cm/kyr.
The uppermost Famennian part of the Kowala section (1500–
3000 cm) is characterized by cm-scale alternations between limestones and shales. These high-frequency rhythmites were analyzed in more detail to assess the astronomical forcing by
precession (Fig. 5). The 40–80 cm cycles are associated with
100-kyr eccentricity while the 8–12 cm cycles are interpreted as
the result of precessional forcing. A band-pass filter centered on
the frequency of precession shows that the amplitude of the
precessional cycles varies significantly (Fig. 5D). The amplitude
envelope of the precessional cycles was calculated using the
Hilbert transform and compares favorably with the 100-kyr
eccentricity cycles. Since the amplitude modulation of precession
by eccentricity is expected from astronomical theory, the astronomical interpretation of the different observed cycles is reinforced. Moreover, the dominance of precession may also explain
amplitude modulation mismatches between long and short
eccentricity cycles, discussed above. Indeed, an important mismatch occurs in that part of the section where precession-related
lithological variations can be firmly grouped in 100-kyr bundles,
but where grouping in 405-kyr ‘‘super-bundles’’ is much less
obvious.
The geochemical data available for this interval is not suitable
for time-series analysis: for the isotopic proxies (Fig. 6G–I),
34
D. De Vleeschouwer et al. / Earth and Planetary Science Letters 365 (2013) 25–37
Fig. 6. Composite plot of the Famennian–Tournaisian interval of the Kowala section showing: (A) Lithological variations; (B) The E2-filtered lithological signal (olive
colored line) and its amplitude envelope (purple dashed line); (C) Magnetic Susceptibility; (D) Weight percentage of CaCO3; (E) Weight percentage of total organic matter
(TOC); (F) Weight percentage of total sulfur (TS); (G) Carbon isotope record of calcite; (H) Oxygen isotope record of calcite and (I) Carbon isotope record of organic matter
(note that only organic carbon data are available for black shale horizons). Blue and red arrows designate climatic cooling and warming respectively. (For interpretation of
the references to color in this figure legend, the reader is referred to the web version of this article.)
resolution and/or dataset-length are too low; for magnetic susceptibility (Fig. 6C), unequal spacing of data makes MTM and
wavelet analysis meaningless. The TOC, TS and CaCO3 concentrations simply reflect lithological variation between shale and
limestone, and therefore do not add to the paleoenvironmental
information that is already contained in the lithological variations
themselves. Excessive burial of organic matter, expressed by the
Dasberg, Kowala and Hangenberg black shales, is clearly indicated
by increased TOC, TS and decreased CaCO3 concentrations
(Fig. 6D–F). All anoxic shales in the studied interval have been
interpreted to be the result of repeated transgressive pulses,
accompanied by the blooming of primary producers (Joachimski
et al., 2001; Kaiser et al., 2008; Marynowski et al., 2010, 2012;
Racka et al., 2010). Enhanced bioproduction is reflected by
positive d13Corg excursions of about 4%, up to values of 25%
in the limestone enclosing the shales (Fig. 6I). Oxygen isotopes
suggest a shift towards warmer conditions during anoxic shale
deposition (Fig. 6H), which is consistent with the transgressive
trend observed during interglacial episodes (Racka et al., 2010).
Rising sea-levels resulted in the flooding of the shallow-water
environment by nutrient-rich but oxygen-depleted waters that
originated from the deeper-shelf (see review in Hallam and
Wignall, 1999). Moreover, relatively warm conditions were most
probably associated with more moist periods. Therefore,
enhanced chemical weathering and continental runoff may
have contributed to eutrophication in the oceans, stimulation of
primary production, oxygen consumption and therefore marine
anoxia (Algeo and Scheckler, 1998). Based on the delineated 405kyr cycles, it is possible to estimate the time-difference between
the well-described black shales that occur in this section. Five and
a half 405-kyr cycles, i.e. 2.2 Myr, separate the Annulata Shale
from the Dasberg Shale. Almost the exact same amount of time
separates the Dasberg Shale from the Hangenberg Shale (Fig. 6).
These values accord with the sequence boundary ages reported by
Haq and Schutter (2008), who estimate a duration of 2.4 and
1.8 Myr for the corresponding 3rd-order sea-level changes. Interestingly, a 2.4 Myr cycle is present in the Earth’s eccentricity.
