GEOLOGICAL SURVEY CIRCULAR 848-B

Biota and Biological Parameters as Environmental Indicators

Biota and Biological Parameters as Environmental Indicators

Edited by Phillip E. Greeson Briefing Papers on Water Quality

GEOLOGICAL SURVEY CIRCULAR 848-B

1981 United States Department of the Interior JAMES G. WATT, Secretary

Geological Survey Doyle G. Frederick, Acting Director

Library of Congress Cataloging in Publication Data Main entry under title:

Biota and biological parameters as environmental indicators.

(Briefing papers on water quality) (Geological Survey circular ; 848-B) Bibliography: p. Supt. of Docs. no.: I 19.4'2:848-A 1. Water quality bioassay. I. Greeson, Phillip E. II. Series. III. Series: Geological Survey circular ; 848-B. QE75.C5 no. 848-B [QH96.8.B5] 557.3s 81-607891 [628.1'61] AACR2

Free on application to Distribution Branch, Text Products Section, U. S. Geological Survey, 604 South Pickett Street, Alexandria, VA 22304 FOREWORD In August 1974, the Water Resources Division of the U.S. Geological Survey introduced the first of a series of briefing papers that were designed to increase the understanding of its employees of the significance of various aspects of water quality. Numerous briefing papers have been prepared by the Quality of Water Branch. Others will be prepared as the need arises. Each paper addresses a separate topic and is written in a nontechnical, easy-to-understand manner for distribution within the organization. Because of the favorable reception that the papers have received and their apparent effectiveness in accomplishing the objective stated above, it would appear that their wider distribution would serve a useful purpose. It is hoped that a wide range of persons, including those interested in the quality of our Nation's water resources, but who have little or no technical training, will find value in reading the papers. Furthermore, it is hoped that the papers will be suitable for supplemental reading in secondary education programs and in beginning college-level courses. The U.S. Geological Survey plans to publish several U.S. Geological Survey Circulars that contain compilations of briefing papers on particular aspects of water quality. This second Circular contains four papers describing selected biota and biological parameters as enviromental indicators.

Philip Cohen Chief Hydrologist

III

CONTENTS

Page Foreword ------III Abstract ------B1 Species diversity and its measurement, by Robert C. Averett------3 Algal growth potential,by W. Thomas Shoaf------7 Estimation of microbial biomass by measurement of adenosine triphosphate (ATP), by W. Thomas Shoaf ------11 Lake classification- Is there method to this madness?,by Dennis A. Wentz______15

ILLUSTRATIONS

Page FIGURE 1. Characteristic growth curves of a unicellular alga in a culture of limited volume___ B7 2. Summary of the major trophic types of lakes------18 3. Hypothetical curve of the course of natural eutrophication of lakes ______18 4. Effect of lake depth on the process of eutrophication------18

v

BRIEFING PAPERS ON WATER QUALITY

Biota and Biological Parameters as Environmental Indicators

Edited by Philip E. Greeson

ABSTRACT tains papers on selected biota and biological parameters as en­ viromental indicators. Briefing papers are included on "Species This is the second of several compilations of briefing papers diversity and its measurement," "Algal growth potential," on water quality prepared by the U.S. Geological Survey. Each "Estimation of microbial biomass by measurement of adenosine briefing paper is prepared in a simple, nontechnical, easy-to­ triphosphate (ATP)," and "Lake classification-Is there method understand manner. This U.S. Geological Survey Circular con- to this madness?"

Bl

Species Diversity and its Measurement

By Robert C. Averett

ABSTRACT environment. In a similar manner, the. presence or Wit? the decrease in popularity of the single-species indicator­ absence of specific types of animals was used to o:gam.sm concept, there arose an understanding in the natural delineate certain climatic conditions. The use of diversity of organisms. This, combined with the science of infor­ these biological indices became more popular as ma~ion theory, led to the development of formulae for com­ more knowledge was gained about the distribution ~u~ng. a diversity index. Interpretation of an index value has hmitatlons; but, when used properly, it is a valuable tool for the and abundance of organisms. assessment of water quality. The first comprehensive work on using aquatic organisms to define water-quality conditions was INTRODUCTION by two German biologists, Kolkwitz and Marsson (1908, 1909). They published findings on a stream Man has always marveled at the number of dif­ that received domestic wastewater. From the ferent types of organisms on Earth. Sketches on standpoint of organism types and diversity, they cave walls are testimony that early man not only were able to divide the stream into four distinct knew about many types of animals, but also knew reaches, which they called zones: the clean water someth~ng about their roles in nature. Today we zone, the septic zone where the sewage is first in­ recognize about 37,000 species of micro­ troduced, the zone of active sewage decomposition, organisms, 341,000 species of plants, and and the zone of recovery. 1.'358,000 species of animals. Taxonomists (scien­ Other workers followed this classification pro­ ~Ists that classify organisms) continually are add­ cedure, or modified it slightly to fit their own find­ I~g to these numbers as new organisms are ings. Each zone of the stream was inhabited by discovered. The insects are the largest single distinct groups of organisms. Moreover, in those grou~. of org~nisms and have been only partially zones where the waste load was heaviest and classified. It IS estimated that if all the different where the stream bottom was covered by waste kinds of insects on Earth were classified at this solids, the number of organisms per unit area was time, they would exceed 2 million species-a figure very high, but the number of different kinds of that exceeds the known types of all other animals. organisms was very low. In other words, there The biological definition of a species is a group of were many organisms, but the diversity (number of actually or potentially interbreeding organisms species) was low. Further, nymphs and larvae of repro~uctively isolated from other such groups of some stream insects (particularly the mayflies, orgamsms. A true species must meet this criterion. stoneflies, and caddisflies) were frequently absent It is this concept of a species that man uses in his from areas receiving heavy loads of organic waste husbandry of domestic animals, and in his use of but were present in clean water reaches. Much of organisms in defining a particular environment. this absence resulted from the alteration of the physical stream bottom by waste sludge. In addi­ CONCEPT OF SINGLE·SPECIES INDICATORS tion, low oxygen concentrations and an altered The use of organisms to indicate certain en­ food suppy prevented them from living in the vironmental conditions stemmed from the work of heavily polluted section of the stream. In contrast, botanists of the 18th and 19th centuries. Plants, some trueflies (for example, midges, gnats), that because they responded to temperature and other were adapted to living in the sludge of a stream climatic vari~b~es, as well as to soil types, were bottom and in water that had a low oxygen concen­ used as an aid m the classification of the physical tration, thrived, and in the absence of competition,

