6. E. L. Thorndike, Animal Intelligence (Mac- 16. 3. Kaas, W. C. Hall, I. T. Diamond, "Dual millan, New York, Neurol. 91, 441 (1949); -, in Biological 1911). representation of the retina in the cortex of and Biochemical Bases of Behavior, H. F. 7. E. D. Adrian, The Physical Background of the hedgehog in relation to architectonic Harlow and C. N. Perception (Oxford Univ. Press, London, Woolsey, Eds. (Univ. of boundaries and thalamic projections," in prep- Wisconsin Press, Madison, 1958); L. T. Dia- 1947). aration. mond, K. L. Chow, W. D. Neff, J. Comp. 8. D. H. Raab and H. W. Ades, Amer. J. 17. R. A. Lende and K. M. Sadler, Brain Res. Neurol. 109, 349 (1958). Psychol. 59, 59 (1946); M. R. Rosenzweig, 5, 390 (1967). 29. L. J. Garey and T. P. S. Powell, Proc. Roy. ibid., p. 127; R. A. Butler, I. T. Diamond, 18. P. Abplanalp and W. Nauta, personal com- Soc. London Ser. B 169, 107 (1967); M. W. D. Neff, J. Neurophysiol. 20, 108 (1957); munication. Glickstein, R. A. King, J. Miller, M. Berkley, R. F. Thompson, ibid. 23, 321 (1960). 19. J. Altman and M. B. Carpenter, J. Comp. J. 9. W. D. Neff and I. T. Comp. Neurol. 130, 55 (1967). Diamond, in Biological Neurol. 116, 157 (1961); D. K. Morest, Anat. 30. The abbreviations are and Biochemical Bases of Behavior, H. F. following used in the Record 151, 390 (1965); E. C. Tarlov and figures and figure legends: A, auditory area of Harlow and C. N. Woolsey, Eds. (Univ. of R. Y. Moore, J. Comp. Neurol. 126, 403 cortex; CM + Pf, centromedian + parafas- Wisconsin Press, Madison, 1958); R. B. Mas- (1966). cicular nuclei; GL, lateral geniculate nucleus; terton and I. T. Diamond, J. Neurophysiol. 20. H. S. Gasser and J. Erlanger, Amer. J. Physiol. GM, medial geniculate nucleus; habe- 27, 15 (1964). 88, 581 Hab, 10. (1929). nula; Hip, hippocampus; L, lateral group of 1. M. Warren, Ann. Rev. Psychol. 16, 95 21. G. H. Bishop, J. Nervous Mental Disease 128, nuclei; LP, lateroposterior nucleus; MD, medi- (1965). 89 (1959). odorsal nucleus; PC, posterior commissure; 11. K. S. Lashley, in Symposia of the Society for 22. C. J. Herrick, Brains of Rats and Men (Univ. Experimental Biology (Academic Press, New Po, posterior nucleus; Pul, pulvinar nucleus; of Chicago Press, Chicago, 1926). R, reticular nucleus; S I, somatic area I; York, 1950), vol. 4. 23. R. P. Erickson, W. C. Hall, J. A. Jane, M. S 12. II, somatic area II; SC, superior colliculus; K. Z. Lorenz, ibid., vol. 4. Snyder, I. T. Diamond, J. Comp. Neurol. 131, area of 13. G. Elliot Smith, T, temporal cortex; TO, optic tract; The Evolution of Man (Ox- 103 (1967). ventral V visual area V ford Univ. Press, London, 1924); V, group; I, I; II, C. J. Her- 24. J. E. Rose and C. N. Woolsey, Electroen- visual area II; VGL, ventral lateral geniculate; rick, Brain of the Tiger Salamander (Univ. cephalog. Clin. Neurophysiol. 1, 391 (1949). VL, ventrolateral nucleus; VM, ventromedial of Chicago Press, Chicago, 1948). 25. Unpublished experiments in our laboratory. nucleus; VP, ventroposterior nucleus; ZI, 14. G. G. Simpson, Bull. Amer. Museum Nat. 26. M. Snyder, W. C. Hall, I. T. Diamond, Psy- zona incerta. Hist. 85, 1307 (1945); A. S. Romer, Vertebrate chonomic Sci. 6, 243 (1966); M. Snyder and 31. The research described was supported by grant Paleontology (Univ. of Chicago Press, Chi- I. T. Diamond, Brain Behavior Evolution 1, MH-04849 from the National cago, 1945); J. Z. Young, The Life of Ver- Institute of 244 (1968). Mental Health. For their helpful comments we tebrates (Oxford Univ. Press, London, 1962). 27. H. Killackey, I. T. Diamond, W. C. Hall, G. thank Doctors G. Bishop, J. Kaas, G. Kimble, 15. W. C. Hall and I. T. Diamond, Brain, Be. Hudgins, Federation Proc. 27, 517 (1968). N. Guttman, J. Jane, R. B. Masterton, and havior Evolution 1, 181 (1968). 28. J. E. Rose and C. N. Woolsey, J. Comp. M. Warren.

