(Falkowski, 1997) and their vertical migration may be critical in obtaining nutrients.

The classic view of the gyres is as superstable, very species-rich but biomass-poor ecosystems, both in the pelagic realm and on the seabed (Hessler and Jumars, 1974). Early in the exploration of the CNP, climax communities were a popular ecological concept, and the gyres looked like end-member examples. For this reason, early exploratory cruises and station locations were named ''Climax." Species diversity is high, and zooplankton samples taken years apart were as similar in species composition as samples from the same cruise (McGowan and Walker, 1985).

Flaws in the idea of oligotrophic gyres as nutrient impoverished appeared in the form of evidence that some phytoplankton were growing at high rates (e.g., Laws et al., 1984). Flaws appeared in the idea of constancy or static stability after the number of visits grew (Venrick et al., 1987), but disintegration of the idea came from time series funded under JGOFS (Joint Global Ocean Flux Study). The CNP showed two seasons of enhanced new production, one in winter based on enhanced physical mixing and one in summer based on enhanced nitrogen fixation, and these pulses showed substantial interannual variability (Karl et al., 1996). Further analysis of the time series and integration with all prior data suggest a doubling of primary production and a shift from dominance by eukaryotes to dominance by prokaryotes in the mid-1970s (Karl, in review). Some evidence links decadal-scale change in the Central North Pacific to the same large-scale ocean-atmosphere interactions that drive El Niño-Southern Oscillation (Karl et al., 1995). Not even this diverse community can resist basin-scale changes imposed by physics and chemistry.

Wind and eddy activity that can be important in bringing nitrate closer to the surface intuitively is unsteady and difficult to integrate over scales in time and space appropriate to the balance of nutrient budgets, and it is not hard to imagine that this component has interannual variability. The perspective shift underway, however, is that unusual lack of physical mixing also leads to enhanced new production, which is based instead on nitrogen fixation (Karl, in review) and on nitrate transport through vertical migration by mats of the diatom Rhizosolenia (Villareal et al., 1999). That is, the CNP's new production is minimized at some intermediate and probably "typical" input of physical energy, and lack of energy input leads to important biological "events." The theme of physical control of functional groups that effect drawdown of nutrients (including CO₂) extends to the Southern Ocean (Arrigo et al., 1999). Margalef (1978) must be pleased.

Time series clearly have power in exploring patterns of temporal variation and cross-correlation and have been key in shifting perspective away from stable, steady climax. They have made central gyres obvious places to improve the global ocean nitrogen budget, making the notion of oligotrophic seas as deserts even less tenable. To what extent nitrogen fixation is limited by phosphorus and trace metals (Falkowski, 1997), and to what extent it occurs in heterotrophic bacteria as well as cyanobacteria, is unclear (Karl, in press). Grazer influences on rates of nitrogen fixation and on the food web fates of microbially fixed nitrogen beg for exploration. The extent to which and reasons why higher trophic levels are more (or less) stable in composition than primary producers are unquantified. The success of Ironex II and newspaper reports of parallel successes in the Southern Ocean make large-scale manipulation of phosphorus and trace-metal concentrations a tantalizing prospect for oligotrophic gyres, and the existing time series can suggest the season and duration that would be effective. Making explicit, mechanistic, a priori predictions of consequences and their time scales is certain to enhance greatly the knowledge gained from any discrepancies observed.

• How are mass and momentum transfer and other environmental forces integrated with information to influence behavior?

• How does performance change with size and form?

Functional ecology is not a well-established term in the popular ecological lexicon, despite the fact that a journal of the British Ecological Society bears this name. Loosely, by contrast to numerical ecology, it refers to the performance of individuals in the context of environmental features, including other organisms. Perhaps the best-known functional responses of individuals are "filtering" and ingestion rates as functions of food concentration: rectilinear, hyperbolic, and sigmoidal responses have been described. As a consequence of rapid advances in understanding of mechanics of food encounter and handling, it can now be argued that it is better to use arrival rate of food items in the sensory field of the forager as the independent variable (instead of food concentration) in predicting ingestion rates. Ambient fluid motion, for example, can alter rate of encounter without any change in food concentration.

Among the most difficult phenomena about which to gain intuition are ones outside human sensory experience. Mechano- and chemosensing at low to intermediate Reynolds numbers may be among the most alien; hence they require accurate description before they can be appreciated and succinct, logical description before they can be intuited. Advances on the front of understanding mechanosensory ecology recently have been stunning; data with broad scatter suddenly have collapsed onto simple curves defined by sys