B3 their numbers per unit areas was very high. These objective of information theory is to measure the findings led to the single-species indicator­ amount of order or disorder in a system. In the organism concept whereby specific groups of biological community, we may wish to measure a organisms were unadapted to pollution, but others number of things, but for our purposes the most were able to tolerate significant changes in the en­ important would be the number of species and the vironment. number of individuals in each species. That is, we The single-species indicator-organism concept may want to look at the richness or variety of the had several failings. First, it was based only upon community structure and ask the question, "How organic (oxygen-demanding) waste loadings. difficult (or easy) will it be to correctly predict the Sublethal levels of toxic materials (some industrial species of the next individual in the sample?" This and agricultural wastes) were ignored, yet they is similar to a message receiver predicting the next also may be preferential as to which organisms symbol, for example, dot or dash, in a Morse code they kill. All organisms, including man, tolerate message. Because it is dot-dash, on-off, or yes-no, varying concentrations of potentially toxic relations can be represented by 0-1 digits, or materials, and a concentration that is toxic to one binary digits. They often are measured in terms of organism, or a group of physiologically similar bits (for number of binary digits). organisms, may not be toxic to others. Conse­ The information theory diversity index IS quently, there may be great variability in the ef­ calculated using the function: fects of toxic materials on the distribution and s abundance of aquatic life. o=- E p;lOgzp;= - E (niln)logz(niln) (1) The second failing was that organisms that make 1=1 up the groups found in the various stream reaches where: o = species diversity or information con­ have had to come from somewhere. That is, they tent of the sample (bits/individual) .. are all natural members of the stream fauna, but, The term o often is denoted by H. when the stream is altered, some find the habitat s =number of species in the sample more suitable and their numbers increase. Those p;=niln=proportion of the total sample belong­ that find the habitat less suitable die or leave the ing to the ith species. area; for example, the presence of truefly larvae in This function, developed simultaneously but in­ a sample does not indicate pollution or water­ dependently, by Wiener (1948) and Shannon and quality degradation. The single-species indicator­ Weaver (1949), is a measure of the amount of organism concept, therefore, is not sound unless uncertainty. The larger the amount of uncertainty practiced by someone thoroughly knowledgeable in and, hence, the greater the diversity of the species, the habitat requirement and physiology of the the greater the diversity or the value of o. For ex­ organisms. As a general biological technique for ample, a message such as aaaa has no uncertainty, assessing water quality, it has severe limitations. and, thus, no diversity and, therefore, o = 0. In a CONCEPT OF SPECIES DIVERSITY sample where all organisms are of the same species (a rare event in nature), the diversity would be At about the time that the single-species zero. In other words, in a single-species sample, indicator-organism concept was losing its populari­ one individual would tell as much about the whole ty (not because it was completely erroneous, but sample as would any other individual, or all of the because it was too often misused), the species­ individuals combined. In a sample of many species, diversity concept was becoming popular. This con­ we would have to identify more organisms, if we cept considered the entire community of aquatic wished to exhaust all of the information from the organisms and, thus, was sound from a biological sample. Thus, a sample in which each organism standpoint. Initially, the concept consisted of sim­ belonged to a different species (again, a rare event ply counting the number and types of organisms in · in nature), would have a high diversity and o would a given sample. Later, the ratio of the number of be a positive number. Refer again to our example organisms to the number of species was used as a where the message aaaa has no uncertainty. In simplified species-diversity index. While this contrast, the message {3acf.>'Y has a high uncertain­ technique also provided a quantitative approach, it ty, and predicting the next letter would be much was difficult to interpret the results because the more difficult, and o would be some positive ratios often changed greatly as the number of number. Note that logarithms to the base 2 are organisms in the sample increase. used so as to obtain bits per individual. Later, the science of information theory was ap­ The diversity index, o, is influenced by the plied to biological sampling problems. Briefly, the number of species in the sample as well as by the B4 evenness of the distribution of the species in the quality changes. Rather, it is a tool, and, like a tool, sample. The measure of evenness is attained first it must be used correctly to be effective. There are by calculating the maximum diversity, o max, as: precautions that must be observed. First, the level s of taxonomic level above species is valid. While omax=- E Glog2!)= log2s (2) desirable, it is not necessary to classify all 1=1 organisms to the species level. Second, the physical The highest value of o max is attained when the habitat, such as velocity of flow and stream bottom number of individuals per species is distributed material must be considered and reported. equally -that is, when all the species are present Streams that have various sizes of bottom material in the same proportion!. In two samples in which will have a greater diversity of organisms than the species are distributed evenly, the sample with streams that have material all of a similar size. the largest number of species will have the Third, the season of the year and life history of the greatest diversity (Pielou, 1969). From equations 1 organisms must be considered. Some organisms and 2, a measure of the evenness of the species in have annual periods of dominance-for example, the sample (J) can be calculated as: immediately after egg hatching, a particular group of organisms may be abundant, but shortly J=--0 3 0 max ( ) thereafter, natural mortality factors take their toll The diversity index, o, has obvious advantages and reduce the number of individuals. Fourth, the over the single-species indicator-organism con­ method of collection and the mesh size of the cept, or the simple method of counting the number screen used to separate the organisms is of utmost of species in a sample. First, it is a dimensionless importance. It is invalid to compare diversity in­ equation, and numbers or biomass (weight per dices where different collecting gear and screen volume or area) can be used. Second, the impor­ sizes have been used. Consequently, comparisons tance of each species in the community is ex­ of diversity indices must be between similar tax­ pressed as a ratio representing the contribution of onomic categories, similar substrates, and at each species to the total diversity or biomass. similar times of the year, using similar collecting Third, and of great importance, diversity indices and separating techniques. derived from information theory are virtually in­ Regardless of these limitations, the diversity in­ dependent of sample size. The occurrence of rare dex (d) is a valuable tool, and with it a more de­ species, as additional samples are collected, does tailed quantitative assessment of water quality can not greatly change the value of o. be made. It does not, however, explain the role of In a group of 10 samples from four habitats (one the individual organism in the community, the of which was a sample board), Wilhm and Dorris physiological stress that water-quality changes (1968) found that the diversity reached 95 percent place upon the organisms, or the physiological of the asymptotic diversity (that value of o that stress that the use of the water may place upon essentially does not change with the addition of man. These processes will never be explained with new samples) before five samples were collected. something as simple as the diversity index. Happi­ In a subsequent paper, Wilhm (1970) reported that ly, continued research is providing information on the diversity index ranged from 0 to 10, but usually the habitat, behavorial, and physiological re­ was less than 5. In streams that do not receive quirements of aquatic organisms and ultimately waste material (clean water zones), it was between will lead to a greater understanding and use of 3 and 4, while, in streams or stream reaches receiv­ species diversity and the diversity index as one tool ing organic waste, the diversity-index was usually for assessing water-quality changes. less than 1. Wilhm (1970) further suggested that the asymptotic diversity value should be used. REFERENCES Generally, as mentioned, this occurs between two Kolkwitz, R., and Marsson, M., 1908, Okologie der and five samples. pflanzlichen Saprobien: Gerichte der Deutschen Botanischen Gesellschaft 26a, p. 505-519. (Translated 1967, Ecology of plant saprobia, p. 47-52, in Keup, L. E., LIMITATIONS OF DIVERSITY INDEX Ingram, W. M., and Mackenthun, K. M., eds., Biology of VALUES water pollution: Federal Water Pollution Control Ad­ ministration, U.S. Department of the Interior, 290 p.). While the diversity index has a number of --, 1909, Okologie der tierischen Saprobien: Beitrage positive attributes, it also has its limitations. It is zur Lehre von der biologischen Gewasserbeurteilung. Inter­ not a panacea for the problem of measuring water- nationale Revue der Gesamten Hydrobiologie and

B5 Hydrogeographie 2, p. 126-152. (Translated 1967, Ecology Warren, C. E., 1971, Biology and water pollution control: of animal saprobia, p. 85-95, in Keup, L. E., Ingram, W. M., Philadelphia, Pa., W. B. Saunders Co., 434 p. and Mackenthun, K. M., eds., Biology of water pollution: Wiener, Norbert, 1948, Cybernetics: Cambridge, Massa­ Federal Water Pollution Control Administration, U.S. chusetts Institute of Technology Press, 194 p. Department of the Interior, 290 p.). Wilhm, J. L., 1970, Range of diversity index in benthic Margalef, Ramon, 1968, Perspectives in ecological theory: macro-invertebrate populations: Journal of the Water Pollu­ Chicago, Ill., University of Chicago Press, 111 p. tion Control Federation, v. 42, no. 5, pt. 2, p. R221, R224. Pielou, E. C., 1969, An introduction to mathematical Wilhm, J. L., and Dorris, T. C., 1968, Biological param­ ecology: New York, John Wiley and Sons, 268 p. eters for water quality criteria: Bioscience, v. 18, no. 6, p. Shannon, C. E., and Weaver, WarrenJ 1949, The mathe­ 477-481. matical theory of communication: Urbana, Ill., University of Illinois Press, 177 p.

B6 Algal Growth Potential

By W. Thomas Shoaf

ABSTRACT desirable. The presence of these algae and Algal growth potential (AGP) is defined as the maximum algal associated micro-organisms suggests that they are mass that can be produced in a natural water sample under suited to survive the environment in which they standardized laboratory conditions. Measurements of AGP can be used for a variety of purposes; however, an understanding of live, and their potential for accelerated growth is the principles of the test and the factors that affect its expres­ of primary interest in many situations. The sion is critical to proper data interpretation. primary difficulty in using indigenous algae is the time required for measuring growth of diverse DEFINITION forms of algae such as colonial, filamentous, or Algal growth potential (AGP) is defined as the univellular forms. In some instances, water maximum algal mass (dry weight) that can be pro­ samples contain few algae, and the introduction of duced in a natural water sample under standard­ foreign algae would be needed to insure a growth ized laboratory conditions. The characteristic response. growth pattern for unicellular algae in a culture of The AGP of a number of samples can be com­ limited volume is shown in figure 1. pared only when variables which affect algal AGP is the algal mass present at stationary growth, such as light and temperature, are stand­ phase and is expressed as milligrams per liter dry ardized and controlled. The need for standardiza­ weight of algae produced. Dry weight was chosen tion of techniques for measuring AGP for routine as a measure of algal growth because it is the most analysis was recognized by the Joint In­ reliable parameter for all sample sources. dustry/Government Task Force on Eutrophication (Bartsch, 1969). A cooperative developmental ef­ STANDARDIZATION OF ANALYSIS fort by the Soap and Detergent Association and the U.S. Environmental Protection Agency to pro­ Determination of the AGP of a water sample us­ duce a test for determining AGP in natural water ing the natural algae present is perhaps most samples was the result (Bartsch, 1971). Selenastrum capricornutum was selected as a test organism for the following reasons: Stationary phase 1. It is a unicellular alga and its growth can be monitored accurately and rapidly with an elec­ tronic particle counter. 2. Selenastrum has been shown to tolerate acidic and alkaline waters as well as oligotrophic (nutrient-poor water) and eutrophic (nutrient­ rich water) conditions in aquatic environments (Forsberg, 1972). 3. It is listed among the most pollution-tolerant algae by Palmer (1969). 4. The AGP determined with S. capricornutum and the potential determined with the species indigenous to the water sample correlate well AGE OF CULTURE (Maloney and others, 1972). FIGURE 1.-Characteristic growth curve of a unicellar alga in a 5. The AGP for various freshwater environments culture of limited volume. using S. capricornutum, as measured by the

