NOTE AND COMMENT Patterns and propensities in reproduction and growth of marine fishes

Science in general, and ecology in particular, often proceed from the identification of patterns or propensities to the construction of theories, then back to the detailed analysis of the patterns, toward the development of more powerful theories (Peters 1992; Levin 1993), or the quantification of the propensities (Popper 1996). Thus, pattern recognition and explanation of the related propensities are keys to increasing ecological knowledge (Bakun 1996; Popper 1996). As advocated by Bascompte and Solé (1998), referring to spatiotemporal patterns in nature, ‘new stages in our understanding of ecological processes can only arise from a fertile interaction between theory and data’. However, questions may arise concerning the ubiquity of patterns, especially whether the same basic patterns occur in both terrestrial and marine environments. Even though the ocean and atmosphere share the same basic fluid dynamics, strong differences in mean density and viscosity are observed (Steele 1995). The sea consists of a fluid medium, and turbulent diffusion acts as the main dispersive mechanism for small organisms, which live in a three-dimensional environment; viscosity is a major constraint but gravity is not important. Furthermore, there is very little oxygen dissolved in even the best-aerated water, as compared with the O2 content of air. Because of inertia in the marine environment, sea temperature is much more predictable, even at small scales, than air temperature; overall, the ocean is less variable or ‘noisy’ than the atmosphere (Steele 1985, 1995). These features have important implications for the life-history of marine organisms, and we shall illustrate these using three examples of linked adaptations generated by these features, and involving the basis of life-history. The links between reproduction and recruitment in relation to the physical environment, between growth and respiration, and between mortality and growth Ecological Research (2000) 15, 101–106