Ralph E Chapman. Encyclopedia of Dinosaurs. Amsterdam: Academic Press, 1997.
Computers are in the process of revolutionizing the way we look at dinosaurs. Early uses of computers with dinosaurs date to the 1960?s, concentrated in research applications (e.g., data analysis) and early attempts by some museums (e.g., the National Museum of Natural History of the United States) to store their collections data electronically. These applications became more and more prevalent, and progressively more sophisticated through the 1970?s and 1980?s. It is during the 1990s, however, that computers finally have expanded to start assisting paleontologists in the field, during the preparation process, and in making a great difference in how dinosaurs are being presented to the general public.
The application of computers and related technology to dinosaurs can be divided into five major areas: field work and specimen collection, specimen preparation, collections and data management, research, and exhibition. We will touch on these subjects in order.
Finding dinosaurs in the field still continues to be done mostly in the same way it has been done for more than a century: finding areas with the right rock from the right time and the right paleoenvironments, and walking the outcrops looking for exposed bones. For the most part this will continue to be the approach taken, but new technology promises to change much of the process of getting into the field, determining where you are once there, and even analyzing what has been found and what is still in the ground.
Formerly, it took much skill to use the available maps and keep track of where you were in the field; many mistakes were made and much of the old location data are inaccurate. This is changing because global positioning systems (GPS) using satellite technology can locate a position within a 100 m or less anywhere on earth and are being used by paleontologists during exploration. For example, a GPS system was used by the American Museum of Natural History in their new Gobi Desert expeditions (McKenna, 1992) to keep track of where they were in unmapped or badly mapped areas. These locality data have yet to be used extensively with geographic information systems (GIS), which combine geographic data with computer databases. In the future this will allow predictive mapping of potential outcrop areas, and the GIS will suggest where new prospecting should be done.
Small-scale geographic data (e.g., quarry maps of bone orientations and positions) were once taken using compasses and were plotted in field notebooks, but now are being taken automatically using electronic distance measurement devices with millimeter accuracy (see Jorstad and Clark, 1995, for work on paleohominid applications) and, in even smaller scale and with higher accuracy, using three-dimensional digitizers (e.g., see Jefferson, 1989, on Pleistocene Rancho La Brea material). Finally, technology is being developed to allow paleontologists to determine, in some cases, the nature of fossils buried in an area. One example is the application of geophysical diffraction tomography by Witten, et al. (1992), who tried to determine the extent of the material buried for a specimen of the sauropod Seismosaurus.
Computers and related technology have had only limited effect on specimen preparation but changes are well on the way. Standard X-rays have been used for years during specimen preparation (see Zangerl and Schultze, 1989), but more advanced imaging methods such as computed tomography (CT) are allowing, in some cases, significantly better indications of what fossil material is present in an unprepared specimen (see Clark and Morrison, 1994). In fact, the CT scanning of dinosaur eggs has become a standard operating procedure (e.g., Hirsch et al., 1989), and many fetal dinosaur fossils are now being found. It may be common in the near future for preparators to have three-dimensional models of the specimens they are preparing as an aid to the process.
Specimen casting also will be changed by three- dimensional computer modeling. Fossils can be digitized using various scanning or related technologies and three-dimensional reconstructions made in wax, plastic, or some other medium for exhibition or study using methods of automated casting (also known as prototyping or, in some cases, stereolithography; see Burns, 1993) thus avoiding more destructive ways of casting specimens. These casts can also be varied in scale to enable them to be viewed at a more manageable scale; very large specimens can be reduced to allow more easy manipulation and very small specimens enlarged to allow them to be viewed without a microscope.
The use and development of collections materials has changed dramatically and will continue to change in the coming years due to computers. The initial transfer of collections data to an electronic format was done in the 1960?s (e.g., the National Museum of Natural History of the United States), with many now using their third, fourth, or later generation of database software, and nearly all having some electronic storage. Where data used to be kept in card catalogs that took long search times to extract simple information, data are now available in large computer databases that can be searched for very complex information almost instantaneously. These databases are also being stored on-line and, in many cases, are available for searching on the Internet.