B7 standard test (Bartsch, 1971), varies widely limiting. The determination of how much and what from 0.1 to 248 milligrams dry weight of algae to spike may be made easier when the concentra­ per liter (National Lake Survey Program, tion of the nutrient(s) of interest is known. Because 1973). algal growth in natural water samples can be limited by one or several nutrients in close succes­ APPLICATION sion, it might be necessary to test spikes of com­ binations of nutrients that could be limiting. The design of a project is developed to meet a When only one nutrient is limiting algal growth, specific objective. All pertinent factors must be it is likely that a correlation will be observed be­ considered in planning if valid results and conclu­ tween the concentration of that nutrient and the sions are to be obtained. It should be possible to AGP. Some AGP studies have shown correlations evaluate AGP data more effectively in light of between orthophosphate and algal growth (Wang other biological, chemical, and physical and others, 1972), while correlations also are likely measurements obtained by sampling at common for other nutrients in other aquatic environments. sites. Various collection methods for streams and lakes have been described by Guy and Norman (1970), Greeson and others (1977), and Greeson INTERPRETATION OF RESULTS (1979). The objectives of AGP tests might include the AGP data which are derived in the laboratory establishment of baseline data or determination of under controlled conditions of light and the growth-controlling nutrient(s), the influence of temperature do not necessarily reflect conditions a nutrient on algal growth, or the source of a in the natural aquatic environment from which the nutrient when several inflows are involved. One of sample was taken. Only the potential at a given the first items that should be determined when time can be measured. In nature, daily and measuring AGP is an understanding of what con­ seasonal variations in the intensity and duration of stitutes a significant change. A certain degree of light occur, which can effect algal growth. Changes variability can be expected with any assay or in other environmental factors can be important as survey; however, once the baseline data are well. For example, the rate of algal growth roughly established, experimentation can be conducted. doubles with every 11 oc (20°F) rise in water The principle of the AGP determination is based temperature between 0°C (32°F) and 32°C (90°F) on Liebig's Law of the Minimum, which recognizes (Ferguson, 1968). Growth can be inhibited by any that the development of a population is essentially suspended material, living or nonliving, which in­ regulated by the substance occurring in minimal terferes with the effective penetration of light quantity relative to the requirements of the popula­ essential to the algae. Grazing by invertebrates tion.AGP measurements have been used frequently and fish, toxic materials entering the water, plant to derive information on the required nutrients or animal excretions, or decay products from the which are limiting algal growth (Leischman and algae themselves are all additional examples of fac­ others, 1979). This type of measurement usually is tors that may inhibit growth potential. designed to examine in detail only a few nutrients In the standard procedure described by Bartsch which preliminary testing indicates may be (1971), Miller and others (1978), and Greeson limiting or in short supply. The approach is to (1979), membrane filtration of the water sample to observe the effect that certain growth promoting remove indigenous algae, bacteria, fungi, and any substances (nutrients) have upon the algae. These other particles which might interfere with the pro­ substances may be of precise known composition cedure is required. These micro-organisms contain (Bartsch, 1971; Miller and Maloney, 1971; Maloney nutrients, which are not available to other algae and others, 1972) or their composition may be while these organisms are living, but later become somewhat less known, such as that of an effluent a source of nutrients as a result of decay after which may flow into a lake or stream from a death. Thus, it is possible to measure a high con­ suspected nutrient source (Brown and others, centration of algae during a "bloom," but observe a 1969). The practice of adding such substances to a low AGP. After cell death and decay, the AGP water sample is called spiking. In the case of probably will increase. This, as well as the other chemical spikes, it is obvious that if increasing factors mentioned above, must be taken into con­ amounts of a limiting nutrient are added to a water sideration when attempting to make predictions sample, eventually another factor becomes about future algal growth.

B8 SUMMARY Greeson, P. E., ed., 1979, A supplement to- Methods for collection and analysis of aquatic biological and Algal growth potential is defined as the max­ microbiological samples (U.S. Geological Survey Techniques imum algal mass (dry weight) that can be produced of Water-Resources Investigations, Book 5, Chapter A4): U.S. Geological Survey Open-File Report 79-1279, 92 p. in a natural water sample under standardized Greeson, P. E., Ehlke, T. A., Irwin, G. A., Lium, B. W., and laboratory conditions. AGP measurements are Slack, K. V., eds., 1977, Methods for collection and analysis designed to establish baseline data, growth of aquatic biological and microbiological samples: U.S. limiting factors (nutrients), and the influence and Geological Survey Techniques of Water-Resources In­ source of various growth promoting nutrients, so vestigations, Book 5, Chapter A4, 332 p. Guy, H. P., and Norman, V. W., 1970, Field methods for the as to provide improved means for predicting and measurement of fluvial sediments: U.S. Geological Survey controlling excessive algal growth in aquatic Techniques of Water-Resources Investigations, Book 3, habitats. Data can be compared only when the Chapter C2, 59 p. variables that control algal growth are standard­ Leischman, A. A., Greene, J. C., and Miller, W. E., 1979, ized. Algal growth potentials derived in the Bibliography of literature pertaining to the Genus Selenastrum: Corvallis, U.S. Environmental Protection laboratory may not reflect natural conditions Agency, EPA-60017-79-021, 191 p. because of insufficient light or temperature, graz­ Maloney, T. E., Miller, W. E., and Shiroyama, Tamotsu, ing by invertebrates or fish, or the presence of tox­ 1972, Algal responses to nutrient addition in natural ic materials. An understanding of the principle of waters. I. Laboratory assays: Nutrients and eutrophication, the test and the factors that affect the expresion of Special Symposium, American Society of Limnology and AGP is critical to proper data interpretation. Oceanography, v. 1,p. 155. Miller, W. E., and Maloney, T. E., 1971, Effects of secondary and tertiary wastewater effluents on algal growth in a lake­ REFERENCES CITED river system: Journal of the Water Pollution Control Federation, v. 43, p. 2361-2365. Bartsch, A. F., 1969, Provisional algal assay procedure: Cor­ Miller, W. E., Green, J. C., and Shiroyama, Tamotsu, 1978, The vallis, Oreg., U.S. Environmental Protection Agency, 61 p. Selenastrum capricornutum Printz algal assay bottle --1971, Algal assay procedure: Bottle test: Corvallis, test-Experimental design, application, and data inter­ Oreg., U.S. Environmental Protection Agency, 82 p. pretation protocol: Corvallis, U.S. Environmental Protec­ Brown, R. L., Bain, R. C., and Tunzi, M. G., 1969, Effects of tion Agency, EPA-600/9-78-018, 125 p. nitrogen removal on the algal growth potential of San Joa­ National Lake Survey Program, 1973, Highlights: Corvallis, quin valley agricultural tile drainage effluents: Paper Environmental Protection Agency, 5 p. presented at the American Geophysical Union National Palmer, C. M., 1969, A composite rating of algae tolerating Meeting, San Francisco, 15 p. organic pollution: Journal of Phycology, v. 5, no. 1, p. Ferguson, F. A., 1968, A nonmyopic approach to the problem of 78-82. excessive algal growth: Environmental Science and Wang, W. C., Sullivan, W. T., and Evans, R. L., 1972, A Technology, v. 2, p. 188-193. technique for evaluating algal growth potential in Illinios Forsberg, C. G., 1972, Algal assay procedure: Journal of the surface water: Urbana, Illinois State Water Survey, Report Water Pollution Control Federation, v. 44, p. 1623-1628. of Investigation 72, 16 p.

B9

Estimation of Microbial Biomass by Measurement of Adenosine Triphosphate (ATP)

By W. Thomas Shoaf

ABSTRACT a biochemical compound) that is present in all liv­ Andenosine triphosohate is present in all living cells and is the ing cells but is not associated with nonliving par­ primary energy donor in celluar processes. The measurement of ticulate material. This cellular substance should the amount of andenosine triphosohate present in a water sam­ have a short survival time after the death of a cell, ple, therefore, is a measurement of the total biomass in the sam­ so that it would be specific for viable (living) ple. The firefly luciferase reaction with luciferin is the common­ ly used method for determining the presence and amount of biomass. It should be present in proportion to some adenosine triphosphate in a sample. The method is very sen­ measure of the total biomass for all micro­ sitive and rapid. organisms: algae, bacteria, fungi, and protozoans. Sensitive and accurate methods for its detection INTRODUCTION should be available. The biochemical compound which seems to meet For many ecological studies, it is important to these requirements best is adenosine triphosphate know the total amount of living material or (ATP). ATP is the primary energy donor in cellular biomass present in an ecosystem. When dealing life processes. Its central role in living cells and its with larger animals and plants, it is relatively easy biological and chemical stability make it an ex­ to measure biomass by conventional methods, such cellent indicator of the presence of living material. as collecting and weighing them. For the micro­ The level of endogenous ATP (that is, the amount organisms, however, it can become very difficult. of ATP per unit biomass) in bacteria (Allen, 1973), Both phytoplankton and zooplankton biomasses algae (Holm-Hansen, 1970), and zooplankton can be estimated by direct microscopic determina­ (Holm-Hansen, 1973) is relatively constant when tion of volume and number. Bacteria can be. compared to the cellular organic carbon content in estimated by various plating techniques (colony several species of organisms. Furthermore, its con­ counting) and extinction dilution techniques. The centration in all phases of a growth cycle remains techniques are often time-consuming, expensive, relatively constant. In studies where cell viability and subject to inherent sources of error. Some of was determined (Hamilton and Holm-Hansen, the major problems include the following: Direct 1967; Dawes and Large, 1970), the concentration microscopic counting of bacteria may yield inac­ of ATP per viable cell remained relatively constant curate biomass estimates because of the difficulty during periods of starvation. The quantity of ATP, in distinguishing bacteria from bacteria-sized inert therefore, can be used to estimate total living particles. There is, also, the inability to differen­ biomass. tiate between living and nonliving cells, and the problems of cell aggregation characteristic of ANALYTICAL DETERMINATION OF ATP bacteria and algae. Low estimates may be pro­ duced by plating and extinction techniques which Very sensitive methods of ATP determination are selective because of the chemical composition have been developed from the finding that of the media and variations in physical parameters luminescence in fireflies has an absolute require­ such as temperature and pressure. The extinction ment for ATP (McElroy, 1947). The ATP is deter­ dilution technique also may be biologically selective mined by measuring the amount of light produced in that one form might overgrow the other. when ATP reacts with reduced luciferin (LH2) and

