The following HTML text is provided to enhance online
readability. Many aspects of typography translate only awkwardly to HTML.
Please use the page image
as the authoritative form to ensure accuracy.
Natural Climate Variability on Decade-to-Century Time Scales
species diversity shifts with respect to hydrography, for such concepts had not yet been well developed. Tont (1981, 1987, 1989), using Allen's 1930-1939 data, looked at seasonality and seasonal anomalies of 23 of the most abundant species of diatoms. He found negative correlations between a few of the species pairs; some of these appear to be significant, whereas others do not. He also cross-correlated seasonal anomalies of total diatoms and dinoflagellates with sea-surface and air-temperature anomalies. All of these yielded negative coefficients, some of which appear to be significant. Tont (1987) states ". . . at very best, however, climatic change, as defined by the variables used in this study, explains only 36 percent of the total variance. . . ." It is clear from his figures that cold years at the SIO pier are more productive than warm ones.
Sugihara and May (1990) have also used Allen's data (along with those for the incidence of chickenpox and measles in New York City from 1949-1972) to test their approach to making short-term predictions about the trajectories of chaotic dynamical systems. They show that "apparent noise associated with deterministic chaos can be distinguished from sampling error and other sources of externally induced environmental noise." They conclude that the method gives sensible answers with the observed time series of measles (deterministic chaos), chickenpox (seasonal cycles with additive noise), and diatoms (a mixture of chaos and additive noise). The dynamics of diatoms near shore reflect at least partial governance by a chaotic attractor. This result is new and valuable information that could only have come from such a time series. It is an indication of the degree to which the variability of plankton populations is driven by external, density-independent forces rather than by the ecosystem's internal nonlinear dynamics.
NATURAL AND ANTHROPOGENIC VARIABILITY
Most of the biological time-series data come from populations that are strongly influenced by humans, either directly by their harvesting forests, fish, and game, or indirectly through pollution or habitat disruption. Our inability to separate the effects of natural processes from anthropogenic ones has proven to be a serious impediment to progress in environmental management, the testing of theories of ecosystem behavior, or determining the role of climate in population variability. Conversi and McGowan (1994) attempted to make this separation through the use of spectral and time-series analyses of three geographically separated sites with similar kinds of sewage discharge and similar water-quality monitoring programs. Her approach was simple: to examine the relationships between the frequency spectra of sewage discharge and those of water transparency, particle concentration, and temperature within and between sites. Each site (Los Angeles City's wastewater treatment plant at Santa Monica Bay, Los Angeles County's at Palos Verdes, and San Diego County's at Point Loma) had a somewhat different 17-year history of discharge and different mass-emission curves. Conversi found that these were, in each case, uncorrelated with the local water-quality monitoring data. But the water quality did vary with time, especially at the lower frequencies. Most of this variations was seasonal (which the discharge was not), and the rest was correlated between sites and with temperature. There are known natural low-frequency coherent temperature variations in the area of the Southern California Bight. It appears that most of the variations in water clarity (a prime water-quality variable) were natural.
Radach et al. (1990) studied a long time series of temperature, salinity, plant nutrients, and phytoplankton samples taken 3 to 5 days per week at a single station in the German Bight in the southeastern North Sea. They observed conspicuous changes in the annual cycles of nutrients, a general increase in phytoplankton biomass, a shift in the ratio of flagellates to diatoms, and other evidence for a strong systematic change in the ecosystem. There is no evidence for concurrent changes in meteorological or hydrographic conditions. They attribute the observed changes to the anthropogenic increase in nutrient loads of the rivers that discharge into the coastal zone. On the basis of the changes over time in the magnitude, proportions, and pulse-like nature of the nutrient inputs, they suggested anthropogenic mechanisms for the observed shift in the Bight's carrying capacities for different kinds of phytoplankton.
Relatively few of the time series in the open ocean are of sufficient length and frequency to permit the exploration of important questions of population and community dynamics or the role of climate in influencing these. Many of the series have such large gaps that aliasing is likely, and some frequencies cannot be determined. Others track populations so affected by human activities that it seems unlikely that climate-induced variations can be separated from anthropogenic changes.
However, there are some marine populations for which the sampling has been long-term and of relatively high frequency, as with the CalCOFI and CPR programs. In both cases long-term, space-averaged data have shown that zooplankton variance has a red spectrum and that it is the interannual and interdecadal changes in abundance that are most important, although both study programs showed clear seasonality as well. The low-frequency variability was coherent over large geographic areas in both the Atlantic and Pacific. Both programs have shown significant correlations of plankton-biomass changes with large-scale climatic variations. There is strong evidence in both cases of comprehensive ecosystem response. The effects of the 1958-1959