Embed Size (px)
Citation preview
Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx–xxx
PALAEO-07483; No of Pages 13
Contents lists available at ScienceDirect
Palaeogeography, Palaeoclimatology, Palaeoecology
j ourna l homepage: www.e lsev ie r .com/ locate /pa laeo
High dependence of Ordovician ocean surface circulation on atmosphericCO2 levels
Alexandre Pohl a,⁎, Elise Nardin b, Thijs R.A. Vandenbroucke c,d, Yannick Donnadieu a
a LSCE, Laboratoire des Sciences du Climat et de l'Environnement, UMR 8212, CNRS-CEA-UVSQ, CEA Saclay, Orme des Merisiers, 91191 Gif-sur-Yvette Cedex, Franceb UMR 5563 Géosciences Environnement Toulouse, Observatoire Midi-Pyrénées, CNRS, Toulouse, Francec Department of Geology and Soil Sciences (WE13), Ghent University, Krijgslaan 281/S8, 9000 Ghent, Belgiumd UMR 8198 Evo-Eco-Paléo, Université de Lille, Sciences et Technologies, Avenue Paul Langevin, bâtiment SN5, 59655 Villeneuve d'Ascq Cedex, France
⁎ Corresponding author.E-mail address: [email protected] (A. Pohl).
http://dx.doi.org/10.1016/j.palaeo.2015.09.0360031-0182/© 2015 Elsevier B.V. All rights reserved.
Please cite this article as: Pohl, A., et al., HiPalaeoclimatol. Palaeoecol. (2015), http://dx
a b s t r a c t
a r t i c l e i n f oArticle history:Received 27 February 2015Received in revised form 21 September 2015Accepted 23 September 2015Available online xxxx
Keywords:OrdovicianOcean surface circulationBiogeographyClimate modellingCO2
Life on Earth radiated dramatically during the Ordovician Period (485–444Ma) at all taxonomic levels during anevolutionary episode known as the ‘Great Ordovician Biodiversification Event’. Abundant scientific literature hasdocumented the pattern of biodiversification for numerous organism groups. However, fully understanding thisevent also requires detailed knowledge about the spatial distribution of the organisms. The full potential of thepalaeobiodiversity patterns remains under-explored due to the lack of knowledge about past ocean circulation.Surface circulation significantly conditions the dispersal of organisms, in creating preferential ocean migrationpathways or, in contrast, by isolating specific regions and thus establishing major physical barriers. These uncer-tainties persist because neither direct observational data, nor proxy data independent from the fossil record, areavailable to reconstruct the ocean surface circulation. Herewe present newmaps of Ordovician ocean surface cir-culation based on simulations using the general circulation model FOAM. This is a coupled ocean-atmospheregeneral circulationmodel and represents a methodological improvement over previously publishedwork. In ad-dition, we use the most up-to-date palaeogeographical reconstructions to increase compatibility with recentpalaeontological databases. In providing clear, synthetic maps of ocean surface circulation in addition to rawmodel outputs, we aim to facilitate data-model comparison, to assist in interpretation of palaeontologicaldatasets, and to promote renewed discussion about Ordovician biogeography. Finally, we investigate thesensitivity of the circulation pattern to the atmospheric CO2 content, which remains poorly constrained in theOrdovician, potentially ranging from more than 15 times the preindustrial atmospheric CO2 level (PAL) to ~5PAL. Maps are systematically constructed for a high (16 PAL), medium (8 PAL) and low (4 PAL) CO2 value. Weshow that the circulation pattern is much more sensitive to CO2 than previously suggested.
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
The Ordovician Period was characterised by major changes inmarine faunal communities. Following the appearance of numerous an-imal phyla during the ‘Cambrian explosion’ (e.g., Zhuravlev and Riding,2001), the Early andMiddle Ordovician reflected their rapid diversifica-tion, notably at genus and species levels, during the ‘Great OrdovicianBiodiversification Event’ (Webby et al., 2004), followed by one of thelargest mass extinction events of the Phanerozoic during the Hirnantian(Sheehan, 2001).
International scientific networks – in particular a series of theInternational Geoscience Programmes (IGCPs) – first focused ondocumenting these biodiversification signals (IGCP No. 410; 2000 –2004), and built a large, global Ordovician palaeontological database(Webby et al., 2004). Follow-up IGCPs 503 (2005 – 2010) and 591
gh dependence of Ordovician.doi.org/10.1016/j.palaeo.201
(2011 – 2016), have concentrated their efforts on understanding andexplaining the causes and context of these palaeodiversity patterns(e.g., Servais et al., 2010; Vandenbroucke et al., 2010; Harper et al.,2013; Rasmussen, 2013; Cramer et al., 2015; Vandenbroucke et al.,2015). To obtain this goal, a much better understanding of regional vs.global biodiversity, faunal migrations and pathways, spatial distributionof sample points, and other aspects of palaeobiogeography is required. Amajor advance towards these goals consists of the compilation ofpalaeobiogeographical data by Harper and Servais (2013), using thestate-of-the-art basemaps of Torsvik and Cocks (2009; 2013a),and representing an important update of the 1990 compilation ofMcKerrowand Scotese (1990). However, amajor lacuna in the interpre-tation of these data is our relatively poor understanding of Ordovicianocean surface circulation (Servais et al., 2014).
Biogeography also constitutes a major tool to reconstruct continen-tal configurations in deep time (Cocks and Fortey, 1982; McKerrowand Scotese, 1990; Lees et al., 2002; Torsvik and Cocks, 2013a). Inparticular, palaeontological affinities conveniently complement
ocean surface circulation on atmospheric CO2 levels, Palaeogeogr.5.09.036
2 A. Pohl et al. / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx–xxx
palaeomagnetismbyproviding information about the longitudinal posi-tion of the continents (Lees et al., 2002; Schmachtenberg, 2011). Somedifficulties, however, remain, notably in detangling the influences ofphysical distance, the potential presence of major oceanic migrationpathways, and physical barriers on the faunal similarity of two localities.In the case where two areas with significant faunal affinities were notnecessarily close to each other, it is important to constrain howcontinental masses and oceanic pathways were connected. Here,again, our poor understanding of ocean surface circulation hampersour overall understanding.
Neither direct observational data, nor proxy data independent of thefossil data discussed, are available to reconstruct Ordovician oceansurface circulation patterns, but climatemodelling provides an unparal-leled opportunity to advance our understanding. Wilde (1991) andChristiansen and Stouge (1999) were the first to suggest conceptualmodels. Based on the fundamental laws of physical oceanography,they postulated the position of the main Ordovician ocean currents.The advent of supercomputers marked a step forward by allowingnumerical modelling of paleocean circulation. Interestingly, the resultsobtained by numerical modelling by Poussart et al. (1999) andHerrmann et al. (2004) are mutually consistent and rather similar tothe conceptual model suggested by Wilde (1991).
Here, we present new maps of Ordovician ocean surface circulationbased on simulations run with the coupled ocean-atmosphere generalcirculation model FOAM (Jacob, 1997). We use up-to-date continentalreconstructions (Torsvik and Cocks, 2009),which represent a significantupdate compared to Poussart et al. (1999) and Herrmann et al. (2004),who used land-sea distributions respectively from Crowley and Baum(1995), and from Scotese and McKerrow (1990, 1991) and Scotese(1997). As such, our models are the first to use the same basemaps as,and to be fully compatible with, the recent and exhaustivepalaeobiogeographical compilation of Harper and Servais (2013). In ad-dition,we also investigate the sensitivity of the circulation pattern to theatmospheric CO2 content, which remains uncertain due to a lack ofOrdovician proxy data (e.g., Yapp and Poths, 1992; Berner, 2006; Nardinet al., 2011). Ourmain goal is to provide syntheticmaps of ocean surfacecirculation that can be used to help interpret palaeontological datasetsand to promote discussion about Ordovician biogeography. EarlyPalaeozoic deep-sea environments are rarely preserved, so ourdiscussion will necessarily focus on shelf and slope settings, i.e., on theupper one to two hundred metres of the water column.
2. Methods
2.1. The coupled climate model FOAM
We used a general circulation model (GCM) with coupled compo-nents for ocean and atmosphere: the Fast Ocean-Atmosphere Model(FOAM) version 1.5 (Jacob, 1997). The atmospheric component is aparallelised version of the National Center for Atmospheric Research's(NCAR) Community Climate Model 2 (CCM2) including the upgradedradiative and hydrologic physics from CCM3 version 3.2 (Kiehl et al.,1998). It was run at a R15 spectral resolution (4.5° × 7.5° × 18 verticallevels). The ocean component is the Ocean Model version 3 (OM3). Itis a 24-level z-coordinate ocean GCM providing a 1.4° × 2.8° resolution.The OM3 explicitly resolves ocean dynamics, which is crucial to investi-gate climate of deep-time periods when the continental configurationwas much different from today (Pohl et al., 2014). FOAM also includesa sea-ice module. It uses the thermodynamic component of theCSM1.4 sea-ice model, which is based on the Semtner 3-layer thermo-dynamic snow/ice model (Semtner, 1976). The coupled model, FOAM,is well designed for palaeoclimate studies. It has no flux correctionsand its quick turnaround time allows for longmillennium-scale integra-tions. FOAM has been widely used in palaeoclimate studies (Poulsenand Jacob, 2004; Donnadieu et al., 2009; Zhang et al., 2010; Nardinet al., 2011; Dera and Donnadieu, 2012; Lefebvre et al., 2012; Ladant
Please cite this article as: Pohl, A., et al., High dependence of OrdovicianPalaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.201
et al., 2014a,b; Licht et al., 2014) including for palaeoceanographicpurposes (Zhang et al., 2010; Dera and Donnadieu, 2012; Lefebvreet al., 2012; Ladant et al., 2014a).
