THE TROPHIC-DYNAMIC ASPECT OF ECOLOGY
RAYMOND L. LINDEMAN
Osborn Zoological Laboratory, Yale University
COMMUNITY CONCEPTS
A chronological review of the major viewpoints guiding synecological thought indicates the following stages: (1) the static species-distributional viewpoint; (2) the dynamic species-distributional viewpoint, with emphasis on successional phenomena; and (3) the trophic-dynamic viewpoint. From either species-distributional viewpoint, a lake, for example, might be considered by a botanist as containing several distinct plant aggre gations, such as marginal emergent, floating-leafed, submerged, benthic, or phytoplankton communities, some of which might even be considered as "climax" (cf. Tutin, '41). The associated animals would be "biotic factors" of the plant environment, tending to limit or modify the development of the aquatic plant communities. To a strict zoologist, on the other hand, a lake would seem to contain animal communities roughly coincident with the plant communities, although the "associated vegetation" would be considered merely as a part of the environment' of the animal community. A more "bio-ecological" species-distributional approach would recognize both the plants and animals as co-constituents of restricted "biotic" communities, such as "plankton communities," "benthic communities," etc., in which members of the living community "co-act" with each other and "react" with the non-living environment (Clements and Shelford, '39; Carpenter, '39, '40; T. Park, '41). Coactions and reactions are considered by bio-ecologists to be the dynamic effectors of succession.
The trophic-dynamic viewpoint, as adopted in this paper, emphasizes the
relationship of trophic or "energy-availing" relationships within the
community-unit to the process of succession. From this viewpoint, which is
closely allied to Vernadsky's "biogeochemical" approach (cf. Hutchinson and
Wollack, '40) and to the "oekologische Sicht" of Friederichs ('30), a lake is
considered as a primary ecological unit in its own right, since all the lesser
"communities" mentioned above are dependent upon other components of the
lacustrine food cycle (cf. figure 1) for their very existence. Upon further
consideration of the trophic cycle, the discrimination between living organisms
as parts of the "biotic community" and dead organisms and inorganic nutritives
as parts of the "environment" seems arbitrary and unnatural. The difficulty of
drawing clear-cut lines between the living community and the non-living
environment is illustrated by the difficulty of determining the status
of a slowly dying pondweed covered with periphytes, some of which are also
continually dying. As indicated in figure 1, much of the non-living nascent
ooze is rapidly reincorporated through "dis-
400 RAYMOND L. LINDEMAN
Ecology Vol. 23, No. 4.
solved nutrients" back into the living "biotic community." This constant organic-inorganic cyclic of nutritive substance is so completely integrated that to consider even such a unit as a lake primarily as a biotic community appears to force a "biological" emphasis upon a more basic functional organization.
This concept was perhaps first expressed by Thienemann ('18), as a result of his extensive limnological studies on the lakes of North Germany. Allee ('34) expressed a similar view, stating: "The picture than finally emerges . . . is of a sort of superorganismic unity not alone between the plants and animals to form biotic communities, but also between the biota and the environment." Such a concept is inherent in the term ecosystem, proposed by Tansley ('35) for the fundamental ecological unit. (footnote 2) Rejecting the terms "complex organism" and "biotic community," Tansley writes, "But the more fundamental conception is, as it seems to me, the whole system (in the sense of physics), including not only the organism-complex, but also the whole complex of physical factors forming what we call the environment of the biome. . . . It is the systems so formed which, from the point of view of the ecologist, are the basic units of nature on the face of the earth.... These ecosystems, as we may call them, are of the most various kinds and sizes. They form one category of the multitudinous physical systems of the universe, which range from the universe as a whole down to the atom." Tansley goes on to discuss the ecosystem as a category of rank equal to the "biome" (Clements, '16), but points out that the term can also be used in a general sense, as is the word "community." The ecosystem may be formally defined as the system composed of physical-chemical-biological processes active within a space-time unit of any magnitude, i.e., the biotic community plus its abiotic environment. The concept of the ecosystem is believed by the writer to be of fundamental importance in interpreting the data of dynamic ecology.
TROPHIC DYNAMICS
Qualitative food-cycle relationships
Although certain aspects of food relations have been known for
centuries, many processes within ecosystems are still very incompletely
understood. The basic process in trophic dynamics is the transfer of energy
from one part of the ecosystem to another. All function, and indeed all life,
within an ecosystem depends upon the utilization of an external source of
energy, solar radiation. A portion of this incident energy is transformed by
the process of photosynthesis into the structure of living organisms. In the
language of community economics introduced by Thienemann ('26), autotrophic
plants are producer organisms, employing the energy obtained by
photosynthesis to synthesize complex organic substances from simple inorganic
substances. Although plants again release a portion of this potential energy in
catabolic processes, a great surplus of organic substance is accumulated.
Animals and heterotrophic plants, as consumer organisms, feed upon this
surplus of potential energy, oxidizing a considerable portion of the consumed
substance to release kinetic energy for metabolism, but transforming the
remainder into the complex chemical substances of their own bodies. Following
death, every organism is a potential source of energy for saprophagous
organisms (feeding directly on dead tissues), which again may act as energy
sources for successive categories of consumers. Heterotrophic bacteria and
fungi, representing the most important saprophagous consumption of energy, may
be conveniently differentiated from animal consumers as
special-
October, 1942 TROPHIC-DYNAMIC ASPECT OF ECOLOGY 401
ized decomposers (footnote three) of organic substance. Waksman ('41) has suggested that certain of these bacteria be further differentiated as transformers of organic and inorganic compounds. The combined action of animal consumers and bacterial decomposers tends to dissipate the potential energy of organic substances, again transforming them to the inorganic state. From this inorganic state the autotrophic plants may utilize the dissolved nutrients once more in resynthesizing complex organic substance, thus completing the food cycle.
The careful study of food cycles reveals an intricate pattern of trophic predilections and substitutions underlain by certain basic dependencies; food-cycle diagrams, such as figure 1, attempt to portray these underlying relationships. In general, predators are less specialized in food habits than are their prey. The ecological importance of this statement seems to have been first recognized by Elton ('27), who discussed its effect on the survival of prey species when predators are numerous and its effect in enabling predators to survive when their
402 RAYMOND L. LINDEMAN Ecology, Vol. 23, No. 4
usual prey are only periodically abundant. This ability on the part of predators, which tends to make the higher trophic levels of a food cycle less discrete than the lower, increases the difficulties of analyzing the energy relationships in this portion of the food cycle, and may also tend to "shorten the food-chain."
Fundamental food-cycle variations in different ecosystems may be observed by comparing lacustrine and terrestrial cycles. Although dissolved nutrients in the lake water and in the ooze correspond directly with those in the soil, the autotrophic producers differ considerably in form. Lacustrine producers include macrophytic pondweeds, in which massive supporting tissues are at a minimum, and microphytic phytoplankters, which in larger lakes definitely dominate the production of organic substance. Terrestrial producers are predominantly multicellular plants containing much cellulose and lignin in various types of supporting tissues. Terrestrial herbivores, belonging to a great number of specialized food groups, act as primary consumers (sense Jacot, '40) of organic substance; these groups correspond to the "browsers" of aquatic ecosystems. Terrestrial predators may be classified as more remote (secondary, tertiary, quaternary, etc.) consumers, according to whether they prey upon herbivores or upon other predators; these correspond roughly to the benthic predators and swimming predators, respectively, of a lake. Bacterial and fungal decomposers in terrestrial systems are concentrated in the humus layer of the soil; in lakes, where the "soil" is overlain by water, decomposition takes place both in the water, as organic particles slowly settle, and in the benthic "soil." Nutrient salts are thus freed to be reutilized by the autotrophic plants of both ecosystems.
