Download Chance Dispersal - Plant Geography - Handout | BOTANY 422 and more Study notes Geography in PDF only on Docsity! ‘Sherwin Carlquist
nN
«|
Instances of chance dispersal have
been observed and reported from
time to time, and many biogeogra-
phers believe that this phenomenon
has played an important role in the
natural distribution of plants and
! animals. Yet other scientists doubt
| that such occurrences as seeds or
snails clinging to birds’ feathers can
account for the development of the
floras and faunas of entire islands,
viewing these events as freak hap-
penings of no significance. This paper
will discuss examples of chance dis-
persal; and will outline another
prominent theory for the spread of
organisms.
Charles Darwin was interested in the
phenomenon of dispersal to islands
because his theory of evolution dic-
tated that a given species orginated
only once, and then spread thereafter.
The alternate idea that a species
could originate independently in
various places throughout the world
tequired an essentially Lamarckian
explanation. Thus in Darwin’s 1859
correspondence with Sir Joseph
Hooker, for example, he mentions
with delight Milner’s discovery that
nestling petrels on the Scottish island
Sherwin Carlquist is Horton Professor of
Botany at Claremont Graduate Schnol and
Pomona College, and is a Research Associate
2 the affiliated Rancho Sante Ana Rotanic
Garden. The author of numerous books and
Papers on island biology and the comparative
Anatomy af flowering plants, he has been
*qually involved in botanical exploration in
“ar:ous parts of the world, especially the
Southern’ Hemisphere, and in taboratory
Study of the materials he has collected. Ad-
tress; Rancho Santa Ana Botanic Garden,
Claremont, CA 9I7IE
Unance Uispersal
Long-distance dispersal of organisms, widely
accepted as a major cause of distribution patterns,
poses challenging problems of analysis
of St. Kilda had West Indian seeds in
their crops.
In this case it seems unlikely that the
petrels themselves carried the seeds
from the West Indies to St. Kilda.
Rather, the adult petrels may have
picked up the seeds along Scottish
beaches and fed them to juveniles; the
Gulf Stream deposits seeds from the
West Indies on the shores of the
British Isles. However, this example
does show the tendency of marine
birds to ingest seeds, and could also
explain how seeds deposited on a
beach by ocean currents could reach
an upland locality more suitable for
their growth. Other examples of ma-
rine birds consuming seeds and fruits
appear in Ridley’s wonderful com-
pendium of instances of seed disper-
sal (1930), which lists so many ob-
served occasions of dispersal of ail
kinds that the chance events seem to
add up to a genuine phenomenon,
that of long-distance dispersal.
Guppy (1906) devoted most of a vol-
ume to his observations on plant dis-
persal in the Pacific. He noted dis-
tributions of plants in the region and
attempted to find a correlation with
means of dispersal. The easiest
method of dispersal to test is that of
flotation in seawater, and Guppy’s
beok contains an overwhelming
amount of data showing which seeds
and fruits float, and for how long. One
can only imagine that Guppy lived for
years surrounded by jars of floating
seeds. His experiments predictably
showed that most beach plants are
dispersed by seawater, and revealed
similar dispersal possibilities for some
inland plants, such as the vines Ent-
ada and Mucuna, which must drop
seeds into rivers or in some places
overhang the seacoast. However,
Guppy’s studies focus mainly on the
dispersal of beach plants, and we
must therefore look elsewhere for
examples of other kinds of chance
dispersal.
Other documented instances of dis-
persal show the surprising range and
variety of the phenomenon. Small
snails have been found adhering to
bird feathers in a number of cases
(Vagvolgyi 1975), and seeds discov-
ered in bird feathers and in the mud
on birds’ feet have been successfully
germinated (Wallace 1895). Rand
(1955) noted that the purple gallinule,
known as a chance visitor to Tristan.
de Cunha for many years, has finally
formed breeding colonies there, and
similar instances of dispersal of bird
species followed by establishment
have been recorded by MacDowall
(1978). Wheeler (1916) reported that
ants survived a journey of about 5 km
in a floating log from the Brazilian
coast to Sao Sebastiano Island.
