Discussion #4 Questions

Description

Read Connell_1978_Diversity_in_Tropical_Rain_Forests_and_Coral_Reefs.pdfFind the discussion questions here: Discussion #4 QuestionsAdd your answers to the Word document (Discussion #4 Questions) and upload them using the TurnItIn link below. Note that Turnitin compares responses to ensure students are turning in original work.Late assignments will incur a penalty of 10% per day and up to 50% if submitted 5 or more days late. DO NOT PLAGIARIZE OR USE AI GENERATED TEXTS

Unformatted Attachment Preview

Diversity in Tropical Rain Forests and Coral Reefs
Joseph H. Connell
Science, New Series, Vol. 199, No. 4335. (Mar. 24, 1978), pp. 1302-1310.
Stable URL:
http://links.jstor.org/sici?sici=0036-8075%2819780324%293%3A199%3A4335%3C1302%3ADITRFA%3E2.0.CO%3B2-2
Science is currently published by American Association for the Advancement of Science.
Your use of the JSTOR archive indicates your acceptance of JSTOR’s Terms and Conditions of Use, available at
http://www.jstor.org/about/terms.html. JSTOR’s Terms and Conditions of Use provides, in part, that unless you have obtained
prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in
the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/journals/aaas.html.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic
journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,
and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take
advantage of advances in technology. For more information regarding JSTOR, please contact [email protected].
http://www.jstor.org
Thu Jan 3 11:01:11 2008
niques is taking place, and new information is already yielding fresh insights into
chemical, physical, and biological systems.
References and Notes
I. B. M. Kincaid, J . Appl. Phys. 48, 2684 (1977); J .
P. Blewett and R. Chasman, ibid., D. 2692.
2 . D. A. G. Deacon, L. R. Elias, J. M.J . Madey,
G. J . Ramian, H. A Schwettman, T. I. Smith,
Phys. Rev. Lett. 38, 892 (1977).
3. D. Iwanenko and I. Pomeranchuk, Phys. Rev.
65. 343 11944).
4 . J . P. ~ l i w e t t i:bid. 69, 87 (1946).
5. J. S. Schwinger, ibid. 70, 798 (1946); ibid. 75,
1912 119491.
6 . B. Alpert and R. Lopez-Delgado, Nature (London) 263, 445 (1976).
7. 0. B. d’Azv. R. Looez-Deleado. A. Tramer.
Chem. phys: 9 , 327 (i975).
8 . L. F. Wagner and W. E. Spicer, Phys. Rela.
Lett. 28. 1381 119721.
9 . D. E. ~ a s t m a nand W. D. Grobman, ibid., p.
1379.
10. I. Lindau, in Synchrotron Radiation Research,
A. N. Mancini and I. F. Quercia, Eds. (International Colloquium on Applied Physics, Istituto Nazionale di Fisica Nucleare, Rome, in
press).
11. R. J. Smith, J. Anderson, G. J. Lapeyre, Phys.
Rev. Lett. 37, 1081 (1976).
12. E. Spiller, R. Feder, J. Topalian, D. Eastman,
W. Gudat, D. Sayre, Science 191, 1172 (1976);
R. Feder, E. Spiller, J. Topaiian, A. N. Broers.
W. Gudat, B. J . Panessa, Z. A. Zadunaisky, J.
Sedat, ibid. 197, 259 (1977); E. Spiller. D. E.
Eastman, R. Feder, W. D. Grobman. W. Gudat.
J . Topalian, J. Appl. Phys. 47, 5450 (1976); E.
Spiller and R. Feder, in X-ray Optics, H . J .
Quesser, Ed. (Springer-Verlag, Berlin, in press):
B. Fay, J. Trotel, Y. Petroff, R. Pinchaux, P.
Thiry, Appi. Phys. Lett. 29. 370 (1976). For a
general review of x-ray lithography with conventional sources see S. E. Bernacki and H. I.
Diversity in Tropical Rain
Forests and Coral Reefs
High diversity of trees and corals is maintained
only in a nonequilibrium state.
