|
Items in cart [1] |
Questions? Call 888-624-8373 |
| Copyright © 2009. National Academy of Sciences. All rights reserved. Terms of Use and Privacy Statement |
Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 16
Colloquium
toss of speciation rate will impoverish
future diversity
Michael L. Rosenzweig*
Deptartment of Ecology and Evolutionary Biology, University of Arizona, Tucson, AZ 85721
Human activities have greatly reduced the amount of the earth's
area available to wild species. As the area they have left declines,
so will their rates of speciation. This loss of speciation will occur for
two reasons: species with larger geographical ranges speciate
faster; and loss of area drives up extinction rates, thus reducing the
number of species available for speciation. Theory predicts steady
states in species diversity, and fossils suggest that these have
typified life for most of the past 500 million years. Modern and
fossil evidence indicates that, at the scale of the whole earth and
its major biogeographical provinces, those steady states respond
linearly, or nearly so, to available area. Hence, a loss of x% of area
will produce a loss of about x% of species. Local samples of
habitats merely echo the diversity available in the whole province
of which they are a part. So, conservation tactics that rely on
remnant patches to preserve diversity cannot succeed for long.
Instead, diversity will decay to a depauperate steady state in two
phases. The first will involve deterministic extinctions, reflecting
the loss of all areas in which a species can ordinarily sustain its
demographics. The second will be stochastic, reflecting accidents
brought on by global warming, new diseases, and commingling
the species of the separate big-provinces. A new kind of conser-
vation effort, reconciliation ecology, can avoid this decay. Recon-
ciliation ecology discovers how to modify and diversify anthropo-
genic habitats so that they harbor a wide variety of species. It
develops management techniques that allow humans to share
their geographical range with wild species.
Einstein pointed out that the essence of science consists in
connecting state variables to their derivatives. This is true of
the state variable called species diversity and the derivative
called speciation rate. We must study them together. And no
examination can succeed if it does not pay at least passing
attention to the other derivative involved the extinction rate.
What determines the number of species alive in a biological
province, such as the Neotropics? Answer: The cumulative
difference between the creative process of speciation and the
destructive process of extinction.
In this paper, I will restrict my attention to the species level of
biodiversity (although I do not thus mean to imply that no other
level is worthy or interesting). Furthermore, I will use an
old-fashioned definition for species i.e., a collection of organ-
isms that can exchange genes. I know that this definition does not
deal adequately with bacteria or interspecies hybrids or jumping
genes. But it will lead us to principles that will illuminate the
status of an important portion of diversity. And, perhaps,
successfully using it may guide us into the world of the multi-
myriads of species whose diversities we all too often ignore.
Further restricting myself, I will focus on diversity at the grand
scale, that of the entire biogeographical province. A biological
province is a self-contained region whose species originate
entirely by speciation within the region (1~. Scales smaller than
a province get their species through some form of dispersal,
echoing provincial diversity (2). Most probably, every real
province has obtained a small proportion of its species by
5404-5410 1 PNAS 1 May 8, 2001 1 vol. 98 1 no. 10
immigration, so the definition constitutes a mathematical ideal.
But percentages in excess of 90% are often found, and that
should be enough to allow speciation dynamics to dominate the
rate processes of a region.
Speciation results from a number of processes, in all of which
species are nurseries for other species. Ultimately, that is why the
future of speciation is bound up inextricably with the future of
species diversity itself. But the connection is even more inter-
esting than that. Both rates speciation and extinction depend
in part on the geographical range sizes of species. Larger ranges
tend to increase the speciation rates and decrease the extinction
rates of otherwise similar species. That gives us a parameter to
examine the area of the biogeographical province available to
life. Other things being equal, the larger the provincial area, the
larger the average species range and speciation rate and the
smaller the extinction rate.
What leads evolutionary ecology to its conclusions about
range size and the components of diversity dynamics?
.
.
.
Larger ranges offer larger targets for geographical isolating
barriers. The formation of geographical isolates begins the
sequence of events leading to allopatric speciation, probably
the source of most new species. Thus, a range that is more
readily subdivided by geographical barriers will spawn more
isolated populations per unit time and have a higher rate of
speciation.
· Widespread species also have more genetic variability than
narrowly distributed species (3~. This may also contribute to
the rate of speciation.
Larger ranges contain species with more subpopulations of
their metapopulation (4~. Ecologists believe that a complex
metapopulation offers some protection from extinction. A1-
though this question continues to be addressed (5), the basic
idea seems straightforward and intuitive. When a metapopu-
lation unit becomes extinct, recolonization may occur from
surviving units (6~. Each unit may survive or become extinct
independently of the others. So, the more units, the smaller
the probability that all will vanish simultaneously. (Notice that
I am not concerned here with the usual question: How does
Increasing fragmentation of a species range effect extinction
probability? Instead, I assume a constant degree of fragmen-
tation and focus on how more fragments alter extinction rate.)
