Population biology and wing color variation in Heliconius erato phyllis (Nymphalidae)

11 Pages·1999·5.5 MB·English

by  Renato Rogner Ramos

JournaloftheLepidoptelists'Society
53(1),1999,11-21
POPULATION BIOLOGYAND WING COLORVARIATION IN
HELICONIUSERATO PHYLLIS (NYMPHALIDAE)
Renato Rogner Ramos
Av.FranciscoGlicerio615/33, 11065-405, Santos, SaoPaulo, Brazil
AND
Andre Victor Lucci Freitas
MuseudeHistoriaNatural, InstitutodeBiologia, UniversidadeEstadualdeCampinas,CP6109, 13083-970,Campinas, SaoPaulo, Brazil*
ABSTRACT. Twenty-sixmonthsofmark-recapturestudyofthreepopulationsofHeliconiuseratophyllisinSE Brazilshowedthesepopu-
lationstoberelativelystablethroughtime.Theadultsshowedhighlongevity,similartothatofothertropicalHeliconius. Sexratiowasmalebi-
ased,andmalesshowedlongerresidencetimesthanfemales.Thenumberofredrayletsontheventralhindwingshowedseasonalvariation,con-
sideredtobenumericpolyphenism. OfallthespeciesofHeliconius, H. eratoisproposedtobethemosttolerantofvariationinhabitat, larval
andadultresources ("ecologicalplasticity").ThesetraitsareproposedasreasonsforthewidedistributionofthisspeciesintropicalAmerica.
Additionalkeywords: mark-recapture, Heliconiini,Passifloraceae,redraylets,polyphenism.
Even ifHeliconiine butterflies are amongthe best- StudySitesand Methods
studiedtropicalinsects, with muchknown abouttheir
systematics, population ecology, behavior, immature A mark-release-recapture (MRR) studywas carried
biology, hostplantrelationships, coevolution, mimicry, out in three areas in Sao Paulo state, southeastern
chemistry, genetics, and conservation (Ehrlich and Brazil. The main study areawas "Morro do Voturua"
Gilbert 1973, Brown & Benson 1974, Gilbert 1975, (MV) (46°22'W, 23°57'S), in the city of Sao Vicente
Araujo 1980, Brown 1981, Turner 1981, Sheppard et (Fig. 1). The site was originally coveredwith lowland
al. 1985, Mallet 1986a, 1986b), data are not available subtropical rainforest (Ururahy et al. 1987). The an-
mm
formost species andgeographicpopulations ofHelico- nual rainfall reaches 2500 andthe average annual
nius; and generalizations based on other well-known temperature is 21°C (Setzer 1949, Prodesan 1969,
species and regions may fail to explain local patterns Nimer 1972), with the mean of the coldest month
andprocesses. Forexample, thecurrentideathatHeli- 18.2°C and of the warmest month 25.3°C (Santos
conius maintain relatively stable population numbers 1965) (Fig. 2, methodsfollowingSantos 1965 andWal-
through time (based mainly on Turner 1971 and ter 1985). Most of the area is covered by secondary
Ehrlich&Gilbert 1973)isnottrueforHeliconiuserato forestonlowhills (100-200melevation). Similarwork
phyllis (F.) near the limits ofits range in temperate was done in two nearby sites: the "Vale do Rio Qui-
southern Brazil (Saalfeld &Araujo 1981). lombo" (VQ, Fig. 1), a road along a river valley with
Heliconius erato is the most widespread species of much secondaryvegetation and flowers, 12 field days
thegenus,presentinseveraldifferenthabitatsandfor- overthreemonths; andthe"MorrodoJapui" (MJ, Fig.
SW MV
est types from Mexico to north Argentina (Emsley 1), alargehillfacingthe ocean of with similar
1964, 1965), and its subspecies phyllis has thewidest vegetation, 47fielddaysoversevenmonths (seedetails
geographic distribution (Brown 1979, Sheppard et al. onthese areas in Freitas 1993).
1985) and environmentaltolerances (Araujo 1980). In Mark-release-recapture studies were made in
the southern populations ofH. erato (30°S), periodic Morro do Voturua during 26 months, from 13 August
variation in several features ofwing colorpattern, es- 1994 to 30 September 1996, 1-4 times perweek, to-
peciallythe hindwing red-raylets, has also been noted taling 153 field days (about 4 hours/day). Butterflies
(Pansera&Araujo 1983, Oliveira &Araujo 1992). The were captured with an insect net, individually num-
present study describes features ofapopulation ofH. bered on the underside ofboth forewings (in the red
eratophyllis in asubtropicalrainforestin southeastern central macula) with a black permanent felt-tipped
Brazil(24°S),6° (660km) farthernorthofthesouthern pen, andreleased. Age (basedonwingwear), forewing
limits ofthe species distribution, and reports cyclical length (in mm), pointofcapture, sexandfoodsources
annualvariationintwowingcolor-pattern elements. were recorded (as in Freitas 1993, 1996). The age of
individual butterflies was estimated based on wing
*Towhomallcorrespondenceshouldbeaddressed. wear, initially using the six categories described by
12 Journalofthe Lepidopterists' Society
SAO PAULO
MORRO DO VOTURUA (MV)
Fig. 1. (A)studyarea(SaoVicenteRegion)inSoutheasternBrazil. Inregionalmap(B),VQ=ValedoRioQuilombo, MJ = MorrodoJapui,
MV= MorrodoVoturua(detailsinC).