Monarch butterflies, which occa-
sionally migrate over the Pacific, were
observed to establish colonies on
Canton Island by Zwaluwenberg
(1942), who also noted the simulta-
neous establishment of the food
plants they needed. When I visited
Pearl and Hermes Reef in Hawaii, 1
noticed that numerous Mucuna seeds
had floated ashore. A few were ger-
minating, but these soon withered in
the hot sun, since Mucuna grows
successfully only in wet forest. This
example, like many of the preceding
ones, indicates that, rare as events of
chance dispersal are, adverse ecolog-
ical conditions may be a greater ob-
stacle to the establishment of a
species in a new location than trans-
port itself.
The ideal site for the study of chance
1981 September-October 509
north cold
temperate 4%
Indo-Pacific 44%
North and South
America 22%
HAWANT
pantropical 13%
~ MARQUESAS
TANT
south cotd
temperate 17%
Figure 1. Evidence suggests that there were 272
hypothetical ancestors of Hawaii’s flowering
plants, arriving from various regions. Biologists
agree that both the patterns of distribution and
dispersal is an island, because an
island represents the ultimate test of
long-distance transport. If a seed
transported by wind fails to the
ground, it can be picked up again, but
if it falls into the sea, the dispersal
event usually ends. An ideal island
would be one newly emerged from the
sea, not previously in contact with
any land mass, and completely devoid.
of any life at its time of origin. It
should be large enough and high
enough to support various types of
life, and should be in a climatically
favorable part of an ocean. The island
should be well removed from conti-
nents and other islands: if dispersal is
too easy, one learns little.
Though no such ideal site is now
available for study, the island of
Surtsey, near Iceland—too close to
Iceland and in too unfavorable a cli-
mate to have shown a great deal since
its emergence in 1963—has never-
theless illustrated some dispersal
phenomena well (Einarsson :1967).
Beach plants dispersed by drift, such
as sea rocket (Cakile), were the first
to arrive. aie
The island closest to the ideal in-the
the relationship of Hawaiian angiosperms to
other angiosperm species point to long-distance
dispersal as the source of the island’s flora.
(Data from Zimmerman 1948.)
past was Krakatau, in the strait be-
tween Java and Sumatra. In 1883
Krakatau erupted with extraordinary
violence. Very few species, if any,
survived. Fortunately, the signifi-
cance of the island was appreciated,
and studies were begun soon after the
eruption. Reports covering fifty years
of recoyery of life on the island have
been prepared, for animals by Dam-
mermann (1948) and for plants by
Docters van Leeuwen (1936).
These studies show that both plant
and animal species appeared slowly in
the years following the eruption; then,
as soils and plant cover formed, the
rate of immigration accelerated.
Later, when the number of species
began to approach what might be
expected on a small island in this re-
gion, there were fewer new coloniza-
tions. Although the’ distances tra-
versed:were small—Krakatau is only
45 km from Java, and an even shorter
distance from othér islands—the or-
ganisms that appeared could all be
said to have good dispersal mecha-
nisms. The studies of Krakatau firmly
support the idea that chance dispersal
over long distances occurs, but since
the istand is not on the scale of, say,
the Hawaiian Islands, some scientists
have been reluctant to concede that
this phenomenon is capable of ex.
plaining the colonization of larger
areas.
Experiments with chance
dispersal
Neither random observations nor
data on the forms, attachment de.
vices, and viability of seeds, fruits,
and eggs offer as broad a picture as
one could wish. Hence some biologists
have attempted to test possibilities of
dispersal in such a way that quanti-
fiable and repeatable data can be
obtained. For example, experiments
have been carried ‘out on the dispersal
of seeds by shorebirds. Vlaming and
Proctor (1968) have shown that
shorebirds.confined in pens will eat
seeds, retaining them in the case of
the killdeer for as long as 120 hours.
Killdeer fly at speeds of 80 to 100 kph,
and could thus disperse seeds over a
range of more than 8,000 km in non-
stop flight. Shorebirds retain large
seeds longer than small seeds, so that
long-distance dispersal of seeds the
size of olive pits is not at all in conflict
with the observed biology of these
birds.
