Results Body Size

The average maximum vesicle diameter of acritarchs displayed non-directional fluctuation through the Proterozoic (Fig. 2). Acritarch body size decreased significantly across the Neoproterozoic-Paleozoic transition, but had increased significantly by the middle/late Cambrian (though not to the size seen in the late Neoproterozoic). The average figure vesicle diameter of acritarchs displayed a similar pattern, though at smaller sizes (Fig. 2). Average maximum diameter and average figure diameter are significantly positively correlated, and figure data generally underestimate maximum vesicle diameter reported in systematic description (Fig. 2 inset). Retrieving body size information from figured specimens appears to be a legitimate approach with Proterozoic and Cambrian acritarchs when one is interested in investigating long-term patterns.

Table 4. Binning structure for morphometric analyses.

Bin

Species Occurrences

P1

104

Ml

11

M2

41

N1

248

N2

13

N3

156

Cl

54

C2

115

C3

36

A .. ^^

iV \ T

Maximum

protero-

zoic

Mesoproterozoic

Neoproterozoic

Paleozoic r-Q 1, C2

PI 1

1 Ml

M2

NI 1 N2 1 N3

1800 1600 1400 1200 1000 800 600 400

1800 1600 1400 1200 1000 800 600 400

Figure 2. Size history of acritarchs. Solid circles represent log-transformed mean maximum vesicle diameter of acritarchs (from species descriptions) through time with 95% CI calculated from 1000 iteration naïve bootstrapped sampling distribution. Hollow triangles represent log-transformed mean vesicle diameter of acritarchs as measured from figured specimens. Inset shows relationship between maximum reported sizes and sizes of figured specimens. R-sq (R-squared) and p-value from Pearson correlation analysis performed in SAS 9.1.

Table 5. Correlation analyses between measures of disparity and body size.

Raw Data

First Differences

Pearson

Pearson

r2

p

r2 p

MDS

0.113

0.38

0.180 0.29

Dissimilarity

0.038

0.62

0.247 0.21

The absence of any notable long-term trend in acritarch body size minimizes the likelihood of mistaking spurious disparity trends due to secular changes in body size (e.g., morphological disparity may increase due to diffusive increase in body size or body size range) with shifts in size-invariant shape disparity. Neither of the disparity metrics analyzed show any significant correlation with body size for all possible comparisons, including both raw data and data corrected for autocorrelations by first differencing (Table 5). This discordance between body size and morphospace occupation suggests that the disparity trends discussed below are not an allometric derivative of changes in body size.

3.2 Morphological Disparity

3.2.1 Dissimilarity

Mean dissimilarity was very low in the Paleoproterozoic (<0.02; Fig. 3A). This low value starkly contrasts with the high species per formation values calculated from our database (Fig. 3C). We interpret this stark contrast as severe taxonomic over-splitting in the Paleoproterozoic. Some caution in such an interpretation may be warranted due to the lack of cell wall thickness data in our matrix. However, reports of cell wall thickness in species descriptions are overwhelmingly qualitative (e.g., thick or thin), and are likely not consistently applied between workers. Moreover cell wall thickness is likely highly susceptible to taphonomic processes such as degradation.

A significant increase in mean dissimilarity occurred between the P1 and Ml bin, with an Ml value of 0.08. Mean dissimilarity coefficients reached a plateau during the M1 bin that would remain through the early Neoproterozoic (M2=0.10 and N1=0.09). A slight, yet significant, decrease in dissimilarity occurred during the Cryogenian (N2=0.08 and upper 95% confidence interval <0.09).

