Experimental Zoogeography of Islands: Defaunation and Monitoring Techniques
Edward O. Wilson; Daniel S. Simberloff
Ecology, Vol. 50, No. 2. (Mar., 1969), pp. 267-278.
Stable URL:
http://links.jstor.org/sici?sici=0012-9658%28196903%2950%3A2%3C267%3AEZOIDA%3E2.0.CO%3B2-J
Ecology is currently published by Ecological Society of America.
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/esa.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 support@jstor.org.
http://www.jstor.org
Thu Mar 20 12:57:34 2008
Early Spring 1969 EXPERIMENTAL DEFAUNATION OF ISLANDS 267
some survive in the desert even during years of der the canopy of xerophytes and in open areas. Comdrought,
without any artificial irrigation with only n'Onwealth Rev. 43: 219-234.
. 19646. Transpiration of Aleppo pine as a func-
80 mm of rainfall, and in the subtropical zone with tion of environment. Ecology 45: 868-873.
200-300 mm of rain. -. 1966. Attraction of atmospheric moisture by
These results and those obtained earlier (Gindel woody xerophytes in arid climates. Commonwealth
1%8b, 1%9) indicate that further study is needed Forest. Rev. 45: 297-321.
to clarify to what extent augmented stornatal -. 1967a. The correlation between cambium activity
and transpiration rate in Aleppo pine. 14th I U F R 0
density and the presence of stomata on the upper Congress 4: 18S207.
epidermis in woody xerophytes grown in the -. 19676. The relationship between transpiration
driest climate and poorest soil serve to increase and six ecological factors. Oecol. Plant. 8: 227-239.
the ability of the tree to absorb atmospheric mois- -- . 1967c. Afforestation of the desert. "Ariel,"
ture, when moisture tension in the soil and within Jerusalem 19: 13-24.
. 1968a. Some ecophysiological properties of three
the plant reach their highest values. tree xerophytes grown in the desert. Oecol. Plant. 3:
49-67.
19686. Dynamic modifications in alfalfa leaves
This work was accomplished with the financial assis- growing in subtropical conditions. Physiol. Plant. 21:
tance of the J.N.F. I express my thanks to Mr. Joseph 1287-1295.
Weitz, pioneer of the 50 years of Israeli afforestation, and 1969. Stomata constellation in the leaves of cotto
Prof. A. H. Halevy for his criticism of the manuscript. ton, alfalfa, maize and wheat plants as a function of
soil moisture and environment. Physiol. Plant. (in
press).
Fahn, A. 1964. Some anatomical adaptations of des- Kohl, F. G. 1886. Transpiration der Pflanzen und ihre
ert plants. Phytomorphology 14: 93-102. Auswirkung auf die Ausbildung pflanzlicher Gewebc.
Gindel, 1. 1952. Some anatomical features of the in- H. Bruhn, Braunsweig. 34 p.
digenous woody vegetation in Israel. Res. Coun. Sorauer, H. 1873. Einfluss der Wasserzufuhr auf die
Israel 2: 1-16. Ausbildung der Gerstenpflanzen. Bot. Z. 31: 145-159.
-. 1957. Acclimatization of exotic woody plants Stocker, 0. 1960. Physiological and morphological
in Israel : The theory of phytoplasticity. Materiae changes in plants due to water deficiency. Arid Zone
Vegetabiles 11: 81-101. Research, UNESCO, 15: 63-104.
1961. The water balance in xerophytes. In- Zelitch, I. 1961. Biochemical control of 'stomata1
tern. Union For. Res. Org. 13 Congress 21: 2-4. opening in leaves. Proc. Nat. Acad. Sci. Wash. 47:
. 1964a. Seasonal flucuations in soil moisture un- 1423-83.
EXPERIMENTAL, ZOOGEOGRAPHY O F ISLANDS : DEFAUNATION 

AND MONITORING TECHNIQUES 

EDWARD0.WILSONAND DANIELS. SIMBERLOFF~
The Biological Laboratories, Harvard University, Cambridge, Massachz~setts02138
(Accepted for publication December 16, 1968)
Abstract. In order to facilitate experiments on colonization, a technique was developed
that permits the removal of the faunas of very small islands. The islands are covered by a
tent and fumigated with methyl bromide at concentrations that are lethal to arthropods but
not to the plants.
Seven islands in Florida Bay, of varying distance and direction from immigrant sources,
were censused exhaustively. The small size (diameter 11-18 m) and ecological simplicity
of these islands, which consist solely of red mangrove trees (Rhizophora mangle) with no
supratidal ground, allowed the location and identification of all resident species. The terrestrial
fauna of these islands is composed almost exclusively of arboreal arthropods, with 20-50
species usually present at any given moment. Surveys of these taxa throughout the Florida
Keys, with emphasis on the inhabitants of mangrove forests, were made during 1967 in order
to estimate the size and composition of the "species pool."
The seven experimetltal islands were defaunated in late 1966 and early 1%7, and the colonists
were monitored for 17-20 man-hours every 18 days. Precautions were taken to avoid artificial
introductions during the monitoring periods.
INTRODUCTION Arthur and Wilson (1963, 1967), who postulated
The first attempt to formulate a quantitative an equilibrium number of species on an island
theory which would unite an ever-increasing mass determined by the intersection of immigration and
of insular biogeographic data was made by Mac- extinction curves drawn as a function of the num'Present
address: Department of Biological Science, ber of present. In addition to
Florida State University, Tallahassee, Florida 32306 speculation on the forms of these curves and ef-
268 EDWARD O. WILSON AND DANIEL S. SIMBERLOFF Ecology, Vol. 50, No. 2
fects of varying both island area and distance from
the source, they derived the equation
mean S
to,go = 1.15 --- (1)X
where Si =the hypothetical equilibrium number
of species on an island
mean$ =the mean of 3 6 for a number of very
similar islands i
X =the extinction rate (or turnover rate,
since an equilibrium requires that extinction
= immigration) in number of
species per time at equilibrium
=time required for number of species
present to increase from 0 to 90% of 3.