For a total biomass determination, it would be oxygen (02) in the presence of firefly enzyme desirable to measure some substance (for example, luciferase and magnesium, producing adenosine

Bll monophosphate (AMP), inorganic pyrophosphate (bacteria, protozoans). Although the ranges of (PPi), oxidized luciferin (L), water (H20), carbon values are yet unestablished, low values of the dioxide (C02), and light (hv), by the following reac­ ATP to chlorophyll ratio should be indicative of tion: autotrophic conditions, whereas high values imply ATP + 0 + LH luciferase AMP+ PPi + L a more heterotrophic community. 2 2 Mg+2 Since ATP can be correlated directly with living +H20+C02+hv. carbon (Holm-Hansen, 1973), another tool for ac­ The bioluminescent reaction is specific for ATP cessing biological water quality is available in the and the reaction rate is proportional to the ATP ratio of living carbon (determined by ATP) to total concentration; one photon of light is emitted for organic carbon. This ratio may be a useful measure each molecule of ATP hydrolyzed. When ATP is in­ of the degree of eutrophication. In an oligotrophic troduced to suitably buffered (with respect to pH) situation, a high percentage of the total organic enzymes and substrates (reduced luciferin and ox­ carbon is in living cells and, thus, the ratio of living ygen), a light flash follows; it decays in an exponen­ carbon (determined by ATP) to total organic car­ tial fashion. Either the peak height of the light bon will be at its highest. As conditions move flash or integration of the area under the decay toward the eutrophic, more nonliving organic car­ curve can be used to form standard curves. Quan­ bon accumulates and the percent of living carbon tities of ATP as low as 10-12 to 10-13 grams of ATP (to the total) decreases. Determination of this ratio can be measured accurately. may, therefore, be a good method for describing the overall biological condition. APPLICATION ATP analysis has been used to estimate the SUMMARY biomass at varying depths in the ocean (Holm­ The determination of biomass (total living Hansen and Booth, 1966). The ATP levels found at material) has been a difficult task for investigators various depths in the ocean indicate bacterial in the field of aquatic microbiology. Measurement populations between 50 and 2,000 times those of ATP provides a very sensitive and rapid found by plating techniques. This may reflect the estimate of biomass. While useful as an estimate of inadequacies of plating techniques referred to biomass alone, it also may be related to determina­ above. tions of chlorophyll and total organic carbon to The effect of chemical or thermal pollution on a provide ratios which give more information about system has been determined by passage of water biological water quality. through plant cooling systems and by monitoring the before and after levels of ATP. In such a study on the Chesapeake Bay, ATP appeared to be a REFERENCES CITED reliable measure of biomass viability (Dupont Cor­ poration, 1970). The effects of industrial plants and Allen, P. D., 1973, Development of the luminescence biometer for microbial detection: Developments in Industrial cities on streams can be evaluated in a similar Microbiology, v. 14, p. 67-73. fashion. Likewise, the effects of suspected (or Dawes, E. A., and Large, P. J., 1970, Effect of starvation ·on known) biocides on micro-organisms also could be the viability and cellular constituents of Zymomonas measured immediately, without the usual delay en­ anaerobia and Zymomonas mobilis: Journal of General countered with plate counts (for bacteria) or Microbiology, v. 60, p. 31-40. Dupont Corporation, 1970, Water ecology/waste water treat­ phytoplankton anaylysis (for algae). ment: Wilmington, Del., Dupont Applications Brief no. Biomass determination by ATP measurement 760BF1 (A 72069), 2 p. can be correlated with measurement of chlorophyll Hamilton, R. D., and Holm-Hansen, 0., 1967, Adenosine where the dominant life form is phytoplankton triphosphate content of marine bacteria: Limnology and (Holm-Hansen, 1969). If bacteria or protozoans are Oceanography, v. 12, p. 319-324. Holm-Hansen, 0., 1969, Determination of microbial biomass present .in large numbers and not evenly in ocean profiles: Limnology and Oceanography, v. 14, p. distributed with the phytoplankton, then a correla­ 740-747. tion with chlorophyll would not be expected (Holm­ . --1970, ATP levels in algal cells as influenced by Hansen and Paerl, 1972). A ratio of ATP (biomass) environmental conditions: Plant Cell Physiology, v. 11, p. to chlorophyll can be useful for determining 689-700. --1973, Determination of total microbial biomass by whether the living matter in a sample is measurement of adenosine triphosphate, in Stevenson, L. predominantly autotrophic (mainly algal) or con­ H., and Colwell, R. R., eds., Estuarine microbial ecology: tains a large number of heterotrophic organisms Columbia, University of South Carolina Press, p. 73-89. B12 Holm-Hansen, 0., and Booth, C. R., 1966, The measurement of and metabolic activity: Memoirs of 1st Italia Idrobioliae, adenosine triphosphate and its ecological significance: Lim­ Supplement 29, p. 149-168. nology and Oceanography, v. 11, p. 510-519. McElroy, W. D., 1947, The energy source for bioluminescence Holm-Hansen, 0., and Paerl, H. W., 1972, The applicability in an isolated system: Proceeding of National Academy of ATP determination for estimation of microbial biomass Sciences, v. 33, p. 342-346.

Bl3

Lake Classification-Is There Method to this Madness?

By Dennis A. Wentz

ABSTRACT (It should be pointed out that the relationships The number of scientific publications discussing lake sometimes are presumed.) Thus, classification is classification schemes has increased considerably since the early the process of defining the groups into which ob­ 1970's. This has resulted primarily from the enactment of the Federal Water Pollution Control Act Amendments of 1972. jects are placed. The process of assigning an object This paper discusses lake classifications based upon geologic, to a previously defined group, on the other hand, hydrologic, and trophic schemes. Many of the schemes have can be described as categorization or identifica­ limited applicability, while others are adaptable to a variety of tion. situations. The next question that logically arises is "Why INTRODUCTION classify?" In general, there seem to be two basic The study of lakes has developed into a popular reasons. The first is that the data base is too large hydrological pursuit in recent years, and there are to consider conveniently. Classification results in a those who would encourage even greater expan­ decreased number of objects that, in turn, are sion in this area. Perhaps first, however, we would more comprehensible and manageable. In this be well advised to pause and reflect on some of the case, the classification process becomes an end in more timely developments of current limnological itself. A typical example might be the taxonomic research. One of these is in the area of lake classification of living things. classification. A second reason for classifying is that the sub­ The number of scientific publications addressing ject matter is not well understood. It is hoped that lake classification has increased considerably since the classification system will reflect the processes the early 1970's. "To what end are these classifica­ that have led to the observed arrangement, thus tions dedicated?," you might ask. Or, "Why has revealing a better understanding of natural laws. there been a sudden interest in classifying lakes?" In this case, classification becomes a means to an The purpose of this paper is to answer these ques­ end-a tool. It is the second reason that seems to tions and to provide some insight into the various have given impetus to many of the attempts at lake lake-classification schemes that have been pro­ classification, particularly the geologically and posed. Although it has not been possible to review hydrologically oriented systems. all lake classifications that have been suggested, Recently, considerations of lake management the bibliographies included herein should provide an and protection strategies-especially those de­ adequate base for anyone interested in further pur­ signed to abate cultural (accelerated) eutrophica­ suing the topic. In particular, those trophic tion-have provided additional, more pragmatic, classifications based on the nutrient loading con­ incentives for devising lake-trophic classifications. cepts introduced by Vollenweider (1968) have been This is particularly true in light of Section 314(a) of specifically excluded, as their consideration would the Federal Water Pollution Control Act Amend­ have made the paper excessively long. ments of 1972 (Public Law 92-500), which dictates: Perhaps the first order of business is to define "Each State shall prepare or establish, and submit exactly what is meant by the term classification. to the Administrator for his approval * * * an According to Sokal (197 4), classification is identification and classification according to " * * * the ordering or arrangement of objects into eutrophic condition of all publicly owned fresh­ groups or sets on the basis of their relationships." water lakes in such State * * * ·"