The same correspondence between third-order sea-level changes
and long-period 2.4 Myr eccentricity has been reported for the
Mesozoic greenhouse world (Boulila et al., 2011). Hence, one
can hypothesize an astronomical influence on the timing of the
development of transgressive–regressive (TR) cycles and the
associated deposition of anoxic black shales (as has already been
suggested for Cretaceous anoxic shales; e.g. Lanci et al., 2010;
Mitchell et al., 2008). During greenhouse periods such as the
Mesozoic and the Devonian, thermoeustasy or variations in
D. De Vleeschouwer et al. / Earth and Planetary Science Letters 365 (2013) 25–37
35
Table 3
Comparison of accumulation rate to the time-scale of Lithological cycles at different Devonian and Early Carboniferous sections.
Lower Mississippian Banff Fm.
Lower Mississippian Lodgepole Fm.
Upper Devonian West Range Limestone
Middle Devonian Denay Fm.
Upper Devonian, Guangxi, South-China
Upper Devonian Kowala section, Interval I
Upper Devonian Kowala section, Interval II
Accumulation rate
(cm/kyr)
Duration of studied interval
(kyr)
Time-scale of lithological cycles
Ref.
8.9–14
6.4–15
5.9–12.7
3.1–5.8
1–3
1.75–3
0.5–0.75
53–83
242–567
45–96
104–194
2000–2400
6000
5300
Millenial
Millenial
Millenial
Millenial
Millenial and Milanković
Milanković
Milanković
Elrick and Hinnov (2007)
Elrick and Hinnov (2007)
Elrick and Hinnov (2007)
Elrick and Hinnov (2007)
Chen and Tucker (2003)
This study
This study
Fm¼ Formation.
storage of groundwater and lakes can account for a maximum of
10 m of sea-level change (Jacobs and Sahagian, 1993; Schulz and
Schafer-Neth, 1997). Yet, it is clear that the amplitude of sea-level
change in the uppermost Devonian was at least 4 to 5 times
higher (Haq and Schutter, 2008). Therefore, we suggest that the
presence and extent of small continental ice sheets on (western)
Gondwana could be strongly influenced by eccentricity. This
assumption is supported by Late Devonian glacial marine deposits
in Peru, Bolivia and Brazil that record important regional glaciations on the western margin of Gondwana at mid-to-high latitudes (40–501S; Blakey, 2008; Caputo et al., 2008; Isaacson et al.,
2008). During periods of low eccentricity, austral summer insolation at the high latitudes of the Gondwana continent would not
have reached high values, which gives these small continental
ice sheets a good chance to survive summer and to continue to
grow during the following winter. Thereby, the Earth’s climate is
gradually cooled (phase 1 in the oxygen isotope record in Fig. 6H).
In contrast, high eccentricity triggers maximal austral summer
insolation when o ¼901 (perihelion in December), causing important melting and ice sheet collapse, associated sea-level rise and
global warming (phase 2). Global warming combined with late
Famennian atmospheric O2 levels exceeding the 13% threshold for
the occurrence of widespread wildfires (Marynowski and Filipiak,
2007; Marynowski et al., 2010; 2012; Rimmer et al., 2004), led to
the establishment of optimal conditions for increased primary
production and bottom-water anoxia. The enhanced burial of
organic carbon reduces the C concentration of the ocean’s surface
layer, that is compensated by additional dissolution of CO2 from
the atmosphere. The climatic cooling resulting from the drawdown of atmospheric CO2 is reflected by temporarily increasing
d18O values during black-shale deposition (phase 3). This coolingphase appears to be short-lived ( 200 kyr) and is probably
counteracted by the astronomical configuration favoring climatic
warming (phase 4). Several hundred of kyrs later, eccentricity and
resultant maximal summer insolation in Gondwana weakens,
enabling ice to survive summer and accumulate once more
(phase 5).
Compared to the Annulata, Dasberg and Hangenberg Shales,
the Kowala anoxic shale is a much thinner organic-rich layer.
Although it occurs exactly halfway between the Dasberg and
Hangenberg Shales, it was probably not triggered by astronomical
forcing but rather as the result of regional tectonic activity.
Indeed, the Kowala Shale is geographically limited to the
Checiny–Zbrza intrashelf basin and is only recorded in the studied
section (Marynowski and Filipiak, 2007).
4.4. On the origins of limestone-shale/marl alternations
The limestone-shale/marl alternations in both studied intervals of the Kowala section demonstrate clear orbital signatures
of precession and eccentricity and it can be assumed that
facies changes reflect a climatic signal. Carbonate-rich layers
(limestones) are interpreted to form under warm and dry conditions with less runoff from land, reduced weathering, less frequent storm events (i.e. low dilution by terrigenous influx) and
enhanced carbonate productivity. Carbonate-poor layers are
thought to result from a lower carbonate productivity under
cooler climatic conditions, associated with wetter conditions
and enhanced clastic input.