Joseph H. Connell
The great variety of species in local
areas of tropical rain forests and coral
reefs is legendary. Until recently, the
usual explanation began with the assumption that the species composition of
such assemblages is maintained near
equilibrium (1). The question thus became: “how is high diversity maintained
near equilibrium’!” One recent answer
communities is a consequence of past
and present interspecific competition, resulting in each species occupying the
habitat o r resource on which it is the
most effective competitor. Without perturbation this species composition persists: after perturbation it is restored to
the original state (3).
Iti recent years it has become clear
Summary. The commonly observed high diversity of trees in tropical rain forests
and corals on tropical reefs is a nonequilibrium state which, if not disturbed further, will
progress toward a low-diversity equilibrium community. This may not happen if gradual changes in climate favor different species. If equilibrium is reached, a lesser degree
of diversity may be sustained by niche diversification or by a compensatory mortality
that favors inferior competitors. However, tropical forests and reefs are subject to
severe disturbances often enough that equilibrium may never be attained.
for tropical bird communities is given as
follows: “The working hypothesis is that,
through diffuse competition, the component species of a community are selected, and coadjusted in their niches
and abundances, so as to fit with each
other and to resist invaders” (2). In this
view, the species composition of tropical
1302
that the frequency of natural disturbance
and the rate of environmental change are
often much faster than the rates of recovery from perturbations. In particular,
competitive elimination of the less efficient o r less well adapted species is not
the inexorable and predictable process
we once thought it was. Instead, other
0036-807517810324-1302$02.0010 Copyright 0 1978 AAAS
Smith, IEEE Trans. Electron De~aices,ED-22,
421 (1975).
13. P. Horowitz and J. A. Howell, Science 178, 608
(1972).
14. B. M. Kincaid, P. Eisenberger, K. 0 . Hodgson,
.
Sci. U.S.A. 72.
S . Doniach. Proc. ~ V a t lAcad.
2340 (1975):
15. K. D. Watenpaugh, L. C . Sieker, J . R. Herriott,
L. H. Jensen, Acra Crystalloyr. Sect. B 29, 943
11971)
,.,, – ,.
16. See, for example, R. G. Shulman, P. Eisenberger, W. E. Blumberg. N. A. Stombaugh,
Proc. hlafl. Acad. Sci. U . S . A . 72, 4003 (1975).
17. R. S . Goody, K. C. Holmes, H. G. Mannherz, J .
Barrington Leigh, G. Rosenbaum, Biophys. J .
15, 687 (1975).
18. T. Tuomi, K. Naukkarian, P. Rabe, Phys. Status SolidiA 25,93 (1974); M. Hart, J. Appl. Crystalloyr. 8, 436 (1975); J . Bordas, S. M . Glazer,
H. Hauser, Philos. Mag. 32,471 (1975); B. Bura s and J . Staun Olsen. N u c . Instrurn. Methods
135, 193 (1976).
19. Research was performed at Brookhaven National Laboratory under contract with ERDA.
forces, often abrupt and unpredictable.
set back, deflect. o r slow the process of
return to equilibrium (4). If such forces
are the norm. we may question the usefulness of the application of equilibrium
theory to much of community ecology.
In this article I examine several hypotheses concerning one aspect of community structure, that is, species richness or diversity (5). I first explore the
view that communities seldom or never
reach an equilibrium state, and that high
diversity is a consequence of continually
changing conditions. Then I discuss the
opposing view that, once a community
recovers from a severe perturbation,
high diversity is maintained in the equilibrium state by various mechanisms.
Here I apply these hypotheses to organisms such as plants or sessile animals
that occupy most of the surface of the
land or the firm substrates in aquatic
habitats. I consider two tropical communities, rain forests and coral reefs, concentrating on the organisms that determine much of the structure. in these
cases, trees and corals. Whether my arguments apply to the mobile species,
such as insects. birds. fish, and crabs,
that use these structures as shelter
or food. or to nontropical regions. remains to be seen. I deal only with variations in diversity within local areas, not
with large-scale geographical gradients
such as tropical to temperate differences. While the hypotheses I present
may help explain them, such gradients
are just as likely to be produced by
mechanisms not covered in the present
article (6).