Later the same authors elaborated upon the general
shapes of immigration and extinction curves,
discussed colonization rates and curves, and predicted
the distribution of survival times of species
that succeed in colonizing an island (MacArthur
and Wilson 1967).
Data to test these hypotheses are scarce. Most
biogeographic information, when combined with
information furnished by the scant fossil record,
can at best lead to descriptions of broad patterns
of distribution and zoogeographic regions and to
hypotheses about the essentially long-term processes
responsible for them (e.g., Darlington
1957). Such theories, though occasionally suggestive
for specific groups, fail to explain the
existing taxonon~icdistributions and the underlying
colonization nlechanisms for most islands.
They neglect the short-term, even daily events that
largely determine the parameters of colonization
on many islands. Such events become decisively
more important than evolution as distance of island
from source area decreases. In particular,
the classical zoogeographic methods do not provide
a test for the existence of an equilibrium
nunlber of species or for the accuracy of equation
(l ) .
The literature on biotic dispersal, both anecdotal
and systematic, provides information on the
relative importance of active and passive transport
and a wealth of data on the specific agents of passive
transport (e.g., Wolfenbarger 1946). Certain
cases of overseas dispersal have even been
traced to specific meteorological events (French
1964). For no island, however, is the time course
of colonization by even a large fraction of the
inhabitants known. The large number of records
of long distance dispersal implies that immigration
rates to islands are probably high, especially for
organisms capable of passive transport by wind.
But the shape and determinants of the immigration
curve cannot be deduced. At best a lower
limit might be given.
In the recent studies of Surtsey, a new volcanic
island near Iceland, there has been a deliberate
attempt to document the colonization process from
FIG.1. The southern tip of Florida and the Florida Keys. The rectangles enclose the experimental areas
shown in detail in Figures 3-5.
Early Spring 1969 EXPERIMENTAL DEFAUNATION OF ISLANDS 269
consist entirely of one to several red mangrove
trees (Rhizoplzora mangle), with rarely a small
black mangrove bush (Avicennia nitida) . Occasionally,
in areas of weak tide, the trees are surrounded
by small areas of supratidal mud and
sand. Such land is only intermittently supratidal,
since it is completely flooded during prolonged
winds of 10 knots or more. Individual islands
may differ in several minor respects such as the
number of large trunks and amount of dead bark
they contain, but on the whole they are remarkably
similar to each other in physical appearance.
A fruitful defamation experiment with these
islands requires the following conditions :
I
i) that there be enough of the islands for repliI
cation and variation in distance to the nearest
FIG.2. Upper: Island El, the second smallest island in source area;
the experimental series. Lower: Island E9, the largest i;) that the animal diversity be sufficient to
island in the experimental series; note also the presence
of supratidal mud. allow statistical treatment, and the organisms be
physically large enough to insure recording inconthe
birth of the island (Fridrikssoll 1967, Her- s~ic"Ous as as Ones;
mannssoll 1967, Lindroth et 1967). But the iii) that the extremely small size of the islands
infrequent natural occurrence of such events tends compensates for their relative nearness to source
to render quantitative biogeographic theory the areas and therefore produces a distance effect.
product of abstract speculation and of enlarge- During an exploratory trip to the Florida Keys
ment and sophistication of untestahle hypotheses. ill 1966, we found that conveniently accessible
What is needed, clearly, is a nlethod either of Pro- sinall islands were numerous at most distances up
ducing new islands similar to natural ones or of to 200 m from the nearest large islands. Beyond
sterilizing preexisting islands. 200 m they were rare, although somewhat larger
islands (diameter 50 m and greater) were plentiTHEEXPERIMENTALISLANDS
ful. A very few small islands suitable for our purAlong
the Overseas Highway of southern Flor- poses were located 0.2-1.4 km from the nearest
ida (U. S. Route 1) there exist a vast array of source area.
small, approximately circular islands situated in A set of such islands were chosen for experishallow
bay water (Fig. 1). Together, they con- mentation and their faunas carefully surveyed.
stitute a potential natural "laboratory" for experi- Since the breeding fauna of islands this size conmental
biogeography. The islands with diameter sists almost entirely of species of insects and spi-
10-20 m (Fig. 2) are 5-10 m tall and usually ders, we made a reference collection of land arTABLE1.
Parameters of experimental islands
.Control island
bWSteppingStone" (SI) 572 m away.
Series 1
Eeries 2
Island
name
El
E3
ST2
E2
E6"
--
E7
-
E8
E9
EIOB
Diameter
(m)
-
11
12
11
12
15%
-
25
18
18
1la
Distance from
nearest source
(m)
533
172
154
2
73"
15
- -
1,18gb
379
37"
Initial no.
arthropod
spec~es
25
31
29
43
30"
29
29
41
20R
Location
off Squirrel Key
in Rattlesnake Lumps
off Saddlebunch Key
off Snipe Keys
in Johnston Key Mangroves
near Manatee Creek
between Calusa Keys and
Bob Allen Keys
between Calusa Keys and
Bob Allen Keys
near Bottle Key
270 EDWARD O. WILSON AND DANIEL S. SIMBERLOFF Ecology, Vol. 50, No. 2
thropods of the Florida Keys with the assistance
of Robert Silberglied. This collection enabled us
to identify the arthropods on the experimental
islands, and it provided a measure of the size ana
composition of the species pool of the presumed
source area, the entire Florida Keys.
Animal diversity on the small mangrove islands
proved to be adequate for statistical purposes.
