the amphibians and reptiles of the estacio´ n biolo´ gica jatun sacha in the lowland rainforest of

30 Pages·2008·1.71 MB·English

USISSN0006-9698
CAMBRIDGE, MASS. 17 DECEMBER 2008 NUMBER 514
THE AMPHIBIANS AND REPTILES OF THE ESTACIO´N BIOLO´GICA JATUN
SACHA IN THE LOWLAND RAINFOREST OF AMAZONIAN ECUADOR:
A 20-YEAR RECORD
GREGORY O. VIGLE1
ABSTRACT. The amphibian and reptile fauna within a small area in a lowland rainforest fragment reserve in
Amazonian Ecuador was intensively surveyed over an initial 2- year period via removal sampling (1986–88),
coincidingwiththeconstructionofaroadthroughtheareaandasubsequentsurgeofincreasedforestconversion
andfragmentationinsurroundingareas.Atime-constrainedtransectsamplingtechniquewasemployedtofacilitate
later,post-fragmentationvisual-encountermonitoringtogatherlong-termdataonspeciesrichnessandcomposition
(onachronologicallycoarsescale).Between1998and2005,anapproximatelyequivalentquantityofsamplingeffort
wasaccumulatedinthesamesmallareatocomparewithdatafrom1986–88,inanattempttoevaluatetheefficacyof
thisreserveinconservingtheherpetofaunalcommunitydocumentedmorethan10yearsearlier.
Basedonatotalof6,722individualrecordsobtainedover1,117effort-hoursofsamplingdividedintotwoprimary
andfivesecondarysamplingperiods,herpetofaunalspeciesrichnesswasamongthehighestyetreportedfromasingle
localityinAmazonia,with84amphibianand82reptilespeciesrecordedfrom1986through2007.Acompletespecies
listisprovided,withcommentaryonnoteworthyrecords.Of73amphibianand46reptilespeciesrecordedin1986–88
sampling,68amphibians(93%)and40reptiles(87%)wererecordedagaininsamplingfrom1998through2007;11
amphibianand33reptilespeciesnotrecordedin1986–88wereaddedfrom1998to2007.Pairwisecomparisonsof
species composition among all sampling periods using a similarity index revealed a . 50% similarity for both
taxonomic groups across all periods; similarity of the two primary periods (1986–88 vs. 1998–2005) was 0.90 for
amphibiansand0.70forreptiles.
The present study provides another single-site reference point for the mega-diverse herpetofauna of the upper
Amazon basin, but is distinct in offering long-term data on species persistence; results from the 20-year span of
samplingsuggestthatthissmallreservehas,since1986,successfullyconservedthisfauna.Theseresultsconcurwith
previouslong-termstudiesfromthecentralAmazonbasininsuggestingthatrelativelysmallreservesandotherforest
remnants with sufficient habitat diversity may substantially contribute to the continued survival of species-rich
Amazonrainforestherpetofaunalcommunitiesamidstwidespreadregionaldeforestation.
KEYWORDS: Amazon;amphibians;biodiversity;conservation;Ecuador;Estacio´nBiolo´gicaJatunSacha;forest
fragmentation;herpetofauna;monitoring;reptiles;reserve;speciesrichness;tropicalrainforest
1School of Life Sciences, Arizona State University, Tempe, AZ 85287-1501, U.S.A.; e-mail: vigle@imap2.
asu.edu,[email protected]
EThePresidentandFellowsofHarvardCollege2008.