Larger ranges are more difficult to contain entirely in the area
covered by any climatic change or anomaly. Any finite climatic
change (such as global warming) or anomaly (such as a severe
storm) will affect a finite amount of the earth's surface. If the
change covers a species' entire range, and degrades all its
appropriate habitat, the species will vanish. And if the anom-
aly destroys all of the individuals that live within its swath, the
This paper was presentec] at the National Acacdemy of Sciences colioqulum, "The Future of
Evolution," heic' March ~ 6-20, 2000, atthe Arnoic] anc! Mabel Beckman Center in Irvine, CA.
Abbreviation: my, million years.
*E-maii: scarab@?u.arizona.eclu.
www.pnas.org/cgi/cloi/ ~ 0. ~ 073/pnas. ~ 01092798
OCR for page 17
species will also vanish. Thus, the larger the area covered by
a species' range, the less chance that range will suffer perma-
nent or temporary obliteration. Again, a larger area dimin-
ishes extinction rates.
· Larger ranges produce species with greater total population
sizes. And, larger populations may have both lower extinction
(7) and higher speciation rates. Many conservation biologists
take its effect on extinction largely for granted. They see it as
the outcome of the improbability of the simultaneous, sto-
chastic demise of an entire large population. Much work
modifies that theory substantially without, however, challeng-
ing the relationship (8-12~. In addition, evidence from islands
supports it (13~. Yet, a recent investigation suggests that the
literature needs substantial revision to achieve a reliable
estimate of how long such extinctions take (14~. It concludes
that such extinctions may take much less time than we once
thought.
No one denies that population size influences speciation rate.
However, the direction of its effect is in doubt. At one time, many
evolutionists, led by Ernst Mayr (15), believed that small isolated
populations provide the crucible for evolution. They believed
that getting speciation started is a matter of breaking up
coadapted complexes of genes in geographical isolates. If that is
correct, small populations would speed up speciation by enhanc-
ing statistical sampling accidents. But an alternative view exists.
Called "centrifugal speciation" (16), it claims that large popu-
lations speed up speciation. Centrifugal speciation also begins
with geographical separation of sister populations. But after
separation, the larger isolates- not the smaller ones do the
changing. The small ones remain as evolutionary relicts.
According to the centrifugal model of speciation, even the
small population sizes of small areas work against evolution.
Some theoretical evidence from population genetics supports
the idea that small populations cannot evolve quickly (17~. So
does some (but not all) evidence from real populations (18, 19~.
And today we know that large populations foster novel gene
functions (17~. Certainly, genetic variability correlates positively
with the total population size of species (3~.
We should not allow our unfinished scientific debates to
distract us from noticing what we have already learned. Both
centrifugal speciation and Mayrian speciation are allopatric
modes and emphasize the importance of isolate formation.
Isolate formation should proceed faster in larger areas. Finally,
no biogeographical patterns suggest that "small" is more pro-
ductive. Instead, small is a recipe for choking off the speciation
process.
The fossil record supports the conclusion that extinction rate
and geographical range are inversely related over long periods
(20-23~. It does not however test the independent significances
of the various separate theoretical influences that I mentioned
above. The relative importance of these factors remains un-
known. Moreover, no one has yet tested the relationship of
speciation rate to range size in the fossil record. Although more
restricted in time scale, studies of modern areas do support both
the correlation of speciation rate with area (24), and the inverse
correlation of extinction rate with area (1~.
Interprovincial Patterns in Steady States of Species Diversity
The influence of diversity on average species' range size pro-
duces a negative feedback system in which species diversity is
self-regulating (1~. The difference between the speciation and
extinction rates of a province ought to approach zero. The net
result is a steady-state value of diversity at which new species
evolve as fast as established ones become extinct (1~.
The steady state of such systems might be only theoretical. The
environmental background within which it must be achieved
might be too fickle and variable for life ever to attain it. Or its
Rosenzweig
rates could be so slow that it has never been consummated. But
paleontological evidence strongly suggests that diversities of
regions are often not very far from such steady states (1, 25-27~.
Alroy's analysis of Cenozoic mammals in North America goes
even farther (27~. He shows that the dynamic elements of a
steady state have been present and active during this time. As
diversity fluctuated, speciation rate and extinction rate re-
sponded, keeping diversity within narrow bounds. In addition,
the two rates intersected over a diversity that did not change
significantly during the 65 million year (my) period. Finally,
although the per-taxon extinction rate did not vary significantly
with diversity, the per-taxon speciation rate did, declining with
increases in S and implicating variations in speciation rate as the
most responsive component of the system's dynamic stability.
Certainly, periods following a mass extinction are exceptions
to the rule of steady states, and so are periods following any
major increase in the value of the steady state. During both types
of periods, diversity tends to rise more or less monotonically.
Nevertheless, during the Phanerozoic Eon, the Earth has expe-
rienced only five mass extinctions and a similar number of major
increases in steady state (e.g., invasion of the land, invasion of the
muddy sea floor, radiation of angiosperms). Each such event
appears to have generated no more than 10 my of response in
species diversity. Most of the rest of the Phanerozoic Eon is a
record of long plateaus interspersed by quick episodes of ex-
tinction and recovery (28, 29~. Thus, I estimate that species
diversity has spent at least 90% of the past 500 my near some
steady state. (It would not surprise me to learn that the true
proportion is closer to 95% or even 98%.)