2.2. Experimental setup
Simulations were carried out on three time slices, i.e., the Early,Middle, and Late Ordovician. Continental reconstructions from Torsvikand Cocks (2009) are available at 5 Ma increments, and we used theirreconstructions at 480 Ma (Early Ordovician – Tremadocian), at460 Ma (Middle Ordovician – Darriwilian) and at 440 Ma (SilurianLlandovery – Aeronian), the latter serving as an approximation for theLate Ordovician Hirnantian (~445–444 Ma). These reconstructions wereselected formaximal compatibilitywith recent palaeobiogeographical da-tabases in the compilation volume of Harper and Servais (2013), wherethey served as basemaps for the majority of contributions. Reconstruc-tions from Torsvik and Cocks (2013a), also used by some authors inHarper and Servais (2013), represent an update of the same maps butare not substantially different, except for the position of modern Asia onthe Early and Middle Ordovician configurations (Cocks and Torsvik,2013). The reconstructions from Torsvik and Cocks (2009) show the po-sition of the ancient continental blocks together with the boundaries ofmost of the modern continents (Fig. 1), facilitating the localisation offield-data. However, no information about ocean palaeo-bathymetrywas provided. We reconstructed bathymetry based on specific studiesfor Gondwana (Torsvik and Cocks, 2013b), Laurentia (Cocks andTorsvik, 2011), Baltica (Cocks and Torsvik, 2005), Siberia (Cocks andTorsvik, 2007) and for modern Asia (Cocks and Torsvik, 2013). These re-gional reconstructions distinguish between land, shallow shelves, deepshelves and open ocean. In our study, we prescribed bathymetric classesas follows: (i) we defined shallow shelves as photic zone, wave-agitatedshoreface environments and consequently used a water depth of–100 m, as a mean value for fair weather wave base; (ii) in the sameway, the deep shelf was defined to include offshore environments, andwater depth was set to−200m; and (iii) in the absence of documentedOrdovician ocean ridges, the ocean model bathymetry was defined as aflat-bottom ocean, which is a choice known not to critically impactocean surface circulation (Losch and Wunsch, 2003). Its depth was setto present-day mean seafloor depth, i.e., −4000 m. Ordovician land to-pography is uncertain, especially given that the rare palaeogeographicalmaps that reconstruct topography do not necessarily concur with eachother (e.g., Blakey, 2015 and Scotese, 2015). We adopted a conservativeview of what could have been land relief during the Ordovician. Wemainly followed published global reconstructions, with additional infor-mation from selected regional studies, for major continents includingLaurentia (Miall and Blakey, 2008; Wise and Ganis, 2009; Murphyet al., 2010; Cocks and Torsvik, 2011), Baltica (Torsvik and Rehnström,2001; Cocks and Torsvik, 2005; Torsvik and Cocks, 2005), Avalonia(Cocks et al., 1997; Torsvik and Rehnström, 2003; Murphy et al., 2008),Siberia (Cocks and Torsvik, 2007), Gondwana (Cocks, 2000; Stampfliet al., 2002; Finney et al., 2003; Cocks and Torsvik, 2006; Torsvikand Cocks, 2013b), and Asia (Metcafle, 1998, 2002; Cocks and Torsvik,2013). We distinguished peneplained shields and recently upliftedareas from younger sediments, and assigned constant elevations ofrespectively 1200 m and 200 m to these terrains. Only active orogens –along subduction zones and the well-documented CaledonianOrogen – were given a higher elevation of 4200 m. The resultingpalaeomaps are shown in Fig. 1. The numerical maps used as input tothe FOAM model are displayed in Fig. 2.
The nature of the land surface affects climate through albedo feed-backs. Continental vegetation was restricted to early, non-vascularplants during the Ordovician (Steemans et al., 2009; Rubinstein et al.,2010). Because the biomass of these pioneer plants and consequentlytheir albedo effect cannot be estimated, we followed Nardin et al.(2011) and Pohl et al. (2014) in imposing a uniform bare soil (rockydesert) cover on the continents. Alternative vegetation styles with
ocean surface circulation on atmospheric CO2 levels, Palaeogeogr.5.09.036
(a) 480 Ma
(b) 460 Ma
(c) 440 Ma
Fig. 1. Reconstructed palaeomaps showing present-day continental blocks boundaries and associated palaeoenvironments on the continental configurations for the Early (a), the Middle(b) and the Late (c) Ordovician (see Section. 2.2 for references). Emerged lands are in dark grey, shallow-water environments in light grey and deep-water environments are in white.These palaeomaps were secondly converted to the FOAM ocean grid resolution (Fig. 2). No continent was situated beyond 45° N.
3A. Pohl et al. / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx–xxx
associated albedo feedbacks cannot be excluded and range fromcontinents partially covered by land plants to a fully vegetated world.However, sensitivity tests show that even the extreme case of anOrdovician continental surface that is completely covered by non-vascular plants would not critically impact our findings. Global annualmean surface air temperature modelled at 440 Ma and 8 PAL is21.20 °C (Table 1) with a rocky desert-like albedo (0.24). Using atundra-like albedo (0.1) – a reasonable estimate for early Palaeozoiclichen and bryophytes – leads to a warmer climate (22.49 °C) butocean surface circulation remains stable between these experiments.Model runs using an Ordovician land surface completely colonised byland plants would shift our pCO2–temperature relationship towardsslightly warmer values. In no circumstance, however, should our simu-lation outputs be taken at face value. For instance, the CO2 levels used
Please cite this article as: Pohl, A., et al., High dependence of OrdovicianPalaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.201
here (i.e., 16 PAL, 8 PAL and 4 PAL) are strictly indicative, and we en-courage the reader to consider them as representative of upper, inter-mediate and lower estimates for Ordovician pCO2 rather than exactvalues.
The climate boundary conditions of the Ordovician are not knownwith certainty. The solar constant for the Ordovician was decreased by3.5% compared to its present value according to themodel of solar phys-ics provided by Gough (1981). This value is slightly lower than the 4.5%reduction in solar luminosity used by Herrmann et al. (2004), thoughboth estimates remain in the range of 3.5% to 5% proposed by Endaland Sofia (1981). The orbital configuration was chosen to bring anequal amount of energy to each hemisphere throughout a year(eccentricity: 0, obliquity: 22°, longitude of perihelion: 90°). There areno precise constraints on the Ordovician atmospheric carbon dioxide
ocean surface circulation on atmospheric CO2 levels, Palaeogeogr.5.09.036
Mid
dle
Ord
ovi
cian
E. O
rdo
Lat
e O
rdo
vici
anL
lan
do
very
Dap
ingi
anH
irn.
Rhu
dd.
Kat
ian
Aer
.S
andb
ian
Dar
riwili
an
Ord
ovi
cian
Mid
dle
Ord
ovi
cian
Ear
ly O
rdo
vici
anL
ate
Ord
ovi
cian
Lla
nd
ove
ryD
ap.
H.
Rh.
Kat
ian
Tely
c.A
er.
San
dbia
nD
arriw
ilianO
rdo
vici
anS
iluri
an
470
478
485
467
458
453
445444
441
439
Flo
ian
Trem
adoc
ian
(a)
(b)
(c)
Active orogeny (4200 m)Ancient orogeny (1200 m)
Epicontinental sea (-100 m)
Peneplane continent (200 m) Abyssal plain (-4000 m)Continental shelf (-200 m)
Fig. 2. Continental reconstructions converted to the FOAM ocean grid resolution for the 440 Ma (a), 460 Ma (b) and 480 Ma (c) time slices. The colour code gives the topographic classesused in this study. The Ordovician chronostratigraphic chart is adapted from the International Chronostratigraphic Chart v2014/10 (Cohen et al., 2013 updated). Dap.: Dapingian;H.: Hirnantian; Rh.: Rhuddanian; Aer.: Aeronian; Telyc.: Telychian.
4 A. Pohl et al. / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx–xxx
partial pressure (pCO2). Climate proxies and continental weatheringmodels provide estimates ranging from more than 15 times the prein-dustrial atmospheric CO2 level (PAL, 1 PAL = 280 ppm) (Yapp andPoths, 1992; Tobin and Bergström, 2002; Berner, 2006; Young et al.,
Table 1Globally-averaged, annual mean surface air temperature for each simulation discussed inthis study.
Time slice (Ma) pCO2 (PAL) Global temperature (°C)
440 16 24.3440 8 21.2440 4 12.1460 16 24.7460 8 21.3460 4 9.7480 16 25.5480 8 22.4480 4 11.2
Please cite this article as: Pohl, A., et al., High dependence of OrdovicianPalaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.201
2008) to ~5 PAL (Rothman, 2002; Vandenbroucke et al., 2010). Inorder to cover every possibility in thiswide spectrum, and to investigatethe sensitivity of the ocean surface circulation to atmospheric forcing,simulations were run for each time slice at 16 PAL, 8 PAL and 4 PAL.FOAM was systematically integrated over 2000 years to reach deep-ocean equilibrium. During the last 100 years of each model run, noapparent drift is observed in the upper-ocean temperature (betweenthe surface and 300mdepth), and globally averaged ocean temperaturevariations are less than 0.002 °C/year.
2.3. From raw GCM outputs to synthetic ocean circulation maps
Water flow directions computed by the GCMwere available at everyoceanic grid point from the surface to the bottom of the ocean. Weaveraged the ocean currents over the first 100 m of the water column,taking into account the effect of Ekman transport (Ekman, 1905). Theresulting maps (Figs. S2–S10) effectively represent patterns of waterflow among shallow-water environments (epicontinental seas,
ocean surface circulation on atmospheric CO2 levels, Palaeogeogr.5.09.036
5A. Pohl et al. / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx–xxx
continental shelves). To improve readability, the main ocean currentswere summarised in synthetic maps (Figs. 4–6).