A rapid morphological diversification occurred in the early Ediacaran period, resulting in a mean dissimilarity coefficient significantly higher than any seen in previous bins (N3= 0.15). This increase in morphological disparity, though dramatic, was short-lived. The first appearance of the Ediacara organisms (~575 Ma) corresponds in time with a dramatic decrease in acritarch disparity. We did not create a separate geochronological bin (i.e., 575-542 Ma) due to low data density. All known acritarchs from this time are simple sphaeromorphic leiosphaerid-like vesicles. Moreover, it would be impossible to calculate a dissimilarity coefficient (much less mean dissimilarity) in such a bin as our data base contains only one known named

species (Leiosphaeridia sp.) in this time interval (Fig. 3C); although other species (e.g., Bavlinella faveolata) may also be present in this interval (Germs et al., 1986). Therefore the dramatic decrease in disparity associated with the first appearance of Ediacaran organisms and the rapid diversification seen in the pre-trilobite early Cambrian (C1) is much more dramatic than Fig. 3A suggests.

Mean dissimilarity coefficients increased monotonically through the Cambrian. Pre-trilobite Early Cambrian mean dissimilarity (C 1=0.11) reflects the morphological diversification following the late Ediacaran drop in disparity addressed above. Mean dissimilarity coefficients continued to increase significantly through the trilobite-bearing Early Cambrian (C2=0.12) and Middle and Late Cambrian (C3=0.15), achieving the high level of disparity seen in the early Ediacaran (N3).

3.2.2 Non-metric Multidimensional Scaling

Non-metric multidimensional scaling analysis shows significant secular variation in acritarch morphologies (Fig. 3B), and is broadly similar to the dissimilarity pattern. The MDS trend is unlikely a sampling artifact as its overall trajectory falls outside of the randomization's 95% confidence intervals (Fig. 3B inset).

MDS variance was very low in the Paleoproterozoic (P1=0.35; Figs. 3B, 4). This indicator of low disparity is also in stark contrast with high species per formation values (Fig. 3C), and is indicative of taxonomic over-splitting (see above). A significant increase in MDS variance occurred in the early Mesoproterozoic (M1=1.24), signaling the beginning of a disparity plateau that would continue through the early Neoproterozoic (M2=1.49 and N1=1.70). This plateau is apparent in Fig. 3 A-B, but not in Fig. 4. This is because the convex hulls in Fig. 4 to a large extent reflect sampling intensity as well as morphological disparity in each bin.

MDS variance decreased during the Cryogenian (N2=1.48) (Figs. 3B, 4). This morphological contraction, together with taxonomic decrease (Knoll, 1994; Vidal and Moczydlowska-Vidal, 1997; Xiao, 2004a), indicates

Figure 3. (on Page 34) History of acritarch disparity. (A) Mean dissimilarity coefficient ± 1 standard error (Note: the standard error brackets for P1 are smaller than the data point.). (B) Variance from multivariate analyses. Black circles are MDS variance. Black squares are PCA variance. Inset graph displays results of MDS randomization. Center line represents mean variance from 1000 iteration randomization. Lower and upper lines represent 95% confidence intervals. (C) Number of species per formation from this study's database, coded according to sampling intensity (number of processed rock samples) of each formation. Vertical black lines represent era boundaries. The gray box represents the Cryogenian. The vertical light gray line at ~575 Ma represents the first appearance of the Ediacara organisms.

possible acritarch extinction during the Cryogenian. Further analysis of MDS plots and loading reveals the restriction of acritarchs from the right-hand side of the morphospace (Fig. 4), suggesting that the Cryogenian acritarch extinction strongly affected acanthomorphic forms. Acritarchs with hollow, cylindrical, blunt-tipped processes are notably absent in the Cryogenian (Knoll, 1994). The post-Cryogenian recovery of acritarchs resulted in the highest MDS variance seen until that time (N3=2.23).

MDS Loading

MDS Loading

Acritarch Morphology Character

Figure 4. MDS and PCA scatter plots and loading. MDS scatter plots for the nine geochronological bins and the MDS loading chart relating variables to Dim 1 (x-axis) and Dim 2 (y-axis). Solid outlines are convex hulls for bin data. Dashed outlines are convex hulls for pooled data representing maximum realized morphospace. Note how MDS and PCA scatter plots and loadings are mirror images of one another.