Various hypotheses have been proposed to explain how local diversity is
produced or maintained (or both). I have
reduced the number to six, which fall into two general categories:
Joseph H. Connell is a professor of biology at the
University of California, Santa Barbara 931%.
SCIENCE, VOL. 199, 24 MARCH 1978
1) The species composition of communities is seldom in a state of equilibrium. High diversity is maintained only
when the species composition is continually changing. (i) Diversity is higher
when disturbances are intermediate on
the scales of frequency and intensity (the
“intermediate disturbance” hypothesis).
(ii) Species are approximately equal in
ability to colonize, exclude invaders, and
resist environmental vicissitudes. Local
diversity depends only on the number of
species available in the geographical area
and the local population density (the
“equal chance” hypothesis). (iii) Gradual environmental changes, that alter the
ranking of competitive abilities, occur at
a rate high enough so that the process of
competitive elimination is seldom if ever
completed (the “gradual change” hypothesis).
2) The species composition of communities is usually in a state of equilibrium; after a disturbance it recovers to
that state. High diversity is then maintained without continual changes in species composition. (iv) At equilibrium each
species is competitively superior in exploiting a particular subdivision of the
habitat. Diversity is a function of the total range of habitats and of the degree of
specialization of the species to parts of
that range (the “niche diversification”
hypothesis). (v) At equilibrium, each
species uses interference mechanisms
which cause it to win over some competitors but lose to others (the “circular
networks” hypothesis). (vi) Mortality
from causes unrelated to the competitive
interaction falls heaviest on whichever
species ranks highest in competitive ability (the “compensatory mortality” hypothesis).
Nonequilibrium Hypotheses
The intermediate disturbance hypothesis. Organisms are killed or badly damaged in all communities by disturbances
that happen at various scales of frequency and intensity. Trees are killed or broken in tropical rain forests by windstorms, landslips, lightning strikes,
plagues of insects, and so on; corals are
destroyed by agents such as storm
waves, freshwater floods, sediments, or
herds of predators. This hypothesis suggests that the highest diversity is maintained at intermediate scales of disturbance (Fig. 1).
The best evidence comes from studies
of ecological succession. Soon after a severe disturbance, propagules (for example, seeds, spores, larvae) of a few
species arrive in the open space. Diver24 MARCH 1978
HIGH
n
Fig. 1. The “intermediate disturbance”
hypothesis. The patterns in species composition of adults and
young proposed by
Eggeling (8) for the difS U C C ~ S S ~ O ~ ~ ~
ferent
stages of the Budongo
forest are shown diagrammatically at the
bottom.
D l STURBANCES FREQUENT
:
SOON AFTER A DISTURBANCE
— LONG AFTER
r SMALL
DISTURBANCE LARGE
A . COLONIZING
sity is low because the time for colonization is short; only those few species that
both happen to be producing propagules
and are within dispersal range will colonize. If disturbances continue to happen
frequently, the community will consist of
only those few species capable of quickly reaching maturity.
As the interval between disturbances
increases, diversity will also increase,
because more time is available for the invasion of more species. New species
with lower powers of dispersal and slower growth, that were excluded by more
frequent disturbances, can now reach
maturity. As the frequency declines further and the interludes between catastrophes lengthen, diversity will decline, for
one of two reasons. First, the competitor
that is either the most efficient in exploiting limited resources or the most effective in interfering with other species
(or both) will eliminate the rest. Second,
even if all species were equal in competitive ability, the one that is the most
resistant to damage or to death caused
by physical extremes or natural enemies
will eventually fill much of the space.
This process rests on the assumption
that once a site is held by any occupant,
it blocks all further invasion until it is
damaged or killed. Thus it competitively
excludes all potential invaders, which
are by assumption incapable of competitively eliminating it (7).
