Bollettino della Società Paleontologica Italiana Modena, Novembre 1999

Neognathodus and the species concept in conodont paleontology

Glen K. MERRJLL Deparrment ofNatural Sciences, University ofHouston- Downtown

KEYWORDS- Neognarhodus, Conodonts, Evolution, Morphotypes, Biostratigraphy.

ABSTRACT- Evolution has two products; speciation and phylomorphogeny. True biologie species are genetically defined whereas morphotypes are produced by phylomorphogenetic change and have the primary utility in biostratigraphy. Confusion between morphotypes and species is commonplace, does nothing to enhance biostratigraphy, indeed may impede it, at the same time confusing the taxonomy and imparting an undeserved taxonomic legitimacy to the morphotypes. Resolution ofthe paleontologic problem of recognizing species when only morphologies are available is di{fìcult, but can be more closely approximated than is common practice. Neognarhodus, with its rapidly evolving sui te ofmorphotypes is an idea[ taxon to which to apply this eone ept for revised speciation based upon distributions ofmorphotypes. Althoufé increased numbers enhance statistica! probabilities, redefining species in this manner base d on large collections permits reasonably conjident identijications ofspecies and determinations ofgeologie age in small to very small collections.

RJASS UNTO- [Neognathodus e il concetto di specie in paleontologia] -L'evoluzione porta a due effetti: la !J.eciazione e la filomorfogenesi. Le specie biologiche sono definite geneticamente, a differenza dei morfotipi che sono prodotti dal cambiamento jilomorfogenetico e rivestono un ruolo primario in biostratigrafia. E comune fare confosione tra morfotipi e specie e ciò non porta alcun vantaggio in biostratigrafia, anzi potrebbe creare problemi, allo stesso modo in cui complica la tassonomia, dando una l.egittimazione immeritata ai morfotipi. Risolvere ilprobkma paleontologico del riconoscimento di una specie quando si ha a che fare solo con morfotipi è difficile, tuttavia è possibile avvicinarsi alla soluzione più di quanto non avvenga nella pratica comune. Neognathodus è un taxon ideale per applicare questo "nuovo" concetto di speciazione basata sulla distribuzione dei morfotipi, a causa della sua sequenza di morfotipi a rapida evoluzione. Per quanto numeri alti favoriscano la probabilità statistica, la ridefinizione della specie in questo modo basandosi su collezioni abbondanti permette ai ottenere specie attendibili e di determinare l'età di collezioni anche molto piccole.

INTRODUCTION convey the useful information we wish them to. This problem is aggravateci by our desire, worthwhile and "I must conclude, therefore, that the most important indeed noble, to use organic evolution as a measuring rule ofpaleontologic nomenclature is no t the law ojpriority, stick against which to calibrate Earth History. Such as isso often stated, but the Red Queen's Rule that: 'When calibration, biostratigraphy, the effort to use fossils to I use a word, it means exactly what I intend it to mean - establish temporal relationships within and between no more and no lessi" (Shaw, 1969, p. 1094). rocks, is an important task for science. Far too often, These words were uttered as part of Alan B. Shaw's this laudable goal has been allowed to be the driving Presidential Address to rhe Paleontological Society: force behind our systematics; an effort to fìnd Adam and Eve, Paleontology, and the non-objective arts. differences attributable to the passage of rime has forced O ne can safely assume that much of what he had to taxonomic enti ti es upon us that are really divorced from say his audience really did not want to hear for he biologie meaning. We have all been guilty of this to outlined how species concepts, however well and some degree and this so-called "nuts and bolts" honesdy proposed, have a serious failing in their ability approach to paleobiology has usually failed for the to actually transmit rhe useful information they are biostratigraphy it was intended to serve while intended to convey. Shaw further posited that a more simultaneously being devoid ofbiologic reality. lt is an useful procedure would involve some numerica! or approach, to quote another line from Shaw, albeit on other encoding system to transmit information about another topic, "so widely condemned and so faithfully morphologic changes through rime produced by practiced.' (Shaw, 1964, p. 51). evolution. He ended his address with the stirring rallying cry, "Help stamp out species!" (p. l 098). Most of all, rhis certainly was not the message an audience THE NATURE OF SPECIES of sysrematic paleontologists wanted to hear, despite the fact that Homo sapiens has been most effective at In spire of Shaw's very severe reservations, species doing just that, especially this century. exist - rhey are the fundamental operational units in Shaw's principle message and his main points biology as well as taxonomy an d their genetic definiti o n contain much truth; species concepts frequendy do no t has remained fìxed, stable, and generally accepted since 