Unravelling extrinsic and intrinsic factors of the early ...

Viewer
Transcript

Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

Contents lists available at ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p a l a e o

Unravelling extrinsic and intrinsic factors of the early Palaeozoic diversiﬁcation of blastozoan echinoderms Elise Nardin a,b,⁎, Bertrand Lefebvre c a b c

UMR 5563 LMTG, Université Paul Sabatier, Toulouse, France UMR 5561 Biogéosciences, Université de Bourgogne, Dijon, France UMR 5125 PEPS, Université Lyon 1, France

a r t i c l e

i n f o

Article history: Received 21 April 2009 Received in revised form 27 November 2009 Accepted 2 January 2010 Available online 11 January 2010 Keywords: Early Palaeozoic Evolution Blastozoa Echinodermata Palaeogeography Diversity

a b s t r a c t The Subphylum Blastozoa represents more than one third of the early Palaeozoic echinoderm fauna. A comprehensive database including all records of blastozoans was built to provide quantitative analyses of palaeogeography and diversity patterns and processes, for the 10 classes currently included in this subphylum during the early Palaeozoic. The global pattern of taxonomic diversity shows two peaks during the Cambrian Series 3 and the Late Ordovician intervals. In Cambrian times, the high taxonomic diversity seems to be related with a high turnover rate and a high endemicity of blastozoan genera, whereas in Ordovician times, the rise in diversity is associated with a low endemicity and a low turnover rate. The gradual increase of Ordovician diversity can be explained by the progressive geographic extension of blastozoan genera onto most palaeocontinental margins, associated to the long-term gradual sea level rise. This global pattern is subdivided into local trends to compare the relative inﬂuence of regional distributional patterns. Blastozoans occurred mainly on Gondwanan and Laurentian margins during the Cambrian. In both Ordovician and Silurian times, blastozoans colonised different provinces, characterised by varied latitudinal position and environmental conditions: the Laurentian province was mainly located in the equatorial latitudes (temperate to warm waters, mainly carbonate platforms), and both the northwestern periGondwanan and Asian provinces, in the intermediate to high latitudes (temperate to cold water, mainly siliciclastic environments). After the Hirnantian mass extinction, blastozoans ﬁnally recovered chieﬂy on Laurentian margins, where climate and environmental conditions were probably more favourable. The two diversiﬁcation events recorded for blastozoans in the early Palaeozoic concerned two distinct evolutionary faunas (Cambrian Evolutionary Fauna vs. Palaeozoic Evolutionary Fauna). They were driven by distinct processes, even if the Ordovician diversiﬁcation had indeed started during the middle Furongian with a medium origination event. Difference in processes might be related to the evolution of global environmental/ climatic conditions coupled with the position of continental masses. © 2010 Elsevier B.V. All rights reserved.

1. Introduction In recent years, palaeobiogeographic aspects and biodiversity patterns (diversity, disparity) of the Cambro–Ordovician radiation of metazoans have been extensively investigated (Sepkoski, 1979, 1981; Miller, 1997; Droser and Finnegan, 2003; Webby et al., 2004a; Munnecke and Servais, 2007; Servais et al., 2008, 2009). However, most studies focused on the same three taxa of marine invertebrates (articulate brachiopods, molluscs, and trilobites; Cope and Babin, 1999; Connolly and Miller, 2001; Adrain et al., 2004; Harper, 2006; Owen, 2008), whereas other signiﬁcant components of early Palaeozoic biota, such as echinoderms, remained largely neglected. Recent studies on early Palaeozoic echinoderms focused mostly on the Cambrian ⁎ Corresponding author. UMR 5563 LMTG, Université Paul Sabatier, Toulouse, France. E-mail addresses: [email protected] (E. Nardin), [email protected] (B. Lefebvre). 0031-0182/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.palaeo.2010.01.005

revolution and diversiﬁcation patterns of Ordovician assemblages (Sprinkle and Guensburg, 1995, 2003; Lefebvre and Fatka, 2003; Sprinkle and Guensburg, 2004; Lefebvre 2007a). Few analyses dealt with the ecology of primitive echinoderms (Dornbos, 2006), and they often concerned a limited geographic area and/or number of taxa (Sumrall et al., 1997, Nardin et al., 2009a). A few others combined two or three parameters, i.e., diversity, disparity, palaeogeography to investigate the diversiﬁcation of primitive echinoderms, such as crinoids and blastozoans (Foote, 1992) or stylophorans (Lefebvre et al., 2006). All these studies have suggested relatively different diversiﬁcation patterns in Laurentia and in the northwestern Gondwanan margin, and a provincial distribution as for the majority of echinoderms (Paul, 1976). The Subphylum Blastozoa is a major component of early echinoderm assemblages, being composed of 10 classes (blastoids, cinctans, coronoids, ctenocystoids, diploporans, eocrinoids, parablastoids, paracrinoids, rhombiferans, and solutes) (Sprinkle, 1973; Parsley, 1997, 1998; David et al., 2000; Sprinkle and Guensburg, 2004; Nardin et al.,

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

2009b). However, its phylogenetic status is still in question. Two distinct hypotheses exist: (1) one suggests that this group is paraphyletic and the ancestor of crinoids (Paul and Smith, 1984; Paul, 1988; Ausich, 1998; Smith, 2004, 2005; Zamora and Smith, 2008), and (2) the other one supports the monophyly of both blastozoans and crinoids (Sprinkle, 1973; David et al., 2000; Sprinkle and Guensburg, 2001, 2004; Frest, 2005). Moreover, very few studies have encompassed all classes at the same time for the Ordovician (e.g., Sprinkle and Guensburg, 2004). As a consequence, both palaeogeographic aspects and biodiversity patterns of the Cambrian–Ordovician diversiﬁcation of blastozoans are still poorly known. For this study, a comprehensive database including all records of blastozoans sensu David et al. (2000) and Nardin et al. (2009b) was built to provide a global pattern of taxonomic diversity for this subphylum during the early Palaeozoic (i.e., a 125 m.y. time interval). With one improbable blastoid (Broadhead, 1984), all blastozoan classes originated during this time span. Their dynamics has been so far little studied with analytical methods and mostly for the Ordovician period (Sprinkle and Guensburg, 2003, 2004). The main goal of this paper is to describe the diversity dynamics of early blastozoans in early Palaeozoic times, so as to, ultimately, discuss the possible inﬂuence of both intrinsic and extrinsic causes on this signal. 2. Materials and methods 2.1. Material The database was compiled from an exhaustive review of all published papers dealing with early Palaeozoic echinoderms or faunal communities. This main source of information was complemented by personal observations, and oral communications from P. Courville (Rennes University), O. Fatka (Charles University, Prague), and D. Vizcaïno (Carcassonne). Taxonomy follows the current consensus (Sprinkle, 1973; Parsley, 1997, 1998; David et al., 2000; Nardin et al., 2009b). In this phylogenetic scheme, the Subphylum Blastozoa contains ten classes (364 genera), including the controversial cinctans, ctenocystoids, and solutes (but see Jefferies, 1990; Smith, 2004, 2005; Zamora and Smith, 2008 for alternative interpretations of the last three classes). The 221 blastozoan genera currently recognised for the Cambro– Ordovician periods, have been listed with their stratigraphic, geographic, and environmental occurrences (see Appendix A). Only wellreferenced occurrences were used for the statistical analyses. Eleven geographic units were identiﬁed based on the deﬁnitions and the palaeogeographic reconstructions of Cocks and Fortey (1988), Cocks et al. (1997) and Cocks and Torsvik (2002), updated by recent works on speciﬁc areas (Cocks and Torsvik, 2005, 2007) (Fig. 1). The geographic units Alborz, Avalonia, Baltica, Laurentia, Sibumasu, South China have been used as well deﬁned in the literature (Cocks et al., 1997; Cocks and Torsvik, 2002). The northwestern peri-Gondwanan margin, sensu Cocks and Torsvik (2002) encompasses the North African, Western European and Central European terranes. The unit “North African Terranes” corresponds to the Moroccan, Algerian, Lybian basins; the unit “Western Europe terranes,” to the Iberian Peninsula (except the southwest Portugal), the French Armorican Massif, Massif Central, and Montagne Noire; and the unit “Central Europe terranes,” to the south Poland and Barrandian basins (Cocks and Torsvik, 2002). The two last units, “West Gondwana” and “East Gondwana” are represented by the western part of South America (including Argentina, Bolivia, west Brazil) and by the Australia and Tasmania, respectively (Cocks and Fortey, 1988). Stratigraphy follows updated versions of the Cambrian timescale of Geyer and Shergold (2000), the Ordovician timescale of Webby et al. (2004b) and Bergström et al. (2009), and Grahn (2006) for the Silurian timescale; all of them are using international stages and graptolite/ chitinozoan biozones for correlation among continents (Ogg et al., 2008). Consequently, in this study, several Ordovician time slices have been grouped in 11 time intervals to blend the interval duration between the three periods: early Tremadocian [Tr1-2], late Tremadocian

143

[Tr3], early Floian [Fl1-2], late Floian [Fl3], Dapingian [Dp1-3], early Darriwilian [Da1-2], late Darriwilian [Da3], Sandbian [Sa1-2], early Katian [Ka1-2], late Katian [Ka3-4], and Hirnantian [H1-2] (times slices from Bergström et al., 2009). Collection intensity is highly uneven. Information on Furongian faunas is underrepresented because of numerous sedimentologic and taphonomic biases (McGowan and Smith, 2007), as well as poorly-studied time intervals in the Asian terranes (Late Ordovician), the West and East Gondwanan regions (Cambrian Series 3 and Late Ordovician). In contrast, Laurentia, Baltica, and northwestern peri-Gondwanan terranes have been intensively documented (Miller, 1997). 2.2. Unstandardised methods The calculation of the generic diversity metric is based on the theoretical (RT-MSD) range for each genus (Alroy, 2000). Theoretical range is estimated by counting the number of time slices existing between the ﬁrst and the last appearance events (FAE and LAE, respectively). Results obtained with this dataset are consistent with those given by the observed range dataset (not shown here). Rareﬁed diversity calculation method is based on the occurrences number (Raup, 1975), slightly underestimated here because of the selection of the more precise ones. Normalised diversity (N-MSD), calculated from Cooper (2004), is assumed to minimise the impact of genera with a short stratigraphic extension and to reduce the proportion of sampling bias in the diversity signal (Alroy, 2000; Cooper, 2004). Per interval origination (p), extinction (q), and turnover (turn) rates were calculated according to the method of Sepkoski (1981). It takes into account all types of genera (three and two timers, and singletons). Turnover is the subtraction of origination and extinction rates. These metrics were also calculated for each geographic unit, using the range-through diversity (RT-MSD). Endemicity was estimated by counting the number of genera restricted to one geographic unit per time interval. Quality of regional databases was tested by the comparison of their rareﬁed diversity trajectories (Raup, 1975). Volatility is the absolute value of net geographical exchanges in an interval. This index was modiﬁed from its original expression, developed by Alroy (2000) for taxonomic purposes. 2.3. Statistical methods Statistical correlations between metrics were performed using cross-correlation. For each signal, new distribution results from the subtraction of the value at time t and the value at time t− 1. Subsequently, Kendal correlation processes (τ and Kp) have been applied on these new distributions, so as to avoid the effects of temporal series. Comparisons between two metrics were analysed by the Kolmogorov–Smirnov tests (KS) for paired variables. 2.4. Sampling standardisation Sampling artefact is minimised by the subsampling and standardisation processes applied to diversity and evolutionary analyses. Subsampling procedure corresponds to the random sampling of each time interval from the initial database. Diversity and evolution analyses were both performed for each subsampled dataset. The subsampling was repeated 1000 times. The new matrix was then used for the calculation of average values and their conﬁdence intervals (Manly, 1997). The standardisation process uses a random order of genera leading to a random presence/absence counting for each time slice. The standardised diversity values are referenced with the letter s (e.g., sRT-MSD). 2.5. Ordination methods Faunal similarities were estimated by calculating the value of the Dice index for each pair of geographical units. Cluster analysis