In line with our objectives and anticipated palaeontological applica-tions, where sea surface temperature is an essential constraint on thedispersion of marine organisms (e.g., Vandenbroucke et al., 2009,2013), we set a threshold on the seawater temperature at 18 °C aver-aged over the top 100 m of the water column. Above this threshold,‘warm’ currents were drawn in red, below it, ‘cold’ currents weredrawn in blue (Figs. 4–6). The 18 °C threshold is a critical temperaturefor marine organisms: it marks the shift from mixed photozoan-heterozoan to heterozoan-dominated carbonate producers (Lees,1975; James, 1997), and it also corresponds to the thermal boundary be-tween modern subtropical and transitional foraminifer provinces(Kucera, 2007).
3. Climate results
Pohl et al. (2014) investigated the response of Ordovician climate toa CO2 decrease. Using the general circulationmodel FOAM, they demon-strated major nonlinearities in the responses of the Ordovician climatesystem. They showed that there is a threshold CO2 value, below whichclimate abruptly shifts from a warm climatic equilibrium with no seaice in the Northern Hemisphere, to a much colder state with sea ice ex-tending down to the mid-latitudes. They attributed this climatic insta-bility to the particular ocean dynamics developing in the mainlyoceanic Ordovician Northern Hemisphere. The same climatic responseis observed in our simulations (Table 1; Fig. S1). For each time slice,the simulations with high (16 PAL) and medium (8 PAL) CO2 levelsare above the threshold CO2 value for the climatic instability whereasthe simulations conducted at 4 PAL belong to the colder climatic state.
The response of the atmospheric forcing to a changing CO2 concen-tration is logarithmic. In a present-day continental configuration, simi-lar global temperature changes are thus expected for any CO2 halving,be it from16 PAL to 8 PAL, or from8 PAL to 4 PAL. The cooling associatedwith a halving of CO2 (i.e., the climatic sensitivity) was calculated, usingvarious climate models in different time slices, to vary between 2.1 and4.8 °C (PALAEOSENS Project, 2012; IPCC, 2013). In our Ordovician study,decreasing CO2 from 16 PAL to 8 PAL causes a ~3 °C cooling (in linewithpredictions), while the particular Ordovician climatic response resultsin the crossing of the CO2-threshold with a CO2 decrease from 8 PALto 4 PAL, actually leading to a ~11 °C cooling (Table 1). We refer toPohl et al. (2014) for an in-depth discussion of the causal mechanismsfor this climatic instability.
4. Ordovician ocean surface circulation
4.1. Terminology
Fig. 3 introduces themain terms employed hereafter throughout thediscussion.We follow the terminology defined by Poussart et al. (1999)and later used by Herrmann et al. (2004). The ocean situated betweenGondwana, Baltica, Asia and Siberia has taken many names in theliterature, but we here use the common term Palaeo-Tethys, followingScotese (2015).
Global Ordovician circulation is driven by easterlies at the Equatorand westerlies at the mid-latitudes (~40° N and 40° S), with largeocean gyres in between. Poussart et al. (1999) defined four oceangyres (No. 1 to 4 in Fig. 3a–c): (1) the north Panthalassic convergence,(2) the south Panthalassic convergence, (3) the south Palaeo-Tethysconvergence and (4) the north Palaeo-Tethys convergence. Gyres No.1 and 4 display a clockwise circulationwhereas gyres No. 2 and 3 are an-ticlockwise, following the fundamental principles of oceanography(Wilde, 1991). We use these terms, and we introduce a high-latitudeaustral clockwise gyre south of the 40° S westerlies (No. 5) that weterm the Rheic convergence. Poussart et al. (1999) also introducednames for ocean currents: the Iapetus Current (IC) between Laurentia,
Please cite this article as: Pohl, A., et al., High dependence of OrdovicianPalaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.201
Avalonia and Baltica; the Antarctica Current (AC) along the easterncoast of Gondwana from East Antarctica down to South America; thePanthalassic Circumpolar Current (PCC) in the Northern Hemisphere;and theNorth and South Equatorial Currents (NEC and SEC). In addition,we introduce the Gondwana Current (GC) along the southernmargin ofGondwana (northwestern Africa), the Southern Laurentia Current (SLC)along the coast of southwestern Laurentia, and the SouthernWesterlies(SW) at ~40° S.
Below, we provide a description of the ocean surface circulationmodelled with FOAM. The main features of each continental configura-tion (or time slice) are first explained, and then the influence of thepCO2 is investigated as we highlight specificities for each model run.We do not wish to provide an exhaustive discussion. Instead, we aimto help the readers to interpret the maps and to guide their attentionto some key features. Further information can be derived for specificcase studies, regions and time slices, and details of the patterns are pro-vided on the raw FOAM output maps (Figs. S2–S10).
4.2. Early Ordovician (480 Ma)
In general for this time slice (Fig. 4), because of the absence of conti-nentalmasses in the Northern Hemispheremid- and high-latitudes, thePanthalassic Circumpolar Current is well developed at 30° N. By con-trast, in the Southern Hemisphere, the presence of Baltica obstructsthe SouthernWesterlies at 40° S. Between 30° N and 40 °S, two regionshave to be considered separately: (i) between 0° and 30° N, the oceancirculation is mainly zonal. Intense equatorial currents are observed be-tween western Laurentia and eastern Gondwana; (ii) between 0° and40° S, the circulation is much more complex due to the presence ofnumerous landmasses. In the east, the latitudinal position of Gondwana,and especially of South America, prevents the Antarctica Currentfrom descending further south, and the Antarctica Current mergeswith the Southern Westerlies, thus establishing communication be-tween eastern Gondwana and northern Baltica through the IapetusOcean. At higher latitudes (N40° S), Baltica tends to isolate the watermasses in the Rheic–Iapetus–Tornquist region from those of thesouthern Palaeo-Tethys.
At 16 PAL in the Northern Hemisphere (Fig. 4a), the surface circula-tion is essentially zonal: beyond 30° N, the eastward Panthalassic Cir-cumpolar Current prevails, whereas at lower latitudes westward zonalcurrents establish a one-way communication around the world. East-ward currents appear only at low latitudes from eastern Australia towestern Laurentia, as theNorth Equatorial Current counter-flow. In con-trast, in the Southern Hemisphere, between 0° and 40° S, meridionalcurrents prevail, reflecting the deviation of zonal flow by Laurentia,Baltica and Siberia. In the west, in particular, the Southern LaurentiaCurrent conveys warm waters along the western and southern coastsof Laurentia before merging with the Southern Westerlies. BetweenLaurentia, Baltica and Siberia, water masses are efficiently exchangedthrough an anticlockwise circulation that develops from northern Sibe-ria to Laurentia and then to northern Baltica. An additional flow linkseastern Laurentia to southwestern Siberia. In the Palaeo-Tethys, watermasses emerging from North Australia reach India via eastward flowsalong South China or turning south of North China and Annamia. Be-tween 20° S and 40° S, most of the circulation is eastward. At ~40° S,the Southern Westerlies are inhibited by Baltica, and undergo a north-ward deflection into the Aegir Ocean that allows them to reach thePalaeo-Tethys realm and continue to eastern Gondwana. Along theirway to Gondwana, they gather water masses coming from Laurentiaand Siberia. Just before reaching the Gondwana coastline, they splitinto two independent limbs, a warm one going north to Annamia, anda colder one going further south to eventually join the GondwanaCurrent. Beyond 40° S, the circulation is characterised by a large and rel-atively well-developed Rheic convergence (gyre No. 5 in Fig. 2)favouring water exchanges between Baltica and northwestern Africa.
ocean surface circulation on atmospheric CO2 levels, Palaeogeogr.5.09.036
0°
40°N
40°S
80°S
(a) 480 Ma
2 2
PCC
AC
14
3
1
GC
SW
SLC
5
ACIC
South America
South Africa
Est Antarctica
Australia
India
Arabia
NE Africa
NW Africa
NW Africa
SouthAmerica
L
B
ST
NC
SC
An
A
AAC
GONDWANA
NW Africa
Panthalassa
Iapetus
RheicTornquist
Aegir
Palaeo-Tethys
0°
40°N
40°S
80°S
(b) 460 Ma
PCC
AC
AC
1 4
3
1
GC
SW
SLC
5
2 2IC
L
B
S
NC
A
SCTAn
Australia
Est Antarctica
NW Africa
Arabia
IndiaSouth Africa
NE Africa
SouthAmerica
NW Africa
GONDWANA
AAC
NW Africa
South America
Rheic
Tornquist
Panthalassa
Iapetus
Aegir
Palaeo-Tethys
0°
40°N
40°S
80°S
(c) 440 Ma
PCC
1 4
2 23
1
GC
SW
SLC
5
IC
AC
AACS
LB
NC SC
T
An
Arabia
India
Est Antarctica
Australia
South Africa
GONDWANA
KO
Panthalassa
Iapetus
Aegir
Palaeo-TethysA
NEC and SEC NEC and SEC
NEC and SEC NEC and SEC
NEC and SEC NEC and SECNEC and SEC
Fig. 3. Terminology used for the description of the ocean circulation at 480 Ma (a), 460 Ma (b) and 440 Ma (c). Each potential component of the ocean circulation is displayed withoutconsidering any precise CO2 level. These componentsmayor not be actually observed in the various simulations. Emerged lands are represented indark-grey, shallow-water environmentsin light-grey and deep-water environments are in white. Ordovician continental blocks names are in bold: Gondwana, Laurentia (L), Baltica (B), Siberia (S), Avalonia (A), Arctic Alaska–Chukotka (AAC), Kolyma–Omolon (KO), Tarim (T), North China (NC), South China (SC) and Annamia (An). To make the maps more legible, only selected present-day regions are indicated(South America, NE Africa…). Please see Fig. 1 for more information about the position of present-day continents during the Ordovician. Ocean names are in italic. Rectangular labels areocean currents names: the Panthalassic Circumpolar Current (PCC), the Iapetus Current (IC), the Southern Laurentia Current (SLC), the North and South Equatorial Currents (NEC and SEC), theAntarctica Current (AC), the SouthernWesterlies (SW) and the Gondwana Current (GC). Circles finally represent ocean gyres: (1) the north Panthalassic convergence, (2) the south Panthalassicconvergence, (3) the south Palaeo-Tethys convergence, (4) the north Palaeo-Tethys convergence, and (5) the Rheic convergence.