The Three Scales of Species-Area Relationships
The theory of provincial diversity predicts that larger provinces
will have higher steady state diversities. And they do. But in the
power equation that describes species-area curves, their z-values
hover near unity (30~. Thus, species-area curves among prov-
inces quite unlike those of islands or of subsamples of a single
province exhibit a lack of curvature in an arithmetic coordi-
nate space.
Other scales of the species-area relationship exist (31~.
Among islands of an archipelago, species diversity is governed by
the dynamics of immigrations to islands often a slow process,
but rarely as slow as speciation. Thus, they have more species
than would an equivalent-sized province. This reduces the
z-values among islands of an archipelago; they vary between 0.25
and 0.55. Among sample areas within a biogeographical prov-
ince, species diversity is governed both by the sample of habitats
in the area and by rates of local dispersal to sink populations.
Such rates are quite high compared with those of immigration.
Consequently, they have more species than would an equivalent-
sized island. Having more species further reduces the z-values of
areas within a single province compared with those of islands;
they vary between 0.1 and 0.2. In summary, we see a multiscale
picture of species-area curves with regularly varying z-values
(Fig. 1~.
No mathematical theory explains or predicts the fact that
diversity among provinces should be approximately a linear
function of area. Yet, several lines of evidence bolster our
confidence that the interprovincial z-value is close to unity. First,
the linear pattern also exists deep in the fossil record, as we can
see from examining some unusually good data about seed plants
in the northern hemisphere.
Tiffney and Niklas (32) compiled the numbers of seed plant
species known from 11 periods during the past 408 my. The last
of them (mid-Miocene) ended some 11.3 my ago. These periods
have two characteristics. We know their fossils rather well and
we also know the extent to which shallow seas flooded the land
during their tenure. The 11 periods have durations from 3.1 to
PNAS | May 8, 2001 | vol. 98 | no. 10 | 5405
OCR for page 18
Fig. 1. The three biological scales of species-area curves. Interprovincial
curves have z-values near unity; archipelagic curves have z-values that vary
from about 0.25 to 0.55; intraprovincial curves have the shallowest logarith-
mic slopes: their z-values vary from about 0.1 to 0.2. After Rosenzweig (1): Fig.
9.11.
40 my, and cover a total of 225.6 my. (The rest of the 408 my is
too poorly known for reliable analysis.)
Displayed against time, these data indicate a substantial rise
in species diversity (Fig. 2~. The most thorough analyses would
also note the long period of little change (ca. -330 my to -140
my) and infer the dominance of a steady state during that
interval, at least. But area also has a significant effect on these
data. The multiple linear regression:
logS = 1.78 + 0.33 logt + 1.00 logA
(R2 = 0.94; p' = 0.001; PA < 0~05)
where t is the amount of time since -408 my.
The most remarkable feature of that equation is the coeffi-
cient of area. It is unity, just as it is (approximately) for
regressions of modern provincial diversities.
The linear response of diversity to area among provinces
appears to explain the total absence of certain groups of native
species on small continents, such as ponerine and cerapachyine
ants on Hawaii. Two studies give us a picture of the diversity of
these subfamilies in tropical southeast Asia, Oceania, and New
Zealand (33, 34~. These fall nearly on a straight line in arithmetic
space (logS = -4 + 0.94 logs) and enable us to estimate the area
of a province that would have a single species (Fig. 3~. It is 18,354
km2. Hawaii, having only 10,378 km2, falls short.
A similar problem faces bracken-eating insects (35~. None live
in Hawaii, although the fern does grow there, occupying a range
of roughly 1,500 km2. Based on the species-area curve of such
insects in five other provinces, I estimate that one bracken-eating
insect species cannot be sustained on less than 10,361 km2.
Obviously, with so few data, such an estimate can give only the
order of magnitude of the surface required. However, this is one
order greater than what is available.
Finally, echo patterns of species diversity provide additional
confidence that we understand the scale differences among
species-area relationships. Echo patterns relate local diversity to
regional diversity. (In ideal case, regional diversity is equivalent
to the diversity of a biogeographical province, although the ideal
is rarely met in published examples of the phenomenon.) Based
on the scale differences of Fig. 1, theory predicts near-linear
Echo patterns (2~. Most often, that is what we see (36~.
5406 1 www.pnas.org/cgi/doi/10.1073/pnas.101092798
Fig. 2. Species of known fossils of seed plants in the northern hemisphere.
Diversity in Shrunken Natural Provinces
Ecologists question the amount of the earth's nonglaciated
terrestrial surface that people now use. Myers et al. (37) find that
some 88% of the world's richest habitats have been taken.
Huston (38) gives a figure of 95% for the world's nonglaciated
habitats; Vitousek et al. (39) estimate only 40 or 50%. Some of
these differences arise from disparate definitions of use, and they
are all honest attempts to deal with a somewhat amorphous
quantity. But no ecologist would disagree with the qualitative
judgment that we are taking a lot.