About 75 insect species (of an estimated 500 that
inhabit mangrove swamps and an estimated total
of 4,000 that inhabit all the Keys) commonly live
on these small islands. There are also 15 species
of spiders (of a total Keys complement of perhaps
125 species) and a few scorpions, pseudoscorpions,
centipedes, millipedes, and arboreal isopods.
At any given moment 20-40 species of insects and
2-10 species of spiders exist on each of the islands.
Many birds visit the islands either to roost or
to forage, but very few nest there. The only birds
that bred on our islands during the course of the
experiment (1966-68) were one or two pairs each
of the Green Heron, White-crowned Pigeon, and
Gray Kingbird. Snakes (mostly water snakes of
the genus Natrix) occasionally swim to small
mangrove islands, and raccoons visit islands located
on shallow mud banks. These vertebrates
were not included in the censuses.
On the basis of their distance and direction
from the nearest source area, their size, and their
accessibility, the islands listed in Table 1 were
chosen for defaunation. Two islands (E6 and
ElO), one each in the vicinity of the two groups
of experimental islands, were selected to serve
as control islands; and they were censused at the
same time as the experimental islands. In order
to ascertain whether any long-range change had
occurred in the general fauna of small mangrove
islands in Florida Bay during the course of the
-
Meters
I000 0 1000 2000 3000 4000 5000
FIG.3. Map showing the location of the experimental islands of series 1.
271
0""

Early Spring 1969 EXPERIMENTAL DEFA
ColusaKgs FURIDA
'- E8
Fbxtrot Keys 0
O.
Bob Allen Keys 0
0
- -0
5000 Meters
FIG. 4. Map showing the location of the experimental
islands of series 2.
I - - - - - I
0 1/2 I 2 Miles
I------ I
0 1000 2000 Meters
FIG. 5. Map showing the location of the original test
island E7.
experiment, a second census of the control islands
was made at the close of the experiment.
The two series of experimental islands form two
widely separated groups. Series 1 lies within the
Great White Heron National Wildlife Refuge,
in a region north of Sugarloaf Key (Fig. 3). Series
2 is in the Everglades National Park near the
Calusa Keys north of Islamorada (Fig. 4). E7,
which was chosen primarily to test a method of
defaunation, is located in Barnes Sound south of
the Glades Canal (Fig. 5).
The islands were also selected to show as much
variation as possible in the degree of isolation
from the immigrant sources. The coefficient of
correlation for island series 1 between "distance
from nearest source" and "number of arthropod
species'' is -0.83, indicating the operation in the
UNATION OF ISLANDS
original faunas of the distance effect as defined
by MacArthur and Wilson (1%7), Qualitative
differences in faunas of increasingly distant islands
were even more striking than quantitative ones.
For many groups-especially ants and spidersthe
distance effect is so regular that one can guess
not only the species number but also the approximate
species composition. For example, centipedes
and pseudoscorpions were found almost
exclusively on islands of less than 200 m distance
from the nearest large island.
Islands of this minute size can be remarkably
long lived. All Rhizophora mangle trees, and
particularly those on low, small islands, have
numerous aerial roots : both brace roots growing
out from the main stem and prop roots growing
down from branches. The latter can emerge from
as high as 4 rn above the water ( ~ a ~ i eand
Muzik 1954) and in time can become as thick
as the main stem. Rhizophora trees reproduce
viviparously, their seeds growing on the tree until
they become large, long, pointed seedlings. When
they drop, frequently from the upper canopy, some
evidently plant themselves where they fall (Davis
1940). Mangrove swamps are often quite thick,
and on small islands there may be several trees.
Stems, thickened prop roots, and brace roots in-
,Blake
-l K~lorneter
FIG.6. Map showing El and surrounding region as it
appeared in 1851-57.
FIG.7. The same region shown in Figure 6 as it appeared
in 1964.
272 EDWARD O.WILSON AND DANIEL S. SIMBERLOFF Ecology, Vol. 50, No. 2
termingle and reproduce themselves so that it is
impossible to distinguish how many trees are
present, which are which, which is the initial one
(even if it is still present), and which are the main
stems.
Maps of the Florida Keys, especially of the
Sugarloaf Key-Key West area, have been remarkably
detailed and accurate with respect to small
mangrove islands for at least a century. These
tell us that E7 and all the islands in series 1 have
existed at least 25 years. The other experimental
islands were not mapped in sufficient detail to
determine longevity. The size and shape of individual
islands may change, even drastically, but
in general the islands survive the climatic catastrophes
to which they are repeatedly subjected.
For example, Figure 6 shows E l (Blake's Key)
and surrounding areas as it appeared in 1851-57,
while Figure 7 shows the same island region on
the Coast and Geodetic Survey map of 1964. It
therefore seems reasonable to assume that the
faunas found initially on the experimental islands
were the product of many years' colonization. Our
experiments have subsequently shown that the
minimum ages of the islands greatly exceed the
time required to reach species equilibrium from a
start of zero species (Simberloff and Wilson
1969).
Our first attempt to defaunate islands made use
of an insecticide spray of short-lived residual effect.
On July 9-10, 1966, two experimental
islands ( E l and E2) were sprayed until dripping
with 60 g parathion, 240 g diazinon and 180 g
Pylac sticking agent per 100 1fresh water. When
the islands were examined 1 day later, all the
surface and bark fauna and most borers were
dead. However, the following live animals were
found in thin hollow twigs. E l : one adult wasp,
Scleroderma mcrogaster (Bethylidae) ; workers
and larvae of Cremtogmter ashmeadi (Formicidae).
E2: two larvae of ?Stylolef-tus biustus.