6 A. Pohl et al. / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx–xxx
At 8 PAL (Fig. 4b), no major change occurs in the Northern Hemi-sphere in comparison to the 16 PAL configuration. In the SouthernHemisphere, between 0° and 40° S, the communication betweenLaurentia, Baltica and Siberia weakens. The anticlockwise circulationstill exists, but the export of waters from Laurentia to Siberia ceases. Incontrast, Baltica now directly receives incoming waters from bothLaurentia and Siberia. South of North China and Annamia the currentsare reversed, now flowing from the northeast to the southwest. Thesouth Palaeo-Tethys gyre features in the 8 PAL model run, but doesnot promote exchange between emerged lands or shallow-water
Please cite this article as: Pohl, A., et al., High dependence of OrdovicianPalaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.201
environments. The Southern Laurentia Current keeps bringing warmwaters from the Equator to the western and southern coasts ofLaurentia. At 40° S, the Southern Westerlies are now definitivelyinterrupted by Baltica. In the southern high-latitudes, the GondwanaCurrent intensifies in its western part and disappears in its easternpart. Two independent gyres replace the previously single Rheic gyre:afirst one in the Palaeo-Tethys that does not induce anywater exchangebetween shelfal environments, and another in the Rheic, Iapetus andTornquist oceans that tightly connects northwestern Africa, SouthAmerica, Avalonia and Baltica.
ocean surface circulation on atmospheric CO2 levels, Palaeogeogr.5.09.036
0°
40°N
40°S
80°S
(a) 480 Ma: 16 PAL
0°
40°N
40°S
80°S
(b) 480 Ma: 8 PAL
0°
40°N
40°S
80°S
(c) 480 Ma: 4 PAL
Fig. 4. Early Ordovician (480Ma) synthetic ocean surface circulation at 16 PAL (a), 8 PAL (b) and 4 PAL (c). If ocean temperatures are warmer than 18 °C, currents are drawn in red. If not,they are in blue (see Section. 2.3 for motivation). Beyond 45° N, the ocean circulation is purely zonal.
7A. Pohl et al. / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx–xxx
When CO2 is further decreased from 8 PAL to 4 PAL (Fig. 4c), aclimatic threshold is crossed and a sharp cooling occurs (Table 1; alsosee Section. 3), dramatically affecting the ocean surface circulation. Be-tween the Equator and 30° N, water exchanges are now restricted tothe very low latitudes where the circulation remains essentially zonal.A new connection is established between eastern Siberia and Australiathrough an equatorial, clockwise gyre centred on Tarim and SouthChina. The North Equatorial Current counter-flow becomes disconnect-ed from western Laurentia. Northwards of 15° N, water masses flownorthwestwards. In the Southern Hemisphere, westward currents pre-vail between the Equator and the Southern Westerlies. The SouthernLaurentia Current, in particular, is now reversed. It conveys watermasses equatorward and finally merges with the South EquatorialCurrent, which becomes stronger and establishes communicationfrom the southwestern Laurentian platform to eastern Gondwana.
Please cite this article as: Pohl, A., et al., High dependence of OrdovicianPalaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.201
Between Laurentia, Baltica and Siberia, the same pattern is observedas in the 8 PAL run, except that there is no longer a direct connectionfrom Laurentia to Baltica. The south Panthalassic gyre appears and thesouth Palaeo-Tethys gyre intensifies. At 40° S, the Southern Westerliesare still inhibited by Baltica, resulting in the splitting of the Rheic con-vergence into two smaller gyres. In the Rheic Ocean, a clockwise gyreconveys waters from Avalonia to Baltica and then to northwesternAfrica, but there is no longer a counter-flow from northwestern Africato Avalonia. The other convergence, situated in the southern Palaeo-Tethys, intensifies. The Gondwana Current now operates from Arabiato South America.
In summary, when CO2 levels are changed from 16 PAL to 4 PALthere are global trends emerging that include: (i) the gradual weaken-ing of water-mass exchanges between Laurentia, Baltica and Siberia;(ii) the reversal of communication between eastern Gondwana and
ocean surface circulation on atmospheric CO2 levels, Palaeogeogr.5.09.036
8 A. Pohl et al. / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx–xxx
western Laurentia, which operates through the eastward NorthEquatorial Current counter-flow at 16 and at 8 PAL, and through thewestward South Equatorial Current at 4 PAL; (iii) the establishmentand then intensification of large ocean gyres in the SouthernHemisphere; (iv) the gradual reinforcement of the Gondwana Currentalong the northern margin of Gondwana; and (v) the reversal of theSouthern Laurentia Current (from 8 PAL to 4 PAL).
4.3. Middle Ordovician (460 Ma)
In the Middle Ordovician Northern Hemisphere (Fig. 5), circulationis predominantly zonal, with westward currents between the Equatorand 30° N and the well-developed Panthalassic Circumpolar Currentto the north. Interesting, however, is the establishment of a southeast-ward counter-current from Arctic Alaska–Chukotka to Siberia. In the
0°
40°N
40°S
80°S
(a) 460 Ma: 16 PAL
0°
40°N
40°S
80°S
(b) 460 Ma: 8 PAL
0°
40°N
40°S
80°S
(c) 460 Ma: 4 PAL
Fig. 5. Middle Ordovician (460 Ma) synthetic ocean surface circulation
Please cite this article as: Pohl, A., et al., High dependence of OrdovicianPalaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.201
Southern Hemisphere, the circulation pattern is complicated by thepresence of tropical landmasses, similar to the situation at 480 Ma. Inspite of the clustering of the continental blocks, a zonal current developsat ~10° S from Annamia to eastern Laurentia, making its way betweenNorth China and Tarim: this current is deflected by the eastern coastof Laurentia, where it divides into northern and southern limbs. In thecentral region between Laurentia, Baltica and Siberia, an anticlockwisecirculation still exists but with a smaller latitudinal extent than at480 Ma. Water masses connect south Siberia to equatorial easternLaurentia and finally to northwestern Baltica. In addition, Siberia is di-rectly connected to Baltica. Another similar, anticlockwise circulationdevelops between Laurentia, Avalonia and Baltica, in the Iapetus andTornquist oceans. Baltica also receives water masses originating frommodern Asia. In the west, the south-Panthalassic gyre is observed ateach CO2 level. Along the eastern coast of Gondwana, the circulation is
at 16 PAL (a), 8 PAL (b) and 4 PAL (c). Same colour code as Fig. 4.
ocean surface circulation on atmospheric CO2 levels, Palaeogeogr.5.09.036
9A. Pohl et al. / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx–xxx
mostly unchanged compared to the Early Ordovician. The equatorialcurrents cross the Panthalassa until they reach East Antarctica, wherethe South Equatorial Current merges with the Antarctica Current,which then flows along the Gondwana coastline before merging inturn with the Southern Westerlies. The latter intercept Avalonia andcontinue further eastwards until they eventually reach the Arabianshallow-water environments. At higher latitudes a single, large Rheicconvergence developed as Baltica moved equatorwards. The Rheic con-vergence connects most of the North African shallow-water platforms.The Gondwana Current is observed solely in its central and westernparts.
Specifically for the 16 PALmodel run (Fig. 5a), themajor supplemen-tary component of the Northern Hemisphere ocean surface circulation,compared to the Early Ordovician with the same atmospheric forcing,is the southwestward current connecting Arctic Alaska–Chukotka tonorthern Laurentia. Between 0° and 40° S, two domains can be distin-guished. From 0° to 10° S, the circulation is predominantly westward;the equatorial currents are disconnected from western Laurentia.From 10° S to 40° S, the water masses flow southeastwards until theyjoin the Southern Westerlies. Avalonia is now at lower latitudes, sothe southeastward Southern Laurentia Current establishes a connectionfrom Laurentia to Avalonia through the Iapetus Ocean and then fromAvalonia to Baltica through the Tornquist Ocean. Between Laurentia,Siberia and Baltica, an additional link is observed from Baltica toLaurentia, as part of a small-scale anticlockwise gyre. In the Palaeo-Tethys, no south Palaeo-Tethys gyre is observed but a large anticlock-wise circulation originating between South China and Annamia reachessouthern Siberia, eastern Baltica and finally the Arabian platform. In thecentre of this proto-gyre, the circulation is oriented southeastwardsfrom modern Asia to the Indian platform. At the southern high-latitudes, the Rheic convergence does not reach the southern coasts ofBaltica. The Gondwana Current is restricted to northwestern Africa.
The global ocean surface circulation remains very stable when de-creasingCO2 from16PAL to 8 PAL (Fig. 5b). In the Palaeo-Tethys, the cir-culation pattern is modified within the proto-gyre between Annamiaand the coasts of Arabia and India. This new circulation promotes recip-rocal exchanges between Annamia and Gondwana, whereas previouslytheywere unidirectional. The southern limb of the Palaeo-Tethys proto-gyre also penetrates further into the shallow-water environments alongthe coasts of India. A second change occurs in the northern limb of theRheic convergence, which reaches southern Baltica and establishes atwo-way communication between Baltica and northern Gondwana.