What does the pervasive impact of civilization do to wild
species? It shrinks the area of their province that they can use for
themselves. Will that reduce their diversity? Many careful ecol-
ogists are not so sure. To explain their doubt, they point to two
real phenomena biotic reserves and diversity hot spots and
the hope associated with combining them.
Biotic reserves constitute a collection of relictual habitats
scattered all over the world. They vary greatly in size, but all exist
to preserve a remnant of a once vast network of Edens. As an
ensemble, they are hyperdispersed in biotic space. That is to say,
they form a deliberately varied collection. We spend our limited
resources to save the dwindling last bits of long-leaf pine forest
or tall-grass prairie or Arizona riparian bottomland. We delay
reserving that which is presently common. Later, I will explain
that this approach has indeed given us precious time, although
it cannot forever avert the evil decree.
Fig. 3. The interprovincial species-area relationship for cerypachine and
ponerine ants extrapolated to the area of a province that would sustain a
single native species. That area is 18,354 km2. Hawaii has only 10,378 km2 and
has no native species of these ants.
Rosenzweig
OCR for page 19
Fig. 4. The three phases of mass extinction that follow a severe reduction
in area. Endemic extinctions occur because species lose all their habitat.
Sink species disappear because all their source populations lose their
habitats. Accidental extinctions remove species that suffer a run of bad
luck despite being able to sustain their populations during an average
generation.
What is the apparent significance of hot spots of diversity?
They tantalize us with the promise that we can minimize the
amount of reservation needed to prevent a mass extinction. For
example, about 44% of all vascular plant species and 35% of all
species of mammals, birds, reptiles, and amphibians live confined
to only about 1.4% of its land in 25 hot spots (37~. Clearly, this
1.4% is critical. If it goes, so do all those species. But, it is very
tempting to reverse that argument: Let us save the 1.4%.
Perhaps, then, we will save all those species now living in it. That
is the combination of reservation and hot spots that offers hope
to some conservation biologists.
Unfortunately, the interprovincial pattern of diversity ques-
tions that hope. It underscores the fact that diversity is the
outcome of dynamical processes, whereas hot spots constitute a
static view of diversity. The differences among the scales of
species-area relationships and their z-values provide a dynamic
basis for predicting what will happen to species diversity and to
speciation in the future.
Following a reduction in province size, the three species-
area relationships predict a mass extinction in three phases
(Fig. 4~. First, we lose the endemics those species whose
habitat gets entirely expropriated (40~. These extinctions will
be deterministic and virtually instantaneous. Second, we lose
the sink species- those which get restricted to marginal hab-
itats (i.e., habitats in which their death rates exceed their birth
rates). These extinctions will also be deterministic (41, 42~;
however, they will take a while because some individuals of
sink species not only survive, but may also reproduce. Third,
we lose not particular species, but the diversity in excess of the
new steady state. These extinctions will be stochastic, and
predicting their rate even their average rate to an order of
magnitude may not be possible as I will soon point out.
Nevertheless, it is possible to say that the diversity lost because
of these extinctions will be deterministically and irreplaceably
lost after all, the predicted loss constitutes however many
species it takes to restore a steady-state diversity to the earth.
Stages I and 11: The Deterministic Extinctions
The coefficients of species-area equations at smaller scales
permit us to estimate the proportions of species that are en-
demics or sink species. The prediction that we lose the sum of
Rosenzweig
Cli_l
Fig. 5. Declining area reduces the proportion of species maintained at
steady state. The amount of the reduction depends on the z-value. This figure
shows the result for z-values of 1.5 (---), 1 (—), 0.9 (- · · -), and 0.3 (- - -).
Assuming the latter, which is a typical archipelagic z, we would predict the
least severe losses. But data show that interprovincial z-values are close to
unity.
these is the well known and often cited equation of island
biogeography:
S =A03
where S end A are proportions of diversity and natural area that
remain. I plot this relationship in Fig. 5.
Some impatient types have complained that these predicted
extinctions have not happened. In some cases, no doubt, their
analyses have been flawed (43~. But another factor has been at
work. Conservation's two strategies have stayed the imposition
of the island equation, and may even have reduced its severity.
Focusing on preserving the habitats most likely to vanish
entirely, we considerably reduced the likelihood that any
species would lose everything, including its sink habitats. That
transferred some extinction of endemics to the category of
extinction of sink species. And, although the island equation
assumes so, we have not lost natural area randomly. Instead,
we diversified our biotic reserves, making them a stratified
sample of our habitats and the species they harbor. Although
we still have no analytical theory to predict the power coef-
ficient of the island equation, a stratified sample has to have
preserved more habitats and their species than a random
sample would have. That is the good news.
Other news is more upsetting. Extinction, like speciation, is a
dynamical process. Extinction, like speciation, takes time (44).
Relaxation to island equilibrium is likely to take a number of
human generations (see ref. 37 and references 34-43 therein).
