The Stability of Species in Taxonomy STOR ® Nigel C. Hughes; Conrad C. Labandeira Paleobiology, Vol. 21, No. 4 (Autumn, 1995), 401-403. Stable URL: http://links.jstor.org/sici?sici=0094-8373%28199523%2921%3A4%3C401%3ATSOSIT%3E2.0.CO%3B2-X Paleobiology is currently published by Paleontological Society. Your use of the JSTOR archive indicates your acceptance of JSTOR' s Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of ajournai or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/joumals/paleo.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is an independent not-for-profit organization dedicated to creating and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact [email protected]. http://www.j stor.org/ Mon Oct 18 09:50:13 2004 Paleobiology, 21(4), 1995, pp. 401-403 MATTERS OF THE RECORD The stability of species in taxonomy Nigel C. Hughes and Conrad C. Labandeira Nigel C. Hughes. Geier Collections and Research Center, Cincinnati Museum of Natural History, 1720 Gilbert Avenue, Cincinnati, Ohio 45202 Conrad C. Labandeira. Department of Paleobiology, National Museum of Natural History, MRC-121, Smith- sonian Institution, Washington, DC 20560 Accepted: July 8, 1995 Many questions of evolutionary pattern and dex fossils, and here we explore the history mechanism must be addressed at the species of a striking example of the instability of spe- level, because species are a fundamental unit cies definitions within the Trilobita. In each of biology. Even studies which examine pat- of these three cases it has been concluded that terns of morphologic (as opposed to taxic) the species-level taxonomy was previously diversity rely on species as a basic unit of substantially over split, and the implications analysis. Species also provide the backbone of these observations are potentially far- for both biostratigraphy and biogeography. reaching. By detailing the history of a well Consequently, the definion of individual spe- documented example below we aim to draw cies is an important and influential aspect of attention to the problem, and to suggest pos- paleontology. Most fossil species have been sible solutions. erected using brief descriptions of small num- The Late Cambrian asaphide trilobite Di- bers of specimens. The vast majority of these kelocephalus occurs in large numbers in the St. species, even those that characterize biostrati- Lawrence Formation of Wisconsin, Minne- graphic units, are poorly known. Recovery of sota, and Iowa. Since its first description in new material often necessitates re-evaluation 1852 the systematics has received substantive of species definitions as specimen numbers revision approximately once every 30 years rise and variations among specimens are as- (fig. 1). Over the same time period there has sessed more rigorously. How stable are spe- been an exponential increase in the numbers cies definitions in the light of increasing of specimens reposited in museums (fig. 1). knowledge? The result of this prolonged series of studies Answering this question will allow us to is that Dikelocephalus is one of the best-doc- estimate confidence limits on geological or umented trilobite taxa. The most recent re- paleobiological interpretations which utilize vision (Hughes 1994), based on analysis of fossil species. To address this issue we need patterns of variation within and among col- well-known species that have been subject to lections from single bedding planes, has re- intensive and sustained research effort. At sulted in synonymizing all St. Lawrence For- present the numbers of fossil taxa that have mation specimens into a single species, Di- been subject to such detailed analysis is min- kelocephalus minnesotensis. While there are imal. However, recent morphometric studies considerable morphological differences of ammonites (Hohenegger and Tatzreiter among the 2750 specimens analyzed, these 1992) and foraminifera (Tabachnick and variations are concentrated within collec- Bookstein 1990) have shown that in both these tions made from single beds, and are best cases characters once thought to define bio- interpreted as intraspecific variations. This stratigraphically significant taxa failed closer revision forms a basis from which we review scrutiny. Trilobites are commonly used as in- the taxonomic history of the taxon. © 1995 The Paleontological Society. All rights reserved. 