However, the exact same response mechanism is used to
explain millennial-scale paleoclimate rhythms, expressed by
individual limestone-shale couplets (Chen and Tucker, 2003;
Elrick and Hinnov, 2007), despite the order of magnitude difference between the sub-Milanković’s and Milanković time-scale.
To address this discrepancy, accumulation rate, duration of the
studied interval and time-scales of the distinguished paleoclimatological rhythms in lithology are compared to the Devonian and
lower Carboniferous sections in Elrick and Hinnov (2007), the F–F
section in Chen and Tucker (2003) and both intervals from this
study (Table 3). From this table it is apparent that a high
accumulation rate (several cm’s/kyr) is needed for lithological
variations to record millennial-scale climate change. In sections
like Kowala, where accumulation occurs at an order of magnitude
more slowly, limestone-marl/shale alternations represent Milanković-scale climate change. In summary, terrigenous input to the
basin and carbonate productivity vary at both time-scales, but
only in rapidly accumulating settings millenial-scale rhythmicity
can be registered as lithological variations. On the other hand,
only sections characterized by a sufficiently long duration of
deposition (e.g. Kowala) are suitable for the detection of Milanković-scale cyclicity. Orbital forcing is known to be a powerful
driver of climate change, whereas the driver of millennial-scale
climate change remains a mystery. The present research reinforces this mystery since the same type of lithological rhythm,
driven by similar climate changes, occurs at time scales with an
order of magnitude difference.
5. Conclusions
The lithological variations between limestones, shales and
marls in the Kowala section (Holy Cross Mountains, Poland)
record astronomical forcing. Precession and eccentricity control
the cyclical environmental changes that are responsible for the
lithological rhythms. The expression of obliquity in the Kowala
sedimentary record is weak and intermittent, as is expected
during greenhouse worlds (Boulila et al., 2011). The association
of lithological alternations with astronomical parameters enables
the construction of a cyclostratigraphic framework that suggests a
800 kyr time-gap between the LKE and the UKE and a 400 kyr
duration for low oxygen levels during the UKE. Isotopic data
(d13C, d18O and d13Corg) across the UKE exhibit a strong positive
shift associated with paleoenvironmental change and destabilization of the carbon cycle. Despite the severe environmental
36
D. De Vleeschouwer et al. / Earth and Planetary Science Letters 365 (2013) 25–37
perturbations associated with the UKE, the imprint of precession
in these isotopic records is clear. The sharp positive isotope shift
in d13C and d13Corg is interpreted to be the result of enhanced
primary production, promoted by increased nutrient-supply coeval with sea-level rise and upwelling of anoxic waters. The
consequent burial of large amounts of organic carbon resulted
in a drop of atmospheric pCO2, as is suggested by the 1% d18O
positive excursion (4 to 5 1C cooling). A 600 kyr recovery for the
carbon cycle after the þ3% UKE excursion is suggested by the
cyclostratigraphic framework. Time-differences between the
Annulata, Dasberg, Kowala and Hangenberg anoxic black shales
are estimated at 2.2, 1.2 and 1.2 Myr respectively. It is proposed
that the formation of the late Famennian Annulata, Dasberg and
Hangenberg Shales is related to maxima of the 2.4-Myr eccentricity cycle. These astronomical configurations would have provoked sea-level rise by triggering the collapse of small continental
Gondwanan ice-sheets. Thereby, an explicit link between astronomical forcing, eustatic transgression and marine anoxia for the
Late Devonian is proposed.
Acknowledgments
This work is supported by a Ph.D. fellowship awarded by the
Research Foundation – Flanders (FWO) assigned to DDV, and a
Ph.D. fellowship of the Ministry of Science and Higher Education –
Poland (Grant N N307 4272 34) assigned to MR. We would like to
thank P. Wignall, M. Zatoń, L. Marynowski and A. Pisarzowska for
participation in the field and geochemical analyses which were
partly funded by a NERC Ph.D. studentship to DB. M. Joachimski
kindly provided unpublished isotope data. PC thanks the FWO
(project G.A078.11), Hercules foundation and VUB research funds
for their contribution. The authors thank Dr. Stephen R. Meyers
and Dr. Maya Elrick for their constructive reviews and Dr. A.-C. Da
Silva for positive discussions.