The striking absence of terrestrial "life-forms" analogous to planters (footnote four) (cf. figure 1) indicates that the terrestrial food cycle is essentially "mono-cyclic" with macrophytic producers, while the lacustrine cycle, with two "life-forms" of producers, may be considered as "bicyclic." The marine cycle, in which plankters are the only producers of any consequence, may be considered as "mono-cyclic" with microphytic producers. The relative absence of massive supporting tissues in plankters and the very rapid completion of their life cycle exert a great influence on the differential productivities of terrestrial and aquatic systems. The general convexity of terrestrial systems as contrasted with the concavity of aquatic substrata results in striking trophic and successional differences, which will be discussed in a later section.
Productivity
Definitions.--The quantitative aspects of trophic ecology have been commonly expressed in terms of the productivity of the food groups concerned. Productivity has been rather broadly defined as the general rate of production (Riley, '40, and others), a term which may be applied to any or every food group in a given ecosystem. The problem of productivity as related to biotic dynamics has been critically analyzed by G. E. Hutchinson ('42) in his recent book on limnological principles. The two following paragraphs are quoted from Hutchinson's chapter on "The Dynamics of Lake Biota":
The dynamics of lake biota is here treated as primarily a problem of energy transfer . . . the biotic utilization of solar energy entering the lake surface. Some of this energy is transformed by photosynthesis into the structure of phytoplankton organisms, representing an energy content which may be expressed as A1 (first level). Some of the phytoplankers will be eaten by
October, 1942 TROPHIC-DYNAMIC ASPECT OF ECOLOGY 403
zooplankters (energy content A2)~ which again will be eaten by plankton predators (energy content A3). The various successive levels (i.e., stages5) of the food cycle are thus seen to have successively different energy contents (A1A2~ A3 etc.).
Considering any food-cycle level An, energy is entering the level and is leaving it. The rate of change of the energy content An therefore may be divided into a positive and a negative part:
dAn/dt = An + An'
where ~n is by definition positive and represents the rate of contribution of energy from An-l (the previous level) to An, while Anis negative and represents the sum of the rate of energy dissipated from An and the rate of energy content handed on to the following level An+l. The more interesting quantity is An which is defined as the true productivity) of level An. In practice, it is necessary to use mean rates over finite periods of time as approximations to the mean rates L0, l1, l2 . . .
In the following pages we shall consider the quantitative relationships of the following productivities: L0 (rate of incident solar radiation), ~, (rate of photosynthetic production), L2 (rate of primary or herbivorous consumption), -L3(rate of secondary consumption or primary predation), and ~ (rate of tertiary consumption). The total amount of organic structure formed per year for any level An~ which is commonly expressed as the annual "yield," actually represents a value uncorrected for dissipation of energy by (1) respiration, (2) predation, and (3) post-mortem decomposition. Let us now consider the quantitative aspects of these losses.
Respiratory corrections.--The amount of energy lost from food levels by catabolic processes (respiration) varies considerably for the different stages in the life histories of individuals, for different levels in the food cycle and for different seasonal temperatures. In terms of annual production, however, individual deviates cancel out and respiratory differences between food groups may be observed.
Numerous estimates of average respiration for photosynthetic producers may be obtained from the literature. For terrestrial plants, estimates range from 15 per cent (Putter, re Spoehr, '26) to 43 per cent (Lundegardh, '24) under various types of natural conditions. For aquatic producers, Hicks ('34) reported a coefficient of about 15 per cent for Lemna under certain conditions. Wimpenny ('41) has indicated that the respiratory coefficient of marine producers in polar regions (diatoms) is probably much less than that of the more "animal-like" producers (peridinians and coccolithophorids) of warmer seas, and that temperature may be an important factor in determining respiratory coefficients in general. Juday ('40), after conducting numerous experiments with Manning and others on the respiration of phytoplankters in Trout Lake, Wisconsin, concluded that under average conditions these producers respire about ~ of the organic matter which they synthesize. This latter value, 33 per cent, is probably the best available respiratory coefficient for lacustrine producers.
Information on the respiration of aquatic primary consumers is obtained from an illuminating study by Ivlev ('39a) on the energy relationships of Tubifex. By means of ingenious techniques, he determined calorific values for assimilation and growth in eleven series of experiments. Using the averages of his calorific values, we can make the following simple calculations: assimilation (16.77 cal.)--growth (10.33 cal.) = respiration (6.44 cal.), so that respiration in terms of growth = 6.44/10.33 = 62.30 per cent. As a check on the growth stage of these worms, we find that growth/assimilation = 61.7 per cent, a value in good agreement with the classical conclusions of Needham ('31, III, p. 1655) with respect to embryos: the efficiency of all developing embryos is numerically similar, between 60 and 70 per cent, and
404 RAYMOND L. LINDEMAN Ecology, Vol. 23, No. 4
independent of temperature within the range of biological tolerance. We may therefore conclude that the worms were growing at nearly maximal efficiency, so that the above respiratory coefficient is nearly minimal. In the absence of fur- ther data, we shall tentatively accept 62 per cent as the best available respiratory coefficient for aquatic herbivores. The respiratory coefficient for aquatic predators can be approximated from data of another important study by Ivlev ('39b), on the energy transformations in predatory yearling carp. Treating his calorific values as in the preceding paragraph, we find that ingestion (1829 cal.)--defecation (454 cal.) = assimilation (1375 cal.), and assimilation--growth (573 cal.) = respiration (802 cal.), so that respiration in terms of growth 802 = 573 = 140 per cent, a much higher coefficient than that found for the primary consumer, Tubifex. A rough check on this coefficient was obtained by cal- orific analysis of data on the growth of yearling green sunfishes (Lepomis cy- anellus) published by W. G. Moore ('41), which indicate a respiratory coefficient of 120 per cent with respect to growth, suggesting that these fishes were growing more efficiently than those studied by Ivlev. Since Moore's fishes were fed on a highly concentrated food (liver), this greater growth efficiency is not surprising. If the maximum growth efficiency would occur when growth/assimilation = 60-70 per cent (AEE of Needham, '31), the AEE of Moore's data (about 50 per cent) indicates that the minimum respiratory coefficient with respect to growth might be as low as 100 per cent for certain fishes. Food-conversion data from Thompson ('41) indicate a minimum respiratory coefficient of less than 150 per cent for young black bass (Huro salmoides) at 70deg. F., the exact percent-age depending Upon how much of the ingested food (minnows) was assimilated. Krogh ('41) showed that predatory fishes have a higher rate of respiration than the more sluggish herbivorous species; the respiratory rate of Esox under resting conditions at 20deg. C. was 3 1/2 times that of Cyprinus. The form of piscian growth curves (cf. Hile, '41) suggests that the respiratory coefficient is much higher for fishes towards the end of their normal life-span. Since the value obtained from Ivlev (above) is based on more extensive data than those of Moore, we shall ten- tatively accept 140 per cent as an aver- age respiratory coefficient for aquatic predators.
Considering that predators are usually more active than their herbivorous prey, which are in turn more active than the plants upon which they feed, it is not surprising to find that respiration with respect to growth in producers (33 per cent), in primary consumers (62 per cent) and in secondary consumers (~100 per cent) increases progressively. These differences probably reflect a trophic principle of wide application: the per- centage loss of energy due to respiration is progressively greater for higher levels in the food cycle.