When further decreasing CO2 from 8 PAL to 4 PAL (Fig. 5c), in theNorthern Hemisphere, the zonal component weakens in favour ofnorthwestward currents, similar to the situation at 480 Ma. The south-westward current between Arctic Alaska–Chukotka and northernLaurentia becomes broken, and the link from Arctic Alaska–Chukotkato Siberia intensifies, driven by a very local anticlockwise gyre. An addi-tional eastward counter-current connects Tarim to South China andAustralia, unlike in the 16 PAL and 8 PAL runs. Along the Equator, theNorth and South Equatorial Currents are now connected to westernLaurentia, establishing communication between the latter and easternGondwana. In the Southern Hemisphere, between 0° and 40° S, theSouthern Laurentia Current is reversed. The south Panthalassic gyresignificantly intensifies (but it still does not reach the surrounding plat-forms) and the communication from Baltica to Laurentia is interrupted.In the Palaeo-Tethys, the circulation is almost zonal, up to 20° S. Furthersouth, the south Palaeo-Tethys convergence appears and constitutes amajor connection between Baltica, India, Annamia and North China.Once again, no major change affects the Antarctica Current and the cir-culation along the eastern coast of Gondwana remains the same. At 40°S, the SouthernWesterlies are still uninterrupted but they now split intotwo branches that respectively merge with the south Palaeo-Tethysconvergence andwith the Rheic convergence. Consequently, the South-ern Westerlies do not any longer reach the Arabian platform. The Rheicgyre is unchanged compared to 8 PAL except within its eastern part,
Please cite this article as: Pohl, A., et al., High dependence of OrdovicianPalaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.201
where an additional, anticlockwise gyre appears along the coast ofnortheastern Africa.
In summary, the Middle Ordovician ocean circulation in the North-ernHemisphere is not very sensitive to the changes in CO2 levels, exceptfor the connection between Arctic Alaska–Chukotka and northernLaurentia. The Middle Ordovician palaeogeographical configuration fa-vours the onset of ocean gyres, which intensify at low CO2 levels. Aproto south Palaeo-Tethys convergence is observed even at high CO2
levels, before turning into an anticlockwise strong gyre when atmo-spheric forcing is decreased. Simultaneously, the Rheic convergence ex-pands into lower latitudes and more directly reaches southern Baltica.At 16 PAL and 8 PAL, the communication between Laurentia, Balticaand Siberia is characterised by water masses export from Baltica toLaurentia. When CO2 is decreased to 4 PAL this connection disappears,the Southern Laurentia Current and the gyre along northeastern AfricaandArabia are reversed, and the equatorial currents thatwere previous-ly disconnected from Laurentia newly establish communication be-tween the latter and eastern Gondwana.
4.4. Late Ordovician (440 Ma)
Following the northward drift of the tropical continental massesfrom the Middle to the Late Ordovician, the ocean-surface circulationin the Northern Hemisphere at the mid-latitudes is no longer zonal(Fig. 6). Northern Siberia deflects the Panthalassic Circumpolar Currentat 40° N. Between the Equator and40°N, the surface circulation nowhasa significant meridional component. Along the Equator, equatorial cur-rents develop both from western Laurentia to eastern Gondwana andfrom modern Asia to eastern Laurentia. When the latter water massesreach the eastern Laurentia coastline, they split into three branches:(i) a southern branch flowing into the Southern Hemisphere IapetusOcean (as the Iapetus Current); (ii) a central branch that penetratesinto the Laurentian epicontinental sea and reaches the southwesterncoast of Laurentia, where it eventually merges with the SouthernLaurentia Current; and (iii) a northern branch that flows along northernLaurentia, then to Arctic Alaska–Chukotka and along the western coastof Siberia. This clockwise circulation constitutes a rare configuration inwhich water masses are conveyed from Laurentia to Siberia. Laurentiaexports water masses on each side of the Equator to Siberia and Baltica,without any connection between the two latter. The Panthalassic limbof the equatorial currents connects western Laurentia to easternGondwana through the South Equatorial Current. This current is thendeflected and flows southwards along the eastern coast of Gondwanadown to the South American volcanic arcs, as the Antarctica Current. Fi-nally, the Antarctica Current deviates to form the Southern Westerliesthat flow back through the southern Panthalassa, with one partreaching Laurentia and Avalonia and the other part continuing furthereastward. In the Rheic Ocean, the Rheic convergence is well developedbut it does not reach Baltica, which has moved too far to the north.The Rheic gyre connects the platforms and epicontinental seas of north-western Africa. The Gondwana Current constitutes its southern,westward limb.
In the Northern Hemisphere at 16 PAL (Fig. 6a) five main oceangyres are observed: (i) in the Panthalassic Ocean at very low latitudes(0° – 10° N); this first gyre connects western Laurentia and easternGondwana through the westward South Equatorial Current and theeastward North Equatorial Current backflow; (ii) at higher latitudes(~20° N – 40° N), the north Panthalassic convergence links theGondwanan east coast volcanic arcs to Arctic Alaska–Chukotka; (iii) be-tween Australia and Siberia, a proto north Palaeo-Tethys convergenceappears; (iv) a small-scale clockwise gyre develops in the Aegir ocean,bringing waters from North China and conveying together the watermasses originating from Kolyma-Omolon and from southern Siberia;(v) finally, a proto south Palaeo-Tethys gyre is outlined by the anticlock-wise circulation established between 0° and ~35° S, originating in NorthChina, reaching Baltica by deflection of the equatorial currents and
ocean surface circulation on atmospheric CO2 levels, Palaeogeogr.5.09.036
0°
40°N
40°S
80°S
(a) 440 Ma: 16 PAL
0°
40°N
40°S
80°S
(b) 440 Ma: 8 PAL
0°
40°N
40°S
80°S
(c) 440 Ma: 4 PAL
Fig. 6. Late Ordovician (440 Ma) synthetic ocean surface circulation at 16 PAL (a), 8 PAL (b) and 4 PAL (c). Same colour code as Fig. 4.
10 A. Pohl et al. / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx–xxx
flowing back to the east towards Gondwana. In the sector of modernAsia, the circulation is oriented southwestwards in the NorthernHemisphere and southeastwards in the Southern Hemisphere. Watermasses originating in Australia are exported to South China and theyare then deflected towards the Arabian coast. The latter also receiveswater masses coming from Baltica through the Palaeo-Tethys. At 40°S, the Southern Westerlies merge with this Palaeo-Tethysian currentconnecting Baltica to Gondwana. The Rheic convergence is large and af-fects the northeastern and northwestern coasts of Africa.
From 16 PAL to 8 PAL (Fig. 6b), the only significant changes reflect:(i) the establishment of a two-way connection between North Chinaand South China, where it was previously unidirectional from SouthChina to North China; and (ii) an interruption of the connection be-tween Baltica and Gondwana. The Southern Westerlies do not reachthe western coast of Gondwana any longer.
Please cite this article as: Pohl, A., et al., High dependence of OrdovicianPalaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.201
At 4 PAL (Fig. 6c) in the Northern Hemisphere, the north Palaeo-Tethys gyre, the Aegir gyre and the two gyres previously observed inthe Panthalassic Ocean disappear. Although the South EquatorialCurrent still connects western Laurentia to eastern Gondwana, theNorth Equatorial Current backflow does not cross the Panthalassaanymore. A clockwise gyre becomes established in the northernPanthalassa at ~25° N, but it does not induce any communication be-tween shelfal environments. In the east, water mass export fromAustralia is now restricted to South China. The two-way communicationbetween North China and South China and the clockwise cell betweenLaurentia, Alaska–Chukotka and Siberia are both still active. An east-ward current also develops at ~30° N from Arctic Alaska–Chukotka to-wards Australia through the epicontinental seas of Siberia. The wholesurface circulation is modified in the Asiatic region. In particular, thecurrents that previously linked North China to Arabia now operate in
ocean surface circulation on atmospheric CO2 levels, Palaeogeogr.5.09.036
11A. Pohl et al. / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx–xxx
the opposite direction. A small gyre also appears between Annamia andIndia, allowing bidirectional exchanges. The connection between BalticaandArabia recovers due to a vigorous south Palaeo-Tethys convergence.The Southern Westerlies are still deflected northward along the south-ern coast of Baltica, and they continue eastwards until they mergewith the south Palaeo-Tethys convergence. In the west, the southPanthalassic convergence appears but it does not reach the surroundingcontinents. The Southern Laurentia Current is reversed only in its verysouthern part, not along the western coast of Laurentia. At higher lati-tudes, the Rheic convergence is shifted westwards, extending to SouthAmerica but at the same time disappearing in its eastern part alongnortheastern Africa, where it is replaced by a small clockwise gyre. Be-tween South America and the volcanic arcs of southern Gondwana, anadditional clockwise gyre appears.
In summary, at 440Ma, the circulation pattern is quite stable from 16PAL to 8 PAL. When CO2 is further decreased to 4 PAL, the main changesconsist of: (i) profound modifications in the ocean circulation in theAsiatic region, notably including a decreasing influence of Australia;(ii) a weakening of the ocean gyres in the Northern Hemisphere and astrengthening of the south Panthalassic and south Palaeo-Tethys gyres,associated with a westward shift of the Rheic convergence; and (iii) anincomplete reversal of the Southern Laurentia Current: thewarmcurrentflowing southwards along the western coast of Laurentia remains un-changed across our entire range of pCO2 values.
4.5. Ocean temperature and endemism
Although it bears biological significance (see Section. 2.3), theboundary set between warm and cool waters is somewhat arbitrary(James, 1997). It should be interpreted cautiously because:(i) absolute temperatures are highly model-dependent (see, for exam-ple, Braconnot et al., 2012); (ii) absolute temperatures also depend onthe boundary conditions employed (e.g., the solar constant, the landsurface type, etc.); and (iii) the limitingwater temperature for organismdispersal is taxon specific, drifts with time, and may co-vary with otheroceanographic parameters. This discussion highlights only the existenceof other barriers that have the potential to inhibitmigrations outside thepurely geographical configuration.