Appendix A. Supplementary materials
Supplementary data associated with this article can be found in
the online version at http://dx.doi.org/10.1016/j.epsl.2013.01.016.
References
Algeo, T.J., Scheckler, S.E., 1998. Terrestrial–marine teleconnections in the Devonian: links between the evolution of land plants, weathering processes, and
marine anoxic events. Philos. Trans. R. Soc. B-Biol. Sci. 353, 113–128.
Bai, S.L., 1995. Milankovitch cyclicity and time scale of the middle and upper
Devonian. Int. Geol. Rev. 37, 1109–1114.
Bathurst, R.G.C., 1987. Diagenetically enhanced bedding in Argillaceous platform
limestones—stratified cementation and selective compaction. Sedimentology
34, 749–778.
Becker, R.T., Gradstein, F.M., Hammer, O., 2012. The Devonian Period. In:
Gradstein, F.M., Ogg, J.G., Schmitz, M.D., Ogg, G. (Eds.), The Geologic Time
Scale 2012. Elsevier, Amsterdam, pp. 559–601.
Belka, Z., 1990. Thermal maturation and burial history from Conodont color
alternation data, Holy Cross Mountains, Poland. Cour. Forsch.-Inst. Senckenberg 118, 241–251.
Berkowski, M., 2002. Famennian Rugosa and Heterocorallia from Southern Poland.
Palaeontol. Pol. 61, 3–88.
Blakey, R.C., 2008. Gondwana paleogeography from assembly to breakup-A
500 m.y. odyssey. Geological Society of America Special Papers 441, 1-28.
Bond, D., Zatoń, M., 2003. Gamma-ray spectrometry across the Upper Devonian
basin succession at Kowala in the Holy Cross Mountains (Poland). Acta Geol.
Pol. 53, 93–99.
Bond, D., Wignall, P.B., Racki, G., 2004. Extent and duration of marine anoxia
during the Frasnian-Famennian (Late Devonian) mass extinction in Poland,
Germany, Austria and France. Geol. Mag. 141, 173–193.
Bond, D.P.G., Wignall, P.B., 2008. The role of sea-level change and marine anoxia
in the Frasnian–Famennian (Late Devonian) mass extinction. Palaeogeogr.
Palaeoclimatol. Palaeoecol. 263, 107–118.
Boulila, S., Galbrun, B., Miller, K.G., Pekar, S.F., Browning, J.V., Laskar, J., Wright, J.D.,
2011. On the origin of Cenozoic and Mesozoic ‘‘third-order’’ eustatic
sequences. Earth-Sci. Rev. 109, 94–112.
Brett, C.E., Zambito, J.J., Schindler, E., Becker, R.T., 2012. Diagenetically-enhanced
trilobite obrution deposits in concretionary limestones: the paradox of
‘‘rhythmic events beds’’. Palaeogeogr. Palaeoclimatol. Palaeoecol.
Caputo, M.V., de Melo, J.H.G., Streel, M., Isbell, J.L., 2008. Late devonian and early
carboniferous glacial records of South America. Geological Society of America
Special Papers 441, 161–173.
Cecil, C.B., 1990. Paleoclimate controls on stratigraphic repetition of chemical and
siliciclastic rocks. Geology 18, 533–536.
Chen, D.Z., Tucker, M.E., 2003. The Frasnian–Famennian mass extinction: insights
from high-resolution sequence stratigraphy and cyclostratigraphy in South
China. Palaeogeogr. Palaeoclimatol. Palaeoecol. 193, 87–111.
Chen, D.Z., Qing, H.R., Li, R.W., 2005. The Late Devonian Frasnian–Famennian (F/F)
biotic crisis: insights from delta C-13(carb), delta C-13(org) and Sr-87/Sr-86
isotopic systematics. Earth Planet. Sci. Lett. 235, 151–166.
Coniglio, M., 1989. Neomorphism and cementation in ancient deep-water limestones, Cow Head Group (Cambro-Ordovician), Western Newfoundland,
Canada. Sediment. Geol. 65, 15–33.
Crick, R.E., Ellwood, B.B., Feist, R., El Hassani, A., Schindler, E., Dreesen, R., Over, D.J.,
Girard, C., 2002. Magnetostratigraphy susceptibility of the Frasnian/Famennian boundary. Palaeogeogr. Palaeoclimatol. Palaeoecol. 181, 67–90.
De Vleeschouwer, D., Da Silva, A.C., Boulvain, F., Crucifix, M., Claeys, P., 2012a.