Ehrlich and co-workers (Ehrlich & Davidson 1960, als present per day" (NIPD). To estimate the NIPD,
Brussard & Ehrlich 1970, Ehrlich & Gilbert 1973). recaptured individuals were considered to be present
Thesesixcategories (freshlyemerged, new, intermedi- in the population on allprevious days since the dayof
ate, old, veryold, tattered) were regrouped into three first capture (=marked animals at risk).
categories: new = freshly emerged and newindividu-
Results
als,intermediate = intermediateindividuals, andold =
old, very old, tattered (as in Freitas 1993, 1996). Age Population Dynamics
structurewascalculatedasthe dailyproportionofeach InMorrodoVoturua,theNICDformalesvariedfrom
category, and grouped into monthly means. Three oneto 13 (mean = 5.9, SD = 2.4; n = 153 days),with 19
wing-color-pattern elements were recorded (Fig. 3): dayswith n = 3 and 13 dayswith n = 10 males (Fig. 4).
the number ofredbasal spots, the number and shape The number ofnewly marked individuals captured in
of the red raylets on the ventral hindwing, and the eachmonth(=monthlyrecruitment)alsovaried,morein
color and size ofthe spot in the inner angle ofspace the first year than in the second (Fig. 5). In nearby
Cul-Cu2 ("cubital spot") on the dorsal forewing Morro do Japui, the male NICD ranged from 1 to 5
(nomenclature following Miller 1970). Data from Au- (mean=2.3, SD = 1.1,n=47days).Theareacoveredin
MV
gust and September 1994 at were regardedas the thesetwo sitesis similar (Fig. 1, see alsoFreitas 1993).
"Winter 1994" sample; each ofthe following seasons InVale do Rio Quilombo, the NICD for males was
represents observations for3 consecutive months. greaterthanin Morro doVoturua (mean = 14.5, SD =
The MRRdatafrom MVwereanalyzedbytheJolly- 4.1, n = 12 days) (Fig. 6A). The area covered in this
Sebermethod (Southwood 1971) forestimatingpopu- site was about six times greater than in the previous
lation parameters (software developed by R. B. two, corresponding closely to the NICD ratio for
Francini, UNISANTOS). In most cases, only males Morro do Japui (6.3) and suggesting an equivalent
were analyzed because ofthe lownumberoffemales. density, 40% ofthatin Morro doVoturua.
Dailyresultsweretabulatedas "numberofindividuals Jolly-Seber analysis for males in Morro do Voturua
capturedper day" (NICD), and "numberofindividu- (Fig. 7) gave estimatednumbers from 3 to 88 individ-
Volume 53, Number 1 13
new), even ifthis number was considered greater in
males (%2 = 7.6, df= 1, P < 0.005). Mostofthebutter-
flies captured on each day (afterthe first) were previ-
ouslymarkedindividuals (mean = 71.4%, SD = 23.4, n
= 152 days). The age structure during the 26 months
MV
in was not stable, with decreases in proportion of
"new" individuals in thewinter (Fig. 9).
In VQ, the proportion of recaptures/day was high
(Fig. 6C), andtheage structure quite stable (Fig. 6D),
but these results shouldbe taken with care due to the
shortperiodofstudy.
ResidenceTime
JAi Si OlNi Di Ji Fi Mi Ai MiJiJiAS ONDJi Fi MiAiMiJi Ji AiS In MV, males had alonger residence time (mean =
1994 1995 (MONTHS) 1996 37.6 days, SD = 25.8, n = 154) than females (mean =
Fig. 2. Climaticdiagramofthe SaoVicente Regionduringthe 22.6 days, SD = 23.0, n = 30) (t = 2.96, df= 182, p =
study period (format following Santos 1965 and Walter 1985). 0.004). Estimated residence time of males ("life ex-
Hatched=humidperiods,black=superhumidperiods. pectancy"ofCooketal. 1967)was28.1 days. The max-
imum individual permanence (survival) was at least
uals (mean = 13.5, SD = 9.9, n = 152), with most days 127days foramale and89days forafemale (Table 2).
(n = 90) between 10 and 30. The NIPD for males in In MJ, residence time was calculated only for males
the same site ranged from 1 to 20 (mean - 9.8, SD - (mean = 18 days, SD = 14.6, n = 22), since the recap-
3.8, n = 154). Themalepopulationwas stablethrough- turerateoffemaleswaslow(only3recaptures). InVQ
outtheyear,withsmallfluctuationsinnumberofindi- maximumpermanencecouldnotbecalculated(dueto
viduals apparently not related to season, but with the short period of study and low number of recap-
lower numbers ofindividuals observed at the end of tures), but the residence time for males (mean = 20.1
winterinthe threeyears (Figs. 4, 5). days, SD = 12.8, n = 32) and females (mean = 15.0
Theresults showthat NIPD isbetterthan NICD as days, SD = 10.6, n = 11) were not significantly differ-
anindexofpopulationsize. Inspecieswithstablepop- ent (t = 1.17, df= 41, p = 0.248).
ulations composed oflong-lived resident individuals,
efficient marking and recapture can give nearlyequal
Vagility
population size estimates by the Jolly-Seber method
andsimple NIPD. In MV, most adults showed home range behavior,
beingobservedin the same site forseveral days, often
Sex Ratio using the same flower resources. The total distance
The sex ratio of individuals captured and marked
was male biased in all sites (Table 1, Figs. 6B and 8).
MV
In males dominated in all but one month ofthe
study; all the captured individuals were male in four
months, including April ofboth years (Fig. 8). In all,
263 males and 74 femaleswere captured and marked.
Males were recaptured from one to 16 times and fe-
males from one to five times; 154 males and 30 fe-
males were recaptured at least once. The proportion
ofrecaptured males (58.5%) was statisticallythe same
as that offemales (40.1%) (%2 = 2.6, df= 1, P > 0.20).