2 BREVIORA No.514
RESUMEN. Los anfibios y reptiles que se encuentran en un a´rea pequen˜a dentro de una reserva fragmentada de
bosquelluviosoenlaAmazon´ıaecuatorianahansidointensivamenteinvestigadosenunper´ıodoinicialdedosan˜os
mediantelarecoleccio´ndemuestras(1986–88).Estaactividadcoincidio´ conlaconstruccio´ndeuncaminoatrave´s
dela´reaylasubsiguienteoleadadeconversio´nyfragmentacio´ndelbosqueenlasa´reascircundantes.Seempleo´ una
te´cnicademuestreodetransectosdetiemporestringidoparafacilitarluegoelmonitoreodeencuentrovisualpost-
fragmentacio´nyreunirdatosdelargoplazosobrelariquezaycomposicio´ndelasespecies(enunaescalacronolo´gica
desigual). Entre 1998 y 2005 se acumulo´ un esfuerzo de investigacio´n y muestreo equivalente en la misma a´rea
reducidaparacompararlaconlainformacio´nde1986a1988,enunintentoporevaluarlaeficienciadeestareserva
enlaconservacio´ndelafaunaherpetolo´gicadocumentadama´sde10an˜osatra´s.
Basada en un total de 6,722 registros individuales obtenidos durante 1,117 horas de recoleccio´n de muestras
divididas en dos per´ıodos principales y cinco per´ıodos secundarios de muestreo, la diversidad herpetofaun´ıstica
estuvoentrelasma´saltasreportadashastaelmomentoenunau´nicaubicacio´nenlaAmazon´ıa,con84especiesde
anfibios y 82 especies de reptiles registrados de 1986 a 2007. Se provee una lista completa de especies, con
comentariosenlosregistrosma´ssignificativos.Delas73especiesdeanfibiosy46especiesdereptilesregistradasde
1986a1988,68anfibios(93%)y40reptiles(87%)fueronregistradosnuevamenteenelmuestreode1998a2007;11
especies de anfibios y 33 especies de reptiles no registrados entre 1986 y 1988 fueron an˜adidos en 1998 a 2007.
Medianteunana´lisisdecomposicio´ndeparejaentretodoslosper´ıodosdemuestreousandounindicedesimilitud
revelaronun.50%desimilitudparaambosgrupostaxono´micosentreper´ıodos;lasimilituddeestosdosper´ıodos
principales(1986–88vs.1998–2005)fue0.90poranfibiosy0.70porreptiles.
Elpresenteestudioproveeotropuntodereferenciadesitiou´nicoparalaherpetofaunamegadiversadelazonaalta
de la cuenca amazo´nica, pero es diferente ya que ofrece datos de largo plazo en la persistencia de especies; los
resultados de un lapso de muestreo de 20 an˜os sugieren que esta pequena reserva, desde 1986ha conservado
exitosamenteestafauna.Estosresultadoscoincidenconestudiospreviosdelargoplazadelacuencacentraldela
Amazon´ıaensugerirquelasreservasrelativamentepequen˜asyotrosremanentesdebosqueconsuficientediversidad
de ha´bitat pueden contribuir sustancialmente a la continua supervivencia de la alta diversidad de comunidades
herpetofaun´ısticasdelosbosqueslluviososamazo´nicosenmediodelaextendidadeforestacio´nregional.
INTRODUCTION Lamar, 2004; Dixon and Soini, 1986; Duell-
man, 1999; Frost, 2007; Peters and Donoso-
The rainforests of the Amazon basin Barros, 1970; Peters and Orejas-Miranda,
harbor the highest concentration of biodi- 1970).Reportsonsingle-sitestudieshavebeen
versityonEarth,yettheyarebeingdestroyed reviewedbyDuellman(1988,2005)andDuell-
byhumanactivitiesatanalarmingpace;one man and Thomas (1996) (also see Doan and
prominent threat is deforestation and the Ariza´bal, 2002; Lynch, 2005). Within the
fragmentation of previously continuous for- immediate area of the Ecuadorian Amazon,
estintosmallerislandremnantsofremaining published studies on the entire herpetofauna
habitat (Bierregaard et al., 2001; Lovejoy et are available for several single or clustered
al., 1984; Myers, 1984). sites, including Almenda´riz (1987), Cisneros-
Regionally,andonthebasisofsurveysfrom Heredia (2003), Duellman (1978), Duellman
singlelocalities,thelowlandrainforestsofthe andMendelson(1995),Izquierdoetal.(2000),
westernAmazonbasinofEcuador,Peru,and andLescureandGasc(1986).