144

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

Fig. 1. Reconstruction of the palaeogeography of the southern hemisphere at 480 Ma. The geographic units used in this work are named and coloured in dark grey. Modiﬁed from Cocks and Torsvik, 2002.

used the dissimilarity distance of Raup–Crick (Raup and Crick, 1979) and the agglomerative nesting with group averaging method (corresponding to UPGMA) (Legendre, 1998). Non-parametric ordination technique (NMDS) was used to verify the expected relationships, in the case of small and non-normal datasets (Clarke, 1993). NMDS is also based on Raup–Crick distances. NeighbourJoining cluster and metric ordination techniques for presence/ absence data (PCO) were also calculated but are not shown here. Their results are consistent with UPGMA cluster and NMDS. All these analyses were computed with the libraries cluster, stats and vegan for R2.8.1 (R Development Core Team, 2008).

3. Diversity dynamics 3.1. Generic diversity patterns All calculated diversity curves display a pulsed increase of global diversity throughout the early Palaeozoic with two main peaks (Fig. 2): the ﬁrst one occurs in the Cambrian Series 3 (Stage 5Drumian), and the second one, in the early Late Ordovician (Sandbian–early Katian). The two peaks are relatively narrow, although a regular rise in diversity can be observed from the Furongian to the Darriwilian. Both diversiﬁcation events are followed

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

145

Table 1 Kolmogorov–Smirnov test (bottom triangular matrix) and Kendall correlation (top triangular matrix) between the main metrics of diversity, evolution and geography. NS: not signiﬁcant. *: signiﬁcant. ***: highly signiﬁcant. Kendall coefﬁcients are given if probabilities are at least signiﬁcant.

by a crisis. The late Cambrian crisis could correspond to a sampling artefact, the Furongian being poorly represented in the sedimentologic record (Sumrall et al., 1997; Vizcaïno et al., 2001; McGowan and Smith, 2007). The second crisis ends with the late Ordovician (Hirnantian) mass extinction. However, the decrease in diversity begins as early as the beginning of the Katian stage. Then the late Ordovician drop in diversity seems to correspond to a medium-term event rather than to an instantaneous one (Fig. 2). Finally, the postcrisis recovery is apparently relatively slow, with no signiﬁcant increase in diversity before the Telychian. Standardised and unstandardised curves exhibit similar distributions through the early Palaeozoic (KS test non-signiﬁcant, not shown). RT-MSD shows the highest diversity for each period, compared to the other methods (Fig. 2). The rareﬁed diversity (RAR-MSD) presents a similar pattern to RT-MSD. Their high correlation degree suggests that the global diversity pattern could be considered as a poorly biased signal (Table 1). The late Ordovician decrease depicted by the RAR-MSD metric is restricted to the Hirnantian interval, suggesting that the expression of the late Ordovician crisis is not an artefact for blastozoans. Rarefaction signiﬁcantly diminishes the Ordovician peak of about one third of the genera (from 96 to 62 genera during the Sandbian). This could

indicate that the late Ordovician diversity is controlled by the fossil record quality and by the occurrence number (also expected in Sprinkle and Guensburg, 2004). The latter hypothesis is conﬁrmed by the signiﬁcant correlation between the RT-MSD and RAR-MSD and the occurrence number per time interval (Table 1). In contrast to the curves of raw-diversity, normalised process (NMSD) globally decreases the number of genera for each time interval (Fig. 2). However the non-signiﬁcant Kolmogorov–Smirnov test between the raw (RT-MSD), the rareﬁed (RAR-MSD), and the normalised diversity metrics and their signiﬁcant cross-correlations (Table 1) indicate that these metrics reveal the same diversity signal even if they slightly differ in amplitude. The Cambrian peak declines from 32 (RT-MSD) to 16 genera (N-MSD). The slope of the Furongian– Early Ordovician rise is less pronounced in N-MSD curve. The Ordovician peak is smaller (from 96 to 65 genera for N-MSD) and more extended through time, from the early Darriwilian to the early Katian (as for RAR-MSD), compared to the peak obtained with the RTMSD signal (Sandbian to early Katian). The diversity is stable through the Llandovery and remains at a low level, ending with a slight decline. Standardised normalised diversity curves illustrate a clear shift downward, resulting in the quasi-disappearance of the Cambrian peak and of the Furongian–Early Ordovician rise (Fig. 2). StN-MSD and

Fig. 2. Generic diversity of the Subphylum Blastozoa through the early Palaeozoic. Unstandardised (Unst, raw line) and standardised (St, stippled line) diversities: rareﬁed (RAR, rectangles), theoretical range (RT, dots), and normalised diversity with singletons (norm, lozenges). Error bars represent the conﬁdence interval at 95%.

146

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

N-MSD show similar but shifted patterns. Random model signiﬁcantly predicts less diversity than the unstandardised pattern. Only two other diversity curves have been published so far for blastozoans, one for the Palaeozoic generic diversity (Foote, 1992) and one for Ordovician species richness (Sprinkle and Guensburg, 2004). The global curve of Foote (1992) shows a small rise in diversity and in disparity during the Cambrian, followed up by a rapid one during the Ordovician, with a peak during the Late Ordovician, and an abrupt decline during the Silurian. The detailed curve produced by Sprinkle and Guensburg (2004) shows a gradual increase in diversity from the latest Early Ordovician to the late Sandbian and a slow decrease until the Hirnantian. Both published patterns are similar with the one obtained here, and thus conﬁrm major trends in blastozoan diversity, even if Foote's study was not based on the whole subphylum. The resulting global blastozoan diversity pattern differs slightly from the generic diversity signals obtained for other marine invertebrate groups (Sepkoski, 1990; Miller, 1997; Webby et al., 2004a; Zhuravlev and Naimark, 2005; Servais et al., 2008). The Cambrian part of the blastozoan diversity signal appears mostly similar to that obtained for reef builder taxa, also characterised by a slow but regular increase during the Cambrian, interrupted by a slight decrease at the end of the Cambrian Series 3 (Zhuravlev and Naimark, 2005). These two patterns are comparable, even if archaeocyaths disappeared during the Furongian, because most of the blastozoan genera contributing to the Cambrian signal also disappeared before the late Cambrian Series 3. As for blastozoans, the diversity patterns described so far for other invertebrate taxa all exhibit a regular increase in the Early–Middle Ordovician time interval. However, they differ in the Late Ordovician, with either (1) an increase (e.g., nautiloids, sponges), (2) a plateau (e.g. brachiopods, bivalves) or (3) a gradual decline (bryozoans, graptolites) (Webby et al., 2004a; Servais et al., 2008). None of these patterns show the same major diversity peak and its consecutive rapid decline in the Late Ordovician. Bryozoans seem to exhibit the most similar signal to that of blastozoans in Ordovician times (Taylor and Ernst, 2004). As a consequence, both for the Cambrian and Ordovician time intervals, the diversity dynamics of blastozoans shows most similarities with that of ﬁlter-feeding organisms (reefs builders and then bryozoans), using hard substrate for attachment (Taylor and Ernst, 2004). 3.2. Evolutionary rates Origination, extinction, and turnover rates were calculated per time interval (p, q, and turn, respectively). Per interval turnover rate (turn) reaches a high plateau during the Cambrian period (0.36 ± 0.19), with a peak during the Paibian and a signiﬁcant drop at the Furongian–Tremadocian boundary (Fig. 3). Turn is low and stable during the Ordovician and the Llandovery (0.14 ± 0.07), in spite of a smaller late Katian–Hirnantian peak. Per interval origination rate (p) is high during the Cambrian. It drops slowly at the beginning of the Ordovician and remains at a medium level during the rest of the study period. Per interval extinction rate (q) shows a similar pattern with few differences: a lower level during the Cambrian–Middle Ordovician time interval and a peak during the late Katian–Hirnantian (Fig. 3A). The signiﬁcant cross-correlation between origination and extinction rates and the turnover rate suggests that the two ﬁrst ones contribute sub-equally to the latter one (Table 1). Two periods of extinction can be identiﬁed: (1) during the Cambrian Series 3 (Drumian to Guzhangian); and (2) during the Late Ordovician (Katian and Hirnantian). The latter fully explains the turnover peak and corresponds to the Hirnantian mass extinction. However it appears here that this event could be more gradual than previously expected for blastozoans (Sprinkle and Guensburg, 2004). For the rest of the study period, origination rates are higher than extinction ones, suggesting a continuous but slow diversiﬁcation event. Each diversiﬁcation period is followed by an extinction event and a clear faunal