at the level as a means of emphasizing those aspects of that can be accepted on the on March 3, 2011 basis of present knowledge, those that require more study, and those that have The Strategy of special relevance to human . Ecosystem Development Definition of Succession Ecological succession may be defined An understanding of ecological succession provides www.sciencemag.org in terms of the following three param- a basis for resolving man's conflict with . eters (1). (i) It is an orderly process of development that is rea- sonably directional and, therefore, pre- Eugene P. Odum dictable. (ii) It results from modifica- tion of the physical environment by the community; that is, succession is com- munity-controlled even though the phys- Downloaded from The principles of ecological succes- gists to regard "succession" as a single ical environment determines the pattern, sion bear importantly on the relation- straightforward idea; in actual fact, it the rate of change, and often sets limits ships between man and nature. The entails an interacting complex of proc- as to how far development can go. (iii) framework of successional theory needs esses, some of which counteract one It culminates in a stabilized ecosystem to be examined as a basis for resolving another. in which maximum (or high man's present environmental crisis. Most As viewed here, ecological succession information content) and symbiotic ideas pertaining to the development of involves the development of ; function between organisms are main- ecological systems are based on descrip- it has many parallels in the develop- tained per unit of available . tive data obtained by observing changes mental biology of organisms, and also In a word, the "strategy" of succession in biotic communities over long periods, in the development of human society. as a short-term process is basically the or on highly theoretical assumptions; The ecosystem, or ecological system, is same as the "strategy" of long-term very few of the generally accepted hy- considered to be a unit of biological evolutionary development of the bio- potheses have been tested experimental- organization made up of all of the or- sphere-namely, increased control of, ly. Some of the confusion, vagueness, ganisms in a given area (that is, "com- or homeostasis with, the physical en- and lack of experimental work in this munity") interacting with the physical vironment in the sense of achieving area stems from the tendency of ecolo- environment so that a flow of energy maximum protection from its pertur- The author is director of the Institute of Ecol- leads to characteristic trophic structure bations. As I illustrate below, the strat- ogy, and Alumni Foundation Professor, at the and material cycles within the system. egy of "maximum protection" (that is, University of Georgia, Athens. This article is based on a presidential address presented before It is the purpose of this article to sum- trying to achieve maximum support of the annual meeting of the Ecological Society of marize, in the form of a tabular model, complex biomass structure) often con- America at the University of Maryland, August 1966. components and stages of development flicts with man's goal of "maximum 262 SCIENCE, VOL. 164 production" (trying to obtain the high- ter and biomass (B) will accumulate in early seasonal bloom characterized by est possible yield). Recognition of the the system (Table 1, item 6), with the rapid growth of a few dominant species ecological basis for this conflict is, I result that ratio P/B will tend to de- being followed by the development later believe, a first step in establishing crease or, conversely, the BIP, BIR, or in the season of high B/P ratios, in- rational land-use policies. BIE ratios (where E = P + R) will creased diversity, and a relatively steady, The earlier descriptive studies of suc- increase (Table 1, items 2 and 3). Theo- if temporary, state in terms of P and R cession on sand dunes, grasslands, for- retically, then, the amount of standing- (4). Open systems may not experience ests, marine shores, or other sites, and crop biomass supported by the available a decline, at maturity, in total or gross more recent functional considerations, energy flow (E) increases to a maximum , as the space-limited micro- have led to the basic theory contained in in the mature or climax stages (Table 1, cosms do, but the general pattern of bio- the definition given above. H. T. Odum item 3). As a consequence, the net energetic change in the latter seems to and Pinkerton (2), building on Lotka's community production, or yield, in an mimic nature quite well. (3) "law of maximum energy in bio- annual cycle is large in young nature These trends are not, as might at first logical systems," were the first to point and small or zero in mature nature seem to be the case, contrary to the out that succession involves a funda- (Table 1, item 4). classical limnological teaching which mental shift in energy flows as increas- describes lakes as progressing in time ing energy is relegated to maintenance. from the less productive (oligotrophic) Margalef (4) has recently documented Comparison of Succession in a to the more productive (eutrophic) this bioenergetic basis for succession and Laboratory Microcosm and a Forest state. Table 1, as already emphasized, has extended the concept. refers to changes which are brought Changes that occur in major structur- One can readily observe bioenergetic about by biological processes within the al and functional characteristics of a changes by initiating succession in ex- ecosystem in question. Eutrophication, developing ecosystem are listed in Table perimental laboratory microecosystems. whether natural or cultural, results when 1. Twenty-four attributes of ecological Aquatic microecosystems, derived from are imported into the lake systems are grouped, for convenience of various types of outdoor systems, such from outside the lake-that is, from the discussion, under six headings. Trends as ponds, have been cultured by Beyers watershed. This is equivalent to adding are emphasized by contrasting the situ- (5), and certain of these mixed cultures nutrients to the

laboratory on March 3, 2011 ation in microecosys- early and late development. The are easily replicated and maintain them- tem or a field; the system is degree of absolute fertilizing change, the rate of selves in the climax state indefinitely on pushed back, in successional terms, to a change, and the time required to reach defined media in a flask with only light younger or "bloom" state. Recent stud- a steady state may not vary only with input (6). If samples from the climax ies on lake sediments (10), as well as different climatic and physiographic sit- system are inoculated into fresh media, theoretical considerations (11), have in- uations but also with different ecosystem succession occurs, the mature system dicated that lakes can and do attributes in progress the same physical environ- developing in less than 100 days. In to a more oligotrophic condition when ment. Where good data are available, Fig. 1 the general pattern of a 100-day the input from the watershed rate-of-change curves are usually con- autotrophic succession in a microcosm slows or ceases. Thus, there is hope that www.sciencemag.org vex, with changes occurring most rapid- based on data of Cooke (7) is compared the troublesome cultural at eutrophication ly the beginning, but bimodal or cy- with a hypothetical model of a 100- of our can be reversed if the in- clic patterns may also occur. year forest succession as presented by flow of nutrients from the watershed Kira and Shidei (8). can be greatly reduced. Most of all, During the first 40 to 60 days in a however, this situation that Bioenergetics of Ecosystem emphasizes typical microcosm experiment, daytime it is the entire drainage or catchment Development net production (P) exceeds nighttime not the lake or basin, just stream, that Downloaded from respiration (R), so that biomass (B) must be considered the ecosystem unit Attributes 1 5 in through Table 1 accumulates in the system (9). After an if we are to deal successfully with our represent the bioenergetics of the eco- early "bloom" at about 30 days, both pollution problems. Ecosystematic system. In the early stages of ecological rates decline, and they become approxi- study of entire landscape catchment succession, or in "young nature," so to mately equal at 60 to 80 days. the B/P units is a major goal of the American speak, the rate of ratio, in terms of grams of sup- plan for the proposed International or total (gross) (P) ex- ported per gram of daily carbon produc- the obvious ceeds Biological Program. Despite the rate of community respiration tion, increases from less than 20 to more of such a it is (R), logic proposal, proving so that the PIR ratio is greater than 100 as the steady state is reached. surprisingly difficult to get tradition- than 1. In the special case of organic Not only are autotrophic and hetero- bound scientists and granting agencies pollution, the PIR ratio is typically less trophic metabolism balanced in the cli- to look beyond their specialties toward than 1. In both cases, however, the max, but a large organic structure is the support of functional studies of theory is that PIR approaches 1 as suc- supported by small daily production and large units of the landscape. cession occurs. In other words, energy respiratory rates. fixed tends to be balanced by the energy While direct projection from the small cost of maintenance (that is, total com- laboratory microecosystem to open na- Food Chains and Food Webs munity respiration) in the mature or ture may not be entirely valid, there is "climax" ecosystem. The PIPR ratio, evidence that the same basic trends that As the therefore, should be ecosystem develops, subtle an excellent func- are seen in the laboratory are character- changes in the network pattern of food tional index of the relative maturity of istic of succession on land and in large chains be the system. may expected. The manner in bodies of water. Seasonal successions which are linked So long as P exceeds R, organisms together organic mat- also often follow the same pattern, an through food tends to be relatively simn IS APRIL 1969 263 a species-number ratio or a species- area ratio, tends to increase during the early stages of community development. A second component of species diver- sity is what has been called equitability, or evenness (18), in the apportionment of individuals among the species. For example, two systems each containing 10 species and 100 individuals have the same diversity in terms of species-num- ber ratio but could have widely differ- ent equitabilities depending on the ap- portionment of the 100 individuals among the 10 species-for example, 91-1-1-1-1-1-1-1-1-1 at one extreme or 10 individuals per species at the other. The Shannon formula,

-2 Nni log2 -ni

where ni is the number of individuals in each species and N is the total num- ber of individuals, is widely used as a di- 0 20 40 60 80 100 versity index because it combines the Days variety and equitability components in Fig. 1. Comparison of the energetics of succession in a forest and a laboratory inicro- one approximation. But, like all such cosm. PG, gross produiction; PN, net production; R, total community respiration; B, total lumping parameters, Shannon's formula biomass. may obscure the behavior of these two on March 3, 2011 rather different aspects of diversity. For example, in our most recent field ex- ple and linear in the very early stages tem of these protective mechanisms and periments, an acute stress from insecti- of succession, as a consequence of low allow irruptive, cancerous growths of cide reduced the number of species of diversity. Furthermore, heterotrophic certain species to occur, as man too insects relative to the number of individ- utilization of net production occurs pre- often finds to his sorrow. An example uals but increased the evenness in the dominantly by way of grazing food of a stress-induced pest irruption oc- relative abundances of the surviving curred at Brookhaven National Labora- species (19). Thus, in this case the chains-that is, plant--carni- www.sciencemag.org vore sequences. In contrast, food chains tory, where oaks became vulnerable to "variety" and "evenness" components become complex webs in mature stages, aphids when translocation of sugars would tend to cancel each other in with the bull of biological energy flow and amino acids was impaired by con- Shannon's formula. following pathways (Table 1, tinuing gamma irradiation (15). While an increase in the variety of item 5). In a mature forest, for ex- Radionuclide tracers are providing a species together with reduced domi- ample, less than 10 percent of annual means of charting food chains in the nance by any one species or small group net production is consumed (that is, intact outdoor ecosystem to a degree of species (that is, increased evenness) grazed) in the living state (12); most is that will permit analysis within the con- can be accepted as a general probability Downloaded from utilized as dead matter (detritus) cepts of network or matrix algebra. For during succession (20), there are other through delayed and complex pathways example, we have recently been able to community changes that may work involving as yet little understood animal- map, by use of a radiophosphorus against these trends. An increase in the interactions. The time tracer, the open, relatively linear food size of organisms, an increase in the involved in an uninterrupted succession linkage between plants and insects in an length and complexity of life histories, allows for increasingly intimate associa- early old-field successional stage (16). and an increase in interspecific compe- tions and reciprocal adaptations between tition that may result in competitive plants and animals, which lead to the exclusion of species (Table 1, items 12- development of many mechanisms that Diversity and Succession 14) are trends that may reduce the reduce grazing-such as the develop- number of species that can live in a ment of indigestible supporting tissues Perhaps the most controversial of the given area. In the bloom stage of suc- (cellulose, lignin, and so on), feedback successional trends pertain to the com- cession organisms tend to be small and control between plants and plex and much discussed subject of to have simple life histories and rapid (13), and increasing predatory pressure diversity (17). It is important to distin- rates of reproduction. Changes in size on herbivores (14). Such mechanisms guish between different kinds of diversi- appear to be a consequence of, or an enable the biological community to ty indices, since they may not follow adaptation to, a shift in nutrients from maintain the large and complex organic parallel trends in the same gradient or inorganic to organic (Table 1, item 7). structure that mitigates perturbations of developmental series. Four components In a mineral nutrient-rich environment, the physical environment. Severe stress of diversity are listed in Table 1, items small size is of selective advantage, es- or rapid changes brought about by out- 8 through 11. pecially to , because of the side forces can, of course, rob the sys- The variety of species, expressed as greater surface-to-volume ratio. As the 264 SCIENCE, VOL. 164 ecosystemrr develops, however, inorganic lize the growth and composition of the Nutrient Cycling nutrients tend to become more and ecosystem. Such metabolites may, in more tied up in the biomass (that is, to fact, be extremely important in prevent- An important trend in successional become intrabiotic), so that the selec- ing populations from overshooting the development is the closing or "tighten- tive advantage shifts to larger orga- equilibrial density, thus in reducing os- ing" of the biogeochemical cycling of nisms (either larger individuals of the cillations as the system develops sta- major nutrients, such as , phos- same species or larger species, or both) bility. phorus, and (Table 1, items 15- which have greater storage capacities The cause-and-effect relationship be- 17). Mature systems, as compared to de- and more complex life histories, thus tween diversity and stability is not clear veloping ones, have a greater capacity are adapted to exploiting seasonal