Figure 4. MDS and PCA scatter plots and loading. MDS scatter plots for the nine geochronological bins and the MDS loading chart relating variables to Dim 1 (x-axis) and Dim 2 (y-axis). Solid outlines are convex hulls for bin data. Dashed outlines are convex hulls for pooled data representing maximum realized morphospace. Note how MDS and PCA scatter plots and loadings are mirror images of one another.

1800

1600

1600

Paleo-

protero-zoic

Mesoprolerozoic

efe f&C3

PI 1

1 Ml 1 1

vl2

Nl

N7

Ni

[

o o qom® • • it

2*

4m • 5« —

7» •

M ••

• ID

11*

«>

12*

M* «

- (D

13*

• O

14*

M M

r

•QD

15*

EL

• •

-co

17» •

HIM

O

•o

18* • 19» •

VI

• •

B

22*

m iR1

• •

•• •»

•I,

« o -©

23«

*

• ••

®

24*

• • •

©

25» •

• •

-o

26»

• •

• ®

27»

m

28«

•••

•o

29*

• o

30

•0

31

• ®

Figure 5. Stratigraphic occurrences of morphological characters utilized in this study: 1) spherical vesicle; 2) ellipsoidal vesicle; 3) barrel-shaped vesicle; 4) bulb-shaped vesicle; 5) polyhedral vesicle; 6) medusoid vesicle; 7) cylindrical process; 8) hemispherical process; 9) tapered process; 10) hair-like process; 11) triangular process; 12) rounded-tip process; 13) capitate-tip process; 14) blunt-tip process; 15) pointed-tip process; 16) funnel-tip process; 17) hollow process; 18) interior of process communicates with interior of vesicle; 19) branching process; 20) processes fuse at tip; 21) enveloping membrane; 22) excystment-like structure; 23) internal bodies in vesicle; 24) concentric ornamentation on vesicle surface; 25) plates on vesicle; 26) multi-celled appearance (vesicles contained in a larger envelope); 27) colonial appearance (aggregation of vesicles); 28) pores in vesicle wall; 29) flange ornamentation; 30) crest ornamentation; 31) costae meshwork surrounding vesicle.

MDS variance decreased between the early Ediacaran and the pre-trilobite Early Cambrian (C 1=2.04), concurrent with the diversification of Ediacara organisms. MDS variance increased monotonically through the remainder of the Cambrian in step with the taxonomic diversification of acritarchs and animals (C2=2.16, C3=2.90).

The dissimilarity coefficient and MDS results, described above, are broadly supported by the geochronological distribution of morphological characters (Fig. 5). Paleoproterozoic acritarchs typically had spherical vesicles with the occasional medial split (e.g., Schizofusa sinica), enveloping membrane (e.g., Pterospermopsimorpha pileiformis), internal bodies (e.g., Nucellosphaeridium magnum), or concentric surface ornamentation (e.g., Thecatovalvia annulata and Valvimorpha annulata). The Mesoproterozoic saw the first appearance of elliptical vesicles (e.g., Fabiformis baffinensis), ten process-related characters (e.g., Shuiyousphaeridium macroreticulatum and Tappania plana), vesicle plates (e.g., S. macroreticulatum and Dictyosphaera delicata), pores in vesicle walls (e.g., Tasmanites volkovae), and multi-celled and colonial appearance (e.g., Majasphaeridium sp. and Satka squamifera). Of the thirty-one characters identified in this study, fifteen first appeared in the Mesoproterozoic (nine in the M1 bin and six in the M2 bin). Many more acritarch body plans evolved in the Neoproterozoic. Polyhedral vesicles (e.g., Octoedryxium truncatum), bulb-shaped vesicles (e.g., Sinianella uniplicata), medusoid vesicles (e.g., Multifronsphaeridium pelorium), barrel-shaped vesicles (e.g., Artacellularia kellerii), triangular and hair-like processes (e.g., Cymatiosphaera wanlongensis and Dasysphaeridium trichotum), funnel-tipped processes (e.g., Briareus borealis), processes that fuse at the tips (e.g., ectophragm acanthomorph from Butterfield and Rainbird 1998), and flange ornamentation about the vesicle equator (e.g., Simia simica) all appear for the first time in the Neoproterozoic. Two new morphological characters appeared in the Cambrian: a costae meshwork that surrounds the vesicle (e.g., Retisphaeridium brayense) and crest-ornamentation—equatorial ornamentation that is similar to flange but does not circumvent the vesicle resulting in wing-like structures (e.g., Pterospermella solida). It should be noted that our data for Cambrian acritarchs were not as exhaustive as our Proterozoic data, and that further investigation would likely reveal more first appearances of characters in the Cambrian than what we report. Another caveat is that the same morphological characters in different taxa or in different geochronological bins may not be homologous. Such simple characters may have evolved multiple times.

