Mesozoic mammaliaform diversity: The effect of sampling corrections on reconstructions of evolutionary dynamics

https://doi.org/10.1016/j.palaeo.2014.07.017Get rights and content

Highlights

  • We quantitatively examine Mesozoic mammaliaform taxonomic diversity.

  • Sampling biases influence observed diversity in many Mesozoic intervals.

  • We accounted for biases using subsampling and modelling approaches.

  • Sampling-corrected diversity shows peaks in the Late Jurassic and Latest Cretaceous.

  • But results also suggest that uneven spatial sampling drives many apparent patterns.

Abstract

Recent years have witnessed an explosion of new fossil discoveries and analyses documenting the unappreciated ecological and morphological diversity of Mesozoic Mammaliaformes. In contrast, the taxonomic diversity dynamics through the first 165 million years of mammal evolution have not yet been rigorously analysed, leaving patterns of diversification during this important period open to conjecture. Here, we present a comprehensive statistical analysis of global mammaliaform diversity spanning from the Late Triassic appearance of mammaliaforms (~ 230 million years ago [hereafter, mya]) to the end Cretaceous mass extinction (66 mya). We analysed 691 occurrences representing 367 genera and 550 species in standard time bins of approximately 10 million years in duration. Significant correlations between diversity and sampling proxies suggest sampling biases in the mammaliaform fossil record. Shareholder quorum subsampling and model-based approaches were used to mediate these biases. After applying these methods, the following patterns were supported: low standing diversity during the Late Triassic–Early Jurassic evolution of early Mammaliaformes (e.g., morganucodonts) was followed by high standing diversity during the Late Jurassic due primarily to the diversification of Eutriconodonta, Multituberculata and Cladotheria. This peak was followed by a fall in diversity during the middle of the Hauterivian–Barremian interval, suggesting that extinctions typically associated with the Jurassic–Cretaceous boundary may instead have occurred later, during the Early Cretaceous. Standing diversity recovered through several fluctuations during the ‘mid’ Cretaceous (approximately Barremian to Albian), leading to a second peak in the Campanian that reflects the diversification of key clades, including therians. Analyses of geographically restricted datasets illustrate a significant spatial heterogeneity in sampling, with several intervals dominated by North American occurrences. Uneven sampling effort and geographic heterogeneities in the fossil record are significant factors affecting reconstructions of Mesozoic mammaliaform diversity, and correcting these biases can markedly alter observed patterns and their interpretation.

Introduction

For decades, the focus of synapsid palaeontology has been on the therapsid response and recovery to the end Permian mass extinction, and the mammalian radiation following the K/Pg mass extinction, which heralded the “Age of Mammals”. However, the fossil record of Mesozoic mammaliaforms spans ~ 2.5 times the duration of the comparatively well-studied record of Cenozoic mammals. Within Mammaliaformes, 11 major clades or functional grades (following Kielan-Jaworowska et al., 2004, and detailed below) formed an ecologically diverse Mesozoic assemblage from the Carnian (Late Triassic, ca 235–229 mya) onwards (Luo, 2007a). Recent fossil discoveries highlight a complex evolutionary history for Mesozoic Mammaliaformes (e.g. Luo et al., 2011), with the traditional scenario of a linear acquisition of mammalian characters being challenged by multiple evolutionary origins of key morphological features such as the tribosphenic molar (Luo et al., 2001) and middle ear ossicles (Luo et al., 2011). Moreover, in contrast to common depictions of early mammaliaforms as small terrestrial and scansorial insectivores, new fossils demonstrate that Mesozoic mammaliaforms invaded a variety of ecological niches, from semi-aquatic to gliding forms, and even dog-sized forms that preyed on juvenile dinosaurs (Hu et al., 2005, Luo and Wible, 2005, Ji et al., 2006, Meng et al., 2006, Luo, 2007a). Despite the great attention paid in recent years to this previously unappreciated morphological and ecological diversity of early mammaliaforms, and a series of recent quantitative studies of taxonomic diversity in more basal synapsids (Brocklehurst and Fröbisch, 2014, Brocklehurst et al., 2013, Fröbisch, 2013), there has been little rigorous analysis of mammaliaform diversity dynamics prior to the K/Pg mass extinction (Rose, 2006). Previous approaches have been either broad and qualitative assessments of subclades (Luo, 2007b) or geographically restricted to the North American record (Alroy, 2009) and more specific localities (Wilson, 2005, Wilson, 2013).