466 G.K MERRILL the work of Ernst Mayr (1942). That a species is a organisms show little or no morphologic change potentially interbreeding population of organisms essentially throughout the period. Por example, capable of producing fertile offspring is well accepted. Sturgeon & Hoare (1968) illustrate ranges of seven Phrased in more modern jargon i t might be stated that brachiopod "species" such as Composita subtilita whose speciation in eukaryotic organisms reflects sufficient ranges can reasonably be interpreted as from 22 to 35 separation of their DNA in two populations that their million years in the Pennsylvanian of Ohio and some members cannot transmit their combined DNA are interpreted with stili longer ranges elsewhere. through a succession of generations. Those definitions Ranges of severa! other Pennsylvanian brachiopods are fit with modern eukaryotes and have probably been nearly as long and other groups of invertebrates record true since eukaryotes first appeared. The reality that similarly long ranges. These 'observations" are at odds species exist and have existed cannot be avoided simply with a series of estimates of the life spans of species because of inconvenience in applying the concepts of summarized by Valentine (1970) as from roughly five paleontologists. Perhaps someone will devise a method to ten million years. These estimates are always open to convey morphologic information much in the to question and may, in fact, be too high, but it manner that Shaw envisioned, but that has not emphasizes that we should no t expect individuai species happened and quantified or unquantified morphology to range throughout a period. We should expect that must be used instead - both for systematics and speciation really was a continuing process, at least biostratigraphy. As far as Mayr's definiti o n is concerned, sympatrically, but that there was little or no perceptible the genetic testing by interbreeding of the vas t majority morphologic change. Lineages that show rapid of Earth's severa! million extant eukaryotic species has evolutionary changes that are easily recognized, never been attempted in spite of their being recognized probably underwent both speciation and phylo- as "species," and extinct species are totally untestable morphogenetic changes through allopatry. Lineages (barring the unlikely circumstance that fossils of extinct exhibiting little change in morphology could still have species will yield suffìcient DNA to reconstruct their had speciation events involving isolate gene pools, but entire genomes). Thus neontologic species are primarily o n es retaining generalized an d successful morphologies. probabilistic constructs and paleontologic ones are Long-term genetic shift to satisfy the definition of entirely so. The best we can ever hope to achieve is the speciation within sympatric populations remains a establishment of what was probably an extinct possibility, of course, because no gene pool will remain interbreeding population. How well or how poorly we constant through time. Nevertheless, it seems likely do that is based on many factors, but clearly individuai that a fair proportion of the debate about punctuated taxonomic philosophies play a key role. equilibria versus gradualism is really a matter of differing Evolution has two products that are commonly emphasis upon speciation and morphologic change confused as if they were one (Turner, 1986, p. 203). through time (Turner, 1986). Thus speciation maywell Evolution produces species, which simply involves be allopatric and the resulting speciation punctuated, genetic isolation. This can follow a variety of patterns; but the differing pace of phylomorphogeny and its bifurcation, cladogenesis, anagenesis (phyletic rather tenuous link to speciation permits it to appear evolution), etc. (Wagner & Erwin, 1995). It also sympatric and graduai and for practical purposes, i t is. produces phylomorphogenetic change through time. Speciation is discrete, usually punctuated, generally Phylomorphogeny can be defined as the inherited allopatric, cryptic, and irreversible (Eldredge & Gould, change in form through time within a genetically 1972). Phylomorphogeny is frequently continuous, related lineage. These changes may closely reflect the generally graduai, may be or appear to be sympatric, genetic change or not and assumption that they are obvious, and reversible (Boucot, 1978). parallel is not justified in some cases (Stern, 1949). Although both genetic and morphological changes are products of evolution i t is only the second that we can MORPHOLOGY INTIME AND SPACE directly observe an d a very large part of the uncertainty in taxonomic reconstructions stems from the uncertain Morphologic variation within lineages stems from correlation between the two. Indeed the best that we a variety of sources. Some certainly represents nothing can hope to do is to predict that when sufficient more than simple genetic variation within a population morphologic change has taken place that speciation whether involving mutation or not. Within three- probably will have occurred as well. We must dimensional space we can add possible paleobio- nevertheless keep in mind that these two products of geographic differences (probably varieties in the evolution occurred at different rates in most cases. taxonomic sense) an d ecophenotypic variations. Qui te There have been times in Earth History when obviously, this complexity is compounded by changes phylomorphogenetic change seems to have occurred in populations through time. Unlike speciation with especially slowly and o ne of these was the Pennsylvanian its discrete units, phylomorphogeny commonly where, for whatever reason(s), many invertebrate produces a continuous range ofvariables. Dividing such NEOGNATHODUS AND THE SPECIES CONCEPT IN CONODONT PALEONTOLOGY 467

