Experimental Zoogeography of Islands: Defaunation and Monitoring Techniques
Author(s): Edward O. Wilson and Daniel S. Simberloff
Source: Ecology, Vol. 50, No. 2 (Mar., 1969), pp. 267-278
Published by: Ecological Society of America
Stable URL: http://www.jstor.org/stable/1934855 .
Accessed: 18/01/2011 13:24
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. 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/action/showPublisher?publisherCode=esa. .
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.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
Ecological Society of America is collaborating with JSTOR to digitize, preserve and extend access to Ecology.
http://www.jstor.org
EXPERIMENTAL
Early Spring 1969
DEFAUNATION
267
OF ISLANDS
der the canopy of xerophytesand in open areas. Comsome survivein the desert even duringyears of
monwealth Forest. Rev. 43: 219-234.
drought,withoutany artificialirrigationwithonly
. 1964b. Transpiration of Aleppo pine as a func80 mmof rainfall,and in the subtropicalzone with
tion of environment. Ecology 45: 868-873.
200-300 mmof rain.
. 1966. Attraction of atmospheric moisture by
These resultsand thoseobtainedearlier(Gindel
woody xerophytes in arid climates. Commonwealth
Forest. Rev. 45: 297-321.
1968b,1969) indicatethatfurtherstudyis needed
. 1967a. The correlationbetween cambium activity
to clarify to what extent augmented stomatal
and transpiration rate in Aleppo pine. 14th I U F R 0
densityand the presenceof stomataon the upper
Congress 4: 188-207.
epidermis in woody xerophytesgrown in the
. 1967b. The relationship between transpiration
driest climateand poorest soil serve to increase and six ecological factors. Oecol. Plant. 8: 227-239.
-.
1967c. Afforestation of the desert. "Ariel,"
the abilityof the treeto absorbatmosphericmois19: 13-24.
ture,whenmoisturetensionin the soil and within Jerusalem
. 1968a. Some ecophysiological propertiesof three
theplantreachtheirhighestvalues.
tree xerophytes grown in the desert. Oecol. Plant. 3:
49-67.
ACKNOWLEDGMENTS
1968b. Dynamic modificationsin alfalfa leaves
This work was accomplished with the financial assisgrowing 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 constellationin the leaves of cotto Prof. A. H. Halevy for his criticismof the manuscript.
ton, alfalfa, maize and wheat plants as a function of
soil moisture and environment. Physiol. Plant. (in
LITERATURE
CITED
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 Gewebe.
H. Bruhn, Braunsweig. 34 p.
Gindel, I. 1952. Some anatomical features of the indigenous 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. Biochemical control of stomatal
. 1961. The water balance in xerophytes. In- Zelitch, I.
tern. Union For. Res. Org. 13 Congress 21: 2-4.
opening in leaves. Proc. Nat. Acad. Sci. Wash. 47:
1423-83.
. 1964a. Seasonal flucuationsin soil moisture un-
EXPERIMENTAL
ZOOGEOGRAPHY
AND MONITORING
OF ISLANDS:
TECHNIQUES
DEFAUNATION
EDWARD 0. WILSON AND DANIEL S. SIMBERLOFF1
The Biological Laboratories, Harvard University,Cambridge, Massachusetts 02138
(Acceptedforpublication
December16, 1968)
Abstract. In order to facilitateexperiments
on colonization,a techniquewas developed
thatpermitsthe removalof the faunasof very small islands. The islandsare coveredby a
tentand fumigatedwith methylbromideat concentrations
that are lethal to arthropodsbut
notto theplants.
Seven islands in Florida Bay, of varyingdistanceand directionfromimmigrant
sources,
were censusedexhaustively.The small size (diameter11-1i m) and ecological simplicity
of these islands,whichconsistsolely of red mangrovetrees (Rhizophoramangle) with no
supratidalground,allowedthe locationand identification
of all residentspecies. The terrestrial
fauna of these islands is composedalmost exclusivelyof arboreal arthropods,with 20-50
speciesusuallypresent.
at any givenmoment. Surveysof these taxa throughout
the Florida
Keys, withemphasison the inhabitants
of mangroveforests,were made during1967 in order
to estimatethe size and composition
of the "speciespool."
The sevenexperimental
islandsweredefaunated
in late 1966and early1967,and thecolonists
weremonitored
for17-20man-hours
every18 days. Precautionswere takento avoid artificial
introductions
duringthemonitoring
periods.
INTRODUCTION
Arthurand Wilson (1963, 1967), who postulated
The firstattemptto formulatea quantitative an equilibriumnumber of species on an island
and
of immigration
theorywhichwould unitean ever-increasing
mass determinedby the intersection
numthe
of
as
a
function
drawn
curves
extinction
of insular biogeographicdata was made by MacPresent address: Departmentof Biological Science, ber of species already present. In addition to
speculationon the formsof these curves and efFlorida State University,
Tallahassee, Florida 32306
1
268
EDWARD
0. WILSON
AND DANIEL
fectsof varyingbothislandarea and distancefrom
the source,theyderivedthe equation
1.15 mean
to-0
where9mean
-
X
-
(1)
the hypotheticalequilibriumnumber
of species on an island
the mean of 9i for a numberof very
similarislandsi
the extinctionrate (or turnoverrate,
since an equilibriumrequiresthat extinction immigration)in numberof
species per timeat equilibrium
time required for numberof species
presentto increasefrom0 to 90% ofD.
-
to1.o
Later thesame authorselaborateduponthegeneral shapes of immigration
and extinctioncurves,
discussed colonizationrates and curves,and predictedthe distribution
of survivaltimesof species
that succeed in colonizingan island (MacArthur
and Wilson 1967).