Colombiaarehometothehighestconcentra- From these studies, there has been an
tionofamphibianandreptilespeciesrichness increasing awareness of the geographic and
knownfromanyareaonEarth(Campbelland ecological heterogeneity of the Amazonian
2008 AMAZONFORESTFRAGMENTHERPETOFAUNA 3
herpetofauna, and the need to survey addi- ous studies from single upper Amazonian
tional localities to better understand both sites were conducted over relatively short
local and regional distribution patterns of spans of time (5 years or less), and within
this mega-diverse but still poorly known largely undisturbed expanses of continuous
fauna (Doan and Ariza´bal, 2002; Duellman, forest. The present paper is distinct in
2005; Lynch, 2005). presenting data on the long-term composi-
The Estacio´n Biolo´gica Jatun Sacha tion of the herpetofaunal community within
(EBJS) is a small remnant of lowland a forest fragment remnant; as such, it is
tropical wet forest located at the western offered as another single-site reference point
edge of the Amazon Basin in eastern for the upper Amazon, in the form of a
Ecuador, established as a reserve in 1985 traditional species checklist, and as a neces-
shortly before the construction of a road sary preface to forthcoming accounts on
through the area. species-level abundance and distribution
In 1986, coinciding with the onset of road derived from the same sampling efforts.
construction, an intensive survey of the This study was undertaken several years
amphibian and reptile fauna was initiated. before widespread interest in reports of
Although the first goal was to conduct a amphibian declines (Blaustein, 1994; Stuart
standard inventory via removal sampling, et al., 2004); despite concerns over various
the ultimate fate of this new reserve as ‘‘enigmatic’’ declines with causal factors
eventually becoming an isolated habitat ranging from epidemic disease to global
island was recognized at the time, and a climate change, the effects of habitat change
simple time-constrained transect sampling are understudied and perhaps underempha-
method was implemented to facilitate future
sized for both amphibians and reptiles
long-term monitoring via visual-encounter
(Gardner et al., 2007), especially in the
surveys (VES) and to attempt to assess the
Neotropicallowlands,wherelong-termmon-
long-term efficacy of the reserve in conserv-
itoring data remain very rare (Duellman,
ing the herpetofauna. A general initial
2005:49; Whitfield et al., 2007:8355). The
prediction was that species richness would
EBJS provides an established lowland re-
eventually decline or otherwise exhibit obvi-
serve with a well-known history as a forest
ous changes in community composition.
fragment, and is an appropriate site for
Between1998and2005,asecondphaseof
monitoring the effects of habitat change on
sampling was undertaken, employing the
an exceptionally species-rich Neotropical
same methods over the same small area,
herpetofauna.
but with most individual records gathered
via VES efforts and few collected aside from
METHODS
road-kill specimens. The 1986–88 and
1998–2005 periods are treated as equivalent Thefollowingsummaryofmethodologyis
‘‘primary’’ units for comparative purposes, divided into three sections, progressing from
and each is further divided into equivalent (1) a detailed description of the physical
shorter ‘‘secondary’’ periods on the basis of characteristics, history, and climate of the
an arbitrary (and partially climate-based) EBJS; to (2) an account of the long-term
division of total sampling effort (see Meth- division of sampling effort and major goals
ods). within the context of this history; and to (3)
With the exception of Cusco Amazo´nico thedescriptionofspecificsamplingprotocols
in Peru (Duellman, 1995, 2005), most previ- employedinthefieldaswellasthescopeand
4 BREVIORA No.514
deposition of specimens and records thus
obtained. Online material will be available
through the journal’s website.
I) Description of the study site: forest,
history, and climate
The EBJS is located in eastern Ecuador in
NapoProvinceatelevationsof400to450 m
(016049S, 776379W at main station complex;
see Figs. 1–4); the reserve is owned and
administered by Fundacio´n Jatun Sacha
(FJS), an Ecuadorian conservation founda-
tion. The EBJS was the first of several
reserves established by FJS and originally
(1985–88) consisted of 230 ha of preserved
forestowned bythreeindependentlandown-
ers.Afteraninitial150-haexpansionfunded
by the Grateful Dead music group in 1989
(via their Rex Foundation), subsequent
acquisitions of additional land have greatly
increased the size of the reserve, and current
total land holdings are 2,334 ha.
