thing." I cannot claim to have originated the ideas because I was present both for discussions of future research by all 40 of the participants and for the task of writing them up. The vision necessarily is clouded, however, by the second and far less pleasant surprise of the meeting, that anthropogenic effects on marine ecosystems are ubiquitous and probably have been since the advent of commercial whaling. The third surprise on which I focus here is how quickly perspectives in all areas of ocean ecology have been changing since the meeting—and in the direction of making long-difficult problems suddenly more tractable.

For OEUVRE we chose to summarize research opportunities in subject matter categories that are cross-cutting and untraditional. I use the same headings here to facilitate the reader' s testing of my assertion of rapid progress against the OEUVRE report. Immediately after the headings, I reproduce the questions that OEUVRE identified under these headings as being both pressing and well poised for progress in the next two decades, and I spend the first few paragraphs in each section explaining the topic heading. I devote most space, however, to developing one or two examples of striking progress since the February 1998 meeting. I make no pretense of balanced coverage of the topic area or questions; scientific progress rarely is even across all fronts. Further, each participant or other ocean ecologist would be likely to choose somewhat different examples. My examples take highly variable space and referencing to develop, depending on the background provided in the OEUVRE report.

The exercise reveals several symptoms of substantial shift in perspectives. Unsuccessful search for a superstable marine ecosystem has ended. Along with this failure comes a new focus on how and why marine ecosystems change over time—and on which changes may contain an anthropogenic component. Coherent, succinct models are emerging of sensory systems and behaviors at spatial scales and Reynolds numbers for which humans have no native intuition at the same time that high-technology sensor systems are being deployed that allow unprecedented spatial and temporal resolution in human exploration of the sea. For many reasons, some of the most revealing exploration now is in time rather than space. Indeed this essay focuses more on how perspectives are changing than on summarizing old or new perspectives and is clearly derivative in that sense as well.

• How do environmental and biotic factors determine the distributions and activities of key species or functional groups important to biogeochemical cycles in space and time?

• What are the important interactions among marine biota, global climate, and biogeochemistry?

Species diversity has received broad attention for many good reasons. The intent of the topic heading was to focus on function and the mapping of biological diversity onto functional diversity in biogeochemical transformations. Analytic and predictive models of ecosystem function for the foreseeable future require some aggregation of organisms into functional groups (e.g., bacterivores or sulfate reducers). The level of aggregation that is useful depends on the question and discriminatory ability at hand, but it is clear that much effort remains to be spent on assigning organisms to biogeochemically functional groups, with due attention to taxonomy and physiology. Excitement is palpable about the maturation of DNA methodologies for both species identification and identification of potential to catalyze specific reactions (e.g., presence of genes that code for nitrogenase) and the maturation of RNA technologies that can assess whether that potential is being realized.

The central gyres of the ocean present many interesting questions. Collectively they constitute the largest habitat type on Earth and one of the oldest. Among them, the Central North Pacific Gyre (CNP) individually is the largest ecosystem on Earth, and my focus is primarily on this example. A small mistake in understanding of geochemical processes in such habitats can integrate into a large problem with global budgets. One geochemical problem of long standing is an inability to balance the fixed nitrogen budget for the global ocean. In most summaries, the loss terms exceed production substantially. Central gyres present a problem both individually and collectively, in that they seem to use more nitrogen in new production than can be accounted for (e.g., McGillicuddy and Robinson, 1997).

A parallel, long-running argument between geochemists and biologists has been about whether the oceans ultimately are limited in primary production by nitrogen or by phosphorus. The argument is partly a semantic one about what is meant by "ultimately" (i.e., time scale) and by "limited" (i.e., abundance or production of any or most phytoplankton), but it revolves around the issue that phosphorus has no substantial atmospheric source, whereas nitrogen is the largest component of the atmosphere. At its most simplistic, the assertion sometimes is made that because nitrogen-fixing organisms exist, the oceans cannot ultimately be nitrogen limited. Nitrogen fixation is notoriously expensive in energy and phosphorus, however (partly because it must be done anaerobically), and more sophisticated arguments revolve around whether rates of nitrogen fixation are slow enough to make phosphorus the effectively limiting nutrient because its availability restricts nitrogen fixation. There has long been evidence of phosphate as well as nitrate limitation in gyres, even of less demanding taxa than nitrogen fixers (e.g., Perry, 1972, 1976). Moreover, nitrogen fixers like Trichodesmium may well be limited or co-limited by iron