VQ
In the proportions ofrecaptures ofmales (30.8%)
andfemales (30.5%) were equal.
Age Structure Fig.3. WingpatternvariationinHeliconiuscratophyllis(back-
groundblack, solid = red, dotted =yellow). Coloredspotinthein-
Comparingthenewindividualsvs. intermediate+old, ner angle ofspace Cul-Cu2 (cubital spot, CS) present (left, red
most ofthe first captures in MVwere of"new" indi- m(aRcBuSl)aaanldmossitxfruesdedrawyiltehtsth(isR)R,)orsahbaspeendtl(irikgehts)m.aFllivleibneassal(lerfet)d;sfpootusr
viduals from both males (71%), and females (54% basalredspotsandfourredravletsshapedlikedots(right).
r
14 Journalofthe Lepidopterists' Society
14-
12
10
e
6-
4-
2 —
i
A S N D, J
1994 1995 (MONTHS)
Fig.4. Dailycapturesformalesof//, eratophyllisin MorrodoVoturua,August 1994toSep-
tember1996.
flown by adults varied among individuals. In general, The number of red raylets in butterflies of both
two classes ofindividuals couldbe identified: those al- sexesshowedabimodaldistribution, rangingfrom one
ways observed in the same site (residents) and those to seven, with peaks at three and five (Table 4). The
seen in two or more separate points in the study area averagenumberincreasesinsummerandautumn and
(vagile). Most ofthe males (71%) and females (85%) decreases inwinterandspring (Table4).
recaptured were vagile, while among the residents, The shape ofthe red raylets varied from a rounded
some individuals appeared to have tightly restricted dot to a small red fine. Individuals with red raylets
home ranges. One male was captured 17 times (twice shaped like dots (65.8%) were more frequent than
onthe same day) atthe sameplace, andatleasthalfof thosewithfines (34.2%). Theproportions ofeachtype
the resident individuals were captured more than 4 alsovariedthroughouttheyear (Fig. 10A).
times. The distance traveledbyvagile individuals var- Thecubitalspotonthe dorsalforewingcouldbeab-
iedfrom 50to 660 m (mean = 270.6, SD = 122.9 m, n sent, present, or fused with the transverse subapical
= 97) formales, and80to 600 m (mean = 236.3, SD = redbar. Ifpresentandnotfused, thespotcouldbered
140.9, n = 23) forfemales (Table 3). Some individuals oryellow. This characterwas veryvariable, but there
showedgreat mobilityin asingle day. In onecase, one was aclearmajorityof"spottobepresentandyellow",
male captured at 0900 was observed 500 m away at with no regular pattern of variation throughout the
m
1000 h, returning at 1030 h to within 50 ofthe first year (Fig. 10B).
site ofcapture. Males were commonly seen nearnec- The forewing length ofmales ranged from 27 to 42
tarsourcesandfemaleswere mostlyobservednearthe mm, the average varying alongthe year, with greatest
larvalfoodplants. values beingobserved duringthe summer (Fig. 11).
The average forewing length of males (mean =
Wing Size andColorPatternin Morro doVoturua 36.25 mm, SD = 2.99, n = 250) was greaterthan that
Four or five redbasal spots are present on the ven- offemales (mean = 35.21 cm, SD = 2.96, n = 71; t =
tralhindwingofHeliconiuserato butterflies. This trait 2.58, df= 319,p = 0.01). Thisdifferencewasnotfound
wasverystable, asonlythreeoutof337individualsan- whenonlyrecapturedindividualswere considered if =
alyzed (males andfemales) showedthe fifthbasalspot 1.31, df= 181, p = 0.191). Whencomparingthe recap-
(Fig. 3). turedvs. notrecapturedwithinsexes, recapturedindi-
viduals were smaller. This difference was almost sig-
nificant for males (t = 1.933, df= 248, p = 0.054), but
notforfemales it = -0.172, df= 2.83, p = 0.864).
Natural HistoryofAdults
Adultsbegan activityaround0800hin summerand
0930 h in winter, but this could varywithin a season
according to the weather. On some cold days during
winter, activitybeganonlyafter 1030h.
1995 MONTHS Beak marks (from handling and release by, or es-
( )
capefrom, birdpredators)werefoundonthewings of
Fig. 5. Monthly recruitment of males ofH. erato phyllis in
MorrodoVoturua,August1994toSeptember1996,astotalnumber individuals from all populations, but a different pro-
ofindividualsmarkedforthefirsttimeineachmonth. portion of individuals with damaged wings was ob-
Volume 53, Number 1 15
Table 1. Sexratioofmarkedindividuals ofH. eratophyllis in Table2. PermanenceofmarkedH. eratophyllisinthe"Morro
thethreestudysites. doVoturua". Dayselapsedbetweenmarkingandlastrecapturerep-
resenttheminimumpermanence(MP)foreachindividual.
X'valu
MorrodoVoturua 263 74 3.5:1 105.9
MorrodoJapui 50 6 8.3:1 34.6 1-20 51 33.1 18 60.0 69
ValedoRioQuilombo 104 36 3:1 33 21-40 38 24.7 5 16.7 43
41-60 34 22.0 5 16.7 39
61-80 21 13.6 1 3.3 22
servedineachsite. Beakmarkswereseenin MVon 18 8>11-01000 91 06..60 1 30..30 101
males (6.8%; 14laterrecaptured) and 1 female (1.3%);
Total 154 100.0 30 100.0 184
in MJ, on 10males (20%) andnofemales;inVQ, on 18
males (17%) and8 females (22%).