Data to testthesehypothesesare scarce. Most
when combinedwith
biogeographicinformation,
information
furnishedby the scant fossil record,
can at best lead to descriptionsof broad patterns
of distribution
and zoogeographicregionsand to
hypothesesabout the essentiallylong-termprocesses responsible for them (e.g., Darlington
1957). Such theories,though occasionallysugrrpqtil-p
fnr%
rnri*i
e vr t-,L
foi
toorn- --
i the--1
Ecology, Vol. 50, No. 2
S. SIMBERLOFF
existingtaxonomic distributionsand the underlying colonizationmechanismsfor most islands.
even dailyeventsthat
They neglectthe short-term,
largelydeterminethe parametersof colonization
on many islands. Such eventsbecome decisively
more importantthan evolutionas distanceof island fromsource area decreases. In particular,
the classical zoogeographicmethodsdo not provide a test for the existence of an equilibrium
numberof species or for the accuracy of equation ( 1).
The literatureon biotic dispersal,both anecon the
providesinformation
dotal and systematic,
relativeimportanceof activeand passive transport
and a wealthof data on the specificagentsof passive transport(e.g., Wolfenbarger1946). Certain cases of overseas dispersal have even been
traced to specificmeteorologicalevents (French
1964). For no island,however,is thetimecourse
of colonizationby even a large fractionof the
inhabitantsknown. The large numberof records
of long distancedispersalimpliesthatimmigration
rates to islands are probablyhigh,especiallyfor
organismscapable of passive transportby wind.
of the immigraBut the shape and determinants
tion curve cannot be deduced. At best a lower
limitmightbe given.
In the recentstudiesof Surtsey,a new volcanic
island near Iceland, there has been a deliberate
attemptto documentthe colonizationprocessfrom
EI~...
X E BXMb.
CAPC
FIG. 1. The southerntip of Florida and the Florida Keys.
shown in detail in Figures 3-5.
C
The rectangles enclose the experimental areas
EXPERIMENTAL
Early Spring 1969
DEFAUNATION
NWFIG. 2.
Upper: Island El, the second smallest island in
the experimental
series. Lower: Island E9, the largest
island in the experimental series; note also the presence
of supratidal mud.
the birthof the island (Fridriksson 1967, Hermannsson1967, Lindrothet al. 1967). But the
infrequent
naturaloccurrenceof such eventstends
to render quantitativebiogeographictheorythe
product of abstract speculationand of enlargementand sophisticationof untestablehypotheses.
What is needed,clearly,is a methodeitherof producing new islands similarto natural ones or of
sterilizingpreexistingislands.
THE
EXPERIMENTAL
ISLANDS
Along the Overseas Highway of southernFlorida (U. S. Route 1) thereexist a vast array of
small, approximatelycircular islands situated in
shallowbay water (Fig. 1). Together,theyconstitutea potentialnatural"laboratory"forexperimentalbiogeography. The islands with diameter
10-20 m (Fig. 2) are 5-10 m tall and usually
TABLE
Series 2
island
aControl
b"Stepping
Stone"(SI) 572maway.
consist entirelyof one to several red mangrove
trees (Rhizophora mangle), with rarely a small
black mangrovebush (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 completelyflooded during prolonged
winds of 10 knots or more. Individual islands
may differin several minorrespectssuch as the
numberof large trunksand amountof dead bark
theycontain,but on the whole they are remarkably similarto each otherin physicalappearance.
A fruitfuldefaunationexperimentwith these
islands requiresthe followingconditions:
i) thattherebe enoughof the islands forreplication and variation in distance to the nearest
source area;
ii) that the animal diversitybe sufficientto
allow statisticaltreatment,and the organismsbe
physicallylarge enoughto insurerecordinginconspicuousformsas well as conspicuousones;
iii) thatthe extremelysmall size of the islands
compensatesfor theirrelativenearnessto source
areas and therefore
producesa distanceeffect.
During an exploratorytripto the Florida Keys
in 1966, we found that convenientlyaccessible
small islandswere numerousat most distancesup
to 200 m fromthe nearestlarge islands. Beyond
200 m theywere rare, althoughsomewhatlarger
islands (diameter50 m and greater) were plentiful. A veryfewsmallislandssuitableforour purposes were located 0.2-1.4 km fromthe nearest
sourcearea.
A set of such islands were chosen for experimentationand their faunas carefullysurveyed.
Since the breedingfauna of islands this size consists almostentirelyof species of insectsand spiders, we made a referencecollectionof land ar-
Parameters of experimental islands
Diameter
Distance from
nearestsource
Initial no.
arthropod
species
El
E3
ST2
E2
E65
11
12
11
12
15a
533
172
154
2
73a
25
31
29
43
30a
offSquirrelKey
in RattlesnakeLumps
offSaddlebunchKey
offSnipe Keys
in JohnstonKey Mangroves
E7
25
15
29
near Manatee Creek
E8
18
1, 188b
29
E9
18
379
41
Elope
lha
betweenCalusa Keys and
Bob Allen Keys
betweenCalusa Keys and
Bob Allen Keys
near Bottle Key
Island
name
Series 1
1.
269
OF ISLANDS
(i)
(i)
37a
208
Location
270
EDWARD
0. WILSON
AND DANIEL
Ecology,Vol. 50,No. 2
S. SIMBERLOFF
thropodsof the Florida Keys with the assistance that bred on our islands duringthe course of the
of Robert Silberglied. This collectionenabledus experiment(1966-68) were one or two pairs each
to identifythe arthropodson the experimental of the Green Heron, White-crownedPigeon, and
islands,and it provideda measureof the size and Gray Kingbird. Snakes (mostlywater snakes of
compositionof the species pool of the presumed the genus Natrix) occasionally swim to small
mangroveislands, and raccoons visit islands losourcearea, the entireFlorida Keys.
Animal diversityon the small mangroveislands cated on shallow mud banks. These vertebrates
proved to be adequate for statisticalpurposes. were not includedin the censuses.