or and needs to be investigated from many to entrap and hold nutrients for cycling periodic releases of nutrients or other angles. If it can be shown that biotic within the system. For example, Bor- resources. The question of whether the diversity does indeed enhance physical mann and Likens (24) have estimated seemingly direct relationship between stability in the ecosystem, or is the re- that only 8 kilograms per hectare out organism size and stability is the result sult of it, then we would have an im- of a total pool of exchangeable calcium of positive feedback or is merely for- portant guide for conservation practice. of 365 kilograms per hectare is lost per tuitous remains unanswered (21). Preservation of hedgerows, woodlots, year in stream outflow from a North Thus, whether or not noneconomic species, noneutrophicated Temperate watershed covered with a continues to increase during succession waters, and other biotic variety in man's mature forest. Of this, about 3 kilo- will depend on whether the increase in landscape could then be justified on grams per hectare is replaced by rain- potential niches resulting from increased scientific as well as esthestic grounds, fall, leaving only 5 kilograms to be biomass, stratification (Table 1, item 9), even though such preservation often obtained from weathering of the under- and other consequences of biological or- must result in some reduction in the lying rocks in order for the system to ganization exceeds the countereffects of production of food or other immediate maintain mineral balance. Reducing the increasing size and . No one needs. In other words, is va- volume of the vegetation, or otherwise has yet been able to catalogue all the riety only the spice of life, or is it a setting the succession back to a younger species in any sizable area, much less necessity for the long life of the total state, results in increased water yield follow total species diversity in a suc- ecosystem comprising man and nature? by way of stream outflow (25), but this cessional series. Data are so far available on March 3, 2011 only for segments of the community (trees, birds, and so on). Margalef (4) Table 1. A tabultr model of ecological succession: trends to be expected in the development postulates that diversity will tend to of ecosystems. peak during the early or middle stages Ecosystem attributes Developmental of succession and then decline in the stages Mature stages climax. In a study of bird populations along a Community energetics successional gradient we found 1. Gross production/community Greater or less Approaches I a bimodal pattern (22); the number of respiration ratio) than I (PIR www.sciencemag.org species increased during the early stages 2. Gross production/standing crop High Low biomass (P/B ratio) of old-field succession, declined during 3. Biomass supported/unit energy Low High the early forest stages, and then in- flow (BIE ratio) creased again in the mature forest. 4. Net community production (yield) High Low 5. Food chains Linear, predom- Weblike, predom- Species variety, equitability, and strat- inantly grazing inantly detritus ification are only three aspects of di- Community structure versity which change during succession. 6. Total organic matter Small Large Perhaps an even more important trend 7. Inorganic nutrients Extrabiotic Intrabiotic Downloaded from 8. Species diversity-variety component Low High is an increase in the diversity of organic 9. Species diversity-equitability Low High compounds, not only of those within component the 10. Biochemical diversity Low High biomass but also of those excreted 11. Stratification and spatial Poorly organized Well-organized and secreted into the media (air, , heterogeneity (pattern diversity) water) as by-products of the increas- Life history ing community metabolism. An increase 12. Niche specialization Broad Narrow 13. Size of organism Small Large in such "biochemical diversity" (Table 14. Life cycles Short, simple Long, 1, item complex 10) is illustrated by the increase Nutrient cycling in the variety of plant pigments along a 15. Mineral cycles Open Closed successional gradient in aquatic situa- 16. Nutrient exchange rate, between Rapid Slow organisms and environment tions, as described by Margalef (4, 23). 17. Role of detritus in nutrient Unimportant Important Biochemical diversity within popula- regeneration tions, or within systems as a whole, has Selection pressure not yet been systematically studied to 18. Growth form For rapid growth For feedback control ("r-selection") ("K-selection") the degree the subject of species diver- 19. Production Quantity Quality sity has been. Consequently, few gen- Overall homeostasis eralizations can be made, except that 20. Internal Undeveloped Developed it seems safe to say that, as succes- 21. Nutrient conservation Poor Good 22. Stability (resistance to external Poor Good sion progresses, organic extrametabo- perturbations) lites probably serve increasingly impor- 23. Entropy High Low tant functions as regulators which stabi- 24. Information Low High 18 APRIL 1969 265 greater outflow is accompanied by increase and K is the upper asymptote one is impressed with the importance of greater losses of nutrients, which may or equilibrium , we may , , , com- also produce downstream eutrophica- say that "r selection" predominates in mensalism, and other forms of symbi- tion. Unless there is a compensating in- early colonization, with "K selection" osis. Partnership between unrelated spe- crease in the rate of weathering, the prevailing as more and more species and cies is often noteworthy (for example, exchangeable pool of nutrients suffers individuals attempt to colonize (Table 1, that between coral coelenterates and gradual depletion (not to mention pos- item 18). The same sort of thing is even algae, or between mycorrhizae and sible effects on soil structure resulting seen within the species in certain "cy- trees). In many cases, at least, biotic from erosion). High fertility in "young clic" northern insects in which "active" control of grazing, population density, systems" which have open nutrient genetic strains found at low densities and nutrient cycling provide the chief cycles cannot be maintained without are replaced at high densities by "slug- positive-feedback mechanisms that con- compensating inputs of new nutrients; gish" strains that are adapted to tribute to stability in the mature system examples of such practice are the con- crowding (27). by preventing overshoots and destruc- tinuous-flow culture of algae, or inten- Genetic changes involving the whole tive oscillations. The intriguing question sive agriculture where large amounts of biota may be presumed to accompany is, Do mature ecosystems age, as orga- are imported into the system the successional gradient, since, as de- nisms do? In other words, after a long each year. scribed above, quantity production char- period of relative stability or "adult- Because rates of leaching increase in acterizes the young ecosystem while hood," do ecosystems again develop un- a latitudinal gradient from the poles to quality production and feedback control balanced metabolism and become more the equator, the role of the biotic com- are the trademarks of the mature system vulnerable to diseases and other per- munity in nutrient retention is especially (Table 1, item 19). Selection at the eco- turbations? important in the high-rainfall areas of system level may be primarily interspe- the subtropical and tropical latitudes, cific, since species replacement is a including not only land areas but also characteristic of successional series or Relevance of Ecosystem Development estuaries. Theoretically, as one goes seres. However, in most well-studied Theory to Human Ecology equatorward, a larger percentage of the seres there seem to be a few early suc- available nutrient pool is tied up in the cessional species that are able to persist Figure 1 depicts a basic conflict be- on March 3, 2011 bionfass and a correspondingly lower through to late stages. Whether genetic tween the strategies of man and of na- percentage is in the soil or sediment. changes contribute to adaptation in such