Adultswerealmostalwaysobservedfeedingonflow- resultedindisappearanceofsomeindividualsfromthe
ers, but a few individuals were observed feeding on studyareaforuptotwoweeks.
damagedfruits ofRubus rosaefolius Smith (Rosaceae), FourPassifloraspecieswererecordedaslarvalfood-
and in one case a captured individual was observed plants in MV Passiflora alata Dryand, P. edulis Sims
bearing a seedofthisMplVant onits proboscis. The most (bothusedrarely), PjilekiWawra, andP. capsularis L.
visited flower in the site was Lantana camara L. Otherspecies maybeusedinthesame areaandinthe
(Verbenaceae) (more than 150 records). Twovarieties neighboringsites. Thebehaviors observedin oviposit-
ofL. camara occurinthestudyarea, onewithred-yel- ing females and larvae were the same as reported in
lowflowers andthe other (much more common) with theliterature (see Brown 1981).
white flowers. Some individuals were also observed
visiting flowers ofRubus rosaefolius (30 records), As- Discussion
clepias curassavica L. (Asclepiadaceae) (13 records),
Epidendrumfulgens Brongn. (Orchidaceae) (3records), Populational EcologyandAdult Biology
Gurania sp. (Cucurbitaceae) (3 records), andImpatiens The population parameters ofH. erato phyllis seen
wallerianaHook.f. (Balsaminaceae) (2records). Pollinia inthis studyagreewiththose reportedforothertropi-
ofEpidendrumfulgens andAsclepiascurassavicawere &
cal Heliconius (Turner 1971, Ehrlich Gilbert 1973,
observed attached to the proboscis ofsome captured &
Araujo 1980, Mallet Jackson 1980, Ehrlich 1984,
individuals. Anewmalewas onceobservedvisitingthe Mallet et al. 1987, Quintero 1988). The MV popula-
inflMoVrescence ofHeliconia sp. (Musaceae). Individuals tionwas reasonablystable in numbers during the two
in were rarely observed feeding on flowers of years ofstudy ifcomparedwith species ofother sub-
Asteraceae, although several species ofMikania, Eu- familiessuchas IthomiinaeandTroidini,bothshowing
patorium, and Vernonia are common there. The few pronounced fluctuations in numbers through time
Asteraceae that were observed being visited include (Vasconcellos-Neto 1980, Brown et al. 1981, Francini
Mikania lundiana D.C. (3 records), Trixis antimenor- 1989, Freitas 1993, 1996, Tyler et al. 1994, Pinto &
rhoea Mart. ex. Baker (3 records), Eupatorium laevi- Motta 1997). Other features observed forH. erato in
gatumD.C. (3records), andEmiliasonchifolia D.C. (2 the present study were the continual recruitment of
records). In the VQ site (where L. camara is not so newindividuals, lowdensityofadults, absence ofsud-
common), the Asteraceae Bidenspilosa L. and Titho- denincreases anddecreasesinnumbersofindividuals,
nia speciosa Hook ex Gris. were the mostvisited spe- and non-cyclical age structure. The population seems
cies with 90 and45 feeding records respectively. Also to be less affected by climate than those observed by
winasthicsosmimteoSntlaychuysteadrp(h1e5tarepcoorldysu)r.aS(oL.m)e(VaedurlbtesnaicneMaeV) JBaennesiorono(r19S7a8a)lffeolrdpaonpudlAatriaounjsoi(n19a8d1r)iefrorsiptoepiunlRatiioodnes
appeared to show "trap-line" (learned sequence) be-
havior for gatheringpollen and nectar (see Ehrlich &
Gilbert 1973 andEhrlich 1984). Table3. Maximumdistance(m)traveledbyadultsofH. erato
phyllisinMV.
Over a two-month period, one male was observed
visiting a dense Lantana patch at about 1000 h, and Residents Vagile
then in another patch 150 m away after 30-40 min- upto50 51-150 150-500 >500
utes. Several individuals showed similar patterns Males 39 17 77 3
throughout the two years ofstudy. Occasional distur- Females 4 8 13 2
Total 43 25 90 5
banceonforestedges (removalofsomeflowerpatches)
16 Journalofthe Lepidopterists' Society
Table4. Frequencydistributions,samplesize(N), mean(X),andstandarddeviation(SD)ofthenumberof"redraylets"formonthoffirst
captureinbutterfliesofbothsexescapturedinthe"MorrodoVoturua".