We know that the past 10,000 years have not sufficed to allow
mammals on newly formed islands to attain new steady states
(e.g., Fig. 6~. Lizard species diversity in the wheatbelt reserves of
Western Australia has declined to a z-value of 0.26, but has yet
to reach the value of 0.36 found among the state's true islands
(Fig. 7~. And on new islands in the Sea of Cortez, lizard species
diversity continued to decline for at least 12,000 years (45~. We
must not be too impatient.
Stage 111: The Stochastic Extinctions
A true island reaches its steady state as a result of the balance
between immigration and extinction. But our shrunken natural
world will see no immigrations. Species that become extinct
cannot immigrate from the past to recolonize the world of the
future. So, even after our new world reaches its levels of
island-like diversity, its diversity will continue to diminish.
PNAS | May 8, 2001 | volt. 98 | no. 10 | 5407
OCR for page 20
Fig. 6. Mammal diversity on Sunda Islands (circles) and southwestern U.S.
mounta i ntops (triang les). These isolates formed about 10,000 yea rs ago at the
end of the Pleistocene. In both archipelagos, larger islands have experienced
proportionately less extinction. After Rosenzweig's (1) figure 6.5.
Relaxation below the island-like steady state will come from
inflated extinction rates. But the species of our shrunken island
world will all begin with at least one source population. They will
all have enough habitat to keep them going in perpetuity-
provided that nothing ever changes in those habitats. But
something always changes. So, species will vanish simply because
they lose one roll of the dice. Global warming may push their
remaining habitats out of all reserves and into cornfields or the
sea (46~. New parasites and diseases will appear to take their toll.
A series of bad weather events may be too much for some
dwindled populations.
The Depauperate Steady State
Accidents that eradicate successful species have always accom-
panied life. However, in ordinary times, life has replaced such
losses by speciation. Not this time. This time the loss of area will
have also depressed the speciation rate curve. Constricted
geographic ranges will have fewer isolates. Many species will be
restricted to a single reserve with no chance of further allopatric
speciation. The loss of ecological theater will change the evo-
lutionary play. New speciations will not be able to keep up with
the losses.
_~_l
Fig. 7. Lizard species diversity in the 22 isolated reserves of Western Aus-
tralia's wheatbelt (triangles) and its marine islands (circles). The reserve species
area curve shows some decline from what would be expected in pre-European
times as its z-value is 0.26. But it has not yet reached the value of 0.36 found
among the true islands.
5408 1 www.pnas.org/cgi/doi/10.1 073/pnas.101 092798
Fig. 8. Diversity dynamics of three provinces of different area: S (small, ~ -);
M (medium,—); and L (large, - - -). Extinction rates have positive second
derivatives; speciation rates have negative second derivatives. The steady
state of each province is indicated at the point where speciation and extinction
balance (diamond, small; box, medium; circle, large).
Yes, eventually, a balance will be restored. But before that
happens, the total extinction rate must decline to the level of the
total speciation rate (Fig. 8~. Thus, at the steady state of the
future, when life is replacing its losses by speciation, it will not
build back the diversity it now enjoys. The new balance will occur
with life decimated of its richness.
Like any evolutionarily independent province, our future
world must seek its steady state along the interprovincial spe-
cies-area curve. The z-value of this curve is approximately unity,
so our losses of species should be approximately linear (Fig. 5~.
Lose 10% of the natural world's surface and we save about 90%
of its species. Lose 95% and save only 5% of the species.
Diversity in provinces appears to have been following such a law
for hundreds of millions of years. We have no evidence to
indicate that law has been repealed.
This prediction is much more severe a decree than the usual
proclamation of doom. The usual bad news predicts the loss of
about half the world's species. But such rosy predictions rest on
the z-value of island species-area curves, which is only about 0.3.
How long does it take to restore a steady state in species
diversity? Perhaps a very long time- although, in truth, we do
not know. One thing we can say. The process of re-achieving a
steady state after this biotic crisis will not resemble any previous
recovery from a mass extinction.
· First, as we have seen, life will not recover anything like its
previous levels of diversity.
· Second, some taxa will vanish forever. The new world's area
will fall below the threshold they require for any speciation
(24), and so it will be unable to sustain even one of these
species, just as Hawaii cannot sustain even one species of
native insect that specializes in eating bracken, or even one ant
of certain subfamilies (see above).
Third, previous mass extinctions were a violent interruption
and perturbation of steady-state diversities. Afterward,
background conditions returned and life recovered its steady
states through the action of speciation. In this case, however,
the mass extinctions represent a gradual relaxation to a new
steady state dictated by the shrunken area available to
nature. Thus, recovery of steady-state dynamics will occur as
soon as the mass extinction is over i.e., after complete
relaxation. The trajectory of extinctions not the trajectory
Rosenzweig
OCR for page 21
of originations will determine how long the process will
take.
We can determine an upper limit to relaxation time by realizing
that, as a dynamic process, its time scale should resemble its flux
rates. Most of the time speciation-extinction flux is glacial. Esti-
mates of background rates come from both the fossil record and
modern phylogenetic reconstructions. Mammal species appear to
turn over on a scale of 105 yr. Shelly marine invertebrates on a scale
of 106 yr. But these are normal background rates.