ture. The, "bloom-type" relationships, as This theory, however, needs testing, species has not been determined, so far exhibited by the 30-day microcosm or since data to show such a geographical as I know, but studies on population the 30-year forest, illustrate man's pres- trend are incomplete. It is perhaps sig- genetics of Drosophila suggest that ent idea of how nature should be di- nificant that conventional North Tem- changes in genetic composition could rected. For example, the goal of agri- perate row-type agriculture, which be important in population regulation culture or intensive forestry, as now represents a very youthful type of eco- (28). Certainly, the human population, generally practiced, is to achieve high system, is successful in the humid if it survives beyond its present rapid rates of production of readily harvest- www.sciencemag.org tropics only if carried out in a system growth stage, is destined to be more able products with little standing crop of "shifting agriculture" in which the and more affected by such selection left to accumulate on the landscape-in crops alternate with periods of natural pressures as adaptation to crowding be- other words, a high P/B efficiency. Na- vegetative redevelopment. Tree culture comes essential. ture's strategy, on the other hand, as and the semiaquatic culture of rice pro- seen in the outcome of the successional vide much better nutrient retention and process, is directed toward the reverse consequently have a longer life expect- Overall Homeostasis efficiency-a high B/P ratio, as is de- Downloaded from ancy on a given site in these warmer picted by the relationship at the right in latitudes. This brief review of ecosystem de- Fig. 1. Man has generally been preoc- velopment emphasizes the complex na- cupied with obtaining as much "pro- ture of processes that interact. While duction" from the landscape as possible, Selection Pressure: one may well question whether all the by developing and maintaining early Quantity versus Quality trends described are characteristic of successional types of ecosystems, usu- all types of ecosystems, there can be ally monocultures. But, of course, man MacArthur and Wilson (26) have re- little doubt that the net. result of com- does not live by food and fiber alone; viewed stages of colonization of islands munity actions is symbiosis, nutrient he also needs a balanced C02-02 at- which provide direct parallels with conservation, stability, a decrease in mosphere, the climatic buffer provided stages in ecological succession on conti- entropy, and an increase in information by oceans and masses of vegetation, nents. Species with high rates of repro- (Table 1, items 20-24). The overall and clean (that is, unproductive) water duction and growth, they find, are more strategy is, as I stated at the beginning for cultural and industrial uses. Many likely to survive in the early uncrowded of this article, directed toward achieving essential life-cycle resources, not to stages of island colonization. In contrast, as large and diverse an organic struc- mention recreational and esthetic needs, selection pressure favors species with ture as is possible within the limits set are best provided man by the less "pro- lower growth potential but better capa- by the available energy input and the ductive" landscapes. In other words, the bilities for competitive survival under prevailing physical conditions of exist- landscape is not just a supply depot but the equilibrium density of late stages. ence (soil, water, climate, and so on). is also the oikos-the home-in which Using the terminology of growth equa- As studies of biotic communities be- we must live. Until recently mankind tions, where r is the intrinsic rate of come more functional and sophisticated, has more or less taken for granted the 266 SCIENCE, VOL. 164 gas-exchange, water-purification, nutri- Table 2. Contrasting characteristics of young moderate and moderate yield on and quality ent-cycling, and other protective func- mature-type ecosystems. all the landscape, or we can deliberately tions of self-maintaining ecosystems, Young Mature plan to compartmentalize the landscape chiefly because neither his numbers nor so as to simultaneously maintain highly his environmental have Production Protection and manipulations Growth Stability productive predominantly protec- been great enough to affect regional and Quantity Quality tive types as separate units subject to global balances. Now, of course, it is different management strategies (strate- painfully evident that such balances are gies ranging, for example, from inten- being affected, often detrimentally. The sive cropping on the one hand to wilder- "one problem, one solution approach" is no effective mechanism whereby ness management on the other). If is no longer adequate and must be re- negative feedback signals can be re- ecosystem development theory is valid placed by some form of ecosystem anal- ceived and acted on before there has and applicable to planning, then the ysis that considers man as a part of, not been a serious overshoot. Thus, today so-called multiple-use strategy, about apart from, the environment. there are governmental agencies, spur- which we hear so much, will work only The most pleasant and certainly the red on by popular and political en- through one or both of these ap- safest landscape to live in is one con- thusiasm for dams, that are putting on proaches, because, in most cases, the taining a variety of crops, forests, lakes, the drawing boards plans for damming projected multiple uses conflict with streams, roadsides, marshes, seashores, every river and stream in North one another. It is appropriate, then, to and "waste places"-in other words, a America! examine some examples of the com- mixture of communities of different Society needs, and must find as promise and the compartmental strat- ecological ages. As individuals we more quickly as possible, a way to deal with egies. or less instinctively surround our houses the landscape as a whole, so that ma- with protective, nonedible cover (trees, nipulative skills (that is, technology) shrubs, grass) at the same time that we will not run too far ahead of our under- Pulse Stability strive to coax extra bushels from our standing of the impact of change. Re- cornfield. We all consider the cornfield a cently a national ecological center out- A more or less reguLlar btut acute "good thing," of course, but most of us side of government and a coalition of physical perturbation imposed from on March 3, 2011 would not want to live there, and it governmental agencies have been pro- without can maintain an ecosystem at would certainly be suicidal to cover the posed as two possible steps in the some intermediate point in the develop- whole land area of the biosphere with establishment of a political control mental sequence, resulting in, so to cornfields, since the boom and bust os- mechanism for dealing with major speak, a compromise between youth and cillation in such a situation would be environmental questions. The soil con- maturity. What I would term "fluctuat- severe. servation movement in America is an ing water level ecosystems" are good The basic problem facing organized excellent example of a program dedi- examples. Estuaries, and intertidal zones society today boils down to determining cated to the consideration of the whole in general, are maintained in an early, www.sciencemag.org in some objective manner when we are farm or the whole watershed as an eco- relatively fertile stage by the tides, which getting "too much of a good thing." logical unit. Soil conservation is well provide the energy for rapid nutrient This is a completely new challenge to understood and supported by the public. cycling. Likewise, freshwater marshes, mankind because, up until now, he has However, soil conservation organiza- such as the Florida Everglades, are had to be concerned largely with too tions have remained too exclusively held at an early successional stage by little rather than too much. Thus, con- farm-oriented, and have not yet risen to the seasonal fluctuations in water levels. crete is a "good thing," but not if half the challenge of the urban-rural