Although the transport of seeds over
distances as great as 8,000 km is rare,
we should remeimber that the rare
does occur. If such events were com-
mon, the flora and fauna of each cli-
matic’ zone would long since have
been homogenized the world over—
which is not the case. The fact that so
many: Hawaiian: plant and animal
species have evolved new character-
istics. distinguishing: them from
species in the source areas indicates
that most of them were introduced in
Hawaii only once, then never rein-
troduced by a'second natural event of
long-distance dispersal.
Williams and Williams (1978) have
used sources of evidence such as radar
to document the large numbers o!
individuals that participate in bird
migrations. At Palo Alto Marsh in
California as many as 11,700 migrat-
ing shorebirds have been counted on
a peak day; 4,000 to 5,000 individuals
is not. an unusual number (Jurek and
Leach 1971, 1972). Given these large
bird feathers
noteven: necessary,.
invoke ‘numerous
agents; only a:few
believable. It is
this rate, to
ies of birds as
ié8 of birds and
terrestrial mammals that live at
yund level near the seeds. However,
mammals obviously cannot tra-
yerse long oceanic distances, so the
feathers of migratory birds provide a
substitute animal surface where dis-
rsal to islands is involved. Any
plant with these adaptations that
grows near the nesting or foraging
grounds of shorebirds or marine birds
is a likely candidate for dispersal in
this fashion. Since most of these
nesting or foraging areas tend to be
located in lowlands rather than in wet
upland forest, externally transported
seeds and fruits in the Hawaiian flora
are mainly found at lower eleva-
tions.
Herbs of North and
South America
Long-distance dispersal must also
have operated in the case of a series of
herbs—mostly annuals—found in
Chile and California or nearby regions
(Fig. 3). Constance (1963) and others
have documented this series of plants
(more than 100 species are involved),
in which closely related spectes—or in
afew instances the same species—
occur in small pockets of Chile and
North America, with enormous dis-
tances in between. The great simi-
larity of the species in the two regions
can only mean that they have been
distributed recently—probably less
than 5 million years ago at most—
since as time elapses the herbs will
naturally evolve into quite different.
species or genera.
Not only must this disjunction be the
tesult of a rather recent dispersal; the
dispersal agent must be a good one.
Humans, who are notoriously good
dispersal agents, can be ruled out
here, because the species involved do
Rot follow the patterns of weeds or
other plants known to be carried
around the world by humans. Nor
does tectonic plate movement seem to
be a possible explanation: the pattern
's too recent, the distances too great,
and the crustal movements of the
farth have, in fact, worked contrary
‘othis distribution. North and South
America have been moving slowly
toward each other, not. diverging. The
fonnection of South America to
orth America via Panama has come
into existence only recently, and the
£ap of ocean between the two conti-
"ents in the Pliocene and earlier
“ould have made migration more
difficult.
However, at least seventeen species of
shorebirds and marine birds are
known to migrate over this route
every year. Of these, ten shorebird
species have been seen to feed on
seeds, berries, or other plant material
(Collins 1974). Two of the shorebird
species are often seen well inland, so
that distribution of inland as well as
coastal plants can be attributed to
birds. Moreover, during wetter peri-
ods of the Pleistocene, swampy, low-
lying areas attractive to shorebirds
were widespread in both Chile and
California. Plants now more re-
stricted in distribution probably grew
along the muddy shores of ponds,
swamps, or even inland seas such as.
the one then occupying the San Joa-
quin Valley in California.
Transport of seeds and fruits on or in
birds can account for virtually all of
the disjunctions of the Chile-Cali-
fornia herbs (Fig. 4). Only a single
instance, the morning glory Calyste-
gia soldanella, requires another ex-
planation. Although the seeds of this
plant float in seawater, it would have
been virtually impossible for them to
have floated from North to South
America or vice versa, because the
currents in the intervening area
would have swept floating seeds
westward into the equatorial Pacific.
However, the seeds could have drifted
from a Pacific locality to both Nerth
and South America; Calystegia
soldanella also occurs in Australia,
New Zealand, and Japan, as well as in
the Atlantic.
Figure 4. The presence in temperate South
America of 106 herb species either identical
with or closely related to North American
species has been attributed to the migratory
patterns of birds. Herb seeds or fruits with
various adaptations that enhance the possi-
bility of transport by birds are shown here.
