Paleozoic Ammonoid Suture Morphology

PCA identified six factors with eigenvalues >1.0. Ordination and original spectral data for each sampled suture are listed in Allen (2005) and are available upon request. The first three components account for nearly 92% of the total variation in the data set. PC1 itself explains ~83% of the total variation and was loaded primarily by the first 20 harmonics. This is not surprising as the majority of Paleozoic suture morphologies are low-frequency sinusoids; relatively few cases exhibit significant contributions from mid-range (bifurcations-trifurcations) and higher-order folding (ammonitic detail and serrations). Because of this simplicity, the general pattern and frequency of most Paleozoic ammonoid sutures can be described solely by the lower portion of the spectrum and by the first principal component. However, the variation represented by the second and third principal components (~7% and 2%, respectively) is also informative, since these components represent variation in the higher-order harmonics (lobe elaboration, subdivision, and serration).

Two sets of bivariate plots, generated for each period, nicely summarize the tempo of morphospace exploration by the Paleozoic and basal-Triassic ammonoids (PC1-2, Fig. 8.1, and PC2-3, Fig. 8.2). As power spectra are nonlinear summaries

C20 P

m%

i ]

35

L.Triassic

33

Fig. 8.1 Bivariate suture morphospace and density contours. PC1 correlates to lobe number and amplitude; PC2 to lobe subdivision, with elaboration inversely correlated to score. Numbers refer to sutures illustrated in either Figs. 8.1 or 8.2 and correspond to taxa listed in Table 8.1. Sutures were reconstructed from sampled landmarks. Extinction survivors are indicated by gray squares. L.-M. Devonian morphospace (A; n = 90) is characterized by sutures possessing few broad, shallow, sinusoidal lobes, with the exception of a few unique, regular-sinusoidalmultilobate forms, e.g., Beloceras [4]. F-F extinction survivors fell within the mode of morphospace occupation during the L.-M. and U. Devonian (B; n = 110). End-Devonian extinction and Carboniferous morphospace (C; n = 245) reveals a significant shift away from typical Devonian forms; extinction survivors fell at the extremes of the U. Devonian morphospace, indicated by the dashed gray line. No ammonoid extinction is associated with the C-P boundary; holdover Pennsylvanian genera are numerous (n = 17) and cover the full range of Carboniferous morphologies. Novel taxa appearing within the Permian (D; n = 151) radiated near these points, establishing a bimodal distribution. Taxa surviving the P-T extinction fell at the extremes of the Permian occupied space. Postextinction recovery (E; n = 73) failed to recover the Permian mode corresponding to PC1 < 0, forms that had persisted since the L. Carboniferous.

4

3

L

.-M.Devonian

10 -8 -6 -4 -2 C PC3

2 4 6

16 21 C 21

15

Carboniferous

10

(Vv-B

12

U.Devonian

PC2 0

PC20

PC2 0

Fig. 8.2 Bivariate plots and density contours summarizing PC2-3 suture morphospace. Conventions as in Fig. 8.1. Although PC3 accounts for ~2% of the variation in the data set, it is still informative, differentiating self-similar (-PC3) from more irregular morphologies (+PC3). As in the PC1-2 space, F-F survivors (A, B) possessed modal L.-M. Devonian suture morphologies and new forms were generated within the limits established before the extinction. Also, recovery from the end-Devonian event(C) resulted in a near complete shift away from preexisting morphologies similar to that seen in PC1-2 space. During the Permian (D), the most extreme suture morphologies seen in the Paleozoic appeared, including ammonitic, yet sinusoidal forms, such as Timorites [28], which occupied a region of morphospace noticeably avoided by all other examined taxa (PC2 < 0 and PC3 < 0). Survivors of the P-T extinction were also distantly placed in PC2-3 morphospace, and recovery from the extinction (E) was typified by irregular suture forms and the failure to regenerate the multilobate, yet regularly sinusoidal morphologies that appeared in the late Paleozoic (PC3 < 0).