Numbeioftedraylets
Month/year 1 2 3 4 5 6 7 N X»SD
Aug/1994 2 5 2 8 2 1 20 3.3± 1.4
Sep/1994 3 6 4 2 15 3.3± 1.0
Winter/1994 2 8 8 12 4 1 35 3.3 ± 1.2
Oct/1994 4 1 3 8 3.9± 1.0
Nov/1994 2 6 6 4 3 21 4.0 ± 1.2
Dec/1994 2 7 2 5 1 17 3.7± 1.2
Spring/1994 4 17 9 12 4 46 3.9± 1.2
Jan/1995 4 8 5 10 6 33 4.2 ± 1.3
Feb/1995 1 1 2 5 9 5.0 ± 1.6
Mar/1995 1 3 4 8 3 1 20 4.5 ± 1.4
Summer/1995 2 4 11 10 20 14 1 62 4.4± 1.4
Apr/1995 1 5 3 1 9 5,3 ±1.1
May/1995 1 2 3 3 8 17 3.9 ± 1.3
Jun/1995 2 3 4 5 2 16 4.1 ± 1.3
Autumn/1995 1 4 7 7 18 5 1 43 4.3 ± 1.3
Jul/1995 2 4 9 4 5 1 25 3.4± 1.3
Aug/1995 2 1 1 1 1 6 3.3 ±2.1
Sep/1995 2 2 5 1 1 11 2.7± 1.2
Winter/1995 6 6 15 6 7 2 42 3.2 ± 1.4
Oct/1995 2 3 1 2 8 3.6± 1.6
Nov/1995 1 1 3 6 1 1 13 3.6± 1.6
Dec/1995 2 1 1 1 5 3.6± 1.8
Spring/1995 1 5 7 7 2 4 26 3.6 ± 1.4
Jan/1996 2 1 1 5 2 11 4.4± 1.4
Feb/1996 1 4 5 10 5.3 ± 1.0
Mar/1996 1 4 5 5 15 4.6± 1.4
Summer/1996 3 6 1 14 12 36 4.7± 1.3
Apr/1996 2 3 4 6 1 16 5.1 ± 1.2
May/1996 1 1 1 1 4 4.0± 1.2
Jun/1996 1 2 1 2 1 7 4.0± 1.4
Autumn/1996 2 5 4 7 8 1 27 4.6± 1.4
Jul/1996 1 3 2 1 2 9 3.0 ± 1.5
Aug/1996 1 2 3 4.7±2.3
Sep/1996 2 3 3 8 2.2 ±0.8
Winter/1996 3 7 5 1 2 2 20 2.9± 1.5
Total 15 43 81 57 86 52 3 337 3.9± 1.4
much farther south. In the latter populations, three in the same ornearbysites (Table 5; see alsoYoung &
distinctphaseswere reported: aperiodofrapidgrowth Moffett 1979, Vasconcellos-Neto 1980, Brown et al.
in early spring, aperiod ofmaximum densityin sum- 1981, Freitas 1993, 1996, Tyler et al. 1994, Pinto &
merthrough autumn, and an abrupt decline inwinter Motta 1997). Thehighvalues forresidence arenotre-
(Saalfeld &Araujo 1981, Romanowskyet al. 1985). In lated to geographic location ofthe populations, since
the periods ofmaximum density, the number ofindi- they are higherin Heliconius erato than in butterflies
viduals captured in these populations could be very ofothergroupsinthe sameandnearbysites (Table5).
high (morethan 150individuals reportedinone roost- InpartthehighvaluesofresidencereportedinHe-
ingsite) inrelationtothemaximumvaluesobtainedin liconius could be related to the movements and dis-
the present study (from 20to 30individuals) (Saalfeld persaloftheadults. SomeHeliconius andEueidesbut-
&Araujo 1981, Romanowskyet al. 1985). terflies are known to have restricted home ranges
The average time of residence and the life expec- throughout their lives, and even ifadults can fly dis-
tancyreportedinthepresent studyare bothhigh, like tances up to 5 km within the forest (Brown 1981,
those ofotherHeliconius (Turner 1971, Benson 1972, Pansera&Araujo 1983, Romanowskyetal. 1985, Mal-
Ehrlich & Gilbert 1973, Araujo 1980, Quintero 1988), let et al. 1987), small areas ofopen field could act as
including the populations from southern Brazil (Ro- efficient barriers to dispersal (Romanowsky et al.
manowsky et al. 1985). The values are higher than 1985). This pattern contrasts with that of "nomadic"
those ofmany Ithomiinae, Pieridae, and Papilionidae butterfliessuchas IthomiinaeandTroidini (seediscus-
Volume 53, Number 1 17
C. Dias & K. S. Brown, pers. comm.), several studies
showed low rates of recapture and low permanence
20- valuesforspeciesinthissubfamily,probablyasaresult
ofthe high dispersal ofthe individuals (Gilbert 1993).
10- But even ifmovement explains in part the residence
values of adults, almost certainly the high actual
longevityofthe adults is the main factoraffectingthis
O parameterinHeliconius.
I<- In the present study, sexratiowas male biased; sev-
a.
x 50%- eralprevious fieldstudies reportedmalebiasedsexra-
LU
if) tios even when the sex ratio in the laboratorywas 1:1
& &
(Ehrlich Gilbert 1973, Mallet Jackson 1980,
Ehrlich 1984, Ehrlich et al. 1984, Freitas 1993). This
differenceincapturebetweensexesmaybeduetodif-
50%- ferential behavior ofthe adults, with males generally
fliyinginthesameplacesasthecollectors, andfemales
moredispersedthroughoutthehabitatlookingforhost
plants (Freitas 1996).
Contrasting with these features, which agree with
those foundin other species ofHeliconius, the use of
50%
tou food resources by H. erato phyllis in the Sao Vicente
< region is unlike that ofmanyothertropical species of
thesamegenus. Thebutterfliesseemtopreferfeeding
alongforestedgesoncommonsecond-growthflowers,
even though a species ofRubiaceae and Gurania sp.
RioFiQgu.il6o.mboPo,puMlaartcihontaolJpuanream1e9t9e5r.sA,ofdHai.lyercaatpotuprhiejslloifsminalVeasle(tdoop (Cucurbitaceae) flowers are common in clearings in
line)andfemales; B,sexratioaspercentofmalesineachday'scap- the forests, where theyare heavilyused byHeliconius
tures;C,percentofcapturesofmales; D, agestructureofmalesas numata rohigus Weymer. It is interesting that sym-
inFig.4.
patric H. sara apseudes (Hiibner) and H. ethilla nar-
sion in Mallet et al. 1987) and probably other more caea (Godart) haveflowerpreferences similartothose
primitive genera of Heliconiini like Dryas, Agraulis of//, eratophyllis (pers. obs.),perhaps relatedtotheir
andDione (K. Brown, pers. comm.). For example, al- specialization on small pollen grains (Boggs et al.
thoughsome Ithomiinae arereportedtoliveas longas 1981) andtheuseoflarvalhostplantspeciestypicalof
10 months (e.g., Mcclungiasalonina (Hewitson), R. S. successionalhabitats (Gilbert 1991).
