Palaeogeography, Palaeoclimatology, Palaeoecology 248 (2007) 82–95 www.elsevier.com/locate/palaeo

The space-time relationship of taxonomic diversity and morphological disparity in the Middle ammonite radiation ⁎ Sébastien Moyne a, , Pascal Neige b

a Centre de Sédimentologie-Paléontologie and FRE CNRS 2761, Université de Provence, Bâtiment de Sciences Naturelles, Case 67, Place Victor Hugo, F-13 331 Marseille cedex 3, France b Centre des Sciences de la Terre and UMR CNRS 5561 Biogéosciences, Université de Bourgogne, 6 bd Gabriel, F-21000 Dijon, France Received 9 May 2006; received in revised form 17 October 2006; accepted 24 November 2006

Abstract

The Middle Jurassic ammonite radiation (from the late Aalenian to the end of the mid-Bathonian) is traced using combined analyses of morphological disparity and taxonomic diversity. The global signals of disparity and diversity are compared. These signals are then broken down by paleogeographical provinces to detect any heterogeneity in the radiation. An examination of the global signals reveals three biodiversity crises (discordances between signals) where morphological disparity grows while taxonomic diversity declines. The subdivision of the signals indicates the radiation was heterogeneous between provinces: the global signal is an aggregate of signals from each province. The three biological crises have different paleogeographical signatures: the first is visible in a few provinces only while the other two are visible in most provinces. First-order crises can be distinguished from second-order ones by studying the number of paleogeographical provinces affected. © 2006 Elsevier B.V. All rights reserved.

Keywords: Paleogeography; Morphological disparity; Taxonomic diversity; Ammonites; Jurassic

1. Introduction groups of taxa with equal diversity may display very different morphological variability. Therefore, taxonom- Taxonomic diversity is generally used to characterize ic diversity analyses have been supplemented exten- the evolutionary history of clades and particularly sively over the last decade by morphological disparity extinctions and radiations. It may be used both to detect analyses for various taxa (e.g. Wills et al., 1994; Foote, major biological crises (Sepkoski, 1981; Sepkoski and 1999; Ciampaglio et al., 2001; Navarro et al., 2005), Hulver, 1985) and to help model the diversification of particularly because the combined use of taxonomic different biological groups (Benton, 2001). Although diversity and morphological disparity signals can only interesting in itself, taxonomic diversity has proved a lead to a better understanding of macroevolutionary poor proxy for morphological variability (see Foote processes (Wills, 2001). 1993a; Neige et al., 1997; Neige 2003). Indeed, two Interestingly, the geographical aspects of diversity/ disparity relationships have come in for a little study. ⁎ Corresponding author. This is probably because studies have concentrated on E-mail address: [email protected] (P. Neige). analysing large macroevolutionary patterns covering

0031-0182/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.palaeo.2006.11.011 S. Moyne, P. Neige / Palaeogeography, Palaeoclimatology, Palaeoecology 248 (2007) 82–95 83 long time spans and under the assumption that the long-standing use of morphometrics for ammonites, phenomena under study (e.g. radiations, crises) were which ensures a robust quantification of shape and fundamentally worldwide. There is a presupposition, disparity (see Raup, 1967; Dommergues et al., 1996; then, that observing diversity/disparity relationships Neige et al., 2001; Korn and Klug, 2003; Navarro et al., independently for several paleogeographical areas will 2005). The second is the wealth of knowledge paleontol- add nothing to our understanding of the radiations and ogists have of these creatures because they are abundant crises involved. This may be true, but only if large within Jurassic marine outcrops and also because they macroevolutionary patterns are independent of any have been extensively studied and figured for biostrati- environmental context. Most studies of large-scale graphical purposes. biological radiations generally present just one global diversity (or disparity) signal over time for the 2. The radiation in a global context considered taxa. They fail to take account of paleogeo- graphical homogeneity or heterogeneity. Some studies, Middle Jurassic ammonites are found worldwide in though, have shown that a radiation may be paleogeo- the three great oceanic realms (Tethyan, Pacific and graphically heterogeneous. Miller (1997a,b) reported Boreal). To trace the geographical variations in diversity that diversification patterns during the great and disparity, 16 paleogeographical provinces have been radiation varied between provinces. A study of bivalve individualized (Fig. 1). This subdivision is derived from radiation for the same period (Babin, 1993) identified that proposed by Dommergues et al. (2001) for the the geographical shifting of the diversification as a Lower Jurassic and modified for the needs of the Middle major process. These observations indicate that a Jurassic (Moyne et al., 2004). Present-day geographic radiation cannot be described from a global signal units belonging to the paleogeographic provinces are alone. A global signal is an aggregate measure of described in Table 1. These provinces are defined by information compiled from signals from different their fauna and geology (see Dommergues et al., 2001). paleogeographical provinces. It has to be dissected to The studied period (late Aalenian to mid-Bathonian) be properly interpreted, and in particular to determine is subdivided into 15 biozones in the Tethyan Realm whether or not a radiation occurs uniformly worldwide. (Contini et al., 1997; Rioult et al., 1997; Mangold and This pilot study computes and compares signals of Rioult, 1997). This zonation is available for the Western taxonomic diversity and morphological disparity. These Tethys only and it has been correlated with parallel signals are divided between provinces to determine scales established in other countries (Krymholts et al., whether paleogeography needs be taken into account if a 1988; Hillebrandt et al., 1992; Callomon, 1994; Enay biological radiation is to be understood in a global and Mangold, 1994; Hillebrandt, 2001). All results context. presented in this study refer to the 15 biozones of the The data used are for the Middle Jurassic ammonite scale proposed for the Tethys. This period covers a total radiation, which was marked by substantial faunal span of 8 Ma (Gradstein et al., 2004, but which may be turnover of the Ammonitina sub-order near the Aale- extended to 15 Ma in reference to Odin and Odin, 1990). nian/ boundary. A single super-family went forth Unfortunately, there are no radiometric data for this and multiplied giving rise to all the groups that dominated period for accurate time calibrations. The results are the Late Jurassic (House, 1985; Page, 1996). The exact therefore presented here with a relative biostratigraphic relationships between these groups are hard to understand scale. and many phylogenies have been proposed (see Moyne and Neige, 2004 for a recent review). This key period has 2.1. Fluctuation of taxonomic diversity over time already been studied, but only for its taxonomic diversity (O'Dogherty et al., 2000; Sandoval et al., 2001, 2002; The faunal turnover at the Aalenian/Bajocian Moyne et al., 2004). Those studies show that taxonomic boundary is marked by a large increase in the number diversity was at its lowest in the Aalenian and then surged of Ammonitina species: some 80 species for each during the Bajocian. In this paper, we study all groups of biozone of the Aalenian versus 179 species in the the Ammonitina from the late Aalenian to focus on faunal Laeviuscula Zone (Fig. 2). Species diversity peaks at the turnover. We then trace the radiation until the end of the end of the early Bajocian and then remains high until the mid-Bathonian. All species belonging to the Ammonitina onset of the Bathonian (Zigzag Zone). Thereafter sub-order from the late Aalenian (Bradfordensis Zone) to species diversity plummets, recovering slightly at the the end of the mid-Bathonian (Bremeri Zone) are covered. start of the mid-Bathonian. It remains at the same level Two factors facilitate the study of this radiation. One is the until the end of the mid-Bathonian, at which point, 84 S. Moyne, P. Neige / Palaeogeography, Palaeoclimatology, Palaeoecology 248 (2007) 82–95