B15 Such trophic-classification schemes are to be not necessary that a lake possess all these used to assess trends in lake quality and to aid ad­ characteristics to be considered eutrophic, nor is it ministrators in environmental decisionmaking. necessary that all eutrophic lakes possess any one Properly designed and employed, these schemes of these characteristics. The nonuniformity of the will allow for optimum management and protec­ trophic-classification scheme makes it particularly tion of the lake resource. difficult to apply, although this is not its only disad­ vantage, as is explained below. CHARACTERISTICS OF Uttormark and Wall (1975) discussed several CLASSIFICATION SCHEMES characteristics of classification systems that also are worth summarizing. In post-set systems, Perhaps the most important distinction to make classes are defined based on an analysis of the data in regard to classification schemes is between base. Such systems are termed relative -lakes are monothetic and polythetic systems (Sokal, 197 4). classified with respect to one another, and the Monothetic systems require that classes differ structure of the system may change when new from one another by a single property that is lakes are added. This introduces an element of uniformly different among classes. Thus, lakes can complexity when trying to apply such systems to be classified according to their stratification prop­ data different from that used to derive the erties, as measured by the number of times the classification. water column circulates annually. For example, Preset systems, on the other hand, are composed the following scheme was proposed by Hutchinson of the predefined classes that do not depend on the (1957) and summarized by Wetzel (1975): data base. These systems are simpler to use than Class designation Description postset systems; a new lake ean be added and Amictic Noncirculating; perennially frozen; Antarctic and categorized without changing the structure of the some high-altitude lakes Cold monomictic One circulation period, in summer; temperatures classification. never greater than 4 oc; Arctic and high-altitude lakes Uttormark and Wall (1975) described analytic Dimictic Two circulation periods, in spring and fall; temperate lakes and high-altitude lakes in subtropics systems-classes defined by a statistical or other Warm monomictic One circulation period, in winter; temperatures never less than 4 °C; subtropical lakes numerical analysis of the data base-and empirical Oligomictic Circulating rarely and at irregular intervals; temperatures well above 4 oc; tropical lakes systems, whose class boundaries were defined ar­ Polymictic Circulating frequently or continuously; lakes with bitrarily. large surface areas in windy regions of low humidity or high altitude; tropical lakes. In general, preset systems are empirical, and The above classes are subdivisions of either postset systems are analytic. holomictic lakes, wherein the entire water column Finally, one should be aware of the differences can circulate, or meromictic lakes, wherein a lower between single-purpose and multi-purpose layer remains permanently stratified because of in­ classification systems. Multi-purpose systems are creased density from high dissolved-solids concen­ more difficult to develop and generally become trations. quite generalized. Such systems sacrifice utility for In polythetic classification systems, classes are applicability to a broad spectrum of situations. On defined on the basis of sharing a large proportion the other hand, the usefulness of lake classification of similarities, but not necessarily having any prop­ often can be enhanced, if it is allowed to become erty in common (Sokal, 1974). An excellent exam­ dynamic, whereby single-purpose schemes are ple of such a system is the classical scheme that has developed for specific applications. One must developed from the terms oligotrophic and beware, however, of expending great amounts of eutrophic, as applied to lakewater types by resources trying to design an ad hoc scheme for Naumann (1919). Brief histories of the develop­ each new situation. In reality, a compromise ap­ ment of this scheme have been presented by proach probably is best. Hutchinson (1969) and Rodhe (1969). Identification of lakes as eutrophic conjures up visions of (1) large concentrations or rates of input of nutrients, par­ DESCRIPTIONS OF SOME LAKE-CLASSIFICATION SCHEMES ticularly phosphorus; (2) high rates of primary pro­ ductivity, with associated large standing crops of Most of the lake-classification schemes that have algae (often blue-greens) and small Secchi-disc been developed probably can be described best as transparencies; (3) high rates of organic sedimen­ geologic, hydrologic, or trophic. Examples of each tation; and (4) hypolimnia devoid of oxygen. It is are summarized below.

B16 GEOLOGIC SCHEMES In developing their classification, Born and One of the most widely accepted lake classifica­ others (1974) allowed for input of data or, because tions is that of lake-basin origin proposed by actual measurements often were not available, for Hutchinson (1957). By considering the geologic input of descriptive information. The latter is processes responsible for lake-basin origin, the basically an inference or best guess as to the func­ classification has been somewhat regionalized, as tioning of the hydrologic system based on proxy certain processes act predominantly on certain criteria established by Born and others (1974). The areas of the earth's surface. Hutchinson (1957) available hydrologic data base was judged too described 76 major basin types formed by tectonic sparse to adequately assess the performance of the activity, volcanic activity, landslides, glacial pro­ proxy criteria, and it was concluded that " * * * a cesses, rock solution, stream processes, wind, useful and workable classification cannot be shoreline processes, organic accumulation, ac­ adopted at this time." System efficiency seemed to tivities of higher organisms (including man), and be the most difficult of the descriptors to evaluate meteoric impact. Twenty of the major basin types on the basis of the proxy criteria. are of glacial origin, and Hutchinson (1957) noted In Wisconsin, extreme lake levels leave boat that, "No lake-producing agent can compare in im­ docks high and dry in some years and cause flooded portance with the effects of the Pleistocene glacia­ shorelines in others. In order to arrive at a better tions." understanding of these stage fluctuations, A second geologically oriented scheme, that may Novitzki and Devaul (1978) devised a hydrologic be useful in some areas, is based on altitude classification similar to, but simpler than, that of (Pennak, 1966). Four limnological zones were Born and others (1974). Lakes were designated as recognized in northern Colorado: (1) the plains surface-water flow-through (major surface inlet zone, below 1, 700 m (5,600 ft); (2) the foothills zone, and outlet), ground-water discharge (surface outlet 1, 700-2,500 m (5,600-8,200 ft); (3) the montane only), and ground-water flow-through or land­ zone, 2,500-3,200 m (8,200-10,500 ft); and (4) the locked (no surface inlet or outlet). Landlocked alpine zone, above 3,200 m (10,500 ft). Generally lakes have the largest water-level fluctuations, and speaking, lakes at higher altitudes are colder, less ground-water discharge lakes have the smallest productive, and contain lower concentrations of fluctuations. The implications of using such a dissolved solids than lakes at lower altitudes. scheme, based on simple observations, for manage­ ment purposes are obvious and illustrate the prac­ tical utility of lake classification. HYDROLOGIC SCHEMES Both of the above schemes employ a somewhat Born and others (197 4) devised a hydrologic arbitrary determination as to which hydrologic classification based on 63 glacial lakes located variables should be used to classify lakes: the mostly in the north-central United States and classifications are preset. Winter (1977) used a south-central Canada. The incentive for the more general, postset, analytic approach to deter­ classification was based on the premise that cer­ mine which variables are important. His study of tain hydrologic data are required to adequately 150 randomly selected natural lakes in north­ define the nature and causes of lake-quality prob­ central United States assumed that there existed lems, and, thus, to outline appropriate steps for no established variables upon which to base a lake management and protection. hydrological classification. Winter (1977) selected Lakes were classified on the basis of three all the variables " * * * believed to control the at­ descriptors: (1) regime dominance, that is, whether mospheric, surface-, and ground-water inter­ a lake's water budget is surface-water dominated change with a lake." His list of 13 variables includ­ (drainage lakes) or ground-water dominated ed 1 relating to atmospheric-water interchange, 3 (seepage lakes); (2) system efficiency, or flushing relating to surface-water interchange, 6 relating to rate of water through the lake; and (3) position of ground-water interchange, and 3 relating directly the lake in the regional ground-water flow system. to water quality and indirectly to ground-water in­ The latter determination is important for seepage terchange. Thus, the input data were weighted lakes; it describes whether ground-water flows heavily toward the ground-water component of the primarily from the aquifer to the lake (discharge lake's hydrologic budget. lake), from the lake to the aquifer (recharge lake), Numeric data were available for some of the or into the lake on one side and from the lake on variables, the rest were subdivided based on the opposite side (flow-through lake). qualitative descriptions, and numeric values were

B17 assigned to each subgroup. The entire data set was A second important point is that both subjected to principal-component analysis- a eutrophication and dystrophication are con­ multivariate statistical technique (see, for exam­ tinuums. Note, for example, the graphical ple, Rummel, 1970) designed to disclose the impor­ representation of the eutrophication process of­ tant independent variables controlling the fered by Hasler (1947) (fig. 3). The result of this variance in a large complex data matrix. It was continuum realization has been an expansion of the concluded (Winter, 1977) that precipitation­ terminology of Naumann (1919) to include evaporation balance, ground-water dissolved­ mesotrophy, for those lakes having characteristics solids concentration, stream outflow, relation of of both oligotrophy and eutrophy, and, more drainage-basin area to lake area, position (altitude) recently, ultraoligotrophy, hypereutrophy in the regional ground-water flow system, regional (presumably describing the effects of cultural or slope, and local relief of the lake's drainage basin accelerated eutrophication), and other variants. all should be considered in a general hydrologic classification of lakes. Stream inflow and drift tex­ ture (a measure of hydraulic conductivity) might be considered important, depending on how the results of the principal-component analysis are in­ terpreted. Thus, the number of variables required was reduced from 13 to 7 or 9.