Fig.7. Estimatedpopulationsize(Jolly-Seber)forH.eratophyllis(males)inMorrodoVoturua,August1994toSeptember1996(middle).
Themaximumnumberofindividualsisgivenastheestimateplustheerror(topline),andtheminimumnumberisgivenastheNIPD(bottom
line),assumingthatthepopulationcouldnotbelowerthanthisnumber(seealsoFreitas 1993, 1996).
—
r
18 Journalofthe Lepidopterists' Society
a 50%-
50%-
———— — —— ——— ————————
-r-i i i i Fi—iMAr—MiJi iJ—iASi OiNr—i i Fi Mi Ai MiJi Ji AS
A S N D J J ———————————————
1994 (995 (MONTHS) 1996 A~\SOi Ni Di Ji Fi Mi Ai Mi Ji Ji AI SI Oi NiDiJiFiMiAiMiJiJiArS~
Fig. 8. SexratioforH. eratophyllis markedin MorrodoVotu-
rua, August 1994 to September 1996 (based on monthly recruit-
ment),aspercentofmalesineachmonth'scaptures.
WingColorPattern 50%-
Although the numberofredbasal spots is relatively
constantinthispopulationofH. eratophyllis, otherel-
ements of color pattern are not: the number of"red
raylets" varies seasonally, and the shape of the red
1995 MONTHS
raylets andthe cubitalspotonthe dorsalforewingvary ( )
duringtheyearwithnoclearpattern. The numberand Fig. 10. Monthly variation of two wing pattern elements in
shape ofthebasal spots ontheventralhindwingis one adultsofbothsexesofH. eratophyllisin MorrodoVoturua,August
1994toSeptember1996.A,ratiobetweenredrayletsofbothshapes
ofthe distinctivecharacters usedinHeliconius classifi- aspercentageofindividualswith redrayletsshapedlikedots. B, ra-
cation. This characterwas discussedby Emsley (1965) tioamongthe fourdifferentphenotypes ofthe coloredspotin the
as probably important in courtship as a recognition innerangleofspaceCul-Cu2:white=presentyellow,black=pres-
ent red, hatched = absent, dotted = fused with the subapical red
mechanism. Ifthis is true, the constancy ofthis char- macula.
acteris easilyexplained.
The patternobservedin the shape ofthe redraylets
cold winter, and the variation in the number of red
and in the cubital spot is not easyto explain, but ran-
dom variation is the most probable hypothesis. There rayletscouldbeexplainedthroughthe resultingbottle-
arenoknownpressuresactingonthesetwocharacters. necks (A. M. Araujo, pers. comm.). In the SaoVicente
Oliveira and Araujo (1992) proposed that the mean region (24°S), low temperature does not kill adults in
number of red raylets is related to temperature: the winter, and genes are not eliminated in this way. Al-
though color pattern frequencies are not stable, the
hot months shouldhave ahighproportion ofindividu-
alswith numerous redraylets. The genetic determina- population itself shows a relative stability in number
throughouttheyear,indicatingthatthisvariationoccurs
tion of this character was defined by Pansera and
Araujo (1983), but numericpolyphenism could not be without reduction in population size, suggesting that
the variation could be due to simple numerical poly-
discarded. Polyphenism occurs in several species in
seasonal environments (Brakefield & Larsen 1984, phenism.
Braby 1994, Windig et al. 1994). In southern Brazil
(30°S), adult populations greatly decline during the
iiiT
h
in
Q Id
SON MAMJJASON J F M A M J J ASONDJFMAMJ JASONDJFMAMJ JAS
1994 1995 (MONTHS) 1996 1994 1995 1996
Fig.9. AgestructureformaleH. eratophyllisin MorrodoVo- Fig. 11. Mean forewinglength ofmales ofH. eratophyllis in
turua, August 1994 to September 1996 (black = fresh individuals, Morro do Voturua, August 1994 to September 1996 (based on
hatched=intermediate,white =wornindividualsas%ofeachday's monthlyrecruitment). Histogrambars = monthlymeans,lineexten-
captures). sions =standarddeviation.
Volume 53, Number 1 19
Table5. Residencevalues(indays)forsomebutterflyspeciesinSoutheasternBrazil. Meanmr= meanminimumresidence,mo= maximum
foroneindividual.
—«
meanmr
iteRegion
Nymphalidae
Heliconiuseratophyllis 'MorrodoVoturua 37.6 ±25.8 127 22.6±23.0 89
Heliconiuseratophyllis 'MorrodoJapui 18.04± 14.6 47 22
Melinaealudovicaparaiya Reakirt2 14.1 ± 14.1 55
MelinaeaethraGodart2 13.7± 12.9 47
Placidulaeuryanassa2-3 8.4±8.3 45 7.2±5.7 23
Mechanitislysimnialysimnia Fabricius2 15.5± 15.4 67 18.3± 16.2 72
Dircennaderoceltina Burmeister2 9.4±6.9 23 10.9 ± 10.5 49
HeterosaisedessaHewitson2 12.9± 15.1 65 13.5± 10.3 39
Hypothyrisninoniadaeta (Boisduval)4 16.3± 11.1 40 14.2 ± 16.2 52
Pseudoscadaerruca(Hewitson)4 9.9±9.6 25 33
Ithomiadryrno Hiibner4 7.0± 7.3 17 12,3 ± 12.1 26
Actinotepelleneapellenea Hiibner5 3.2—± 2.6 12 6
ActinotebryllaOberthiir5 16 16
Papilionidae
Paridesanchisesnephalion Godart6 14.1 ± 8.2 30 9.0±3.6 12
Otherregions
Nymphalidae
Heliconiuseratophyllis' 30.7±29.0 112
AeriaolenaolenaWeymer8 8.7±5.9 24 11.7±2.5 14
Actinotezikani D'Almeida9 3.9—± 1.3 7
Pierellalamia Sulzer10 60
Pieridae
Euremaelathea (Cramer)11dryseason 10.9 ±9.1 52 9.7±9.8 54
Euremaelathea11wetseason 8.8 ±5.6 28 6.9 ±4.2 21
Superscriptnumbers:1,thisstudy;2,Freitas1996;3,Freitas1993;4,Freitas,unpubl.datafromMorrodoJapui;5,Francini1989;6,Freitas,unpubl.datafrom
MorrodoVoturua;7, Francini, unpubl. datafrom Lavras, MG, Brazil; 8,VaniniandFreitas, unpubl. datafromCampinas, SP;9, Francini, unpubl. datafrom
Paranapiacaba,SP; 10,Freitas,unpubl.datafromSeteBarras,SP;11Vanini,BonatoandFreitas,unpubl.resultsfromCampinas,SP.
