BULLETIN OF THE ALLYN MUSEUM
THE ALLYN MUSEUM OF ENTOMOLOGY
Number 13 26 July 1973
THE HELICONIANS OF BRAZIL
PART V. THREE NEW SUBSPECIES FROM
MATO GROSSO AND RONDONIA
Keith S. Brown, Jr.
(entro de Pesquisas de Produtos Naturnis
Centro de ('Cincias Midicas, U.F.R.J.
Ilha do Fundao. Rio de Janeiro Z'-:12. HBrasil
In the course of an intensive systematic, biological, and chemical investiga-
tion of aposematic butterflies of the nymphaline tribe Heliconiini, the author has
had occasion to examine a large number of museum collections, in Brazil and in
other American countries. In the Museu Nacional in Rio de Janeiro, there was dis-
covered an unrecognized form of Heliconius astraea Staudinger 1896, represented
by a small number of specimens from southern Amazonas and Rondonia, mixed
with the known Heliconius egeria hyas Weymer 1884 (for the reasons for separa-
tion of H. astraea from H. egeria (Cramer), see part II of this series, Brown and
Mielke, 1972). A further specimen of this new form was examined in the Allyn
Museum of Entomology in Sarasota, Florida, originally in the W. J. Kaye collection,
now part of that Museum.
In this last collection, as well as in the U. S. National Museum in Washington,
and illustrated in the literature (Collenette and Talbot, 1928), there was found a
new subspecies of Heliconius ethilla Godart 1819, quite similar to, but distantly
separated geographically from, the Colombian H. e. metalilis Butler 1873.
During excursions to central and central-western Brazil to obtain sufficient
material of these two heliconians to permit their formal description, the author
discovered two additional new forms of Heliconiini. One of these, confined to
central-western Mato Grosso, proved to be an unknown subspecies of Heliconius
aoede (HUbner 1809-13). A number of specimens transitional between this new form
and the subspecies to the northwest, H. aoede faleria Fruhstorfer 1910, were later
seen in the Allyn Museum and the British Museum (Natural History). The latter
collection also contained representatives of the new subspecies of astraea and
ethilla, mentioned above.
The second new form discovered was a black-suffused extreme phenotype
of the common and widespread Dionejuno (Cramer 1779-80), present in pure popu-
lations in a variety of areas in the Distrito Federal. This was originally regarded
as a new subspecies to be here described together with the three new Heliconius
forms. It was later decided, however, to validate the name suffumata Hayward
1931 and apply it to these uniform juno populations in Brasilia. The name was
originally published to denote an aberrational specimen from Paraguay; the
original description and figure of this specimen, however, fit the Brasilia form so
closely that it does not seem justified to introduce a new name. This form is there-
fore redescribed and illustrated here under the validated name Dionejuno suffumata,
which receives a new status as a recognizable and isolated subspecies.
The existence of these four forms was mentioned, and the astraea was
illustrated, in Part II of this series (Brown and Mielke, 1972). No names were applied
to the three new Heliconius subspecies, which are to be considered as described in
the present paper. The name suffumata was used in a subspecific sense in the first
Appendix of Part II and may be considered as validated as of that publication,
though its redescription and official new status are included here.