change. At the Cambrian Series 3-Furongian boundary, few genera or classes survived (e.g., extinction of cinctans, ctenocystoids, gogiid eocrinoids). Most extinct genera possessed plesiomorphic morphologies (e.g., a stalk and an unorganised theca in gogiid eocrinoids), whereas the survivors or the newly appeared groups have derived characters (e.g., a derived stem and a reduced number of brachioles in solutes). The beginning of the second diversiﬁcation in Furongian times coincides with the appearance/expansion of genera with derived morphology, i.e., the most diverse blastozoan classes (e.g., stemmed eocrinoids, rhombiferans and diploporans), and later on, coronoids and paracrinoids. The beginning of the second extinction period corresponds to the rapid diversity decline of four classes (coronoids, eocrinoids, parablastoids, and paracrinoids), even if both evolutionary rates and diversity metrics are not statistically dependent (Table 1). Blastozoans have lower evolution rates compared to most other early Palaeozoic benthic marine invertebrate fauna, including crinoids (Webby et al., 2004a; Peters and Ausich, 2008). These low rates can be explained by several hypotheses: (1) the blastozoan fossil record could be more biased than that of all other invertebrates (blastozoans having been less studied than most other marine taxa); (2) Furongian and Lower Ordovician strata are less represented in the sedimentary records in several regions (McGowan and Smith, 2007); and (3) blastozoans possibly had a low metabolism associated with medium genetic capacities to survive in varied environments with a reduced need of genus or species originations. 3.3. Endemicity Endemicity is the proportion of genera restricted to one geographic unit. It strongly ﬂuctuates through the early Palaeozoic (Fig. 3B). Endemicity is high and varied for Cambrian genera (average at 80.1 ± 20.6%). The Cambrian Series 3 medium level coincides with an increase in diversity, even if these two signals are clearly different and not correlated (Table 1). This small drop in endemicity could be driven by the higher dispersion of some genera of cinctans and ctenocystoids (e.g., Ctenocystis, 5 units) and gogiid eocrinoids (e.g., Gogia, 4 units), being immediately cosmopolitan after their ﬁrst appearance. Endemicity is lower during the rest of the study period, suggesting that Ordovician and Silurian faunas rapidly became more widely distributed. Endemicity remains relatively stable during the Early–Middle Ordovician (20.3 ± 2.9%) and reaches a slightly higher plateau from the Late Ordovician to the Llandovery (39.1 ± 2.5%). Continental connectivity was high in the Cambrian Series 2–3. This could explain why some genera were able to become cosmopolitan soon after their origination. Some other genera remained endemic, probably because they could have been specialised to speciﬁc environmental conditions. Endemicity should have increased through the Cambrian as continental break-up continued. However, the opposite pattern is observed here for blastozoans. Their endemicity decreases while the habitats become more partitioned. Blastozoans could have developed abilities to colonise closed marine basins, even separated by geographic barriers, gradually in the late Cambrian and more quickly in the Early–Middle Ordovician, mostly in the periGondwanan margins (Lefebvre and Fatka, 2003). The Hirnantian mass extinction apparently had no major impact on the endemicity signal. This result is a bit unexpected, as the Hirnantian regression (and the associated habitat partitioning) should have led to higher values for endemicity (Fig. 3B). However, endemicity is here a proportion of genera. The geographic dynamics of the crisis survivors was possibly affected by the strong environmental changes. 3.4. Discussion The Cambrian rise in diversity coincides with both high per interval origination rate and high endemicity. In contrast, the

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

147

Fig. 3. A. Fluctuations of the evolutionary rates through the early Palaeozoic. p: origination rate; q: extinction rate; turn: turnover rate. The two kinds of rates have been calculated according to Alroy (2000) and Foote (2000). B. Proportional endemicity of genera in all palaeocontinental margins.

Drumian–Guzhangian drop in diversity coincides with both a period of high extinctions and a high endemicity. These results indicate that the ﬁrst peak in blastozoan diversity is driven by the generic turnover in restricted geographic areas. This interpretation is consistent with the generally admitted postulate, which suggests that main processes of the Cambrian diversiﬁcation are the evolutionary rates (Sepkoski,

1979, 1981; Zhuravlev and Riding, 2001; Lieberman, 2003). The Furongian–Ordovician rise in diversity starts with relatively high origination rates and continues with a clear diversiﬁcation event, even if both origination rate and endemicity decrease (Figs. 2, 3A–B). The Furongian–Ordovician increase in diversity was possibly driven ﬁrst by high generic originations, and later on, by both background

148

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

Fig. 4. Rarefaction of the generic diversity as a function of occurrences, subdivided into geographical units. Av: Avalonia; Ba: Baltica; CE: Central Europe; EG: East Gondwana; La: Laurentia; NA: North Africa; SC: South China; Si: Sibumasu; WE: Western Europe; WG: West Gondwana. Grey areas represent the 95% conﬁdence intervals.

originations and an increase of the generic resistance because of the wide geographic extension. The probability of taxa survival increases with the possession of a large geographic range (Powel, 2007). A genus occurring in several geographic units is more likely to resist local extinction events and thus, to keep contributing to the global generic diversity. This hypothesis is coherent with (1) the gradual diversity decrease through the Late Ordovician linked with a high endemicity (less resistance to environmental changes) and (2) the medium-level diversity during mass extinction, which could be explained by the survival of widespread genera restricted to protected geographic areas (e.g., Heliocrinites, Holocystites). Recovery after the Hirnantian crisis seems to be slow, as suggested by the low values observed for diversity, evolution rates, and endemicity. The late Ordovician mass extinction impacted blastozoans during the whole Llandovery. This observation differs from patterns based on the Paleobiology database published by Krug and Patzkowsky (2007) (brachiopods, trilobites, anthozoans, and bivalves). They show that diversity and evolution rates increased since the Aeronian. The evidence of two successive diversiﬁcation events (linked or not) in blastozoans is in good agreement with their frequent identiﬁcation as members of both the Cambrian Evolutionary Fauna (CEF) and the Palaeozoic Evolutionary Fauna (PEF) (Sprinkle, 1995; Sprinkle and Guensburg, 1995, 2004). In this scheme, CEF taxa would be characterised by a high turnover, a high endemicity and then it would mainly exist in the Cambrian Series 2–3. The PEF would correspond to taxa ranging from the Furongian to the Llandovery, with a low turnover, and a medium to low endemicity. However, the distinction between CEF and PEF is not as clear as in other coeval taxa (e.g., trilobites; Adrain et al., 2000, 2004). The PEF can be already present in the Cambrian fossil record and reaches high diversity values during the rest of Palaeozoic times, whereas the CEF is mostly

restricted to Cambrian–Ordovician times, and shows its highest diversity during the Cambrian (Sepkoski, 1990; Sprinkle, 1995; Sheehan, 2001). The two diversiﬁcation events in blastozoans reﬂect the separation between CEF and PEF blastozoan taxa. An additional clue is provided by the similarity of the signals of blastozoan diversity and evolution with those of typical CEF elements during the Cambrian (archaeocyaths), and PEF taxa during the Ordovician–Llandovery (bryozoans; Sheehan, 2001). The timing of the second increase in diversity suggests that the Ordovician diversiﬁcation of blastozoans was already well under way as early as the Furongian. Turnover rates at that time were as high as those documented for the Cambrian diversiﬁcation. Moreover, a signiﬁcant part of the Furongian genera shows strong afﬁnities with younger Ordovician assemblages (PEF), with the ﬁrst records of typical, more derived morphologies: (1) bowl-shaped holdfasts in Spanish and Iranian stemmed blastozoans from the Cambrian Series 3 (Rozhnov, 2002; in a different sense of Clausen, 2004), and (2) abundant remains of solutes, glyptocystoids rhombiferans, and stemmed eocrinoids in the Furongian (e.g., Barroubiocystis, Tatonkacystis; Sumrall et al., 1997; Ubaghs, 1998). All Cambrian blastozoans belonging to the CEF become extinct at the Cambrian Series 3– Furongian boundary (e.g., cinctans, gogiid-like eocrinoids, lepidocystids). All these clues suggest a certain continuation between the Cambrian diversiﬁcation event and the Ordovician one (as proposed by Droser and Finnegan, 2003). Moreover, as already pointed out, the low values observed for the late Cambrian blastozoan diversity coincide with a very poor fossil and sedimentary record in the Furongian. Consequently, the possibility that the distinction between the two biodiversiﬁcation events could be purely artiﬁcial (and caused by strong taphonomic bias) cannot be entirely avoided. This alternative view will need to be tested by further investigations.

Fig. 5. Regional generic diversity. A. Geographic volatility. B. Proportional diversity for each geographic unit. C. Origination (p) and extinction (q) rates for each geographical unit. Grey squares represent the origination rate, and black lozenges, the extinction rates. Same abbreviations as in Fig. 4 caption. Al: Alborz. Scale bars correspond to the 95% conﬁdence interval.

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

149

150

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

4. Palaeogeographical signal 4.1. Regional databases Rarefaction curves for all nine geographic units are presented in Fig. 4. Their comparison gives the possibility to identify three pools of units, based on the virtual coincidence of their trajectory after adjustments for sample size. Different aggregations indicate that diversity is not static between pools. The ﬁrst pool, including Laurentia (La), Western and Central Europe (WE and CE), contains the most investigated units, with the most diverse and supposedly better-known fossil record. The second pool includes Baltica (Ba), Avalonia (Av), North Africa (NA), and less obviously Sibumasu (Si) and the East Gondwanan margin (EG). These regions share a similar trajectory (same slope). Their diversity is lower and equal or probably less complete to that of regions of the ﬁrst pool. However, the values obtained for the three ﬁrst regions of the second pool (Ba, Av, and NA) can be considered as relatively realistic, because of the shape of the curve (elongate and nearly reaching the ﬁnal plateau). Recent works about Si and EG suggest that their future enriched fossil record may resemble the diversity of the three ﬁrst units (Cocks et al., 2005; Lefebvre et al., 2005). The last pool contains the less-known regions (South China (SC) and West Gondwana (WG)), although signiﬁcant advances were recently made in the knowledge of their blastozoan faunas (Aceñolaza and Gutiérrez-Marco, 2002; Zhao et al., 2007). When interpreting the regional diversity and evolution patterns, the artiﬁcial high diversity in Laurentia needs to be compared with the situation in other members of its pool, or with the artiﬁcially underestimated diversity of Si, SC, EG and WG. In these areas, short and low-sloped rarefaction curves reﬂect either single but extensive records of some epochs (e.g., Early Ordovician in WG, Middle Ordovician in Si and WG) or more continuous, but non-extensive records (e.g., SC). 4.2. Regional patterns and volatility Regional diversity was estimated based on the normalisation process (N-MSD), which seems to offer the best compromise between all different tested methods. Volatility reveals dissimilarity between all geographic units between two successive time intervals (Fig. 5). This index is high from the Cambrian to the early Tremadocian, with three peaks in the Cambrian Stage 5, the Paibian and the lowermost Tremadocian. Then it drops to a low plateau for the rest of the study period. This result indicates that the occupation of geographic units during the Cambrian was highly dissimilar, whereas it was relatively homogeneous from the late Tremadocian to the Llandovery (Fig. 4). This interpretation is consistent with the low level of endemicity observed for Ordovician times (Fig. 3B). The volatility index does not seem to be affected by the Hirnantian mass extinction, whereas it clearly shows the decline in diversity from the Furongian to the late Tremadocian. This conﬁrms that the late Cambrian diversity pattern is strongly biased by the poor quality of its corresponding sedimentary and fossil records. This is also in good accordance with the above-mentioned hypothesis that genera with a large geographic range had better survival selectivity to the Hirnantian crisis. Although all geographic units are not equally well documented, it is remarkable that the diversity patterns observed for the less wellknown units are, in spite of a more patchy fossil record, relatively consistent with those obtained for better-studied units (Fig. 5). Earliest blastozoans were restricted to Laurentia, Baltica, South China, and North African and West European terranes margins. Most of them are endemic genera (medium to low regional originations rates; Fig. 5). Moreover, Laurentia and peri-Gondwana associated to Baltica were clearly separated by the Iapetus Ocean during the Cambrian Series 2 and 3 (Cocks and Torsvik, 2002). As a consequence, these