Predation corrections.--In considering the predation losses from each
level, it is most convenient to begin with the highest level, An. In a
mechanically perfect food cycle composed of organ- ically discrete levels, this
loss by preda- tion obviously would be zero. Since no natural food cycle is so
mechanically constituted, some "cannibalism" within such an arbitrary level can
be expected, so that the actual value for predation loss from An probably will
be somewhat above zero. The predation loss from level An_l will represent the
total amount of assimilable energy passed on into the higher level (i.e., the
true productivity, )~n) ~ plus a quantity representing the average
content of substance killed but not assimilated by the predator, as will be
discussed in the following section. The predation loss from level An_2
will likewise represent the total amount of assimilable energy passed on to
the next
October. 1942 TROPHIC-DYNAMIC ASPECT OF ECOLOGY 405
level (i.e., Ln-1), plus a similar factor for unassimilated material, as illustrated by the data of tables II and III. The various categories of parasites are somewhat comparable to those of predators, but the details of their energy relationships have not yet been clarified, and cannot be included in this preliminary account.
Decomposition corrections.--In conformity with the principle of Le Chatelier, the energy of no food level can be completely extracted by the organisms which feed upon it. In addition to the energy represented by organisms which survive to be included in the "annual yield," much energy is contained in "killed" tissues which the predators are unable to digest and assimilate. Average coefficients of indigestible tissues, based largely of the calorific equivalents of the "crude fiber" fractions in the chemical analyses of Birge and Juday ('22), are as follows:
Nannoplankters ........ca. 5%
Algal mesoplankters.....5--35 %
Mature pondweeds.............ca. 20%
Primary consumers............ca. 10%
Secondary consumers. . . ca. 8%
Predatory fishes ca 5%
Corrections for terrestrial producers would certainly be much higher. Although the data are insufficient to warrant a generalization, these values suggest increasing digestibility of the higher food levels, particularly for the benthic components of aquatic cycles.
The loss of energy due to premature death from non-predatory causes usually must be neglected, since such losses are exceedingly difficult to evaluate and under normal conditions probably represent relatively small components of the annual production. However, considering that these losses may assume considerable proportions at any time, the above "decomposition coefficients" must be regarded as correspondingly minimal.
Following non-predated death, every organism is a potential source of energy for myriads of bacterial and fungal saprophages, whose metabolic products provide simple inorganic and organic
solutes reavailable to photosynthetic producers. These saprophages may also serve as energy sources for successive
levels of consumers, often considerably supplementing the normal diet of herbivores (ZoBell and Feltham, '38). Jacot
('40) considered saprophage-feeding or coprophagous animals as "low" primary consumers, but the writer believes that
in the present state of our knowledge a quantitative subdivision of primary consumers is unwarranted.
Application.--The value of these theoretical energy relationships can be illustrated by analyzing data of the three ecosystems for which relatively comprehensive productivity values have been published (table I). The summary account of Brujewicz ('39) on "the dynamics of living matter in the Caspian Sea" leaves much to be desired, as bottom animals are not differentiated into their relative food levels, and the basis for determining the annual production of phytoplankters (which on theoretical grounds appears to be much too low) is not clearly explained. Furthermore, his values are stated in terms of thousands of tons of dry weight for the Caspian Sea as a whole, and must be roughly transformed to calories per square centimeter of surface area. The data for Lake Mendota, Wisconsin, are
406 RAYMOND L. LINDEMAN Ecology, Vol. 23, No. 4
taken directly from a general summary (Juday, '40) of the many productivity studies made on that eutrophic lake. The data for Cedar Bog Lake, Minnesota, are taken from the author's four-year analysis (Lindeman, '41b) of its food-cycle dynamics. The calorific values in table I, representing annual production of organic matter, are uncorrected for energy losses.
TABLE 11. Productivity values for the Cedar
Bog Lake food cycle, in g-cal/cm2/year, as corrected
by using the coefficients derived in the preceding
sections.
* This value includes the productivity of the small cyprinoid fishes found in the lake.
Correcting for the energy losses due to respiration, predation and decomposi tion, as discussed in the preceding sec tions, casts a very different light on the relative productivities of food levels. The calculation of corrections for the Cedar Bog Lake values for producers, primary consumers and secondary consumers are given in table II. The application of similar corrections to the energy values for the food levels of the Lake Mendota food cycle given by Juday ('40), as shown in table I I I, indicates that Lake Mendota is much more productive of producers and primary consumers than is Cedar Bog Lake, while the production of secondary consumers is of the same order of magnitude in the two lakes.
In calculating total productivity for Lake Mendota, Juday ('40) used a blanket correction of 500 per cent of the annual production of all consumer levels for "metabolism," which presumably in cludes both respiration and predation. Thompson ('41) found that the "carrying-capacity" of lakes containing mostly carp and other "coarse" fishes (primarily A3), was about 500 per cent that of lakes containing mostly "game" fishes (primarily A;), and concluded that "this difference must be about one complete link in the food chain, since it usually requires about five pounds of food to produce one pound of fish." While such high "metabolic losses" may hold for tertiary and quaternary predators under certain field conditions, the physiological experiments previously cited indicate much lower respiratory coefficients. Even when predation and decomposition corrections are included, the resultant productivity values are less than half those obtained by using Juday's coefficient. Since we have shown that the necessary corrections vary progressively with the different food levels, it seems probable that Juday's "coefficient of metabolism" is much too high for primary and secondary consumers.
Biological efficiency
The quantitative relationships of any food-cycle level may be expressed in terms of its efficiency with respect to lower levels. Quoting Hutchinson's ('42)
TABLE III. Productivity values for the Lake Mendota food cycle, in g-cal/cm'/year, as corrected by using coefficients derived in the preceding sections, and as given by Juday ('40).
* Hutchinson ('42) gives evidence that this value is probably too high and may actually be as low as 250.
+ Apparently such organisms as small " forage " fishes are not included in any part of Juday's balance sheet. The inclusion of these forms might be expected to increase considerably the productivity of secondary consumption.
October, 1942 TROPHIC DYNAMIC ASPECT OF ECOLOGY 407
definition, "the efficiency of the productivity of any level (An) relative to the productivity of any previous Level (Am) is defined as lambda n/lambdam *100. If the rate of solar energy entering the ecosystem is denoted as Ao, the efficiencies of all levels may be referred back to this quantity lambda0." In general, however, the most interesting efficiencies are those referred to the previous level's productivity (lamdan-1) or those expressed as lamdan/lamdan-1*100. These latter may be termed the progressive efficiencies of the various food-cycle levels, indicating for each level the degree of utilization of its potential food supply or energy source. All efficiencies discussed in the following pages are progressive efficiencies, expressed in terms of relative productivities (lamdan/lamdan-1* 100) . It is important to remember that efficiency and productivity are not synonymous. Productivity is a rate (i.e., in the units here used, cal/cm2/year), while efficiency, being a ratio, is a dimensionless number. The points of reference for any efficiency value should always be clearly stated.
The progressive efficiencies (lamdan/lamdan-1* 100) for the trophic levels of Cedar Bog Lake and Lake Mendota, as obtained from the productivities derived in tables II and III, are presented in table IV. In view of the uncertainties concerning some of the Lake Mendota productivities, no definite conclusions can be drawn from their relative efficiencies(s. The Cedar Bog Lake ratios, however, indicate that the progressive efficiencies increase from about 0.10 per cent for production, to 13.3 per cent for primary consumption, and to 22.3 per cent for secondary consumption. An uncorrected efficiency of tertiary consumption of 37.5 per cent ~ 3.0 per eent (for the weight ratios of "carnivorous" to "forage" fishes in Alabama ponds) is indicated in data pub- lished by Swingle and Smith ('40). These progressively increasing efficiencies may well represent a fundamental trophic principle, namely, that the consumers at progressively higher levels in the food cycle are progressively more efficient in the use of their food supply.