Fig. 8.2 Bivariate plots and density contours summarizing PC2-3 suture morphospace. Conventions as in Fig. 8.1. Although PC3 accounts for ~2% of the variation in the data set, it is still informative, differentiating self-similar (-PC3) from more irregular morphologies (+PC3). As in the PC1-2 space, F-F survivors (A, B) possessed modal L.-M. Devonian suture morphologies and new forms were generated within the limits established before the extinction. Also, recovery from the end-Devonian event(C) resulted in a near complete shift away from preexisting morphologies similar to that seen in PC1-2 space. During the Permian (D), the most extreme suture morphologies seen in the Paleozoic appeared, including ammonitic, yet sinusoidal forms, such as Timorites [28], which occupied a region of morphospace noticeably avoided by all other examined taxa (PC2 < 0 and PC3 < 0). Survivors of the P-T extinction were also distantly placed in PC2-3 morphospace, and recovery from the extinction (E) was typified by irregular suture forms and the failure to regenerate the multilobate, yet regularly sinusoidal morphologies that appeared in the late Paleozoic (PC3 < 0).

of suture shape, it is difficult to assign morphological states to the principal axes. However, the first principal component seems to correlate with the relative frequency and amplitude of lobes along the suture length, with number of lobes increasing, and amplitude decreasing, with larger scores. The second principal component correlates most strongly with the shape of the ventral portion of the suture; positive values correspond mostly to sutures possessing pronged ventral lobes, (e.g., Pseudopronorites [suture 17]; Fig. 8.2C) or ventral saddles (e.g., Schartymites [suture 21]; Fig. 8.2C), and the negative values to sutures with ventral lobes (e.g., Costimitoceras [suture 15]; Fig. 8.1C). Presence or absence of very-high-frequency subdivisions also load on PC2, with negative values coinciding with more highly serrated forms. The third principal component discriminates sutures resembling regular, periodic sinusoids from irregular forms, with irregularity increasing with increasing PC3 score.

Density contours, calculated using JMP-IN's default algorithm, were added to the PC plots to highlight patterns of space occupation through time. Also illustrated are sutures representative of the extremes and modes of the sampled geometries, with their taxonomic affiliations listed in Table 8.1. These data reveal a strikingly dynamic pattern of suture evolution: the occupied region of suture morphospace shifts significantly across period and extinction boundaries. These changes are described in more detail below.

Frasnian-Famennian extinction. Five genera survived the Frasnian-Famennian extinction (376 Ma), all of which possessed the modal suture morphotypes of the lower-middle Devonian (Figs. 8.1A, B, 8.2A, B; survivors indicated as gray squares on upper Devonian plots, Figs. 8.1B, 8.2B). After the extinction, the majority of the lower-middle Devonian suture morphospace was reoccupied, with the modal morphotypes being maintained. Noticeable differences between the two distributions included the failure of highly multilobate forms (e.g., Beloceras [suture 4]; Fig. 8.2A) to reevolve, and the appearance of irregular clymeniid morphotypes (e.g., Falciclymenia [suture 8], Otoclymenia [suture 10], Discoclymenia [suture 11]; Figs. 8.1B and 8.2B).

End-Devonian extinction. In contrast, the rebound and subsequent filling of suture morphospace in the Carboniferous after the end-Devonian extinction saw a near-complete shift away from pre-existing geometries. Extinction survivors fell at the extreme of the Devonian suture morphospace (survivors indicated as squares on Carboniferous plots; Figs. 8.1C, 8.2C). These taxa also fell well outside the main Carboniferous-occupied space. When compared to suture complexity, as measured by SCI, this morphological changeover is somewhat surprising. Both intervals were dominated by ammonoids with very simple sutures; ~80% of Devonian and ~60% of Carboniferous genera had SCI values less than 5 (i.e., ~3 lobes in the sampled region), and average Mississipian suture complexities closely mirrored those of middle and late Devonian taxa (Saunders et al., 1999; see also Allen, 2005). Despite this, overlap between the two faunas in all three dimensions of the PC space is minimal. Devonian taxa had broader, shallower lobes than their Carboniferous counterparts. Additionally, the lobes of Devonian sutures were more rounded (resembling a true sine wave; on average, values less than zero on PC3) than those of the Carboniferous forms, which tapered more strongly to form a point, i.e., typical "goniatitic" suture geometries (Figs. 8.1C, 8.2C).