Figure 1. CurrentlandholdingsoftheEBJSreserve
The EBJS forest is classified as Tropical
(gray shaded areas within dark borders); location in
Wet Forest (Canadas, 1983; Holdridge, eastern Ecuadoris indicatedby arrowon inset mapat
1967) and lies between the confluence of top left. Star indicates EBJS main station complex at
the Napo River to the north and smaller approximately 016049S, 776379W. The area located
between land tracts A and B comprises the Central
Arajuno River to the south (Figs. 1, 4).
Study Area (CSA; see Fig.2); tract A is cleared cattle
Forest within reserve boundaries consists of pasture, whereas tract B is a mixture of agricultural
approximately 70% primary forest, with the fields and various secondary forest successional stages.
remaining 30% composed of various succes- AlsoseetextandFigures3,4.Approximatelytoscale.
sional stages of secondary forest growth,
including approximately 95 ha of open-for- (datingfrom2000and2001)arepresentedin
mation riparian habitats (Fig. 3). Figure 4 and online Figure 1.
AlthoughtheEBJSpreservesasubstantial Before the early 1980s, the region was
tract of lowland rainforest, since 1985 the sparsely populated and accessible only by
upper Napo region has undergone substan- trails or river transport, although forest
tialchangesresultingfromdeforestationand conversion through small-scale logging and
forest fragmentation because of increasing agriculturalclearinghadalreadyaffectedthe
colonization and development. Most adja- area for many years, especially along the
cent bordering areas have been logged or banks of both major rivers. In March 1986,
affected by agricultural activities, although the construction of a road (from the west)
noprecisesurveydataareavailable.Satellite reached the north-central area of the EBJS
images showing general overviews of forest (Figs. 1–4), and ongoing construction now
cover at the EBJS and in adjacent areas extends approximately 30 km further east.
2008 AMAZONFORESTFRAGMENTHERPETOFAUNA 5
reserve is relatively small, the larger regional
effect of increased development is the pre-
dominant threat, as the EBJS has increas-
inglybecomean‘‘island’’reservesurrounded
by fragmented forest.
The present study reports primarily on
samplingconductedinasmall(approximate-
ly 100 ha) central study area (CSA) border-
ing the Napo River (Figs. 1–4); the CSA
includesallmajorhabitatsrepresentedinthe
area (primary forest, and various river-
floodplain forests, swamps, and secondary
forest; Fig. 3), in contrast to land acquisi-
tions since 1990, which have focused almost
entirely on additional primary forest to the
south and west. This concentration of forest
types provides the most appropriate site to
sample the overall amphibian and reptile
diversity, in that some species occupy only
limited portions of the various habitats
within the EBJS as a whole.
The CSA is also the area most directly
affected by the road (and forestclearing and
Figure 2. The Central Study Area (CSA) in the fragmentation in adjacent areas) for the
north-central region of the EBJS located between two
longestperiodoftime.Twodeforestedtracts
deforestedlandtracts,AandB(alsoseeFigs.1,4,and
of land, in total area approximately equal in
text).Allsamplingeffortsassummarizedinthetextand
onlineTable1wereconductedontrailroutes(irregular sizetotheCSA,borderthisareadirectlyand
linesincentralarea)extendingnorthandsouthfromthe entirely to both the east and west (see
EBJS main station complex (star) and north from the Figs. 1, 2, 4); both of these adjacent areas
LibrariusStudyCentercomplex(solidcircle),aswellas
had already been partially cleared before
along the margins of the road to a distance of
1986, although the largest (to the west) had
approximately 300m to the east and west of the main
station. The gray-shaded area is shown in detail in undergone 10 years of secondary growth
Figure3.Approximatelytoscale. before being cleared again in 1990.