At first sight, this generalization of increasing efficiency in higher consumer groups would appear to contradict the previous generalization that the loss of energy due to respiration is progressively greater for higher levels in the food cycle. These can be reconciled by remembering that increased activity of predators considerably increases the changes of encountering suitable prey. The ultimate effect of such antagonistic principles would present a picture of a predator completely wearing itself out in the process of completely exterminating its prey, a very improbable situation. However, Elton ('27) pointed out that food-cycles rarely have more than five trophic levels. Among the several faetors involved, increasing respiration of suecessive levels of predators contrasted with their successively increasing efficiency of predation appears to be important in restricting the number of trophic levels in a food cyclic.
The effect of increasing temperature is alleged by Wimpenny ('41) to cause a decreasing consumer/producer ratio, pre- sumably because he believes that the "acceleration of vital velocities" of con- sumers at increasing temperatures is more rapid than that of producers. He
TABLE IV. Productivities and progressive efficiencies in the Cedar Bog Lake and LakeMendota food cycles, as g-cal/cm2/year
*Probably too high, see footnote of table 111.
+Probably too low; see footnote of table III.
408 RAYMOND L. LINDEMAN Ecology, Vol. 23, No. 4
cites as evidence Lohmann's ('12) data for relative numbers (not biomass) of Protophyta, Protozoa an-l Metazoa in the centrifuge plankton of "cool" seas (741:73 :1) as contrasted with tropical areas (458:24:1). Since Wimpenny himself emphasizes that many metazoan plankters are larger in size toward the poles, these data do not furnish convincing proof of the allegation. The data given in table IV, since Cedar Bog Lake has a much higher mean annual water temperature than Lake Mendota appear to contradict Wimpenny's generalization that consumer/producer ratios fall as the temperature increases.
The Eltonian pyramid
The general relationships of higher food-cycle levels to one another and to community structure were greatly clarified following recognition (Elton, '27) of the importance of size and of numbers in the animals of an ecosystem. Beginning with primary consumers of various sizes, there are as a rule a number of food-chains radiating outwards in which the probability is that predators will become successively larger, while parasites and hyper-parasites will be progressively smaller than their hosts. Since small primary consumers can increase faster than larger secondary consumers and are so able to support the latter, the animals at the base of a food-chain are relatively abundant while those toward the end are progressively fewer in number. The resulting arrangement of sizes and numbers of animals, termed the pyramid of Numbers by Elton, is now commonly known as the Eltonian Pyramid. Williams ('41), reporting on the "floor fauna" of the Panama rain forest, has recently published an interesting example of such a pyramid, which is reproduced in figure 2.
The Eltonian Pyramid may also be expressed in terms of biomass. The weight of all predators must always be much lower than that of all food animals, and the total weight of the latter much lower than the plant production (Bodenheimer, '38). To the human ecologist, it is noteworthy that the population density of the essentially vegetarian Chinese, for example, is much greater than that of the more carnivorous English.
The principle of the Eltonian Pyramid has been redefined in terms of productivity by Hutchinson (unpublished) in the following formalized terms: the rate of production cannot be less and will almost certainly be greater than the rate of primary consumption, which in turn cannot be less and will almost certainly be greater than the rate of secondary consumption, which in turn . . .. etc. The energy-relationships of this principle may be epitomized by means
October, 1942 TROPHIC-DYNAMIC ASPECT OF ECOLOGY 409
of the productivity symbol L, as follows:
L0 > L1 > L2 . . . > Ln
This rather obvious generalization is confirmed by the data of all ecosystems analyzed to date.
TROPHIC-DYNAMICS IN SUCCESSION
Dynamic processes within an ecosystem, over a period of time, tend to produce certain obvious changes in its species-composition, soil characteristics and productivity. Change, according to Cooper ('26), is the essential criterion of succession. From the trophicdynamic viewpoint, succession is the process of development in an ecosystem, brought about primarily by the effects of the organisms on the environment and upon each other, towards a relatively stable condition of equilibrium.
It is well known that in the initial phases of hydrarch succession (oligotrophy ~ eutrophy) productivity increases rapidly; it is equally apparent that the colonization of a bare terrestrial area represents a similar acceleration in productivity. In the later phases of succession, productivity increases much more slowly. As Hutchinson and Wollack ('40) pointed out, these generalized changes in the rate of production may be expressed as a sigmoid curve showing a rough resemblance to the growth curve of an organism or of a homogeneous population.
Such smooth logistic growth, of course, is seldom found in natural succession, except possibly in such cases as bare areas developing directly to the climax vegetation type in the wake of a retreating glacier. Most successional seres consist of a number of stages ("recognizable, clearly-marked subdivisions of a given sere"--W. S. Cooper), so that their productivity growth-curves will contain undulations corresponding in distinctness to the distinctness of the stages. The presence of stages in a successional sere apparently represents the persistent influence of some combination of limiting factors, which, until they are overcome by species-substitution, etc.., tend to decrease the acceleration of productivity and maintain it at a more constant rate. This tendency towards stage-equilibrium of productivity will be discussed in the following pages.
Productivity in hydrarch succession
The descriptive dynamics of hydrarch succession is well known. Due to the essentially concave nature of the substratum, lake succession is internally complicated by a rather considerable influx of nutritive substances from the drainage basin surrounding the lake. The basins of lakes are gradually filled with sediments, largely organogenic, upon which a series of vascular plant stages successively replace one another until a more or less stable (climax) stage is attained. We are concerned here, however, primarily with the productivity aspects of the successional process.
Eutrophication.--Thienemann ('26) presented a comprehensive theoretical discussion of the relation between lake succession and productivity, as follows: In oligotrophy, the pioneer phase, productivity is rather low, limited by the amount of dissolved nutrients in the lake water. Oxygen is abundant at all times, almost all of the synthesized organic matter is available for animal food; bacteria release dissolved nutrients from the remainder. Oligotrophy thus has a very "thrifty" food cycle, with a relatively high "efficiency" of the consumer populations. With increasing influx of nutritives from the surrounding drainage basin and increasing primary productivity (L1) oligotrophy is gradually changed through mesotrophy to eutrophy, in which condition the production of organic matter (L1) exceeds that which can be oxidized (L1') by respiration, predation and bacterial decomposition. The oxygen supply of the hypolimnion becomes depleted, with disastrous effects on the oligotroph-
410 RAYMOND L. LINDEMAN Ecology, Vol. 23, No. 4
conditioned bottom fauna. Organisms especially adapted to endure temporary anaerobiosis replace the oligotrophic species, accompanied by anaerobic bacteria which during the stagnation period cause reduction rather than oxidation of the organic detritus. As a result of this process, semi-rcduced organic ooze, or gyttja, accumulates on the bottom. As oxygen supply thus becomes a limiting factor of productivity, relative efficiency of the consumer groups in utilizing the synthesized organic substance becomes correspondingly lower.