Species with barbed or bristly seeds or fruits,
such as Cardionema ramosissima {(tup), ac-
count for 42.4% of the 106 cases of dispersal.
Scirpus nevadensis is a member of a group of
seeds or fruits consumed by birds as fodder and
later excreted in new locations; of the 106
species, 19.9% are believed to have traveled in
this fashion. Viscid seeds or fruits thal become
attached to birds’ feathers, here represented by
Carpobrotus inequilateris, account for an ad-
ditional 18.9%, while small seeds like Ortho-
carpus attenuatus, easily carried in mud or
sticky substances on birds’ feet or on their
feathers, account for 15.1%. Less important
methods of transport account for the remaining
3.7%. The fruits and seeds shown are respec-
lively about 4, 2, 1.4, and 1.6 mm in size. (Per-
centages are frum Carlquist, unpubl.)
1983
September-October
513
Vicariance biogeography
The nonreproducible nature of
chance dispersal events makes their
analysis by use of statistical methods
difficult, if not impossible. However,
some biogeographers who have felt
the need for rigorous precision have
invented a methodology known as
vicariance biogeography. Does this
method indeed solve problems of
distribution, or is it merely applicable
to cases where organisms move only
short distances over long periods of
time?
Vicariance biogeography assumes
that. patterns of distribution follow
geographical and climatic events such
as the breaking apart of land
masses—continental drift—and
major shifts of wet and dry, cold and
warm. If several groups of plants or
animals independently show the same
distribution patterns, the proposed
explanation is seen as more likely.
Replicate patterns have been appre-
ciated since the time of Hooker
(1860), who thought that the southern
continents must have once been in-
terconnected because of the many
organisms they share. However, vi-
cariance biogeography as a method-
ology may be said to have begun with
the writings of Croizat (1958, 1962).
A lucid account of the methods of vi-
cariance biogeography has been of-
fered by Wiley (1980), on whose ac-
count the folowing summary of pro-
cedure is based. .
The biogeographer first collects data
on the distribution of particular plant
and animal groups, and these data are
then plotted on maps.to produce
patterns of distribution called
“tracks.” Tracks that simulate or
replicate each other are now sought.
A common pattern of distribution for
several different groups would be
called a “generalized track.” This
type of activity, known as “track
synthesis,” is considered a reduction
of basic data.
Next, the biogeographer identifies
areas of endemism, or the restriction
of species to particular areas. By
grouping species restricted to these
areas, it is possible to analyze the
phylogenetic relationships of the
various species in an attempt to form
probable evolutionary trees for the
organisms. The investigator asks two
questions: What is the scheme of re-
lationships among species of the or-
fish ©
1 2 3 1 2 3
Figure 5. In this area cladogram, the phyloge-
netic relationships of two groups, fish and
moss, confined to three areas of endemism are
analyzed. After the ranges of the two groups
have heen determined (top), possible evolu-
tionary trees are formulated and compared
(middie). The area occupied by each species is
then substituted for the species name (bottom).
in this imaginary example, there is a complete
match between the phylogenetic tree of each
group and that for the areas in which they
‘occur, suggesting that common factors affected
the evolution and distribution of the two
groups. (From Wiley 1980.)
ganisms occupying the endemic
areas? Do the interrelationships of
the organisms reflect the geologic
histories of the areas?
The biogeographer then formulates
various possible evolutionary trees,
trying to find a match with the pat-
tern of areas inhabited. To do this,
the area in which a species is found is
substituted for the species name in
the evolutionary hypothesis. Such a
hypothesis is called an “area clado-
gram.” Figure 5 analyzes the phylo-
genetic relationships of two groups,
fish and moss, confined to three areas
of endemism. Patterns will be similar
to the extent that common or general
factors affected the evolution and
distribution of two or more groups of
organisms. [n order to find such sim-
ilarity, the biogeographer must be
able to separate unique factors found
in a single group from the common
factors present in the evolution of all
the groups considered.
Does this scheme work in practice?
