Veldmeijer Cretaceous, toothed pterosaurs from Brazil. A reappraisal
1. Introduction
Campos & Kellner (1985b) related that references to flying reptiles from Brazil (not from the Araripe Basin) were made as early as the 19th century, but the first find from Chapada do Araripe was described as late as the 1970s (Price, 1971, post–cranial remains of Araripesaurus castilhoi). Wellnhofer (1977) published the description of a phalanx of a wing finger of a pterosaur from the Santana Formation and named it Araripedactylus dehmi. Since then, much has been published on the pterosaurs from Brazil, and there has been an increasing interest in the material from this area, resulting in an increase in scientific interest in pterosaurs in general. The plateau of the Araripe Basin, in northeast Brazil on the boundaries of Piaui, Ceará and Pernambuco (figure 1.1) was already famous for its well preserved fossils, escpacially fish (e.g. Maisey, 1991), long before the area became the most important source of Cretaceous pterosaur fossils. At present, it is the most important area for Cretaceous pterosaurs globally, although an increasing number of finds are reported from China (e.g. Lü & Ji, 2005; Wang & Lü, 2001 and Wang & Zhou, 2003). Some of the Brazilian material is severely compacted (Crato Formatin; Frey & Martill, 1994; Frey et al., 2003a, b; Sayão & Kellner, 2000) and preserved on a laminated limestone comparable to that of Solnhofen. (The type locality of most, if not all, pterosaur fossils from the Araripe Basin is uncertain, because no systematic, scientically based excavations or even surveys have been done in this area. In stead, the fossils are obtained from the local people. For the geological setting of the Araripe Basin, the reader is referred to Beurlen (1971), De Buisonjé (1980); Kellner & Tomida (2000); Martill et al. (1993); Maisey (1991); Pons et al. (1990); Wellnhofer (1977, 1985) and Wellnhofer et al. (1983). Fossils in nodules are typical for the Rornualdo Member of the Santana Formation [Martill, pers. com.]). However, many pterosaur fossils from Brazil are preserved as three–dimensional parts of skeletons (e.g. De Buisonjé, 1980; Campos & Kellner, 1985a, b, 1997; Dalla Vecchia, 1993; Dalla Vecchia & Ligabue, 1993; Fastnacht, 2001; Kellner, 1984, 1995a, 1996a; Kellner & Campos, 1989, 1994; Kellner & Hasegawa, 1993; Leonardi & Borgomanero, 1985, 1987; Martill & Frey, 1999; Price, 1971; Veldmeijer, 2002; Wellnhofer, 1977, 1985, 1987, 1991b; Wellnhofer & Kellner, 1991; Wellnhofer et al., 1983), or almost complete skeletons (Kellner & Tomida, 2000; Veldmijer, 2003a; Wellnhofer, 1985, 1991b). Apart from descriptions palaeobiological studies of Brazilian pterosaurs include those of Bennett (1990, functional morphology; 1992, sexual dimorphism; 1993, ontogeny), Campos et al., (1984, soft tissue), Frey & Martill (1998, ontogeny), Frey et al., 2003c; Kellner (1994a, soft tissue; 1994b, paleoecology; 1995b, 1996b, phylogeny; 1996c, soft tissue), Kellner & Tomida (1996, phylogeny), Martill & Frey (1998, diversity), Martill & Unwin (1989, soft tissue) and Wellnhofer (1988, functional morphology). The increasing interest however, has not lead to much more clarity in the systematics of the prehistoric animals. The Brazilian toothed taxa are linked with the pterosaurs from the Cambridge Greensands, Cambridge, England, which include two type species (Coloborhynchus and Criorhynchus). The material however, is extremely fragmented and the systematics is much complicated and discussed (see Unwin, 2001 for an exhaustive list of publications concerning this material). Even the most recent evaluation (ibidem) does not enlighten the situation much. The material presented here (Coloborhynchus in chapter 2; Criorhynchus in chapter 3; Thalassodromeus and Anhanguera in chapter 4; Brasileodactylus in chapter 5 and 6) not always proved to be a new species (only in the case of Coloborhynchus, a new species could be assigned). Nevertheless, the evaluation of this previously unpublished material as well as
10
Veldmeijer Cretaceous, toothed pterosaurs from Brazil. A reappraisal the first hand study of most of the published fossils, has resulted, not only in alternative systematics, differing from the systematics advocated by Kellner (for example Kellner & Tomida, 2000) and Unwin (for example Unwin, 2001), but also in the creation of more detailed diagnoses. Does this however mean that problems have been solved now? Far from that. It is striking that systematic palaeontology is often based on very small pieces (the material from the Cambridge Greensands is particularly notorious, but quite a few holotypes from Brazil are only small fragments too whereas almost complete specimens are hitherto unpublished), partially prepared specimens (Th. sethi for example), specimens collected over long period of time (again Th. sethi) and so forth. Fossils, seemingly belonging to already existing taxa cannot rely on much interest even though these specimens might be much more complete than the holotypes. This is unfortunate, as the present works shows, because not only will it result in more detailed and reliable diagnoses (which, in turn, will increase the reliability of the phylogenetic studies), but also statistic studies increase in reliability. For obvious reasons, first hand study of material is inevitable. But despite the fact that material in collections should be available for study, this is unfortunately not always the case, either because of unwillingness of people in charge or some other, often unclear reason (for instance closed down institutes or institutes which do not exist [yet]). Fortunately, some institutes already started to put their collection on the worldwide web, including pictures (AMNH New York) but these pictures are in low resolution. There is another major problem attached to the material from Chapada do Araripe. All material, as far as I am aware, has been obtained through fossil dealers. This means that valuable information on the site and stratigraphy and so on has been lost.1 Furthermore, the production of fake fossils has become regular practice (see also Martill, 1993). For pterosaurs in particular this means that specimens, apparently complete, are in fact composites. If the owner is lucky, he/she knows what belongs to what, but unfortunately most of the time nobody knows (anymore). Not only the Brazilian fossil hunters are guilty of this kind of practice (the creation of composite skeletons) but well known preparators in Europe do not hesitate to create their fossils. Although sometimes they say the specimen is a composite, important details on what exactly was put together, lacks. The present work presents descriptions of new material from Brazil of most toothed taxa (Criorhynchus, Coloborhynchus, Anhanguera and Brasileodactylus). Most of the chapters in this work have been published previously as papers in various journals; the reader is therefore requested to refer to the original papers, rather than the edited chapters herein. In the present work, the papers have been inserted largely intact. However, if necessary alterations had to be made, this is either done by means of endnotes, or indicated in an endnote but altered in the text proper. Minor alterations such as corrections of grammar have not been indicated; references have been unified in one section ‘Cited literature’. Furthermore, the layout of the various different papers have been abandoned in order to unify them in this work. Joining the papers also resulted in inclusion of a list of abbreviations in each chapter because of the use of different abbreviations in various papers. No references to figures in other chapters are made.
11