The validity of Thicnemann's interpretation, particularly regarding the trophic mechanisms, has recently been challenged by Hutchinson ('41, '42), who states that three distinct factors are involved: (1) the edaphic factor, representing the potential nutrient supply (primarily phosphorus) in the surrounding drainage basin; (2) the age of the lake at any stage, indicating the degree of utilization of the nutrient supply; and (3) the morphometric character at any stage, dependent on both the original morphometry of the lake basin and the age of the lake, and presumably influencing the oxygen concentration in the hypolimnion. He holds that true eutrophication takes place only in regions well supplied with nutrients, lakes in other regions developing into "ideotrophic types." The influx of phosphorus is probably very great in the earliest phases, much grater than the supply of nitrogen, as indicated by very low N/P ratios in the earliest sediments (Hutchinson and Wollack, '40). A large portion of this phosphorus is believed to be insoluble, as a component of such mineral particles as apatite, etc., although certainly some of it is soluble. The supply available nitrogen increases somewhat more slowly, being largely dependent upon the fixation of nitrogen by microorganisms either in the lake or in the surrounding soils. The photosynthetic productivity (L1) of lakes thus increases rather rapidly in the early phases, its quantitative value for lakes with comparable edaphic nutrient supplies being dependent on the morphometry (mean depth). Since deep lakes have a greater depth range for plankton photosynthesis, abundant oxygen and more chance for decomposition of the plankton detritus before reaching the bottom, such deep lakes may be potentially as productive as shallower lakes, in terms of unit surface area. Factors tending to lessen the comparative productivity of deep lakes are (1) lower temperature for the lake as a whole, and (2) greater dilution of nutrients in terms of volume of the illuminated "trophogenic zone" of the lake. During eutrophication in a deep lake, the phosphorus content of the sediment falls and nitrogen rises, until a N/P ratio of about 40/1 is established. "The decomposition of organic matter presumably is always liberating some of this phosphorus and nitrogen. Within limits, the more organic matter present the easier will be such regeneration. It is probable that benthic animals and anion exchange play a part in such processes" (Hutchinson, '42). The progressive filling of the lake basin with sediments makes the lake shallower, so that the oxygen supply of the hypolimnion is increasingly, and finally completely, exhausted during summer stagnation. Oxidation of the sediments is less complete, but sufficient phosphorus is believed to be regenerated from the ooze surface so that productivity in terms of surface area remains relatively constant. The nascent ooze acts as a trophic buffer, in the chemical sense, tending to maintain the productivity of a lake in stage-equilibrium typological equilibrium of Hutchinson) during the eutrophic stage of its succession.
The concept of eutrophic stage-equilibrium seems to be partially confused (cf. Thienemann, '26; Hutchinson and Wollack, '40) with the theoretically ideal condition of complete trophic equilibrium, which may be roughly defined as the dynamic state of continuous, complete
October, 1942 TROPHIC-DYNAMIC ASPECT OF ECOLOGY 411
utilization and regeneration of chemical nutrients in an ecosystem, without loss or gain from the outside, under a peri odically constant energy source--such as might be found in a perfectly balanced aquarium or terrarium. Natural eco systems may tend to approach a state of trophic equilibrium under certain con ditions, but it is doubtful if any are sufficiently autochthonous to attain, or maintain, true trophic equilibrium for any length of time. The biosphere as a whole, however, as Vernadsky ('29, '39) so vigorously asserts, may exhibit a high degree of true trophic equilibrium.
The existence of prolonged eutrophic stage-equilibrium was first suggested as a result of a study on the sediments of Grosser Ploner See in Germany (Groschopf, '36). Its significance was recognized by Hutchinson and Wollack ('40), who based their critical discussion on chemical analyses (ibid.) and pollen analyses (Deevey, '39) of the sediments of Linsley Pond, Connecticut. They reported a gradual transition from oligotrophy to eutrophy (first attained in the oak-hemlock pollen period), in which stage the lake has remained for a very long time, perhaps more than 4000 years. They report indications of a comparable eutrophic stage-equilibrium in the sediments of nearby Upper Linsley Pond (Hutchinson and Wollack, unpublished). Similar attainment of stage-equilibrium is indicated in a preliminary report on the sediments of Lake Windermere in England (Jenkin, Mortimer and Pennington, '41). Every stage of a sere is believed to possess a similar stage equilibrium of variable duration, although terrestrial stages have not yet been defined in terms of productivity.
The trophic aspects of eutrophication cannot be determined easily from the sediments. In general, however, the ratio of organic matter to the silt washed into the lake from the margins affords an approximation of the photosynthetic productivity. This undecomposed organic matter, representing the amount of energy which is lost from the food cycle, is derived largely from level Al, as plant structures in general are decomposed less easily than animal structures. The quantity of energy passed on into consumer levels can only be surmised from undecomposed fragments of organisms which are believed to occupy those levels. Several types of animal "microfossils" occur rather consistently in lake sediments, such as the carapaces and postabdomens of certain cladocerans, chironomid head-capsules, fragments of the phantom-midge larva Chaoborus, snail shells, polyzoan statoblasts, sponge spicules and rhizopod shells. Deevey ('42), after making comprehensive microfossil and chemical analyses of the sediments of Linsley Pond, suggested that the abundant half-carapaces of the planktonic browser Bosmina afford "a reasonable estimate of the quantity of zooplankton produced" and that "the total organic matter of the sediment is a reasonable estimate of the organic matter produced by phytoplankton and littoral vegetation." He found a striking similarity in the shape of the curves representing Bosmina content and total organic matter plotted against depth, which, when plotted logarithmically against each other, showed a linear relationship expressed by an empirical power equation. Citing Hutchinson and Wollack ('40) to the effect that the developmental curve for organic matter was analogous to that for the development of an organism, he pressed the analogy further by suggestting that the increase of zooplankton (Bosmina) with reference to the increasc of organic matter (L1) fitted the formula y = bxk for allometric growth (Huxley, '32), "where y = Bosmina, x = total organic matter, b = a constant giving the value of y when x = 1, and k = the 'allometry constant,' or the slope of the line when a double log plot is made." If we represent the organic matter produced as L1 and further assume that Bosmina represents the primary con sumers (L2), neglecting benthic browsers,
412 RAYMOND L. LINDEMAN Ecology, Vol. 23, No. 4
the formula bccomes L2 = bL1k . Whether this formula would express the relationship found in other levels of the food cyclc, the devclopmctlt of othcr stages, or othcr ccosystcms, rcmains to be demonstrated. footnote 6 Stratigraphic analyses in Cedar Bog Lake (Lindeman and Lindeman, unpublished) suggest a roughly similar increase of both organic matter and Bosmina carapaces in the earliest sediments. In the modern senescent lake, however, double logarithmic plottings of the calorific values for L1 against L2, and L2 against L3, for the four years studied, show no semblance of linear relationship, i.e., do not fit any power equation. If Deevey is correct in his interpretation of the Linsley Pond micro fossils, allometric growth would appear to characterize the phases of pre-equilibrium succession as the term "growth" indeed implies.