Fig. 1. Location of the 16 paleogeographical provinces (A–P) from the end of the Aalenian to the Bathonian (Middle Jurassic). Letters refer to provinces named and defined in Table 1 (Paleogeographical reconstruction compiled from data of Owen, 1983; Scotese, 1991, 1997; Thierry et al., 2000; Vrielynck and Bouysse, 2001). diversity dwindles to the level recorded during the (Fig. 2). An initial increase in diversity at the end of the Aalenian. Aalenian stabilizes in the Lower Bajocian. Genus diversity Genus diversity follows a roughly similar pattern of peaks in the late Bajocian before falling off rapidly during fluctuation to species diversity over the same period of time the early Bathonian and remaining low thereafter.

Table 1 Name and description of the 16 paleogeographical units used in the analysis (see main text for further explanations) No Name Description A Northern Mediterranean Threshold Austroalpine outcrops of Italy, Switzerland, Austria and Hungary (e.g. Bakony Mountains) B Central Mediterranean Threshold Algarve, the BeticRange, the Appennines, the Hellenids, a Sicilian section (Calabria) and Serbia C Southern Mediterranean Threshold Morroco to Tunisia D Southern Tethyan border Libya, Egypt, Southern Turkey, Southern Iran, Saudi Arabia, Madagascar and India E South America and Antarctica Chile, Peru, Argentina and the Antarctic peninsula F Central America Mexico G North America Western Interior of the United States of America, Western Canada and Southern Alaska H Japan and Eastern Russia Honshu, Shikoku, Kyushu and South-eastern Russia I Tibet and South-East Asia Tadjikistan, Tibet and Thailand J North-eastern Tethyan border Pontic range, Crimea, Caucasus, Northern Iran, Turkmenistan, Ouzbekistan, Kazakhstan K Northern Tethyan border Romania, Hungary p.p. (Villany and Mecsek), Balkans (Bulgaria and Eastern Serbia) and Carpathian mountains L South of the North-Western Southern Germany, Jura, Burgundy, Paris Basin, Causses, Western Alps, Ibersian Basin Tethyan platform M Lusitanian Basin Lusitanian Basin N North of the North-Western British Isles, Northern Germany, Scandinavia and Poland (platform) Tethyan platform O Boreal Siberia, Eastern Greenland, Northern Canada and Northern Alaska P Australia Western Australia, New Zeland, New Guinea (Irian Jaya) and Sula Islands S. Moyne, P. Neige / Palaeogeography, Palaeoclimatology, Palaeoecology 248 (2007) 82–95 85

most of the morphological variability is explained by the first axes (see Harper and Owen, 1999). Morphospaces can be obtained by combining the first two axes. Morphological disparity may be also evaluated arith- metically from indices. There are many indices available for quantifying morphological disparity (see Ciampaglio et al., 2001; Navarro, 2003). A single index is used to describe morphological disparity: the sum of variances. This index corresponds to the sum of univariate variances from all axes computed by correspondence analysis (Foote, 1991). This index is relatively insensi- tive to sample size but it can be sensitive to taxonomic practice. To obtain an unbiased estimate of variance, a bootstrap procedure is used (Efron, 1979; Efron and Tibshinari, 1993). Data are resampled randomly with replacement (1000 replications) and a standard deviation is calculated reflecting an estimation of analytical error (Foote, 1993b; Eble, 2000). For our study, the sum of variances is computed with the first 10 axes, which represent more than 90% of total variance.