About 75 insectspecies (of an estimated500 that
On the basis of their distance and direction
inhabitmangroveswamps and an estimatedtotal fromthe nearestsource area, theirsize, and their
of 4,000 thatinhabitall the Keys) commonlylive accessibility,the islands listed in Table 1 were
on these small islands. There are also 15 species chosen for defaunation. Two islands (E6 and
of perhaps E10), one each in the vicinityof the two groups
of spiders (of a total Keys complement
125 species) and a few scorpions,pseudoscor- of experimentalislands, were selected to serve
pions,centipedes,millipedes,and arborealisopods. as controlislands; and theywere censusedat the
At any givenmoment20-40 speciesof insectsand same time as the experimentalislands. In order
2-10 speciesof spidersexiston each ofthe islands. to ascertainwhetherany long-rangechange had
Many birds visit the islands eitherto roost or occurredin the generalfauna of small mangrove
to forage,butveryfewnestthere. The onlybirds islands in Florida Bay duringthe course of the
Barracuda)Ys
~~M affordKey
e
J
= ORSqirref
17<
Kay
daldn&Key
Nappjj
JackKeDy/
U
H
y
~~~~~Drequez
S
<7-=
ST2
Sugarloat KV
wa
10 1
FIG. 3.
0
X
i
~~~~~~~~~~~~~~~Rattlesnake
Lump
~
MiJes
3000
$000M
2000
400
00
Map showing the location of the experimental islands of series 1.
0
Early Spring 1969
EXPERIMENTAL
DEFAUNATION
271
OF ISLANDS
original faunas of the distance effectas defined
by MacArthurand Wilson (1967). Qualitative
distantislands
in faunasof increasingly
differences
FLORIDA
Ca/usa
Xgsr
were even more strikingthan quantitativeones.
For many groups-especially ants and spiders6''.
B~.AY
effectis so regularthatone can guess
the-distance
not onlythe species numberbut also the approxi-.
ES
mate species composition. For example, centiFoxtrot
X5ys
pedes and pseudoscorpionswere found almost
exclusivelyon islands of less than 200 m distance
BobAllenK~ys
fromthe nearestlarge island.
Islands of this minutesize can be remarkably
long lived. All Rhizophora mangle trees, and
particularlythose on low, small islands, have
numerousaerial roots: both brace roots growing
0
5000 Meters
FIG. 4. Map showingthe locationof the experimental out fromthe main stem and prop roots growing
downfrombranches. The lattercan emergefrom
islandsof series2.
as high as 4 m above the water (LaRue and
Muzik 1954) and in time can become as thick
MANATEE
as the main stem. Rhizophora trees reproduce
theirseeds growingon the tree until
viviparously,
theybecomelarge,long,pointedseedlings. When
Bqy Point
E7
fromtheuppercanopy,some
theydrop,frequently
evidentlyplantthemselveswheretheyfall (Davis
1940). Mangrove swamps are oftenquite thick,
and on small islands theremay be several trees.
Stems,thickenedprop roots,and brace roots in-
~~
0
0
FIG.
;
SKH
CrossKey,
Bf13ake~s
K~y
ElI
2 Miles
1/2
1000
~:
2000 Meters
5. Map showingthe locationof the originaltest
island E7.
a secondcensus of the controlislands
experiment,
was made at the close of the experiment.
islandsformtwo
The two seriesof experimental
widelyseparatedgroups. Series 1 lies withinthe
Great White Heron National Wildlife Refuge,
in a regionnorthof SugarloafKey (Fig. 3). Series 2 is in the EvergladesNational Park near the
Calusa Keys northof Islamorada (Fig. 4). E7,
which was chosen primarilyto test a methodof
defaunation,is located in Barnes Sound south of
the Glades Canal (Fig. 5).
The islandswere also selectedto show as much
variation as possible in the degree of isolation
of
fromthe immigrantsources. The coefficient
correlationfor island series 1 between"distance
fromnearestsource" and "numberof arthropod
species" is -0.83, indicatingthe operationin the
IKilometer
FIG.
IV
6. Map showingEl and surrounding
regionas it
appearedin 1851-57.
,SquirreIl'~
E1I
II W~st;IA
Kilometer/
FIG.
7. The sameregionshownin Figure6 as it appeared
in 1964.
272
EDWARD
0. WILSON
AND DANIEL
S. SIMBERLOFF
Ecology,Vol. 50.No. 2
termingleand reproducethemselvesso that it is cannot be expected to penetrateall the hollow
impossible to distinguishhow many trees are twigsof a Rhizophoraisland. Some twigdwellers
present,whichare which,whichis the initialone will probablysurvive,particularlyif they inhabit
(even ifit is stillpresent),and whichare the main narrow twigs (Tapinoma and Xenomyrmex)or
else pack theirtunnelstightlywith powderyexstems.
Maps of the Florida Keys, especially of the creta (Styloleptus). Had we attemptedto kill
SugarloafKey-Key West area, have been remark- thesesurvivorsby usinga spraywitha long-lived
ably detailed and accurate with respectto small or more powerfulresidue,we would have postmangroveislands for at least a century. These poned the beginningof the colonizationcurve,
tell us thatE7 and all the islands in series 1 have since immigrantsmay be killed by the residue.
existedat least 25 years. The otherexperimental Because we had to be certainthat all colonists,
islands were not mapped in sufficient
detail to or at least all but thosebelongingto a small numdeterminelongevity. The size and shape of indi- ber of known taxa, would be destroyedimmevidual islands may change, even drastically,but diately,we abandonedthe spraytechnique.
We turnednextto the moredifficult
alternative
in general the islands survivethe climaticcatastrophesto which they are repeatedlysubjected. technique of fumigation. The method has the
For example,Figure 6 shows El (Blake's Key) obvious advantages that the residual effectof a
and surroundingareas as it appeared in 1851-57, gas is negligible,and even the inhabitantsof holwhile Figure 7 shows the same island region on low twigs are contacted. This is the standard
the Coast and GeodeticSurvey map of 1964. It method used by professionalexterminatorson
thereforeseems reasonable to assume that the termitesand otherwood-boringinsectpests.