TROPHIC SCHEMES Before discussing trophic classification schemes, it would be appropriate to examine briefly the lake trophic-state concept, as traditionally perceived, and some of its implications. AGE OF LAKE First, a useful distinction to be made is that be­ FIGURE 3.-Hypothetical curve of the course of natural tween eutrophication and dystrophication. eutrophication of lakes (modified from Hasler, 1947). Reviews of the details of these processes have been presented by Rodhe (1969) and Wetzel (1975). It is Finally, one must be aware of the controls that important to note that the two processes follow can be imposed on the eutrophication process by separate pathways originating from a similar morphometry, particularly lake depth. Odum oligotrophic beginning (fig. 2); one process is not (1971) has presented the conception outlined in an extension of the other. When the rate of supply figure 4. Thus, for example, a lake can receive of organic matter to a lake comes predominantly large quantities of nutrients; but if the lake is deep, from the watershed (allotrophy), the resulting the volume of water in the lake will keep the con­ process is dystrophication. When the rate of centrations low enough so problems do not organic-matter supply results primarily from develop. primary productivity of phytoplankton in the lake All the above points are reflected in current lake (autotrophy), eutrophication occurs. trophic classifications. Any such scheme ideally should accommodate both eutrophication and dystrophication; however, as the latter process is rarer, produces fewer esthetically displeasing results, and, most importantly, is not accelerated SMALL LARGE CONCENTRATIONS CONCENTRATIONS Dystrophy OF NUTRIENTS OF NUTRIENTS > ::t: 0.. 0 SHAL- Morphometric a:: ')I Eutrophy t- LOW eutrophy 0 ...J ...J <( t t 01 igotrophy Eutropby 01 igotrophy Morphometric ol igotrophy DEEP AUTOTROPHY FIGURE 2. -Summary of the major trophic types of lakes FIGURE 4.-Effect of lake depth on the process of eutrophication (modified from Rodhe, 1969, and Wetzel, 1975). (modified from Odum, 1971).

B18 by man's activities, it normally has not been con­ TSI's ranged from 1.3 to 22.1; they were divided sidered. Also, most recent trophic classifications into five groups and labeled as follows: have taken a quantitative approach, in an attempt Trophic state TSI to avoid some of the subjectivity involved in deter­ Ultraoligotrophic TSI 2 mining when a lake changes from oligotrophic to < Oligotrophic 3 > TSI;;::: 2 eutrophic. Morphometric oligotrophy (fig. 4) Mesotrophic 7 > TSI;;::: 3 creates particular problems when trying to apply Eutrophic 10 > TSI;;::: 7 some of the trophic-classification schemes. Hypereutrophic TSI;;::: 10 Classifications of lake trophic-state might be The classification of Shannon and Brezonik based on physical, chemical, or biological (1972) is a postset system: adding more lakes to the measurements, or on various combinations of data base could change the classification structure. these. In one scheme requiring only chemical Thus, this system cannot be applied directly to measurements, Zafar (1959) proposed a differen­ other areas. One also should appreciate the fact tiation of trophic states using Pearsall's (1922) cat­ that the same variables used by Shannon and + + ion ratio, ([Na] [K])/([Ca] [Mg]), where the con­ Brezonik (1972) may not be appropiate trophic­ centrations are expressed on an equivalent basis: state indicators in other regions; only the approach Pearsall's may be applicable. Trophic state cation ratio Bortleson (1978) used a similar technique to Basically classify lakes in the State of Washington. He oligotrophic >2.0 calculated a trophic-state index, called a Basically characteristic value (CV), based on principal­ mesotrophic 1.2-2.0 component analysis of 617 lakes: Basically eutrophic < 1.2 CV =59.4 (si-0.661) +74.9(UON -0.697) This classification depends on Pearsall's (1922) + 126(UP- 0.124)+ 104.1, conclusion that waters with high ratios normally contain small concentrations of nitrogen, silica, where: SD = Secchi-disc transparency, in meters, carbonate, and phosphorus, whereas, waters hav­ UON =total organic-nitrogen concentration ing low ratios usually are rich in these nutrients. in the upper waters, in milligrams per Shannon and Brezonik (1972) used several liter of nitrogen, and multivariate statistical procedures to classify 55 UP= total phosphorus concentration in the lakes in north and central Florida. An expression upper waters, in milligrams per liter for a trophic-state index (TSI) was derived using of phosphorus. principal-component analysis: Values of CV ranged from 1 to 1,259. Bortleson (1978) did not label CV ranges according to tradi­ TSI=0.919 s1J+0.800COND+0.896TON tional trophic-state designations; however, lakes + 0. 738TP + 0.942PP + 0.862CHA with larger values of CV normally would be con­ +0.634 ci+5.19, sidered more eutrophic. Bortleson's (1978) where: SD = Secchi-disc transparency, in meters, classification is a postset scheme. COND =specific conductance, in micro mhos A somewhat simpler approach to the trophic per centimeter at 25°C, classification of lakes has been presented by the TON= total organic-nitrogen concentration, U.S. Environmental Protection Agency (1974) and in milligrams per liter of nitrogen, was based on 209 lakes sampled by the National TP =total phosphorus concentration, in Eutrophication Survey staff in 1972. A trophic in­ milligrams per liter of phosphorus, dex number (TIN) was derived from annual median PP =primary productivity, in milligrams concentrations of total phosphorus · (mg P/L ), per cubic meters per hour of carbon, dissolved phosphorus (mg P/L), and inorganic CHA= chlorophyll a concentration, in nitrogen (mg N/L), 500 minus the annual mean milligrams per cubic meter, and Secchi-disc transparency (inches), 15 minus the CR =Pearsall's (1922) cation ratio. minimum hypolimnetic dissolved-oxygen concen-

B19 tration (mg/L), and the annual mean chlorophyll a dissolved-oxygen concentration (0-6); (b) Secchi­ concentration (Jtg/L). For each lake, each variable disc transparency (0-4); (c) history of fishkills was ranked according to the percentage of lakes (0-4); and (d) assessment of recreational use im­ having a greater value. Thus, for example, a lake pairment by algal blooms or emergent vegetation having a low total phosphorus concentration would (0-9). A larger rating implies a more eutrophic con­ have a high percentile rank for that variable. The dition. The ratings from each of the four variables TIN is the sum of the percentile ranks for each of are summed to yield a lake-condition index (LCI). the six variables; the more eutrophic lakes have The technique of Uttormark and Wall (1975) is lower TIN's. much less data-intensive than that of Bartleson Using a data set with 209 observations, the TIN and others (197 4). In addition, the LCI technique . can vary from 594 (six variables x 99 percent) to 0. was designed to allow for substitution, when Actual values were in the range 573-41 and were necessary, of subjective information for actual labeled as follows: data. This characteristic permitted a considerably Trophic state TIN larger number of Wisconsin lakes to be classified than would have been possible otherwise. Oligotrophic > 500 According to Uttormark and Wall (1975) Mesotrophic 499-420 " * * * LCI values are not synonymous with Eutrophic < 420 trophic status * * *;"however, the following com­ The U.S. Environmental Protection Agency (1974) parisons were offered: emphasized: " * * * the [trophic] index number Trophic state LCI ranges * * * are only valid when used with the data base of 209 lakes and reservoirs considered in Very oligotrophic 0-1 this report. If a different data base were Oligotrophic 2-4 used, * * * the resulting index number might be in­ Meso trophic 5-9 terpreted quite differently * * *." This classifica­ Eutrophic 10-12 tion is also a postset system. Very eutrophic >13 An illustration of a preset system is provided by LCI's for Wisconsin lakes ranged from 0 to 22; Bartleson and others (1974) in an early attempt at and, for 42 lakes classified both by U ttormark and classifying Washington lakes according to trophic Wall (1975) and by the U.S. Environmental Protec­ condition. Each of 24 separate variables was rated tion Agency (1974), the following results were from 1 to 5, based on actual measurements, higher noted: values indicating more eutrophic conditions. The variables were separated into: (1) natural physical Number of Lakes factors, predominantly measures of lake mor­ TIN method (U.S. Environ­ phometry; (2) culturally related factors, which LCI method mental represent nutrient-supply rates from various (Uttormark and Protection sources; and (3) trophic-state indicators, which are Trophic state Wall, 1975) Agency, 1974) measures of chemical, physical, and biological con­ Oligotrophic ______5 2 ditions in the lake. A summed score was deter­ Mesotrophic ______16 5 mined for each of the three groups of factors, and Eutrophic ______21 35 each score was interpreted separately. Total ______42 42 Because of the number of variables involved, ap­ plication of the system requires an exhaustive Thus, the LCI method seems to be more lenient data-collection effort. Bartleson and others (197 4) for assessing trophic state than the TIN method. applied the scheme to seven Washington lakes to Uttormark and Wall (1975) also evaluated the demonstrate the technique; however, they did not possibility of using the LCI technique in other critically evaluate its effectiveness or its parts of the United States. They determined that transferability to other regions. the method would have direct applicability in the The classification of 1, 14 7 Wisconsin lakes was northeast and northern midwest, and would least accomplished by Uttormark and Wall (1975) using likely be useable in the south and southwest. They an approach similar to that of Bartleson and others indicated that the method might be useable, with (197 4). The variables considered (with their rating some modification, throughout the rest of the scales in parentheses) are: (a) hypolimnetic country.