2.2. Fluctuation of morphological disparity through time The sum of variances shows that morphological disparity is lowest in the Bradfordensis Zone at the beginning of the period under study (Fig. 3). Thereafter disparity increases until the middle of the early Bajocian (Laeviuscula Zone). It remains high until the end of the Fig. 2. Variation in taxonomic diversity of Ammonitina over time. Bajocian and throughout the Bathonian. However, this These variations are represented at two taxonomic levels (species period of high disparity is marked by two minor crises diversity and genus diversity). Error bars estimated as ± √D, where D is the number of species or genera. (falls in disparity): one at the beginning of the late

2.2. Morphological disparity

2.1. Quantification of morphological disparity The ammonites in this study display wide ranges of morphologies. For this reason just one representative specimen per genus is used to describe the morpholog- ical disparity. This procedure is considered reliable at this scale of study (Eble, 1998, 2000; Foote, 1999). The morphology of these ammonites is evaluated using 18 characters. Some characters are continuous variables describing mainly the general shape of the shell. They include the three parameters established by Raup (1966, 1967) and two other shape indices. Two other con- tinuous variables are also used to quantify ribbing den- sity. The remaining characters are discrete and are qualitatively defined by having generally four or five states. These measurements are made at the end of the phragmocone so as to ensure they are homologous. Fig. 3. Variation of morphological disparity of Ammonitina over time. The data are treated by correspondence analysis. This The index used is the sum of variances. Error bar represents a standard method yields new axes from original characters and deviation calculated from 1000 bootstrap replicates. 86 S. Moyne, P. Neige / Palaeogeography, Palaeoclimatology, Palaeoecology 248 (2007) 82–95

Fig. 4. Fluctuation of diversity and disparity over time for each paleogeographical province. For each province, taxonomic diversity is representedby the shaded curve while morphological disparity is represented by the black curve. Time is represented by numbers at the bottom of the figure (01 for Bradfordensis Zone through to 15 for Bremeri Zone). S. Moyne, P. Neige / Palaeogeography, Palaeoclimatology, Palaeoecology 248 (2007) 82–95 87

Fig. 4 (continued ).

Bajocian (Niortense Zone) and one at the end of the Others indices, not reported here, were also computed studied period (Morrisi and Bremeri Zones). These two and yielded similar results for the global trend: a marked falls in disparity are slight and disparity does not rise in disparity in the first part of the period followed by collapse to the level observed at the end of the Aalenian. stabilization until the end of the mid-Bathonian. 88 S. Moyne, P. Neige / Palaeogeography, Palaeoclimatology, Palaeoecology 248 (2007) 82–95

Fig. 5. Spatial and temporal fluctuation of taxonomic diversity and morphological disparity represented on maps. Each map represents a transition between two biozones. Summary information on diversity/disparity relationships is given for each province.

2.3. Signal comparison and display of discordances both the diversity and disparity signals display similar increases. The late Bajocian is marked by high Comparison of the diversity curves with the diversity and disparity, despite small crises (well disparity indices reveals discordances between the marked species diversity crisis in the Garantiana two signals. Initially (late Aalenian and early Bajocian) Zone – well marked disparity crisis for the sum of S. Moyne, P. Neige / Palaeogeography, Palaeoclimatology, Palaeoecology 248 (2007) 82–95 89

Fig. 5 (continued ). variances in the Niortense Zone). In the Bathonian, 3. Impact of the paleogeographical pattern on diversity plunges to a low in the Progracilis Zone. This diversity and disparity signals dramatic fall is not visible in the disparity index. Indeed, in the Bathonian, disparity remains high before Previous studies have shown that the Bajocian declining slightly at the very end of the mid-Bathonian. radiation of the Ammonitina sub-order was not To summarize, diversity and disparity signals are pretty paleogeographically homogeneous (see Moyne et al., much concordant during the Bajocian but discordant 2004 for a taxonomic diversity study). The total during the Bathonian. distribution area of these ammonites was subdivided 90 S. Moyne, P. Neige / Palaeogeography, Palaeoclimatology, Palaeoecology 248 (2007) 82–95

The graphs display signal fluctuations over time (late Aalenian to mid-Bathonian). Obviously they are biased to varying extents by outcrop abundance (no signal for a given time may be simply an absence of outcrops), but discordance between the two signals may be seen as a genuine pattern based on well-documented data.

3.1.1. Province A (Northern Mediterranean Threshold) ThemarkeddevelopmentoftheHammatocerataceae and Hildocerataceae at the end of the Aalenian led to a surge in diversity and disparity alike. The extinction of the Hildocerataceae in the early Bajocian resulted in a fall in diversity that was not offset by the faunal turnover of the Hammatocerataceae (see Moyne and Neige, 2004). This period (early Bajocian) is marked by low numbers of species but with contrasted morphologies as reflected by the continued rise in the disparity signal. The late Bajocian saw a recovery of diversity coupled with a downturn in disparity. This fall in disparity reflects the presence of morphologically more homogeneous faunas. Finally, diversity fell during the early Bathonian while disparity remained at its late Bajocian level.