Another big change due to computer technology is the nature of data being made available; computer databases are not restricted to just text anymore. Image scanners are making it possible to store pictures of specimens as well as the text information that goes along with it. Furthermore, with the scanning technology being developed, three-dimensional computer images can now be stored and viewed from a variety of angles. A first attempt at storing such images was made by Rowe et al. (1993) for high-resolution CT scans of the skull of the cynodont Thrinaxodon. The CD-ROM released contains both the important descriptive literature and CT scan data that can be viewed from front-to-back, side-to-side, or top-to-bottom.
The use and storage of bibliographic data on dinosaurs is changing rapidly. The Bibliography of Fossil Vertebrates is being made available in electronic format and the next release of A Bibliography of the Dinosauria (Chure and McIntosh, 1989) will be made available in electronic format as well. Many dinosaur paleontologists have developed and maintained their own electronic bibliographic databases using their personal computers.
Computers have had a major impact on the types of research being done in the natural sciences. The relatively unquantitative approaches taken by scientists in the early days mostly have been replaced by quantitative ones as methods have become more and more rigorous. Computer-based studies of dinosaurs are still in their infancy, however, because most of the research done on them still proceeds mostly in a qualitative fashion. This is due, to a large part, to the difficulties involved in doing research on a group represented by relative few individuals that are often incomplete and fragmentary. However, there are a number of important exceptions and we will discuss them within four major areas of research: morphometrics, mathematical and/or statistical studies of variation used to solve taxonomic or evolutionary problems; phylogenetic analyses, the analysis of the relationships among taxa; functional morphology, studies of the biomechanics and locomotion of dinosaurs; and distributional analyses, studies of the distribution of dinosaurs through time and space.
Morphometrics is the quantitative analysis of shape. Before the availability of computers, paleontologists were limited to analyses of two or three variables, usually within the context of allometry, the study of size and its consequences. One classic study of this type is the analysis of bivariate (two-variable) allometry in groups of ceratopsian dinosaurs by Gray (1946). Another morphometric approach, the application of D?Arcy Thompson?s (1942) transformation grids, could be done without computers and grids were generated for a number of dinosaur groups (e.g., Lull and Gray, 1949, for ceratopsians).
Computers allowed calculations to be done much faster than ever before, which opened the door for multivariate analyses, those using many variables simultaneously (e.g., Dodson, 1975, 1976; Chapman, et al., 1981), as well as very sophisticated geometric methods of shape analysis (e.g., Chapman, 1990). These methods have led to a much better understanding of growth in dinosaurs, have allowed sexual dimorphs to be described in some cases (e.g., see Chapman, et al., 1981, for the pachycephalosaurid Stegoceras), and have started to be used more within studies of phylogeny and functional morphology. The next step will be to do even more sophisticated analyses of shape in three-dimensions and use high-level computer graphics to show the results.
Phylogenetic Analyses try to determine relationships among the dinosaur taxa being studied. Here, we will concentrate on numerical cladistic analyses, which attempt to reconstruct these relationships on trees, called cladograms, using the principle of parsimony; looking for the shortest trees by minimizing the number of evolutionary steps needed to generate the tree and minimizing instances of convergence or parallelism.
Originally done by hand, most cladistic analyses of dinosaurs, as well as all other groups of organisms, are carried out on computers using programs such as Phylogenetic Analysis Using Parsimony (PAUP) (Swofford and Begle, 1993) and MacClade (Maddison and Maddison, 1992). These packages search for the shortest trees (cladograms) that account for the characters forming the database supplied by the researcher, while reducing the number of reversals and convergences in these characters. Computer programs are necessary because the number of possible trees increases exponentially as more taxa are studied.
