Nodule morphology and growth model.

To the Editor: A recent article in PNAS by Asikainen and Werle (1) reports on the occurrence and morphology of ferromanganese nodules found in a New Hampshire lake. The authors offer data on the shapes, banding, composition, and distribution of the nodules and crusts. They also state that my paper (2) claims that the morphology of such nodules is controlled by elemental concentrations and that this idea is not supported by their data. There is no such claim in ref. 2, or in ref. 3, which is the original reference. The model in ref. 3 explains the disk-shape of some lake nodules by the particular mass-transfer processes operating at the sedimentor nucleus-water interface. The model predicts that the morphology and growth rate will depend on the position of the nodule, the water velocity (boundary layer thickness) over the nodule for the period of growth, and the precipitating metal concentration in the water flowing over the nodule. Given this data, the model makes specific growthrate predictions. For example, for the conditions in Lake Charlotte, Nova Scotia, Canada, the prediction is 50 mm/ 100 years for continuous growth, which is not inconsistent with the limited evidence. To test the model in ref. 3, the stated inputs are required, and the resulting model predictions must be compared with observed shapes and rates. Asikainen and Werle (1) report none of the data needed in the model and make no comparisons. Logically, they cannot say that this model is not verified by their data or, by inference, that this model is wrong.