a morphologic continuum is commonly an arbitrary more reliable than others. Because we are dividing a practice, especially so as differing parts of the set of continuous variables into classes (forcing them morphology may not be changing in anything like a to become discrete variables) this becomes a nominai regular, unidirectional manner. Some aspects of the scale (Sachs, 1984, p.132). With such count data, the morphology may show directional change, however, mode, especially when modified by secondary mode(s) and it is these that offer the prospect of utility in or next most abundant morphotype(s), is probably a biostratigraphy. At a minimum to be useful, however, more reliable tool than is the mean and standard some differentiation must be made of parts of the deviation (McClave et al., 1998, p.58). Through time continuum. This frequently is done by designating these modes should shift from dominance by one morphotypes although application of morphotypes is morphotype to the next, and the distribution of the generally an artificial attempt to divide the continuous other morphotypes should reflect this shift. The variables into discrete variables. Such a compromise product commonly approximates a classi c "bell" curve. need not have adverse effects - no more serious than Ideally, we would like to see such curves that are highly dividing the continuum that is Geologie Time into leptokurtic as this should suggest a distribution of arbitrary segments that we cali eras, periods, and epochs, morphotypes that is clustered tightly about the for example- but only as long as we recognize that this dominant one, simplifying the recognition of species, is an approximation of variables that are truly an d enhancing the likelihood of precise biostratigraphic continuous. correlation. Such increased confidence levels are not Morphotypes, produced by random genetic, always achieved, however, because the transition geographic, ecologie, or phylomorphogenetic changes between species would logically in valve more morpho- are not species, but are often confused with them, types and produce curves that are more platykurtic. especially if they are highly idiomorphic and no t clearly Biologie populations are usually "noisy," in a statistica! continuous (Futuyama, 1986). sense and this "noise" may also increase platykurtosis. As is generally recognized, a species may consist of Skewness is helpful in determining from which part of severa! discrete or continuous, but generally definable the range of a chronospecies a particular population and identifiable, morphotypes. What is less commonly was derived and thereby reflect sometimes-subtle recognized is that, because of differing rates of differences in geologie ages. speciation and phylomorphogeny (Schopf et al., 1975), For Neognathodus large collections (approximately i t is extremely likely that indistinguishable morpholo- 11,000 specimens) were used to define "species" (now gies, i. e., morphotypes, may be shared among at least a morphotypes) from rocks of the Appalachians and pair of ancestor-descendant species. Where this is true, Illinois Basin spanning approximately 15 million years one consequence could be that an individuai specimen of geologie time (Merrill & King, 1971; Merrill, 1972, could not be assigned with confidence to a particular 1975). Several thousand specimens have been added species or that a species cannot be positively identified subsequendy and used to test the originai criteria and Jrom a single individua! specimen. This is in particular as biostratigraphic tools in intra- and inter-basinal contrast with the cherislied, but flawed, "index fossil correlations. The challenge has been to both recognize concept" (Shaw, 1964). Consequendy, finding of a species and interpret geologie age from smaller and particular morphotype may be of greater or lesser smaller samples. As will be shown, this can be done biostratigraphic utility, but species should be with a high probability of success for the former an d a t probabilistic reconstructions based upon identifications least some reasonable level of success for rhe latter. of populations of fossil organisms, or populations of morphotypes. Additionally, for mobile (pelagic?) organisms with widely distributed gene pools, most THE BASIS FOR SPECIATION AMONG CONODONTS species