For each continental configuration at high (16 PAL) and medium (8PAL) CO2 levels, the position of the 18 °C threshold broadly is at the lat-itude of the Northern and Southern Westerlies (between 30° and 40°;Figs. 4a,b, 5a,b, and 6a,b). With the exception of the Early Ordovician(480 Ma), when Baltica and Avalonia were positioned at high latitudesand thus outside the 40° N – 40° S latitudinal range, all palaeocontinentsexcept Gondwana are within the warm domain, favouring warm-waterorganismmigrations. At latitudes higher than 40°, cool-water is prefer-entially exchanged along the margin of Gondwana. At 480 Ma, Balticaand Avalonia belong to the cool-water domain, favouring their isolationfrom Laurentia, Siberia and modern Asia, but at the same time promot-ing exchange with Gondwana. Avalonia and Baltica, being tangent tothe boundary between the two thermal domains during most of theOrdovician, potentially benefit from mixed faunal migrations.
When CO2 is decreased to 4 PAL, global climate cools (see Section. 3),narrowing the latitudinal extent of the warm domain (Figs. 4c, 5c and6c). For each time slice, two regions remain warm (i.e., N 18 °C) evenat very low CO2 levels (i.e., at 4 PAL): (i) the equatorial PanthalassicOcean between western Laurentia and eastern Gondwana; and (ii) atleast one pathway between the Equator and 20° S, with a warm, west-ward current originating from Asia and bringing warm waters toLaurentia. These two areas could constitute preferential migrationroutes for warm-water organisms even in an icehouse climate.
5. Discussion
The modelled global ocean surface circulation patterns (Figs. 4–6)are consistent with the conceptual model of Wilde (1991). They also
Please cite this article as: Pohl, A., et al., High dependence of OrdovicianPalaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.201
have many features in common with the results obtained by Poussartet al. (1999) and later by Herrmann et al. (2004). Poussart et al.(1999) coupled an energy-moisture balance model to an ocean generalcirculationmodel and to a sea-icemodel. Herrmann et al. (2004) forcedthe same ocean model – the Geophysical Fluid Dynamics Laboratorymodular ocean model (Pacanowski, 1996) – with the outputs from anatmospheric general circulation model (GENESIS v.2.0). Although theirmodelling designs somewhat differ, it is noteworthy that both usedthe same ocean model. Not surprisingly, their results were very similar.Our FOAM modelling results provide independent insight into Ordovi-cian ocean surface circulation. They confirm the findings fromPoussart et al. (1999) and Herrmann et al. (2004), which allows us tosuggest that the main patterns of the ocean circulation are not overlymodel-dependent.
As already suggested by Herrmann et al. (2004), we confirm thatcontinental configuration is a major control on the surface circulation.By considering a wide range of CO2 values (16 PAL – 4 PAL, i.e., 4480 –1120 ppm), we demonstrate, however, that atmospheric forcing alsoprofoundly impacts the ocean circulation pattern. On this point, ourconclusions differ from those of Herrmann et al. (2004) who notedthat “changes in atmospheric pCO2 have almost no effect on the globalocean circulation pattern” (Herrmann et al., 2004, p. 390). This discrep-ancy arises from two main differences: (i) Herrmann et al. (2004) con-ducted their climatic runs for a narrower range of CO2 values (8 – 18PAL); (ii)more significant is the fact that they did not obtain the climaticinstability described in Section. 3, that induces the sudden globalcooling in FOAM, when CO2 is decreased from 8 PAL to 4 PAL (~11 °Ccooling, Table 1, Fig. S1). Their climatic sensitivity, computed with theatmospheric model GENESIS, remained around 2.5 °C per halving ofCO2. The temperature drop associated with their full range of modelledpCO2 (~5.5 °C) was consequently much weaker than the cooling ob-served in our climatic runs between 16 PAL and 4 PAL (~14 °C,Table 1). In our study, the most significant changes in the surface circu-lation occurwhen climate abruptly cools by crossing a tipping point thatcauses climatic instability. Without the latter, the ocean circulation ob-served by Herrmann et al. (2004) remained very stable. We refer hereto the study by Pohl et al. (2014) for an in-depth discussion about themechanisms accounting for the differential behaviour betweenGENESIS (i.e., the climatic model employed by Herrmann et al., 2004)and FOAM.
Christiansen and Stouge (1999) proposed that orbital variations anda faster spinning Ordovician Earth potentially induced major shifts inthe Hadley circulation. The authors estimated that during the Ordovi-cian, the planetary temperate low-pressure zones were located at 50°latitude and the subtropical high-pressure zones at 25° latitude,compared to respectively ~60° and 30° today. The possible shift of theHadley cell has not been tested here and constitutes a future researchtarget. It is however noteworthy that the circulation patterns modelledby Poussart et al. (1999), using an early Silurian length-of-day estimateof 21.5 h provided by Berger et al. (1989), do not substantially differfrom our findings.
6. Conclusions
For the first time, we detail ocean surface circulation patterns pro-duced by the coupled ocean-atmosphere general circulation modelFOAM. Our study confirms the patterns of earlier conceptual and nu-merical climate models of the Ordovician oceans. Modelled oceangyres and surface currents are consistent with those modelled byPoussart et al., (1999, Fig. 1, p. 544) for the latest Ordovician and byHerrmann et al. (2004, Fig. 4, pp. 392–393) for the Middle and Late Or-dovician. We demonstrate the largely model-independent nature ofthese circulation patterns. In addition, by running our models for awide range of pCO2 (i.e., 16 – 4 PAL) representative of Ordovician CO2
levels, as suggested by various recent papers and warranted by the
ocean surface circulation on atmospheric CO2 levels, Palaeogeogr.5.09.036
12 A. Pohl et al. / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx–xxx
virtual absence of pCO2 proxy data, we demonstrate that Ordovicianocean circulation was significantly pCO2-dependent.
The FOAM output constitutes a series of ocean-surface circulationmaps presented here in a synthetic manner. They are built on themost up-to-date palaeogeographical reconstructions of Torsvik andCocks (2009, 2013a) in order to facilitate comparison between themodels and palaeobiogeographical data, and include recently demon-strated Ordovician climatic behaviour (Pohl et al., 2014). The maps areavailable for the Early (480 Ma), the Middle (460 Ma) and the Late(440 Ma) Ordovician and for a wide range of CO2 values coveringmost of the range proposed for the Ordovician, in what we hope catersto the needs of the Ordovician deep-time research community.
Acknowledgements
This article is a contribution to the IGCP Project-591, ‘The Early toMiddle Paleozoic Revolution’. The authors acknowledge access grantedto the HPC resources of TGCC under the allocation 2014-012212 madeby GENCI. This research was funded through a CEA PhD grant CFR.T.R.A.V. acknowledges financial support from the French ‘AgenceNationale de la Recherche’ through grant ANR-12-BS06-0014‘SEQSTRAT-ICE’ and from the CNRS (action SYSTER). The authorsthank Mark Williams (University of Leicester) for editing a pre-production version of the manuscript. We also thank two anonymousreferees for their constructive reviews, and the guest-editors of this vol-ume for the invitation.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.palaeo.2015.09.036.
References
Berger, A., Loutre, M.F., Dehant, V., 1989. Influence of the changing lunar orbit on the as-tronomical frequencies of pre-Quaternary insolation patterns. Paleoceanography 4(5), 555–564. http://dx.doi.org/10.1029/PA004i005p00555.
Berner, R.A., 2006. GEOCARBSULF: a combined model for Phanerozoic atmospheric O2
and CO2. Geochim. Cosmochim. Acta 70, 5653–5664. http://dx.doi.org/10.1016/j.gca.2005.11.032.
Blakey, R., 2015. Global Paleogeography [WWW Document]. (URL: https://www2.nau.edu/rcb7 (accessed 12.06.15)).
Braconnot, P., Harrison, S.P., Kageyama, M., Bartlein, P.J., Masson-Delmotte, V., Abe-Ouchi,A., Otto-Bliesner, B., Zhao, Y., 2012. Evaluation of climate models using palaeoclimaticdata. Nat. Clim. Chang. 2, 417–424. http://dx.doi.org/10.1038/nclimate1456.
Christiansen, J.L., Stouge, S., 1999. Oceanic circulation as an element in palaeogeographicalreconstructions: the Arenig (early Ordovician) as an example. Terra Nova 11, 73–78.http://dx.doi.org/10.1046/j.1365-3121.1999.00229.x.
Cocks, L.R.M., 2000. The early Palaeozoic geography of Europe. J. Geol. Soc. 157, 1–10.http://dx.doi.org/10.1144/jgs.157.1.1.
Cocks, L.R.M., Fortey, R.A., 1982. Faunal evidence for oceanic separations in the Palaeozoicof Britain. J. Geol. Soc. 139, 465–478. http://dx.doi.org/10.1144/gsjgs.139.4.0465.
Cocks, L.R.M., Torsvik, T.H., 2005. Baltica from the late Precambrian to mid-Palaeozoictimes: the gain and loss of a terrane's identity. Earth Sci. Rev. 72, 39–66. http://dx.doi.org/10.1016/j.earscirev.2005.04.001.
Cocks, L.R.M., Torsvik, T.H., 2006. European geography in a global context from theVendian to the end of the Palaeozoic. In: Gee, D., Stephenson, R.A. (Eds.), EuropeanLithosphere Dynamics. Geological Society, London, pp. 83–95 http://dx.doi.org/10.1144/GSL.MEM.2006.032.01.05.
Cocks, L.R.M., Torsvik, T.H., 2007. Siberia, the wandering northern terrane, and its chang-ing geography through the Palaeozoic. Earth Sci. Rev. 82, 29–74. http://dx.doi.org/10.1016/j.earscirev.2007.02.001.
Cocks, L.R.M., Torsvik, T.H., 2011. The Palaeozoic geography of Laurentia and westernLaurussia: a stable craton with mobile margins. Earth Sci. Rev. 106, 1–51. http://dx.doi.org/10.1016/j.earscirev.2011.01.007.