0095-8373/95/2104-0001/$1.00 402 NIGEL C. HUGHES AND CONRAD C. LABANDEIRA 1900 1950 Year number of species number of specimens 1900 1950 2000 FIGURE 1. Numbers of species (left y axis) and numbers Year of specimens (right y axis) of trilobites now assigned to FIGURE 2. Log of ratio of average number of specimens Dikelocephalus minnesotensis, since its first description in per species now assigned to D. minnesotensis, since its first 1852. Only those papers in which new taxa were erected description in 1852. Least squares regression line shows or synonymized are shown, and include Owen (1852), the general increase in the ratio through time. Hall (1863), Walcott (1914), Ulrich and Resser (1930), Raasch (1951) and Hughes (1994). Estimates of numbers of specimens available were made either from statements in publications or accession data from U.S. museums. was able to note that variation within Dike- locephalus is continuous, and consequently he The taxonomic history of specimens now reduced the species diversity drastically. The assigned to D. minnesotensis shows marked in- eight species he recognized were defined us- stability, with a peak of 26 species erected in ing stratigraphie criteria to draw arbitrary 1930 (fig. 1). Initial slow growth in the num- boundaries within ranges of continuous vari- bers of collected specimens of D. minnesotensis ation. Once again, the desire to define zonal was succeeded by rapid growth after 1914 (fig. taxa generated a taxonomic scheme that was 1). This initial period is also characterized by based on perceived stratigraphie utility rath- slow growth in the numbers of species, with er than intrinsic patterns of biology. one "variety" described by Hall in 1863, and The discussion above explains the taxo- another species added by Walcott in 1914. The nomic history of Dikelocephalus in terms of stability in species numbers during this early two factors: (1) patterns of morphologic vari- phase may have been due in part to the small ation evident at particular sample sizes, and but relatively constant specimen numbers, (2) the historical agendas of particular sci- because proliferation of species in Ulrich and entists. What was the relative contribution of Resser's 1930 monograph was coincident with these factors? We have approached this ques- a sharp, seven-fold rise in specimen numbers tion by assessing how far particular works (fig. 1). This dramatic rise in species descrip- deviate from a general trend in the taxonomic tions was attributable to: (1) an expanded history of Dikelocephalus. We calculated the sample that presented a significantly greater ratio of specimens available to species de- variety of morphologies than had been avail- scribed for each major revision, took the log able previously, and (2) the authors' zest for of this ratio because of the exponential growth defining new species that reflected an explicit of specimen numbers, and plotted it against commitment to provide tools for high reso- time (fig. 2). The steady increase in the ratio lution biostratigraphy. Their taxonomy was of specimens to species is summarized by a also influenced by the fact that the sample regression line. Ulrich and Resser's 1930 they examined, despite containing 350 spec- monograph forms a clear outlier to this gen- imens, was too small to demonstrate the con- eral trend. The average ratio of specimens per tinuity of variation evident among larger species was 14:1 in that work, whereas Wal- numbers of specimens. By 1951 the numbers cott in 1914 had a ratio of about 25:1, and had risen sharply to about 1250 specimens, Raasch in 1951 about 160:1. Hence the spec- and it proved impossible to apply the species imen support for each of Ulrich and Resser's definitions proposed in 1930 to the new ma- species was, on average, anomalously low, and terial. With this specimen base Raasch (1951) the deviation of their study from the regres- MATTERS OF THE RECORD 403 sion line may approximate their particular of variation at low taxonomic levels, and doc- predilection for defining new species. How- ument the effect of these studies on previous ever, while Ulrich and Resser were notorious taxonomic schemes. The second is to explore taxonomic splitters the regression suggests whether particular workers had characteristic that other contemporary workers might have "taxonomic signatures." If, for example, a recognized an inflated number of species giv- constant specimen/species ratio is character- en the same sample (the regression predicts istic throughout Ulrich and Resser's work, and four or five species among 350 specimens). if this is significantly different from that of This result compliments the contention that other workers, it may be possible to compute insufficient sampling may give rise to the ap- error estimates that can be applied to their pearance of discrete morphologic character work when using their data in comprehen- states, which can be misinterpreted to have sive analyses of taxic diversity. While partic- taxonomic significance (Labandeira and ular personalities and philosophies have had Hughes 1994). This pitfall is likely most dan- marked effects on taxonomic interpretations gerous in cases where finding discontinuous of the fossil record, analysis of the historical variation is a primary object of the analysis.