1. Heliconius. astraea rondonia, K. Brown, new subspecies
(Plate I, Figs. 1, 4; Plate IV, top; Text-Figures 1-2)
Adult 6 morphologically identical (Text-Figure 1) to Heliconius astraea
astraea, differing from the nominate subspecies (Plate I, Figure 2) only in the
presence of large yellow spots in the outer third of the forewing cell and the inner
half of forewing space Cu1-Cu2, causing the yellow forewing median band to be
square and centered over the end of the cell (Plate I, Figure 1), rather than rec-
tangular and distal to the cell. Differs from the superficially similar H. egeria hyas
(Plate I, Figure 3) morphologically (see Emsley, 1965 and Brown and Mielke, 1972)
and in the less elongated forewing, and clearer and usually longer rays on the
hindwing, with the hindwing cubital vein always black. As in other astraea and
egeria forms, the dennis and rays are very bright red in fresh specimens, fading
rapidly to orange-red in older specimens or after death, and the underside (Plate I,
Figure 1) is extensively washed over with silvery-gray, fading to gray-brown after
death, and bears a long yellow costal streak on the hindwing and a shorter yellow
costal spot on the forewing. 9 wing-pattern similar to that of a; morphology (Text-
Figure 2) as reported for egeria 9 (Emsley, 1965). Forewing 42-48 mm (base to apex);
head markings all white; thorax markings yellow dorsally, mostly white ventrally;
abdomen markings and ventral stripe yellow.
HOLOTYPE 6 and ALLOTYPE 9 donated by the author to the Museu
Nacional, Rio de Janeiro; Km. 70 of the Vilhena-Pimenta Bueno highway, Rond6nia
(woods and creek flowing down to northeast of road), elevation 400 m., October 19,
1971, K. Brown leg.
PARATYPES: two 6 and one 9, same locality, 19-X-71, collected and retained
by K. Brown; one pair, creek flowing down to north of same highway at Km. 81,
19-X-71, and one S, 22-VI-71, collected and retained by K. Brown; three 5, ManicorB,
Text-Figure 1: Heliconius astraea rondonia, Paratype S from Manicor6,
Amazonas, in the Museu Nacional, internal aspect of left genital valve.
Rio Madeira (generalized locality?), in the Museu Nacional, Rio; one 6, collected
by the Commissao Rondon in southeastern Rond6nia, in the Museu Nacional, Rio,
and one 6, same data, ceded by the Museu Nacional to K. Brown; one 3, probably
captured near Vilhena by B. Steingruber, in the collection of K. Brown; three 8,
11 Km.west of Vilhena on the road to Pimenta Bueno, Rond6nia, elevation 600m.,
11-VII-72, in the collection of W. W. Benson, Rio de Janeiro; one pair, same date and
locality, donated by W. W. Benson to the Museu de Zoologia da Universidade de
Sdo Paulo; one &, same locality, 6-VII-72, and one 6, Km. 57 of the Vilhena-Pimenta
Bueno highway (elevation 450 m.), on sand where clothes being washed, 6-VII-72,
in the collection of the Departamento de Zoologia, Universidade Federal do Para-
na, O. Mielke leg.; one a from the "Cuyabd-CorumbA River System" (for discussion
of this locality, probably in northern Mato Grosso or eastern Rondonia in the
Amazon drainage, see Part II) in the Allyn Museum of Entomology, Sarasota,
Florida (ex W. J. Kaye collection); and five 6, same data, in the British Museum
Additional specimens of this subspecies have been seen, sufficiently closely
to permit positive identification, flying over the Vilhena-Pimenta Bueno highway
13 and 30 Km. northwestt of Velhena, in heavy woods and open cerradao, respective-
ly. The known collection points and probable range of the subspecies are shown in
the Map; it is to be expected in hilly forested areas over much of the "dry island"
of the southeastern Rio Madeira region (southern half of the middle Amazon Basin),
where it probably evolved during a recent dry climatic cycle-and remains isolated
Sympatric Heliconius species at the type-locality (Plate IV) include H. b.
burneyi (Hiibner), wallacei flavescens Weymer, xanthocles paraplesius Bates
x melete Felder, doris doris (Linn6) and form "delila" (Htibner), numata (Cramer)
and varieties including "silvana" Cramer and "mirus" Weymer, ethilla eucoma
(Hiibner) (Plate II, Figure 14), erato venustus Salvin and transitions to e. amazona
Staudinger, and sara thamar (Hiibner).