3.1.2. Province B (Central Mediterranean Threshold) Diversity and disparity signals in this province are relatively concordant throughout the period under study. Both rose gradually during the late Aalenian and early Bajocian. Thereafter disparity remained high until the mid-Bathonian while diversity plunged in the Garanti- ana and Aurigerus Zones.

3.1.3. Province C (Southern Mediterranean Threshold) Both diversity and disparity surged in the late Aalenian. Disparity peaked at the end of the Aalenian Fig. 6. Comparison of diversity and disparity over time at a worldwide (Concavum Zone) while diversity continued to rise scale. Biological crises (discordant disparity and diversity signals) are during the early Bajocian (peaking in the Laeviuscula highlighted. Zone). From the late Bajocian, both the diversity and into 16 paleogeographical provinces (the same pro- disparity curves exhibit similar fluctuations. vinces as described in Table 1) and taxonomic diversity signals computed for each province. Signals from 3.1.4. Province D (Southern Tethyan border) several provinces are similar to the signal computed From the end of the Aalenian to the end of the mid- for the global radiation but other provinces display very Bathonian, diversity and disparity curves follow similar different signals. The global signals of taxonomic trends although with different amplitudes. Diversity diversity and morphological disparity presented in the peaks in the Niortense Zone before plummeting in the first part of this study were broken down and compared late Bajocian and sliding further in the Bathonian. Over for the 16 paleogeographical provinces. the same period, disparity dips and then recovers at the end of the Bajocian. It trails off during the Bathonian. 3.1. Signal fluctuations over time in each province 3.1.5. Province E (South America and Antarctica) Fig. 4 shows a plot of the taxonomic diversity signal Diversity and disparity surge until the end of the early (species diversity) and the morphological disparity signal Bajocian. The late Bajocian sees a collapse of diversity (sum of variances) for each paleogeographical province. while disparity peaks in the Garantiana Zone. Thereafter S. Moyne, P. Neige / Palaeogeography, Palaeoclimatology, Palaeoecology 248 (2007) 82–95 91 diversity remains weak during the Bathonian while climbs steadily in the Bajocian and after a drop in the disparity slumps over this period. number of species at the onset of the mid-Bathonian, it increases again. By contrast, disparity peaks at the end of 3.1.6. Province F (Central America) the Aalenian when diversity is low. Thereafter disparity Ammonites are found in four biozones only (Propin- declines gradually until the end of the mid-Bathonian. quans to Garantiana Zones) in this province. Diversity is extremely low in the Propinquans Zone and then surges in 3.1.12. Province L (South of the North-Western Tethyan the following two biozones before plummeting in the Platform) Garantiana Zone. Disparity follows a different pattern, Diversity and disparity curves follow a similar rising in the early Bajocian, falling in the Niortense Zone pattern throughout the Bajocian. After an increase and recovering in the Garantiana Zone. during the early Bajocian, both signals fall at the onset of the late Bajocian and then level off until the end of the 3.1.7. Province G (North America) Bajocian. In the Bathonian, there is discordance bet- The diversity and disparity curves for this province ween the two signals with diversity falling while dis- exhibit a similar pattern to that for South America. Both parity remains stable. increase initially. Diversity peaks at the end of the early Bajocian. In the late Bajocian, diversity plunges while 3.1.13. Province M (Lusitanian Basin) disparity climbs. The Bathonian is marked by a down- The diversity and disparity signals for this province turn in disparity. are very peculiar. Diversity surges at the beginning of the Bajocian and this trend is matched by disparity. 3.1.8. Province H (Japan and Eastern Russia) Diversity then collapses until the onset of the late The diversity and disparity curves exhibit similar Bajocian. Disparity also plummets during the same patterns. After a crisis at the onset of the Bajocian, period even more steeply. Thereafter the two signals rise both recover during the early Bajocian. The dip in again until the early Bathonian where diversity collapses diversity in the mid-Bajocian (Humphriesianum Zone) again while disparity remains high. is not followed by disparity, which continues to rise in this biozone. Both diversity and disparity peak in the 3.1.14. Province N (North of the North-Western Tethyan late Bajocian. Platform) Signals for this province are similar to those observed 3.1.9. Province I (Tibet and South-East Asia) for the South of the North-Western Tethyan Platform The early Bajocian is marked by low diversity (a single (Province L). Diversity rises steadily until the end of the species for each biozone). Thus no disparity signal can be early Bajocian and remains high until the onset of the computed (disparity is ascribed a zero value when only early Bathonian despite a dip in the Garantiana Zone. one species occurs). Diversity and disparity increase in the Diversity then plunges and remains low through the late Bajocian. In the early Bathonian diversity dips while mid-Bathonian. Disparity increases until the end of the disparity continues to climb. The mid-Bathonian sees a early Bajocian and then remains high until the end of the surge in diversity while disparity remains unchanged. mid-Bathonian in spite of minor variations.

3.1.10. Province J (North-East Tethyan border) 3.1.15. Province O (Boreal) The onset of the Bajocian (Discites Zone) is Some morphologically well-diversified species are represented by a small number of morphologically dis- reported in the early Bajocian (Propinquans Zone). The parate species. The next biozone (Laeviuscula Zone) sees late Bajocian is marked by a surge in diversity while a slight dip in diversity combined with a plunge in disparity remains low. This period in this province is disparity. The end of the early Bajocian is marked by an marked by the presence of many endemic species increase in both signals. Thereafter diversity peaks during exhibiting little morphological variation. Diversity the late Bajocian while disparity dwindles. The Bathonian declines in the Bathonian while disparity rises because is marked by a plunge in diversity while disparity remains of incursions of scarce Tethyan faunas with different stable. morphologies.