results question the possibility (1) of two distinct geographic origins for blastozoan faunas (Laurentia vs. Gondwana + Baltica), (2) that the lack of fossil record older than late Cambrian Series 2 could hide the existence of a single origination point, and/or (3) homogeneous blastozoan faunas could have occurred on both sides of the Iapetus Ocean during Cambrian Series 2 and then evolved independently, while the ocean became mature. A peak in global diversity is observed in all geographic units from the Cambrian stage 4 to the stage 5. This time interval is characterised by a high proportion of endemic genera in both Laurentia and the peri-Gondwanan regions, and a lower proportion of them in both Baltica and Avalonia (Fig. 5). The paucity of cosmopolitan taxa (e.g., Ctenocystis, Gogia) during this time interval suggests that migrations were limited between neighbouring geographic units. The high proportion of endemic genera possibly indicates that the geographic dynamism of Cambrian blastozoans was low. Early Furongian faunas were apparently restricted to the margins of both Laurentia and Baltica. This distribution will need to be conﬁrmed by future ﬁeld investigations (possibility of strongly biased sedimentary and fossil records). The second global rise in diversity seems to begin on Laurentian margins (medium p), and to continue mostly in the epicontinental seas of Central–Western Europe and North Africa, and at a lesser degree on the Avalonian part of the peri-Gondwanan margin. Sibumasu, South China, and Alborz have yielded mostly endemic faunas, allied with few typical peri-Gondwanan genera (e.g., Cheirocystella, Macrocystella, Rhopalocystis). Most genera have restricted geographic ranges suggesting a certain provincialism for Lower Ordovician genera. After the late Early Ordovician, diversity increases in low to intermediate latitude regions (e.g., Baltica, Laurentia, Avalonia), whereas it remains stable in high latitude, polar regions (e.g., Central and Western Europe, North Africa). This observation conﬁrms previous qualitative works (Paul, 1976; Smith, 1990; Sprinkle, 1992). The geographic distribution of blastozoans in major regions remains about the same in Middle and Late Ordovician epochs. This is consistent with the low volatility and the medium origination and extinction rates. The global rise in diversity more likely results from migrations and wide geographic distribution of genera rather than from their turnover (e.g., distribution of Asteroblastus, Echinosphaerites, Tholocystis through peri-Gondwanan terranes, and Baltica; Cheirocrinus, and Pleurocystites in Asian regions, Baltica, Laurentia, and peri-Gondwanan margins). Increase in faunal exchanges between the different geographic units during Middle to Late Ordovician times was also reported for other invertebrate marine taxa (Gutiérrez-Marco et al., 1999; Cocks and Torsvik, 2002). The Hirnantian crisis had strong effects on blastozoan faunas from Baltica, Central Europe, and North Africa, and a more limited effect on those from Avalonia and Laurentia (Fig. 5). This turnover event coincides with the diversity rise in Laurentia and Avalonia, and with a clear decrease in the other regions from the Katian to the Hirnantian. As expected, blastozoans occurring in peri-Gondwanan regions were more affected by icehouse conditions in the early Katian-early Rhuddanian time interval (Le Héron and Craig, 2008). In low to intermediate latitude regions, blastozoans possibly beneﬁted from warmer seawater temperatures in epicontinental seas (Kolata et al., 1996). The recovery during the Silurian is slow, and it seems to occur chieﬂy in the two main less impacted regions (Laurentia and Avalonia; Fig. 5).

4.3. Palaeogeographic blastozoan distribution Time intervals were grouped in larger stratigraphic units to synthesise information. Faunal similarity between geographic units was calculated using Dice similarity index (Tables 2–5), hierarchical clusters, and multivariate ordination analysis (NMDS). For some periods, clusters and NMDS seem to show different patterns.

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

151

Table 2 Dice similarity index between geographic units colonised during the Cambrian. Cambrian Series 2–3

Western Europe

Avalonia

Baltica

East Gondwana

Laurentia

North Africa

Central Europe

South China

Western Europe Avalonia Baltica East Gondwana Laurentia North Africa Central Europe South China

0.00 0.18 0.09 0.26 0.19 0.11 0.13 0.00

0.00 0.17 0.36 0.11 0.25 0.11 0.00

0 0.18 0.10 0.20 0.10 0

0.00 0.21 0.00 0.11 0.00

0.00 0.00 0.07 0.00

0.00 0.00 0.00

0.00 0.00

0.00

Furongian

Western Europe

Baltica

Laurentia

Central Europe

Sibumasu

West Gondwana

Western Europe Baltica Laurentia Central Europe Sibumasu West Gondwana

0.00 0.00 0.00 0.00 0.00 0.00

0.00 0.00 0.00 0.00 0.00

0.00 0.00 0.00 0.25

0.00 0.50 0.40

0.00 0.40

0.00

4.3.1. Cambrian Blastozoan faunas from peri-Gondwanan regions seem to be relatively homogeneous during Cambrian series 2 and 3, except for Central Europe (Table 2, Fig. 6). This conclusion is supported by the existence of few cosmopolitan genera (e.g., Ctenocystis and Gogia), and also because blastozoans were mostly endemic during this period (e.g., primitive solutes in Laurentia; Sinoeocrinus and Globoeocrinus in South China). The shared genera show the existence of faunal exchanges between Laurentia and the northwestern peri-Gondwana margin. During the early Cambrian Series 3, blastozoans are newly originated in most occupied regions, except in North Africa and Western Europe, where invasive or surviving taxa contribute signiﬁcantly to the diversity (Fig. 7). The question of the geographic origin of blastozoans is still in question. Typical Cambrian genera occur contemporaneously in Laurentia, South China, and Central– Western Europe. However, faunal exchanges were apparently reduced between distant units, and much stronger between neighbouring units. This could indicate that blastozoans preferentially colonised close geographic regions rather than distant ones, suggesting a poor dispersal ability for these genera. Other taxa also show this pattern (e.g., trilobites, small shelly fossils; Álvaro et al., 2007). One exception is the contemporaneous presence of few cosmopolitan genera between the two sides of the Iapetus ocean (Table 2, Fig. 7), also shown by trilobites and brachiopods (Palmer, 2005). Faunal similarity between the two sides of the Iapetus Ocean increases during the Cambrian, even if many genera remain endemic. The importance of the oceanic dispersion is still in question even if faunal connection exists. As noted above, the Iapetus Ocean does not seem to constitute a natural barrier for blastozoans during the Cambrian. They could have been present earlier in the different regions but not yet detected in the fossil record; alternatively, cosmopolitan genera could have had developed special abilities to

cross a mature ocean (such as cold-water adaptation or larval dispersal as suggested for crinoids; Rozhnov, 2007). Other hypotheses are the presence of small intra-Iapetan terranes, which could have served as stepping stones for the faunal migration (Harper et al., 1996) or strong oceanic currents (Palmer, 2005). The Furongian blastozoan geographic distribution is reduced to six units. Relationships between geographic units clearly differ between Cambrian Series 2–3 and Furongian. 4.3.2. Early Ordovician The Early Ordovician is characterised by a gradual reorganisation of faunal associations, as well as by a signiﬁcant increase in the quality of available data. In Early Ordovician times, blastozoans were distributed on ten geographic units with (1) well-identiﬁed, endemic assemblages in Baltica, (2) relatively homogeneous faunas on northwestern peri-Gondwanan margins, and (3) strong links between peri-Gondwanan faunas from the Middle Eastern Terranes (e.g., Alborz), South China, and Sibumasu (Table 3, Fig. 7). The strong generic afﬁnities between blastozoan faunas from North Africa, Central, and Western Europe were used to deﬁne a “Mediterranean province” (sensu Havlíček et al., 1994). Strong endemism in Baltica deﬁnes the Baltic province, already recognised for brachiopods (Harper et al., 1996). Faunal similarity of regions at intermediate latitudes (Laurentia, Avalonian peri-Gondwana, and the pool of eastern peri-Gondwanan regions) indicates that blastozoan geographic distribution could correspond to a latitudinal cline with migrations occurring between relatively distant units. Provinces clearly differ by their faunal contents. The Mediterranean province houses mostly eocrinoids, solutes, and diploporans, whereas Laurentian and Baltic provinces are more dominated by eocrinoids and rhombiferans. Medium-level endemism still exists within extended peri-Gondwanan regions for blastozoans, as well as for other

Table 3 Dice similarity index between geographic units colonised during the Early Ordovician. Early Ordovician

Alborz

Western Europe

Avalonia

Baltica

East Gondwana

Laurentia

North Africa

Central Europe

Sibumasu

South China

Alborz Western Europe Avalonia Baltica East Gondwana Laurentia North Africa Central Europe Sibumasu South China

0.00 0.22 0.40 0.17 0.33 0.25 0.15 0.20 0.40 0.50

0.00 0.20 0.21 0.31 0.27 0.33 0.13 0.33 0.22

0.00 0.15 0.29 0.22 0.14 0.18 0.33 0.40

0.00 0.14 0.22 0.17 0.11 0.15 0.17

0.00 0.33 0.13 0.17 0.44 0.33

0.00 0.12 0.14 0.36 0.25

0.00 0.26 0.14 0.15

0.00 0.18 0.20

0.00 0.40

0.00

152

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

Table 4 Dice similarity index between geographic units colonised during the Middle–Late Ordovician. Middle Ordovician

Alborz

Western Europe

Baltica

Laurentia

North Africa

Central Europe

South China

Sibumasu

West Gondwana

Alborz Western Europe Baltica Laurentia North Africa Central Europe South China Sibumasu West Gondwana

0.00 0.19 0.22 0.00 0.31 0.25 0.39 0.24 0.00

0.00 0.17 0.13 0.25 0.19 0.18 0.16 0.00

0.00 0.10 0.06 0.14 0.18 0.21 0.07

0.00 0.00 0.00 0.07 0.09 0.13

0.00 0.15 0.26 0.11 0.00

0.00 0.14 0.17 0.00

0.00 0.38 0.00

0.00 0.00

0.00

Sandbian/Katian

Western Europe

Avalonia

Baltica

East Gondwana

Laurentia

North Africa

Central Europe

South China

Sibumasu

Western Europe Avalonia Baltica East Gondwana Laurentia North Africa Central Europe South China Sibumasu

0.00 0.22 0.24 0.11 0.09 0.27 0.28 0.15 0.15

0.00 0.31 0.14 0.08 0.21 0.19 0.19 0.14

0.00 0.10 0.10 0.15 0.25 0.14 0.14

0.00 0.00 0.13 0.00 0.00 0.22

0.00 0.07 0.07 0.07 0.03

0.00 0.26 0.13 0.13

0.00 0.19 0.19

0.00 0.33

0.00

Table 5 Dice similarity index between geographic units colonised during the latest Ordovician– Llandovery. Hirnantian