The fumigationof livingplants is a relatively
faunas foundinitiallyon the experimentalislands
weretheproductofmanyyears'colonization.Our unchartedarea. The biochemicaleffectsof fumiexperimentshave subsequentlyshown that the gants on plants are poorly known,damage usuminimumages of the islands greatlyexceed the ally being describedin termsof the physicalaptime requiredto reach species equilibriumfroma pearance of the plant shortlyafter fumigation
start of zero species (Simberloffand Wilson (Page and Lubatti 1963). To our knowledge,the
only priorfumigationof red mangrovetrees was
1969).
by B. P. Stewart(pers. comm.), who used methyl
DEFAUNATION
bromideon small plants. With 32 kg/1000 m3
Our firstattemptto defaunateislands made use for 2 hr at 320C and 40 kg/1000m3 for 2 hr at
of an insecticidespray of short-livedresidual ef- 270C he found that all insects apparentlywere
fect. On July 9-10, 1966, two experimental killed,althoughsome did not die until48 hr after
islands (El and E2) were sprayeduntildripping fumigation. Mortality of cockroach eggs was
with 60 g parathion,240 g diazinon and 180 g unrecordedin Stewart's study. Damage to the
Pylac stickingagentper 1001 freshwater. When plant was limitedprimarilyto young leaves and
the islands were examined 1 day later, all the twigs,a resultthat is in accord with the general
surface and bark fauna and most borers were findingof Page and Lubatti (1963) thatthegrowdead. However, the followinglive animals were ing stages of bothplantsand insectsare the most
foundin thinhollowtwigs. El: one adult wasp, susceptibleto fumigation.
Sclerodermamacrogaster(Bethylidae); workers
In our preliminarytrials we were limitedby
and larvae of Crematogasterashmeadi (Formici- the fact that fumigantshighlysoluble in water,
dae). E2: two larvae of ?Styloleptusbiustus. such as hydrogencyanide, could not be used.
(Cerambycidae); two coloniesof Paracryptocerus These substances would be expected either to
varians (Formicidae); and one colony of Tapi- leave the fumigationtent throughthe water or
noma littoral (Formicidae). A more through to concentrate
to an unknownextentin the water
examination9 days after sprayingproduced the under the island. Either effectwould greatly
followinglive and apparentlyhealthyinsects. El:
complicatethe maintenanceand measurementof
two larvae of ?Styloleptusbiustus (Cerambyci- gas concentration. Consequently,the following
dae). E2: one larvae of ?Styloleptusbiustus four relativelyinsoluble fumigantswere tried:
(Cerambycidae); one lepidopteranlarva; and two Pestmaster Soil Fumigant-1 (methyl bromide
coloniesof Xenomyrmexfloridanus(Formicidae). 98% wt, chloropicrin2% wt); Acritet 34-66
The large populations of the ant colonies and (acrylonitrile34% vol, carbontetrachloride
66%
advanced developmentof the beetle larvae pre- vol); Dow Ethylene Oxide (ethylene oxide
cludedtheseinsectshavingimmigratedduringthe 100%) ; and Vikane (sulfurylfluoride95% wt,
9 days followingspraying.
inertingredients5% wt). Field tests were perThe resultson El and E2 showedthat a spray formedon small Rhizophora trees in Matheson
Early Spring 1969
EXPERIMENTAL
DEFAUNATION
OF ISLANDS
273
Hammock Park, Coral Gables, by National Ex- canopy. Instead, whole sectionsof the tree beterminatorsof 'Miami, Florida, under the direc- came completelybrown while others remained
tion of Steve Tendrich in consultationwith the almostunscathed. The upper canopywas always
authors. Additionaltestson themortality
of man- more severelybrowned,but there seemed to be
grove colonistswere performedat National Ex- no otherconsistentdamage pattern.
Abscission of the dead leaves began withina
terminators'
laboratoryin Miami.
With the temperatureoutside the fumigation week and, oftenaided by wind,was usuallycomtentremainingat 240-290C, we determinedwhich pletewithin8 weeks. At thistimedamagedtrees
duration-concentration
intensitiescaused no more looked superficiallydead except for remaining
than bearable damage to the tree yet were lethal green sections. Closer examinationshowed that
to all the mangrovearthropods. Acritet,Vikane, the damaged trees could be convenientlydivided
and ethyleneoxide caused extensive,apparently intothreeclasses, as follows:
irreversibledamage to red mangroveand were
For those trees where browningof leaves and
rejected. However, methylbromide at 22 kg/ shoots appeared completeby 72 hr the damage
1000 M3 for 2 hr caused browningof only about was irreversible.Withina fewmonthsbarkbegan
10% of the leaves 1 week aftertreatment,with to peel and twigs with green wood could not be
no visiblelatereffects.This durationand concen- found. A year latertherewas no evidenceof life.
tration killed all the mangroveinhabitants(in- Island E8 is in thisclass. In the secondcategory,
cluding such resistantformsas roach eggs and which includes only E7, a large majorityof the
lepidopteranpupae) with one possible exception. leaves and shoots were brownedand died, while
When data fromall tests at this duration-concen-a small section remained green. In the third
trationare lumpedtogether,only 95% of ceram- category,the trees were less severely damaged
bycidsand 80% of weevils were dead when dug (60% or fewerbrownedleaves). Green sections
out of theirburrows 1-10 days afterfumigation. remainedunchanged,whileleafabscissionin most
The remainderall died within2 days of removal. of the damaged sectionswas accompaniedby the
It is temptingto ascribe the death of these few sproutingof new shoots and leaves fromalmost
larvae to a delayedeffectof the fumigant. How- all branches. Within2 monthsthe treesappeared
ever, a controlexperimentconsistingof the re- normal except for occasional leafless patches in
moval of five unfumigatedcerambycid larvae the extremeupper canopy. Island El, which 2
(Styloleptus biustus) fromtheir respectivebur- days afterfumigation
appearedto have about 50%
rows caused death within 5 days without any brownleaves, recoveredafter1 monthto the exadditionaltreatment. Monro and Delisle (1943) tentthatit did not appear to have incurredmore
report that many insects are active soon after damagethansimilartreesdo duringheavystorms,
exposureto methylbromidebut succumbin time and 2 monthsafterdefaunationseemedquite nor(some lepidopterous larvae surviving for 2 mal except for a small bare patch in the upper
months),and thatthis delayedeffectis especially canopy. At no timedid its barkcrackor peel.