Precessional and half-precessional climate forcing of Mid-Devonian monsoonlike dynamics. Clim. Past 8, 337–351.
De Vleeschouwer, D., Whalen, M.T., Day, J.E., Claeys, P., 2012b. Cyclostratigraphic
calibration of the Frasnian (Late Devonian) time scale (Western Alberta,
Canada). Geol. Soc. Am. Bull..
Dzik, J., 2006. The Famennian ‘‘Golden Age’’ of conodonts and ammonoids in the
Polish part of the Variscan sea. Palaeontol. Pol. 63, 1–359.
Elrick, M., Hinnov, L.A., 2007. Millennial-scale paleoclimate cycles recorded in
widespread Palaeozoic deeper water rhythmites of North America. Palaeogeogr. Palaeoclimatol. Palaeoecol. 243, 348–372.
Elrick, M., Berkyova, S., Klapper, G., Sharp, Z., Joachimski, M., Fryda, J., 2009.
Stratigraphic and oxygen isotope evidence for My-scale glaciation driving
eustasy in the Early-Middle Devonian greenhouse world. Palaeogeogr. Palaeoclimatol. Palaeoecol. 276, 170–181.
Filipiak, P., 2009. Lower Famennian phytoplankton from the Holy Cross Mountains, Central Poland. Rev. Palaeobot. Palynol. 157, 326–338.
Filipiak, P., Racki, G., 2010. Proliferation of abnormal palynoflora during the endDevonian biotic crisis. Geol. Q. 54, 1–14.
Fischer, A.G., Bottjer, D.J., 1991. Orbital forcing and sedimentary sequences.
J. Sediment. Petrol. 61, 1063–1069.
Garcia-Alcalde, J.L., Ellwood, B.B., Soto, F., Truyols-Massoni, M., Tomkin, J.H., 2012.
Precise timing of the Upper Taghanic Biocrisis, Geneseo Bioevent, in the
Middle–Upper Givetian (Middle Devonian) boundary in Northern Spain using
biostratigraphic and magnetic susceptibility data sets. Palaeogeogr. Palaeoclimatol. Palaeoecol. 313–314, 26–40.
Ghil, M., Allen, M.R., Dettinger, M.D., Ide, K., Kondrashov, D., Mann, M.E.,
Robertson, A.W., Saunders, A., Tian, Y., Varadi, F., Yiou, P., 2002. Advanced
spectral methods for climatic time series. Rev. Geophys. 40, 1–41.
Giles, P.S., 2012. Low-latitude Ordovician to Triassic brachiopod habitat temperatures (BHTs) determined from delta O-18([brachiopod) (calcite]): a cold hard
look at ice-house tropical oceans. Palaeogeogr. Palaeoclimatol. Palaeoecol. 317,
134–152.
Giorgioni, M., Weissert, H., Bernasconi, S.M., Hochuli, P.A., Coccioni, R., Keller, C.E.,
2012. Orbital control on carbon cycle and oceanography in the mid-Cretaceous
greenhouse. Paleoceanography 27, 12.
Hallam, A., Wignall, P.B., 1999. Mass extinctions and sea-level changes. Earth-Sci.
Rev. 48, 217–250.
Haq, B.U., Schutter, S.R., 2008. A chronology of Paleozoic sea-level changes. Science
322, 64–68.
Hinnov, L.A., 2000. New perspectives on orbitally forced stratigraphy. Annu. Rev.
Earth Planet. Sci. 28, 419–475.
House, M.R., 1991. Devonian sedimentary microrhythms and a Givetian time scale.
Proc. Ussher Soc. 7, 392–395.
House, M.R., 1995. Devonian precessional and other signatures for establishing a
Givetian timescale. In: House, M.R., Gale, A.S. (Eds.), Orbital forcing timescales
and cyclostratigraphy. Geological Society of London. Special Publications,
London, pp. 37–49.
House, M.R., Gale, A.S., 1995. Orbital forcing timescales and cyclostratigraphy.
Geological Society of London. Special Publications, London, p. 204.
House, M.R., 2002. Strength, timing, setting and cause of mid-Palaeozoic extinctions. Palaeogeogr. Palaeoclimatol. Palaeoecol. 181, 5–25.
Isaacson, P.E., Diaz-Martinez, E., Grader, G.W., Kalvoda, J., Babek, O., Devuyst, F.X.,
2008. Late Devonian–earliest Mississippian glaciation in Gondwanaland and
its biogeographic consequences. Palaeogeogr. Palaeoclimatol. Palaeoecol. 268,
126–142.