land- The dry-season drawdown speeds up the world is covered with it. Insecticides scape, where lie today's most serious aerobic of accumulated Downloaded from are "good things," but not when used, problems. We do, then, have potential organic matter, releasing nutrients that, as they now are, in an indiscriminate mechanisms in American society that on reflooding, support a wet-season and wholesale manner. Likewise, water could speak for the ecosystem as a bloom in productivity. The life histories impoundments have proved to be very whole, but none of them are really of many organisms are intimately cou- useful man-made additions to the land- operational (29). pled to this periodicity. The wood stork, scape, but obviously we don't want the The general relevance of ecosystem for example, breeds when the water whole country inundated! Vast man- development theory to landscape plan- levels are falling and the small fish on made lakes solve some problems, at least ning can, perhaps, be emphasized by the which it feeds become concentrated and temporarily, but yield comparative little "mini-model" of Table 2, which con- easy to catch in the drying pools. If the food or fiber, and, because of high trasts the characteristics of young and water level remains high during the evaporative losses, they may not even be mature-type ecosystems in more general usual dry season or fails to rise in the the best device for storing water; it terms than those provided by Table 1. wet season, the stork will not nest (30). might better be stored in the watershed, It is mathematically impossible to ob- Stabilizing water levels in the Everglades or underground in aquafers. Also, the tain a maximum for more than one by means of dikes, locks, and impound- cost of building large dams is a drain on thing at a time, so one cannot have ments, as is now advocated by some, already overtaxed revenues. Although both extremes at the same time and would, in my opinion, destroy rather as individuals we readily recognize that place. Since all six characteristics listed than preserve the Everglades as we now we can have too many dams or other in Table 2 are desirable in the aggre- know them just as surely as complete large-scale environmental changes, gov- gate, two possible solutions to the di- drainage would. Without periodic draw- ernments are so fragmented and lacking lemma immediately suggest themselves. downs and fires, the shallow basins in systems-analysis capabilities that there We can compromise so as to provide would fill up with organic matter and 18 APRIL 1969 267 succession would proceed from the pres- Prospects for a Detritus Agriculture investment in the culture of oysters and ent pond-and-prairie condition toward other seafoods would also provide the a scrub or swamp forest. As indicated above, heterotrophic best possible deterrent against pollution, It is strange that man does not readily utilization of primary production in ma- since the first threat of damage to the recognize the importance of recurrent ture ecosystems involves largely a de- pollution-sensitive oyster industry would changes in water level in a natural situ- layed consumption of detritus. There is be immediately translated into political ation such as the Everglades when simi- no reason why man cannot make greater action! lar pulses are the basis for some of his use of detritus and thus obtain food or most enduring food culture systems other products from the more protective (31). Alternate filling and draining of type of ecosystem. Again, this would The Compartment Model ponds has been a standard procedure in represent a compromise, since the short- fish culture for centuries in Europe and term yield could not be as great as the Successful though they often are, the Orient. The flooding, draining, and yield obtained by direct exploitation of compromise systems are not suitable soil-aeration procedure in rice culture is the grazing . A detritus agri- nor desirable for the whole landscape. another example. The rice paddy is thus culture, however, would have some com- More emphasis needs to be placed on the cultivated analogue of the natural pensating advantages. Present agricul- compartmentalization, so that growth- marsh or the intertidal ecosystem. tural strategy is based on selection for type, steady-state, and intermediate-type Fire is another physical factor whose rapid growth and edibility in food ecosystems can be linked with urban periodicity has been of vital importance plants, which, of course, make them and industrial areas for mutual benefit. to man and nature over the centuries. vulnerable to attack by insects and dis- Knowing the transfer coefficients that Whole biotas, such as those of the ease. Consequently, the more we select define the flow of energy and the move- African grasslands and the California for succulence and growth, the more ment of materials and organisms (in- chaparral, have become adapted to peri- effort we must invest in the chemical cluding man) between compartments, it odic fires producing what ecologists control of pests; this effort, in turn, in- should be possible to determine, through often call "fire climaxes" (32). Man uses creases the likelihood of our poisoning analog-computer manipulation, rational fire deliberately to maintain such cli- useful organisms, not to mention our- limits for the size and capacity of each on March 3, 2011 maxes or to set back succession to some selves. Why not also practice the reverse compartment. We might start, for ex- desired point. In the southeastern strategy-that is, select plants which are ample, with a simplified model, shown coastal plain, for example, light fires of essentially unpalatable, or which pro- in Fig. 2, consisting of four compart- moderate frequency can maintain a pine duce their own systemic insecticides ments of equal area, partitioned accord- forest against the encroachment of older while they are growing, and then con- ing to the basic biotic-function criterion successional stages which, at the present vert the net production into edible prod- -that is, according to whether the time at least, are considered economi- ucts by microbial and chemical enrich- area is (i) productive, (ii) protective, (iii) or cally less desirable. The fire-controlled ment in food factories? We could then a compromise between (i) and (ii) (iv), www.sciencemag.org forest yields less wood than a tree farm devote our biochemical genius to the urban-industrial. By continually refining does (that is, young trees, all of about enrichment process instead of fouling up the transfer coefficients on the basis of the same age, planted in rows and har- our living space with chemical poisons! real world situations, and by increasing vested on a short rotation schedule), but The production of silage by fermenta- and decreasing the size and capacity of it provides a greater protective cover for tion of low-grade fodder is an example each compartment through compu- the landscape, wood of higher quality, of such a procedure already in wide- ter simulation, it would be possible and a home for game birds (quail, wild spread use. The cultivation of detritus- to determine objectively the limits turkey, and so on) which could not sur- eating fishes in the Orient is another that must eventually be imposed on Downloaded from vive in a tree farm. The fire climax, example. each compartment in order to maintain then, is an example of a compromise By tapping the detritus food chain regional and global balances in the ex- between production simplicity and pro- man can also obtain an appreciable har- change of vital energy and of materials. tection diversity. vest from many natural systems without A systems-analysis procedure provides It should be emphasized that pulse greatly modifying them or destroying at least one approach to the solution of stability works only if there is a com- their protective and esthetic value. Oys- the basic dilemma posed by the question plete community (including not only ter culture in estuaries is a good exam- "How do we determine when we are plants but animals and ) ple. In Japan, raft and long-line culture getting too much of a good thing?" Also adapted to the particular