will prove to be chronospecies and, unlike distinct or intergrading morphotypes, more than one In spite of both predictions and fears that is unlikely to occur in a given space at a given time multielement reconstructions would drastically revise (Ager, 1963, pp. 254-259). Given that we can never conodont taxonomy, most workers would probably genetically test fossil organisms to see if they meet now agree that the turbulence produced has been less Mayr's criterion of being conspecific, species should destructive than was feared. Too often reconstruction be distinguished probabilistically by significant changes of the apparatuses is performed as an almost mandatory in the distribution of morphotypes. Rather than calling exercise, followed by assignment of these apparatuses morphotypes "species," we should thus use changes in as parts of species, genera, and families originally the ranges an d abundances of morphotypes in an effort defined on single-element criteria. This is especially to establish where speciation events are likely to have unfortunate at the generic- and familial-levels where occurred. Various statistica! schemes to measure and much information is contained in parts of the apparatus quantify the distribution of morphotypes are possible. other than the Pa elements (Merrill, et al., 1990). At Some measurements of centrai tendencies are probably the specific level this practice is probably less harmful 468 G.K MERRILL because, unlike genera and families, most if not all of analysis remains uncertain, but it is clear that an the differentiated morphologic characters occur in the analysis performed solely on "adult" specimens should Pa elements. Thus, traditional practice of basing be more reliable. This question is somewhat addressed taxonomy on the Pa elements is more justifiable at the later in this paper under Morphotypes, Species, and specific level. Although there seems to be a reluctance Biostratigraphy of Neognathodus. Nevenheless, in any to do so, "genera" containing species that have phylogeny that features morphologic simplification significandy different, phylomorphogenetically stable, through time, juveniles will bear a strong resemblance distinctive Pb, M, and S elements are ripe candidates to their descendants. This situation in Neognathodus for division, into separate genera or subgenera at the exacdy parallels the one encountered by Carruthers very least. That was suggested several years ago as likely (1910), summarized by Moore et al. (1952) dealing outcome given the genetic diversity probably present with Lower Carboniferous zaphrentid corals from among elements belonging to Palmatolepis (Merrill et Britain. Carruthers found adult corals of ever-increasing al., 1990, p. 399). morphologic simplicity through succeeding strati- graphic levels. At any given level he found a distriburion of several different morphotypes that he referred to as ONTOGENY AND PHYLOMORPHOGENY OF individua! species. Moreover, a t each stratigraphic level NEOGNATHODUS Carruthers found ontogenetic successions featuring increasing morphologic complexity during growth, a In common with the Pa elements of many genera, predictable enough situation. Not so predictable, those of Neognathodus show an ontogeny in which the however, was the realization that, phylogenetically, there less mature elements are generally simplified versions was simplification in the morphology of the adults so of the more mature elements. Element size does seem that descendant adults resemble the juveniles of their to be a good predictor of the maturity of the animals ancestors (paedomorphosis). The situation in that bore them, however, and terms such as "juvenile" Neognathodus is exacdy parallel. Ontogenetically, some and "mature" seem to be justified (Merrill & Powell, amplification in morphology occurred so that juveniles 1980). Just how simplified the juvenile Pa elements of are less reliable for placement into morphotypes and Neognathodus are in comparison with the adults is not species or for use in biostratigraphy than are the adults. uniform. Although many are more simplified others Through time the morphology of the adults simplified are essentially miniatures of the adults and in general in a highly predictable manner being reminiscent of they are reasonably good predictors of how the final the juveniles among their ancestors. The earliest known morphology will appear. How much imprecision a large species of Neognathodus in the Morrowan (N bassleri) number of juveniles will introduce into morphotype has wide platforms with widely flaring parapets,