Thus, although the total area of preserved
The discovery of oil reserves in the region forestistodayfarlargerbytheacquisitionof
hasacceleratedtheprocessofcolonizationin additional land, in keeping with the long-
recentyears, constitutinganincreasedthreat term goals of the study begun in 1986 (see
totheEBJSreserveintheformofadditional below), subsequent sampling efforts have
road traffic, and deforestation in nearby focusedalmostexclusivelyonthissamesmall
areas. In November 2003, a high-pressure area.
crudeoilpipelinewasinstalledalongtheside The Ecuadorian Amazon region receives
oftheroad(buriedtoadepthof2 m),andin one ofthehighest levelsofrainfall within all
early2006,theroadwaswidenedby4 mand of lowland Amazonia. In contrast to more
paved with asphalt. Although the immediate easterly areas, however (exhibiting pro-
area affected by the road itself within the nounced wet and dry seasons), climate at
6 BREVIORA No.514
Figure 3. Map (bottom, approximately to scale), and diagrammatic cross section (top) showing relief and
habitatdistributionwithintheCSA(alsoseeFigs.1,2,4).All1986–88periodsamplingwasconductedwithinthis
narrowplotandalongthemarginsoftheroadtoadistanceof,300mtotheeastandwest;approximately70%of
1998–2005periodsamplingeffortwasalsoconductedwithinthissamesmallarea.AreasAtoDformthefloodplain
oftheRioNapo,areaEincludestheroad(R)andearlysuccessionalgrowthandephemeralponds(EP)alongthe
roadmargins,andareaFisclosed-canopyprimaryforestonsteeplyrollinghillsbisectedbysmallstreams(S).Within
thefloodplainregion,areaAisanephemeralfluvialswampsurroundedbycroptreesandsuccessionalgrowth;area
Bwas(until1990)clearedandplantedwithcornevery,6months;areaCisoldersecondaryforestwithmanylarge
trees;areaDisapermanent(flowing)fluvialswampwithlowHeliconiagrowth.Asidefromincreasedsecondary
growthinareasA,B,andC(secondaryforestinareaBisnowcontinuouswithareaC),thishabitatdistributionhas
remainedlargelyunchangedsince1986.
theEBJSisfarlessseasonal,andrainfallcan and 210 mm, respectively) was lower than
be highly irregular within any given year. both November and December (335 mm in
The yearly distribution of rainfall is relevant each month).
to the distribution of sampling effort for the
purpose of comparing faunal records be- II) Temporal distribution of sampling effort
tween sampling periods (discussed below). and major goals
RainfallhasbeenrecordedattheEBJSsince
1986,althoughdataareincompleteforsome A Coarse-Scale Sequence from ‘‘Before’’
years; for one 10-year period with complete to ‘‘After.’’ Collections and records of
data for each month (1994–2003; see online observations of the amphibian and reptile
Table 1), total rainfall per year averaged fauna of the EBJS include two major
3,711 mm(rangingfrom3,051to4,593 mm). (primary) sampling periods, each including
This record also indicates that March– approximately equal sampling effort, the
August are on average the wettest months, first from 15 March 1986 to 15 February
and September–February show the lowest 1988 (536.0 hours total effort) and the
average rainfall. Within any single year, second spanning 1 July 1998 through De-
however, this general pattern can vary cember 2005 (581.0 hours). Each primary
greatly; for example, in 1996 January had samplingperiodisarbitrarilysubdividedinto
the highest rainfall of any month (510 mm), secondary sampling periods (discussed be-
and in 1997 rainfall for June and July (255 low). The two primary sampling periods
2008 AMAZONFORESTFRAGMENTHERPETOFAUNA 7
Figure 4. Landsat 7 satellite image (30m/pixel resolution), captured 9 September 2001, showingforest cover
withintheEBJSreserveandcloselyadjacentareas.RefertoFigures1–3forlocationofmajorlandscapefeatures
relativetoreserveborders;thetwooutlinedrectanglesapproximatelyindicatedeforestedareasAandBtotheeast
andwestoftheCSA.AlsoseeonlineFigure1showingforestcoveroveralargerarea.Atright,adultRanitomeya
duellmani(Dendrobatidae),shown1.53naturalsize.