Heliconius astraea rondonia is one of the largest and highest-flying of the
heliconians; it has been observed in normal social chasing over the tops of 30-
Text-Figure 2: H. a. rondonia, Allotype 9 from Km. 70, Vilhena-
Pimenta Bueno highway, Rond6nia, bursa copulatrix, abdominal process, and
meter trees, and almost never descends below five meters from the ground in
ordinary flight. For puposes of collection, it may be easily attracted down to the
ground by a large bright red or orange cloth, as can other members of the wallacei-
group of the genus Heliconius (the author is grateful to Dr. W. W. Benson for informa-
tion on this method, long used by collectors in French Guyana for H. egeria,
which permitted the collection of the type-series of H. a. rondonia at Km. 70). In
nature, it flies from 7:30 AM to 4:00 PM, visiting flowers (especially high vines
of Gurania (Curcubitaceae) and Passiflora glandulosa) frequently during all this
period, but delaying on them most in mid-morning. The flight is very fast, powerful,
and fluttery, interspersed with rapid turns and spectacular dives; the descent to
the red cloth is made by a vertical drop on half-closed wings or a direct spiralling
dive, followed by brief hovering near the cloth and a rapid climb to exit. The cloth
distracts both males and females from all their normal activities, including
chasing, flower visiting, and pre-oviposition behavior. Both sexes (especially males)
are also known to come down to wet sand during the warmer hours of the day.
The subspecies probably prefers restricted high humid wooded patches near
medium-sized streams within relatively dry cerrado areas, and has been seen flying
over open cerrado and cut-over wooded areas where flowers are abundant. It
may occur south over much of northwestern Mato Grosso into the upper
Paraguay River system, and north along the Rio Madeira to the Manicore area,
but its center of abundance is apparently on the northwestern escarpment of the
Serra dos Parecis in extreme southeastern Rond6nia.
Although a female was observed in pre-oviposition fluttering before a
growing shoot of Passiflora (Distephana) coccinea at Km. 70, she did not lay eggs,
and a caterpillar obtained from an egg expressed from this female (Plate I,
Figure 4) did not grow well on P. coccinea, though some feeding activity was
evidenced. The foodplant in Vilhena is almost surely P. (D.) glandulosa, with which
the species is closely associated there. Possibly, like some populations of the closely
related H. burneyi, the caterpillar of astraea locally uses primitive passion vines in
the subgenus Astrophea (W. W. Benson, personal communication). The lowest-flying
member of the group, H. wallacei Reakirt, is always closely associated with members
of P. (Distephana) glandulosaa, coccinea, quardriglandulosa, and vitifolia),
while H. burneyi is usually found associated with these and has been discovered and
bred through on P. coccinea.
The subspecies name is considered a substantive in the feminine gender, in
apposition with the genus and species names, and is chosen to indicate the principal;
area where the butterfly is found, the Brazilian territory of Rondonia. This name in
turns honors the Brazilian hero Merchal CAndido Mariano da Silva Rondon, world-
famous for his humanitarian work in exploring the southwestern Amazon Basin
and making lasting peace with the Indian nations of the area, without damage to
their culture, dignity, or traditional lands.
2. Heliconius aoede eurycleia K. Brown, new subspecies
(Plate I, Figures 5, 8, 10; Text-Figures 3-4)
Adult a morphologically identical (Text-Figure 3) to H. aoede lucretius
Weymer 1890, differing from this subspecies principally in the distal truncating of
the forewing yellow median band, the reduction of the dennis-marking on the fore-
wing to three basal lines of variable length and width, and the deeper red color
of these lines and of the hindwing rays (Plate I, Figure 5). The color-pattern
closely resembles those of the sympatric species H. melpomene penelope Staudinger
1894 (Plate I, Figure 9) and H. erato venustus Salvin 1871 (Plate IV), which also
bear truncated forewing bands and reduced and reddened dennis markings. For
comparison, H. aoede lucretius Weymer, 1890 (Plate I, Figure 6), with a full fore-
wing band and orange dennis and rays, flies together with the very similar H.
melpomene vicina MWnetriBs 1857 and H. erato reductimacula Bryk 1953 and lati-
vitta Butler 1873 in the northwestern and western Amazon. Just northwest of
eurycleia populations, aoede may be found as the form faleria, with a full orange
dennis and a less constricted yellow forewing band, flying with the very similar
H. melpomene madeira Riley, 1919 and H. erato form "constricta" Joicey & Kaye,
1917 (transition amazona/venustus; see Plate IV). The hindwing rays in eurycleia
are moderately well expressed on both wing surfaces (Plate I, Figures 5, 8).