Text-fìgure l - (all fìgures X55) Neognathodus sp. cf. symmetricus (Lane)- Magoffin Member, Breathitt Formation, (sample from D. Chesnut, his locality 1422), N. side ofDaniel Boone Parkway, KY 23-I-73 2350FSLX1550FEL. 3r 10'23'"N 83°27'19"W Leslie Co. , KY. Neognathodus "preatokaensis" - Stoney Fork Member, Breathitt Formation, type section, strip mine W. of KY Rte. 1201, KY Coordinates Sourh 211,250N 2,640,550E, 36°53'49"N 83 °33 '24"W Beli Co., KY. Neognathodus atokaensis Grayson- "Boggs" Member, Pottsville Formation, along OH Rte. 93 N. ofMt. Pleasant, 17SLP74806268 39°23'53"N 82°24'32"W Hocking Co., OH .Neognathodus sp. cf. columbiensis (Stibane) - Lower Mercer Member, Pottsville Formation, bed of Blunt Run, 40°01'22"N 81 °57'40"W Muskingum Co., OH. Neognathodus bothrops Merrill- Upper Mercer Member, Pottsville Formation, bed ofLittle Beaver Creek, 40°41 '38 "N 80°31 '54 "W Columbiana Co., OH. Neognathodus medadultimus Merrill - Putnam Hill Member, Allegheny Formation, type section, Putnam Hill in Zanesville, 39o56'15"N 82°00'52"W Muskingum Co., OH. Neognathodus medexultimus Merrill - Obryan Member, Allegheny Formation, strip mine S. of U.S. Rte. 50 E. of McArthur, 17SLP73614444 39°14'32"N 82°27'54"W Vinton Co., OH. Neognathodus roundyi (Gunnell) - St. David Member, Carbondale Formation, Jubilee College (Merrill, 1979, Stop 1), 16TBA65402221 40°49'12"N 89°46'52"W Peoria Co., IL. Neognathodus dilatus (Stauffer and Plummer) - Sparland Member, Carbondale Formation, along creek W. of Smithville, 16TBA61670443 40°38'54"N 89°56'48"W Peoria Co., IL. Neognathodus oligonodosus Merrill - Lonsdale Member, Modesto Formation, waterfall locality (Merrill, 1979, Stop 7), 16TBA50940360 40°38'54"N 89°56'48"W Peoria Co., IL. Oak Grove Grouping- Oak Grove Member, Carbondale Formation, Wolf Bridge (Merrill, 1979, Stop 3), 15TYR43592665 40°51'23"N 90°06'36"W Knox Co., IL. Obryan Grouping- Obryan Member, Allegheny Formation, along U.S. Rte. 52, 17SLN55836429 38°31 '08 "N 82°49'16"W Lawrence Co., OH. NEOGNATHODUS AND THE SPECIES CONCEPT IN CONODONT PALEONTOLOGY 469 470 G.K MERRILL generally higher than the carina (Lane, 1967). The MORPHOTYPES, SPECIES, AND BIOSTRATIGRAPHY OF carina usually does no t extend to the posteri or junction NEOGNATHODUS of the parapets. Later forms are narrower, more symmetrical, but still retained high and somewhat The various points of biologie theory and reality flaring parapets Neognathodus sp.cf. symmetricus o n text- that are discussed above were generally well known to fìg. l (Lane, 1967; Grayson, 1984). This part of the me when I studied Neognathodus in 1972 and 1975. phylogeny is least well known, at least to me. The To reiterate, these were: Atokan forms continue the wide flare of the parapets l. Congeneric species are unlikely to co-occur; with some reduction in their length (compare with the 2. Species frequently consist of populations composed form referred to informally as Neognathodus of several morphotypes; "preatokaensis" o n text-fìg. l). I t was followed by forms 3. Morphotypes are not equivalent to species and a considered transitional to the younger Atokan forms single morphotypes may be shared among several that showed reduced flare and lowering of the outer related species; parapet, maintaining an inner parapet much like their 4. Morphotypes can be used in an attempt to recognize ancestors N atokaensis (Grayson, 1984). The younger species, but because of 3., an individuai specimen Atokan phylogeny features parapets reduced in latera! may not unequivocally identify a species. flare and more consistently, reduced in height to, or even below, the height of the carina. Additionally, the It is this last point that caused me to depart from carina was shortened (or the parapets were extended!) what I considered sound biologie practice and to give so that the carina does not extend to the posterior Linnaean binomina to morphotypes. This was done junction of the parapets. This form is somewhat in the belief that i t was preferable to have a workable homeomorphic with N bassleri an d I erroneously used abstraction with biologie flaws than it was to erect that name in earlier works. It appears to be what sound biologie categories that departed dramatically Stibane ( 1967) called "Streptognathodus colombiensis" from common paleontologic practice. I have had a and is presently assigned diere provisionally. Later in quarter of a century to regret having made that choice. the Atokan the parapets fused with the carina Neognathodus illustrates a long geologie range with posteriorly (N bothrops). It was followed later, mostly a huge amount of phylogenetic change. Therefore i t is in Desmoinesian time, by shortening of the outer not reasonable to yield to the idea that only a single, parapet by progressively more anterior fusion with the highly variable species was involved. Rather it is a carina (N medadultimus and N medexultimus) This question of trying to select places that are likely points shortening continued until the outer parapet was where genetic separation was suffìcient that parts of reduced to a few nodes, commonly a single node (N the morphologic continuum should be recognized as roundyz) and was fìnally eliminateci altogether in N distinct species. The number of species involved that dilatus (Merrill, 1972). Following the loss of the outer can be redefìned on the basis of the distribution of parapet, the inner one began to disappear in an almost their morphotypes should not exceed the number of identical progressive fashion, fìrst by fusing farther an d morphotypes and may be