commonin fumigations
at the low concentrations In mostdamagedtreesof the firsttwo classesused in our own fieldtests. This peculiarityis whereeitherthe wholetreeor all but a small secconfirmed
by Chisholm(1952) forJapanesebeetle tion was killed-there was copious oozing of sap
larvae and other insects. Of at least equal im- for up to a week afterfumigation. Such oozing
portance,theresultsfromour experimental
islands was neverobservedin untreatedtrees,nor in those
(Simberloffand Wilson 1969) stronglysuggest fumigatedtrees which showed good leaf replacethatthe cerambycidswere eliminatedby the fumi- ment.
gationtreatment.We therefore
The action of methylbromideon livingplants
concludethateven
the deep-boringbeetle larvae were very probably has been studied rather extensivelyby several
killedby the treatment.
authors. Mainwaring(1%1) examinedthe physThe overt physical damage to the trees con- iological responses of several plants to methyl
sisted of a browningof leaves and shoots and bromide,and his numerous observationsimply
occasional heavy oozing of sap. At 61 kg/1000 that its major effectis auxin inhibition. The
m3 the leaf browningwas manifestupon tent re- damage we observedseemed too drasticand cermoval, but at 35 kg/1000 m3 and less there was tainly too immediateto be attributedsolely to
littleor no immediatebrowning. The effectbe- auxin inhibition,althoughthe presenceof such a
came evidentat about24 hr and worsenedin terms biochemicalaction cannotbe ruled out.
of fractionof leaves brownedthrough72 hr, by
Whatever the effectof methyl bromide on
whichtime up to 90%o of the leaves were brown Rhizophoramangle,it apparentlyhad the expected
and after which no furtherbrowningoccurred. Q1o of 2-3. Our initialtests in MathesonHamBrowningdid not occur randomlythroughoutthe mockwere conductedon small plantswell shaded
274
EDWARD
0. WILSON
AND DANIEL
S. SIMBERLOFF
Ecology,Vol. 50,No. 2
by large trees, and the temperatureinside the TABLE 2. Arthropod species found on E7 just prior to
and following fumigation
fumigation
tentremainedwithin30 of the air temperature (240-29?C). The experimentalislands
could not be shaded, and even on overcast days INSECTS
gen. sp.a
Embioptera:
the temperatureinside the dark tent was at least
Latiblattellan. sp.
Orthoptera:
Cycloptilum
sp.
320C. On sunnydays the innertemperature
was
Tafalisca luridaa
so high that the presence of the coveringalone
PseudoacallesSp.b
Coleoptera:
Leptostylus
sp.
caused browningof leaves. It seems probable
Styloleptus
biustusb
thatthegenerallymoreseveredamageto theupper
Tricorynussp.
canopythroughout
this experimentwas caused by
Archipsocuspanama
Psocoptera:
Pseudococcussp.
Hemiptera:
highertemperature,especially since methylbroNemapogonsp.a
Lepidoptera:
mide gas has a density (3.974 g/literat 20'C)
Casinaria texana
Hymenoptera:
considerablygreaterthan that of air.
Camponotusfloridanusa
Camponotusplanatue
Air humidityis much less importantthan temParacryptocerus
varianse
perature in its effecton fumigationdamage to
Pseudomyrmex
elongatusl
Xenomyrmex
floridarusa
plants (Page and Lubatti 1963); our experiments
were almost all conductedwithin a narrow hu- ARACHNIDS
miditvrange (65 %-90%) and the effectsof varyAriadna arthuri
Araneae:
Eustala sp.
ing this parameterwere not studied.
Leucaugevenustaa
The firsttwo experimental
islands (E7 and El)
Acarina:
Galumnasp.
were fumigatedduringdaytimeand damaged. In
order to reduce heat damage we thereforefumi- OTHERChilopoda:
Orphnaeusbrasilianus'
gatedall otherislandsat night. The concentration
Rhyscotussp.a
Isopoda:
was increasedfrom22 kg/1000m3 to 28-30 kg/
discovered dead afterfumigation
1000 n13withno observableadditionaldamage.
discovered mostlydead afterfumigation(see text)
The chemicalmeans for defaunationnow having been chosen and tested,we were leftwithan tionalside piecesforlargerislands) ontoan island,
overwhelmingphysicalproblemof how to fumi- the tree could not support the weight without
gate an entire island. Assuming that we could major limb breakage,since even our special halfsomehowget the fumigation
tent (a 25-mby 25-m weighttentweighed275 kg. It was evidentthat
piece of plastic-impregnated
canvas with addi- some sortof superstructure
was needed,one which
would surroundand rise over the islandto remove
mostor all of the tentweightfromthe tree. Constructinga permanentframeworkand lowering
it overthe islandsby helicopterwas too expensive.
t~~~~~~~~~~~
An alternativemethodemployinga temporary
full frameworkconstructedon the site was used
on October 10, 1966 to defaunateE7. The procedure was as follows (see also Fig. 8): First,
planks to preventthe frameworkfrom sinking
intothe mud were placed underwaterin a square
around
the islands and held manuallyuntil the
~
~~
firsttier of scaffolding(which was sufficiently
heavyto keep themfromfloating)was placed on
top. Then the remainingtiers were added, one
side at a time,untilthe island was surroundedby
a cube of scaffolding.Fig. 8 (upper) shows the
frameworkcompleteexcept for the upper tier of
in the rightrear cornerand the square
scaffolding
walkway of planks around the very top of the
structure.