The two proxies of morphological disparity used in this study, mean dissimilarity coefficient and MDS variance, resulted in coherent histories of acritarch morphological disparity (Fig. 3). The agreement of the two methods, the independent verification of the significance of the trends found by each method (i.e. randomization for MDS and calculation of standard error for mean dissimilarity), the elimination of allometry as a confounding factor, and the invariance of disparity estimates relative to unequal binning characters (e.g., temporal duration of bin, number of formations per bin, number of sampling localities per bin, number of processed rock samples per bin, and number of species occurrences per bin) (Table 6) all attest to the robustness of our interpreted history of acritarch morphological evolution.

Table 6. Correlation analyses between measures of disparity and unequal binning characters.

Table 6. Correlation analyses between measures of disparity and unequal binning characters.

MDS Variance

Mean Dissimilarity Coefficient

Spearman

Spearman

r2

p

r2

p

Duration of Bin

-0.639

0.06

-0.571

0.11

Number of Formations

0.367

0.33

0.428

0.25

Number of Locations

0.092

0.81

0.185

0.63

Number of Samples

0.067

0.86

0.100

0.80

Species Occurrences

0.333

0.38

0.317

4.1 Comparative Histories of Morphological Disparity and Taxonomic Diversity

The morphological disparity of acritarchs (as approximated by mean dissimilarity coefficients, MDS variance, and stratigraphic ranges of individual morphological characters) initially increased significantly by the early Mesoproterozoic (Fig. 3, 5). In contrast, the first taxonomic radiation did not occur until the early Neoproterozoic (Fig. 1). This increase in disparity preceded the first major taxonomic radiation by approximately 500 million years. This statement remains true even if the diversity curve (Knoll, 1994; Vidal and Moczydlowska-Vidal, 1997) is updated with more recent data (Xiao etal., 1997; Yin, 1997; Javaux etal., 2001), although the addition of the exuberantly over-split Paleoproterozoic taxa (Fig. 3C) to the diversity curve may significantly change the picture. However, as discussed earlier and implied in previous compilations of acritarch diversity (Knoll, 1994; Vidal and Moczydlowska-Vidal, 1997), such over-splitting is not justified.

The pattern of high morphological disparity early in the history of acritarchs is very similar to patterns seen in the evolution of multi-celled organisms in the Phanerozoic. Many groups of organisms in the Phanerozoic display high morphological disparity early in their history: Cambrian metazoa (Thomas et al, 2000), marine arthropods (Briggs et al., 1992), Paleozoic gastropods (Wagner, 1995), seed plant leafs (Boyce, 2005), and Cenozoic ungulate teeth (Jernvall et al., 1996). Thus, high morphological disparity in the early evolutionary history appears to be a prevailing, although not universal, pattern among many groups (Foote, 1997). As far as we know, this study documents the first example of a similar pattern in protists and in the Precambrian. It is becoming apparent that morphological diversification preceding taxonomic diversification may be a prevailing pattern in eukaryote evolution.