fewer. Thus one might farther anteriorly (N metanodosus and N polynodosus) redefìne Neognathodus medadultimus as a population until i t was reduced to a few nodes (N oligonodusus) to whose most common morphotype (mode) agrees a fìnal few specimens encountered that had lost all closely with the description of it as a species by Merrill vestiges of either parapet (N anodosus) virtually at the (1972, pp. 824-825) and that also contains lesser end of Desmoinesian time (Merrill, 1975). These are numbers of morphotypes with such names such as the last neognathodontids known and the extinction bothrops and medexultimus, columbiensis and roundyi, of the genus should be a leading candidate for the etc. Other species would be redefìned in a similar paleontologic placement of the Desmoinesian- fashion. Redefìning species based upon centrai Missourian boundary. The phylomorphogenetic tendencies of their morphologic ranges rather than progression from approximately the middle of the single, fìxed, morphologies is no t entirely a new concept Atokan to essentially the end of the Desmoinesian is (Wagner & Erwin, 1995) and has little negative impact quite clear: it was one of continuous morphologic other than to make a few specifìc assignments simplifìcation. The rate and the constancy of rate of uncertain. Some units have yielded so few specimens this simplifìcation is hard to determine, but there does belonging to the genus that specifìc determination must seem to be an acceleration in the simplifìcation in the remain questionable. This is especially true near the later part of the range of the genus, although this may termination of the range of the genus. However, this in part be an artifact of the paucity of marine units may not preclude making a reasonable biostratigraphic representing the necessary times to show the fìnal interpretation from such faunas. Such redefìnition of simplifìcation in as much detail as the part preceding species should lead to no reduction in the biostra- i t. tigraphic utility of Neognathodus. The distributions of NEOGNATHODUS AND THE SPECIES CONCEPT IN CONODONT PALEONTOLOGY 471 morphotypes have been, and will continue to be, the that would be assigned to roundyi, and five specimens criteria used for age determination. The precision that would be assigned to dilatus. This is the apparent possible by this method is extremely high. Brown et skewness toward the younger end of the phylogeny. al., (1998) in their analysis of some Desmoinesian That this is an artifact is borne out by the fact that i t is marine units used a numeric analysis of the means of the five smallest specimens that have the dilatus morphotype distribution to demonstrate differences in morphology, o ne of these having a platform that is well age among units previously considered coeval. Rexroad under 0.1 mm in length. In a similar context the next et al. (1998) made use of these ratios in correlation two smallest specimens from this sample are the ones and more recendy did a numerica! analysis in the same with the roundyi morphology and the largest half dozen manner as Brown et al. (1998). Application of modes exhibit medadultimus an d medexultimus morphologies. seems a more robust procedure, but we are stili lndeed, if the six largest specimens are arrayed from attempting to determine the best way to apply these largest to smallest they would correspond to these chronologically very sensitive tools. morphotypes: medexultimus - medadultimus - The following examples may help to illustrate both medadultimus - medexultimus - medexultimus - concepts of speciation based upon the distribution of medadultimus. Analysis of many more specimens from morphotypes and the biostratigraphic application of many more samples and localities in the Obryan shows those morphotypes. To do this, two samples were that the single most common morphotype (mode) for chosen not quite at random. The first is from the these samples is in fact the medexultimus one with a Obryan , formerly "southern" Vanport Limestone strong secondary mode of the medadultimus (Allegheny Formation) in southern Ohio at a well- morphotype (Merrill, 1972). Thus a specific known locali ty (Ferm et al., 1971; Rice et al., 1992). assignment of ali of these forms to the species The second is the transitional shale at the top of the Neognathodus medexultimus is justified. black shale at Wolf Bridge, Oak Grove Beds Applying a similar analysis to the Oak Grove (Carbondale Formation), northwestern Illinois. This specimens shows that the half dozen largest specimens is also a well-known conodont locality (Merrill & King, from largest downward in size belong to the roundyi - 1971; Merrill, 1975, 1979; von Bitter &Merrill, 1998). roundyi - medexultimus - medexultimus - roundyi - Each was chosen no t only because they are well-known, medexultimus morphotypes. As stated above, this but also because each sample yielded a convenient sample has lower juvenility than the one from the number of neognathodontid Pa elements Obryan and continuing with successively smaller (coincidentally 13 in each