Separating “unique” from “common”
factors might be a troubling task.
possibly involving arbitrary or intui-
tive assessments. However, propo-
nents of vicariance biogeography
point to the compelling nature of
replicated patterns, once these have
been located.
Even supposing that an objective or
less subjective method is used to lo-
cate unifying patterns, does this work
for all situations? Chance dispersa:
over time, widening the area occupiec
by a species or suddenly causing «
species to skip to new areas, woul
destroy identity of patterns. If one o
the species of moss in Figure 5 begin:
to colonize new areas by dispersal
whereas the fish do not, the general
ized track would at first be altered
and eventually might be obscured t«
the point where an original patter:
could no longer be discerned. A weed:
species would, in fact,.be expected ts
spread in just suth an irregula
fashion.
The majority of vicariance biogeo
graphers concede that long-distanc
dispersal exists, but they feel that i
creates at most a minor “noise leve}
superimposed on the basic pattern
dictated by geological history, such a
tectonic plate movements or the ris
ing of a mountain chain. They believ
that most species respond to geolog
ical and climatic changes similarly
and that species leave traces of thei
responses in their distribution pai
terns. The sum of similar response:
in the opinion of these scientist:
amounts to statements of statistic:
likelihood and attendant verificatio
or falsification. But is dispersal b
plants or animals to new areas so slo
or infrequent that basic patterns a)
not obscured? Can this “noise leve!
become so loud that no basic patter
other than long-distance dispers:
itself is evident?
Vicarianee biogeography seems i
come closest to explaining the situ
tion when an organism is capable -
dispersing only over very short di
tances, and therefore shifts to ne
ecological zones very rarely or e
tremely slowly. Freshwater fish n
adaptable to brackish water or se
water are such organisms. It is no a
ident that the most vocal proponent
{ vicariance biogeography, Rosen
1975), studies such fish. Those who
westigate organisms capable of
ing-distance dispersal and those who
wdy islands or other areas that were
robably populated by long-distance
ispersal have not adopted the
xethods of vicariance biogeo-
raphy.
‘ceanic islands and areas such as
wse colonized by the disjunctive
erbs we have described above are not
ie only places where long-distance
ispersal may prevail. Many groups
ft organisms characteristically excel
+ dispersal and are found in multiple
avironments. Algae, fungi, and pro-
woa are among groups with little
ademism: a species found in Europe
often also found in Australia or
outh Africa and at any suitable lo-
ation in between. Can these organ-
ams, or relatively dispersible groups
ich as flowering plants, ferns, in-
acts, and Jand snails, illustrate the
ovement of tectonic plates? Only a
:w families of flowering plants have
istributions that suggest continental
tift (Thorne 1972), indeed, only
Ider groups of flowering plants, be-
ause of their time of origin, could be
xpected to show such distribu-
ons. :
is also apparent that we may find
jues to the reason for a particular
istribution pattern by taking into
ecount factors other than phyloge-
etic trees and endemic areas. A good
xample of this is seen in the plants of
ye African volcanoes, which are
sainly related to European plants,
nd those of the Andean volcanoes,
hich show a strong affinity with
tants of temperate North America.
‘oth of these high-altitude environ-
tents require plants adapted Lo cold.
‘ropical plants are apparently unable
» make this adaptation readily be-
ause of the long period of time nec-
ssary for the numerous changes re-
uired, whereas Europe and temper-
te North America are rich in plants
dapted to cold. The plants of these
¥o areas, moreover, tend to have
volved good dispersal mechanisms
ecause of the mountainous, discon-
nuous terrain in which they live.
‘he appearance of these plants in the
ew African and Andean locations
iggests that these mechanisms are
tpable of operating over very long
istances as well as over fairly short
nes.
A new synthesis?
Most modern investigators combine
tectonic plate movement with chance
dispersal in some fashion to explain
present-day distributions {e.g., Raven
and Axelrod 1974). The problem is
not whether chance dispersal occurs,
but the extent to which the patterns
we see have been caused by it, and
how, if such patterns occur fre-
quently, we may analyze them.
One possibility involving both tec-
tonic plate movement and long-dis-
tance dispersal should be mentioned:
Was there interchange among conti-
nents while they were moving apart?