Thus, diversity will decline during
long interludes between disturbances unless other mechanisms, such as those
given in the other hypotheses below, intervene to maintain diversity. Disturbances interrupt and set back the pro-
INFREQUENT
B,
MIXED
C , CLIMAX
cess of competitive elimination, or remove occupants that are competitively
excluding further invaders. Thus, they
keep local assemblages in a nonequilibrium state, although large geographic
areas may be stable in the sense that species are gained or lost at an imperceptible rate.
Evidence that this model applies to
tropical rain forests comes from several
sources. Eggeling ( 8 ) classified different
parts of the Budongo forest of Uganda
into three stages: colonizing, mixed, and
climax stands. Using observations made
many years apart, he showed that the
colonizing forest was spreading into neighboring grassland. In these colonizing stands the canopy was dominated
by a few species (class A in Fig. l), but
the juveniles (class B in Fig. 1) were of
entirely different species. Adults of the
class B species occurred elsewhere as
canopy trees in mixed stands of many
species. In these mixed stands, the juveniles were also mainly of different species (class C , Fig. I), those with even
greater shade tolerance. Adults of class
C species occurred in the canopy of other climax stands where a few species
dominated (mainly ironwood, Cynometra alexandrei, which comprised 75
to 90 percent of the canopy trees). However, in these stands, the understory was
composed mainly of juveniles of the canopy species. Thus, an assemblage of selfreplacing species (that is, a climax
community of low diversity) had been
achieved. This is not a special case; the
Budongo forest is the largest rain forest
in Uganda, and one-quarter of it is dominated by ironwood. Later and more ex-
Table 1 Mortal~tyof young trees (between 0 2 and 6 1 meters tall) In relatton to the~rabundance
for two ram forests In Queensland Not all spectes had enough young trees to analyze, only
those whose adults were capable of reach~ngthe canopy and that had at least SIX young trees are
Included The mortal~tyrate between 1965 and 1974 was plotted agalnst the or~glnalnumbers
mapped In 1965, the least-squares regresston slope and correlat~oncoeffic~entare shown
– –
–
– –
Regress~onof
mottal~ty(‘3)
on abundance
Number
S~te
of
specles
Trop~cal,North Queensland, 16″s
Subtrop~cal,South Queensland, 26″s
Slope
49
46
—
–
–
rr
0.039
0.217
0 002
0
—
*Neither correlat~oncoefficient 1s significantly different from zero at P < 05 tensive surveys (9) showed that the proportion so dominated in other forests in Uganda is even higher and have confirmed that, where Cynornetra dominates the canopy, its juveniles also dominate the understory Another excellent example is the work of Jones (10) in Nigeria. In this diverse tropical forest, many of the larger trees, aged about 200 years, were dying. They probably became established in abandoned fields in the first half of the 18th century, a time when the countryside was depopulated by the collapse of the Benin civilization. These trees had few offspring; most regeneration was by other species, shade-tolerant and of moderate stature. This mixed forest was in fact an "old secondary" forest that had invaded after agricultural disturbances. It was in about the same state as Eggeling's mixed forest in Uganda. In both Nigeria and Uganda, high diversity was found in a nonequilibrium intermediate stage in the forest succession. In many studies of forest dynamics, the abundance of juvenile stages constitutes the evidence as to whether a species is expected to increase or to die out. Such inferences are of course open to the criticism that, if the mortality rate of juveniles increases with their abundance, it is not necessarily a good indicator of more successful recruitment. I tested this for young trees in two rain forest plots in Queensland that several colleagues and I have been studying since 1963 (11). Over a 9-year period, mortality showed no correlation with abundance (Table 1). Thus, it seems safe to assume that species which now have many offspring will be more abundant in the next generation of adult trees as compared to those species which now have few offspring. In most of the mixed, highly diverse stands of tropical rain forests that have been studied, some species are represented by many large trees with few or no offspring, whereas others have a superabundance of offspring (8,lO-12). (Of course, many species are so rare as adults that one would not expect to find many offspring.) My interpretation of this finding is that these mixed tropical forests represent a nonequilibrium intermediate stage in a succession after a disturbance, in which some species populations are decreasing whereas others are increasing. Since mixed rain forests are common in the tropics, this hypothesis suggests that disturbance is frequent enough to maintain much of the region in the nonequilibrium state. If this is so, tropical forests dominated by a single canopy species that has abundant offspring