Human pressure may greatly accelerate the relaxation process
by increasing accidental extinction rates. Various human activ-
ities suggest this. We increasingly commingle evolutionarily
separate provincial biotas, creating the New Pangaea and intro-
ducing native species to predatory and competitive threats from
exotics (47~. We rapidly transport novel diseases and parasites
around the world. We simplify biotic temporal regimes (for
example by limiting disturbances such as fire). And we are
warming the globe. The National Research Council (44) impli-
cates exotic species or lack of adequate disturbance as the root
cause in endangering a significant proportion of threatened U.S.
species. But global warming may constitute the worst threat of
all: by altering the basic abiotic conditions of reserves, it can
destroy their ability to do much of their job. When the earth was
covered with contiguous tracts of natural habitat, species could
track such changes, moving to keep up with the shifts in location
of their favored habitats and so avoiding extinction (48-50~. But
today, with natural habitats restricted to patches of reserves, this
is not possible. Meanwhile, we show little sign of abandoning
the destruction of habitat that brings deterministic extinction
to species.
So, how long could it take to achieve a world so poor in species
that its speciation rates once again counterbalance its accidental
extinctions? Perhaps a million years, perhaps a hundred.
Reconciliation Ecology
Not all species are losing ground to us. Some live with us and
prosper. In German, they are known as kulturfolger—culture
followers. One might dream that these kulturfolger would speci-
ate rapidly and take up the slack. But the population isolates of
kulturfolger cannot make much evolutionary headway. To do so
would be to adapt to local circumstances, and that adaptation
requires time. Thus, it requires that the isolates face some
stability in the environmental challenges we cast at them. But
Technological Man does not allow stable intervals; we constantly
change our habitats and exert new pressures on the kulturfolger.
They may not disappear, but they are also unlikely to radiate into
a wide variety of new species.
Can we find a way to change all this? In the felicitous metaphor
of Norman Myers, can we find a way to save Darwin's Genie and
the treasures it has given us? We can. But not if we accept today's
dominant strategy of conservation biology. For historical reasons,
conservation biology has become mired in an attitude of confron-
tation: The green forces of nature versus the green forces of money.
Conservation divides the world into pristine habitats and ruined
habitats. It tries to save and restore the former while preventing
further loss.
Let us ignore the fact that all of us must live in one world,
accepting both its economic environment and its ecology. Let us
also ignore the fact that if we view nature as an embattled victim,
we are ultimately consigning her to defeat at the hands of the
superior forces of human population pressure, human guile, and
human greed. Let us instead concentrate on the science.
Science insists that area is an intrinsic property of natural
ecosystems. To maintain their diversity, they must have their area.
Thus, conservation biology has to address itself to the habitats in
which human beings live, work, and play. Conservation biology has
to learn how to share anthropogenic habitats with wild species. It
Rosenzweig
needs to discover how to modify and diversify those habitats so that
they harbor a wide variety of species. I call this sort of conservation
biology reconciliation ecology.
Reconciliation ecology seeks techniques of management that
turn more species into kulturfolger, giving them back their
geographical ranges without taking away ours. It takes advantage
of the fact that the earth remains as large as ever, but that we
have devised new habitats in which most species cannot function
at all. If we meet wild species half way, many will adapt to our
world. They will spread in it and reestablish their potential for
speciation and their resistance to extinction.
A growing number of examples demonstrates that reconcil-
iation ecology can work. Countryside biogeography is showing
that some styles of land use are already compatible with the
needs of many species (51-534. Projects devoted to particular
species are bringing them back in anthropogenic, economically
productive habitats (54-56~. Other projects focus on develop-
ing and managing habitats (57-60~. The examples I cite are but
a small sample of those I have collected over the past 3 years.
One project created a single patch of salt marsh at a critical
point in the migratory flyway of perhaps a third of all of the bird
individuals in Europe and western Asia. In doing so, it saved at
least a fraction of the 257 species that use that flyway (614. Until
30 years ago, a nearby 12-km2 natural salt marsh had done this
job. But the natural salt marsh was totally destroyed by resort
development. Today's patch of salt marsh little resembles its
predecessor or any natural habitat. It is carefully built up,
contoured, and planted on a refuse dump. And it is regularly
irrigated with treated, nutrient-rich sewage water.
Another project set out to save a long-leaf pine forest and its
endangered species such as the red-cockaded woodpecker (62~.
Through novel, carefully studied, and continuous management
in a large, important Air Force base, it has multiplied by two
orders of magnitude the area covered by a sustainable long-
leaf pine forest. Meanwhile, residential, recreational, timbering,
and all military uses continue. This is not restoration. Never
before has there been a long-leaf pine forest like this one.
As the millennium changes, the media engorge on visions of
the future. I have yet to hear one that does not involve technol-
ogy. My own vision of the year 2100 is fundamentally different.
We will have the technology, but we will be using it in magnif-
icent surroundings that help to bring us peace and fulfillment,
and to discharge our responsibilities to nature. I see my great-
grandchildren opening their doors on a world of wonders.
