Bijdragen tot de Dierkunde, 64 (4) 193-213 (1995) SPB Academie Publishing bv, The Hague Arthropod pattern theory and Cambrian trilobites Frederick A. Sundberg Research Associate, Invertebrate Paleontology Section, Los Angeles County Museum of Natural History, 900 Exposition Boulevard, Los Angeles, California 90007, USA Keywords: Arthropod pattern theory, Cambrian, trilobites, segment distributions 4 Abstract ou 6). La limite thorax/pygidium se trouve généralementau niveau du node 2 (duplomères 11—13) et du node 3 (duplomères les les 18—20) pour Corynexochides et respectivement pour Pty- An analysis of duplomere (= segment) distribution within the chopariides.Cette limite se trouve dans le champ 4 (duplomères cephalon,thorax, and pygidium of Cambrian trilobites was un- 21—n) dans le cas des Olenellides et des Redlichiides. L’extrémité dertaken to determine if the Arthropod Pattern Theory (APT) du corps se trouve généralementau niveau du node 3 chez les proposed by Schram & Emerson (1991) applies to Cambrian Corynexochides, et au niveau du champ 4 chez les Olenellides, trilobites. The boundary of the cephalon/thorax occurs within les Redlichiides et les Ptychopariides. D’autre part, les épines 1 4 the predicted duplomerenode (duplomeres or 6). The bound- macropleurales, qui pourraient indiquer l’emplacement des ary between the thorax and pygidium generally occurs within gonopores ou de l’anus, sont généralementsituées au niveau des node 2 (duplomeres 11—13) and node 3 (duplomeres 18—20) for duplomères pronostiqués. La limite prothorax/opisthothorax corynexochids and ptychopariids, respectively. This boundary des Olenellides est située dans le node 3 ou près de celui-ci. Ces occurs within field 4 (duplomeres21—n) for olenellids and red- résultats indiquent que nombre et distribution des duplomères lichiids. The termination of the body generally occurs within des Trilobites du Cambrien ont été enquelque sorte sous la con- node 3 for corynexochids and within field 4 for olenellids, red- strainte d’un certain programme de patterning génétique. lichiids, and ptychopariids. In addition, the location of macro- Cependant, l’emplacementfréquent des limites ailleurs que dans pleural spines, which may indicate the location of the gonopores les endroits pronostiqués, ainsi que le déplacementpossible des or anus, generallyfalls at the predicted duplomeres. The bound- nodes suggèrent que d’autres facteurs ont été aussi impliqués ary between the prothorax and opisthothorax of olenellids oc- dans le nombre des duplomères des diverses parties du corps. curs within or near node 3. These results indicate that the num- Cette variabilité rend plausible l’idée que les Arthropodes du ber and distribution of duplomeres within Cambrian trilobites à la différence de avaient Cambrien, ceux récents, un pro- were somewhat constrained by some genetic patterning pro- gramme génétique plus simple, ce qui facilitait l’apparition de gram. However, the common distribution ofboundaries outside modifications dans le Bauplan. of the locations and the of predicted possible shifting nodes sug- gest that other factors were also controlling the number of duplomereswithin the body parts. This variation supports the Introduction idea that Cambrian arthropods, unlike modern arthropods, had which allowed for in a simpler genetic program, easily changes the Bauplan. Interpreting morphological characters to decipher phylogenies of organisms is always a challenge. This is especially true when organisms are extinct, Résumé where there are no soft-body or biochemical re- mains to supplement the limited amount of infor- Une analyse a été entreprise de la distribution des duplomères mation provided by skeletal remains. Thus, when dans le le le des (= segments) céphalon, thorax, et pygidium be made for generalizations on Baupläne can a Trilobites du Cambrien, afin de vérifier si l’Arthropod Pattern specific group, guidelines can be generated to help Theory (APT) proposée par Schram & Emerson (1991) s’ap- the observed and plique à ces fossiles. La limite céphalon/thorax se trouve, interpret morphological patterns comme pronostiqué,au niveau du node duplomère 1 (duplomère to construct phylogenies. Such a generalization has 194 F.A. Sundberg - Arthropod pattern theory and Cambrian trilobites recently been proposed by Emerson & Schram APT implications to trilobite Baupläne and Schram & Emerson (1990) (1991). They sug- gested that the location of tagmata transitions, As a prediction of APT, changes in arthropod the and terminations gonopores, anus, body occurs Baupläne should occur in quantum jumps, from within the at specific regions Bauplan of arthro- node to node. This prediction appears to hold true pods. This Arthropod Pattern Theory (APT) was for those extinct and