Western Europe

Western Europe Avalonia Baltica Laurentia North Africa Central Europe

0.00 0.17 0.00 0.00 0.00 0.29

Llandovery

Western Europe

Western Europe Avalonia Baltica Laurentia Central Europe

0.00 0.00 0.00 0.00 0.40

Avalonia

Baltica

Laurentia

0.00 0.13 0.00 0.00 0.15

0.00 0.00 0.00 0.00

0.00 0.00 0.00

North Africa

Central Europe

0.00 0.00

0.00

Avalonia

Baltica

Laurentia

0.00 0.00 0.10 0.00

0.00 0.18 0.00

0.00 0.00

Central Europe

0.00

invertebrate faunas (e.g., brachiopods, trilobites; Harper et al., 1996; Tychsen and Harper, 2004; Owen, 2008). However, blastozoan distribution does not show the same strong relationships between East Gondwana, Laurentia, and North China, as evidenced for brachiopods, graptolites, and pelagic trilobites (Fortey and Cocks, 2003). 4.3.3. Middle Ordovician–Katian The Dice indices calculated for the Middle Ordovician suggest that blastozoans were proportionally distributed in all geographic regions (Table 4). A large proportion of shared genera existed between eastern peri-Gondwanan regions, South China, and Sibumasu; this conﬁrms the faunal identity of this pool of regions (called here Asian province) and the latitudinal impact on blastozoan geography. The Asian Province is still connected with the Mediterranean one and also with Baltica (Fig. 6). Low endemism and relatively high average faunal similarity both reﬂect an increase of migrations, mostly from the Mediterranean Province to the Asian one (e.g., Aristocystites, Glypto-

sphaerites, Tholocystis) and to Baltica (e.g., Echinosphaerites). Avalonia and Laurentia keep their faunal identity acquired since the Early Ordovician with high endemism and low Dice index (Table 4). In Middle Ordovician times, the palaeogeographic distribution of most supra-generic taxa appears relatively well delimited in space, with rhombiferans (e.g., cheirocrinids) still present in Laurentian and Baltic provinces, and diploporans dominating peri-Gondwanan assemblages (e.g., calixids, sphaeronitids). Moreover, two new classes of blastozoans (parablastoids and paracrinoids) very likely originated within the Laurentian Province in late Early Ordovician times (also shown by Sprinkle and Guensburg, 2004). However, the existence of a few cosmopolitan genera discovered in Alborz, Baltica, Central and Western Europe, and Sibumasu suggests the existence of migration paths from the polar regions and/or Baltica to the Asian province (Lefebvre et al., 2005). The Laurentian Province remained isolated during the early Late Ordovician (few invasive taxa, Figs. 7,8), with some taxa restricted to this region (e.g., paracrinoids). One new taxonomic group originates in this area (coronoids). This distinct identity of Laurentian faunas was also evidenced for other marine invertebrates, such as nautiloids (Cricks, 1990). Mediterranean and Asian provinces appear as clearly distinct by all statistical analyses (Figs. 6–8, Table 2). Analyses also tend to conﬁrm the debated geographic proximity of the Central Europe terrane (also called Bohemia, Barrandian terrane, or Perunica) with the other regions from the peri-Gondwana margin, as suggested by Krs et al. (2001) (but see Havlíček et al., 1994; Cocks and Torsvik, 2002). Regional diversity seems to be equally explained by surviving, newly originated, and invasive taxa (Fig. 7). This could suggest an increase of faunal exchanges between the different provinces, consistent with the low global endemicity (Fig. 3B). In contrast to older time intervals, diploporans, rhombiferans, and solutes are present in all regions. Since the Sandbian, the distribution of numerous genera indicates a preferential migration of warm-water faunas from the Asian Province and Baltica to the Mediterranean regions (e.g., Caryocrinites, Heliocrinites, Stichocystis) and stronger afﬁnities between Avalonia and Baltica. The latter can be explained by the continental drift of Avalonia towards Baltica, initiated by the

Fig. 6. Cluster and ordination analyses of the dataset for the Cambrian to Middle Ordovician periods. Dendrograms from cluster analysis are shown at left, with percentages of node robustness. Ordination results at right, with generic scores plotted by geographic units. Cluster analysis agglomerative coefﬁcients and NMDS stresses are added under each analysis. Grey areas underline the strongest similarities.

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

153

154

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

Fig. 7. Proportional contribution of survivorship, genus origination, and invasion to standing diversity in early Cambrian Series 3, early Sandbian, early Katian, and early Rhuddanian. Error bars are 95% conﬁdence intervals.

closing of the Rheic Ocean during the late Middle–Late Ordovician (Cocks and Torsvik, 2002). During the middle–late Katian, new migration pathways are used between the peri-Gondwanan regions and Baltica (e.g., Corylocrinus, Protocrinites). Province distribution and intensiﬁed migrations during the Middle–Late Ordovician were already reported for blastozoans based on qualitative study (Paul, 1976; Smith, 1990; Lefebvre and Fatka, 2003; Sprinkle and Guensburg, 2003; Lefebvre, 2007a), and for stylophorans (Lefebvre, 2007b).

They are also observed for other invertebrate marine taxa (e.g., bryozoans, brachiopods, sponges, trilobites; Carrera and Rigby, 1999; Anstey et al., 2003; Tychsen and Harper, 2004). 4.3.4. Hirnantian-Llandovery (early Silurian) Blastozoan geographic dispersion is strongly reduced during the Hirnantian glaciation event, with an increase of raw endemism (Figs. 3, 8; Table 5). Blastozoans can be grouped into two main

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

155

Fig. 8. Cluster and ordination analyses of the dataset for Late Ordovician and Llandovery. Dendrograms from cluster analysis are shown at left, with percentages of node robustness. Ordination results at right, with generic scores plotted by geographic units. Cluster analysis agglomerative coefﬁcients, and NMDS stresses are added under each analysis. Grey areas underline the strongest similarities.

assemblages: (1) genera adapted to cold or deep water in siliciclastic platforms were restricted to high latitudes regions (Central and Western Europe); and (2) genera adapted to temperate to warm water in carbonate platforms survived in the less affected ecosystems of the low to intermediate latitude regions (Avalonia, Baltica, and Laurentia). Recovery from the Hirnantian crisis occurred in the same geographic units (Fig. 8; Table 5). The Rhuddanian diversity results from the combination of the presence of surviving taxa in Avalonia,

Central and Western Europe, along with the appearance of new taxa in Baltica and Laurentia (Fig. 7). The clear distinction between low and high latitude regions and their low diversity persists during the whole Llandovery. This suggests that blastozoans slowly overcame the effects of the Hirnantian glaciation. This situation is in marked contrast with that described in other taxa (such as brachiopods, trilobites), which became rapidly cosmopolitan in continental seas (Harper and Rong, 2008). Finally, a strong correlation can be shown between the surviving to the extinction event and the wide

156

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

geographic distribution of some genera (e.g., Gomphocystites, Pleurocystites, Stephanocrinus). 4.4. Discussion The medium to high diversity observed for most geographic units during Cambrian Series 2–3 is possibly related to (1) the individualisation of several distinct palaeocontinents (following the split up of the Rodinia supercontinent), (2) their relative geographic proximity

(absence of large oceans), and (3) the Cambrian Substrate Revolution (the appearance of bioturbated substrates favouring the creation of new ecologic niches) (Bottjer et al., 2000; Dornbos, 2006). In the uppermost part of the Cambrian Series 3, the global drop in diversity and associated high turnover rate could be result of the habitat partitioning (and a low sea level) and to unstable geo-environmental conditions: enhanced continental drift, increase of orogeny index, deep changes in seawater chemistry (Fig. 8), and the last effects of the Cambrian Substrate Revolution. During the Furongian the advent of

Fig. 9. Comparison between the ﬂuctuations of the normalised generic diversity and the turnover rate, and some geological parameters through early Palaeozoic. Orogeny index data are from Khain and Seslavinsky (1994); platform ﬂooding data, from Algeo and Seslavinsky (1995); sea level variations data, from Haq and Schutter (2008); pCO2, from Goddéris et al. (2001); mobile mean 87Sr/86Sr ratio and δ13C isotopes, from Veizer et al. (1999); possible global climate, from Boucot and Gray (2001); carbon isotope excursions, from Kump and Arthur (1999); global mean temperature is extrapolated from Goddéris et al. (2001); ocean chemistry is based on Stanley and Hardie (1999) and Kiessling et al. (2008). G: greenhouse period, I: icehouse period. A: aragonite sea, C: calcite sea.

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

the “calcite sea” led to the formation of abundant and widespread carbonate hardgrounds (Stanley and Hardie, 1999). Its combination with a major global sea level rise and an increased continental drift led to the creation of new available ecological habitats and more favourable environmental conditions in shallow epicontinental seas. It is possible that the enhancement of both seaﬂoor production and orogeny may have contributed to the establishment of global greenhouse climatic conditions, and thus to a more intense weathering leading to an increase in nutrient supply into seawater (Fig. 9 and references herein). Moreover, the ﬂooding of large continental areas provides variable, nutrient-rich habitats by creating additional upwelling zones. All these events possibly triggered the establishment of more favourable conditions for the diversiﬁcation of benthic macrofauna and the onset on the Great Ordovician Biodiversiﬁcation Event (Wilson et al., 1992; Sprinkle and Guensburg, 1995; Guensburg and Sprinkle, 2001; Rozhnov, 2001; Palmer and Wilson, 2004; Sprinkle and Guensburg, 2004). Similar processes were also suggested for the Mesozoic diversiﬁcation of marine phytoplankton (Katz et al., 2004). The intensive colonisation of numerous geographic units of the widespread peri-Gondwanan margins from the Early Ordovician until the earliest Late Ordovician was possibly triggered by (1) the important sea level rise (creation of new habitats; Fig. 9), (2) the global warming (warmer temperature favours oxygen dissolution in seawater; Paul, 1976), and (3) less global siliciclastic and fresh water supply (low strontium ratio, low orogeny index; Fig. 9). However high provincialism could be explained by the variety of local environmental conditions during the Early–Middle Ordovician (Ocloyic and Carapronensis orogenies in west Gondwana; Grampian Orogeny in west Avalonia; Restrepo-Pace et al., 1997; Thomas and Astini, 2007). On the other hand, Laurentian and Baltic provinces were characterised by warm water in carbonate platforms, whereas the northwestern and eastern peri-Gondwanan regions were characterised by temperate water in siliciclastic platforms (Barnes, 1999; Cocks and Torsvik, 2002). These two distinct kinds of environments could explain why their faunal contents were clearly distinct (see above; Fig. 6), and why faunal exchanges existed only within these two regions (e.g., from the Mediterranean province to the Asian one). Finally, the dramatic faunal changes observed in Avalonia are related to its changing palaeogeography through the Ordovician: this region rifted away from the northwestern margin of Gondwana (late Early Ordovician), and drifted regularly towards Baltica (Middle to Late Ordovician), before ﬁnally colliding with it (latest Ordovician). Blastozoan faunas from Avalonia reﬂect this evolution, with a transition from typical periGondwanan assemblages (Cambrian to earliest Ordovician) to mixed faunas (Middle Ordovician), and ﬁnally, assemblages similar to those of Baltica (Late Ordovician). Moreover, in Middle to Late Ordovician times, Middle Eastern and Asian peri-Gondwanan terranes (e.g., Sibumasu, South China) drifted northwards to reach intermediate to low latitudes, and developed carbonate platforms with temperate to warm water (Cocks and Torsvik, 2002). Faunal similarity analyses suggest that warm wateradapted blastozoans migrated from these regions to the high latitude peri-Gondwanan ones during the Late Ordovician (Lefebvre, 2007a). The late Katian short expansion of more temperate environments and faunas to northwestern peri-Gondwanan regions (the so-called “Boda event”; Fortey and Cocks, 2005) attracted warmer water faunas from lower latitude regions (but see Cherns and Wheeley, 2007). As a consequence, in late Katian times, cool-adapted blastozoan assemblages of high latitude peri-Gondwanan regions either disappeared or were restricted to deeper water settings (Lefebvre, 2007a). In Katian times, the Taconic Orogeny in the eastern Laurentian margin reaches its paroxysmal phase (orogeny index rise). Local environment evolved from a carbonate to a mixed-siliciclastic platform with probably high turbidity (Kolata et al., 1996). The strong endemism of Laurentian fauna and limited exchanges with neighbouring geo-