(Cerambycidae) ; two colonies of Paracryptocerus
varians (Formicidae) ; and one colony of Tapinoma
littorale (Formicidae) . A more tE6rough
examination 9 days after spraying produced the
following live and apparently healthy insects. El :
two larvae of ?Styloleptus biustus (Cerambycidae).
E2: one larvae of ?Styloleptus biustus
(Cerambycidae) ;one lepidopteran larva ; and two
colonies of Xenomyrmex jloridanus (Formicidae) .
The large populations of the ant colonies and
advanced development of the beetle larvae precluded
these insects having immigrated during the
9 days following spraying.
The results on El and E2 showed that a spray
cannot be expected to penetrate all the hollow
twigs of a Rhizophora island. Some twig dwellers
will probably survive, particularly if they inhabit
narrow twigs (Tapinowza and Xenowzyrwzex) or
else pack their tunnels tightly with powdery excreta
(Styloleptus). Had we attempted to kill
these survivors by using a'spray with a long-lived
or more powerful residue, we would have postponed
the beginning of the colonization curve,
since immigrants may be killed by the residue.
Because we had to be certain that all colonists,
or at least all but those belonging to a small number
of known taxa, would be destroyed immediately,
we abandoned the spray technique.
We turned next to the more difficult alternative
technique of fumigation. The method has the
obvious advantages that the residual effect of a
gas is negligible, and even the inhabitants of hollow
twigs are contacted. This is the standard
method used by professional exterminators on
termites and other wood-boring insect pests.
The fumigation of living plants is a relatively
uncharted area. The biochemical effects of fumigants
on plants are poorly known, damage usually
being described in terms of the physical appearance
of the plant shortly after fumigation
(Page and Lubatti 1963). To our knowledge, the
only prior fumigation of red mangrove trees was
by B. P. Stewart (pers. comm.), who used methyl
bromide on small plants. With 32 kg/1000 m3
for 2 hr at 32°C and 40 kg/1000 m3 for 2 hr at
27°C he found that all insects apparently were
killed, although some did not die until 48 hr after
fumigation. Mortality of cockroach eggs was
unrecorded in Stewart's study. Damage to the
plant was limited primarily to young leaves and
twigs, a result that is in accord with the general
finding of Page and Lubatti (1963) that the growing
stages of both plants and insects are the most
susceptible to fumigation.
In our preliminary trials we were limited by
the fact that fumigants highly soluble in water,
such as hydrogen cyanide, could not be used.
These substances would be expected either to
leave the fumigation tent through the water or
to concentrate to an unknown extent in the water
under the island. Either effect would greatly
complicate the maintenance and measurement of
gas concentration. Consequently, the following
four relatively insoluble fumigants were tried :
Pestmaster Soil Fumigant-1 (methyl bromide
98% wt, chloropicrin 2% wt) ; Acritet 34-66
(acrylonitrile 34% vol, carbon tetrachloride 66%
vol) ; Dow Ethylene Oxide (ethylene oxide
100%) ; and Vikane (sulfuryl fluoride 95% wt,
inert ingredients 5% wt). Field tests were performed
on small Rhizophora trees in Matheson
273Early Spring 1969 E X P E R I M E N T A L DEFA.UNATION OF ISLANDS
Hammock Park, Coral Gables, by National Exterminators
of Miami, Florida, under the direction
of Steve Tendrich in consultation with the
authors. Additional tests on the mortality of mangrove
colonists were performed at National Exterminators'
laboratory in Miami.
With the temperature outside the fumigation
tent remaining at 24'-29"C, we determined which
duration-concentration intensities caused no more
than bearable damage to the tree yet were lethal
to all the mangrove arthropods. Acritet, Vikane,
and ethylene oxide caused extensive, apparently
irreversible damage to red mangrove and were
rejected. However, methyl bromide at 22 kg/
1000 m3 for 2 hr caused browning of only about
10% of the leaves 1 week after treatment, with
no visible later effects. This duration and concentration
killed all the mangrove inhabitants (including
such resistant forms as roach eggs and
lepidopteran pupae) with one possible exception.
When data from all tests at this duration-concentration
are lumped together, only 95% of cerambycids
and 80% of weevils were dead when dug
out of their burrows 1-10 days after fumigation.
The remainder all died within 2 days of removal.
It is tempting to ascribe the death of these few
larvae to a delayed effect of the fumigant. However,
a control experiment consisting of the removal
of five unfumigated cerambycid larvae
(Styloleptzts biustus) from their respective burrows
caused death within 5 days without any
additional treatment. Monro and Delisle (1943)
report that many insects are active soon after
exposure to methyl bromide but succumb in time
(some lepidopterous larvae surviving for 2
months), and that this delayed effect is especially
common in fumigations at the low concentrations
used in our own field tests. This peculiarity is
confirmed by Chisholm (1952) for Japanese beetle
larvae and other insects. Of at least equal importance,
the results from our experimental islands
(Simberloff and Wilson 1 x 9 ) strongly suggest
that the cerambycids were eliminated by the fumigation
treatment. W e therefore conclude that even
the deep-boring beetle larvae were very probably
killed by the treatment.
The overt physical damage to the trees consisted
of a browning of leaves and shoots and
occasional heavy oozing of sap. At 61 kg/1000
m3 the leaf browning was manifest upon tent removal,
but at 35 kg/1000 m3 and less there was
little or no immediate browning. The effect became
evident at about 24 hr and worsened in terms
of fraction of leaves browned through 72 hr, by
which time up to 90% of the leaves were brown
and after which no further browning occurred.
Browning did not occur randomly throughout the
canopy. Instead, whole sections of the tree became
completely brown while others remained
almost unscathed. The upper canopy was always
more severely browned, but there seemed to be
no other consistent damage pattern.