3.1.11. Province K (Northern Tethyan border) 3.1.16. Province P (Australia) Diversity and disparity curves exhibit contrasting The early Bajocian is marked by great diversity trends throughout the period under study. Diversity associated with considerable disparity. Values of both 92 S. Moyne, P. Neige / Palaeogeography, Palaeoclimatology, Palaeoecology 248 (2007) 82–95 signals fall during the Bajocian and level off until the end signals. The crisis, visible from the global signal for of the early Bathonian. The mid-Bathonian sees a this time, is marked in most provinces and no decrease in these values, a jump in disparity in the conflicting signals are observed. Progracilis Zone notwithstanding. From the Garantiana Zone to the Zigzag Zone (Fig. 5, Maps 8–9), while global signals are largely concordant 3.2. Paleogeographical heterogeneity of signals (Figs. 2 and 3), many provinces exhibit either increasing diversity and decreasing disparity, or the reverse To illustrate diversity/disparity relationships over arrangement. Once again, large differences are found time and space on a worldwide scale, but using data for in the provincial patterns which are masked by the the separate provinces, we have mapped each transition global signals of diversity and disparity. between two biozones (Fig. 5). Relative variations in the The period from the Zigzag Zone to the Progracilis diversity and disparity curves are indicated for each Zone is marked by a strong global decline in diversity province for the time-span covered by each map (i.e. the (Fig. 2). During this time, global disparity increases transition from one biozone to the next): slightly (Fig. 3). These observations are reflected in Maps 10 and 11 of Fig. 5. Many provinces during this 1. When both curves change in the same direction (both time see an increase in disparity while diversity falls. increase or both decrease), a barred square is plotted on From the Progracilis Zone to the Subcontractus Zone the map. (Fig. 5, Map 12), diversity and disparity signals change 2. When the two curves are discordant, a circle is plot- in the same direction in most provinces. However, ted on the map. A black one if disparity increases during this time, some provinces display increasing while diversity decreases, and a white one if diversity diversity while their disparity decreases. increases while disparity decreases. Black and white From the Subcontractus Zone to the Bremeri Zone circles are smaller if either signal remains at the same (Fig. 5 (13–14)), diversity and disparity signals change level while the other changes. in the same direction for most provinces. Only a few provinces have signals varying in opposing directions. Heterogeneity between provinces in terms of the Spatial and temporal analyses of diversity and dis- relative variation in diversity and disparity signals can parity signals show marked differences in signal pattern thus be easily read from the maps. This information is between provinces. The global signals of diversity and deduced from the plots in Fig. 4. disparity presented previously (Figs. 2 and 3)presentan From the Bradfordensis Zone to the Concavum Zone, aggregate picture of what is happening in the different diversity and disparity change in the same direction in paleogeographical provinces. all provinces (Fig. 5, Map 1). Signal fluctuations are paleogeographically homogeneous for this period, 4. Discussion which is marked by increasing diversity and disparity worldwide. The various studies conducted in this paper show From the Concavum Zone to the Niortense Zone how useful it is to take account of paleogeographical (Fig. 5 Maps 2–6), the diversity and disparity signals are context when trying to interpret a radiation. very variable from one province to another. Although First of all, morphological disparity signals are not signals in most provinces change in the same direction, always correlated with taxonomic diversity signals. some provinces exhibit increasing diversity associated While the signals sometimes vary in the same direction, with decreasing disparity while others display the reverse they are often discordant. The significance of discor- pattern. Over this time, the global signals of diversity and dances between diversity and disparity patterns has been disparity exhibit generally similar fluctuations (Figs. 2 largely studied and subject to debates (Foote, 1993a, and 3), indicating that the global signal masks the 1999; Villier and Eble, 2004; Nardin et al., 2006). One paleogeographical heterogeneity of the radiation. interpretation suggests that the discordance is merely The transition from the Niortense Zone to the due to bias (e.g. linked to taxonomic level sampling, Garantiana Zone is marked by a substantial fall in temporal scale, or chosen morphological characters) global species diversity (Fig. 2) while disparity whereas the opposite suggests that the discordance recovers after hitting a low in the Niortense Zone reveals a true biological pattern. One main source of bias (Fig. 3). This pattern is visible in Map 7 of Fig. 5, with that potentially affects diversity and disparity patterns is many provinces displaying increasing disparity while the quality of the fossil record. Because, our study is diversity decreases. Other provinces show concordant based on a fully revised and updated dataset, we S. Moyne, P. Neige / Palaeogeography, Palaeoclimatology, Palaeoecology 248 (2007) 82–95 93 consider these two signals as being homogeneous (i.e. Bradfordensis Zone and Concavum Zone), the (because they are based on the same dataset). To resolve diversity and disparity signals are liable to differ the question of bias affecting taxonomic diversity or considerably between