reported here can be interpreted as ‘‘pre- to assess the overall stability and similarity
fragmentation’’ (1986–88) and ‘‘postfrag- of the community over time (on a chrono-
mentation’’ (1998–2005) data on the EBJS logically coarse scale). This is an ongoing
fauna, although as previously noted, forest process; hypothetically, each sampling pe-
conversion processes in the area had been riod(andallperiodscombined)canprovide
ongoing even before 1985. a predictive ‘‘before’’ reference point to
An initial program of removal sampling compare with subsequent equivalent sam-
from 1986 to 1988 was undertaken as a pling efforts ‘‘after,’’ as one measure of the
baseline inventory of the fauna to facilitate efficacy (or lack thereof) of the EBJS
subsequent monitoring of the same area in reserve in preserving the herpetofauna at
lateryears(1998–2005)viaVESefforts,with this increasingly isolated site (the third
three major goals. The first goal was to major goal, with potential larger scale
generate an initial species list (and a record relevance for conservation in the region as
of museum specimens), while employing a a whole). Sorenson’s quotient of similarity
sampling technique that could be accurately (SQ 5 2J[A + B]) is used to compare the
repeated in later years. percent similarity among all sampling
A second goal was to accumulate equiv- periods (see Results and Table 2). This
alent amounts of sampling effort for each simple formula estimates similarity inde-
sampling period, with the use of results pendent of the individual abundance of
fromeachtomonitortrendsinthepresence each species, where A is the number of
orabsence(persistence)ofthemanyspecies species in sample 1, B is the number in
8 BREVIORA No.514
TABLE1. DISTRIBUTIONOFNUMBERSOFAMPHIBIANANDREPTILESPECIES(TOP)ANDINDIVIDUALSPECIMENRECORDS
(BOTTOM)FORALLSAMPLINGPERIODSFROMTHEEBJS(ALSOSEETEXT,APPENDIX1,ANDONLINETABLE1).SPECIES
TOTALSFORTHE1998–2005PRIMARYPERIODINDICATETHENUMBEROFSPECIESRECORDED,FOLLOWEDBYTHENUMBEROF
ADDITIONALSPECIES(+)RECORDEDLATERFROM2006OR2007.FARRIGHTCOLUMNPRESENTSTHETOTALNUMBEROF
SPECIESFROMALLSOURCES(INCLUDING‘‘R’’SPECIES)FROM1986THROUGH2007,ASLISTEDINAPPENDIX1.INDIVIDUAL
SPECIMENTOTALSINDICATETHENUMBEROFINDIVIDUALSRECORDEDINEACHPERIOD;NUMBERSINPARENTHESESINDICATE
THEINDIVIDUALSPEREFFORT-HOURFOREACHPERIOD.
No.bySamplingPeriod
Total Total Total:AllSources
1986–19871987–19881986–19881998–20012002–2003 2004–2005 1998–2005* 1986–2007*
Species:
Amphibian: 63 63 73 63 54 61 78(+1) 84
Reptile: 33 39 46 34 31 55 66(+7) 82
Total: 96 102 119 98 86 116 144(+8) 166
Individuals:
Amphibian: 1,298 1,329 2,627 1,407 989 831 3,227 5,854
(4.87) (4.92) (4.90) (6.54) (5.47) (4.47) (5.55) (5.24)
Reptile: 152 170 322 157 155 234 546 868
(0.57) (0.62) (0.60) (0.73) (0.85) (1.26) (0.93) (0.77)
Total: 1,450 1,499 2,949 1,564 1,144 1,065 3,773 6,722
*Individualtotalsinthe1998–2005and1986–2007columnsdonotincluderecordsfrom2006and2007.
sample 2, and J is the number of species high rainfall (but also see Summary and
common to both samples. These pairwise Discussion).