Abscission of the dead leaves began within a
week and, often aided by wind, was usually complete
within 8 weeks. At this time damaged trees
looked superficially dead except for remaining
green sections. Closer examination showed that
the damaged trees could be conveniently divided
into three classes, as follows:
For those trees where browning of leaves and
shoots appeared complete by 72 h r the damage
was irreversible. Within a few months bark began
to peel and twigs with green wood could not be
found. A year later there was no evidence of life.
Island E8 is in this class. In the second category,
which includes only E7, a large majority of the
leaves and shoots were browned and died, while
a small section remained green. In the third
category, the trees were less severely damaged
(60% or fewer browned leaves). Green sections
remained unchanged, while leaf abscission in most
of the damaged sections was accompanied by the
sprouting of new shoots and leaves from almost
all branches. Within 2 months the trees appeared
normal except for occasional leafless in
the extreme upper canopy. Island E l , which 2
days after fumigation appeared to have about 50%
brown leaves. recovered after 1 month to the extent
that it did not appear to have incurred more
damage than similar trees do during heavy storms,
and 2 months after defaunation seemed quite normal
except for a small bare patch in the upper
canopy. At no time did its bark crack or peel.
In most damaged trees of the first two classeswhere
either the whole tree or all but a small section
was killed-there was copious oozing of sap
for up to a week after fumigation. Such oozing
was never observed in untreated trees, nor in those
fumigated trees which showed good leaf replace-
ment.
The action of methyl bromide on living plants
has been studied rather extensively by several
authors. Mainwaring ( 1 x 1 ) examined the physiological
responses of several plants to methyl
bromide, and his numerous observations imply
that its major effect is auxin inhibition. The
damage we observed seemed too drastic and certainly
too immediate to be attributed solely to
auxin inhibition, although the presence of such a
biochemical action cannot be ruled out.
Whatever the effect of methyl bromide on
Rhizophora mangle, it apparently had the expected
Qlo of 2-3. Our initial tests in Matheson Hammock
were conducted on small plants well shaded
274 EDWARD O.WILSON AND DANIEL S. SIMBERLOFF Ecology, Vol. 50, No. 2
by large trees, and the temperature inside the TABLE2. Arthropod species found on E7 just prior to
fumigation tent remained within 3" of the air tem- and
perature (24"-29" C). The experimental islands
could not be shaded, and even on overcast days INSECTS
Embioptera: gen. sp.'
the temperature inside the dark tent was at least Orthoptera: Latiblattelh n. sp.
32°C. b n sunny days the inner temperature was
so high that the presence of the covering alone
caused browning of leaves. It seems probable
that the generally more severe damage to the upper
canopy throughout this experiment was caused by
higher temperature, especially since methyl bromide
gas has a density (3.974 g/liter at 20°C)
considerably greater than that of air.
Air humidity is much less important than temperature
in its effect on fumigation damage to
plants (Page and Lubatti 1963) ; our experiments
were allnost all conducted within a narrow humidity
range (65%-90%) and the effects of varying
this parameter were not studied.
The first two experimental islands (E7 and E 1)
were fumigated during daytime and damaged. In
order to reduce heat damage we therefore fumigated
all other islands at night. The concentration
was increased from 22 kg/1000 m3 to 28-30 kg/
1000 1n3wit11 no observable additional damage.
The chemical means for defamation now having
been chosen and tested, we were left with an
overwhelming physical problem of how to fumigate
an entire island. Assuming that we could
somehow get the fumigation tent (a 25-n~by 25-m
piece of plastic-impregnated canvas with addiFIG.8.
Upper: The scaffolding constructed around E7,
complete except for the top walkway. Lozver: the fumigation
tent over E7.
Coleoptera:
Psocoptera:
Hemiptera:
Lepidoptera:
Hymenoptera:
ARACHNIDS
Araneae:
Acarina:
OTHER
Chilopoda:
Isopoda:
Cycloptilum sp.
Tafaliscalurida"
Pseudoacdh sp.b
Leptostylus sp.
Styloleptus 6iustusb
Tricorynus sp.
Archipso~lspanama
Pseudococcus sp.
Nemapogon sp."
Casinaria tezana
Camponotus$oridanuss
Camponolus planatu.9
Paracryplocerus variad
Pseudomyrmez ebngatw
Xenomyrmez $oridaruk
Ariadna arthuri
Eustala sp.
Leuuauge vcnustaa
Galumna sp.
Orphnaeus brasilianusa
Rhyscotus sp."
discovered dead after fumigation
"discovered mostlydead after fumigation (see text)
tional side pieces for larger islands) onto an island,
the tree could not support the weight without
major limb breakage, since even our special halfweight
tent weighed 275 kg. It was evident that
some sort of superstructure was needed, one which
would surround and rise over the island to remove
most or all of the tent weight from the tree. Constructing
a permanent framework and lowering
it over the islands by helicopter was too expensive.
An alternative method employing a temporary
full framework constructed on the site was used
on October 10, 1966 to defaunate E7. The procedure
was as follows (see also Fig. S) : First,
planks to prevent the framework from sinking
into the mud were placed underwater in a square
around the islands and held manually until the
first tier of scaffolding (which was sufficiently
heavy to keep them from floating) was placed on
top. Then the remaining tiers were added, one
side at a time, until the island was surrounded by
a cube of scaffolding. Fig. 8 (upper) shows the
framework complete except for the upper tier of
scaffolding in the right rear corner and the square
walkway of planks around the very top of the
structure.
The tent was then raised by block and tackle to
the top of the framework and lowered over the
island. Methyl bromide was introduced from
45 kg cylinders through the tent wall at opposite
sides of the island, 2.5 m above the water, and
into a metal tub mounted with an electric fan to
Early Spring 1969 EXPERIMENTAL DEFA.SINATION OF ISLANDS 275
disperse the gas. Chloropicrin (tear gas), added
to the odorless methyl bromide as a safety measure,
was allowed to drip into burlap sheets at the
bottom of the tubs. Concentration of the methyl
bromide was measured frequently by a Gow-Mac
thermal conductivity meter with a self-contained
pump attached to a testing station 2.5 m above
the water and well away from the two shooting
stations.