Perhaps this vision is mere fantasy. Perhaps it is my way to
cope with the unthinkable, because, indeed, the alternative is too
disheartening to contemplate. Evidence indicates that we cannot
preserve the large scale at the tiny scale (63~. Area constitutes
a basic inherent property of every biome, a property crucial to
the dynamical functioning of its components. So it is an oxymo-
ron to imagine a pristine biome that retains only 2 or 5 or even
10% of its original size.
In particular, area helps set the provincial diversity of a ma-
ture province. In turn, provincial diversity determines local diver-
sity local diversity can only echo provincial diversity. Thus, if the
area available to native species remains very low or declines even
farther, the loss of speciation rate will prevent even our biotic
preserves from maintaining their diversities for very long.
For an evolutionary ecologist, the year 2100 lies in plain
sight, a short trip into the future. At the end of this path, we
could find a splendid set of human environments, clean and
brimming with a richness of life that will invigorate and restore
our spirits. But this will happen only if we take advantage of
the ecological opportunities we still have.
Thanks to Norman Myers and Andrew Knoll for their vision. Useful
suggestions came from Andrew Knoll, John Alroy, Harald Beck-King, an
anonymous reviewer, and discussions at the colloquium itself.
PNAS | May 8, 2001 | vol. 98 | no. 10 | 5409
OCR for page 22
1. Rosenzweig, M. L. (1995) Species Diversity in Space and Time (Cambridge Univ.
Press, Cambridge, U.K.).
2. Rosenzweig, M. L. & Ziv, Y. (1999) Ecography 22, 614-628.
3. Mitton, J. B. (1997) Selection in Natural Populations (Oxford Univ. Press,
Oxford).
4. Hanski, I. & Gilpin, M. E., eds. (1997) Metapopulation Dynamics: Ecology,
Genetics and Evolution (Academic, New York).
5. Harrison, S. (1998) in Biodiversity Dynamics: Turnover of Populations, Taxa, and
Communities, eds. McKinney, M. L. & Drake, J. A. (Columbia Univ. Press, New York).
6. Allen, J. C., Schaffer, W. M. & Rosko, D. J. (1993) Nature (London) 364, 229-232.
7. Maurer, B. A. & Nott, M. P. (1998) in Biodiversity Dynamics: Turnover of
Populations, Tauca, and Communities, eds. McKinney, M. L. & Drake, J. A.
(Columbia Univ. Press, New York), pp. 31-50.
8. Leigh, E. G. (1981) J. Theor. Biol. 90, 213-239.
9. Goodman, D. (1987) in Viable Populations for Conservation, ed. Soule, M. E.
(Cambridge Univ. Press, Cambridge, U.K.), pp. 11-34.
10. Gilpin, M. (1992) J. Theor. Biol. 154, 1-8.
11. Lande, R. (1993) Am. Nat. 142, 911-927.
12. Lynch, M., Conery, J. & Burger, R. (1995) Am. Nat. 146, 489-518.
13. Pimm, S. L., Jones, H. L. & Diamond, J. (1988) Am. Nat. 132, 757-785.
14. Belovsky, G. E., Mellison, C., Larson, C. & van Zandt, P. A. (1999) Science 286,
1 175-1 177. 46
15. Mayr, E. (1954) in Evolution as a Process, eds. Huxley, J. S., Hardy, A. C. &
Ford, E. B. (Allen and Unwin, London), pp. 157-180.
16. Brown, W. L., Jr. (1957) Q. Rev. Biol. 32, 247-277.
17. Walsh, J. B. (1995) Genetics 139, 421-428.
18. Greenbaum, I. F., Baker, R. J. & Ramsey, P. R. (1978) Evolution 32, 646-654.
19. Templeton, A. R., Robertson, R. J., Brisson, J. & Strasburg, J. (2001) Proc.
Natl. Acad. Sci. USA 98, 5426-5432.
20. Hansen, T. A. (1980) Paleobiology 6, 193-207.
21. Jablonski, D. (1986) Science 231, 129-133.
22. Jackson, J. B. C. (1974) Am. Nat. 108, 541-560.
23. McKinney, M. L. (1997) in The Biology of Rarity, eds. Kunin, W. E. & Gaston,
K. J. (Chapman & Hall, London), pp. 110-129.
24. Losos, J. B. (1996) Philos. Trans. R. Soc. London B 351, 847-854.
25. Boucot, A. J. (1975) Evolution and Extinction Rate Controls (Elsevier Scientific
Publications, Amsterdam).
26. Miller, A. I. & Foote, M. (1996) Paleobiology 22, 304-309.
27. Alroy, J. (1998) in Biodiversity Dynamics: Turnover of Populations, Taxa and
Communities, eds. McKinney, M. L. & Drake, J. A. (Columbia Univ. Press, New
York), pp. 232-287.
28. Brett, C. E., Ivany, L. C. & Schopf, K. M. (1996) Palaeogeogr. Palaeoclimatol.
Palaeoecol. 127, 1-20.