Cladistic research on dinosaurs is burgeoning, and nearly every major group has been analyzed to some degree. Much of this work was begun in the 1980?s, spearheaded by the research of Gauthier (1984, 1986), and it has become the standard for reconstructing phylogenetic relationships.
As more numerical cladistic analyses are done, we have begun to get a better understanding of how the different dinosaur groups are related and a much better understanding of the anatomy of dinosaurs and why they look the way they do. An additional way to use phylogenetic analysis in research is to follow the evolution of a single character (an anatomical feature) on a tree to see how it varies across a dinosaur taxon. Other ways are to superimpose geographic locations, ecological characteristics, or time on trees to see how they have influenced the history of dinosaurs. As computers get stronger, phylogenetic analyses will provide better information.
Functional morphology is the study of how organisms work. These studies are still relatively rare for dinosaurs but such analyses are now becoming more common. Functional analyses typically make use of architectural (e.g., Weishampel, 1993) and/or machine analogies to understand the evolution and operation of a particular anatomical structure. They commonly use physical models, graphical representations, mathematical computations, computer simulations, and thought experiences to analyze this anatomy. Examples include Alexander?s (1989) analyses of dinosaur locomotion and Weishampel?s (1981) study of the nasal systems of lambeosaurine hadrosaurids. Computers can help increase the level of sophistication possible in such studies, especially through the use of high-level computer graphics, and should provide a strong impetus for a great increase in the number of functional studies on dinosaurs.
To date, computer applications in dinosaur biomechanics have been limited to studies of feeding mechanisms and locomotion. For example, Weishampel (1984) has used a three-dimensional kinematics computer program (developed by engineers) to analyze a series of ornithopod skulls as chewing machines. Heinrich et al. (1993) studied locomotion in the Late Jurassic iguanodontian Dryosaurus lettowvorbecki by modeling the femur as a beam. Bone cross-sections provided indications of both strength and patterns of loading on the living dinosaurs. Studying juvenile and adult specimens allowed them to postulate changes in locomotory patterns with age for that species.
Distributional studies analyze the distribution of organisms through time and space. Dinosaur studies of this kind have been very limited so far because of the nature of the fossil record for dinosaurs, but a number of paleontologists are now actively studying distributional problems with some success. To date, studies using computer databases have been able to track dinosaur diversity through space and time (Weishampel and Norman, 1989; Dodson, 1990) as well as the rate of study of dinosaurs by paleontologists (Dodson and Dawson, 1991). The large compilation by Weishampel (1990) of dinosaur localities is making it possible to analyze dinosaur paleobiogeography quantitatively for the first time (R. Chapman and D. Weishampel, work in progress). Such studies, however, can only be done using computers because they involve the mathematical and statistical analysis of large matrices of data. Clearly, this is one area of research that will be expanding greatly because of computers and related technology.
The interface between the general public and dinosaurs is one area of great change because of computers. Computers make available a wide range of educational approaches for teaching people about dinosaurs, especially using CD-ROM and multi-media technology. More people are also gaining access to data about dinosaurs through the Internet using online computer databases.
One of the biggest effects will be in changing the ability of the public to visualize what dinosaurs looked like. Conventional approaches of reconstructing dinosaurs (e.g., Paul, 1987) are being supplemented by sophisticated three-dimensional computer graphics that use computer visualization technology (e.g., Nielson and Shriver, 1990) to help generate lifelike dinosaurs such as those seen in the film Jurassic Park (Shay and Duncan, 1993) and can support the development of more life-like robotic dinosaurs (e.g., Poor, 1991). The development and distribution of better systems for virtual reality will allow researchers and the public alike to tour a dinosaur?s morphology and even view it from the inside [(see Fr?hlich, et al., (1995) and Stevens (1995) for discussions in the field of biology and medicine).
Once these approaches are developed, they will be used more in conjunction with the original fossil material within exhibitions. Most modern exhibits on dinosaurs include some computer technology and this will increase more with time. Clearly, computers can vastly improve how the general public is introduced to dinosaurs and increase their general knowledge.