Jacobs, D.K., Sahagian, D.L., 1993. Climate-induced fluctuations in sea-level during
non-glacial times. Nature 361, 710–712.
Joachimski, M.M., Ostertag-Henning, C., Pancost, R.D., Strauss, H., Freeman, K.H.,
Littke, R., Damste, J.S.S., Racki, G., 2001. Water column anoxia, enhanced
productivity and concomitant changes in delta(13)C and delta(34)S across the
Frasnian–Famennian boundary (Kowala Holy Cross Mountains/Poland). Chem.
Geol. 175, 109–131.
D. De Vleeschouwer et al. / Earth and Planetary Science Letters 365 (2013) 25–37
Joachimski, M.M., Pancost, R.D., Freeman, K.H., Ostertag-Henning, C., Buggisch, W.,
2002. Carbon isotope geochemistry of the Frasnian–Famennian transition.
Palaeogeogr. Palaeoclimatol. Palaeoecol. 181, 91–109.
Joachimski, M.M., Breisig, S., Buggisch, W., Talent, J.A., Mawson, R., Gereke, M.,
Morrow, J.R., Day, J., Weddige, K., 2009. Devonian climate and reef evolution:
insights from oxygen isotopes in apatite. Earth Planet. Sci. Lett. 284, 599–609.
Kaiser, S.I., Steuber, T., Becker, R.T., 2008. Environmental change during the Late
Famennian and Early Tournaisian (Late Devonian–Early Carboniferous): implications from stable isotopes and conodont biofacies in southern Europe. Geol.
J. 43, 241–260.
Kaufmann, B., 2006. Calibrating the Devonian Time Scale—a synthesis of U–Pb
ID-TIMS ages and relative, time-linear conodont stratigraphy. Earth-Sci. Rev. 76,
175–190.
Kazmierczak, J., Kremer, B., Racki, G., 2012. Late Devonian marine anoxia
challenged by benthic cyanobacterial mats. Geobiology 10, 371–383.
Kuiper, K.F., Deino, A., Hilgen, F.J., Krijgsman, W., Renne, P.R., Wijbrans, J.R., 2008.
Synchronizing rock clocks of Earth history. Science 320, 500–504.
Kutzbach, J.E., 1981. Monsoon climate of the Early Holocene—climate experiment
with the earths orbital parameters for 9000 years ago. Science 214, 59–61.
Kutzbach, J.E., Liu, Z., 1997. Response of the African monsoon to orbital forcing and
ocean feedbacks in the middle Holocene. Science 278, 440–443.
Lanci, L., Muttoni, G., Erba, E., 2010. Astronomical tuning of the Cenomanian
Scaglia Bianca formation at Furlo, Italy. Earth Planet. Sci. Lett. 292, 231–237.
Laskar, J., Robutel, P., Joutel, F., Gastineau, M., Correia, A.C.M., Levrard, B., 2004. A
long-term numerical solution for the insolation quantities of the Earth. Astron.
Astrophys. 428, 261–285.
Laurin, J., Meyers, S.R., Sageman, B.B., Waltham, D., 2005. Phase-lagged amplitude
modulation of hemipelagic cycles: a potential tool for recognition and analysis
of sea-level change. Geology 33, 569–572.
Lourens, L.J., Becker, J., Bintanja, R., Hilgen, F.J., Tuenter, E., van de Wal, R.S.W.,
Ziegler, M., 2010. Linear and non-linear response of late Neogene glacial cycles
to obliquity forcing and implications for the Milankovitch theory. Quat. Sci.
Rev. 29, 352–365.
Malec, J., 1995. Devonian/Carboniferous boundary, Development of the Variscan
Basin and Epi-Variscan Cover at the Margin of the East European Platform
(Pomerania, Holy Cross Mts., Kraków Upland): Guide to Excursion A2, August
23–August 27, 1995. In: Proceedings of the XIII International Congress on
Carboniferous–Permian (XIII ICC-P), August 28–September 2, 1995, Kraków,
Poland. Państwowy Instytut Geologiczny.
Marynowski, L., Czechowski, F., Simoneit, B.R.T., 2001. Phenylnaphthalenes and
polyphenyls in Palaeozoic source rocks of the Holy Cross Mountains, Poland.
Org. Geochem. 32, 69–85.