provinces. The first conclusion morphological disparity independently, some authors we can draw is that global signals provide only an have assessed their robustness using various statistical aggregate picture of what is happening in the different methods. Results for diversity show that a potential bias paleogeographical provinces. Paleogeography must be is the use of paraphyletic taxa when working at high taken into account when interpreting a radiation. Three taxonomic level (see Patterson and Smith, 1987). Others biological crises are detected from comparison of have argued that this is not a problem (Sepkoski and global signals (Fig. 6), corresponding to discordances Kendrick, 1993 and see Robeck et al., 2000 for a com- between diversity and disparity signals. These crises plementary point of view). Here, our analysis is based have different signatures in the different provinces on species counts, thus avoiding the paraphyly problem. (Fig. 5). For the first crisis (Laeviuscula Zone– In a recent paper, Villier and Eble (2004) – following Propinquans Zone), variations in diversity and dispar- Foote (1993a, 1999) – have tested the robustness of ity signals between provinces are concordant for most disparity signal. Using spatangoid echinoids, they have provinces but discordant for some; the discordances are shown that disparity signal was not significantly of the two possible types (black and white circles). The affected when various morphometric descriptors or second crisis (Niortense Zone–Garantiana Zone) is temporal resolution are used. Finally, and also because marked by half the provinces having concordant ammonites under study beneficiate from a long history signals and the other half having discordant signals, of research, we assume here that both taxonomic and all in the same direction (black circles, indicating morphological signals (respectively diversity and dis- increasing disparity and decreasing diversity). The parity) may be considered as reflecting true biological third crisis, which lasts longer (Zigzag Zone–Progra- patterns, except for the obvious bias when no ammonite cilis Zone), is marked by concordant signals for some record occurs (see before and Fig. 4). The observed provinces and by discordant signals always in the same discordances between the two may thus be interpreted as direction (black circles, indicating increasing disparity reflecting biological crises and not mere taxonomic or and decreasing diversity) except for two provinces sampling artefacts. (Province O, Zigzag Zone–Aurigerus Zone; Province The radiation in the Bajocian is marked initially B, Aurigerus Zone–Progracilis Zone) where discor- overall by signals of increasing diversity and disparity; dance is slight. The second part of this third crisis this may be directly linked to the fact that the signals (Aurigerus Zone–Progracilis Zone) is marked by as are at their lowest level at the Aalenian/Bajocian many as half the provinces exhibiting discordant boundary. The second stage of the radiation, during the signals. These three global crises do not have the Bathonian, is marked by discordance between diversity same signatures in the provinces: (1) the first crisis and disparity: morphological disparity remains high or (Laeviuscula Zone–Propinquans Zone) may be con- continues to increase while taxonomic diversity sidered a second-order (minor) biological crisis declines. From a depiction of these signals (Fig. 6), because it occurred in a small number of provinces three biodiversity crises can be identified (i.e. discor- and opposing signals are found in some provinces; (2) dant disparity and diversity signals): (1) from the the other two crises (Niortense Zone–Garantiana Zone Laeviuscula Zone to the Propinquans Zone; (2) from andZigzagZone–Progracilis Zone) may be considered the Niortense Zone to the Garantiana Zone; (3) from first order (major) biological crises because they the Zigzag Zone to the Progracilis Zone. These three occurred in at least half of the provinces and there episodes are characterized by varying degrees of are no opposing signals in the other provinces. decline in taxonomic diversity combined with an This study has shown the utility of a morphological upturn in morphological disparity. These crises can disparity signal. Such a signal provides additional in- be understood as lineages becoming extinct and which formation for interpreting biological radiations as has are nonselective in terms of morphology; morpholog- been already shown with other biological groups (e.g. ical disparity remains high and may even increase as Foote, 1991, 1999; Wills et al., 1994; Eble, 1998). taxonomic diversity falls (Foote, 1993a). Similarly, the breakdown of global signals among The breakdown of global signals of diversity and paleogeographical provinces reveals the heterogeneity disparity into signals for each paleogeographic prov- of a radiation. These results confirm those of Miller ince reveals that the radiation was not homogeneous, (1997b). They confirm the utility of taking account of except in the initial period. After the first two biozones paleogeography when interpreting a worldwide 94 S. Moyne, P. Neige / Palaeogeography, Palaeoclimatology, Palaeoecology 248 (2007) 82–95 radiation. Such breakdowns may distinguish between Efron, B., 1979. Bootstrap methods: another look at the Jacknife. – biological crises of varying intensity: (1) first order Annals of Statistics 7, 1 26. Efron, B., Tibshirani, R.J., 1993. An Introduction to the Bootstrap. crises may be considered as true global biological Chapman and Hall, New York. 436 pp. crises affecting a large number of provinces. These Enay, R., Mangold, C., 1994. Première zonation par ammonites du crises correspond