The tentwas thenraised by blockand tackleto
the top of the frameworkand lowered over the
island. Methyl bromide was introduced from
45
kg cylindersthroughthe tent wall at opposite
FIG. 8. Upper: The scaffolding
constructed
aroundE7,
sides
of the island, 2.5 m above the water, and
completeexceptforthe top walkway. Lower: the fumiinto a metal tub mountedwith an electricfan to
gationtentoverE7.
.
,in
......N
INS-
_.
l_;
...
Early Spring 1969
EXPERIMENTAL
DEFAUNATION
275
OF ISLANDS
dispersethe gas. Chloropicrin(tear gas), added
to the odorless methylbromideas a safetymeasure,was allowed to drip intoburlapsheetsat the
bottomof the tubs. Concentrationof the methyl
bromidewas measuredfrequently
by a Gow-Mac
thermalconductivitymeterwith a self-contained
pump attached to a testingstation2.5 m above
the water and well away fromthe two shooting
stations.
The methylbromideconcentrationwas gradually builtup from0 to 22 kg/1000m3over 30 min,
thenkeptat 22-25 kg/1000m3for2 hr. Outside
air temperatureduringthis time remainedsteady
at about 270C. At the end of the fumigation
the
gas was released throughseam openingsfor 45
min,and the tentwas thenremoved.
Immediatelyfollowingthe dissipation of the
gas, E7 was exploredfor a total of 6 man-hours.
The resultsgenerallyconfirmed
the evidencefrom
the Matheson Hammock tests. Dead individuals
representing11 of the 23 species observedbefore
fumigationwere found,frequentlyin large numbers and in all life stages. The details are given
in Table 2. The only live animals encountered
were larvae of two deep-boringbeetles. These
*..-...~~~~~~~~~~~~~~~~~~~......
formedonly a small fractionof the originalpopulation,as shown by the followingrecoverydata:
of 59 larvae of Styloleptusbustus (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 cerambycidswas
extremelysluggishand died an hourafterremoval.
The othercerambycidand the two weevilsseemed
..:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.
healthy,thoughone of the weevils became torpid
......
.
and died 9 hr afterremoval.
The tree was severely damaged as described
previously. Within2 months85%o of the foliage
was dead, althoughthe single living section has
remainedgreen until at least May 1968. Thus
the procedureworkedessentiallyas expected,and
tentbeingopenedover
this aspect of the fumigationwas leftunchanged FIG. 9. Upper: the fumigation
in subsequentdefaunations,except that the work E2 from a tower support. Middle: E2 partially covered.
E9 coveredby the fumigation
tenton a tower
was henceforthdone at night in all but one in- Lozccr:
support.
stance (El).
The cubical scaffolding
framework
had disadvantages,chief among them being the
largecrewnecessaryto erectit (at least fivemen) nerable to high winds, but it has the great ad anand the numberof pieces and weightof the equip- tage of requiring less equipment and a smaller
ment. This necessitatednumerousboat trips be- crew to assemble it. National Exterminators
tween E7 and our staging area near Manatee contracted a steeplejack to perfect the tower
Creek,sincethe shallowwateralongthe routepre- method. After a month of experimenting on iscluded the use of large craft. This problemwas lands near Key Largo, the following procedure
greatlyaggravatedon the otherislands,whichare was adopted and used for the remaining islands.
located considerablyfurtherfrom a convenient A collapsible 10-m triangular steel tower was
stagingarea. We thereforeset out to devise a erected in the center of the island, either propped
towerto be locatedin the centerof the island and on a large root or forced deep into the ground.
risinghighenoughso thatmostof the tentweight The steeplejack then climbed to the top and threw
restedon it. This device provedto be more vul- out three guy wires which were fastened down
....
....:.
..
.......
276
EDWARD
TABLE 3.
Island
0. WILSON
AND DANIEL
S. SIMBERLOFF
Ecology,Vol. 50,No. 2
Extent of damage to foliage and subsequent recovery on experimental islands
Fumigation
date
Time
Extent of
damage
(burntleaves)
Damage location
El..............
3/13/67
day
50%
general
E2..............
E3..............
E7..............
3/7/67
3/17/67
10/10/66
night
night
day
10%
5%
85%
E8 ..............
E9..............
4/17/67
4/7/67
night
night
100%
20%
ST2.............
4/20/67
night
5%
upper canopy
upper canopy
all exceptone
section
complete
uppercanopy and
one other
section
scattered
Recoveryoffoliage
completein 2 months,
exceptforextreme
uppercanopy
completein 2 months
completein 2 months
none
none
littlein upper canopy
completeelsewherein
2 months
completein 2 months
withmudscrewsor stakes,dependingon the sub- proctinus ?bartschi) were found. Althoughwe
strate. Using a blockand tackle,one man on the have no furtherdirectevidenceand know of no
towerand one on the island liftedthe rolledtent workon methylbromidefumigation
of millipedes,
until it formeda triangleover the island. The we suspectthatall millipedesnot destroyedimmetent was then carefullyunrolleddown the guide diatelywere killedby the delayedeffectof methyl
wires until the island was covered (Fig. 9). It bromidediscussedearlier.
was impossibleto unfurlthe tentfromthe tower
In sum, we feel that this fumigationtechnique
in winds greater than 8 knots. Sandbags and killed all mangrovearthropodswith the single
stakeswereused to keepthetentedgessubmerged. remotelypossible exceptionof deep-boringbeetle
Equipmentused in the towermethodwas iden- larvae.