157

graphic units (Avalonia and Baltica) both possibly result from this major palaeoenvironmental disturbance. Regional decrease in periGondwanan diversity could be related to the local climatic deterioration (temperature and humidity decrease; Ramstein et al., 2004) linked with the ice sheet development on Gondwana since the middle Katian (Le Héron and Craig, 2008). The Hirnantian decline in diversity is observed in all regions. The glacio-eustatic decline in sea level (strong regression), the productivity increase, and the associated anoxia and temperature drop caused by the North African glaciation have been widely invoked as the main triggers of the extinction event (Patzkowsky et al., 1997). The short duration of the total glaciation and its localisation on Gondwana could explain why faunas survived in intermediate to low latitude regions, where environmental conditions remained relatively suitable for the survivorship of the marine invertebrate fauna, including blastozoans (cool to warm water, carbonate platforms). In the northwestern peri-Gondwanan regions, the prolonged anoxia could have prevented the survival of blastozoans, and their rapid (re) colonisation of the area. Modelling and geologic data interpretations both suggest that environmental recovery is relatively rapid after the glaciation event (Ramstein et al., 2004), in spite of the persistence of a thick ice sheet on west Gondwana during the Rhuddanian (Diaz-Martinez and Grahn, 2007), and widespread anoxia during the Llandovery (Veizer et al., 1999). These environmental disturbances did possibly slow down the blastozoan recovery in terms of both diversity and geographic distribution. 5. Conclusion Diversity and evolution dynamics of blastozoans were partly related to palaeogeographic and palaeoenvironmental changes through the early Palaeozoic. However, the two peaks in diversity and their following extinctions are apparently driven by different factors. In Cambrian times, the ﬁrst peak in diversity is associated with a high turnover. The high endemicity is conﬁrmed by the small faunal correspondence between most of geographic units, except between those located in the northwestern peri-Gondwanan area. This could be linked with environmental instability (e.g., Cambrian substrate revolution, establishment of calcite-seas) and the relatively low sea level (accentuating habitat partitioning). In most regions (with the exception of northwestern Gondwanan epicontinental seas), blastozoan taxa were restricted to the geographic unit of their origination. Diversity increase is explained by a higher origination rate. High turnover may result from local environmental crises causing the extinction of genera restricted to a single geographic area (i.e. the majority of blastozoan taxa reported from Cambrian Series 2–3). These blastozoans are still dominated by genera with primitive morphology (CEF) inhabiting mostly Neoproterozoic-like ﬁrm substrates, but also included a signiﬁcant number of taxa able to colonise Phanerozoic-type bioturbated substrates. In contrast, Furongian blastozoans are mainly composed of derived genera (PEF) with a true stem for attachment on hard substrates (mostly in Laurentia) or for locomotion on soft substrates (mostly in peri-Gondwana and Baltica). In spite of a biased Furongian fossil (and sedimentary) record, the transition between the two faunas seems to be real. It reﬂects a dramatic change in the composition and ecology of blastozoan assemblages, with (1) a ﬁrst diversiﬁcation event correlated to the progressive colonisation of bioturbated substrates, along with the regular decline of taxa adapted to Neoproterozoic-like substrates (Cambrian series 2–3), and (2) a second diversiﬁcation event associated with the colonisation of carbonate hardgrounds in low palaeolatitude (Furongian) (Sumrall et al., 1997). Since the Furongian and through the whole Ordovician, both climatic and environmental conditions (e.g., sea level rise, greenhouse climate, plate movements, increase of nutrient supply) led to the

158

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

onset of new habitats with more stable and homogeneous environments over large regions: carbonate or mixed environments in low to intermediate latitude areas vs. siliciclastic environments in high latitude regions. Statistical and qualitative analyses show a clear faunal distinction between these two pools of regions (high provincialism). Thus, the latitudinal factor often invoked to explain ﬁrst-order diversity patterns impacts the blastozoan geographic distribution and its diversity (Krug et al., 2009). However, increasing migrations between the regions within each pool or between them may be also caused by other factors, such as sea level rise, oceanic circulation, and/or continental drift of peri-Gondwanan terranes towards the equator. The Laurentian province presents a different dynamic, with medium turnover rates and limited faunal exchanges with the other regions. This could result from its isolated geographic position. Ordovician genera have low turnover rates and a relatively low endemicity. Increasing faunal afﬁnities between all regions conﬁrm the growing complexity of faunal exchanges and the cosmopolitan status of many Late Ordovician genera, characterised by a high geographic dynamism. Thus, the Ordovician global rise in generic diversity may result from the resistance of blastozoan taxa to local environmental crisis, because of their wide geographic range and ability to migrate, rather than the low turnover rate (except in Laurentia). However, blastozoans were dramatically affected in both their diversity and geographic distribution by the Hirnantian glaciation. In early Silurian times, a low diversity was maintained based on surviving genera located in geographic units colonised before the crisis. The persistence of widespread inhospitable environmental conditions (anoxia) during the whole Llandovery may have altered the dynamism of blastozoans. These results suggest that the two diversiﬁcation events evidenced for blastozoans in early Palaeozoic times concerned two distinct evolutionary faunas (CEF vs. PEF). They were driven by distinct processes, even if the Ordovician diversiﬁcation had indeed started during the middle Furongian with a high turnover event. Differences in processes might be related to the evolution of global environmental/climatic conditions coupled with the position of continental masses. Acknowledgments The authors are grateful to the editors of the volume for their invitation to produce this paper, and to M. Reich (Geoscience Centre of the University of Göttingen) and J. Sprinkle (University of Texas) for their helpful comments. EN would like to thank J. Alroy and the team of the Paleodatabase, for the intensive courses provided in 2005. This is a contribution to the team “Forme, Evolution, et Diversité” of the UMR 5561 Biogéosciences (EN), to the team ‘Paléoenvironnement’ of the UMR 5563 LMTG (EN), and to the team “Primitive Life” of the UMR 5125 PEPS (BL). This paper is also a contribution to the IGCP project 503 “Ordovician Palaeogeography and Palaeoclimate.” Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.palaeo.2010.01.005. References Aceñolaza, G.F., Gutiérrez-Marco, J.C., 2002. Ordovician Echinoderms of Argentina. Ser. Correl. Geol. 16, 121–130. Adrain, J.M., Westrop, S.R., Chatterton, B.D.E., Ramsköld, L., 2000. Silurian trilobite alpha diversity and the end-Ordovician mass extinction. Paleobiology 26, 625–646. Adrain, J.M., Edgecombe, G.D., Fortey, R.A., Hammer, Ø., Laurie, J., McCormick, T., Owen, A.W., Waisfeld, B.G., Webby, B.D., Westrop, S.R., Zhou, Z., 2004. Trilobites. In: Webby, B.D., Paris, F., Droser, M.L., Percival, I.G. (Eds.), The Great Ordovician Biodiversiﬁcation Event. Columbia University Press, New York, pp. 231–254. Chapter 24.

