Dynamics of the Ordovician Radiation: a comment on Westrop and Adrain

    An extensive body of recent research on the Ordovician Radiation has shown that the continuous, period-long transition among evolutionary faunas exhibited at the global level (Sepkoski 1981) was anything but continuous when evaluated at more local levels (see Droser et al. 1996, Miller 1997a). Local and regional transitions were characterized by abrupt changes in the diversities and abundances of faunal elements, often in association with equally dramatic changes to paleoenvironmental conditions induced by local, regional, or global phenomena (Patzkowsky and Holland 1993, 1996; Droser and Sheehan 1995; Droser et al. 1995, 1996; Miller and Mao 1995; Miller 1997b).
    While the sum of these transitions produced a synoptic decrease in the relative contribution of trilobites to global diversity and concomitant increases among elements mainly of the Paleozoic Evolutionary Fauna, regional and local shifts in trilobite diversity and abundance typically did not mirror this global average.  Indeed, where environmental conditions were favorable, it was not uncommon for trilobites, bivalve molluscs, and other elements of the Cambrian and Modern Evolutionary Faunas to dominate the biota (e.g., Bretsky 1969, 1970a,b; Babin et al. 1982; Babin and Gutirrez-Marco 1991;  Babin 1993, 1995; Frey 1987a,b; Miller 1989; Sßnchez and Waisfeld 1995; Cope 1996; Waisfeld and Sßnchez 1996).
    In this context, Westrop and Adrain (1998)  have demonstrated convincingly that  trilobites were major, stable contributors to alpha diversity, throughout the Ordovician,  among Ordovician biotic assemblages located mainly in tropical, carbonate environments in western Canada, the Canadian Arctic, and Newfoundland.  However,  we are concerned that their characterization of these biotas in the context of the Ordovician Radiation and "onshore-offshore"  patterns of faunal diversification is misleading for several reasons:
        1.  While they acknowledge that biotic patterns exhibited by the paleocontinent of Laurentia might not be indicative of broader, global patterns, Westrop and Adrain downplay this concern and imply that their documented patterns reflect trilobite diversities worldwide.  However, throughout the Ordovician, biotic compositions appear to have varied markedly from paleocontinent-to-paleocontinent (Miller 1997b, Miller and Mao in press), and no single paleocontinent reflects a global microcosm.  To be fair, it must be acknowledged that the "onshore-offshore" interpretations that Westrop and Adrain call directly into question (Sepkoski and Sheehan 1983; Sepkoski and Miller 1985) may be compromised similarly because they are based almost entirely on data from Laurentia (Mao and Miller 1994; Mao 1995; Miller and Mao in press).
         2. Given their geographic limitations within North America, Westrop and Adrain's data do not reflect regional variations throughout Laurentia.  Westrop and Adrain allude to this concern by suggesting, for example, that trilobites were affected negatively by environmental transitions in central and eastern Laurentia associated with the Taconic Orogeny.   The limited diversity of trilobites in these settings can perhaps best be illustrated with data compiled for two regions in the midcontinent affected by the onset of orogenic activity: the Cincinnati Arch and Nashville Dome.  Data from Upper Ordovician strata of the Cincinnati Arch are summarized in Table 1, compiled at the scale of  Holland and Patzkowsky's (1996) six Late Ordovician sequences (C1 through C6). These trilobite diversities are based on known reports of trilobites from the region and represent over 100 years of intensive collection by professionals and amateurs.  In addition, this region has been studied by workers with well-known reputations for taxonomic splitting, so it is unlikely that these diversities significantly underestimate true diversity.  The first number in each column reports this total diversity and is based on fieldwork, museum collections, and published literature.  Based on our field experience, many of these trilobite species are either so rare or stratigraphically restricted that they would not be encountered in most samples.  Thus, the second number (in parentheses) reflects the species diversity that would likely be encountered from each of these units in a sample of  90 trilobites, equivalent to the rarified sample sizes of Westrop and Adrain. A perusal of these data indicates that Cincinnati Arch trilobites are uniformly less diverse than nearly any of the Upper Ordovician samples reported by Westrop and Adrain.
    The situation is as extreme or even more so on the Nashville Dome.  The trilobite diversity data shown in Table 2 are based on field work (Patzkowsky and Holland unpublished data) and published literature and represent all reported species from the Middle Ordovician (M1 through M6) and Late Ordovician (C1 through C5) of the Nashville Dome.  Most of these species are sufficiently common to be encountered in large samples, so they are comparable to the rarified samples of Westrop and Adrain.  Again, these diversities are consistently at the low end of the values reported by Westrop and Adrain.
    While Westrop and Adrain would probably not quarrel with these numbers, they would argue that these patterns were endemic to portions of Laurentia overprinted by the Taconic Orogeny. However, this begs the following question: how large an area must be affected by the by-products of tectonic activity before such an overprint is viewed as more than just an exception to a larger pattern? Taken together, the Appalachian Basin, Cincinnati Arch, Nashville Dome, and other regions affected by the Taconic Orogeny and volcanism constituted a significant areal percentage of Laurentia in the Middle and Late Ordovician.  Further, the effects of orogeny and volcanism were by no means limited to eastern and central Laurentia; tectonic activity was on the rise in several venues around the world, and these areas may have been hotbeds of diversification (Miller and Mao 1995; see below).