Table 8.1 Genera associated with sutures referred to in text and illustrated in Figs. 8 1, 2. L.- M. Devonian: 1-5; U. Devonian: 6-12; Carboniferous: 13-21; Permian: 22-28. L. Triassic: 29-36. ID corresponds to specimen number in Saunders et al., 1999, 2004; Allen, 2005.

No.

Genus

Order

ID

1.

Delephinites

Anarcestida

57

2.

Anetoceras

Anarcestida

1

3.

Sphaeromanticoceras

Anarcestida

52

4.

Beloceras

Anarcestida

88

5.

Mimagoniatites

Anarcestida

14

6.

Prionoceras

Goniatitida

201

7.

Uraloclymenia

Clymeniida

96

8.

Falciclymenia

Clymeniida

133

9.

Oxytornoceras

Goniatitida

178

10.

Otoclymenia

Clymeniida

112

11.

Discoclymenia

Clymeniida

181

12.

Mimitoceras

Goniatitida

209

13.

Nuculoceras

Goniatitida

333

14.

Marathonites

Goniatitida

479

15.

Costimitoceras

Goniatitida

205

16.

Vidrioceras

Goniatitida

486

17.

Pseudopronorites

Prolecanitida

542

18.

Eothalassoceras

Goniatitida

403

19.

Kazakhastania

Goniatitida

216

20.

Artinskia

Prolecanitida

554

21.

Schartymites

Goniatitida

351

22.

Atsabites

Goniatitida

456

23.

Shangracoceras

Goniatitida

232

24.

Almites

Goniatitida

480

25.

Artioceratoides

Prolecanitida

563

26.

Sosiocrimites

Goniatitida

523

27.

Tapashanites

Ceratitida

573

28.

Timorites

Goniatitida

496

29.

Subinyoites

Ceratitida

1201

30.

Beatities

Ceratitida

1235

31.

Pseudoaspenites

Ceratitida

1236

32.

Tirolites

Ceratitida

1128

33.

Rhacophyllites

Ceratitida

1293

34.

Proavites

Ceratitida

1149

35.

Clypopoceras

Ceratitida

1168

Carboniferous-Permian. By the onset of the Permian, the relative proportion of simple sutured taxa was greatly reduced (~12% with SCI < 5; Saunders et al., 1999). In the PC-suture morphospace, this deficit is reflected by thinning in the region of modal Carboniferous morphospace occupation across the Carboniferous-Permian (C-P) boundary (Figs. 8.1C, D, 8.2C, D). The 17 ammonoid genera that crossed the C-P boundary were scattered evenly throughout the Carboniferous suture morphospace, falling both within the mode and at the extremes (holdovers indicated as gray squares on Permian plots; Figs. 8.1D, 8.2D).

During the Permian, a wide range of novel suture morphologies originated, most of which fell within the boundaries established by the Carboniferous ammonoids. This morphological radiation led to a bimodal distribution in PC1-PC2 morphospace with each mode centered near the holdover morphotypes. This differentiation highlights the separation between groups characterized by serrated and fractal-like ammonitic lines (more positive PC1 values; Carboniferous extremes) and more typical "Paleozoic" sutures (Carboniferous mode; Figs. 8.1D, 8.2D).

Permian-Triassic. The early Triassic saw a continuation of the evolutionary shift initiated in the Permian. The three genera that survived the end-Permian extinction event (Xenodiscus, Otoceras, and Episageceras) fell at the extremes of the Permian morphospace (survivors indicated as gray squares on Triassic plots; Figs. 8.1E, 8.2E). Xenodiscus and its descendants formed the rootstock of the Triassic Ammonoidea, while Otoceras and its descendants had short-lived success and the prolecanitid Episageceras survived only briefly (McGowan, 2004).

The early Triassic was typified by increasing lobe serration, but also saw the reemergence of both "simple" forms (albeit with irregular folding; e.g., Beatities [suture 30]) and subammonitic ones (e.g., Rhacophyllites [suture 33]). There are also notable gaps in the distribution. In particular, sutures with PC1 scores less than zero rarely reappeared, indicating a failure to regenerate Devonian and most Carboniferous morphotypes (Fig. 8.1E).

Was this article helpful?

0 0

Post a comment