Reconstructing diversity dynamics over deep time is a core theme of palaeobiology (Jablonski, 1999, Raup, 1972, Valentine, 1985). Although the potential effects of geological and anthropogenic biases on accurate taxon counts have been discussed for decades (Raup et al., 1975), it is only more recently that substantial efforts have been made to correct these biases (Alroy, 2000, Alroy, 2008, Alroy, 2010, Alroy et al., 2001, Alroy et al., 2008, Behrensmeyer et al., 2005, Peters and Foote, 2001, Smith and McGowan, 2007, Smith et al., 2012). A growing number of studies have focussed in particular on biases introduced by differences in outcropping rock area (Crampton et al., 2003, Smith and McGowan, 2007), preservation potential of fossil organisms (Smith, 2001), or evenness and fairness of sampling during standard intervals (Alroy, 2010, Alroy et al., 2001, Alroy et al., 2008). These studies suggest that many features of observed diversity curves could be artefacts of changes in fossil preservation, geological sampling, or anthropogenic sampling rather than true biotic signals (e.g. Smith, 2007, Smith et al., 2012). Complex Earth system interactions such as sea level change may drive both sedimentation and ancient biodiversity in the marine realm, suggesting that covariation of fossil taxon counts and potentially biasing factors is not always causal (Peters, 2005, Benson and Butler, 2011, Hannisdal and Peters, 2011; but see Smith and Benson, 2013). However, terrestrial processes may be simpler, with factors such as rock area and collection effort directly biasing taxon counts (e.g. Benson and Upchurch, 2013, Benson et al., 2013, Butler et al., 2011a, Butler et al., 2011b, Upchurch et al., 2011). Here, we present the first quantitative investigation of the global taxonomic palaeodiversity of Mesozoic Mammaliaformes, applying robust sampling-correction approaches to account for geological and anthropogenic biases and reassessing diversity dynamics in early mammal evolution.

Section snippets

Mammaliaform taxa

We have attempted to maximise coverage of Mesozoic mammaliaform occurrence data in the Palaeobiology Database (Alroy et al., 1998), with an extensive literature review and comparison with data in Kielan-Jaworowska et al. (2004). Mammaliaformes was considered as a monophyletic clade, consisting of all descendants of the most recent common ancestor of Morganucodonta and crown Mammalia (Luo et al., 2002, Rowe, 1988, Zhou et al., 2013). Morganucodonts, docodonts and kuehneotherids are successively

Observed Mesozoic mammaliaform taxonomic diversity

We first calculated observed in-bin species counts (species taxonomic diversity estimate; STDE) and generic counts (generic taxonomic diversity estimate; GTDE) (Table 1). The uncorrected mammaliform fossil record displays an apparent long-term increase in diversity through the Mesozoic, punctuated by four peaks occurring in the Late Triassic (Triassic 4), Late Jurassic (Jurassic 6), early Late Cretaceous (Cretaceous 3) and Late Cretaceous (Cretaceous 6–7) (Fig. 1a). These peaks are separated by

Sampling bias in the mammaliaform fossil record

The significant correlations between observed mammaliaform diversity and multiple fossil sampling proxies suggest that some patterns in apparent diversity may be explained by temporal variation in rock record quality rather than by evolutionary dynamics. As a result, caution is needed when interpreting uncorrected data. Both sets of analyses described here do support discrete series of statistically robust shifts in Mesozoic mammaliaform diversity. However, it is important to note that these

Acknowledgements

We acknowledge the substantial work by contributors of the Paleobiology Database (http://paleodb.org), especially John Alroy and Matthew Carrano. We thank John Alroy for his considerable guidance when initially running the SQS analyses. This work was supported by a Leverhulme Research Project Grant (RPG-129) to PU and AG. Finally, we thank all those who reviewed our original manuscript. This paper is Palaeobiology database publication number #204.

References (85)

  • J. Alroy

    Geographical, environmental and intrinsic biotic controls on Phanerozoic marine diversification

    Palaeontology

    (2010)
  • J. Alroy et al.