in the understory must not have been disturbed for several generations. Such forests, similar to the ironwood climax of Eggeling (8), also occur commonly elsewhere in Africa as well as in tropical America and Southeast Asia (13). Two lines of evidence indicate that they have been less frequently disturbed than have mixed forests. First, the only papers that 1 have found in which the incidence of storms was described in relation to single-dominant forests state that destructive storms "never occur" in these regions (8, 14). Second, many of these forests are unlikely to have been disturbed by man, because they lie on poor soils, in swamps or along creek margins, on steep stony slopes, or on highly leached white sands (12, 15). All of these are soils that the farmers of shifting cultivation in forests avoid since they produce very poor crops (12). Such agriculture is confined mainly to the welldrained good soils, and these are the soils where the mixed diverse forests exist. Thus, mixed forests occur in the places most likely to have been disturbed by man, whereas single-dominant forests occur in those least likely to have been disturbed. Since single-dominant forests often (though not always) lie on poor soils, it has usually been assumed that this is because only a few species have evolved adaptations to tolerate them (12, 15). However, the difference between forests on good soils and those on poor soils lies in the dominance of a single species in the canopy rather than in the total number of species. Thus, in comparing plots in rain forests in Guyana, the commonest species constituted 16 percent of the large trees (more than 41 centimeters in diameter) on good soils and 67 percent on poor soils (leached white sands), yet the number of species of trees more than 20 centimeters in diameter was 55 and 49, respectively [table 27 in (12)l. Thus, a large number of moderate- to large-sized tree species occupy poor soils, even though only a few are common. The best evidence that single-species dominance is not necessarily due to poor soils is the example of the Budongo forest. Here, various forest stands, ranging from ones of mixed high diversity to those with single-species dominance, each occur on similar soils. Single-species dominance seems to be explained more satisfactorily by the absence of disturbance rather than by poor soil quality. On coral reefs, the relation between disturbance and diversity is similar to that in tropical forests. At Heron Island, Queensland, the highest number of species of corals occurs on the crests and outer slopes that are exposed to damaging storms. Since I began studying this reef in 1962, two hurricanes have passed close to it, one in 1967 and one in 1972. Each destroyed much coral on the crest and outer slopes but failed to damage another slope protected by an adjacent reef. The disturbed areas have been recolonized by many species after each hurricane, but colonization has not been so dense that competitive exclusion has yet begun to reduce the diversity (Fig. 2A). Other workers on corals have witnessed the same phenomenon; disturbances caused both by the physical environment and by predation remove corals and then recolonization by many species follows (16, 17). In contrast, in permanently marked quadrats observed over several years without disturbance at Heron Island, I found that competitive elimination of neighboring colonies was a regular feature, either by one colony overshadowing or overgrowing another, or by direct aggressive interactions (18). Here competition is by interference, rather than by more efficient exploitation of resources. On one region of the south outer slope, protected from storm disturbance by an adjacent reef, huge old colonies of a few species of "staghorn" corals occupy most of the surface (Fig. 2B). Since these are able to overshadow neighbors (18) at a height sufficient to be out of reach of the mesenteric filaments used as defenses (19), I infer that such staghorns have in fact competitively SCIENCE, VOL. 199 eliminated many neighbors during their growth. Here competitive elimination has apparently gone to completion, with a consequent reduction in local diversity. A similar situation has been described for Hawaii (16) and for the Pacific coast of Panama (17). The discussion so far has concerned mainly the frequency of disturbance. However, the same reasoning applies to variations in intensity and area perturbed; diversity is highest when disturbances are intermediate in intensity or size, and lower when disturbances are at either extreme. For example, if a disturbance kills all organisms over a very large area, recolonization in the center comes only from propagules that can travel relatively great distances and that can then become established in open, exposed conditions. Species with such propagules are a small subset of the total pool of species, so diversity is low. In contrast, in very small openings, mobility is less advantageous: the ability to become established and grow in the presence of resident competitors and natural