to global changes in the evolutionary Jurassique d'Arabie Saoudite, une référence pour la province dynamics of the taxa under consideration; (2) second- arabique. Géobios. Mémoire spécial 17, 161–174. order crises cannot be considered as global biological Foote, M., 1991. Morphological and taxonomic diversity in a clade's history: the Blastoid record and stochastic simulations. Contribu- crises. They are visible in a small number of provinces tions from the Museum of Paleontology 28, 101–140. and sometimes in one alone. They generally result from Foote, M., 1993a. Discordance and concordance between morpholog- the influence of a change in one province with a large ical and taxonomic diversity. Paleobiology 19, 185–204. number of species. The ‘global’ signal is affected by Foote, M., 1993b. Contributions of individual taxa to overall – the signal for the one province alone and so is a local morphological disparity. Paleobiology 19, 403 419. Foote, M., 1999. Morphological diversity in the evolutionary radiation crisis and not a worldwide one. of Paleozoic and Post-Paleozoic crinoids. Paleobiology memoirs 1. Paleobiology 25, 1–115 (suppl. to no. 2). Acknowledgments Gradstein, F.M., Ogg, J.G., Smith, A.G., Bleeker, W., Lourens, L.J., 2004. A new geological time scale with special reference to – This research was supported by the ECLIPSE II Precambrian and Neogene. Episodes 27, 83 100. Harper, D.A.T., Owen, A.W., 1999. Quantitative and morphometric (Centre National de la Recherche Scientifique, CNRS) methods in taxonomy. In: Harper, D.A.T. (Ed.), Numerical research program. This study is a contribution to Equipe Palaeobiology. John Wiley and Sons Ltd., pp. 1–39. Forme, Evolution, Diversité of UMR CNRS 5561 Hillebrandt, A.v., 2001. Ammonite stratigraphy of the Bajocian in Biogéosciences, to FRE CNRS 2761 and to GDR 2474 northern Chile. Hantkeniana 3, 49–87. Morphométrie et évolution des formes. Hillebrandt, A.v., Smith, P., Westermann, G.E.G., Callomon, J.H., 1992. Ammonite zones of the circum-Pacific region. In: Westermann, G.E.G. (Ed.), The Jurassic of the Circum Pacific. Cambridge University References Press, pp. 247–272. House, M.R., 1985. Cephalopoda. In: Murray, J.W. (Ed.), Atlas of Babin, C., 1993. Rôle des plates-formes gondwaniennes dans les Invertebrate Macrofossils. The Palaeontological Association, diversifications des mollusques bivalves durant l'Ordovicien. pp. 114–152. Bulletin de la Société Géologigue de France 164, 141–153. Korn, D., Klug, C., 2003. Morphological pathways in the evolution of Benton, M.J., 2001. Patterns of diversity. In: Briggs, D.E.G., Early and Middle ammonoids. Paleobiology 29, Crowther, P.R. (Eds.), Palaeobiology, vol. II. Blackwell Science 329–348. Ltd, pp. 211–220. Krymholts, G.Y., Mesezhnikov, M.S., Westermann, G.E.G., 1988. The Callomon, J.H., 1994. Jurassic ammonite biochronology of Greenland Jurassic ammonite zones of the Soviet Union. Special Paper of the and the Arctic. Bulletin of the Geological Society of Denmark 41, Geological Society of America, vol. 223, 116 pp. 128–137. Mangold, C., Rioult, M., 1997. Bathonien. In: Cariou, E., Hantzpergue, Ciampaglio, C.N., Kemp, M., McShea, D.W., 2001. Detecting changes P. (Eds.), Groupe français d'étude du Jurassique. Biostratigraphie du in morphospace occupation patterns in the fossil record: character- Jurassique ouest-européen et méditerranéen: zonations parallèles et ization and analysis of measures of disparity. Paleobiology 27, distribution des invertébrés et microfossiles. Bull. Centre Rech. Elf 695–715. Explor. Prod., Mém., vol. 17, pp. 55–62 (Coords.). Contini, D., Elmi, S., Mouterde, R., Rioult, M., 1997. Aalénien. In: Miller, A.I., 1997a. Dissecting global diversity patterns: examples Cariou, E., Hantzpergue, P. (Eds.), Groupe français d'étude du from the Ordovician radiation. Annual Review of Ecology and Jurassique. Biostratigraphie du Jurassique ouest-européen et Systematics 28, 85–104. méditerranéen: zonations parallèles et distribution des inverté- Miller, A.I., 1997b. Comparative diversification dynamics among brés et microfossiles. Bull. Centre Rech. Elf Explor. Prod., palaeocontinents during the Ordovician radiation. Geobios. Mém, vol. 17, pp. 37–40 (Coords.). Mémoire Spécial 20, 397–406. Dommergues, J.L., Laurin, B., Meister, C., 1996. Evolution of Moyne, S., Neige, P., 2004. Cladistic analysis of the Middle Jurassic ammonoid morphospace during the Early Jurassic radiation. ammonite radiation. Geological Magazine 141, 115–123. Paleobiology 22, 219–240. Moyne, S., Neige, P., Marchand, D., Thierry, J., 2004. Répartition Dommergues, J.L., Laurin, B., Meister, C., 2001. The recovery and mondiale des faunes d'ammonites au Jurassique moyen (Aalénien radiation of Early Jurassic ammonoids: morphologic versus supérieur à Bathonien moyen): relations entre biodiversité et palaeobiogeographical patterns. Palaeogeography, Palaeoclimatol- paléogéographie. Bulletin de la Société Géologique de France 175, ogy, Palaeoecology 165, 195–213. 513–523. Eble, G.J., 1998. Diversification of disasteroids, holasteroids and Nardin, E., Rouget, I., Neige, P., 2006. Tendencies in paleontological spatangoids in the Mesozoic. In: Mooi, R., Telford, M. (Eds.), practice when defining species, and consequences on biodiversity Echinoderms: San Francisco. Balkema, Rotterdam, pp. 629–638. studies. Geology 33, 969–972. Eble, G.J., 2000. Contrasting evolutionary flexibility in sister groups: Navarro, N., 2003. MDA: a MATLAB-based program for morpho- disparity and diversity in Mesozoic atelostomate echinoids. space-disparity analysis. Computers and Geosciences 29, Paleobiology 26, 56–79. 655–664. S. Moyne, P. Neige / Palaeogeography, Palaeoclimatology, Palaeoecology 248 (2007) 82–95 95