The methyl bromide concentration was gradually
built up from 0 to 22 kg/1000 m3 over 30 min,
then kept at 22-25 kg/1000 m3 for 2 hr. Outside
air temperature during this time remained steady
at about 27°C. At the end of the fumigation the
gas was released through seam openings for 45
min, and the tent was then removed.
Immediately following the dissipation of the
gas, E7 was explored for a total of 6 man-hours.
The results generally confirmed the evidence from
the Matheson Hammock tests. Dead individuals
representing 11 of the 23 species observed before
fumigation were found, frequently in large numbers
and in all life stages. The details are given
in Table 2. The only live animals encountered
were larvae of two deep-boring beetles. These
formed only a small fraction of the original population,
as shown by the following recovery data:
of 59 larvae of Styloleptus biustus (Cerambycidae),
2 were alive and 57 dead; of 10 larvae of
Pseudoacalles sp. (Curculionidae), 2 were alive
and 8 dead. One of the two live cerambycids was
extremely sluggish and died an hour after removal.
The other cerambycid and the two weevils seemed
healthy, though one of the weevils became torpid
and died 9 hr after removal.
The tree was severely damaged as described
previously. Within 2 months 85% of the foliage
was dead, although the single living section has
remained green until at least May 1%8. Thus
the procedure worked essentially as expected, and
this aspect of the fumigation was left unchanged
in subsequent defaunations, except that the work
was henceforth done at night in all but one instance
( E l ). The cubical scaffolding framework
had disadvantages, chief among them being the
large crew necessary to erect it (at least five men)
and the number of pieces and weight of the equipment.
This necessitated numerous boat trips between
E7 and our staging area near Manatee
Creek, since the shallow water along the route precluded
tne use of large craft. This problem was
greatly aggravated on the other islands, which are
located considerably further from a convenient
staging area. \Ve therefore set out to devise a
tower to be located in the center of the island and
rising high enough so that most of the tent weight
rested on it. This device proved to be more vulFIG.9.
Upper: the fumigation tent being opened over
E 2 from a tower support. Middle: E2 partially covered.
Lower: E9 covered by the fumigation tent on a tower
support.
nerable to high witids, but it has the great advantage
of requiring less equipment and a smaller
crew to assemble it. National Exterminators
contracted a steeplejack to perfect the tower
method. After a month of experimenting 011 islands
near Key Largo, the following procedure
was adopted and used for the remaining islands.
A collapsible 10-m triangular steel tower was
erected in the center of the island, either propped
on a large root or forced deep into the ground.
The steeplejack then climbed to the top and threw
out three guy wires which were fastened down
276 EDWARD O. WILSON AND DANIEL S. SIMBERLOFF Ecology, Vol. 50, No. 2
TABLE3. Extent of damage to foliage and subsequent recovery on experimental islands
I I I I
Island Fumigation Time Extent of
damage
(burnt leaves)
- --
Damage location
day 50% general
I
Recovery of foliage
complete in 2 months,
except for extreme
upper canopy
oomplete in 2 months
oomplete in 2 months
none
none
little in upper canopy
complete elsewhere in
2 months
complete in 2 months
I
night
night
day
night
night
night
with mudscrews or stakes, depending on the substrate.
Using a block and tackle, one man on the
tower and one on the island lifted the rolled tent
until it formed a triangle over the island. The
tent was then carefully unrolled down the guide
wires until the island was covered (Fig. 9). It
was inlpossible to unfurl the tent from the tower
in winds greater than 8 knots. Sandbags and
stakes were used to keep the tent edges submerged.
Ecluipment used in the tower method was identical
to- that described already for the scaffolding
method, and was similarly situated. The procedure
was identical except that gas was administered
from 0.5-kg cans and concentration and
duration were increased to 28-33 kg/1000 m3 for
2.5 hours on all islands but El. After fumigation
the methyl bromide was released through tent
seams as with scaffolding, but in the tower system
tent removal was more difficult. At low wind
speeds the tent could be laboriously refurled and
gradually lowered to the boat. W e fortuitously
discovered that high winds, instead of causing
extensive damage as we expected, could actually
aid tent removal. The windward side of the tent
was lifted by poles to about 3 n~ above water,
whereupon sufficient wind entered the tent to lift
it up over the tree and to drop it, partly extended,
into the water on the leeward side of the island.
E l was fumigated in the daytime. For all remaining
islands the tent was unrolled no earlier
than 60 min before sundown, the fumigation was
conducted between 8:00 and 12:00 pm, the gas
was removed by 1 :00 am, and the tent was either
off the island or refurled within 90 mi11 of dawn.
Effects on the mangrove trees of the fumigated
islands are noted in Table 3.
All islands were examined immediately after
tent removal. No living animals at all were found
on El, E2, E8, and E9. A single live curculionid
larva was found on E3. On ST2, 1 living and
at least 100 dead polyxenid millipedes (Lophoupper
canopy
upper canopy
all except one
section
complete
upper canopy and
one other
section
scattered
I
proctinus ?bartschi) were found. Although we
have no further direct evidence and know of no
work on methyl bromide fumigation of millipedes,
we suspect that all millipedes not destroyed immediately
were killed by the delayed effect of methyl
bromide discussed earlier.
In sum, we feel that this fumigation technique
killed all mangrove arthropods with the single
remotely possible exception of cleep-boring beetle
larvae.
Precautions were taken at all monitorings2 to
prevent contamination of the experimental islands.