29. Miller, A. I. (1998) Science 281, 1157-1160.
30. Rosenzweig, M. L. (1998) in Biodiversity Dynamics: Turnover of Populations,
Taxa, and Communities, eds. McKinney, M. L. & Drake, J. A. (Columbia Univ.
Press, New York), pp. 311-348.
31. Rosenzweig, M. L. (1999) in Advanced Theoretical Ecology: Principles and Appli-
cations, ed. McGlade, J. (Blackwell Science, Oxford), Vol. 80, pp. 249-281.
32. Tiffney, B. H. & Niklas, K. J. (1990) in Causes of Evolution: A Paleontological
Perspective, eds. Ross, R. M. & Allmon, W. P. (Univ. of Chicago Press,
Chicago), pp. 76-102.
5410 1 www.pnas.org/cgi/doi/10.1073/pnas.101092798
33. Wilson, E. O. (1961) Am. Nat. 95, 169-193.
34. Wilson, E. O. & Taylor, R. W. (1967) The Ants of Polynesia (Hymenoptera:
Formicidae) (Bernice P. Bishop Museum, Honolulu).
35. Lawton, J. H. (1984) in Ecological Communities: Conceptual Issues and the
Evidence, eds. Strong, D. R., Simberloff, D. S., Abele, L. G. & Thistle, A. B.
(Princeton Univ. Press, Princeton).
36. Lawton, J. (1999) Oikos 84, 177-192.
37. Myers, N., Mittermeier, R. A., Mittermeier, C. G., da Fonseca, G. A. B. & Kent,
J. (2000) Nature (London) 403, 853-858.
38. Huston, M. (1993) Science 262, 1676-1680.
39. Vitousek, P., Mooney, H., Lubchenco, J. & Melillo, J. (1997) Science 277,
494-499.
40. Harte, J. & Kinzig, A. (1997) Oikos 80, 417-427.
41. Patterson, B. D. (1990) Oikos 59, 330-342.
42. Patterson, B. D. & Atmar, W. (2000) in Isolated Vertebrate Communities in the
Tropics; Proc. 4th Intern. Symp., ed. Rheinwald, G., (Bonn. Zool. Monogr.),
Vol. 46, pp. 93-108.
43. Brooks, T. M., Pimm, S. L. & Collar, N. J. (1997) Conserv. Biol. 11, 382-394.
44. National Research Council (1995) Science and the Endangered Species Act
(Natl. Acad. Press, Washington, DC).
Wilcox, B. A. (1978) Science 199, 996-998.
Peters, R. L. & Darling, J. D. S. (1985) Bioscience 35, 707-717.
Mooney, H. A. & Cleland, E. E. (2001) Proc. Natl. Acad. Sci. USA 98,
5446-5451.
48. Brett, C. E. (1998) Palaios 13, 241-262.
49. Coope, G. R. (1987) in Organization of Communities Past and Present, eds. Gee,
J. H. R. & Giller, P. S. (Blackwell Scientific, Oxford), pp. 421-438.
50. Davis, M. B. (1983) in Late-Quaternary Environments of the United States, ed.
Wright, H. E., Jr. (Univ. of Minnesota Press, Minneapolis), pp. 166-181.
51. Daily, G., Ehrlich, P. R. & Sanchez-Azofeifa, G. A. (2001) Ecol. Appl.
11, 1-13.
52. Vandermeer, J. & Perfecto, I. (1995) Breakfast of Biodiversity (Food First,
Oakland, CA).
53. Greenberg, R., Bichier, P. & Sterling, J. (1997) Biol. Conserv. 80, 235-247.
54. Denton, J. S., Hitchings, S. P., Beebee, T. J. C. & Gent, A. (1997) Conservation
Biology 11, 1329-1338.
Gaby, R., McMahon, M. P., Mazzoti, F. J., Gillies, W. N. & Wilcox, J. R. (1985)
J. Herpetol. 19,189-198.
56. Yosef, R. & Grubb, T. C., Jr. (1994) Auk 111, 465-469.
57. Bormann, F. H., Balmori, D. & Geballe, G. T. (1993) Redesigning theAmerican
Lawn (Yale Univ. Press, New Haven, CT).
58. Esselink, H., Geertsma, M., Kuper, J., Hustings, F. & van Berkel, H. (1995)
Proc. West. Found. Vertebr. Zool. 6, 287-293.
59. Edwards, V. M. (1995) Dealing in Diversity: America's Market for Nature
Conservation (Cambridge Univ. Press, Cambridge, U.K.).
60. Western, D. (2001) Proc. Natl. Acad. Sci. USA 98, 5458-5465.
61. Cherrington, M. (1999) Earthwatch May/June, 39-47.
62. McWhite, R. W., Green, D. R., Petrick, C. J., Seiber, S. M. & Hardesty, J. L.
(1993) Natural Resources Management Plan, Eglin Air Force Base, Florida (U.S.
Dept. of the Air Force, Eglin Air Force Base, FL).
63. Harrison, S. & Bruna, E. (1999) Ecography 22, 225-232.
Representative terms from entire chapter:
species diversity