tical to that describedalreadyfor the scaffolding
MONITORING
TECHNIQUE
method,and was similarlysituated. The procePrecautionswere taken at all monitoring2to
dure was identicalexcept that gas was adminisof the experimentalislands.
tered from 0.5-kg cans and concentrationand preventcontamination
durationwere increasedto 28-33 kg/1000m3 for Boats were never broughtin contactwiththe is2.5 hourson all islandsbut El. Afterfumigation lands and only seldomtied to them; usuallythey
the methylbromide was released throughtent wereanchoredat least 12 m away. Beforewading
seams as withscaffolding,
but in the towersystem to an island, experimentersexamined clothing,
tent removal was more difficult.At low wind equipment,and persons for animals,which were
speeds the tent could be laboriouslyrefurledand destroyed. To lessen the chance of phoresy,ingraduallylowered to the boat. We fortuitouslyvestigatorssimplyimmersedthemselvesbrieflyor
discovered that high winds, instead of causing sprayed themselveswith "Off" insect repellent.
extensivedamage as we expected,could actually All equipmentused on the island was sprayed
aid tentremoval. The windwardside of the tent weeklywith a short-livedinsecticide.
The absence of supratidalland, except around
was liftedby poles to about 3 m above water,
E9,
greatlysimplifiesthe environment
of the exwhereuponsufficient
wind enteredthe tentto lift
it up over the treeand to drop it,partlyextended, perimentalislands. The arboreal substratemay
be arbitrarilydivided into hollow twigs, living
intothe wateron the leewardside of the island.
and twigs, dead bark and tree holes,
branches
E1 was fumigatedin the daytime. For all reand
leaves
flowers,green shoots,and fruits. Beislands
maining
the tent was unrolledno earlier
smallsize oftheexperimenthan 60 min beforesundown,the fumigationwas cause ofthe extremely
conductedbetween8:00 and 12:00 pm, the gas tal islands it was possibleto studymost of these
all but exhaustivelyfor the pre-dewas removedby 1 :00 am, and the tentwas either microhabitats
A typicalexample is the surcensuses.
faunation
offthe island or refurledwithin90 min of dawn.
Effectson the mangrovetrees of the fumigated vey of E3, where we broke and examined 3,500
of an estimated5,000 smallhollowtwigs,collected
islandsare notedin Table 3.
All islands were examined immediatelyafter ? The term "monitoring" is used to designate one or
tentremoval. No livinganimalsat all were found more of the censuses of species, conducted for a period of
on El, E2, E8, and E9. A singlelive curculionid 17-20 hr during the course of the experiment. In the
present series of articles, monitoring,census, and censuslarva was found on E3. On ST2, 1 living and ing are used interchangeablyto refer to the listing of
at least 100 dead polyxenidmillipedes (Lopho- species.
Early Spring 1969
EXPERIMENTAL
DEFAUNATION
OF ISLANDS
277
90% of the fruits,examined virtuallyall of the mately 2 days every 18 days. At each census
leaves,and lookedunderat least 80% ofthepieces period approximately10% of all hollow twigs
of dead bark. Rhizophora has numerouslimbs were broken,and with this treatmentthe numand our islands were relativelylow, so that we ber of hollow twigs seemed to remain constant
were able, by climbingcarefully,to examine the throughoutthe experiment. Only when gentle
liftingfailed to allow vision was dead bark recanopyrightto the top.
The fumigationprocedure itself gave added moved,and thenonlya partwas peeledoff. Small
and ammonia,and
evidenceof the completenessof our originalsur- amountsof carbontetrachloride
veys. The fumigantincluded 2% chloropicrin, tappingwitha steel probewere used to driveaniwhich drives out even deep-boringand hollow- mals out fromunderneathbark and inside hollow
arthropods(cf. Lauck 1965). The twigs withoutdamage to tree or colonist. Altwig-dwelling
searchingmethodshad
animals either fell into the water or remained thoughthe pre-defaunation
hangingfrombranchesby spinesor silk,tenuously been drastic,care was taken followingdefaunaand conspicuously. Those which fell generally tion to preventthe censusingfrombeing destrucremainedafloatand, whena tentseam was opened tive and from altering the proportionsof the
flowedout in masses wherethey respectivemicrohabitatson an island.
afterfumigation,
Most colonistscould be identifiedin the field
were collectedon a screen.
techniquewas especiallyeffec- because of the small subset of the Keys fauna
The chloropicrin
tiveon ES, E9, and ST2, and the resultsindicated which invades small mangrove islands and the
greaterdensitiesof both insectsand spidersthan completenessof our referencecollection. Those
we had suspected. Only on E8, however,were that could not be recognizedwith certaintywere
animals found that had not been noted in the treatedin one oftwo ways. If the animalin quesoriginalsurveys. These consistedof fourbeetles: tion was obviouslypart of a sizable population,
as with many psocopteransand thrips,a small
Staphylinidae:gen. sp.
collectionwas made that did not exceed 2%o of
Carabidae: Bembidionsp. nr. contractum
the estimatedminimumsize of the population.
Carabidae: Tachys occulator
When an unknownarthropodwas not part of a
Oedemeridae: Oxacis sp.
largepopulation,we photographed
it.
Of these the firstthreewere apparentlylivingin
It was quicklydiscoveredthat the rate of disa patchof temporarily
supratidalmud,whichwas coveryof new species duringeach censusingpesubmergedagain 2 weeks later during a strong riod,althoughat firstveryhigh,declinedapproxiwind. The Oxacis were encounteredonly on the matelylinearlyto near 0 at about 14 man-hoursof
fumigationtent,and even their status as "tran- searching. A very few species are expected to
sients"is doubtful.
appear from immigrationwithinthe monitoring
Defaunatedislands were censusedfor approxi- period. Figure 10 depictsthe cumulativenumber
of species observedvs. time for threerepresenta30
tive monitorings. Different sizes of islands,
weatherconditions,and especiallyvaryingpopulationsizes caused minordifferences
in the curves
up) 25fromisland to island and fromtime to time; but
.0
despite great changes in the numbersof species
'U
the course of the experiment,the curves
through
Z3
2-0
'U
remainedremarkablysimilarin shape. The cena.