comparisons result in SQ values between 0 Although the division of sampling effort
and 1, 0 indicating no similarity and 1 presented in online Table 1 results in two
indicating complete similarity (SQ 3 100 5 primaryperiodsapproximatelyequalintotal
% similarity). effort-hours, the second primary period
Additionalgoalsweretoassesspopulation spanned a far longer number of years. The
fluctuations by gathering time-series abun- arbitrary division of sampling effort as
dance (and ecological distribution) data for shown in Table 1 and online Table 1 is a
all species individually at the EBJS. The compromise that uses theavailable sampling
description and evaluation of species-level effort and climate data to provide the most
patterns of abundance and distribution are, equivalentperiodsforcomparativepurposes.
however,beyondthescopeofthispaper,and The first primary sampling period is
Table 1 presents data solely on total num- divided into two secondary periods (see
bers of individuals in each major group online Table 1), providingtwo approximate-
(amphibians and reptiles) from each sam- ly equal units (in total effort-hours); this
pling period. Interpretation of this coarse- division also approximately corresponds to
scale data is problematic, in that multiple each of two annual climate periods (see
variables influence abundance of the com- below).
ponent species independently, especially The second, much longer primary sam-
among amphibians, with fluctuations in pling period from 1998 to 2005 is divided
activity and apparent abundance often de- intothree secondary periods,varyinggreatly
pendent on brief and irregular periods of in the monthly distribution of effort-hours
2008 AMAZONFORESTFRAGMENTHERPETOFAUNA 9
TABLE2. PAIRWISECOMPARISONSOFALLPRIMARYANDSECONDARYSAMPLINGPERIODSFROMTHEEBJSFORAMPHIBIANS
ANDREPTILES.THETOTALNUMBERSOFSPECIESFROMEACHPERIODAREINDICATEDINBOLDFACEINTHECOMMONCELL,
ANDTHENUMBERSOFSPECIESSHAREDBYPERIODSAREINDICATEDATTHEUPPERRIGHT.SIMILARITYQUOTIENT(SQ)VALUES
AREINDICATEDINITALICSATLOWERLEFT.
SamplingPeriod 1986–1987 1987–1988 1998–2001 2002–2003 2004–2005 1986–1988 1998–2005
Amphibians
1986–1987 63 53 53 47 51 — 59
1987–1988 .85 63 53 45 49 — 60
1998–2001 .83 .83 64 45 48 59 —
2002–2003 .80 .77 .76 54 51 51 —
2004–2005 .82 .79 .77 .89 61 57 —
1986–1988 — — .86 .80 .85 73 68
1998–2005 .84 .85 — — — .90 78
Reptiles
1986–1987 33 26 23 17 25 — 30
1987–1988 0.72 39 25 19 34 — 34
1998–2001 0.69 0.68 34 18 27 27 —
2002–2003 0.53 0.54 0.55 31 24 20 —
2004–2005 0.57 0.72 0.61 0.56 55 33 —
1986–1988 — — 0.67 0.52 0.65 46 39
1998–2005 0.61 0.65 — — — 0.70 66
each year (online Table 1); this division, (consisting mostly of amphibians) and those
however, results in the most equal partition- diurnal species of both amphibians and
ing of the total sampling effort from the reptiles that typically rest at night within
1998–2005 period. the areas searched.
Relative to annual climate variation, both Sampling at night consisted of carefully
the 1986–88 and 1998–2005 primary periods searching the ground and vegetation, with
each include in total approximately equal spotlights,alongtheperimetersofestablished
proportions of sampling effort expended trails(transects)fromgroundleveltoaheight
duringthewettermonths(MarchtoAugust) of approximately 2 m aboveground and to a
and drier months (September to February) distance of approximately 2 m on either side
(see online Table 1). of each trail. Although the actual lengths of
trail segments searched per time unit of
sampling effort were generally consistent
III) Field sampling protocol, museum and
(typically on the order of approximately 300
VES records, and taxonomy
linearmetersperhour),thebaseunitusedfor
Most amphibian species at the EBJS are comparisonsbetween samplingperiodsisthe
nocturnal; some reptiles are also nocturnal, person-effort-hour, defined as the hourly
whereas others may encountered resting at searcheffortexpendedbyasingleexperienced
night, and sampling effort was almost person using the methods noted above. This
exclusively concentrated on searches at wasthusaformoftime-constrainedsampling
night, with very little total effort (24 hours; technique (Campbell and Christman, 1982;
2.1%) expended by day (these hours and Crump and Scott, 1994), applied to both
records are lumped together with night removal and VES sampling efforts to facili-
sampling efforts). Thus, this study focused tate comparisons between sampling periods.