enemies is critical. In addition, recolonizing propagules are more likely to come from adults adjacent to the small opening. Therefore, colonizers will again be a small subset of the available pool of species, and diversity will tend to be low. When disturbances create intermediate-sized openings, both types of species can colonize and the diversity should be higher than at either extreme. Not only the size, but also the intensity of disturbances makes a difference. If the disturbance was less intense so that some residents were damaged and not killed, in a large area recolonization would come both from propagules and from regeneration of survivors, so that diversity would be greater than was the case when all residents were killed and colonization came only from new propagules. Direct evidence linking diversity with variations in intensity and total area of disturbance in tropical communities is meager. However, there is evidence that the processes described above do occur. For example, a 40-kilometer-wide swath of reef in Belize was heavily damaged by a hurricane in 1961, with lesser damage on both sides. Four years later, in the middle of the swath, new colonies of a few species were present, but the only significant frame-building corals, mainly Acropora palmata , were the survivors of the original storm (20). Ten years later many of the new colonies were of this species. In contrast, in the zone of lesser damage, colonies and broken fragments of many species had survived the storm 24 MARCH 1978 and had regenerated quickly so that recovery was complete. Likewise, in rain forests, the size and intensity of a disturbance influences the process of recolonization. In a long-term study of a small experimental opening made in a Queensland rain forest, the most successful colonists after 12 years were either stump sprouts from survivors of the initial bulldozing or seedlings that came from adult trees at the edge of the clearing (21). Farther from the forest edge, in a much larger clearing, only species with great powers of seed dispersal had colonized (22). It has recently been suggested (23) that in a nonequilibrium situation, any conditions that increase the population growth rates of a community of competitors should result in decreased diversitv (since faster growth produces faster competitive displacement). In places with a lower rate of competitive elimination, there is also a greater chance for interruption by further disturbances. This "rate of competitive displacement" hypothesis is an extension of the intermediate disturbance hypothesis and should be true, other things being equal. How relevant it is for explaining differences in local diversity remains to be seen. However, present evidence from tropical communities does not support it. Forests on extreme soils (such as Fig. 2. Species diversity of corals in the subtidal outer reef slopes at Heron Island, Queensland. (A) Changes over 11 years on one of the permanently marked plots on the north slope. The number at each point gives the years since the first census at year 0 (no censuses were made inyears3.5,andlO). The dashed lines indicate changes caused by hurricanes in 1967 and 1972. (B) Results from line transects done 3 to 4 months after the 1972 hurricane. (A)Data from the heavily damaged north slopes; (0)data from the undamaged south slope; the line drawn by eye. Where disturbances had either great or little effect (very low or high percent cover, respectively) there were few species, with maximum numbers of soecies at intermediate levels of disturbance. leached white sands, heavy silt, or steep stony slopes) that have slower growth rates than those on less extreme soils have either few species or strong singlespecies dominance (24). Likewise, coral diversity shows little correlation with growth rates. Coral diversity varies with increasing depth, sometimes decreasing, sometimes increasing, or sometimes being greatest at intermediate depths. Coral growth rates tend to be faster at intermediate depths (24). Thus, among neither tropical rain forest trees nor corals is there a consistent correlation of diversity and growth rates, as predicted by the hypothesis. In summary, variations in diversity between local stands of these tropical communities are more likely to be due to differences in the degree of past disturbances than to differences in the rate of competitive displacement during recovery from the disturbances. The high diversities observed in tropical rain forest trees and in corals on reefs appear to be a consequence of disturbances intermediate in the scales of frequency and intensity. The equal chance hypothesis. In contrast to the previous model, let us assume that all species are equal in their abilities to colonize empty spaces, hold them against invaders, and survive the vicissitudes of physical extremes and natu- 24 1 A ! Square meter plot 1963 - 1974 18{ i & " 6 / / / r 2 24 I I B A 2 0 meter line transects O -- . ' $ l8 = z O 0 ' 0 b, 0 0, A 1 I I 10 I % Cover I 50 30 of I 70 I I 90 L ~ v e Coral 1305 I ral enemies. Then