Our comparative analysis of disparity and diversity does differ from the results of other studies. Morphological disparity typically approaches its maximum realized value early in the history of other clades [e.g., Paleozoic crinoids (Foote, 1995)], but our analysis reveals periodic expansions of realized morphospace (Fig. 3). Because the Group Acritarcha is undoubtedly polyphyletic and includes organisms from many phyla or divisions (Butterfield, 2004, 2005), the periodic expansion of acritarch morphospace is best interpreted as a result of the evolutionary appearance of new clades. In particular, fluctuation of acritarch morphospace in the Neoproterozoic and Cambrian may represent the coming and going of different eukaryote groups.

4.2 Linking Morphological Disparity with Geological and Biological Revolutions

4.2.1 Morphological Constraints, Convergence, and Nutrient Stress in the Mesoproterozoic

The ~1500 Ma (M1) morphological expansion was followed by a prolonged plateau of morphological disparity until ~800 Ma (N1). Constraints on protist morphology likely played a dominant role in a significant part of protist history from 1500 Ma to 800 Ma. The increasingly populated morphospace during this period suggests that either the morphological history of acritarchs was characterized by convergent morphologic evolution of phylogenetically unrelated groups or that diversification was restricted within morphologically similar clades. Either way, it is remarkable that the morphological constraints were not overcome for such a long time given that the Group Acritarcha is polyphyletic and thus includes multiple clades (Butterfield, 2004, 2005).

Buick and others (1995) described the Mesoproterozoic as "the dullest time in Earth's history (p. 153)" and remarked that "never in the course of Earth's history did so little happen to so much for so long (p. 169)". These statements were based upon 513C values that hovered around 0%o with little change for nearly 600 million years (1600-1000 Ma) (Buick et al., 1995; Xiao et al., 1997; Brasier and Lindsay, 1998). The global rate of organic carbon burial relative to inorganic carbon burial, as inferred from the Bangemall Group of northwestern Australia and equivalent carbonates elsewhere, remained unchanged through the Mesoproterozoic, resulting in the static 513C pattern. This was ascribed to relatively little environmental and tectonic changes during this most lackluster era (Buick et al., 1995). Tectonic and environmental tranquility would lead to low bioproductivity through nutrient stress such as phosphorus limitation (Brasier and Lindsay, 1998) and/or the dearth of metabolically important trace metals in the Mesoproterozoic oceans (Anbar and Knoll, 2002).

Our results suggest that Buick and others were only partially correct in their depiction of the Mesoproterozoic as being irksome and tedious. Our quantitative measures of acritarch morphological disparity do suggest a long plateau lasting ~600 million years. Similarly, qualitative data suggest that the taxonomic turnover rate of acritarchs during the Mesoproterozoic and early Neoproterozoic was much lower than that of the late Neoproterozoic (Peterson and Butterfield, 2005). However, the first appearance of nearly half the morphological characters considered in this study (15 of 31) occurred during the early Mesoproterozoic, well within the time of subdued 513C fluctuations, and the plateau continued into the early Neoproterozoic when the carbon cycle fluctuated moderately. Is this plateau indeed related to Mesoproterozoic nutrient stress? The great temporal overlap between acritarch disparity plateau and Mesoproterozoic geochemical stasis is suggestive of a possible causal relationship, but the apparent mismatch in their initiation and termination raises some concerns. At the present, the mismatch cannot be fully addressed because of poor temporal resolution in acritarch and 513C data, as well as poor understanding of the response time (lag time) between the different components of the Earth system.

4.2.2 Neoproterozoic Global Glaciations

The late Neoproterozoic saw perhaps the most dramatic of global climatic events in the history of Earth. It has been hypothesized that multiple global glaciations occurred during this time (~720-630 Ma), even to the extent of glaciers at the equator with tropical sea ice 1 km thick (Kirschvink, 1992; Hoffman et al., 1998; Hoffman and Schrag, 2002). It is reckoned that the "snowball Earth" glaciations lasted for approximately 10

million years (Hoffman et al., 1998; Bodiselitch et al., 2005). With the carbon cycle cut short, due to completely iced-over oceans, the CO2 concentration in the atmosphere (sourced by volcanic out-gassing) would build up, eventually resulting in greenhouse conditions and deglaciation. The deglaciation events were also likely very dramatic, with wind and waves unlike those seen on Earth before or since (Allen and Hoffman, 2005).