The relative duration of eutrophic stage-equilibrium is not yet completely understood. As exemplified by Linsley Pond, the relation of stage-equilibrium to succession is intimately concerned with the trophic processes of (1) external influx and efflux (partly controlled by climate), (2) photosynthetic productivity, (3) sedimentation (partly by physiographic silting) and (4) regeneration of nutritives from the sediments. These processes apparently maintain a relatively constant ratio to each other during the extended equilibrium period. Yet the food cycle is not in true trophic equilibrium, and continues to fill the lake with organic sediments. Succession is continuing, at a rate corresponding to the rate of sediment accumulation. In the words of Hutchinson and Wollack ('40), "this means that during the equilibrium period the lake, through the internal activities of its biocoenosis, is continually approaching a condition when it ceases to be a lake." Senescence.--As a result of long-continued sedimentation, eutrophic lakes attain senescence, first manifested in bays and wind-protected areas. Senescence is usually characterized by such pond-like conditions as (1) tremendous increase in shallow littoral area populated with pondweeds and (2) increased marginal invasion of terrestrial stages. Cedar Bog Lake, which the author has studied for several years, is in late senescence, rapidly changing to the terrestrial stages of its succession. On casual inspection, the massed verdure of pond weeds and epiphytes, together with sporadic algal blooms, appears to indicate great photosynthetic productivity. As pointed out by Wesenberg-Lund ('12), littoral areas of lakes are virtual hot houses, absorbing more radiant energy per unit volume than deeper areas. At the present time the entire aquatic area of Cedar Bog Lake is essentially littoral in nature, and its productivity per cubic meter of water is probably greater than at any time in its history. However, since radiant energy (L0) enters a lake only from the surface, productivity must be defined in terms of surface area. In these terms, the present photosynthetic productivity pales into insignificance when compared with less advanced lakes in similar edaphic regions; for instance, L1, is less than 1/3 that of Lake Mendota, Wisconsin (cf. table IV). These facts attest the essential accuracy of Welch's ('35) generalization that productivity declines greatly during senescence. An interesting principle demonstrated in Cedar Bog Lake (Lindeman, '41b) is that during late lake senescence general productivity (Ln) increasingly influenced by climatic factors, acting through
October, 1942 TROPHIC-DYNAMIC ASPECT OF ECOLOGY 413
water level changes, drainage, duration of winter ice, snow cover, etc., to affect the presence and abundance of practically all food groups in the lake.
Terrestrial stages.--As an aquatic ecosystem passes into terrestrial phases, fluctuations in atmospheric factors in creasingly affect its productivity. As succession proceeds, both the species composition and the productivity of an ecosystem increasingly reflect the effects of the regional climate. Qualitatively, these climatic effects are known for soil morphology (Joffe, '36), autotrophic vegetation (Clements, '16), fauna (Clements and Shelford, '39) and soil microbiota (Braun-Blanquet, '32), in fact for every important component of the food cycle. Quantitatively, these effects have been so little studied that generalizations are most hazardous. It seems probable, however, that productivity tends to increase until the system approaches maturity. Clements and Shelford ('39, p. 116) assert that both plant and animal productivity is generally greatest in the subclimax, except possibly in the case of grasslands. Terrestrial ecosystems are primarily convex topographically and thus subject to a certain nutrient loss by erosion, which may or may not be made up by increased availability of such nutrients as can be extracted from the "C" soil horizon.
Successional productivity curves.--In recapitulating the probable photosynthetic productivity relationships in hydrarch succession, we shall venture to diagram (figure 3) a hypothetical hydrosere, developing from a moderately deep lake in a fertile cold temperate region under relatively constant climatic conditions. The initial period of oligotrophy is believed to be relatively short (Hutchinson and Wollack, '40; Lindeman '41a), with productivity rapidly increasing until eutrophic stage-equilibrium is attained. The duration of high eutrophic productivity depends upon the mean depth of the basin and upon the rate of sedimentation, and productivity fluctuates about a high eutrophic mean until the lake becomes too shallow for maximum growth of phytoplankton or regeneration of nutrients from the ooze. As the lake becomes shallower and more senescent, productivity is increasingly influenced by climatic fluctuations and
414 RAYMOND L. LINDEMAN Ecology, Vol. 23, No. 4
gradually declines to a minimum as the lake is completely filled with sediments. The terrestrial aspects of hydrarch succession in cold temperate regions usually follow sharply defined, distinctive stages. In lake basins which are poorly drained, the first stage consists of a mat, often partly floating, made up primarily of sedges and grasses or (in more coastal regions) such heaths as Chamaedaphne and Kalmia with certain species of sphagnum moss (cf. Rigg, '40). The mat stage is usually followed by a bog forest stage, in which the dominant species is Larix laricina, Picea mariana or Thuja occidentalis. The bog forest stage may be relatively permanent ("edaphic" climax) or succeeded to a greater or lesser degree by the regional climax vegetation. The stage-productivities indicated in figure 3 represent only crude relative estimates, as practically no quantitative data are available.
Elficiency relationships in succession
The successional changes of photosynthetic efficiency in natural areas (with respect to solar radiation, i.e., L1/L0*100 have not been intensively studied. In lake succession, photosynthetic efficiency would be expected to follow the same course deduced for productivity, rising to a more or less constant value during eutrophic stage-equilibrium, and declining during senescence, as suggested by a photosynthetic efficiency of at least 0.27 per cent for eutrophic Lake Mendota (Juday, '40) and of 0.10 per cent for senescent Cedar Bog Lake. For the terrestrial hydrosere, effciency would likewise follow a curve similar to that postulated for productivity. Rough estimates of photosynthetic effciency for various climatic regions of thc earth have been summarized from the literature by Hutchinson (unpublished). These estimates, corrected for respiration, do not appear to be very reliable because of imperfections in the original observations, but are probably of the correct order of magnitude. The mean photosynthetic efficiency for the sea is given as 0.31 per cent (after Riley, '41). The mean photosynthetic efficiency for terrestrial areas of the earth is given as 0.09 per cent + or - 0.02 per cent (after Noddack, '37), for forests as 0.16 per cent, for cultivated lands as 0.13 per cent, for steppes as 0.05 per cent, and for deserts as 0.004 per cent. The mean photosynthetic efficiency for the earth as a whole is given as 0.25 per cent. Hutchinson has suggested (cf. Hutchinson and Lindeman, '41) that numerical efficiency values may provide "the most fundamental possible classification of biological formations and of their developmental stages." Almost nothing is known concerning the efficiencies of consumer groups in succession. The general chronological increase in numbers of Bosmina carapaces with respect to organic matter and of Chaoborus fragments with respect to Bosmina carapaces in the sediments of Linsley Pond (Deevey, '42) suggests progressively increasing efficiencies of zooplankters and plankton predators. On the other hand, Hutchinson ('42) concludes from a comparison of the P: Z (phytoplankton: zooplankton) biomass ratios of several oligotrophic alpine lakes, ca 1: 2 (Ruttner, '37), as compared with the ratios for Linsley Pond, 1: 0.22 (Riley, '40) and three eutrophic Bavarian lakes, 1: 0.25 (Heinrich, '34), that "as the phytoplankton crop is increased the zooplankton by no means keeps pace with the increase." Data compiled by Deevey ('41) for lakes in both mesotrophic (Connecticut) and eutrophic regions (southern Wisconsin), indicate that the deeper or morphometrically "younger" lakes have a lower ratio of bottom fauna to the standing crop of plankton (10-15 per cent) than the shallower lakes which have attained eutrophic equilibrium (22-27 per cent). The ratios for senescent Cedar Bog Lake, while not directly comparable because
October, 1942 TROPHIC-DYNAMIC ASPECT OF ECOLOGY 415
of its essentially littoral nature, are even higher. These meager data suggest that the efficiencies of consumer groups may increase throughout the aquatic phases of succession. For terrestrial stages, no consumer efficiency data are available. A suggestive series of species-frequencies in mesarch succession was published by Vera Smith-Davidson ('32), which indicated greatly increasing numbers of arthropods in successive stages approach ing the regional climax. Since the photosynthetic productivity of the stages probably also increased, it is impossible to determine progressive efficiency relationships. The problems of biological efficiencies present a practically virgin field, which appears to offer abundant rewards for studies guided by a trophic dynamic viewpoint.
In conclusion, it should be emphasized that the trophic-dynamic principles indicated in the following summary cannot be expected to hold for every single case, in accord with the known facts of biological variability. a priori, however, these principles appear to be valid for the vast majority of cases, and may be expected to possess a statistically significant probability of validity for any case selected at random. Since the available data summarized in this paper are far too meager to establish such generalizations on a statistical basis, it is highly important that further studies be initiated to test the validity of these and other trophic-dynamic principles.