Algeo, T.J., Seslavinsky, K.B., 1995. The Paleozoic world: continental ﬂooding, hypsometry, and sea level. Am. J. Sci. 295, 787–822. Alroy, J., 2000. New methods for quantifying macroevolutionary patterns and process. Paleobiology 26, 707–733. Álvaro, J.J., Ferretti, A., Gonzalez-Gomez, C., Serpagli, E., Tortello, M.F., Vecoli, M., Vizcaïno, D., 2007. A review of the Late Cambrian (Furongian) palaeogeography in the western Mediterranean region, NW Gondwana. Earth-Sci. Rev. 85, 47–81. Anstey, R.L., Pachut, J.F., Tuckey, M.E., 2003. Patterns of bryozoan endemism through the Ordovician–Silurian transition. Paleobiology 29, 305–328. Ausich, W.I., 1998. Early phylogeny and subclass division of the Crinoidea (phylum Echinodermata). J. Paleontol. 72, 499–510. Barnes, C.R., 1999. Paleoceanography and paleoclimatology: an Earth system perspective. Chem. Geol. 161, 17–35. Bergström, S.M., Xu, C., Gutiérrez-Marco, J.C., Dronov, A., 2009. The new chronostratigraphic classiﬁcation of the Ordovician system and its major regional series and stages and to δ13C chemostratigraphy. Lethaia 42, 97–107. Bottjer, D.J., Hagadorn, J.W., Dornbos, S.Q., 2000. The Cambrian substrate revolution. GSA Today 10, 9–12. Boucot, A.J., Gray, J., 2001. A critique of Phanerozoic climatic models involving changes in the CO2 content of the atmosphere. Earth-Sci. Rev. 56, 1–159. Broadhead, T.W., 1984. Macurdablastus a Middle Ordovician blastoid from the Southern Appalachians. Univ. Kans. Paleontol. Contrib. 110, 1–9. Carrera, M.G., Rigby, J.K., 1999. Biogeography of Ordovician sponges. J. Paleontol. 73, 26–37. Cherns, L., Wheeley, J.R., 2007. A pre-Hirnantian (Late Ordovician) interval of global cooling — the Boda event re-assessed. Palaeogeogr. Palaeoclimatol. Palaeoecol. 251, 449–460. Clarke, K.R., 1993. Non-parametric multivariate analyses of changes in community structure. Aust. J. Ecol. 18, 117–143. Clausen, S., 2004. New Early Cambrian eocrinoids from the Iberian Chains (NE Spain) and their role in nonreefal benthic communities. Eclogae Geol. Helv. 97, 371–379. Cocks, L.R.M., Fortey, R.A., 1988. Lower Palaeozoic facies and faunas around Gondwana. In: Audley-Charles, M.G., Hallam, A. (Eds.), Gondwana and Tethys: Geological Society Special Publication, London, pp. 183–200. Cocks, L.R.M., Torsvik, T.H., 2002. Earth geography from 500 to 400 million years ago: a faunal and palaeomagnetic review. J. Geol. Soc. Lond. 159, 631–644. Cocks, L.R.M., Torsvik, T.H., 2005. Baltica from the late Precambrian to mid-Palaeozoic times: the gain and loss of a terrane's identity. Earth-Sci. Rev. 72, 39–66. Cocks, L.R.M., Torsvik, T.H., 2007. Siberia, the wandering northern terrane, and its changing geography through the Palaeozoic. Earth-Sci. Rev. 82, 29–74. Cocks, L.R.M., McKerrow, W.S., Van Staal, C.R., 1997. The margins of Avalonia. Geol. Mag. 134, 627–636. Cocks, L.R.M., Fortey, R.A., Lee, C.P., 2005. A review of Lower and Middle Palaeozoic biostratigraphy in west peninsular Malaysia and southern Thailand in its context within the Sibumasu Terrane. J. Asian Earth Sci. 24, 703–717. Connolly, S.R., Miller, A.I., 2001. Global Ordovician faunal transitions in the marine benthos: proximate causes. Paleobiology 27, 779–795. Cooper, A., 2004. Measures of biodiversity. In: Webby, B.D., Paris, F., Droser, M.L., Percival, I.G. (Eds.), The Great Ordovician Biodiversiﬁcation Event. Columbia University Press, New York, pp. 52–57. Cope, J.C.W., Babin, C., 1999. Diversiﬁcation of bivalves in the Ordovician. Geobios 32, 175–185. Cricks, R.E., 1990. Cambrian–Devonian biogeography of nautiloid cephalopods. In: McKerrow, W.S., Scotese, C.R. (Eds.), Palaeozoic Palaeogeography and Biogeography: Geological Society, London, vol. 12, pp. 147–162. David, B., Lefebvre, B., Mooi, R., Parsley, R.L., 2000. Are homalozoans echinoderms? An answer from the extraxial–axial theory. Paleobiology 26, 529–555. Diaz-Martinez, E., Grahn, Y., 2007. Early Silurian glaciation along the western margin of Gondwana (Peru, Bolivia, and Northern Argentina): Palaeogeographic and geodynamic setting. Palaeogeogr. Palaeoclim. Palaeoecol. 245, 62–81. Dornbos, S.Q., 2006. Evolutionary palaeoecology of early epifaunal echinoderms: response to increasing bioturbation levels during the Cambrian radiation. Palaeogeogr. Palaeoclimatol. Palaeoecol. 237, 225–239. Droser, M.L., Finnegan, S., 2003. The Ordovician radiation: a follow-up to the Cambrian explosion? Integr. Comp. Biol. 43, 178–184. Foote, M., 1992. Paleozoic record of morphological diversity in blastozoan echinoderms. Proc. Natl. Acad. Sci. U. S. A. 89, 7325–7329. Foote, M., 2000. Origination and extinction components of taxonomic diversity: Paleozoic and post-Paleozoic dynamics. Paleobiology 26, 578–605. Fortey, R.A., Cocks, L.R.M., 2003. Palaeontological evidence bearing on global Ordovician–Silurian continental reconstructions. Earth-Sci. Rev. 61, 245–307. Fortey, R.A., Cocks, L.R.M., 2005. Late Ordovician global warming — the Boda event. Geology 33, 405–408. Frest, T.J., 2005. Ampheristocystis, a Silurian (Lower Wenlock) eocrinoid (Echinodermata, Blastozoa) from Indiana. Geol. J. 40, 301–329. Geyer, G., Shergold, J.H., 2000. The quest for internationally recognized divisions of Cambrian time. Episodes 23, 188–195. Goddéris, Y., Francois, L.M., Veizer, J., 2001. The early Paleozoic carbon cycle. Earth Planet. Sci. Lett. 190, 181–196. Grahn, Y., 2006. Ordovician and Silurian chitinozoan biozones of western Gondwana. Geol. Mag. 143, 509–529. Guensburg, T.E., Sprinkle, J., 2001. Ecologic radiation of Cambro–Ordovician echinoderms. In: Zhuravlev, A.Y., Riding, R. (Eds.), The Ecology of the Cambrian Radiation, Colombia, pp. 429–444. Gutiérrez-Marco, J.C., Rábano, I., Sarmiento, G.N., Aceñolaza, G., San José, M.A., Pieren, A.P., Couto, H.M., Picarra, J.M., 1999. Faunal dynamics between Iberia and Bohemia during

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160 the Oretanian and Drobotivian (late Middle–earliest Upper Ordovician), and biogeographic relations with Avalonia and Baltica. Acta Univ. Carol., Geol. 43, 487–491. Haq, B.U., Schutter, S.R., 2008. A chronology of Paleozoic sea-level changes. Science 322, 64–69. Harper, D.A.T., 2006. The Ordovician biodiversiﬁcation: setting an agenda for marine life. Palaeogeogr. Palaeoclimatol. Palaeoecol. 232, 148–166. Harper, D.A.T., Rong, J.Y., 2008. Completeness of the Hirnantian brachiopod record: spatial heterogeneity through the end Ordovician extinction event. Lethaia 41, 195–197. Harper, D.A.T., McNiocaill, C.M., Williams, S.H., 1996. The palaeogeography of early Ordovician Iapetus terranes: an integration of faunal and palaeomagnetic constraints. Palaeogeogr. Palaeoclimatol. Palaeoecol. 121, 297–312. Havlíček, V., Vanek, J., Fatka, O., 1994. Perunica microcontinent in the Ordovician (its position within the Mediterranean Province, series division, benthic and pelagic associations). Sb. Geol. Vĕd, Geol. 46, 23–56. Jefferies, R.P.S., 1990. The solute Dendrocystoides scoticus from the Upper Ordovician of Scotland and the ancestry of chordates and echinoderms. Palaeontology 33, 631–679. Katz, M.E., Finkel, Z.V., Grzebyk, D., Knoll, A.H., Falkowski, P.G., 2004. Evolutionary trajectories and biogeochemical impacts of marine eukaryotic phytoplankton. Annu. Rev. Ecol. Evol. 35, 523–556. Khain, V.E., Seslavinsky, K.B., 1994. Global rhythms in the Phanerozoic endogenic activity of the Earth. Stratigr. Geol. Correl. 2, 40–63. Kiessling, W., Aberhan, M., Vilier, L., 2008. Phanerozoic trends in skeletal mineralogy driven by mass extinctions. Nat. Geosci. 1, 527–530. Kolata, D.R., Huff, W.D., Bergström, S.M., 1996. Ordovician K-bentonites of eastern North America. GSA Spec. Pap. 313, 1–84. Krs, M., Pruner, P., Man, O., 2001. Tectonic and paleogeographic interpretation of the paleomagnetism of Variscan and pre-Variscan formations of the Bohemian Massif, with special reference to the Barrandian terrane. Tectonophysics 332, 93–114. Krug, A.Z., Patzkowsky, M.E., 2007. Geographic variation in turnover and recovery from the Late Ordovician mass extinction. Paleobiology 33, 435–454. Krug, A.Z., Jablonski, D., Valentine, J.W., Roy, K., 2009. Generation of Earth's ﬁrst-order biodiversity pattern. Astrobiology 9, 113–124. Kump, L.R., Arthur, M.A., 1999. Interpreting carbon-isotope excursions: carbonates and organic matter. Chem. Geol. 161, 181–198. Le Héron, D.P., Craig, J., 2008. First order reconstructions of a Late Ordovician Saharan ice sheet. J. Geol. Soc. Lond. 165, 19–29. Lefebvre, B., 2007a. Changes in Late Ordovician (pre-Hirnantian) echinoderm assemblages: palaeoenvironmental and palaeobiogeographical implications. Acta Palaeontol. Sin. 46, 241–248. Lefebvre, B., 2007b. Early Palaeozoic palaeobiogeography and palaeoecology of stylophoran echinoderms. Palaeogeogr. Palaeoclimatol. Palaeoecol. 245, 156–199. Lefebvre, B., Fatka, O., 2003. Palaeogeographical and palaeoecological aspects of the Cambro–Ordovician radiation of echinoderms in Gondwanan Africa and periGondwanan Europe. Palaeogeogr. Palaeoclimatol. Palaeoecol. 195, 73–97. Lefebvre, B., Ghobadipour, M., Nardin, E., 2005. Ordovician echinoderms from the Tabas and Damghan regions, Iran: palaeobiogeographical implications. Bull. Soc. Géol. Fr. 176, 231–242. Lefebvre, B., Eble, G., Navarro, N., David, B., 2006. Diversiﬁcation of atypical Paleozoic echinoderms: a quantitative survey of patterns of stylophoran disparity, diversity, and geography. Paleobiology 32, 483–510. Legendre, L., 1998. Numerical Ecology. Elsevier. 321 pp. Lieberman, B.S., 2003. Taking the pulse of the Cambrian radiation. Integr. Comp. Biol. 43, 229–237. Manly, B.F.J., 1997. Randomization, Bootstrap and Monte Carlo Methods in Biology. Chapman and Hall, New York. 399 pp. McGowan, A.J., Smith, A.B., 2007. Are global Phanerozoic marine diversity curves truly global? A study of the relationship between regional rock records and global Phanerozoic marine diversity. Paleobiology 34, 80–103. Miller, A.I., 1997. Comparative diversiﬁcation dynamics among palaeocontinents during the Ordovician radiation. Geobios 20, 397–406. Munnecke, A., Servais, T., 2007. What caused the Ordovician biodiversiﬁcation? Palaeogeogr. Palaeoclimatol. Palaeoecol. 245, 1–4. Nardin, E., Almazán-Vásquez, E., Buitrón-Sánchez, B.E., 2009a. First report of Gogia (Eocrinoidea, Echinodermata) from the Early–Middle Cambrian of Sonora (Mexico), with biostratigraphical and palaeoecological comments. Geobios 42, 233–242. Nardin, E., Lefebvre, B., David, B., Mooi, R., 2009b. La diversiﬁcation des échinodermes primitifs au Paléozoïque inférieur: l'exemple des blastozoaires. CR Palevol. 8, 179–188. Ogg, J.G., Ogg, G., Gradstein, F.M. (Eds.), 2008. The Concise Geologic Time Scale. Cambridge University Press, Cambridge. 184 pp. Owen, A.W., 2008. Ordovician and Silurian environments, biogeography and biodiversity change: an introduction. Lethaia 41, 97-97. Palmer, A.R., 2005. East-Gondwana/Laurentia trilobite connections — what do they tell us? Geosci. J. 9, 75–79. Palmer, T.J., Wilson, M.A., 2004. Calcite precipitation and dissolution of biogenic aragonite in shallow Ordovician calcite seas. Lethaia 37, 417–427. Parsley, R., 1997. The echinoderm classes Stylophora and Homoiostelea: non Calcichordata. In: Waters, J.A., Maples, C.G. (Eds.), Geobiology of Echinoderms: Paleont. Soc. Pap., vol. 3, pp. 225–248. Parsley, R., 1998. The cincta (Homostela) as Blastozoans. In: Carnevali, M.D.C., Bonasoro, F. (Eds.), Echinoderm Research 1998. Proceedings of the 5th Echinoderm Conference. Balkema, Milan, Italy, pp. 369–375. Patzkowsky, M.E., Slupik, L.M., Arthur, M.A., Pancost, R.D., Freeman, K.H., 1997. Late Middle Ordovician environmental change and extinction: Harbinger of the Late Ordovician or continuation of Cambrian patterns? Geology 25, 911–914.

