Palaeontologia Electronica http://palaeo-electronica.org BRYOZOAN GENERIC EXTINCTIONS AND ORIGINATIONS DURING THE LAST ONE HUNDRED MILLION YEARS Frank K. McKinney and Paul D. Taylor Frank K. McKinney. Department of Geology, Appalachian State University, Boone, North Carolina 28608, USA. Honorary Research Fellow, Department of Palaeontology, The Natural History Museum, Cromwell Road, London, UK Paul D. Taylor. Department of Palaeontology, The Natural History Museum, Cromwell Road, London SW7 5BD, UK ABSTRACT Stage-level analysis of extinctions and originations of bryozoan genera of the orders Cyclostomata and Cheilostomata for the Upper Cretaceous and Cenozoic yields some predictable and some unexpected results. Extinction rates in cyclostomes and cheilostomes track one another except for some lower Upper Cretaceous stages in which apparently random extinctions among the small numbers of cheilostome gen- era generated ‘noise’. Both orders experienced high rates of extinction during or at the end of the Maastrichtian (our data cannot resolve position of extinctions within a stage) but, surprisingly, Danian extinction rates were essentially equal to Maastrichtian rates. High extinction rates for the Danian are attributed to loss of the ‘chalk’ sea of northern Europe, which was a centre of bryozoan diversity from early in the Late Cretaceous until its disappearance at the end of the Danian. Origination rates of cyclostomes and cheilostomes were similar and relatively high during the Late Cretaceous, but following the K-T extinction event, cyclostome origination rates dropped and remained at low levels through the Cenozoic, while cheilostome origination rates rebounded by the Eocene and then declined through the Pleistocene. The different Cenozoic diversity trajectories of cyclostomes and cheilostomes appear to derive primarily from differ- ences in origination rather than extinction rates. KEY WORDS: extinctions, originations, Cretaceous-Tertiary, K-T, benthic invertebrates Copyright: Palaeontological Association, 22 June 2001 Submission: 9 February 2001 Acceptance: 23 May 2001 INTRODUCTION mooted in 1980. Substantial information has been amassed and used to address Much attention has been focused on issues relating to the causal mechanism(s) the K-T extinction since the now famous of the extinction (e.g., Alvarez et al. 1980, hypothesis of bolide impact was first Bohor et al. 1984, Venkatesan and Dahl McKinney, Frank K. and Paul D. Taylor, 2001. Bryozoan Generic Extinctions and Originations During the Last One Hundred Million Years. Palaeontologia Electronica, vol. 4, issue 1, art. 3: 26pp., 2.4MB. http://palaeo-electronica.org/2001_1/bryozoan/issue1_01.htm FRANK K. MCKINNEY & PAUL D. TAYLOR: BRYOZOAN EXTINCTIONS AND ORIGINATIONS 1989, Hildebrand and Boynton 1990, Izett fery 1998, for echinoids). Despite these et al. 1991), as well as to patterns of problems in documentation, the complete- extinction and recovery experienced by ness of the fossil record apparently is particular groups of organisms (e.g., rather high for bryozoan genera, with an Jablonski and Raup 1995, Marshall and approximately 70 percent probability of the Ward 1996, Sheehan et al. 1996, preservation of a genus in each strati- MacLeod et al. 1997, d’Hondt et al. 1998, graphic stage throughout its range (Foote Jablonski 1998, Smith and Jeffrey 1998). and Sepkoski 1999). With regard to global With regard to the latter, biological groups patterns, the finest level of analysis that with more utilitarian fossil records (e.g., can reasonably be attempted at present planktonic foraminifera) or broad scientific for bryozoan genera is at the level of the and public appeal (e.g., dinosaurs) have stratigraphical stage/age, as done, for received the most attention, whereas rela- example, by Viskova (1980, 1997). tively little is currently known about extinc- Alternative approaches to understand- tion patterns in many other groups, ing the historical record of bryozoans have including bryozoans. looked at changes in species assemblage Bryozoans are a phylum of colonial diversity and in relative skeletal biomass metazoans with a rich marine fossil record around and across the K-T boundary (Lid- due to the readily fossilisable calcareous gard et al. 1993; McKinney et al. 1998, in skeletons possessed by most species. press). The only bed-by-bed study of bryo- Two clades of calcareous bryozoans are zoan species distributions across the K-T represented in the Cretaceous and Ceno- boundary is that of Håkansson and Thom- zoic, cyclostomes (Class Stenolaemata) sen (1999) at Nye Kløv in Denmark. In this and cheilostomes (Class Gymnolaemata). paper, we track the history of originations Both clades were diverse during the Late and extinctions of Late Cretaceous