Articles

The Nation's Rivers

Science  26 Nov 1971:
Vol. 174, Issue 4012, pp. 905-918
DOI: 10.1126/science.174.4012.905

Abstract

All measures of growth whether population, gross national product, waste loads, or demand for recreation suggest that demands on the country's water resources are increasing at a rate that exceeds the rate of installation of waste treatment facilities (Fig. 9). Presumably it is these demands which constitute the customary measure of "pollution" and upon which prophesies are made regarding the condition of the resource and its future prospects. Pressures posed by population and growth must have an effect on the river systems which, even with regulation, have finite assimilative or carrying capacity for wastes of different kinds. At the outset, then, one might assume, even in the absence of proof, that the very large increase in society's activities should have recognizable consequences. Unfortunately, it is easier to estimate the potential demand on a river system than it is to measure the effect. Certainly many rivers of the United States, like the rivers in Illinois, Colorado, and other places cited here, are not as they were 70 years ago. Nevertheless, while we "know" that conditions must be getting worse, we are hard-pressed to determine precisely the relation between the pressures posed by society and the responses of the river system.

To a large extent observations of the condition of river systems have been confined to measures of the quality of the flow itself and exclude descriptions of the bed, banks, and environs of the river. In some instances the parameters observed, for example, dissolved oxygen or coliform organisms, are surrogates for broader effects of greater interest. Although useful, they provide only partial measures of the river's condition or assimilative capacity (45). Thus in the Hudson River, while it can be shown that the change in dissolved oxygen is exceedingly small since 1922, we have some reason to believe that dissolved oxygen inadequately describes accumulating sludge on the bottom of the river as well as increasing growth of algae. Many years ago Henderson (46) demonstrated that a portion of the Shenandoah River was virtually devoid of life despite high oxygen levels and other quality standards of the water. Intended to surmount this disability, current attempts to quantify biological observations may not in fact do so unless they are associated with true biological surveys (47). Observation programs continue to emphasize measures of water quality to the virtual exclusion of measurements of river quality (48).

Observational programs appear to be particularly weak with regard to the detection of subtle initial changes from a natural to a polluted condition. This is not to suggest that no such observations are being made, for example on wild rivers, but rather that there is no systematic program for the detection of initial changes.

The initiation and progression of change are often considered impossible to prevent and inexorable in character. This leads further to the view that virtually all changes from a pristine to a used condition are irreversible. Such does not appear to be the case. Experience on the Ohio River described above, the reversal of changes in surfactant concentrations in the Illinois River, and the marked reduction of sediment in the Schuylkill River illustrate that it is possible to reverse or to improve the conditions of some rivers with respect to important pollutants. Fish may migrate rapidly to formerly polluted areas when either industrial shutdowns or large cleanup programs occur. The Thames River near London is reported to have fish for the first time in very many decades as a result of a major pollution control program. Indeed, a brief report suggests " that the condition of rivers in England and Wales . . . may well have improved since 1958" (49).

Because the evidence up to now does indicate that the pressure on the water resources posed by population and industrial growth significantly exceeds the rate of investment in control facilities, much higher cost must be incurred if even a rough parallelism between development and control is to be achieved. More important, the kind of technology that will be needed is not only more expensive, but may be distinctly different in kind from secondary treatment that is now projected for most municipal areas.

Even if treatment or pollution control were to maintain parity with pollution pressure, there is good reason to believe that many would continue to hold to the definition that " pollution is that which is getting worse." In the eyes of the beholders (50), the sights and smells of rivers today may be considered unsatisfactory, and while the "quality" of the rivers may be unchanged, they will be viewed as being polluted and becoming more so, as the interests of those living in more crowded areas turn toward rivers to satisfy their recreational and esthetic needs. Many observers have noted historic changes in values in the environment, and there is reason to believe that these play no small part in the current scene (51).

The paucity of information and the handful of investigators concerned with evaluating trends in the quality of the rivers of the United States suggest some specific conclusions. First, none of the observational programs were designed specifically to measure the quality of rivers or the river environment. The sampling programs emphasize the measurement of specific characteristics primarily related to water use by industry and municipalities. The new National Water Quality Network should improve on this single objective orientation (52). Few observational programs combine the necessary hydrology with measurements of water quality, river characteristics, and biology. While some long-term observations exist, the lack of coordinate observations makes longterm comparisons virtually impossible. For this reason, one must resort to the selected or case method described here. In addition, as Dworsky and Strandberg (53) emphasize, interpretation is "the vital part of the task of water quality assessment." Such interpretation requires the knowledge and skill of analysts familiar both with the data and with the changing characteristics of the land use and economy of the drainage basin. The new emphasis on quality of the environment demands continuing assessment and interpretation.

A second conclusion from the available data suggests that surrogate measures of river and water quality as wellas a multiplicity of measurements of easily measured parameters may shed little light on the dynamics of the processes active in river systems and hence such measures may be of limited use in estimating the likelihood of reversing specific observed trends in the absence of a knowledge of their causes. Additional attention must be given to the measurement of parameters related to models of river behavior and to estimates of inputs based on budgets of materials derived from industrial outputs and land use.

Third, while hydrologists have long been concerned with variability of the flow of natural rivers, because of the difficulty of observation, much less attention has been given to the variability of biological activity as well as physical variability associated with natural variations and cycles in rivers. Many measurements of biological effects are done during low and summer flows where measurement is easy, organisms often flourish, and concentrations of various substances in the flow are high. The effect of winter flow on the growth of slimes on the bottom of rivers, for example, and the special significance to the flora and fauna of periodic floods are not well documented. Significantly, however, among the most common trends in river management is the progressive regulation of flow through the provision of storage. Conceivably regulation rather than pollutants alone may have the most far-reaching effects on the character of many river systems. To date, observations have not been designed to measure these effects.

Because the demands on the waters of the rivers of the country are increasing, the concept of threshold and irreversibility must be studied on (i) pristine waters to disclose the nature of the initial, presumably biological, changes which take place and (ii) specific rivers where large scale control or cleanup programs have been initiated. It may well be that observations designed to detect "polluters," that is, observations designed to support the enforcement of standards, may not in themselves provide satisfactory measures of thresholds, trends, or reversals of trends. If one is to judge the effectiveness of the expenditure of large sums of money, observational tools must be designed to evaluate the response of the rivers to these expenditures.