intensity and of oysters has proved to be a very prac- it provides a means of evaluating the frequency of the perturbation. Adapta- tical way to harvest the natural micro- energy drains imposed on ecosystems by tion-operation of the selection process bial products of estuaries and shallow pollution, radiation, harvest, and other -requires times measurable on the evo- bays. Furukawa (33) reports that the stresses (34). lutionary scale. Most physical stresses yield of cultured oysters in the Hiro- Implementing any kind of compart- introduced by man are too sudden, too shima Prefecture has increased tenfold mentalization plan, of course, would re- violent, or too arrhythmic for adaptation since 1950, and that the yield of oysters quire procedures for zoning the land- to occur at the ecosystem level, so severe (some 240,000 tons of meat) from this scape and restricting the use of some oscillation rather than stability results. one district alone in 1965 was ten times land and water areas. While the princi- In many cases, at least, modification of the yield of natural oysters from the ple of zoning in cities is universally naturally adapted ecosystems for cul- entire country. Such oyster culture is accepted, the procedures now followed tural purposes would seem preferable to feasible along the entire Atlantic and do not work very well because zoning complete redesign. Gulf coasts of the United States. A large restrictions are too easily overturned by 268 SCIENCE, VOL. 164 short-term economic and population Earl F. Murphy, in a book entitled pressures. Zoning the landscape would Protective Governing Nature (38), emphasizes that require a whole new order of thinking. // (mature systems) \ the regulatory approach alone is not Greater use of legal measures providing . 2 C rtm ent enough to protect life-cycle resources, for tax relief, restrictions on use, such as air and water, that cannot be scenic easements, and public owner- Productive - 'Compromise allowed to deteriorate. He discusses per- ship will be required if appreciable land (growth systems) - (multiple-use systems) mit systems, effluent charges, receptor and water areas are to be held in the environment renviro nment levies, assessment, and cost-internaliz- "protective" categories. Several states ing procedures as economic incentives (for example, New Jersey and Califor- for achieving Hardin's "mutually agreed nia), where pollution and population upon coercion." pressure are beginning to hurt, have It goes without saying that the tabu- made a start in this direction by enact- lar model for ecosystem development ing "open space" legislation designed to which I have presented here has many get as much unoccupied land as possi- parallels in the development of human ble into a "protective" status so that partitioned according to ecosystem develop- society itself. In the pioneer society, as future uses can be planned on a rational ment and life-cycle criteria. in the pioneer ecosystem, high birth and scientific basis. The United States rates, rapid growth, high economic as a whole is fortunate in that large profits, and exploitation of accessible areas of the country are in national many species of animals avoid crowd- and unused resources are advantageous, forests, parks, wildlife refuges, and so ing and social stress (36). but, as the saturation level is approached, on. The fact that such areas, as well as Since the legal and economic prob- these drives must be shifted to consid- the bordering oceans, are not quickly lems pertaining to zoning and compart- erations of symbiosis (that is, "civil exploitable gives us time for the accel- mentalization are likely to be thorny, rights," "law and order," "education," erated ecological study and program- I urge law schools to establish depart- and "culture"), birth control, and the ming needed to determine what propor- ments, or institutes, of "landscape law" recycling of resources. A balance be- on March 3, 2011 tions of different types of landscape and to start training "landscape lawyers" tween youth and maturity in the socio- provide a safe balance between man and who will be capable not only of clari- environmental system is, therefore, the nature. The open oceans, for example, fying existing procedures but also of really basic goal that must be achieved should forever be allowed to remain drawing up new enabling legislation for if man as a species is to successfully protective rather than productive terri- consideration by state and national gov- pass through the present rapid-growth tory, if Alfred Redfield's (35) assump- erning bodies. At present, society is stage, to which he is clearly well tions are correct. Redfield views the concerned-and rightly so-with hu- adapted, to the ultimate equilibrium- oceans, the major part of the hydro- man rights, but environmental rights density stage, of which he as yet shows sphere, as the biosphere's governor, are equally vital. The "one man one little understanding and to which he www.sciencemag.org which slows down and controls the rate vote" idea is important, but so also is a now shows little tendency to adapt. of decomposition and nutrient regenera- "one man one hectare" proposition. tion, thereby creating and maintaining Education, as always, must play a References and Notes the highly aerobic terrestrial environ- role in increasing man's awareness of 1. E. P. Odum, Ecology (Holt, Rinehart & Winston, New York, 1963), chap. 6. ment to which the higher forms of life, his dependence on the natural environ- 2. H. T. Odum and R. C. Pinkerton, Amer. such as man, are adapted. Eutrophica- ment. we need Scientist 43, 331 (1955). Perhaps to start teaching 3. A. J. Lotka, Elements of Physical Biology tion of the ocean in a last-ditch effort to the principles of ecosystem in the third (Williams and Wilkins, Baltimore, 1925). Downloaded from 4. R. Margalef, Advan. Frontiers Plant Sci. 2, feed the populations of the land could grade. A grammar school primer on 137 (1963); Amer. Naturalist 97, 357 (1963). well have an adverse effect on the oxy- man and his environment could logically 5. R. J. Beyers, Ecol. Monographs 33, 281 (1963). gen reservoir in the atmosphere. consist of four chapters, one for each 6. The systems so far used to test ecological Until we can determine more pre- of the four essential kinds of environ- principles have been derived from sewage and farm ponds and are cultured in half- cisely how far we may safely go in ex- ment, shown diagrammatically in Fig. 2. strength No. 36 Taub and Dollar medium panding intensive agriculture and urban Of the many books and articles that [Limnol. Oceanog. 9, 61 (1964)]. They are closed to organic imput or output but are sprawl at the expense of the protective are being written these days about open to the atmosphere through the cotton landscape, it will be good insurance to plug in the neck of the flask. Typically, man's environmental crisis, I would like liter-sized microecosystems contain two or hold inviolate as much of the latter as to cite two that go beyond "crying out three species of nonflagellated algae and one to three species each of flagellated protozoans, possible. Thus, the preservation of na- in alarm" to suggestions for bringing ciliated protozoans, , nematodes, and tural areas is not a peripheral luxury for about a reorientation of the goals of ostracods; a system derived from a sewage pond contained at least three species of fungi society but a capital investment from society. Garrett Hardin, in a recent arti- and 13 bacterial isolates [R. Gordon, thesis, which we expect to draw interest. Also, cle in Science (37), points out that, since University of Georgia (1967)]. These cul- tures are thus a kind of minimum ecosystem it may well be that restrictions in the the optimum population density is less containing those small species originally use of land and water are our only than the maximum, there found in the ancestral pond that are able to is no strictly function together as a self-contained unit practical means of avoiding overpopula- technical solution to the problem of under the restricted conditions of the labora- tion or too great an exploitation of re- tory flask and the controlled environment of pollution caused by ; a a growth chamber [temperature, 650 to 75°F sources, or both. Interestingly enough, solution, he suggests, can only be (18' to 24'C); photoperiod, 12 hours; illumi- restriction of land use is the analogue of nation, 100 to 1000 footcandles]. achieved through moral and legal 7. G. D. Cooke, BioScience 17, 717 (1967). a natural behavioral control mechanism means of "mutual coercion, mutually 8. T. Kira and T. Shidei, Japan. J. Ecol. 17, 70 (1967). known as "territoriality" by which agreed upon by the majority of people." 