Cocks, L.R.M., Torsvik, T.H., 2013. The dynamic evolution of the Palaeozoic geography ofeastern Asia. Earth Sci. Rev. 117, 40–79. http://dx.doi.org/10.1016/j.earscirev.2012.12.001.
Cocks, L.R.M., McKerrow, W.S., Van Staal, C.R., 1997. The margins of Avalonia. Geol. Mag.134, 627–636.
Cohen, K.M., Finney, S.C., Gibbard, P.L., Fan, J.-X., 2013. The ICS International Chrono-stratigraphic Chart updated, Episodes 36, 199–204 (URL: http://www.stratigraphy.org/ICSchart/ChronostratChart2014-10.pdf).
Cramer, B.C., Vandenbroucke, T.R.A., Ludvigson, G.A., 2015. High-Resolution Event Stratig-raphy (HiRES) and the quantification of stratigraphic uncertainty: Silurian examples
Please cite this article as: Pohl, A., et al., High dependence of OrdovicianPalaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.201
of the quest for precision in stratigraphy. Earth-Sci. Rev. 141, 136–153. http://dx.doi.org/10.1016/j.earscirev. 2014.11.011.
Crowley, T.J., Baum, S.K., 1995. Reconciling Late Ordovician (440 Ma) glaciation with veryhigh (14X) CO2 levels. J. Geophys. Res. 100 (1984–2012), 1093–1101. http://dx.doi.org/10.1029/94JD02521.
Dera, G., Donnadieu, Y., 2012. Modeling evidences for global warming, Arctic seawaterfreshening, and sluggish oceanic circulation during the Early Toarcian anoxic event.Paleoceanography 27, PA2211. http://dx.doi.org/10.1029/2012PA002283.
Donnadieu, Y., Goddéris, Y., Bouttes, N., 2009. Exploring the climatic impact of the conti-nental vegetation on the Mezosoic atmospheric CO2 and climate history. Clim. Past 5,85–96. http://dx.doi.org/10.5194/cp-5-85-2009.
Ekman, V.W., 1905. On the influence of the earth's rotation on ocean currents. Ark. Mat.Astron. Fys. 2, 1–53.
Endal, A.S., Sofia, S., 1981. Rotation in solar-type stars. I—Evolutionary models for thespin-down of the sun. Astrophys. J. 243, 625–640. http://dx.doi.org/10.1086/158628.
Finney, S.C., Gleason, J., Gehrels, G.E., Peralta, S.H., Aceñolaza, G., 2003. Early Gondwanaconnection for the Argentine Precordillera terrane. Earth Planet. Sci. Lett. 205,349–359. http://dx.doi.org/10.1016/S0012-821X(02)01063-4.
Gough, D.O., 1981. Solar interior structure and luminosity variations. Sol. Phys. 74, 21–34.http://dx.doi.org/10.1007/BF00151270.
Harper, D.A.T., Servais, T. (Eds.), 2013. Early Palaeozoic biogeography andpalaeogeography 38. Geological Society of London, Memoirs.
Harper, D.A.T., Rasmussen, C.M.O., Liljeroth, M., Blodgett, R.B., Candela, Y., Jin, J., Percival,I.G., Rong, J.Y., Villas, E., Zhan, R.B., 2013. Biodiversity, biogeography andphylogeography of Ordovician rhynchonelliform brachiopods. In: Harper, D.A.T.,Servais, T. (Eds.), Early Palaeozoic biogeography and palaeogeography. Geological So-ciety of London, Memoirs vol. 38, pp. 127–144. http://dx.doi.org/10.1144/M38.11.
Herrmann, A.D., Haupt, B.J., Patzkowsky, M.E., Seidov, D., Slingerland, R.L., 2004. Responseof Late Ordovician paleoceanography to changes in sea level, continental drift, and at-mospheric pCO2: potential causes for long-term cooling and glaciation. Palaeogeogr.Palaeoclimatol. Palaeoecol. 210, 385–401. http://dx.doi.org/10.1016/j.palaeo.2004.02.034.
IPCC, 2013. Climate Change 2013—The Physical Science Basis. Cambridge University Press,Cambridge, United Kingdom and New York, NY, USA.
Jacob, R.L., 1997. Low Frequency Variability in a Simulated Atmosphere Ocean System.University of Wisconsin-Madison.
James, N.P., 1997. The cool-water carbonate depositional realm. In: James, N.P., Clarke,J.A.D. (Eds.), Cool-water carbonates. SEPM, Spec. Publ. 56, pp. 1–20. http://dx.doi.org/10.2110/pec.97.56.0001.
Kiehl, J.T., Hack, J.J., Bonan, G.B., Boville, B.A., Williamson, D.L., Rasch, P.J., 1998. The NationalCenter For Atmospheric Research Community Climate Model: CCM3. J. Clim. 11,1131–1149. http://dx.doi.org/10.1175/1520-0442(1998)011b1131:TNCFARN2.0.CO;2.
Kucera, M., 1981. Planktonic foraminifera as tracers of past oceanic environments. In:Hillaire-Marcel, C., De Vernal, A. (Eds.), Proxies in Late Cenozoic palaeoceanography.Dev. in Mar. Geol., pp. 213–262 http://dx.doi.org/10.1016/S1572-5480(07)01011-1.3d (Elsevier).
Ladant, J.B., Donnadieu, Y., Dumas, C., 2014a. Links between CO2, glaciation and waterflow: reconciling the Cenozoic history of the Antarctic Circumpolar Current. Clim.Past 10 (6), 1957–1966. http://dx.doi.org/10.5194/cp-10-1957-2014.
Ladant, J.B., Donnadieu, Y., Lefebvre, V., Dumas, C., 2014b. The respective role ofatmospheric carbon dioxide and orbital parameters on ice sheet evolution at theEocene–Oligocene transition. Paleoceanography 29 (8), 810–823. http://dx.doi.org/10.1002/(ISSN)1944-9186.
Lees, A., 1975. Possible influence of salinity and temperature on modern shelf carbonatesedimentation. Mar. Geol. 19, 159–198.
Lees, D.C., Fortey, R.A., Cocks, L.R.M., 2002. Quantifying paleogeography using biogeogra-phy: a test case for the Ordovician and Silurian of Avalonia based on brachiopods andtrilobites. Paleobiology 28, 343–363. http://dx.doi.org/10.1666/0094-8373(2002)0282.0.CO;2.
Lefebvre, V., Donnadieu, Y., Sepulchre, P., Swingedouw, D., Zhang, Z.-S., 2012. Decipheringthe role of southern gateways and carbon dioxide on the onset of the Antarctic Cir-cumpolar Current. Paleoceanography 27 (4), PA4201. http://dx.doi.org/10.1029/2012PA002345.
Licht, A., van Cappelle, M., Abels, H.A., Ladant, J.B., Trabucho-Alexandre, J., France-Lanord,C., Donnadieu, Y., Vandenberghe, J., Rigaudier, T., Lécuyer, C., Terry Jr., D., Adriaens, R.,Boura, A., Guo, Z., Soe, A.N., Quade, J., Dupont-Nivet, G., Jaeger, J.J., 2014. Asian mon-soons in a late Eocene greenhouse world. Nature 513, 501–506. http://dx.doi.org/10.1038/nature13704.
Losch, M., Wunsch, C., 2003. Bottom topography as a control variable in an ocean model.J. Atmos. Ocean. Technol. 20, 1685–1696. http://dx.doi.org/10.1175/1520-0426(2003)0202.0.CO;2.
McKerrow, W.S., Scotese, C.R. (Eds.), 1990. Palaeozoic palaeogeography and biogeogra-phy. Geological Society of London, Memoirs 12.
Metcafle, I., 1998. Palaeozoic and Mesozoic geological evolution of SE Asian region: mul-tidisciplinary constraints and implications for biogeography. In: Hall, R., Holloway,J.D. (Eds.), Biogeography and Geological Evolution of SE Asia. Backhuys Publishers,Leiden, The Netherlands, pp. 25–41.
Metcafle, I., 2002. Tectonic history of the SE Asian–Australian region. Adv. Geoecol. 34,29–48.
Miall, A.D., Blakey, C., 2008. The Phanerozoic tectonic and sedimentary evolution of NorthAmerica. Sediment. Basins World 5, 1–29. http://dx.doi.org/10.1015/S1874-5997(08)00001-4.
Murphy, J.B., Dostal, J., Keppie, J.D., 2008. Neoproterozoic–Early Devonian magmatism inthe Antigonish Highlands, Avalon terrane, Nova Scotia: tracking the evolution ofthe mantle and crustal sources during the evolution of the Rheic Ocean.Tectonophysics 461, 181–201. http://dx.doi.org/10.1016/j.tecto.2008.02.003.
ocean surface circulation on atmospheric CO2 levels, Palaeogeogr.5.09.036
13A. Pohl et al. / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx–xxx
Murphy, J.B., Keppie, J.D., Nance, R.D., Dostal, J., 2010. Comparative evolution of theIapetus and Rheic Oceans: a North American perspective. Gondwana Res. 17 (2–3),482–499. http://dx.doi.org/10.1016/j.gr.2009.08.009.
Nardin, E., Goddéris, Y., Donnadieu, Y., Le Hir, G., Blakey, R.C., Pucéat, E., Aretz, M., 2011.Modeling the early Paleozoic long-term climatic trend. Geol. Soc. Am. Bull. 123,1181–1192. http://dx.doi.org/10.1130/B30364.1.
Pacanowski, R.C., Cannon, C.R. (Eds.), 1996. MOM 2. Documentation, user’s guide and ref-erence manual. GFDL Ocean technical report, No. 3.2. NJ, Princeton.
PALAEOSENS Project, 2012. Making sense of palaeoclimate sensitivity. Nature 491,683–691. http://dx.doi.org/10.1038/nature11574.