Marynowski, L., Filipiak, P., 2007. Water column euxinia and wildfire evidence
during deposition of the Upper Famennian Hangenberg event horizon from
the Holy Cross Mountains (central Poland). Geol. Mag. 144, 569–595.
Marynowski, L., Rakocinski, M., Zatoń, M., 2007. Middle Famennian (Late Devonian) interval with pyritized fauna from the Holy Cross Mountains (Poland):
Organic geochemistry and pyrite framboid diameter study. Geochem. J. 41,
187–200.
Marynowski, L., Filipiak, P., Zatoń, M., 2010. Geochemical and palynological study
of the Upper Famennian Dasberg event horizon from the Holy Cross Mountains (central Poland). Geol. Mag. 147, 527–550.
Marynowski, L., Rakocinski, M., Borcuch, E., Kremer, B., Schubert, B.A., Jahren, A.H.,
2011. Molecular and petrographic indicators of redox conditions and bacterial
communities after the F/F mass extinction (Kowala, Holy Cross Mountains,
Poland). Palaeogeogr. Palaeoclimatol. Palaeoecol. 306, 1–14.
Marynowski, L., Zaton, M., Rakocinski, M., Filipiak, P., Kurkiewicz, S., Pearce, T.J.,
2012. Deciphering the upper Famennian Hangenberg Black Shale depositional
environments based on multi-proxy record. Palaeogeogr. Palaeoclimatol.
Palaeoecol. 346, 66–86.
Meyers, S.R., Hinnov, L.A., 2010. Northern Hemisphere glaciation and the evolution
of Plio–Pleistocene climate noise. Paleoceanography 25.
Mitchell, R.N., Bice, D.M., Montanari, A., Cleaveland, L.C., Christianson, K.T.,
Coccioni, R., Hinnov, L.A., 2008. Oceanic anoxic cycles? Orbital prelude to the
Bonarelli Level (OAE 2). Earth Planet. Sci. Lett. 267, 1–16.
Muller, R.A., MacDonald, G.J., 2000. Ice Ages and Astronomical Causes. Data,
Spectral Analysis and Mechanisms. Springer, London.
Paillard, D., Labeyrie, L., Yiou, P., 1996. Macintosh program performs time-series
analysis. EOS Trans. AGU 77, 379.
Racka, M., Marynowski, L., Filipiak, P., Sobstel, M., Pisarzowska, A., Bond, D.P.G.,
2010. Anoxic Annulata Events in the Late Famennian of the Holy Cross
Mountains (Southern Poland): Geochemical and palaeontological record.
Palaeogeogr. Palaeoclimatol. Palaeoecol. 297, 549–575.
Racki, G., 1993. Evolution of the bank to reef complex in the Devonian of the Holy
Cross Mountains. Acta Geol. Pol. 37, 87–182.
37
Racki, G., Baliński, A., 1998. Late Frasnian Atrypida (Brachiopoda) from Poland and
the Frasnian–Famennian biotic crisis. Acta Palaeontol. Pol. 43, 273–304.
Racki, G., Racka, M., Matyja, H., Devleeschouwer, X., 2002. The Frasnian/Famennian
boundary interval in the South Polish–Moravian shelf basins: integrated
event-stratigraphical approach. Palaeogeogr. Palaeoclimatol. Palaeoecol. 181,
251–297.
Racki, G., 2005. Towards understanding Late Devonian global events: few answers,
many questions. In: Over, D.J., Morrow, J.R., Wignall, P.B. (Eds.), Understanding
Late Devonian and Permian-Triassic Biotic and Climatic Events: Towards an
Integrated Approach, Developments in Paleontology and Stratigraphy. Elsevier, Amsterdam, pp. 5–36.
Radwanski, A., Kin, A., Radwanska, U., 2009. Queues of blind phacopid trilobites
Trimerocephalus: a case of frozen behaviour of Early Famennian age from the
Holy Cross Mountains, Central Poland. Acta Geol. Pol. 59, 459–481.
Raiswell, R., 1988. Chemical-model for the origin of minor limestone-shale cycles
by anaerobic methane oxidation. Geology 16, 641–644.
Rakocinski, M., 2011. Sclerobionts on upper Famennian cephalopods from the
Holy Cross Mountains, Poland. Palaeobiodivers. Palaeoenviron. 91, 63–73.
Richards, B.C., Ross, G.M., Utting, J., 2002. U–Pb geochronology, lithology and
biostratigraphy of tuff in the upper Famennian to Tournaisian Exshaw
Formation: evidence for a mid-Paleozoic magmatic arc on the northwestern
margin of North America. Canadian Society of Petroleum Geologists Memoir
19, 158–207.