(/)
Boa
chosento last between
susingperiodwas therefore
Li.
17 and 20 man-hours.
Because of the simplicityand very small size
w
E9, IV-5-6-68
-2/
-0m
E7, IV-4,8-68
of the experimentalislands, and because several
M 10E3, IV-2,4-68
nocturnalcensuseslocatedno new species,we felt
z
that all colonistshavingcrepuscularor nocturnal
activitywere being recordedduring the diurnal
5-J
monitoring. The anyphaenidspider Aysha sp.
and ant Paracryptocerusvarians are both strictly
0
nocturnal,yet both were recordedduringthe day
in their retreats. Similarlythe crepuscularbra5
30
15
20
25
l0
conidwasp Macrocentrussp. and largelynocturnal
SEARCHING
'NUMBER OF MAN-HOURS
FIG. 10. Cumulativespeciescountsduringindividualcen- ants Camponotusfloridanusand Tapinoma littorale were repeatedlyobserved.
suses on threeof theexperimental
islands.
278
0. WILSON
Ecology,Vol. 50,No. 2
Hermannsson,S. 1967. Introduction. Surtsey Research ProgressReport,III: 1.
We thank the United States Department of the Interior
or permissionto fumigatethe islands of this experiment. LaRue, D. C., and T. T. Muzik. 1954. Growth,regenerationand precociousrootingin Rhizophoramann particular, Jack Watson of the Great White Heron
gle. Pap. Mich.Acad. Sci. Arts Lett. (I) 39: 9-29.
national Wildlife Refuge, Roger W. Allin and William
3. Robertson of the Everglades National Park, and Lauck, D. R. 1965. Chloropicrinfor fast action with
a Berlese funnel. Turtox News 43: 115.
Zobert M. Linn of the National Park Service examined
Lindroth,C. H., H. Andersen,H. Bodvarsson,and S. H.
he plans of the experimentand concluded, with the auRichter. 1967. Report on the Surtsey investigation
hors, that the total mangrove faunas would be unaffected.
in 1966. Terrestrialinvertebrates.SurtseyResearch
;teve Tendrich, Vice-President of National ExterminaProgress Report,III: 59-67.
ors, Inc. (Miami) coped imaginatively and effectively
MacArthur,R. H., and E. 0. Wilson. 1963. An equiviththe unique problemspresentedby the project. W. H.
libriumtheoryof insular zoogeography. Evolution
3ossert and P. B. Kannowski assisted in the original
DANIEL
S. SIMBERLOFF
AND EDWARD
ACKNOWLEDGMENTS
ensuses. Maps and tables for all parts of this series
vere done by Mrs. Helen Lyman. The work was suported by NSF grant GB-5867.
17: 373-387.
MacArthur,R. H., and E. 0. Wilson. 1967. The theory
of island biogeography.PrincetonUniversityPress.
203 p.
CITED
LITERATURE
Mainwaring,A. P. 1961. Some effectsof methylbroChisholm,R. D. 1952. Nature and uses of fumigants, mide on aphids and whiteflyand their host plants.
p. 358. In Insects,The Yearbook of Agriculture1952.
Doctoral thesis,London Univ., London, England.
Darlington,P. J. 1957. Zoogeography:the geographi- Monro,H. A. U., and R. Delisle. 1943. Furtherappliof animals. Wiley,New York. 675 p.
cal distribution
cations of methylbromideas a fumigant. Sci. Agr.
Davis, J. H. 1940. The ecology and geologic role of
23: 546-556.
mangrovesin Florida. Papers fromTortugas Lab.,
Page, A. B. P., and 0. F. Lubatti. 1963. Fumigation
Carnegie Inst. Wash. 32: 307-412.
of insects. Ann. Rev. Entomol.8: 239-264.
French, R. A. 1964 (1965). Long range dispersal of
D. S., and E. 0. Wilson. 1969. Experiinsects in relation to synopticmeteorology. Proc. Simberloff,
mentalzoogeographyof islands. The colonizationof
Int. Congr.Entomol.12th(London). 6: 418-419.
emptyislands. Ecology 50: 278-296.
on the
Fridriksson,S. 1967. Life and its development
D. 0. 1946. Dispersionof small orgavolcanic island, Surtsey. Proc. Surtsey Research Wolfenbarger,
nisms. Amer. Midland Naturalist35: 1-152.
Conf.,1967: 7-19.
ZOOGEOGRAPHY OF ISLANDS:
EXPERIMENTAL
OF EMPTY ISLANDS
COLONIZATION
THE
DANIEL S. SIMBERLOFF1 AND EDWARD' 0. WILSON
Cambridge,Massachusetts02138
The Biological Laboratories,Harvard University,
(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 immigrationand extinction
processes of animal species based on direct observations.
The colonization of six small mangrove islands in Florida Bay by terrestrial arthropods
was monitoredat frequentintervals for 1 year after removal of the original fauna by methyl
bromide fumigation. Both the observed data and climatic considerationsimply that seasonality
had little effectupon the basic shape of the colonization curves of species present vs. time.
By 250 days after defaunation,the faunas of all the islands except the most distant one ("El")
had regained species numbers and compositionsimilar to those of untreatedislands even though
population densities were still abnormally low. Although early colonists included both weak
and strong fliers,the former,particularlypsocopterans,were usually the firstto 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 differentislands.
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 interaction becomes important,and finally,a lasting dynamic equilibrium. Equations are given for
the early immigration,extinction,and colonization curves.
Dispersal to these islands is predominantlythrough 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 commonlyby competitionand predation.
Present address: Department of Biological Science, Florida State University, Tallahassee, Florida 32306