Pohl, A., Donnadieu, Y., Le Hir, G., Buoncristiani, J.-F., Vennin, E., 2014. Effect of the Ordo-vician paleogeography on the (in)stability of the climate. Clim. Past 10 (6),2053–2066. http://dx.doi.org/10.5194/cp-10-2053-2014.
Poulsen, C.J., Jacob, R.L., 2004. Factors that inhibit snowball Earth simulation.Paleoceanography 19, PA4021. http://dx.doi.org/10.1029/2004PA001056.
Poussart, P.F., Weaver, A.J., Barnes, C.R., 1999. Late Ordovician glaciation under high atmo-spheric CO2: a coupled model analysis. Paleoceanography 14, 542–558. http://dx.doi.org/10.1029/1999PA900021.
Rasmussen, C.M.O., 2013. Phylogeography of Ordovician–Silurian rhynchonelliformeanbrachiopods: tracking higher order distributional patterns, radiations and extinctionsin 4D. GFF http://dx.doi.org/10.1080/11035897.2013.847485.
Rothman, D.H., 2002. Atmospheric carbon dioxide levels for the last 500 million years.Proc. Natl. Acad. Sci. 99, 4167–4171. http://dx.doi.org/10.1073/pnas.022055499.
Rubinstein, C.V., Gerrienne, P., de la Puente, G.S., Astini, R.A., Steemans, P., 2010. EarlyMiddle Ordovician evidence for land plants in Argentina (eastern Gondwana). NewPhytol. 188, 365–369. http://dx.doi.org/10.1111/j.1469-8137.2010.03433.x.
Schmachtenberg, W.F., 2011. Paleolongitudinal estimates for paleocontinents derivedfrom interplate distances based on Late Ordovician bivalves. Paleobiology 37,438–444. http://dx.doi.org/10.1666/09015.1.
Scotese, C.R., 1997. Paleogeographic Atlas. University of Texas at Arlington, Arlington, TX,pp. 1–38.
Scotese, C.R., 2015. PALEOMAP Project [WWW Document]. (URL: http://www.scotese.com (accessed 01.06.15)).
Scotese, C.R., McKerrow, W.S., 1990. Revised worldmaps and introduction. In: McKerrow,W.S., Scotese, C.R. (Eds.), Palaeozoic palaeogeography and biogeography. GeologicalSociety of London, Memoirs vol. 12, pp. 1–21. http://dx.doi.org/10.1144/GSL.MEM.1990.012.01.01.
Scotese, C.R., McKerrow, W.S., 1991. Ordovician plate tectonic reconstructions. In: Barnes,C.R., Williams, S.H. (Eds.), Advances in Ordovician geology. Geological Survey ofCanada, Paper 90–9, pp. 271–282.
Semtner Jr., A.J., 1976. A model for the thermodynamic growth of sea ice in numerical in-vestigations of climate. J. Phys. Oceanogr. 6, 379–389. http://dx.doi.org/10.1175/1520-0485(1976)006b0379:AMFTTGN2.0.CO;2.
Servais, T., Owen, A.W., Harper, D.A., Kröger, B., Munnecke, A., 2010. The great ordovicianbiodiversification event (GOBE): the palaeoecological dimension. Palaeogeogr.Palaeoclimatol. Palaeoecol. 294, 99–119. http://dx.doi.org/10.1016/j.palaeo.2010.05.031.
Servais, T., Danelian, T., Harper, D.A.T., Munnecke, A., 2014. Possible oceanic circulationpatterns, surface water currents and upwelling zones in the Early Palaeozoic. GFF136, 229–233. http://dx.doi.org/10.1080/11035897.2013.876659.
Sheehan, P.M., 2001. The late Ordovician mass extinction. Annu. Rev. Earth Planet. Sci. 29,331–364. http://dx.doi.org/10.1146/annurev.earth.29.1.331.
Stampfli, G.M., Von Raumer, J.F., Borel, G.D., 2002. Paleozoic evolution of pre-Variscan ter-ranes: from Gondwana to the Variscan collision. In: Martinez Catalan, J.R., Hatcher,R.D., Arenas, R., Diaz-Garcia, F. (Eds.), Variscan–Appalachian Dynamics: The Buildingof the Late Paleozoic BasementGeological Society of America, Special Paper 363.Boulder, CO, pp. 263–280. http://dx.doi.org/10.1130/0-8137-2364-7.263.
Steemans, P., Le Herissé, A., Melvin, J., Miller, M.A., Paris, F., Verniers, J., Wellman, C.H.,2009. Origin and radiation of the earliest vascular land plants. Science 324, 353.http://dx.doi.org/10.1126/science.1169659.
Please cite this article as: Pohl, A., et al., High dependence of OrdovicianPalaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.201
Tobin, K.J., Bergström, S.M., 2002. Implications of Ordovician (~460 Myr) marine cementfor constraining seawater temperature and atmospheric pCO2. Palaeogeogr.Palaeoclimatol. Palaeoecol. 181, 399–417. http://dx.doi.org/10.1016/S0031-0182(01)00436-9.
Torsvik, T.H., Cocks, L.R.M., 2005. Norway in space and time: a Centennial cavalcade. Nor.J. Geol. 85, 73–85.
Torsvik, T.H., Cocks, L.R.M., 2009. BugPlates: Linking Biogeography and Palaeogeography,Software Manual [WWW Document]. (URL: http://www.geodynamics.no (accessed01.06.15)).
Torsvik, T.H., Cocks, L.R.M., 2013a. New global palaeogeographical reconstructions for theEarly Palaeozoic and their generation. In: Harper, D.A.T., Servais, T. (Eds.), EarlyPalaeozoic biogeography and palaeogeography. Geological Society of London,Memoirs vol. 38, pp. 5–24. http://dx.doi.org/10.1144/M38.2.
Torsvik, T.H., Cocks, L.R.M., 2013b. Gondwana from top to base in space and time.Gondwana Res. 24, 999–1030. http://dx.doi.org/10.1016/j.gr.2013.06.012.
Torsvik, T.H., Rehnström, E.F., 2001. Cambrian palaeomagnetic data from Baltica: implica-tions for true polar wander and Cambrian palaeogeography. J. Geol. Soc. Lond. 158,321–329. http://dx.doi.org/10.1144/jgs.158.2.321.
Torsvik, T.H., Rehnström, E.F., 2003. The Tornquist Sea and Baltica–Avalonia docking.Tectonophysics 362, 67–82. http://dx.doi.org/10.1016/S0040-1951(02)00631-5.
Vandenbroucke, T.R.A., Armstrong, H.A., Williams, M., Zalasiewicz, J.A., Sabbe, K., 2009.Ground-truthing Late Ordovician climate models using the paleobiogeography ofgraptolites. Paleoceanography 24, PA4202. http://dx.doi.org/10.1029/2008PA001720.
Vandenbroucke, T.R.A., Armstrong, H.A., Williams, M., Paris, F., Zalasiewicz, J.A., Sabbe, K.,Nõlvak, J., Challands, T.J., Verniers, J., Servais, T., 2010. Polar front shift and atmospher-ic CO2 during the glacial maximum of the Early Paleozoic Icehouse. Proc. Natl. Acad.Sci. 107, 14983–14986. http://dx.doi.org/10.1073/pnas.1003220107.
Vandenbroucke, T.R.A., Armstrong, H.A., Williams, M., Paris, F., Sabbe, K., Zalasiewicz, J.A.,2013. Late Ordovician zooplankton maps and the climate of the Early PalaeozoicIcehouse. In: Harper, D.A.T., Servais, T. (Eds.), Early Palaeozoic biogeography andpalaeogeography. Geological Society of London, Memoirs vol. 38, pp. 399–405.http://dx.doi.org/10.1144/M38.24.
Vandenbroucke, T.R.A., Emsbo, P., Munnecke, A., Nuns, N., Duponchel, L., Lepot, K.,Quijada, M., Paris, F., Servais, T., Kiessling, W., 2015. Metal-induced malformationsin early Palaeozoic plankton are harbingers of mass extinction. Nat. Commun. 6,7966. http://dx.doi.org/10.1038/ncomms8966.
Webby, B.D., Paris, F., Droser, M.L. (Eds.), 2004. The Great Ordovician BiodiversificationEvent. Columbia University Press, New York.
Wilde, P., 1991. Oceanography in the Ordovician. In: Barnes, C.R., Williams, S.H. (Eds.),Advances in Ordovician Geology. Geological Survey of Canada, pp. 283–298.
Wise, D.U., Ganis, G.R., 2009. Taconic Orogeny in Pennsylvania: a ~ 15–20 m.y. Apennine-style Ordovician event viewed from its Martic hinterland. J. Struct. Geol. 31 (9),887–899. http://dx.doi.org/10.1016/j.jsg.2008.03.011.
Yapp, C.J., Poths, H., 1992. Ancient atmospheric CO2 pressures inferred from natural goe-thites. Nature 355, 342–344. http://dx.doi.org/10.1038/355342a0.
Young, S.A., Saltzman, M.R., Bergström, S.M., Leslie, S.A., XU, C., 2008. Paired δ13Ccarb andδ13Corg records of Upper Ordovician (Sandbian–Katian) carbonates in NorthAmerica and China: implications for paleoceanographic change. Palaeogeogr.Palaeoclimatol. Palaeoecol. 270 (1–2), 166–178. http://dx.doi.org/10.1016/j.palaeo.2008.09.006.
Zhang, Z.-S., Yan, Q., Wang, H.-J., 2010. Has the Drake passage played an essential role inthe Cenozoic cooling? Atmos. Ocean. Sci. Lett. 3, 288–292.
Zhuravlev, A.Y., Riding, R. (Eds.), 2001. The Ecology of the Cambrian Radiation. ColumbiaUniversity Press, New York.
ocean surface circulation on atmospheric CO2 levels, Palaeogeogr.5.09.036