case), permitting a small, specimens would have no effect on the basic but relevant number of specimens to be analyzed. Thus distribution pattern begun with the six largest these entire populations could be fit on a single figure specimens. This once again emphasizes that arraying (Text-fig. 1). The units chosen were not entirely the morphotypes so that more mature specimens are arbitrary either, because through a tortured chain of emphasized is an effective way to analyze phylo- reasoning, they are supposed to be "correlatives" (Moore morphogenetic development and genuine distriburion et al., 1944; Rice, et al., 1994). Even a cursory glance of morphotypes. By diminishing or eliminating the at the two groups ofillustrations on Text-fig. l indicates apparent "younging" effect that juvenilirywould force that they are drawn from different populations. The upon the population, such analysis emphasizes both Obryan population contains more juveniles and a wider the importance of adults in the analysis and the range in morphologic variation, likely resulting in large importance of centrai tendency statistics based upon part from that juvenility. The Oak Grove population the mode(s). has fewer juveniles and that may be largely responsible Clearly, removal of the "younging" effect of for a curve that is extremely leptokurtic; seven juvenility in the Obryan population shows that the specimens would be assigned to roundyi and six to phylomorphogenetic development of the Obryan medexultimus. Indeed, such a distribution, although neognathodontids is more primitive and the Obryan excellent for biostratigraphic determinations, makes is distinctly older in both a phylogenetic and a specific assignment somewhat uncertain. A larger chronological sense than the Oak Grove population. population is necessary to determine the dominant morphotype (Merrill, 1972, et seq.) and these larger populations show that the dominance makes a specific CONCLUSIONS assignment to N roundyi the correct one. For the Obryan sample the curve is far more plarykurtic and Neognathodus remains the preminent conodont, exhibits an apparent skewness that is an artifact of its possibly the preeminent fossil, for biostratigraphy in juvenility. Thus a calculation based on the pure form Lower - Middle Pennsylvanian rocks. Analysis of the of morphotypes would show three specimens that distribution of morphotypes can lead to very precise would be assigned to medadultimus, three specimens age determinations. lt is also the distribution of these that would be assigned to medexultimus, two specimens morphotypes, rather than the single morphotypes 472 G.K MERRILL themselves, that should be used in the identifìcation LANE, H .R., 1967, Uppermost Mississippian and Lower and redefìnition of species. Such redefìnition will not Pennsylvanian conodonts from the type Morrowan region, Arkansas: Journal ofPaleontology, 41: 920-942, pls. 119-123. adversely effect the genus for biostratigraphy, but will MAYR, E., 1942, Systematics and the Origin ofSpecies: Columbia more closely conform to biologie reality. University Press: 334 pp. The situation in Neognathodus where species consist MccLAVE, J.T., BENSON, P.G. & SI NC ICH, T, 1998, Statistics for of several morphotypes and individuai morphotypes Business and Economics, 7'h ed., Prentice Hall, Upper Saddle River, NJ: l 067 pp. to are shared by more than one species is unlikely be M ERRILL, G.K., 1972, Taxonomy, phylogeny and biostratigraphy unique. lndeed, this is probably the norm with most of Neognathodus in Appalachian Pennsylvanian rocks: Journal populations of fossil organisms. lt certainly should of Paleomology, 46: 81 7-829. impact those interpretations where a large number of MERRILL, G.K., 1975, Pennsylvanian conodont biostratigraphy and congeneric species are presendy recognized within paleoecology of northwestern Illinois: Geologica! Society of America Microform Publication, 3: 130 pp., 4 pls. (on 2 single samples. microfìche). M ERRILL, G.K., 1979, Desmoinesian (Pennsylvanian) conodont localities in northwestern Illinois: Guidebook, College of Charleston for the IX International Congress o n the Geology ACKNOWLEDGMENTS and Stratigraphy of the Carboniferous: 48 pp., 4 pls. M ERRILL, G.K. & PowELL, R.J., 1980, Paleobiology of juvenile Steven J. Kivett assisted with the SEM micrography and (nepionic?) conodonts from the Drum Limestone manipulated the fìles from the SEM to produce the layout for (Pennsylvanian, Missourian- Kansas City area) and its bearing Text-fìg. l. Kenneth M. Fletcher, J r. assisted with the calculations on apparatus ontogeny: Journal of Paleomology, 54: 1058- of magnifìcations. Donald Chesnut, Kentucky Geologica! Survey 1074. provided one of the samples used in this study and some very MERRILL, G.K., VON BITTER, P.H., & GRAYSON, R.C., ]r., 1990, valuable locality information. The generic concept in conodont paleontology - growth, changes, and developments in the last rwo decades: Courier Forschungs-Institut Senckenberg, 118: 397-408. MooRE, R.C. (chairman), 1944, Correlation of Pennsylvanian REFERENCES Formations ofNorth America: Geologica! Society of America Bulletin, 55: 657-706. AGER, D.V., 1963, Principles ofPaleoecology, Mc Graw-Hill: 371 MooRE, R.C., LALI CKER, C.G., & FI SC HER, A.G., 1952, pp. Invertebrate Fossils: McGraw-Hill: 766 pp. BnTER, VON, P. H. & MERRILL, G.K., 1998, Apparatus composition Rf.xROAD, C.B. BROWN, L.M, D EVERA, J. & SUMAN, R.]