‘Too often, it seems to have been as-
sumed that interchange essentially
ceased when the rifts were initiated.
However, suitable habitats for a given
organism are often distributed in a
scattered fashion, and there is not
much difference between habitats
scattered on a single continent and
those scattered on two continents still
close to each other. Most families of
flowering plants have patterns in
which dispersal among continents
after the split has been a major factor.
if such a small target area as the Ha-
waiian Islands has received so many
successful colonists in such a short
time, could not a continent receive
many more? Unfortunately, the tools
available to analyze and identify this
phenomenon are poor, and the results
are not in any range of statistical sig-
nificance.
The desire of vicariance biogeogra-
phers for methods that can operate
with precision is understandable,
since biogeography is a field with lit-
tle unanimity in interpretations. The
refinements of geology are now mak-
ing interpretations a little more cer-
tain: the hypothetical land bridges
once erected or obliterated with
abandon by some biogeographers
cannot be entertained in the hard
light of newer geological evidence.
However, our knowledge of any
unique event in the history of life is
still limited, especially if subsequent
complexities destroy the pattern
created by the original event.
For groups like the primary-division
freshwater fish, the use of the meth-
ods of vicariance biogeography seems
defensible, since—at least in the
clearest cases—patterns are not
erased or destroyed. On the other
hand, patterns which circumstances
tell us have probably resulted from
long-distance dispersal, such as the
distributions of Hawaiian organisms
and of the herbs in disjunctive loca-
tions, seem unlikely to yield to the
methods of the vicariance biogeo-
graphers—who indeed have not en-
tered those areas. The biogeographer
who deals with patterns probably
created by long-distance dispersal
must use evidence that is circum-
stantial, indirect, and subjective, and
therefore vulnerable. However, if that
kind of evidence leads to plausible
answers, we cannot afford to rule it
out of court. Long-distance dispersal,
although annoyingly difficult to study
or take into account, appears to be a
persistent theme which will not go
away, and it is to be hoped that future
biogeographers will find a way to in-
corporate it with skill.
References
Carlquist, S. 1974. Island Biology. Columbia
Univ. Press.
Unpubl. intercontinental dispersal.
Collins, R. 1974. Mechanisms for long-distance
dispersal in some amphitropically disjunct.
plant species. M.A. thesis, Claremont
Graduate School.
Constance, L. 1963. Amphitropical relation-
ships in the herbaceous flora of the Pacitic
Coast of North and South America: A sym-
posium. Introduction and historical review.
Quart. Rev. Biol. 38:109-16.
Croizat, L. 1958. Panbiogeogrephy. Caracas:
published by the author.
1962. Space, Time, and Form: The
Biological Synthesis. Caracas: published by
the author.
Dalrymple, G. B., E. A. Silver, and E. D. Jack-
gon. 1973. Origin of the Hawaiian Istands.
Am, Sci. 61:294-308.
Dammermann, K. W. 1948. The fauna of
Krakatau 1883-1933. Verh. Kon. Ned. Akad.
Wetensch. Afd. Nauurk., 2d ser., 44:1
594.
Docters van Leeuwen, W. M. 1936. Krakatau,
1883-1933. Ann. dard. Bot. Buitenzorg
4647:2ii, 506.
Einarsson, E. 1967. The colonization of Surt-
sey, the new valeanic island, by vascular
plants. Aquilo, Ser. Bot. 6:172-82.
Gressitt, J. L., and S. Nakata. 1958. Trapping
of air-borne insects on ships in the Pacific.
Proc. Hawaiian Entomol. Soc. 16:363-65.
Gressitt, J. L., J. Sedlacek, K. A. Wise, and C.
M. Yoshimoto. 1961. A high-speed airplane
trap for air-borne organisms. Pacific Insects
5:549.55.
Guppy, H. B. 1906. Observations of a Natu-
ralist in the Pacific between 1896 and 1899.
Vol. 2: Plont-Dispersat. London: Mac-
milan,
Hooker, J. D. 1860. On the origin and distri-
bution of species. Introductory essay on the
flora of Tasmania. Amer. J. Sci. and Arts,
Ser. 2, 29:1-25, 305-26.
1081 Gentember-Ortoher 515