EcologicalPlasticityinHeliconius? andvegetation. Colonies ofthis subspecies arepresent
in virtuallyanykind ofvegetation in southeastern and
Severalfeatures ofH. eratovaryindifferentparts of southern Brazil, includingprimary and secondary for-
theneotropics. Althoughinthe SaoVicenteregionand est edges, urban parks, plantations ofPinus and Euca-
Trinidad the populations are stable, maintaining con- lyptus, riparianforests, savannas andsandysoilforests,
stant low numbers throughout time (Turner 1971, R. and in tropical, subtropical, and temperate environ-
B. Francini, pers. obs.; this study), in southern Brazil ments (Araujo 1980andpers. obs. oftheauthors). This
theyshowstrongfluctuations overtheyear, with some is probably related to the ability ofthis species (espe-
extinctionincolderyears (Saalfeld&Araujo 1981, Ro- ciallythe subspeciesphyllis) touseagreatrangeoflar-
manowskyet al. 1985). Heliconius erato is reported as valandadultresources, andchangebehaviorandpref-
feedingonmorethan37species and6genera/subgen- erences according to the environment (=ecological
eraofPassifloraceae (Benson et al. 1976, Brown 1981, plasticity).
Spencer 1988), even using Passiflora edulis and P. Such statements could apply not only to H. erato,
alata, host species normally rejected in most popula- but also to H. sara andH. ethilla populations in south-
tions, on forestedges andurban areas (this study A. V. eastern and southern Brazil, bothwith similargeneral-
L. Freitas, pers. obs., andunpublished results by L. S. ist habits, and could help to explain patterns ofdistri-
Otero and K. S. Brown Jr.). The differences in use of bution ofthe species ofthis genus. In contrast to the
flower resources among areas in this study shows that three species cited above, four other Heliconius of
H. erato phyllis rapidly responds to variations in re- southeastern Brazil have much more restricted habits.
source availabilityindifferent sites andseasons. Heliconius melpomene nanna Stichel, H. numata robi-
&
Heliconius erato, especially the subspecies phyllis, gus andH. nattered Felder Felder, are restrictedto
seems to be able to persist in many lands ofclimate limited sectors of forest habitats, the last one with a
20 Journalofthe Lepidopterists' Society
very small range, a poor competitorwith H. sara and Braby, M. F. 1994. Phenotypicvariation in adultMijcalesis Hiib-
other lowland species (Brown 1972, W. W. Benson, ner(Lepidoptera: Nymphalidae:Satyrinae)fromtheAustralian
wet-drytropics. Aust. Entomol.Soc.33:327-336.
pers. comm.); H. besckei Menetries (a co-mimic ofH. Brakefield,P. M.&J.T. B. Larsen. 1984. Theevolutionarysignif-
erato) is typical of Atlantic forest mountains, in pri- icanceofdryandwetseasonformsinsometropicalbutterflies.
mary and secondaryforests at medium high altitudes, Biol.J. Linn. Soc.22:1-12.
descending to sea level only in winter (Brown & BrowNny,mpKh.alSi.daJer.).1P9a7r2t.IITIh.eEcHoelloigcyonainadnsbioofloBgryazoiflHe(lLiecpoindioupstenraat:-
Mielke 1972 andunpublished results bythe authors). tereri,akeyprimitivespeciesnearextinction,andcommentson
Even though environmental tolerance mayhelp ex- theevolutionarydevelopmentofHeliconiusandEueides. Zoo-
logica57:41-69.
plain distributions ofthe species ofHeliconius, other . 1979. Ecologia Geografica e Evolucao nas Florestas
factors need to be investigated. Benson (1978) argues Neotropicais. SaoPaulo, Universidade Estadual deCampinas,
thatthe availabilityofnewhostplant shoots inthe dry Campinas,SPxxxi +265pp., 120pp.app.
1981. ThebiologyofHeliconiusandrelatedgenera.Ann.
winter results in high competition and is responsible Rev. Entomol.26:427-456.
for the absence ofH. melpomene in a seasonal site in Brown,K.S.Jr. &W.W. Benson. 1974. Adaptivepolymorphism
South Brazil. In the SaoVicente region, however, this associatedwith multiple Miillerian mimicryin Heliconius nu-
mata (Lepid. Nymph.). Biotropica6:205-228.
seasonality is much less pronounced, and periods of Brown, K. S. Jr., A. J. Damman & P. P. Feeny. 1981. Troidine
severedroughtarevirtuallyabsent (Fig. 2); newshoots swallowtails(Lepidoptera:Papilionidae)insoutheasternBrazil:
ofPassiflora species seem to be abundant throughout natural history and foodplant relationships. J. Res. Lepid.