The controversial snowball Earth hypothesis has been criticized on biological grounds (Runnegar, 2000; Corsetti et al., 2003; Olcott et al., 2005). The fossil record clearly indicates that several major photosynthetic clades, including green, red, and chromophyte algae (Butterfield et al., 1994; Butterfield, 2000, 2004), evolved prior to the Cryogenian glaciations. If the snowball model is correct then these three algal clades must have survived the global glaciations, either in sea ice cracks, hydrothermal vents, fresh water melt ponds (Hoffman et al., 1998; Hoffman and Schrag, 2002), or perhaps in an ice-free tropic ocean that may have persisted during the snowball Earth events (Hyde et al., 2000; Runnegar, 2000).

Acritarchs did experience significant change in the Cryogenian. Morphological disparity (Fig. 3) as well as global taxonomic diversity (Fig. 1) decreased significantly in the Cryogenian (N2). It is possible that the Cryogenian suffers from fewer acritarch assemblages reported in the literature; however, Cryogenian acritarch assemblages (Knoll et al, 1981; Vidal, 1981; Vidal and Nystuen, 1990; Yin, 1990) do show lower taxonomic diversity and morphological disparity than older and younger assemblages. Large acritarchs and complex acanthomorphic acritarchs are few in the Cryogenian (Fig. 2, 4). This pattern does suggest that, whether the tropical ocean remained ice-free during the snowball Earth events, eukaryotes did suffer significant loss in the Cryogenian.

Runnegar hypothesized about the biological consequences of the various explanations for Cryogenian glaciations (Runnegar, 2000). A strict snowball scenario would result in an evolutionary bottleneck with the extinction of all but a few eukaryotic lineages. A slushball scenario with ice-free tropical seas would result in a blue-water refugium with the selective filtering of eukaryotic lineages favoring planktonic open ocean forms. He also hypothesized a scenario in which global refrigeration would have had mild impact on the biosphere. Paleoenvironmental analysis appears to suggest that acanthomorphic acritarchs tend to occur in near-shore facies as compared to leiosphaerids (Butterfield and Chandler, 1992). If this paleoecological pattern holds true for all Proterozoic acritarchs, then the selective extinction of acanthomorphic acritarchs during the Cryogenian may be taken as evidence in support of Runnegar's blue-water refugium hypothesis (Runnegar, 2000). It remains to be seen whether benthic algae survived Cryogenian, and, if not, whether post-Cryogenian benthic ecosystem recruited from planktonic algae that did survive glaciations.

4.2.3 Ediacara Organisms

The first macroscopic complex organisms in the fossil record are members of the Ediacara biota and first appeared approximately 575 Ma, within 5 million years after the 580 Ma Gaskiers glaciation that lasted no more than one million years (Narbonne, 1998; Narbonne and Gehling, 2003; Narbonne, 2005). The phylogenetic affinity of many of these organisms is controversial, but whether they represent the ancestors of modern organisms (Runnegar and Fedonkin, 1992) or a failed evolutionary experiment (Seilacher, 1992; Buss and Seilacher, 1994) they certainly indicate a basic ecological restructuring of the world previously dominated by prokaryotes and single-celled eukaryotes (Lipps and Valentine, 2004). The varied body plans of Ediacara organisms suggest equally varied trophic strategies, probably including heterotrophy. Evidence for the presence of heterotrophic consumers includes molluscan-grade bilaterians (Fedonkin and Waggoner, 1997), cnidarian-grade metazoans (Runnegar and Fedonkin, 1992), scratch marks interpreted as radular grazing traces (Seilacher, 1999; Seilacher et al., 2003), epifaunal tiering (Clapham and Narbonne, 2002), and boring of mineralized exoskeletons (Bengston and Zhao, 1992; Hua et al., 2003).