159

Paul, C.R.C., 1976. Palaeogeography of primitive echinoderms in the Ordovician. In: Bassett, M.G. (Ed.), The Ordovician System: Proceedings of a Paleontological Association Symposium. University of Wales Press, Cardiff, pp. 553–574. Paul, C.R.C., 1988. The phylogeny of the cystoids. In: Paul, C.R.C., Smith, A.B. (Eds.), Echinoderm Phylogeny and Evolutionary Biology. Clarendon Press, Oxford, pp. 199–213. Paul, C.R.C., Smith, A.B., 1984. The early radiation and phylogeny of echinoderms. Biol. Rev. 59, 443–481. Peters, S.E., Ausich, W.I., 2008. A sampling-adjusted macroevolutionary history for Ordovician–Early Silurian crinoids. Paleobiology 34, 104–116. Powel, M.G., 2007. Geographic range and genus longevity of late Paleozoic brachiopods. Paleobiology 33, 530–546. R Development Core Team, 2008. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. Ramstein, G., Donnadieu, Y., Goddéris, Y., 2004. Les glaciations du Protérozoïque. CR Geosci. 336, 639–646. Raup, D.M., 1975. Taxonomic diversity estimation using rarefaction. Paleobiology 1, 333–342. Raup, D.M., Crick, R.E., 1979. Measurement of faunal similarity in paleontology. J. Paleontol. 53, 1213–1227. Restrepo-Pace, P.A., Ruiz, J., Gehrels, G.E., Cosca, M., 1997. Geochronology and Nd isotopic data of Grenville-age rocks in the Colombian Andes: new constraints for Late Proterozoic–Early Paleozoic paleocontinental reconstructions of the Americas. Earth Planet. Sci. Lett. 150, 427–441. Rozhnov, S.V., 2001. Chapter 11: evolution of the hardground community. In: Zhuravlev, A.Y., Riding, R. (Eds.), The Ecology of the Cambrian Radiation, Colombia, pp. 238–253. Rozhnov, S.V., 2002. Morphogenesis and evolution of crinoids and other pelmatozoan echinoderms in the Early Paleozoic. Paleontol. J. 36, 523–674. Rozhnov, S.V., 2007. Changes in the Early Palaeozoic geography as a possible factor of echinoderm higher taxa formation: delayed larval development to cross the Iapetus Ocean. Palaeogeogr. Palaeoclimatol. Palaeoecol. 245, 306–316. Sepkoski, J.J., 1979. A kinetic model of Phanerozoic taxonomic diversity. II Early Phanerozoic families and multiple equilibria. Paleobiology 5, 222–251. Sepkoski, J.J., 1981. A factor analytic description of the Phanerozoic marine fossil record. Paleobiology 7, 36–53. Sepkoski, J.J., 1990. Evolutionary faunas. In: Briggs, D.E.G., Crowther, P.R. (Eds.), Palaeobiology, a Synthesis. Blackwell Science, London, pp. 37–41. Servais, T., Lehnert, O., Li, J., Mullins, G.L., Munnecke, A., Nützel, A., Vecoli, M., 2008. The Ordovician biodiversiﬁcation: revolution in the oceanic trophic chain. Lethaia 41, 99–109. Servais, T., Harper, D.A.T., Li, J., Munnecke, A., Owen, A.W., Sheehan, P.M., 2009. Understanding the Great Ordovician Biodiversiﬁcation Event (GOBE): inﬂuences of paleogeography, paleoclimate, or paleoecology? GSA Today 19, 4–10. Sheehan, P.M., 2001. History of marine biodiversity. Geol. J. 36, 231–249. Smith, A.B., 1990. Evolutionary diversiﬁcation of echinoderms during the early Palaeozoic. In: Taylor, P.D., Larwood, G.P. (Eds.), Major Evolutionary Radiations. The Systematics Association Special, vol. 42. Clarendon Press, Oxford, pp. 265–386. Smith, A.B., 2004. Echinoderm roots. Nature 430, 411–412. Smith, A.B., 2005. The pre-radial history of echinoderms. Geol. J. 40, 255–280. Sprinkle, J., 1973. Morphology and Evolution of Blastozoan Echinoderms. Museum Comparative Zoology Special Publication. 284 pp. Sprinkle, J., 1992. Radiation of Echinodermata. In: Signor, P.W., Lipps, J. (Eds.), Origin and Early Evolution of the Metazoa, New York, pp. 375–398. Sprinkle, J., 1995. Do eocrinoids belong to the Cambrian or to the Paleozoic Evolutionary Fauna? In: Cooper, J.D., Droser, M.L., Finney, S.C. (Eds.), Ordovician Odyssey: Short Papers for the Seventh International Symposium on the Ordovician System. : Paciﬁc Section. Society of Economic Paleontologists and Mineralogists, pp. 397–400. Sprinkle, J., Guensburg, T.E., 1995. Origin of echinoderms in the Paleozoic evolutionary fauna: the role of substrates. Palaios 10, 437–453. Sprinkle, J., Guensburg, T.E., 2001. Growing a stalked echinoderm within the Extraxial– Axial Theory. In: Barker, F.K. (Ed.), Echinoderms 2000. Proceedings of the 10th International Echinoderm Conference, Dunedin, N.Z., 2000. Swets and Zeitlinger, Lisse, pp. 59–65. Sprinkle, J., Guensburg, T.E., 2003. Major expansion of echinoderms in the early Late Ordovician (Mohawkian, middle Caradoc) and its possible causes. Correl. Geol. 17, 327–332. Sprinkle, J., Guensburg, T.E., 2004. Chapter 26: Crinozoan, Blastozoan, Echinozoan, Asterozoan and Homalozoan echinoderms. In: Webby, B.D., Paris, F., Droser, M.L., Percival, I.G. (Eds.), The Great Ordovician Biodiversiﬁcation Event. Columbia University Press, New York, pp. 266–280. Stanley, S.M., Hardie, L.A., 1999. Hypercalciﬁcation: paleontology links plate tectonics and geochemistry to sedimentology. GSA Today 9, 1–7. Sumrall, C.D., Sprinkle, J., Guensburg, T.E., 1997. Systematics and paleoecology of Late Cambrian echinoderms from the western United States. J. Paleontol. 71, 1091–1109. Taylor, P.D., Ernst, A., 2004. Chapter 16: Bryozoans. In: Webby, B.D., Paris, F., Droser, M.L., Percival, I.G. (Eds.), The Great Ordovician Biodiversiﬁcation Event. Columbia University Press, New York, pp. 147–156. Thomas, W.A., Astini, R.A., 2007. Vestiges of an Ordovician west-vergent thin skinned Ocloyic thrust belt in the Argentine Precordillera, southern Central Andes. J. Struct. Geol. 29, 1369–1385. Tychsen, A., Harper, D.A.T., 2004. Ordovician–Silurian distribution of Orthida (Palaeozoic Brachiopoda) in the greater Iapetus Ocean region. Palaeontol. Electronica 7 15 p, 706KB. Ubaghs, G., 1998. Echinodermes nouveaux du Cambrien supérieur de la Montagne Noire (France méridionale). Geobios 31, 809–829.

160

E. Nardin, B. Lefebvre / Palaeogeography, Palaeoclimatology, Palaeoecology 294 (2010) 142–160

Veizer, J., Ala, D., Azmy, K., Bruckschen, P., Buhl, D., Bruhn, F., Carden, G.A.F., Diener, A., Ebneth, S., Goddéris, Y., Jasper, T., Korte, C., Pawellek, F., Podlaha, O.G., Strauss, H., 1999. 87Sr/86Sr, δ13C and δ18O evolution of Phanerozoic seawater. Chem. Geol. 161, 59–88. Vizcaïno, D., Álvaro, J.J., Lefebvre, B., 2001. The Lower Ordovician of the Southern Montagne Noire. Ann. Soc. Géol. Nord 8, 213–220. Webby, B.D., Cooper, A., Bergström, S.M., Paris, F., 2004a. Chapter 2: stratigraphic framework and times slices. In: Webby, B.D., Paris, F., Droser, M.L., Percival, I.G. (Eds.), The Great Ordovician Biodiversiﬁcation Event. Columbia University Press, New York, pp. 41–46. Webby, B.D., Droser, M.L., Paris, F., Percival, I.G. (Eds.), 2004b. The Great Ordovician Biodiversiﬁcation Event. Columbia University Press, New York, p. 484.

Wilson, M.A., Palmer, T.J., Guensburg, T.E., Finton, C.D., Kaufman, L.E., 1992. The development of an Early Ordovician hardground community in response to rapid sea-ﬂoor calcite precipitation. Lethaia 25, 19–33. Zamora, S., Smith, A.B., 2008. A new Middle Cambrian stem-group echinoderm from Spain: palaeobiological implications of a highly asymmetric cinctan. Acta Palaeontol. Pol. 53, 207–220. Zhao, Y.-L., Parlsey, R.L., Peng, J., 2007. Early Cambrian eocrinoids from Guizhou Province, South China. Palaeogeogr. Palaeoclimatol. Palaeoecol. 257, 317–327. Zhuravlev, A.Y., Naimark, E.B., 2005. Alpha, beta, or gamma: numerical view on the Early Cambrian world. Palaeogeogr. Palaeoclimatol. Palaeoecol. 220, 207–225. Zhuravlev, A.V., Riding, R. (Eds.), 2001. The Ecology of the Cambrian Radiation. Columbia University Press. 525 pp.

The relationship within and between the extrinsic and intrinsic systems ...

Thermal characterization of intrinsic and extrinsic InP ...

intrinsic and extrinsic motivators to study industrial ...

Image Saliency: From Intrinsic to Extrinsic Context - Research at Google

Image Saliency: From Intrinsic to Extrinsic Context - Semantic Scholar

Personality and Intrinsic Motivational Factors in ... - Dr Michal Kosinski

The Effects of Choice on Intrinsic Motivation and ...

Unravelling the pathophysiology of delirium: A focus on ...

Understand The Factors That Affect The Specification Of Materials And ...

The intrinsic Geometry of the CerebraÅ Cortex

Factors Affecting the Caffeine and Polyphenol Contents of Black and ...

The role of thiols, dithiols, nutritional factors and ... - Semantic Scholar

Specific Factors, Learning, and the Dynamics of Trade

Ergonomics Society of the Human Factors and Human ...

Shoemaker's Key, Extrinsic Dispositions, and Causal ...