    The palaeobiology database

    (1998)
  • J. Alroy et al.

    Effects of sampling standardization on estimates of Phanerozoic marine diversification

    Proc. Natl. Acad. Sci. U. S. A.

    (2001)
  • J. Alroy et al.

    Phanerozoic trends in the global diversity of marine invertebrates

    Science

    (2008)
  • J.D. Archibald

    Sedimentology, stratigraphy, and extinctions during the Cretaceous–Paleogene transition at Bug-Creek, Montana — comment

    Geology

    (1986)
  • J.D. Archibald et al.

    Cretaceous extinctions: multiple causes

    Science

    (2010)
  • P.M. Barrett et al.

    Dinosaur diversity and the rock record

    Proc. R. Soc. B Biol. Sci.

    (2009)
  • A.K. Behrensmeyer et al.

    Are the most durable shelly taxa also the most common in the marine fossil record?

    Paleobiology

    (2005)
  • Y. Benjamini et al.

    Controlling the false discovery rate — a practical and powerful approach to multiple testing

    J. R. Stat. Soc. Ser. B Methodol.

    (1995)
  • R.B.J. Benson et al.

    Uncovering the diversification history of marine tetrapods: ecology influences the effect of geological sampling biases

    Geol. Soc. Lond., Spec. Publ.

    (2011)
  • R.B.J. Benson et al.

    Faunal turnover of marine tetrapods during the Jurassic–Cretaceous transition

    Biol. Rev.

    (2014)
  • R.B.J. Benson et al.

    Diversity trends in the establishment of terrestrial vertebrate ecosystems: interactions between spatial and temporal sampling biases

    Geology

    (2013)
  • M.J. Benton et al.

    Assessing the quality of the fossil record: insights from vertebrates

    Geol. Soc. Lond., Spec. Publ.

    (2011)
  • N. Brocklehurst et al.

    The completeness of the fossil record of Mesozoic birds: implications for early avian evolution

    Plos One

    (2012)
  • N. Brocklehurst et al.

    The early evolution of synapsids, and the influence on their fossil record

    Paleobiology

    (2013)
  • S.L. Brusatte et al.

    Dinosaur morphological diversity and the end-Cretaceous extinction

    Nat. Commun.

    (2012)
  • R.J. Butler et al.

    Diversity patterns amongst herbivorous dinosaurs and plants during the Cretaceous: implications for hypotheses of dinosaur/angiosperm co-evolution

    J. Evol. Biol.

    (2009)
  • R. Butler et al.

    The taxonomic diversity and morphological disparity of pterosaurs: untangling sampling biases, the impact of Lagerstatten, and diversification trajectories

    J. Vertebr. Paleontol.

    (2011)
  • R.J. Butler et al.

    Sea level, dinosaur diversity and sampling biases: investigating the ‘common cause’ hypothesis in the terrestrial realm

    Proc. R. Soc. B Biol. Sci.

    (2011)
  • J.S. Crampton et al.

    Estimating the rock volume bias in paleobiodiversity studies

    Science

    (2003)
  • M. dos Reis et al.

    Phylogenomic datasets provide both precision and accuracy in estimating the timescale of placental mammal phylogeny

    Proc. R. Soc. B Biol. Sci.

    (2012)
  • E. Eizirik et al.

    Molecular dating and biogeography of the early placental mammal radiation

    J. Hered.

    (2001)
  • D.E. Fastovsky et al.

    The extinction of the dinosaurs in North America

    GSA Today

    (2005)
  • D.E. Fastovsky et al.

    Shape of Mesozoic dinosaur richness

    Geology

    (2004)
  • M. Foote et al.

    Evolutionary and preservational constraints on origins of biologic groups: divergence times of eutherian mammals

    Science

    (1999)
  • E.F. Freeman

    a Middle Jurassic mammal bed from Oxfordshire

    Palaeontology

    (1979)
  • J. Fröbisch

    Global taxonomic diversity of anomodonts (Tetrapoda, Therapsida) and the terrestrial rock record across the Permian–Triassic boundary

    Plos One

    (2008)
  • I.J. Good

    The population frequencies of species and the estimation of population parameters

    Biometrika

    (1953)
  • A. Goswami

    A dating success story: genomes and fossils converge on placental mammal origins

    Evodevo

    (2012)
  • A. Goswami et al.