Cretaceous; however, knowledge of their through Cenozoic bryozoans at the genus systematics is woefully inadequate—many level. Therefore, we utilise primarily the species have yet to be formally described global record of temporal distribution of and others remain unrevised and are virtu- the genera rather than local species ally unrecognisable using the original assemblage diversities or skeletal biom- descriptions published in the 19th Century. ass. This lack of knowledge is compounded for The existence of two distinct clades of the cyclostomes which, in contrast with the bryozoans crossing the K-T boundary pro- skeletally more complex cheilostomes vides a replicate data-set of extinction and where a close correspondence has been origination patterns, thereby lessening demonstrated between genetically and potential problems caused by factors spe- skeletally determined species (Jackson cific to one clade. Cyclostomes date back and Cheetham 1990), have relatively sim- in the fossil record to the Lower Ordovi- ple and easily confounded morphologies cian, cheilostomes to the Upper Jurassic (McKinney and Jackson 1989; Håkansson (Taylor 1993). Each group is thought to and Thomsen 1999). Ranges of species in have acquired its mineralised skeleton time and space at all scales of resolution, independently from a bryozoan of ctenos- from single stratigraphical sections to glo- tome-grade and to share a common bally, are poorly documented, and almost ancestor no younger than Early Ordovi- nothing is known about phylogenetic rela- cian (Todd 2000). They show important tionships that might aid the interpretation differences in zooidal baupläne, which of extinction patterns (e.g., Smith and Jef- apparently are linked with the greater rate 2 FRANK K. MCKINNEY & PAUL D. TAYLOR: BRYOZOAN EXTINCTIONS AND ORIGINATIONS of suspension feeding, growth rate, and – genera were provided by J. J. Sepkoski, in tropical and temperate waters – ability Jr. from his unpublished global genus to compete for substrate space in cheilos- database. Accuracy of the database was tomes (McKinney 1993, 1995; but see checked by us, resulting in various alter- Barnes and Dick 2000 for reversal of com- ations to about 20 percent of the records, petitive interactions in high latitude largely changes in ranges but also includ- waters). Species belonging to the two ing elimination of some synonyms, addi- clades commonly co-exist and compete tion of some genera, and reassignment of with one another in the same habitats, and a few genera from cyclostomes to cheilos- a comparable range of colony forms is tomes or vice versa. This proportion of developed in the two clades. Cyclostomes errors is consistent with the findings of and cheilostomes, therefore, occupy Adrain and Westrop (2000) for trilobites in broadly similar ‘ecospace’; however, some Sepkoski’s database, and like they, we cheilostomes have invaded brackish found that the error was randomly distrib- waters and a few others have evolved uted, so that results of analyses based on free-living colonies, while cyclostomes the database did not differ greatly between have done neither. Other than these two the original and corrected versions. categories of exceptions, the two clades The analysed data comprised 808 might be expected to show parallel genera (545 cheilostomes; 263 cyclos- responses to the majority of environmental tomes), of which 24 percent are confined changes. to one stage. Origination and extinction To put our consideration of extinction metrics were calculated both with and at the K-T boundary into a broader evolu- without such stage-only genera. Error bars tionary context for the phylum and to iden- for the origination and extinction metrics tify other times of enhanced generic were calculated using the ‘Percentage turnover during the past 100 million Error Bars’ program, freeware available years—whether resulting from changes in from Steve Holland at http://www.uga.edu/ extinction or origination rates - we analy- ~strata/Software.html. The program is sed a new database of uppermost Lower based on a calculation procedure given in Cretaceous to Recent (Albian-Recent) Raup (1991). bryozoans containing genus ranges at The Gradstein and Ogg (1996) times- stage/age-level precision. The three prin- cale was used for stage boundary dates. cipal questions we set out to address ini- Numbers of appearances and disappear- tially were ances, hereby referred to as originations 1. Did generic extinction of Albian- and extinctions respectively, per strati- Recent bryozoans peak in the Maastrich- graphic interval constitute the raw data tian, corresponding with the K-T event? and must be calibrated in order to be of 2. Are temporal patterns of generic most use. The method of calibration, how- extinction and origination