B20 Thus far, one pervading point must be noticed: quired data are simply and cheaply collected. The Except for Zarfar's (1959) scheme based on Pear­ total phosphorus index is most stable seasonally; sall's (1922) cation ratio, Secchi-disc transparency however, this advantage must be weighed against has been incorporated into all the trophic the probability that the general regression equa­ classifications discussed. Therefore, it is not sur­ tion presented by Carlson (1977) actually describes prising that a system has been developed (Carlson, the Secchi-disc total phosphorus relationship for 1977) based entirely on this measurement. The use the particular lake in question. Alternatively, the of a single variable upon which to base a trophic­ actual relationship for any given lake could be state index greatly simplifies data collection, and defined, but this makes the classification effort leads to somewhat less ambiguity than indexes more complex and expensive. based on several variables. The choice of Secchi­ Although Carlson (1977) did not label ranges of disc transparency is logical for a single-variable his index in terms of traditional trophic-state ter­ system, as Carlson (1977) notes that transparency minology, these relationships can be inferred as has been related to hypolimnetic oxygen deficit follows. Vollenweider (1968) suggested a classifica­ (Lasenby, 1975) and to chlorophyll a and total tion of European lakes into ultraoligotrophic, phosphorus concentrations. All these latter oligomesotrophic, mesoeutrophic, eupolytrophic, variables have been incorporated into one or more and polytrophic based on total phosphorus concen­ of the previously discussed indexes. trations. His mesoeutrophic class is defined by Carlson's (1977) trophic-state index is computed total phosphorus concentrations of 10 to 30 mg from the equation TSI(SD)= 10(6-log2 SD), (3) Pfms and corresponds to the mesotrophic label of where: SD = Secchi-disc transparency, in rp.eters. Sakamoto (1966) for Japanese lakes having By taking the . logarithm. to the base 2 of roughly the same range of total phosphorus con­ the 8ecchi-disc transparency, each doubling or centrations. By converting the total phosphorus halving of the transparency, which is assumed to concentrations to values of TSI(TP) (Carlson, reflect a corresponding doubling or halving of the 1977), one arrives at the following tentative bound­ algal biomass, causes the TSI(SD) to change by one aries for trophic states: unit. Subtracting from 6 ( = log264) assures that the Trophic state TSI(TP) index is always positive (a transparency of 64 Oligotrophic <37 meters should never be exceeded) and inversely Mesotrophic 37-53 related to transparency. Multiplying by 10 allows Eutrophic >53 one to carry 2 significant digits without including a decimal point. The index ranges from 0, at a In his review of lake-classification schemes, Secchi-disc transparency of 64 meters, to a prac­ Shapiro (1975) noted that Carlson's index is tical limit of about 100 (at a transparency of 62 " * * * applicable to many lakes." Indeed, it may be mm): the most widely used system currently employed. Applications in Maine (Bailey, 1975), the Great TSI(SD) Secchi-disc transparency, Lakes states (Garn and Parrott, 1979), and Col­ in meters 0 64 orado (L. J. Britton and D. A. Wentz, written com­ 10 32 mun., 1980) are known. Many others undoubtedly 20 16 exist. 30 8 Trophic classifications based primarily on biological data have been much less rigorous and quantitative than those previously discussed. The use of algal indicators has been discussed by 100 0.062 Rawson (1956), who summarized the following major differences between trophic states: Carlson's (1977) index also can be calculated from chlorophyll a or total phosphorus concentra­ tion, based on generalized regression equations Trophic state relating each of these variables to Secchi-disc Algal transparency; indexes calculated in this manner measurement Oligotrophic Eutrophic normally are labeled TSI (Chl) and TSI (TP), Quantity ______Poor Rich respectively. Variety ______Many species Few species An advantage of computing the trophic-state in­ W aterblooms ______Very rare Frequent dex from Secchi-disc transparency is that the re-

B21 Rawson (1956) noted that the few algal species (1969) summarized evidence indicating changes, often thought to be associated with eutrophic con­ due to complex competition-predation interrela­ ditions may be a misnomer resulting from rarer tionships, in copepod populations from the larger species being overlooked in the relatively rich Bosmina coregoni longispina to the smaller B. populations. He pointed out the importance of ob­ longirostris during lake eutrophication. Thus, dif­ taining algal observations throughout the growing ferences in zooplankton populations among lake season, and provided a list of 18. algal indicators, trophic-states have been demonstrated; however, arranged in order from oligotrophy through interpretation of lake trophic-state based on prop­ eutrophy based on dominant species in lakes of erties of zooplankton populations is not a well­ western Canada. Differentiation of trophic-state defined procedure. based on these indicators is not straightforward Shapiro (1975) reviewed several trophic-classi­ and should be approached with caution. fication systems based on benthic macroinverte­ Tarapchak (1973) classified 68 Minnesota lakes brates and on fish production. He noted that these as oligotrophic, mesotrophic-eutrophic, saline­ often are difficult to understand and apply, and eutrophic, and hypereutrophic on the basis of that are not likely to have widespread application. major solutes and the assumption that " * * * it is Finally, mention should be made of recent at­ likely that salinity is in fact a good indirect tempts to classify lake trophic condition from data measure of nutrient level." He investigated collected by Landsat satellites. Multispectral scan­ associated algal communities and found that the ners aboard Landsat measure intensities of light number of phytoplankton species was the only reflected from the earth in four wavelength bands: biological index sufficient for distinguishing the 500-600, 600-700, 700-800, and 800-1,100 lake groups: diversity, evenness, standing crop, nanometers. Boland (1976) developed a trophic­ and various ratios (quotients) of numbers of in­ state index based on principal-component analysis dividuals in certain taxonomic groups did not give of ground truth data from 100 Minnesota, Wiscon­ satisfactory results. Of the 13 algal species Tarap­ sin, Michigan, and New York lakes sampled by the chak (1973) found to be indicative of oligotrophy National Eutrophication Survey staff in 1972. He and the 9 species listed as representing eutrophy, then used multiple regression analysis to predict only 1 species also was listed by Rawson (1956). his trophic-state index (dependent variable) from Primary productivity is a likely algal-associated the intensities of light in the four Landsat variable for classifying lake trophic condition, wavelength bands (independent variables) especially given the concept in figure 2. Wetzel -reflected from within the lakes. In Wisconsin, (1975) provided the following summary, based ap­ 1,380 of 3,100 lakes with surface areas greater proximately on net primary productivity: than 8.1 hectares (20 acres) have been categorized according to trophic status using a similar ap­ Mean "net"primary productivity proach (Holmquist, 1977; Holmquist and Scarpace, Trophic state (mgC/m2/day) 1977; Martin and Holmquist, 1979). Ultraoligotrophic <50 Both the above studies found that Secchi-disc Oligotrophic 50-300 transparencies could be predicted from Landsat Mesotrophic 250-1,000 data. This would indicate that Carlson's (1977) Eutrophic > 1,000 trophic-state index also could be used in conjunc­ tion with Landsat data to classify lakes. The greatest objection to this approach is the im­ plicit assumption that organic matter inputs from CONCLUSIONS littoral macrophytes and from sources in the lake's watershed are small, relative to those from the It is interesting to note that the writing of this phytoplankton. paper has become, itself, an exercise in classifica­ Relatively little has been published regarding tion. Hopefully, the above classification and trophic characterization of lakes on the basis of categorization of lake-classification schemes have zooplankton. Ghilarov (1972) was successful in provided enough illumination to penetrate much of separating seven oligotrophic, morphometrically the murkiness, but not so much as to make the sub­ oligotrophic, mesotrophic, and morphometrically ject totally transparent. Looking back on what has eutrophic lakes north of the Arctic Circle in north­ been said, perhaps the most obvious conclusion is western Russia, using a plot of zooplankton that a large variety of lake-classification schemes biomass versus diversity. In addition, Brooks have been proposed. Some have rather limited ap-