    Moreover, it should not be assumed that the physical effects of orogeny and other tectonic activity were uniformly detrimental to trilobites.  In the Cincinnati region, for example, the primary environmental effect of the Taconic Orogeny was the contribution of fine-grained siliciclastic sediment, ubiquitous in Cincinnatian strata but manifested most dramatically in intermittent claystones known locally as "butter shales."  The two most common Cincinnatian trilobites, Flexicalymene and Isotelus,  exhibit their greatest abundances in these claystones (e.g., Frey 1987a,b).  Coupled with observations that the global diversity histories of orders constituting the class were rather disparate from one another (cf. Adrain 1996), this serves as a reminder that any attempt to characterize broadly the paleontology of an entire class is bound to overlook markedly different attributes exhibited by many of its constituents.
    In the Cincinnati and Nashville regions, orogeny-induced Middle Ordovician changes in paleoceanographic conditions, including upwelling and associated cooling of surface waters, abated during the Late Ordovician.  This was reflected in the return to the area of certain articulate brachiopods, corals, and molluscs that favored tropical, carbonate settings (Patzkowsky and Holland 1993, 1996; Holland 1997).   Although several trilobite genera first appeared on the Cincinnati Arch during this C4-C5 invasion, no net change in trilobite diversity occurred during this interval on either the Cincinnati Arch or the Nashville Dome (Tables 1 and 2).  Furthermore, no change in trilobite diversity was apparent on the Nashville Dome at the onset of the Taconic Orogeny in the M4-M5 interval (Table 2); trilobite diversity was uniformly low in the entire region throughout the Middle and Late Ordovician.  Thus, in the Nashville area, Westrop and Adrain's suggested linkage of trilobite decline to these paleoceanographic effects is questionable.
        3. While they ascribe the precipitous Lower-to-Middle Ordovician decline in trilobite abundance in the Great Basin as manifested in shell beds (Droser et al. 1996) to taphonomic effects, Westrop and Adrain overlook the fact that many trilobite-rich biotas are contained in storm-deposited strata in the Great Basin and elsewhere.  Indeed, trilobite shell beds are common components of Middle Ordovician strata in the Great Basin.  However, at the base of the Whiterock, the absolute number of trilobite-dominated shell beds decreases dramatically and the number of polytaxic shell beds that include trilobites remains constant, even though the total number of shell beds increases, augmented by the addition of shell beds dominated by articulate brachiopods.  Moreover, there is no evidence of a taphonomic transition associated with passage from Lower to Middle Ordovician strata (Li and Droser unpublished data).  In fact, Westrop (1986) demonstrated that trilobite material contained in Upper Cambrian storm-deposits of southern Alberta was sufficiently robust to be preserved in great abundance, despite being physically sorted and transported after death.  Thus, it is difficult to accept the suggestion that the limited abundances of trilobites in Middle Ordovician strata of Great Basin localities resulted mainly from taphonomic effects, especially given that older, storm-deposited strata in these venues frequently contain abundant trilobites.
    Droser et al. (1996) pointed out that trilobite diversity was far from depleted in Middle Ordovician strata of the Great Basin and that the loss of Lower Ordovician trilobite taxa was offset by a substantial Middle Ordovician radiation of trilobites,  including members of higher taxa not seen in older strata anywhere in Laurentia (Adrain 1996; Fortey and Droser 1996).  Thus, the decline in trilobite abundance may have been associated with a decline in the previously incumbent trilobite fauna.  In the Great Basin, the base of the Middle Ordovician marks a significant shift among trilobite clades and corresponds with the initial diversification of a new trilobite fauna.
        4.  Because their analyses were limited to trilobites, it is difficult to know whether the study localities used by Westrop and Adrain were venues in which the Ordovician Radiation was ongoing actively.   If the authors wish to show that trilobite diversity and abundance persisted in the face of the radiation of other faunal elements, it is not sufficient to show that trilobite diversity was maintained in the Middle and Upper Ordovician settings that they investigated.  They must also demonstrate directly the dilution effect that they claim: that in these same strata, the diversity of other faunal elements, including members of the Paleozoic Fauna, was increasing.  In an earlier paper that focused on nearshore environments, Westrop et al. (1995) provided evidence that other taxa (primarily gastropods, rostroconchs, and bryozoans) diversified through an interval in the Cambro-Ordovician during which trilobite diversity remained relatively constant. This kind of analysis was not extended by Westrop and Adrain to the suite of samples and environments included in the present study.   Thus, can they be confident that they have not simply isolated a number of venues in which trilobites continued to persist, while declining elsewhere in locations where the Radiation was going full tilt?  Most of Westrop and Adrain's data come from tropical, carbonate-dominated paleoenvironments in a limited geographic portion of Laurentia.  However, as suggested by Miller and Mao (1995, in press) and Miller (1997a,b), several major Ordovician taxa diversified most appreciably in settings in which terrigenous sedimentation occurred with at least some regularity.
    In most respects, the approach used by Westrop and Adrain is laudable: they provided a synthetic treatment of highly resolved data emanating largely from their own fieldwork and systematic investigations.  If their efforts were repeated for other taxa and in other major venues around the world, the result would almost certainly be a more highly resolved sense of the spatio-temporal texture of Ordovician biodiversity.  Moreover, we wish to emphasize that we are not claiming a role for competitive interaction in the spatio-temporal history of trilobites; if anything, our collective experience points to the importance of physical mechanisms in producing these patterns.   It is simply our view that Westrop and Adrain overstepped their data in making claims about the nature of the Ordovician Radiation that were unwarranted, given the limited geographic and taxonomic coverage of their investigation.