    A radiation of arboreal basal eutherian mammals beginning in the Late Cretaceous of India

    Proc. Natl. Acad. Sci. U. S. A.

    (2011)
  • D.M. Grossnickle et al.

    Mammal disparity decreases during the Cretaceous angiosperm radiation

    Proc. R. Soc. B

    (2013)
  • Ø. Hammer et al.
  • Cited by (30)

    • The Silurian–Devonian terrestrial revolution: Diversity patterns and sampling bias of the vascular plant macrofossil record

      2022, Earth-Science Reviews
      Citation Excerpt :

      P-values lower than 0.05 were considered as significant. To correct diversity estimates, we applied Lloyd’s (2012) implementation of the Smith and McGowan (2007) method, which has been widely used for evaluating sampling impact on apparent diversity (e.g., Mannion et al., 2012; Smith et al., 2012; Brocklehurst et al., 2013; Fröbisch, 2013; Walter and Fröbisch, 2013; Newham et al., 2014; Minter et al., 2017). The Smith and McGowan (2007) method was developed to remove potential bias caused by variations in rock-availability for sampling per time interval, by calculating modelled (or predicted) diversity estimates assuming a linear relationship between apparent diversity and a correction measure.

    • The largest thalattosuchian (Crocodylomorpha) supports teleosaurid survival across the Jurassic-Cretaceous boundary

      2016, Cretaceous Research
      Citation Excerpt :

      The Jurassic-Cretaceous (J-K) transition has been considered a complex phase of global extinctions in both terrestrial and marine faunas, which affected rates of lineage diversification and morphological evolution during the Early Cretaceous (Bakker, 1978, 1998; Sepkoski, 1984; Bardet, 1994; Benton, 2001; Upchurch and Barrett, 2005; Lu et al., 2006; Benson et al., 2010). Whether this event was real (i.e., a complex combination of clade-specific extinction patterns driven by physical and biotic factors) or represents an artefact remains unresolved (Gasparini et al., 2004; Bambach, 2006; Benson et al., 2010; Ruban, 2012; Newham et al., 2014). Among speciose clades of Jurassic marine reptiles, teleosauroid crocodylomorphs stand as the sole that supposedly went extinct at the Jurassic-Cretaceous boundary (Young et al., 2014a), with all purported Cretaceous remains re-interpreted as belonging to other reptilian clades, in particular, to the other thalattosuchian clade, Metriorhynchoidea (Young et al., 2014a,b).

    • End-Triassic nonmarine biotic events

      2015, Journal of Palaeogeography
      Citation Excerpt :

      Trithelodontids cross the TJB, and a problematic group of tooth-based taxa, the “dromatheriids,” disappears in the Rhaetian (Lucas and Hunt, 1994). Note, though, that the origin and diversification of mammaliaforms, which likely include the “dromatheriids,” began in the Late Triassic (e.g., Newham et al., 2014). On present evidence, a TJB synapsid extinction is thus of a handful of taxa, with most taxa extinct by Rhaetian time, but others crossing the TJB.

    • Evidence for a mid-Jurassic adaptive radiation in mammals

      2015, Current Biology
      Citation Excerpt :

      This pattern is consistent with the “short-fuse” model of Middle Jurassic diversification proposed by Luo [2, 3] (usage of the term referring to a combination of ecomorphological and taxonomic diversification) and closely mirrors the bursts in rates associated with major shifts in ecomorphology observed in post-Palaeozoic echinoids [23], suggesting that such bursts may be commonly associated with the exploration of ecospace in novel adaptive zones. Morphological disparity, based on discrete anatomical characters, also peaked in the Middle Jurassic, coinciding with a peak in phylogenetic lineage diversity, but preceding the Late Jurassic peaks in mandibular/dental disparity [5] and subsampled taxonomic diversity [4] estimated from the fossil record. This disjunction is most likely due to the temporal extension of ghost ranges for the diverse mid- to Late Jurassic fauna by our stratigraphic calibration methods.

    View all citing articles on Scopus
    View full text