., 1998, and structure of the Pennsylvanian conodont genus Gondole/la Conodont biostratigraphy and paleoecology of the Perth based o n assemblages from the Desmoinesian of northwestern Limestone Member, Staunton Formation (Pennsylvanian) of Illinois, U.S.A: Journal of Paleontology, 72: 112-132. the Illinois Basin, U.S.A. In Szaniawski, H. (ed.), Proceeding BoucOT, A.]., 1978, Community evolution and the rates of of the Sixth European Conodont Symposium (ECOS VI): cladogenesis: Evolutionary Biology, 11: 545-655. Palaeomologica Polonica, 58: 247-259. BROWN, L. W, RExRoAD, C. B., KELLEY, J ., CHAPLIN,] .R., HEMISH, Rf.xROAD, C.B., WAD E, J.A., MERRJLL, G.K. & PADGETT-ALANO, L.A. & THOMPSON, T. L., 1998, Conodont biostratigraphy of P. , 1999, Conodont biostratigraphy of the Mecca Quarry Shale the Verdigris Limestone (Pennsylvanian, Desmoinesian) from Member and Velpen Limestone Member of the Linton Oklahoma and Missouri: Geologica! Society of America Formation (Pennsylvanian, Desmoinesian) in the eastern part Abstracts with Program, 30 (2): 7-8. of the Illinois Basin, U.S.A.: Geologica! Society of America CARRUTHERS, R. G., 1910, O n the evolurion of Zaphrentis delanouei Abstracts with Program, 31(5): 67. in Lower Carboniferous times: Geologica! Society ofLondon RI CE, C.L., KosANKE, R.M., & HENRY, T.W, 1994, Revision of Quarterly Journal, 66: 523-538, pls. 36-37. nomenclature and correlations of some Middle Pennsylvanian ELDREDGE, N. & Gomo, S.]., 1972, Punctuated equilibria: An units in the northwestern part of the Appalachian basin, alternative to phyletic gradualism. In Schopf, T.J.M. (ed.), Kentucky, Ohio, and West Virginia: Geologica! Society of Models in Paleobiology, Freeman, Cooper & Co., San America Special Paper, 294: 7-26. Francisco: 82-115. RrcE, C.L., MARTINO, R.L, & SLUCHER, E.R., 1992, Regional ERWIN, D.H. & ANSTEY, R.L., 1995, Speciation and the fossi! aspects of Pottsville and Allegheny stratigraphy and record. In Erwin, D.H. &Anstey, R.L. (eds.), New Approaches depositional environments Ohio and Kentucky: U.S. to Speciation in the Fossi! Record, Columbia University Press: Geologica! Survey Open-fìle Report 92-558 prepared for 11-38. Geologica! Society of America Annua! Meeting, Cincinnati, FERM,J.C., HORNE,].C., SWINCHATT,J.P. &WHALEY, P.W, 1971, OH: 67 pp. Carboniferous depositional environments in northeastern SACHS, L. , 1984, Applied Statistics: A Handbook ofTechniques, Kentucky: Geologica! Society ofKentucky Annua! Spring Field 2nd ed., Springer-Verlag, New York. Conference Guidebook: 30 pp. SHAw, A.B., 1964, Time in Stratigraphy: McGraw-Hill: 365 pp. FuTUYAMA, D.J., 1986, Evolutionary Biology, 2nd ed., Sinaeur SHAw, A.B.,l969, Adam and Eve, paleontology, and the non- Assoc., Sunderland, MA: 600 pp. objective ans: Journal of Paleontology, 43: l 085-1098. GRAYSON, R. C., ]r., 1984, Morrowan and Atokan (Pennsylvanian) STERN, C, 1949, Gene and character. In Jepsen, G.L., Mayr, E. & conodonts from the northeastern margin of the Arbuckle Simpson, G.G. (eds.), Genetics, Paleontology, and Evolution, Mountains southern Oklahoma: Oklahoma Geologica! Survey Princeton University Press: 13-22. Bulletin, 136: 41-63. STAUFFER, C.R., & PLUMM ER, H.J.,l932, Texas Pennsylvanian GUNNELL, EH., 1931, Conodonts from the Fon Scott limestone conodonts and their stratigraphic relations: University ofTexas of Missouri: Journal of Paleontology, 5: 244-252, p l. 29. Bulletin 3201: 13-50, 4 pls. GUNNELL, EH., 1933, Conodonts and fìsh remains from the STIBANE, ER., 1967, Conodonten des Karbons aus den niirdlichen Cherokee, Kansas City, and Wabaunsee groups of Missouri Anden Siidamerikas: Neues Jahrbuch fiir Geologie und and Kansas: Journal ofPaleontology, 7: 261-97, pls. 31-33. Palaomologie, Abhandlungen 128: 329-340, pls. 35-37. NEOGNATHODUS AND THE SPECIES CONCEPT IN CONODONT PALEONTOLOGY 473

STU RGEON, M.T & HoARE, R.D., 1968, Pennsylvanian New Approaches to Speciation in che Fossi l Record, Columbia Brachiopods of Ohio, Ohio Division of Geologica! Survey University Press: 11-38. Bulletin, 63: 95 pp. T URNER, J.R.G., 1986, The genetics of adaprive radiarion: A Neo- (manuscript received December 2, 1998 Darwinian theory of puntuational evolurion In Raup, D. & accepted September 21, 1999) Jablonski, D. (eds.) , Patterns and Processes in che Hisrory of Life, Springer Verlag, Berlin: 447 pp. Glen K. MERRILL VALENTI NE, J.W. , 1970, How many marine invertebrate species: Department of Natura! Sciences, A new approximation: Journal of Paleonrology, 44: 410-415. University ofHouston- Downtown W AGNER, P.J. & ERWTN, D.H., 1995, Phylogeneric patterns as test l Main Srreet, Houston, TX 77002 USA of speciation models. In Erwin, D .H . & Anstey, R.L. (eds.) e-mail: merrillg@ zeus.dt.uh.edu