Report on the NATO Advanced Study Institute "Advances in Morpho- metrics", Il Ciocco, Italy, 18-30 July 1993.
During the last two weeks of July, the "hot spot" in morphomet- rics was Il Ciocco, a tourist and conference center north of Lucca in Tuscany, Italy. Those of us who were there certainly will re- member the magnificent view, the swimming pool, the evenings on the hotel terrace, and one of the most remarkable cakes in the history of morphometrics (for those who were not there: the figure is in the orange book). But we will also recall the lectures, the hours spent at the computer, and all the discussions on warps, landmarks, and shape change, which started before breakfast and ended late at night.
The first week started with a panel discussion about whether there is a "revolution in morphometrics -- real or imagined." The answers of the faculty members spanned a broad range, but whether they called the ongoing changes a revolution or synthesis, they all agreed that morphometrics is a very dynamic field. Richard Reyment and Fred Bookstein provided broader historical overviews about the origin of biometrics and the analysis of shape changes.
The first focus was morphometric data acquisition in two and three dimensions. There are numerous methods for fast and conve- nient recording of coordinate data. Some of these, like the video- based techniques, are already well-established in various fields and are used at various scales ranging from individual cells to dinosaur skulls. Others, like automated matching of space curves in computer tomography, are being developed for medical applica- tions and will have to be adapted to other morphometric applica- tions. Various devices were demonstrated, and participants could try them with their own specimens.
The remainder of the first week was mostly devoted to the core of landmark morphometrics: Procrustes analysis and the Thin Plate Spline. Initially, these approaches and all the concepts upon which they are based seemed to be a lot to learn. It was possible, however, because of the four-tiered teaching approach: there were lectures introducing the theory (often from two different view- points: geometric -- Fred Bookstein and algebraic -- Jim Rohlf), presentations of worked examples, demonstrations of software, and the trial-and-error exploration by participants playing on the computers. Affine and non-affine shape change, principal, partial, and relative warps were the buzzwords during coffee breaks and meals, and these topics even dominated the nightly discussions on the hotel terrace, e.g. when a real-life example for the Pinocchio effect was discovered. At the beginning of the second week, all these terms seemed to be a natural part of our vocabulary.
After the theoretical foundations of the new landmark morpho- metrics had been presented, they could be applied to biological problems and the new methods could be compared with other ap- proaches. A number of empirical examples contrasted conventional, distance-based methods with the landmark methods. In addition, outline methods, space curves, and various statistical techniques for analyzing morphometric results were introduced. Fred Bookstein presented his most recent development in landmark morphometrics, the edge elements (edgels), which provide a bridge to the outline methods.
Applications of morphometrics in quantitative genetics, allo- metry, and in phylogenetic analysis were discussed in several ses- sions during the second week. The discussions and examples showed that these are promising areas for both theoretical and empirical explorations. Special interest groups discussed additional appli- cations, e.g., new morphometrics in paleontology, conservation biology, and fluctuating asymmetry. The available methods are pow- erful tools for these fields, but need to be adapted to the spe- cific questions.
Now, after the ASI is over, we ask ourselves again: is there a revolution? The answers probably will still vary. There cer- tainly is, however, an unprecedented variety of methods and ap- proaches available in morphometrics, due to the innovative work in different fields and to the advances of computer technology. Mod- ern morphometrics is a synthesis of statistics, computer science, medical imaging, and evolutionary biology (and maybe more). It provides a framework to discuss the various approaches in a common language, and to illustrate findings with impressive graphics (e.g. David Dean's skull reconstructions or Jeff Walker's software Morphometrica). If the synthesis is to develop further, it will need continued interactions between workers in diverse fields. Meetings like the ASI have an important task: to bring researchers from the various disciplines together and to help them develop a broader understanding of the problems and possible solutions.
On the other hand, more nuts-and-bolts applications of morpho- metric techniques are needed to demonstrate their potential to solve everyday scientific questions. The Il Ciocco proceedings volume will contain a selection of examples, but a lot remains to be done. Also, morphometricians should present their results in journals and at meetings of their areas of specialization (the morphometric contributions at the meeting of the European Society for Evolutionary Biology in Montpellier were a fine example). Suc- cessful morphometric studies that solve old problems and generate new questions are the best way to convince researchers in various fields that it is worthwhile to learn and apply morphometrics.
Finally, I'd like to thank the organizers of the ASI, espe- cially Leslie Marcus, Stanley Marcus, and Marco Corti. (I apolo- gize to all whom I forgot to mention.) They made the conference a great success.