Boats were never brought in contact with the islands
and only seldom tied to them; usually they
were anchored at least 12 m away. Before wading
to an island, experimenters examined clothing,
equipment, and persons for animals, which were
destroyed. To lessen the chance of phoresy, investigators
simply immersed themselves briefly or
sprayed themselves with "Off" insect repellent.
All equipment used on the island was sprayed
weekly with a short-lived insecticide.
The absence of supratidal land, except around
E9, greatly simplifies the environment of the experimental
islands. The arboreal substrate may
be arbitrarily divided into hollow twigs, living
branches and twigs, dead bark and tree holes,
leaves and flowers, green shoots, and fruits. Because
of the extremely small size of the experimental
islands it was possible to study most of these
microhabitats all but exhaustively~forthe pre-defaunation
censuses. A typical example is the survey
of E3, where we broke and examined 3,500
of an estimated 5,000 small hollow twigs, collected
? The term "monitoring" is used to designate one or
more of the censuses of species, conducted for a period of
17-20 hr during the course of the experiment. In the
present series of articles, monitoring, census, and censusing
are used interchangeably to refer to the listing of
species.
Early Spring 1969 E X P E R I M E N T A L DEFA.UNATION OF ISLANDS 277
90% of the fruits, examined virtually all of the
leaves, and lo~okedunder at least 80% of the pieces
of dead bark. Rhizophora has numerous limbs
and our islands were relatively low, so that we
were able, by climbing carefully, to examine the
canopy right to the top.
The fumigation procedure itself gave added
evidence of the completeness of our original surveys.
The fumigant included 2% chloropicrin,
which drives out even deep-boring and hollowtwig-dwelling
arthropods (cf. Lauck 1965). The
animals either fell into the water or remained
hanging from branches by spines or silk, tenuously
and conspicuously. Those which fell generally
remained afloat and, when a tent seam was opened
after fumigation, flowed out in masses where they
were collected on a screen.
The chloropicrin technique was especially effective
on E8, E9, and ST2, and the results indicated
greater densities of both insects and spiders than
we had suspected. Only on E8, however, were
animals found that had not been noted in the
original surveys. These consisted of four beetles:
Staphylinidae : gen. sp.
Carabidae : Bembidion sp. nr. contracturn
Carabidae : Tachys occulator
Oedemeridae : Oxacis sp.
Of these the first three were apparently living in
a patch of temporarily supratidal mud, which was
submerged again 2 weeks later during a strong
wind. The Oxacis were encountered only on the
fumigation tent, and even their status as "transients''
is doubtful.
Defaunated islands were censused for approxi'NUMBER
OF MAN- HOURS SEARCHING
FIG.10. Cumulative species counts during individual censuses
on three of the experimental islands.
mately 2 days every 18 days. At each census
period approximately 10% of all hollow twigs
were broken. and with this treatment the number
of hollow twigs seemed to remain constant
throughout the experiment. Only when gentle
lifting failed to allow vision was dead bark removed,
and then only a part was peeled off. Small
amounts of carbon tetrachloride and ammonia, and
tapping with a steel probe were used to drive animals
out from underneath bark and inside hollow
twigs without damage to tree or colonist. Although
the pre-defamation searching methods had
been drastic, care was taken following defaunation
to prevent the censusing from being destructive
and from altering the proportions of the
respective microhabitats on an island.
Most colonists could be identified in the field
because of the small subset of the Keys fauna
which invades small mangrove islands and the
completeness of our reference collection. Those
that could not be recognized with certainty were
treated in one of two ways. If the animal in question
was obviously part of a sizable population,
as with many psocopterans and thrips, a small
collection was made that did not exceed 2% of
the estimated minimum size of the population.
When an unknown arthropod was not part of a
large population, we photographed it.
It was quickly discovered that the rate of discovery
of new species during each censusing period,
although at first very high, declined approximately
linearly to near 0 at about 14 man-hours of
searching. A very few species are expected to
appear from immigration within the monitoring
period. Figure 10 depicts the cumulative number
of species observed vs. time for three representative
monitorings. Different sizes of islands,
weather conditions, and especially varying population
sizes caused minor differences in the curves
from island to island and from time to time; but
despite great changes in the numbers of species
through the course of the experiment, the curves
remained remarkably similar in shape. The censusing
period was therefore chosen to last between
17 and 20 man-hours.
Because of the simplicity and very small size
of the experimental islands, and because several
nocturnal censuses located no new species, we felt
that all colonists having crepuscular or nocturnal
activity were being recorded during the diurnal
monitoring. The anyphaenid spider Aysha sp.
and ant Paracryptocerus varians are both strictly
nocturnal, yet both were recorded during the day
in their retreats. Similarly the crepuscular braconid
wasp Macrocentrus sp. and largely nocturnal
ants Camponotus floridanus and Tapinoma littorale
were repeatedly observed.
278 DANIEI. S . SIMBERLOFF AND EDWARD O. WILSON Ecology, Vol. 50, No. 2
ACKNOWLEDGMENTS Hermannsson, S. 1967. Introduction. Surtsey Re-

We thank the IJnited States Department of the Interior search Progress Report, 111: 1. 

for permission to fumigate the islands of this experiment. LaRue, D. C., and T. T. Muzik. 1954. Growth, reIn
particular, Jack Watson of the Great White Heron generation and precocious rooting in Rhizophora manNational
Wildlife Refuge, Roger W. Allin and William gle. Pap. Mich. Acad. Sci. Arts Lett. (1) 39: 9-29.
