Origin of planktotrophyFevidence from early molluscs: a response to Nu ¨ tzel et al. (2006) Gary Freeman à and Judith W. Lundelius Section of Integrative Biology, University of Texas, Austin, TX 78712, USA à Author for correspondence (email: GFREE@mail.utexas.edu) There have been recent speculations in the literature about the nature of the trophic mode of Early Paleozoic marine faunas. It has been proposed that lecithotrophy is the primitive tro- phic mode in Cambrian invertebrates and that a shift to planktotrophy took place during the Late Cambrian–Early Ordovician in response to predation pressure or changes in the plankton (Signor and Vermeij 1994; Peterson 2005). Nu¨tzel et al. (2006) attempt to provide empirical evidence for this hypothesis in the Mollusca by analyzing the apices of Cambrian and Ordovician steinkerns (natural casts) repre- senting early ontogenetic stages. They describe the develop- mental criteria that define lecithotrophy and planktotrophy. In extant molluscs, a small shell is formed during embryo- genesis that approximates the size of the egg (protoconch I). Lecithotrophs have only protoconch I; their eggs are larger than planktotrophic ones because they contain a greater amount of yolk to nourish the embryo until metamorphosis. Planktotrophs generally form a smaller protoconch I and an additional shell is laid down during an extended feeding larval phase (protoconch II); together, these two protoconchs will be larger than protoconch I (Jablonski and Lutz 1980). Each phase can usually be identified by changes in ornamentation and a discernable growth discontinuity on the shell. When the larva undergoes metamorphosis, there is yet another discon- tinuity and change in ornamentation. Nu¨tzel et al. use internal casts of the apices of molluscan shells of Cambrian and Ordovician/Silurian age to make in- ferences about the trophic state of molluscan larvae based on size differences of the protoconchs. The external shell surface that would show details of the actual protoconch has been dissolved away by either diagenetic processes or by acid prep- aration. To compensate for this loss of detail, they selected a distance of 100 mm from the apex of the fossil and measured the diameter of the internal mold at that point. They claim that a diameter measurement at this length reflects the size of the protoconch as well as the amount of yolk and the on- togenetic strategy. No literature is cited to support this state- ment. They plotted a histogram of these diameters (their Fig. 3C), presumably as an approximation of the size of protoconch I, and concluded that the post-Cambrian proto- conch diameters are smaller than the protoconchs of Cam- brian specimens and thus represent the eggs of planktotrophs whereas the larger protoconchs of the Cambrian specimens were the consequence of yolkier lecithotrophic eggs. This re- sult would provide factual support for the initial appearance of lecithotrophy in the Mollusca in the fossil record. To make the quantitative comparison that the Nu¨ tzel et al. conclusion depends on, it is essential that the size comparison be made on equivalent protoconch stages and that the mea- surements be comparable estimates of protoconch size. Their methodology is problematic for several reasons: (1) Identifying and measuring protoconch I: Although they describe the development of protoconchs I and II and their significance, these distinctions are not used in evaluating their illustrated specimens or the measured samples. If the mea- surement protocol is applied to the specimen in their Fig. 1C, which has a bulbous apex that is probably protoconch I, there is a roughly 25% disparity between the diameter measure- ment made through the protoconch itself (59 mm) and a di- ameter measurement made at 100 mm from the apex (82 mm), which is beyond the area of protoconch I (Fig. 1). A single diameter measurement is not a good approximation of size especially where there are significant differences in protoconch shape. (2) Comparing Cambrian and post-Cambrian samples: The crux of their argument is the diameter data graphed in their Fig. 3C taken from 110 Cambrian and Ordovocian/Silurian steinkerns that shows a significant size distinction between the Cambrian and Ordovician/Silurian specimens. Although they state that the sample is of gastropods and other univalve molluscs, they do not identify the taxa and how these mea- surements were distributed among the different classes of Lower Paleozoic molluscs. Univalve molluscs included the monoplacophorans Tergomya and Helcionellida, Rostrocon- chia, Gastropoda, and the closely related Hyolitha (Kouchin- sky 2001). These groups had very different shell shapes that become evident at early stages of development. Gastropods and hyoliths have narrow, tubular shells where the aperture expands relatively slowly as the shell grows. The shells of monoplacophorans tend to have a bell shape and the aperture EVOLUTION & DEVELOPMENT 9:4, 307–310 (2007) & 2007 The Author(s) Journal compilation & 2007 Blackwell Publishing Ltd. 307