B22 plicability; others appear to be adaptable to dif­ condition: Madison, Wisconsin Department of Natural ferent situations. When the need for a lake­ Resources, 54 p. classification system arises, the most important Naumann, Einar, 1919, Nagra synpunkter angaende lim­ noplanktons okologi med sarskild hansyn till fytoplankton: point to remember is not to apply these schemes in­ Svensk Botanisk Tidskrift, v. 13, p. 129-163. (Some aspects discriminantly. As in all scientific endeavors, the of the ecology of the limnoplankton, with special reference results obtained through lake classification are to the phytoplankton, English translation by I. B. Tailing, porportional only to the thought directed initially Freshwater Biological Association, no. 49.) to the problem. Novitzki, R. P., and Devaul, R. W., 1978, Wisconsin lake levels-Their ups and downs: Madison, University of Wisconsin-Extension, Geological and Natural History REFERENCES CITED Survey, 11 p. Odum, E. P., 1971, Fundamentals of ecology, 3rd edition: Phila­ Bailey, John, 1975, Proposed trophic classification of the great delphia, W. B. Saunders Co., 574 p. ponds of Maine: Augusta, Department of Environmental Pearsall, W. H., 1922, A suggestion as to factors influencing the Protection Draft Report, 25 p. distribution of free-floating vegetation: Journal of Ecology, Boland, D. H. P., 1976, Trophic classification of lakes using v. 9, p. 241-253. Landsat-1 (ERTS-1) multispectral scanner data: U.S. En­ Pennak, R. W., 1966, Rocky mountain states, in Frey, D. G., vironmental Protection Agency Report EPA- ed., Limnology in North America: Madison, University of 600/3-76-037' 245 p. Wisconsin Press, p. 349-369. Born, S. M., Smith, S. A., and Stephenson, D. A., 1974, The Rawson, D. S., 1956, Algal indicators of trophic lake types: Lim­ hydrogeologic regime of glacial-terrain lakes, with manage­ nology and Oceanography, v. 1, p. 18-25. ment and planning applications: Madison, University of Rodhe, Wilhelm, 1969, Crystallization of eutrophication con­ Wisconsin-Extension, Inland Lake Renewal and Manage­ cepts in northern Europe, in Eutrophication-Causes, con­ ment Demonstration Project Report, 73 p. sequences, correctives: Washington, D.C., National Bortleson, G. C., 1978, Preliminary water-quality charac­ Academy of Sciences, p. 50-64. terization of lakes in Washington: U.S. Geological Survey Rummel, R. J., 1970, Applied factor analysis: Evanston, Ill., Water-Resources Investigations 77-94, 31 p. Northwestern University Press, 617 p. Bortleson, G. C., Dion, N. P., and McConnell, J. B., 1974, A Sakamota, Mitsuru, 1966, Primary production by phyto­ method for the relative classification of lakes in the State of plankton community in some Japanese lakes and its Washington from reconnaissance data: U.S. Geological dependence on lake depth: Archives Hydrobiologiae, v. 62, Survey Water-Resources Investigations 37-74, 35 p. p. 1-28. Brooks, J. L., 1969, Eutrophication and changes in the com­ Shannon, E. E., and Brezonik, P. L., 1972, Eutrophication position of the zooplankton, in Eutrophication- Causes, analysis-A multivariate approach: Journal of the Sanitary consequences, correctives: Washington, D.C., National Engineering Division, American Society of Civil Engineers, Academy of Sciences, p. 236-255. v. 98, Proceedings Paper 8735, p. 37-57. Carlson, R.E., 1977, A trophic state index for lakes: Limnology Shapiro, Joseph, 1975, The current status of lake trophic in­ and Oceanography, v. 22, p. 361-369. dices-A review: Minneapolis, University of Minnesota, Garn, H. S., and Parrott, H. A., 1979, Recommended methods Limnological Research Center, Interim Report No. 15, 39 p. for classifying lake condition, determining lake sensitivity, Sokal, R. R., 1974, Classification-Purposes, principles, prog­ and predicting lake impacts: Milwaukee, U.S. Forest Serv­ ress, prospects: Science, v. 185, p. 1,115-1,123. ice, Hydrology Paper 2, 48 p. Tarapchak, S. J., 1973, Studies of phytoplankton distribution Ghilarov, A. M., 1972,A classification of northern lakes based and indicators of trophic state in Minnesota lakes: Min­ on zooplankton: Hydrobiological Journal, v. 8, p. 1-8. neapolis, University of Minnesota, unpublished Ph. D. Hasler, A. D., 1947, Eutrophication of lakes by domestic dissertation, 204 p. drainage: Ecology, v. 28, p. 383-395. U.S. Environmental Protection Agency, 1974, An approach to Holmquist, K. W., 1977, The Landsat lake eutrophication study: relative trophic index for classifying lakes and reser­ Madison, University of Wisconsin, unpublished M.S. thesis, voirs-Working paper no. 24: U.S. Department of Com­ 94 p. merce, National Technical Information Service PB-242 Holmquist, K. W., and Scarpace, F. L., 1977, The use of 336, 44 p. satellite imagery for lake classification in Wisconsin: Uttormark, P. D., and Wall, J. P., 1975,Lake classification-A Madison, University of Wisconsin, Water Resources trophic characterization of Wisconsin lakes: U.S. En­ Center, Technical Report, WIS WRC 77-08, 35 p. vironmental Protection Agency Report EPA- Hutchinson, G. E., 1957, A treatise on limnology, Vol. I. 660/3-75-033, 165 p. Geography, physics, chemistry: New York, John Wiley & Vollenweider, R. A., 1968, Scientific fundamentals of the Sons, 1,015 p. eutrophication of lakes and flowing waters, with particular --1969, Eutrophication, past and present, in Eutrophica­ reference to nitrogen and phosphorus as factors in tion-Causes, consequences, correctives: Washington, D.C., eutrophication: Paris, Organization for Economic Coopera­ National Academy of Sciences, p. 17-26. tion and Development, 220 p. Lasenby, D.C., 1975, Development of oxygen deficits in 14 Wetzel, R. G., 1975, Limnology: Philadelphia, W.B. Saunders southern Ontario lakes: Limnology and Oceanography, v. Co., 743 p. 20, p. 933-999. Winter, T. C., 1977, Classification of the hydrologic settings of Martin, R. H., and Holmquist, K. W., 1979, Remote sensing as a lakes in the north central United States: Water Resources mechanism for classification of Wisconsin lakes by trophic Research, v. 13, p. 753-767.

B23 Zafar, A. R., 1959, Taxonomy of lakes: Hydrobiologia, ern Spain: Verhandlungen' Internationale V ereinigung Lim- v. 13, p. 287-299. nologie, v. 13, p. 339-349. . Milligan, J. H., Warnick, C. C., and oth~rs, 1973; RecreatiOn SUPPLEMENTARY REFERENCES water classification system and carrymg capacity: Moscow, University of Idaho, Research Technical Completion Anderson, George, 1975, Classification of Wisconsin lakes by Report, 89 p. . . trophic condition: Madison, Wisconsin Department of Moyle, J. B., 1956, Relationships between the chemistry of Mm­ Natural Resources Report, 107 p. nesota surface waters and wildlife management: Journal of Bogoslovsky, B. B., 1966, On water budget lake classification, Wildlife Management, v. 20, p. 303-320. in Hydrology of lakes and reservoirs: Symposium of Garda, Patalas, Kazimierz, 1971, Crustacean plankton communities in International Association of Scientific Hydrology, Publica­ forty-five lakes in the Experimental Lakes Area, North­ tion No. 71, v. 2, p. 910-917. western Ontario: Journal of the Fisheries Research Board Bright, R. C., 1968, Surface-water chemistry of some Min­ of Canada, v. 28, p. 231-244. nesota lakes, with preliminary notes on diatoms: Min­ Pennak, R. W., 1958, Regional lake typology in northern Col­ neapolis, University of Minnesota, Limnological Research orado, U.S.A.: Verhandlungen Internationale Vereinigung Center, Interim Report No.3, 58 p. Limnologie, v. 13, p. 264-283. Dean, W. E., and Gorham, Eville, 1976, Major chemical and Reynoldson, T. B., 1958, Triclads and lake typology in no~thern mineral components of profunda} surface sediments in Min­ Britian-Qualitative aspects: V erh~ndlungen Internat10nale nesota lakes: Limnology and Oceanography, v. 21, p. Vereinigung Limnologie, v. 13, p. p20-330. 259-284. Round, F. E., 1958, Algal aspects of lake typology: Donaldson, J. R., 1969, The classification of lakes: in Mid­ Verhandlungen Intemationale Vereinigung Limnologie, v. dlebrooks, E. J., Maloney, T. E., Powers, C. F., and Kaack, 13, p. 306-310. L. M., eds., Proceedings of the eutrophication­ Ryder, R. A., Kerr, S. R., Loftus, K. H., and Re~er, H .. A., biostimulation assessment workshop: Berkeley, University 1974 The morphoedaphic index, a fish yield of California, Sanitary Engineering Research Laboratory, estim~tor-Review and evaluation: Journal of the Fisheries U.S. Department of Commerce, National Technical Infor­ Research Board of Canada, v. 31, p. 663-688. mation Service, PB-215 604, p. 171-185. Schindler, D. W., and Holmgren, S. K., 1971, Primary produc­ Findenegg, Ingo, 1964, Types of planktic primary production in tion and phytoplankton in the Experimental Lakes Area, the lakes of the Eastern Alps as found by the radioactive N orthwestem Ontario, and other low carbonate waters, and carbon method: Verhandlungen lnternationale V ereinigung a liquid scintillation method for determining 14C activity in Limnologie, v. 15, p. 352-359. photosynthesis: Journal of the Fisheries Research Board of Jarnefelt, H., 1958, On the typology of the northern lakes: Canada, v. 28, p. 189-201. Verhandlungen Internationale Vereinigung Limnologie, v. Szesztay, K., 1974, Water balance and water level fluc­ 13, p. 228-235. tuations of lakes: Hydrological Sciences Bulletin, v. 19, p. Larkin, P. A., and Northcote, T. G., 1958, Factors in lake 73-84. typology in British Colombia, Canada: V erhandlungen In­ Weiss, C. M., and Kuenzler, E. J., 1976, The trophic state of ternationale Vereinigung Limnologie, v. 13, p. 252-263. North Carolina lakes: Raleigh, University of North Likens, G. E., 1975, Primary production of inland aquatic Carolina, Water Resources Research Institute Report No. ecosystems, Lieth, Helmut, and Wittaker, R. H., eds., in 119, 224 p. . .. Primary productivity of the biosphere: New York, Springer­ Zumberge, J. H., 1952, The lakes of Minnesota-Their ~rigm Verlag, p. 185-202. and classification: Minnesota Geological Survey Bulletm 35, Margalef, Ramon, 1958, Trophic typology versus biotic 99 p. typology, as exemplified in the regional limnology of North-

,z U.S. GOVERNMENT PRINTING OFFICE: 1981-341-614/224

B24

{ =