In light of the 580 Ma Gaskiers glaciation and probable consumers in the late Ediacaran (575-542 Ma), it is instructional to explore their possible effects on the primary producers (as represented by most acritarchs). Our data show that acritarch morphological disparity and taxonomic diversity in the late Ediacaran decreased to levels not seen since the Paleoproterozoic (though we didn't construct a separate bin for this time, see Section 3.2.1 and Fig. 3C). During this time, all acritarchs were of simple leiosphaerid-like and Bavlinella-like morphologies, and all acritarchs characteristic of early Ediacaran (so called Doushantuo-Pertatataka acritarchs) disappeared.

To test whether the Gaskiers glaciation, the diversification of Ediacara organisms, or perhaps something else caused the disappearance of Doushantuo-Pertatataka acritarchs, we need to sort out the exact temporal relationship between several geobiological events. In South Australia, the appearance of Doushantuo-Pertatataka acritarchs occurred long after the Marinoan glaciation and shortly after the Acraman Impact, which has been estimated to be 580 Ma (Grey et al., 2003). However, the late appearance of Doushantuo-Pertatataka acritarchs in South Australia was probably due to regional, environmental, or preservational biases. In South China, Doushantuo-Pertatataka acritarchs first appeared about 632 Ma (Condon et al., 2005), shortly after the Nantuo glaciation that is considered equivalent to the Marinoan glaciation in South Australia (Xiao, 2004a; Zhou et al., 2004). Condon et al. (2005) estimated that Doushantuo-Pertatataka acritarchs persisted at least 50 million years and disappeared somewhere between 580 Ma and 550 Ma. If true, both the Acraman Impact and the Gaskiers glaciation predate, perhaps significantly, the disappearance of Doushantuo-Pertatataka acritarchs. Hence, neither the Acraman Impact nor the Gaskiers glaciation may have directly contributed to the disappearance of Doushantuo-Pertatataka acritarchs.

It is more likely that herbivory by, or other ecological interactions with, Ediacara organisms led to the decline of Doushantuo-Pertatataka acritarchs in the late Ediacaran Period. This hypothesis is distinct from a recent hypothesis proposed by Peterson and Butterfield (2005), who suggest that the origin, not the extinction, of Doushantuo-Pertatataka acritarchs was a consequence of ecological interactions with early eumetazoans. Both hypotheses need to be tested against more precise geochronological data and to be examined for possible taphonomic bias against acritarch preservation in the late Ediacaran Period. If either hypothesis survives more rigorous tests in the future, the origin or extinction of Doushantuo-Pertatataka acritarchs would be the first top-down driven macroevolutionary event recorded in the fossil record (Vermeij, 2004).

4.2.4 Cambrian Explosion of Animals

Perhaps the most dramatic event in the history of life began approximately 540 Ma at the beginning of the Cambrian Period. Almost all known metazoan phyla diverged in the Early-Middle Cambrian (Conway Morris, 1998; Levinton, 2001; Zhuravlev and Riding, 2001; Valentine, 2004). The Cambrian explosion resulted in major ecological restructuring of the biosphere (Zhuravlev and Riding, 2001) and alteration of sedimentation patterns (Bottjer et al., 2000; Droser and Li, 2001).

It has been noted by several observers that acritarch diversity increased in step with animal evolution during the Cambrian explosion (Knoll, 1994; Vidal and Moczydlowska-Vidal, 1997; Butterfield, 2001). So did acritarch morphological disparity (Figs. 1, 3). This implies a close link between these two ecological groups during the radiation. The nature of these links, however, is less clear. It has been argued that morphological diversification of phytoplankton, as shown in acritarch morphology, was an ecological response to the evolution of filter-feeding mesozooplankton in the Cambrian (Butterfield, 1997, 2001). It is also possible that the Cambrian metazoan diversification was driven by morphological and ecological radiation of primary producers including most acritarchs (Moczydlowska, 2001, 2002). Further investigation of this matter, including detailed biostratigraphic studies across complete sections of the Proterozoic-Phanerozoic transition, will help determine which of these scenarios most likely occurred.

Was this article helpful?

0 0

Post a comment