extant taxa discussed by both fossil and modern Schram & Emerson with few based on taxa, comparative (1991) only a excep- anatomy of exterior and interior morphology, tions. Trilobites would be a good test of APT in ontogeny, and developmental genetics, although that, unlike most modern arthropod groups, they this has not been tested. APT = theory statistically vary widely in their numberof segments ( duplo- implies that there are only a limited number of ar- meres). If APT is universal for all arthropods, then rangements of segments and other morphological changes in the numberof segments within trilobites features available to arthropods. This could in- would be in jumps, from one node to the next. In the of crease potential convergenceamong distantly addition, the locations of tagmata transitions related groups because the number of segments (cephalon/thorax and thorax/pygidial boundaries) within the body and body parts or the location of should occur within the nodes. the gonopores or anus are limited. Conversely, Other morphological features that may display related taxa also would to have distributions closely be expected APT may be the location of macro- similar distributions of in the and the the segments body. pleural spines boundary between pro- Schram & Emerson (1991) have also proposed thorax and opisthothorax in some Olenellida. Ac- that most biramous arthropods are composed of cording to Harrington et al. (1959, p. 073), macro- each duplomeres, containing two primary seg- pleural spines may indicate the location of the Both of these fused If ments. segments are typically gonopores. they do, then they wouldbe expected together and in some taxa the individual segments to occur eitherwithin a node or at duplomeres 9 or can only be recognized by the internal organization 16. of soft parts. Schram & Emerson recognized two Schram & Emerson's preliminary study sug- of of and types sequences duplomeres, nodes fields gested that trilobites display a pattern where tag- (Fig. 1). They observed that morphological changes mata boundaries occur at nodes. They studied the generally occurred at three nodes within the trilobites Triarthrus, Naraoia, Olenoides, and Ag- Bauplan : node 1 = duplomeres 5 and 6; node 2 = nostus and in each case the cephalon/thorax transi- duplomeres 11-13; and node 3 = duplomeres tion occurred in node 1. In Olenoides and Triar- 18-20. The fields are duplomeres that are relatively thrus, the thorax/pygidium transition occurred in stable because do they not generally show morpho- nodes 2 and 3, respectively. If these patterns hold logical change (field 1 = 1-4, field 2 = 7-10, field true inother trilobites, then the number of thoracic 3 = field 4 = In notable and well the location of 14-17, 21-n). some ex- pygidial segments, as as ceptions, duplomeres 9 and 16 within fields 2 and macropleural spines, could be important in deter- 3, respectively, can possess gonopores or anus; and mining the evolutionary relationship among taxa. several have and terminate in taxa an anus field 4. If the pattern does not hold true, then APT is not The isopod Ligia provides an example of mor- completely applicable to trilobites, which could phological changes and other features occurring potentially make them different from most other with the nodes (from Schram & Emerson, 1991). arthropods. Ligia has its head, thorax, and abdomen terminate at the ends of nodes 1, 2 and 3, respectively. In ad- the female of in the dition, gonopore Ligia occurs Methods first duplomere of node 2, the male gonoporeis in the last of node 2 and the anus is located duplomere This test of APT is limited to Cambrian trilobites because of in last in 3. the duplomere node the difficulty in deciphering their supergeneric classification Bijdragen tot de Dierkunde, 64 (4) - 1995 195 wi frischenfeldi the bargessensis reyndidsiformis is opiki mat ovata, arctica evens geikjei lobat t s s la el lancastrioides kin roddyi Oryctocephalina Oryctocephalops Oryctocephalus Oryctocephalus Oryctocephalus Oryctocephalus Ovatoryctocara a Cheiruroides ibirica Lancastria Oryctocara Sandoveria Ton Tonkinella Fig. 1. Distribution of duplomeres ( = segments) of representatives of Oryctocephalidae(Corynexochida) relative tothe fields and nodes predicted from APT. A (Palmer & Halley, 1979) and their apparent plasticity in mor- bite specimens. total of 348 trilobite specimens representing phology (McNamara, 1986, 1988; Hughes, 1991). This contrasts 291 species from Olenellida, Redlichiida, Corynexochida, and to later trilobite groups, where phylogenies are somewhat better Ptychopariida were studied (Appendix). Only a single specimen understood and fixed in their number of recorded unless of some groups are was per species holaspids a species were thoracic segments