Acknowledgements
We thank the National Aeronautic and Space Administration (Exobiology Grant NAGW-3307 to A. I. M.), the National ScienceFoundation (Grants EAR-9204916 to A. I. M.; EAR-9204445 and EAR-9705732 to S. M. H.; EAR-9219731 and EAR-9706021to M. L. D.; EAR-97-05829 to M. E. P.), the National Geographic Society (M. L. D.), and the Petroleum Research Fund of the American Chemical Society (S. M. H. and A. I. M.)  for supporting our research on Ordovician paleobiology.

Literature Cited
Adrain, J. M.  1996.  Measures of diversity, the Paleozoic evolutionary fauna, and the  reinvention of trilobites.  Geological
    Society of America Abstracts with Programs 28(7): A291.
Babin, C.  1993.  Rle des plates-formes gondwaniennes dans les diversifications des mollusques  bivalves durant l'Ordovicien.
    Bulletin Socit gologique de France 164:141-153.
---.  1995.  The initial Ordovician bivalve mollusc radiations on the western Gondwanan  shelves. Pp. 491-498  in Cooper et
    al. 1995.
Babin, C.,  and J.-C.  Gutirrez-Marco.  1991.  Middle Ordovician bivalves from Spain and their  phyletic and
    palaeogeographic significance.  Palaeontology 34:109-147.
Babin, C., R. Courtessole, M. Melou, J. Pillet, D. Vizcano, and E. L. Yochelson.  1982.  Brachiopodes (Articuls) et
    mollusques (bivalves, rostroconches, monoplacophores,  gastropodes) de l'Ordovicien infrieur (Trmadocien-Arenigien) de la
    Montagne Noire  (France mridionale).  Mmoires de la Socit d' tudes scientifiques de l' Aude, Carcassonne.
Bretsky, P. W.  1969.  Central Appalachian Late Ordovician communities.  Geological Society  of America Bulletin
    80:193-202.
---.  1970a.  Upper Ordovician ecology of the central Appalachians.  Peabody Museum of Natural History Bulletin 34.
    150 pp.
---.  1970b.  Late Ordovician benthic communities in north-central New York.  New York State Museum and Science
    Service Bulletin 414.  34 pp.
Cooper, J. D., M. L. Droser, and S. C.  Finney, eds. 1995. Ordovician odyssey: short papers for the Seventh International
    Symposium on the Ordovician System.  Pacific Section, SEPM, Fullerton, Calif.
Cope, J. C. W. 1996. Early Ordovician (Arenig) bivalves from the Llangynog Inlier, South  Wales.  Palaeontology
    39:979-1025.
Droser, M. L., and P. M. Sheehan.  1995.  Paleoecology of the Ordovician Radiation and the Late  Ordovician extinction
    event.  Pp. 64-106 in J. Cooper, ed.  Ordovician of the Great Basin:  fieldtrip guidebook and volume for the Seventh
    International Symposium on the  Ordovician System. Pacific Section, SEPM, Fullerton,  Calif.
Droser, M. L.,  P. M. Sheehan, R. A. Fortey, and X. Li.  1995.  Pp. 405-408 in Cooper et al. 1995.
Droser,  M. L., R. A. Fortey, and X. Li.  1996.  The Ordovician Radiation.  American  Scientist 84:122-132.
Fortey, R. A., and M. L. Droser.  1996.  Trilobites at the base of the Middle Ordovician, western United States.  Journal of
    Paleontology 70:73-99

Arnold I. Miller. Department of Geology, Post Office Box 210013, University of Cincinnati, Cincinnati, Ohio  45221-0013

Steven M. Holland. Department of Geology, University of Georgia, Athens, Georgia  30602-2501

Mary L. Droser. Department of Earth Sciences, University of California, Riverside, California  92521

Mark E. Patzkowsky. Department of Geosciences, Pennsylvania State University, University Park, Pennsylvania 16802-2714