19:199-226.
theyear,andthoughadetailedstudyneedstobecarried Brown, K. S. Jr. & O. H. H. Mielke. 1972. The Heliconians of
out,thiscouldnotexplaintheabsenceofH. melpomene Brazil (Lepidoptera: Nymphalidae). Part II. Introduction and
southofRiodeJaneiro State. Althoughtheseideasand generalcomments,withasupplementaryrevisionoftheTribe.
Zoologica57:1-40.
many others were discussed by Gilbert (1984), there Brussard,P.F.&P.R.Ehrlich. 1970. Thepopulationstructureof
still seem tobe more questions than answers. Erebiaepipsodea(Lepidoptera:Satyrinae).Ecology51:119-129.
Populationstudiesneedtobeundertakenwithother Cook, L. M., L. P. Brovver&H.J. Croze. 1967. Theaccuracyof
apopulationestimationfrom multiplerecapturedata. Anim.
species ofHeliconiini, especially the more widspread Ecol.36:57-60. J.
species ofHeliconius, such as H. ethilla and H. sara. Drummond,B.A.III. 1976. Comparativeecologyandmimeticre-
The presence of seasonal polyphenism may indicate lationships ofIthomiinebutterflies in eastern Ecuador. Ph.D.
Thesis, UniversityofFlorida,xvi + 361pp.
species with broad tolerances to different environ- Ehrlich,P. R. 1984. Thestructureanddynamicsofbutterflypop-
ments, ashasbeensuggestedforotherpolyphenicbut- ulations,pp.25-40.In:R.I.Vane-Wright&P.R.Ackery(eds.),
terflies (Shapiro 1976, Kingsolver & Wiernasz 1991, EhrlTihceh,biP.olRo.gy&ofSb.utEt.erDfaliveisd.sAocna.de1m9i6c0.PrTeessc,hnLioqnudeosn.forcapture-
VanDycketal. 1997),althoughinHeliconiusthiscould recapture studies of Lepidoptera populations. Lepid. Soc.
J.
be apparentlywithout adaptive consequences. These 14:227-229.
studies couldguidefuture researchinpopulationbiol- Ehrlich, P. R. & L. E. Gilbert. 1973. Population structureand
dynamicsofthetropicalbutterflyHeliconiusethilla. Biotropica
ogyofbutterflies, andhelpinthe understandingofthe 5:69-82.
ecologyoftropicalinsects. Ehrlich,P.R., A.E.Launer&D. D. Murphy. 1984. Cansexra-
tio be defined or determined? The case of a population of
Acknowledgments checkerspotbutterflies.Am. Nat. 124:527-539.
Emsley, M. G. 1964. The geographical distribution ofthe color-
We thank R. B. Francini, K. S. BrownJr., W. W. Benson, B. A. pattern components of Heliconius erato and Heliconius
Drummond, D. Bowers, A. M.Araujo,A. Caldas,J. R.Trigo, L. E. melpomenewithgeneticalevidenceforthesystematicrelation-
Gilbert,J. Mallet,A. Gomes-Filho, G. Machado, M. R. Pie, G.W. shipbetweenthetwospecies.Zoologica49:245-286.
Chaves and C. E. G. Patto for discussions and comments on the 1965. Speciation in Heliconius (Lep., Nymphalidae):
.
manuscript. R. B. Francini helpedinfieldworkand inpopulation Morphologyandgeographicdistribution.Zoologica50:191-254.
analysis. Climatologicaldatawerekindlyprovidedbythe Brazilian Francini, R. B. 1989. Biologiae Ecologiade BorboletasActinote
AirForce.A.V. L. FreitasthankstheCNPqforafellowship. (Lepidoptera: Nymphalidae)naTransicaoSubtropicaldoSud-
estedoBrasil. Ms.Thesis,UniversidadeEstadualdeCampinas,
Literature cited Campinas,SP.236pp.
Freitas, A. V. L. 1993. Biology and population dynamics of
Araujo,A. M. 1980. EstudosgeneticoseecologicosemHeliconius Placidula euryanassa, a relict ithomiine butterfly (Nymphali-
erato (Lepidoptera, Nymphalidae). Actas IVCongr. Latinoam. dae: Ithomiinae). Lepid. Soc.47:87-105.
J.
Genetica2:199-206. . 1996. Populationbiology-ofHeterosaisedessa (Nymphali-
Benson, W. W. 1972. Natural selection for Miillerian mimicryin dae)anditsassociatedAtlanticForestIthomiinaecommunity.J.
HeliconiuseratoinCostaRica. Science 176:936-939. Lepid. Soc.50:273-289.
. 1978. Resource partitioning in passion vine butterflies. Gilbert, L. E. 1975. Ecologicalconsequencesofacoevolvedmu-
Evolution32:493-518. tualism betweenbutterflies andplants, pp. 210-240. In L. E.
Benson,W.W., K.S.BrownJr.&L.E.Gilbert. 1976. devolu- Gilbert&P.R.Raven(eds.),Coevolutionofanimalsandplants.
tionofplantsandherbivores:passionflowerbutterflies. Evolu- UniversityofTexasPress,Austin,Texas.
Boggtsi,onC29L:.,65J.9T-.68S0m.iley&L.E.Gilbert. 1981. Patternsofpollen I.V.an1e9-84W.rigBhiotl&ogyP.oRf.bAucttkeerrfyly(ecdosm.)m,unTihteiebsi,olpopg.y4of1-b6ut1t.erIfnlieRs..
exploitationbyHeliconiusbutterflies.Oecologia48:284-289. AcademicPress,London.