primarily upon the nocturnal herpetofauna Mostnightlysamplingperiodsrangedfrom2
10 BREVIORA No.514
to 5 hours in duration, and were conducted From August 2003–December of 2005, a
between19:00and24:00. large portion of the total sampling effort
Duringthefirstprimaryperiod(1986–88), (231.5 hours) was conducted by Park
most specimens of most species encountered Guards employed by the EBJS. In a project
were collected and preserved, with 2,810 of funded by the Declining Amphibian Popu-
thetotal2,949individualrecordsrepresented lations Task Force (DAPTF), park guards
by museum specimens. Collecting is now weretrainedon-siteinsearchtechniquesand
restricted at the EBJS, and records from the theuseofdigitalcameras.ParkGuardsthen
second primary period (1998–2005) are continued periodic sampling in the author’s
basedalmostentirelyonVESefforts,includ- absence, and images of all individuals
ing a total of 3,773 individuals. With only a encountered were saved on compact disc
fewexceptions(seeAppendix1andResults), (CD) for later review and identification
theauthorpersonallyidentifiedallspecimens (1,065 identified individuals).
(from all sampling periods) via direct obser- Although most of the 6,722 individuals
vation or examination of photographs (see from all sampling periods were recorded
below). during regularly measured sampling efforts,
Of the total 1,117 effort-hours of sam- a total of 376 records (5.6%) represent
pling, 635 hours (56.8%) were conducted individuals that were randomly encountered
personally by the author, whereas the re- outside the 1,117 hours summarized in on-
maining 482 hours (43.2%) were conducted line Table 1. Appendix 1 also includes all
by persons who were trained on-site and species known to the author from the EBJS
supervised by the author, termed student since 1986 from all sources, including addi-
collectors (SCs). Relative proportions of tionalrecordsfrom2006and2007,andafew
sampling effort by the author versus SCs important records from other investigators
were approximately equal for each primary (see Appendix 1, Table 1, and Results).
period. Persons who assisted as SCs (and Themajority(75%)ofspecimenscollected
EBJS Park Guards), and their respective during the 1986–88 removal sampling effort
sampling effort contributions, are noted in are deposited in the herpetological collec-
the Acknowledgments. tions of the Museum of Comparative Zool-
For all sampling before August 2003, the ogy at Harvard University (MCZ); the
author was present on-site with all SCs and remainder are housed in other collections
reviewed the results of each night’s sampling (primarily in Ecuador). The following list of
efforts immediately; for the 1986–88 period, additionalmuseumcollectionsispresentedin
andfrom1998to2001,allspecimensencoun- approximate decreasing order of the total
tered by SCs were captured and then identi- number of included specimens at each
fied (in 1986–88, most were preserved; in museum (with acronyms): Museo Ecuator-
1998–2001,allwerereleasedwithin48 hours). iano de Ciencias Naturales, Quito (MECN);
Beginning in March 2002, the physical Museo de Zoolog´ıa, Pontifica Universidad
capture of specimens by SCs was discontin- CatolicadelEcuador,Quito(QCAZ);Amer-
ued and digital cameras were used to ican Museum of Natural History (AMNH);
photograph every amphibian and reptile Florida Museum of Natural History (UF);
specimen encountered; all other aspects of United States National Museum (USNM);
sampling methods remained the same, and University of Kansas Museum of Natural
identifications were determined solely from History (KU). All specimens collected from
photos. 1998 to 2007 are deposited at the QCAZ (all