Rimmer, S.M., Thompson, J.A., Goodnight, S.A., Robl, T.L., 2004. Multiple controls
on the preservation of organic matter in Devonian–Mississippian marine black
shales: geochemical and petrographic evidence. Palaeogeogr. Palaeoclimatol.
Palaeoecol. 215, 125–154.
Schulz, M., Schafer-Neth, C., 1997. Translating Milankovitch climate forcing into
eustatic fluctuations via thermal deep water expansion: a conceptual link.
Terra Nova 9, 228–231.
Streel, M., 2000. The late Famennian and early Frasnian datings given by Tucker
and others are biostratigraphically poorly constrained. Subcommission on
Devonian Stratigraphy, Newsletter, vol. 17, p. 59.
Streel, M., Caputo, M.V., Loboziak, S., Melo, J.H.G., 2000. Late Frasnian–Famennian
climates based on palynomorph analyses and the question of the Late
Devonian glaciations. Earth-Sci. Rev. 52, 121–173.
Szulczewski, M., 1971. Upper Devonian conodonts, stratigraphy and facial development in the Holy Cross Mts. Acta Geol. Pol. 21, 1–129.
Szulczewski, M., 1995. Depositional evolution of the Holy Cross Mountains in the
Devonian and Carboniferous—a review. Geol. Q. 39, 471–488.
Szulczewski, M., 1996. Devonian succession in the Koala quarry and railroad cut.
In: Proceedings of the Sixth European Conodont Symposium (ECOS VI),
Excursion Guide, Warsaw, pp. 27–30.
Thomson, D.J., 1982. Spectrum Estimation and Harmonic-Analysis. Proc. IEEE 70,
1055–1096.
Torrence, C., Compo, G.P., 1998. A practical guide to Wavelet Analysis. Bull. Am.
Meteorol. Soc. 79, 61–78.
Trapp, E., Kaufmann, B., Mezger, K., Korn, D., Weyer, D., 2004. Numerical
calibration of the devonian-carboniferous boundary: two new U-Pb isotope
dilution-thermal ionization mass spectrometry single-zircon ages from
Hasselbachtal (Sauerland, Germany). Geology 32, 857–860.
Tucker, R.D., Bradley, D.C., Straeten, C.A.V., Harris, A.G., Ebert, J.R., McCutcheon,
S.R., 1998. New U–Pb zircon ages and the duration and division of Devonian
time. Earth Planet. Sci. Lett. 158, 175–186.
Tuenter, E., Weber, S.L., Hilgen, F.J., Lourens, L.J., 2003. The response of the African
summer monsoon to remote and local forcing due to precession and obliquity.
Global Planet. Change 36, 219–235.
Tyszka, J., 2009. Foraminiferal response to seasonality modulated by orbital cycles
in the Cretaceous mid-latitudes: the Albian record from the Lower Saxony
Basin. Palaeogeogr. Palaeoclimatol. Palaeoecol. 276, 148–159.
van Geldern, R., Joachimski, M.M., Day, J., Jansen, U., Alvarez, F., Yolkin, E.A., Ma,
X.P., 2006. Carbon, oxygen and strontium isotope records of Devonian
brachiopod shell calcite. Palaeogeogr. Palaeoclimatol. Palaeoecol. 240, 47–67.
Veizer, J., Ala, D., Azmy, K., Bruckschen, P., Buhl, D., Bruhn, F., Carden, G.A.F.,
Diener, A., Ebneth, S., Godderis, Y., Jasper, T., Korte, C., Pawellek, F., Podlaha,
O.G., Strauss, H., 1999. Sr-87/Sr-86, delta C-13 and delta O-18 evolution of
Phanerozoic seawater. Chem. Geol. 161, 59–88.
Verschuren, D., Damste, J.S.S., Moernaut, J., Kristen, I., Blaauw, M., Fagot, M., Haug,
G.H., 2009. Half-precessional dynamics of monsoon rainfall near the East
African Equator. Nature 462, 637–641.
Vishnevskaya, V., Pisera, A., Racki, G., 2002. Siliceous biota (radiolarians and
sponges) and the Late Devonian biotic crisis: the Polish reference. Acta
Palaeontol. Pol. 47, 211–226.
Westphal, H., Hilgen, F., Munnecke, A., 2010. An assessment of the suitability of
individual rhythmic carbonate successions for astrochronological application.
Earth-Sci. Rev. 99, 19–30.