9. The metabolism of the microcosms was 18 APRIL 1969 269 monitored by measuring diurnal pH changes, grain culture, both the species-to-numbers and level canal joining the Atlantic and Pacific and the biomass (in terms of total organic the equitability indices increased for all oceans [Science 161, 857 (1968)1, calls for matter and total carbon) was determined trophic levels but esoecially for predators a "control commission for environmental by periodic harvesting of replicate systems. and parasites. Only 44 percent of the species manipulation" with "broad powers of ap- 10. P. J. H. Mackereth, Proc. Roy. Soc. London in the natural ecosystem were phytophagous, proving, disapproving, or modifying al ma- Ser. B 161, 295 (1965); U. M. Cowgill and as compared to 77 percent in the grain field. jor alterations of the marine or terrestrial 0. E. Hutchinson, Proc. Intern. Limnol. Ass. 21. J. T. Bonner, Size and Cycle (Princeton environments. .. ." 15, 644 (1964); A. D. Harrison, Trans. Roy. Univ. Press, Princeton, N.J., 1963); P. Frank, 30. See M. P. Kahl, Ecol. Monographs 34, 97 Soc. S. Africa 36, 213 (1962). Ecology 49, 355 (1968). (1964). 11. R. Margalef, Proc. Intern. Limnol. Ass. 15, 22. D. W. Johnston and E. P. Odum, Ecology 31. The late Aldo Leopold remarked long ago 169 (1964). 37, 50 (1956). [Symposium on Hydrobiology (Univ. of 12. J. R. Bray, Oikos 12, 70 (1961). 23. R. Margalef, Oceanog. Marine Biol. Annu. Wisconsin Press, Madison, 1941), p. 17] that 13. D. Pimentel, Amer. Naturalist 95, 65 (1961). Rev. 5, 257 (1967). man does not perceive organic behavior in 14. R. T. Paine, ibid. 100, 65 (1966). 24. F. H. Bormann and G. E. Likens, Science systems unless he has built them himself. 15. G. M. Woodwell, Brookhaven Nat. Lab. 155, 424 (1967). Let us hope it will not be necessary to re- Pub. 924(T-381) (1965), pp. 1-15. 25. Increased water yield following reduction of build the entire biosphere before we recog- 16. R. G. Wiezert, E. P. Odum, J. H. Schnell, vegetative cover has been frequently demon- nize the worth of natural systems! Ecology 48, 75 (1967). strated in experimental watersheds throughout 32. See C. F. Cooper, Sci. Amer. 204, 150 (April 17. For selected general discussions of patterns the world [see A. R. Hibbert, in International 1961). of species diversity, see E. H. Simpson, Na- Symposium on Forest Hydrology (Pergamon 33. See "Proceedings Oyster Culture Workshop, ture 163, 688 (1949): C. B. Williams, J. Press, New York, 1967), pp. 527-543]. Data Marine Fisheries Division, Georgia Game and Animal Ecol. 22, 14 (1953); G. E. Hutchin- on the long-term hydrologic budget (rainfall Fish Commission, Brunswick" (1968), pp. 49- son, Amer. Naturalist 93, 145 (1959); R. input relative to stream outflow) are avail- 61. Margalef, Gen. Systems 3, 36 (1958); R. able at many of these sites, but mineral 34. See H. T. Odum, in Symposium ont Primary MacArthur and J. MacArthur, Ecology 42, budgets have yet to be systematically studied. Productivity and Mineral Cycling in Natural 594 (1961); N. G. Hairston, ibid. 40, 404 Again, this is a prime objective in the "eco- Ecosystems, H. E. Young, Ed. (Univ. of (1959); B. C. Patten, J. Marine Res. (Sears system analysis" phase of the International Maine Press, Orono, 1967), p. 81; , in Found. Marine Res.) 20, 57 (1960); E. G. Biological Program. Pollution and Marine Ecology (Wiley, New Leigh, Proc. Nat. Acad. Sci. U.S. 55, 777 26. R. H. MacArthur and E. 0. Wilson, Theory York, 1967), p. 99; K. E. F. Watt, Ecology (1965); E. R. Pianka, Amer. Naturalist 100, of Island (Princeton Univ. Press, and Resource Management (McGraw-Hill, 33 (1966); E. C. Pielou, J. Theoret. Biol. Princeton, N.J., 1967). New York, 1968). 10, 370 (1966). 27. Examples are the tent caterpillar [see W. G. 35. A. C. Redfield, Amer. Scientist 46, 205 (1958). 18. M. Lloyd and R. J. Ghelardi, J. Animal Ecol. Wellington, Can. J. Zool. 35, 293 (1957)] 36. R. Ardrey, The Territorial Imperative (Athe- 33, 217 (1964); E. C. Pielou, J. Theoret. and the larch budworm [see W. Baltens- neum, New York, 1967). Biol. 13, 131 (1966). weiler, Can. Entomologist 96. 792 (1964)]. 37. G. Hardin, Science 162, 1243 (1968). 19. G. W. Barrett, Ecology 49, 1019 (1969). 28. F. J. Ayala, Science 162, 1453 (1968). 38. E. F. Murphy, Governing Nature (Quad- 20. In our studies of natural succession following 29. Ira Rubinoff, in discussing the proposed sea- rangle Books, Chicago, 1967). on March 3, 2011

that higher intensities are signaled by higher frequencies of discharge of uni- form nerve impulses. In 1931, when C. H. Graham and I sought to apply to an optic nerve the Visual Receptors and technique developed by Adrian and Bronk for isolating a single fiber, we www.sciencemag.org Retinal Interaction made a fortunate choice of experi- mental animal (3). The xiphosuran Haldan Keffer Hartline arachnoid, Limulus polyphemus, com- monly called "Horseshoe crab," abounds on the eastern coast of North Amer- ica (4). These "living fossils" have

lateral compound eyes that are coarsely Downloaded from The neuron is the functional as well first to elucidate properties of single faceted and connected to the brain by as the structural unit of the nervous sensory receptors (1). These studies long optic nerves. The optic nerve in system. Neurophysiology received an laid the foundations for the unitary the adults can be frayed into thin impetus of far-reaching effect in the analysis of nervous function. bundles which are easy to split until 1920's, when Adrian and his colleagues My early interest in vision was just one active fiber remains. The rec- developed and exploited methods for spurred by another contribution from ords in Fig. 1 were obtained from such recording the activity of single neurons Adrian's laboratory: his study, with a preparation. and sensory receptors. Adrian and R. Matthews, of the massed discharge The sensory structures in the eye of Bronk were the first to analyze motor of nerve impulses in the eel's optic Limulus from which the optic nerve function by recording the activity of nerve (2). I aspired to the obvious ex- fibers arise are clusters of receptor single fibers dissected from a nerve tension of this study: application of cells, arranged radially around the den- trunk and Adrian and Zotterman the unitary analysis to the receptors and dritic process of a bipolar neuron (ec- neurons of the visual system. centric cell) (5). Each cluster lies be- Copyright © by the Nobel Foundation. The author is professor of biophysics at the Oscillograms of the action potentials hind its corneal facet and crystalline Rockefeller University in New York City. This in a single nerve fiber are now common- cone, which give it its own, small visual article is the lecture he delivered in Stockholm, Sweden, 12 December 1967, when, with Regnar place. The three shown in Fig. 1 are field (Fig. 2). Each such ommatidium, Arthur Granit and George Wald, he received nerve retinal though not as simple as I once thought, the Nobel Prize in medicine or physiology. It from an optic fiber whose is published here with the permission of the receptor was stimulated by light, the seems to act as a functional receptor Nobel Foundation and also is included in the unit. Restriction of the stimulating light complete volume of Les Prix Nobel en 1967, as relative values of which are given at the well as in the series Nobel Lectures (in Eng- left of each record. One of the earliest to one facet elicits discharge in one lish), published by the Elsevier Publishing Com- neuron pany, Amsterdam and New York. results of -unitary analysis was to show fiber-the axon of the bipolar 270 SCIENCE, VOL. 164