local diversity would forest was an old secondary one, similar simply be a function of the number of to Jones' (10) in Nigeria. species available and the local popuOther characteristics of trees and corlation densities. The species composition als do not satisfy the requirements of the at any site would be unpredictable, de- equal chance hypothesis. For example, pending upon the history of chance colo- dispersal of propagules of many trees nization. and corals is quite restricted so that local What conditions would produce this? recruitment of juveniles may not be as First, for all species the number of young independent of local production of prop(such as larvae and seeds) invading emp- agules as it apparently is in some fish ty places must be independent of the populations. Likewise, species differ in number produced by the parent popu- fecundity, competitive ability, and resistlation. Otherwise, any species that in- ance to environmental stresses, and the creased its production of offspring per differences often result in predictable parent would progressively increase at patterns of species distribution along enthe expense of those with lesser produc- vironmental gradients (27). Therefore, it tion. Second, any occupant must be able seems unlikely that either rain forest to hold its place against invaders until it trees or corals conform to the equal is damaged or killed. Otherwise, any chance hypothesis. species that evolved the ability to oust an The gradi~alchange hypothe;i;. This occupant would also progressively in- model was suggested by Hutchinson (28) crease. Last, all species must be equal in to explain why many species coexist in ability to resist physical extremes and phytoplankton assemblages. Seasonal natural enemies. Otherwise, the most re- changes in, for example, temperature sistant species will gradually increase, as and light, occur in a lake, and different was discussed in the previous hypothe- species are assumed to be competitively sis. superior at different times. It is posDo communities exist that satisfy tulated that no species has time to elimthese conditions? Sale (25) has proposed inate others before its ability to win that certain guilds of coral reef fish do. in competition is reduced below that He assumes that, as with some temper- of another species by changes in the ate fish, recruitment to newly vacated environment. sites is independent of the stock of eggs Climates change on all time scales released into the plankton. One must from seasonal to annual to millennia1 and probably assume that the fecundity and longer, and hence, this hypothesis may mortality of all species are equal. The ju- apply to organisms with any length of veniles grow quickly after they colonize generation. With long-lived organisms to vacant places, and thus they are able such as trees or corals, gradual changes to hold their territory against further in- in climate over several hundred years vasion by smaller juveniles of any spe- represent the same scale as seasons do to cies from the plankton. Space is limiting, a phytoplankton community. Drier perias judged from the rapid colonization of ods producing a savanna vegetation in vacated sites. Since the juvenile fish regions now covered with rain forest ocseem to be generalists in the use of food curred about 3000 and 11,000 years ago and space, Sale suggests that local diver- in the Amazon basin; similar changes ocsity would be a function of chance colo- curred in Africa and Australia (29). As nization from the available pool of spe- Livingstone (30) pointed out, "Climates cies. Clearly, the initial assumption of in- change and vegetational adjustments are dependence of stock and recruitment is not rare and isolated events, they are the critical and needs to be tested for these norm." As climates changed, marine tropical fish. transgressions shifted and altered coral Likewise, for rain forest trees, Aubre- reef environments (31). ville (26) has suggested that many speWhether such gradual transitions cies have such similar ecological require- would also produce the highly interminments that it would be impossible to pre- gled diverse assemblages seen in present dict which subset would occur together forests and reefs depends on the rate of on a site. He based this suggestion on the competitive elimination compared to the observation that some of the commoner rate of environmental change. If the time large trees on his study plot in the Ivory required for one tree species to eliminate Coast had few or no offspring on the another in competition is much shorter plot. He inferred that their offspring than the time taken for an environmental must be elsewhere, so that the species change that reversed their positions in composition of the forest would contin- the hierarchy, they would not coexist. ually shift in space and time. While this Therefore, very slow changes would not might be so, his original observation of maintain diversity, but higher rates few offspring could be explained if the might do so. Equilibrium Hypotheses The niche divers