B. Robertson of the Everglades National Park, and Lauck, D. R. 1965. Chloropicrin for fast action with
Robert M. Linn of the National Park Service examined a Berlese funnel. Turtox News 43: 115.
the plans of the experiment and concluded, with the au- Lindroth, C. H., H. Andersen, H. Bodvarsson, and S. H.
thors, that the total mangrove faunas would be unaffected. Richter. 1967. Report on the Surtsey investigation
Steve Tendrich, Vice-President of National Extermina- in 1966. Terrestrial invertebrates. Surtsey Research
tors, Inc. (Miami) coped imaginatively and effectively Progress Report, 111: 59-67.
with the unique problems presented by the project. W. H. MacArthur, R. H., and E. 0.Wilson. 1963. An equiBossert
and P. B. Kannowski assisted in the original librium theory of insular zoogeography. Evolution
censuses. Maps and tables for all parts of this series 17: 373-387.
were done by Mrs. Helen Lyman. The work was sup- MacArthur, R. H., and E. 0.Wilson. 1967. The theory
ported by N S F grant GB-5867. of island biogeography. Princeton University Press.
203 p.
LITERATURECITED
Mainwaring, A. P. 1961. Some effects of methyl broChisholm,
R. D. 1952. Nature and uses of fumigants, mide on aphids and whitefly and their host plants.
P. 358. In Insects, The Yearbook of Agriculture 1952. ~~~~~~~l thesis, ~~~d~~ univ., ~ ~ ~~ England, d
Darlington, P. J. 1957. Zoogeography: the geographi- Monro, H. A. U.,and Delisle. F~~~~~~1943. applical
distribution of animals. Wiley, New York. 675 p,
cations of bromide as a fumigant. Sci. Agr.
Davis, J. H. 1940. The ecology and geologic role of
mangroves in Florida. Papers from Tortugas Lab., 23: 546-556.
Carnegie Inst. Wash. 32: 307-412. Page, A. B. P., and 0.F. Lubatti. 1963. Fumigation
French, R. A. 1964 (1965). Long range dispersal of insects. Ann. Rev. Entorno'. 8: 239-264.
illsects in relation to synoptic meteorology. Proc. Simberloff, D. S., and E. 0.Wilson. 1969. ~ x ~ e r i Int.
Congr. Entomol. 12th (London). 6: 418-419. mental zoogeography of islands. The colonization of
Fridriksson, S. 1967. Life and its development on the empty is1ands. Ecology 50: 278-296,
volcanic island, Surtsey. Proc. Surtsey Research Wolfenbarger, D.0. 1946. Dispersion of small orgaConf.,
1967: 7-19. nisms. Amer. Midland Naturalist 35: 1-152.
EXPERIMENTAL ZOOGEOGRAPHY OF ISLANDS : THE
COLONIZATION O F EMPTY ISLANDS
The Biological Laboratories, Harvard University, Cambridge, Massachusetts 02138
(Accepted for publication December 16, 1968)
Abstract. We report here the first evidence of faunistic equilibrium obtained through
controlled, replicated experiments, together with an analysis of the immigration and extinction
processes of animal species based on direct observations.
The colonization of six small mangrove islands in Florida Bay by terrestrial arthropods
was monitored at frequent intervals for 1 year after removal of the original fauna by methyl
bromide fumigation. Both the observed data and climatic considerations imply that seasonality
had little effect upon the basic shape of the colonization curves o i species present vs. time.
By 250 days after defamation, the faunas of all the islands except the most distant one ("El")
had regained species numbers and composition similar to those of untreated islands even though
population densities were still abnorn~allylow. Although early colonists included both weak
and strong fliers, the former, particularly psocopterans, were usually the first to produce large
populations. Among these same early invaders were the taxa displaying both the highest
extinction rates and the greatest variability in species composition on the different islands.
Ants, the ecological dominants of mangrove islands, were among the last to colonize, but they
did so with the highest degree of predictability.
The colonization curves plus static observations on untreated islands indicate strongly that 

a dynamic equilibrium number of species exists for any island. We believe the curves are 

produced by colonization involving little if any interaction, then a gradual decline as inter-

action becomes important, and finally, a lasting dynamic equilibrium. Equations are given for 

the early immigration, extinction, and colonization curves. 

Di~persalto these islands is predominantly through aerial transport, both active and passive.
Extinction of the earliest colonists is probably caused chiefly by such physical factors
as drowning or lack of suitable breeding sites and less commonly by competition and predation.
Present address: Department of Biological Science, Florida State University, Tallahassee, Florida 32306
You have printed the following article:
Experimental Zoogeography of Islands: Defaunation and Monitoring Techniques
Edward O. Wilson; Daniel S. Simberloff
Ecology, Vol. 50, No. 2. (Mar., 1969), pp. 267-278.
Stable URL:
http://links.jstor.org/sici?sici=0012-9658%28196903%2950%3A2%3C267%3AEZOIDA%3E2.0.CO%3B2-J
This article references the following linked citations. If you are trying to access articles from an
off-campus location, you may be required to first logon via your library web site to access JSTOR. Please
visit your library's website or contact a librarian to learn about options for remote access to JSTOR.
Literature Cited
An Equilibrium Theory of Insular Zoogeography
Robert H. MacArthur; Edward O. Wilson
Evolution, Vol. 17, No. 4. (Dec., 1963), pp. 373-387.
Stable URL:
http://links.jstor.org/sici?sici=0014-3820%28196312%2917%3A4%3C373%3AAETOIZ%3E2.0.CO%3B2-X
Experimental Zoogeography of Islands: The Colonization of Empty Islands
Daniel S. Simberloff; Edward O. Wilson
Ecology, Vol. 50, No. 2. (Mar., 1969), pp. 278-296.
Stable URL:
http://links.jstor.org/sici?sici=0012-9658%28196903%2950%3A2%3C278%3AEZOITC%3E2.0.CO%3B2-U
http://www.jstor.org
LINKED CITATIONS
- Page 1 of 1 -