Navarro, N., Neige, P., Marchand, D., 2005. Faunal invasions as a Sandoval, J., O'Dogherty, L., Guex, J., 2001. Evolutionary rates of source of morphological constraints and innovations? The Jurassic ammonites in relation to sea-level fluctuations. Palaios 16, diversification of the early Cardioceratidae (; Middle 311–335. Jurassic). Paleobiology 31, 98–116. Sandoval, J., O'Dogherty, L., Vera, J.A., Guex, J., 2002. Sea-level Neige, P., 2003. Spatial patterns of disparity and diversity of the changes and ammonite faunal turnover during the Lias/Dogger Recent cuttlefishes (Cephalopoda) across the Old World. Journal transition in the western Tethys. Bulletin de la Société Géologique of Biogeography 30 (8), 1125–1137. de France 173, 57–66. Neige, P., Marchand, D., Bonnot, A., 1997. Ammonoid morphological Sepkoski, J.J., Kendrick, D.C., 1993. Numerical experiments with signal versus sea-level changes. Geological Magazine 134 (2), model monophyletic and paraphyletic taxa. Paleobiology 19, 261–264. 168–184. Neige, P., Elmi, S., Rulleau, L., 2001. Existe-t-il une crise au passage Scotese, C.R., 1991. Jurassic and plate tectonic reconstruc- Lias – Dogger chez les ammonites? Approche morphométrique tions. Palaeogeography, Palaeoclimatology, Palaeoecology 87, par quantification de la disparité morphologique. Bulletin de la 493–501. Société Géologique de France 172, 257–264. Scotese, C.R., 1997. Paleogeographic atlas. PALEOMAP progress Odin, G.S., Odin, C., 1990. Echelle numérique des temps géologiques. report 90–0497. Department of Geology. University of Texas at Geochronologie 35, 12–20. Arlington, Arlington, Texas, 37 pp. O'Dogherty, L., Sandoval, J., Vera, J.A., 2000. Ammonite faunal Sepkoski, J.J., 1981. A factor analytic description of the Phanerozoic turnover tracing sea-level changes during the Jurassic (Betic marine fossil record. Paleobiology 7, 36–53. Cordillera, southern Spain). Journal of the Geological Society of Sepkoski, J.J., Hulver, M.L., 1985. An atlas of Phanerozoic clade London 157, 723–736. diversity diagrams. In: Valentine, J.W. (Ed.), Phanerozoic Owen, H.G., 1983. Atlas of the Continental Displacement, 200 Million Diversity Patterns. Profiles in Macroevolution. Princeton Univer- Years to the Present. Cambridge University Press, Cambridge. 159 pp. sity, pp. 11–39. Page, K.N., 1996. Mesozoic ammonoids in space and time. In: Thierry J. et al., 2000. Middle Toarcian. In: Dercourt, J, Gaetani, M., Landman, N.H., Tanabe, K., Davis, R.A. (Eds.), Ammonoid Vrielynck, B., et al. (Eds.), PAtlas peri-Tethys, Palaeogeographical Paleobiology. Topics in Geobiology, vol. 13, pp. 755–794. maps. CCGM/CGMW, Paris. map 8, (40 co-authors). Patterson, C., Smith, A.B., 1987. Is the periodicity of extinctions a Villier, L., Eble, G., 2004. Assessing the robustness of disparity taxonomic artefact? Nature 330, 248–251. estimates: the impact of morphometric scheme, temporal scale, and Raup, D.M., 1966. Geometric analysis of shell coiling: general taxonomic level in spatangoid echinoids. Paleobiology 30, problems. Journal of Paleontology 40, 1178–1190. 652–665. Raup, D.M., 1967. Geometric analysis of shell coiling: coiling in Vrielynck, B., Bouysse, P., 2001. Le visage changeant de la Terre. ammonoids. Journal of Paleontology 41, 43–65. L'éclatement de la Pangée et la mobilité des continents au cours Rioult, M., Contini, D., Elmi, S., Gabilly, J., 1997. Bajocien. In: Cariou, des derniers 250 millions d'années en 10 cartes. CCGM/CGMW, E., Hantzpergue, P. (Eds.), Groupe français d'étude du Jurassique. Paris. Biostratigraphie du Jurassique ouest-européen et méditerranéen: Wills, M.A., 2001. Disparity vs. diversity. In: Briggs, D.E.G., zonations parallèles et distribution des invertébrés et microfossiles. Crowther, P.R. (Eds.), Palaeobiology, vol. II. Blackwell Science Bull. Centre Rech. Elf Explor. Prod., Mém, vol. 17, pp. 41–53 Ltd, pp. 495–500. (Coords.). Wills, M.A., Briggs, D.E.G., Fortey, R.A., 1994. Disparity as an Robeck, H.E., Maley, C.C., Donoghue, M.J., 2000. Taxonomy and evolutionary index: a comparison of and Recent temporal diversity patterns. Paleobiology 26, 171–187. arthropods. Palaeobiology 20, 93–130.