SUMMARY
1. Analyses of food-cycle relationships indicate that a biotic community cannot be clearly differentiated from its abiotic environment; the ecosystem is hence regarded as the more fundamental ecological unit.
2. Thc organisms within an ecosystem may be grouped into a series of more or Iess discretc trophic levels (A1,A2, A3. . .An) as producers, primary consumers, secondary consumers, ctc., each successively dependent upon the preceding level as a source of energy, with the producers (Al) directly dependent upon the rate of incident solar radiation (pro ductivity L0) as a source of energy.
3. The more remote an organism is from the initial source of energy (solar radiation), the less probable that it will be dependent solely upon the preceding trophic level as a source of energy.
4. The progressive energy relationships of the food levels of an "Eltonian Pyramid" may be epitomized in terms of the productivity symbol L, as follows:
L0 > L1 > L2 > . . > Ln
5. The percentage loss of energy due to respiration is progressively greater for higher levels in the food cycle. Respiration with respect to growth is about 33 per cent for producers, 62 per cent for primary consumers, and more than 100 per cent for secondary consumers.
6. The consumers at progressively higher levels in the food cycle appear to be progressively more efficient in the use of their food supply. This generalization can be reconciled with the preceding one by remembering that increased activity of predators considerably increases the chances of encountering suitable prey.
7. Productivity and efficiency increase during the early phases of successional development. In lake succession, productivity and photosynthetic efficiency increase from oligotrophy to a prolonged eutrophic stage-equilibrium and decline with lake senescence, rising again in the terrestrial stages of hydrarch succession. 8. The progressive efficiencies of consumer levels, on the basis of very meager data, apparently tend to increase through out the aquatic phases of succession.
ACKNOWLEDGEMENTS
The author is deeply indebted to Professor G. E. Hutchinson of Yale University, who has stimulated many of the trophic concepts developed here, generously placed at the author's
416 RAYMOND L. LINDEMAN Ecology, Vol. 23, No. 4
disposal several unpublished manuscripts, given valuable counsel, and aided the final development of this paper in every way possible. Many of the concepts embodied in the successional sections of this paper were developed independently by Professor Hutchinson at Yale and by the author as a graduate student at the University of Minnesota. Subscquent to an exchange of notes, a joint preliminary abstract was published (Hutchinson and l.indcman, '41). The author wishes to express gratitude to the mentors and friends at the University of Minnesota who encouraged and hclpfully criticized the initial devclopment of these concepts, particularly Drs. W. S. Cooper, Samuel Eddy, A. C. Hodson, D. B. Lawrence and J. B. Moyle, as well as other members of the local Ecological Discussion Group. The author is also indebted to Drs. J. R. Carpenter, E. S. Deevey, H. J. Lutz, A. E. Parr, G. A. Riley and V. E. Shelford, as well as the persons mentioned above, for critical reading of preliminary manuscripts. Grateful acknowledgment is made to the Graduate School, Yale University, for the award of a Sterling Fellowship in Biology during 1941-1942.
LITERATURE CITED
Allee, W. C. 1934. Concerning the organization of marine coastal communities. Ecol. Monogr., 4: 541-554.
Birge, E. A., and C. Juday. 1922. The inland lakes of Wisconsin. The plankton. Part 1. Its quantity and chemical composition. Bull. Wisconsin Geol. Nat. Hist. Surv., o4: 1-222.
Bodenheimer,F. S. 1938. Problems of Animal Ecology. London. Oxford University Press.
Braun-Blanquet, J. 1932. Plant Sociology.
N. Y. McGraw-Hill Co.
Brujewicz, S. W. 1939. Distribution and dy
namics of living matter in the Caspian Sea.
Compt..Rend. Acad. Sci. URSS, 25: 138-141.
Carpenter, J. R. 1939. The biome. Amer.
Midl. Nat., 21: 75-91. --. 1940. The grassland biome. Ecol.
Monogr. 10: 617-687.
Clements, F E. 1916. Plant Succession. Car
negie Inst. Washington Publ., No. 242. and V. E. Shelford. 1939. Bio-Ecologv. N. Y. John Wiley & Co.
Cooper, W. S. 1926. The fundamentals of
vegetational change. Ecology, 7: 391-413.
Cowles, H. C. 1899. The ecological relations
of the vegetation of the sand dunes of Lake
Michigan. Bot. Gaz., 27: 95-391.
Davidson, V. S. 1932. lhe effect of seasonal
variability upon animal species in a dccidu
ous forest sll(ccssioll. E(ol. Monogr., 2:
30.~ - .33~ .
Deevey, E. S. 1~).3~) Stlldics on Col-llccticllt
lake x( dilllcllts: I . A postgldcial clilllatic chro
nology for southerll Ncw Ellgland. Amcr.
Jour. Sci., 237: 691-724.
1941. Limnological studies in Connecti cut: Vl. The quantity and composition of the bottom fauna. Ecol. Monogr., 11: 413-455.
1942. Studies on Connecticut lake sedi ments: 111. The biostratonomy of Linsley Pond. Amer. Jour. Sci., 240: 233-264, 313 338.
Elton, C. 1927. Animal Ecology. N. Y.
Macmillan Co.
France, R. H. 1913. Das Edaphon, Unter
suchungen zur Oekologie der bodenbewohnen
den Mikroorganismen. Deutsch. Mikrolog.
Gesellsch., Arbeit. aus d. Biol. Inst., No. 2.
Munich .
Friederichs, K. 1930. Die Grundfragen und
Gesetzmassigkeiten der land- und forstwirt
schaftlichen Zoologie. 2 vols. Berlin. Ver
schlag. Paul Parey.
Groschopf, P. 1936. Die postglaziale Ent
wicklung des Grosser Ploner Sees in Ost
holstein auf Grund pollenanalytischer Sedi
mentuntersuchungen. Arch. Hydrobiol., 30:
1-84.
Heinrich, K. 1934. Atmung und Assimilation
im freien Wasser. Internat. Rev. ges. Hy
drobiol. u. Hydrogr., 30: 387-410.
Hentschel, E. 1933-1936. Allgemeine Biologie
des Sudatlantischen Ozeans. Wiss. Ergebn.
Deutsch. Atlant. Exped. a. d. Forschungs- u.
Vermessungsschiff "Meteor" 1925-1927. Bd .
XI.
Hicks, P. A. 1934. Interaction of factors in
the growth of Lcmna: V. Some preliminary
observations upon the interaction of tem
perature and light on the growth of Lemna.
Ann. Bot., 48: 515-523.
Hile, R. 1941. Age and growth of the rock
bass Ambloplites rupestris (Rafinesque) in
Nebish Lake, Wisconsin. Trans. Wisconsin
Acad. Sci., Arts, Lett., 33: 189-337.
Hutchinson, G. E. 1941. Limnological studies
in Connecticut: IV. Mechanism of interme
diary metabolism in stratified lakes. Ecol.
Monogr., ll: 21-60. ---.
1942. Recent Advances in Limnology
(in manuscript).
--and R. L. Lindeman. 1941. Biological
efficiency in succession (Abstract). Bull.
Ecol. Soc. Amer., Z2: 44.
and Anne Wollack. 1940. Studies on Connecticut lake sediments: II. Chemical analyses of a core from Linsley Pond, North Branford. Amer. Jour. Sci., 238: 493-517. Huxley, J. S. 1932. Problems of Relative
Growth. N. Y. Dial Press.
Ivle~, V. S. 1939a. Transformation of energy
by aquatic animals. Internat. Rev. ges.
Hydrobiol. u. Hydrogr., 38: 449-458. ----. 1939b. Balance of energy in carps.
Zool. Zhurn. Moscow, 18: 303-316.
Jacot, A. P. 1940. I he fauna of the soil.
Quart. Rev. Biol., 15: 28-58.
October, 1942 TROPHIC-DYNAMIC ASPECT OF ECOLOGY 417
Jenkin, B. M., C. H. Mortimer, and W. Penning
ton. 1941. I he stu(ly of lake deposits.
Nature, 147: 496-500. Joffe, J. S. 1936. Pedology. New Brunswick,
New Jersey. Rutgers Univ. Press.
Juday, C. 1922. Quantitative studies of the
bottom fauna in the deeper waters of Lake
Mendota. Trans. Wisconsin Acad. Sci.,
Arts, Lett., 20: 461-493.
. 1940. The annual energy budget of an inland lake. Ecology, 21: 438-450.
Krogh, A. 1941. I he Comparative Physiology
of Respiratory Mechanisms. Philadelphia.
IJIliv. Pennsylvania Press.
Lindeman, R. L. 1941a. The developmental
history of Cedar Creek Bog, Minnesota.
Amer. Midl. Nat., 25: 101-112.
1941b. Seasonal food-cycle dynamics in a senescent lake. Amer. Midl. Nat., 26: 636-673.
Lohmann, H. 1912. Untersuchungen uber das
I'fianzen- und Tierleben der Hochsee, zu
gleich ein Bericht uber die biologischen Ar
beiten auf der Fahrt der "Deutschland" von
Bremerhaven nach Buenos Aires. Veroffentl.
d. Inst. f. Meereskunde, N.F., A. Geogr.
naturwissen. Reihe, Heft 1, 92 pp.
Lundegardh, H. 1924. Kreislauf der Kohlen
saure in der Natur. Jena. G. Fischer.
Meschkat, A. 1937. Abwasserbiologische Un
tersuchungen in einem Buhnenfeld unterhalb
Hamburgs. Arch. Hydrobiol., 31: 399-432.
Moore, W. G. 1941. Studies on the feeding
habits of fishes. Ecology, 22: 91-95.
Needham, J. 1931. Chemical Embryology.
3 vols. N. Y. Cambridge University Press.
Noddack, W. 1937. Der Kohlenstofl im Haus
halt der Natur. Zeitschr. angew. Chemie,
50: 505-510.
Park, Thomas. 1941. The laboratory popula
tion as a test of a comprehensive ecological
system. Quart. Rev. Biol., 16: 274-293
440-461.
Rigg, G. B. 1940. Comparisons of the devel
opment of some Sphagnum bogs of the At
lantic coast, the interior, and the Pacific
coast. Amer. Jour. Bot., 27: 1-14.
Riley, G. A. 1940. Limnological studies in
Connecticut. III. The plankton of Linslev
Pond. Ecol. Monogr., 10: 279-306.
. 1941. Plankton studies. 111. Long Is land Sound. Bull. Bingham Oceanogr. Coll. 7 (3): 1-93.
Ruttner, F. 1937. Limnologische Studien an
einigen Seen der Ostalpen. Arch. Hydro
biol., 32: 167-319.
Smith-Da~ridson, Vera. 1932. The eflect of
seasonal variability upon animal species in a
deciduous forest succession. Ecol. Monogr.,
2: 305-334.
Spoehr, H. A. 1926. Photosynthesis. N. Y.
Chemical Catalo~ue Co.
Swingle, H. S., and E. V. Smith. 1940. Fb:
periments on thc stocking of fish pon(l~.
'I'rans. North Alllcr. Wildlifc Conf., 5: 267
276.
Tansley, A. G. 1935. The use an(l abuse of
vegetational concepts and terms. Ecologv,
16: 284-307.
Thienemann, A. 1918. Lebensgemeillschaft
und Lebensraum. Naturw. Wochcnschrift,
N.F., 17: 282-290, 297-303.
. 1926. Der Nahrungskreislallf im Was ser. Verh. deutsch. Zool. Ges., 31: 29-79. (or) Zool. Anz. Suppl., 2: 29-79. 1939. Grundzuge einen allgemeincn Oekologie. Arch. Hydrobiol., 35: 267-285.
Thompson, D. H. 1941. 'I'he fish production
of inland lakes and streams. Symposium
on Hydrobiology, pp. 206-217. Madison.
Univ. Wisconsin Press.
Tutin, T. G. 1941. The hydrosere and current
concepts of the climax. Jour. Ecol. 29: 268
279.
Vernadsky, V. I. 1929. La biosphere. Paris.
Librairie Felix Alcan. . 1939 On some fundamental problems of biogeochemistry. Trav. Lab. Biogeochem. Acad. Sci. URSS, 5: 5-17.
Waksman, S. A. 1941 Aquatic bacteria in
relation to the cycle of organic matter in
lakes. Symposium on Hydrobiology, pp.
86-105. Madison. Univ. Wisconsin Press.
Welch, P. S. 1935. Limnology. N. Y. Mc
Graw-Hill Co.
Wesenberg-Lund, C. 1912. l jber einige eigen
tumliche Temperaturverhaltnisse in der Lito
ralregion.... Internat. Rev. ges. Hydro
biol. u. Hydrogr., 5: 287-316.
Williams, E. C. 1941. An ecological study of
the floor fauna of the Panama rain forcst.
Bull. Chicago Acad. Sci., 6: 63-124.
Wimpenny, R. S. 1941. Organic polarity:
some ecological and physiological aspects.
Quart. Rev. Biol., 16: 389-425.
ZoBell, C. E., and C. B. Feltham. 1938. Bac
teria as food for certain marine invertebrates.
Jour. Marine Research, 1: 312-327.
ADDENDUM
While this, his sixth completed paper, was in the press, Raymond Lindeman died after a long illness on 29 June, 1942, in his twenty-seventh year. While his loss is grievous to all who knew him, it is more fitting here to dwell on the achievements of his brief working life. The present paper represents a synthesis of Lindeman's work on the modern ecology and past history of a small senescent lake in Minnesota. In studying this locality he came to realize, as others before him had done, that the most profit able method of analysis lay in reduction of all the interrelated biological events to energetic terms. The attempt to do this led him far
418 RAYMOND L. LINDEMAN Ecology, Vol. 23, No. 4
beyond the immediate problem in hand, and in stating his conclusions he felt that he was providing a program for further studies. Knowing that one man's life at best is too short for intensive studies of more than a few localities, and before the manuscript was completed, that he might never return again to the field, he wanted others to think in the same terms as he had found so stimulating, and for them to collect material that would confirm, extend, or correct his theoretical conclusions. The present contribution does far more than this, as here for the first time, we have the interrelated dynamics of a biocoenosis presented in a form that is amenable to a productive abstract analysis. The question, for instance, arises, " What determines the length of a food chain? " the answer given is admittedly imperfect, but it is far more important to have seen that there is a real problem of this kind to be solved. That the final statement of the structure of a biocoenosis consists of pairs of numbers, one an integer determining the level, one a fraction determining the efficiency, may even give some hint of an undiscovered type of mathematical treatment of biological communities. Though Lindeman's work on the ecology and history of Cedar Bog Lake is of more than local interest, and will, it is hoped, appear of even greater significance when the notes made in the last few months of his life can be coordinated and published, it is to the present paper that we must turn as the major contribution of one of the most creative and generous minds yet to devote itself to ecological science. G. EVELYN HUTCHINSON. YALE UNIVERSITY.