The 9 is similar to the 3, but has a heavier abdomen, more rounded forewings, and
no wide anterior androconial patch on the hindwings; the morphology is as reported
previously for the species (Text-Figure 4; Emsley, 1965). Forewing length 29-33 mm
(males), 31-35 mm (females); head markings all white; forewing ventral costal
spot red; hindwing ventral costal streak yellow; hindwing paired submarginal
white dots poorly developed to absent; antenna tips yellow.
HOLOTYPE a and ALLOTYPE 9, donated by the author to the Museu
Nacional, Rio de Janeiro; Areia Branca, 203 Km. south of Vilhena on the highway
to Cuiaba (Km. 575, Cuiaba-P6rto Velho), gallery forest to east of road, 650 m.
elevation, October 16, 1971, K. Brown leg.
PARATYPES: One 6 and two 9, same locality, 16-X-71; three 3 and a 9,
same locality, 23-VI-71 (Plate I, Figure 8); one 3 and two 9, same locality, 5-VII-72;
two a and two 9, same locality, 11-VII-72; one 3, same locality, 12-VII-72, all collect-
ed and retained by K. Brown; one pair, same locality, 23-VI-71, ceded to the British
Museum (Natural History), K. Brown leg; one pair, same locality, 11-VII-72, ceded
to the Allyn Museum of Entomology, K. Brown leg.; one 3, same locality, 11-VII-72
in the collection of W. W. Benson, Rio de Janeiro; two 9, l1-VII-72, and four 3
and two 9, 12-VII-72, same locality, in the collection of the Departamento de
Zoologia, Universidade Federal do Parand, O. Mielke leg.; one pair, same locality,
llpVII-72, ceded to the Museu Nacional, Rio, K. Brown leg.; one 6, 14 Km. south
of the Posto Uirapuru, Km. 562 of the Cuiabi-P6rto Velho highway, 14-VII-72,
collected and retained by O. Mielke; one 8, Rio Branco where crossed by the
trail from Salto do Ceu to Rio Vermelho, upper Rio Cabacal (Paraguay drainage),
northwest of CAceres, Mato Grosso, 400 m. elevation, 6-VI-71, collected and retained
by K. Brown (Plate I, Figure 8).
A pair in the Allyn Museum, and five 3 and one 9 in the British Museum
(Natural History), from the "CuyabA-Corumba River System" (see discussion of
this locality above), are apparently transitional between eurycleia and faleria,
possessing a more complete dennis and a less truncated yellow forewing band than
any of the paratypes (Plate I, Figure 7). An identical form (9) was captured by the
author in Vilhena on 6-VII-72.
This subspecies is expected to occur locally in deep riparian woods through-
out the forested parts of the Serra dos Parecis in central-western Mato Grosso
(Map), and perhaps may be discovered flying in other parts of the range of H.
melpomene penelope in the forests of northern Bolivia, although H. aoede lucretius
is the form present in areas just east of Lake Titicaca.
Like other known subspecies of H. aoede, eurycleia is a deep forest insect,
preferring spottily lit humid areas near streams and flying and landing principally
Text-Figure 3: Heliconius aoede eurycleia, Holotype 3 from Areia
Branca, Km. 575 of the CuiabA-P6rto Velho highway, Mato Grosso, internal aspect
of left genital valve.
in the shade. The males survey a small territory from a moderately high perch,
but are only mildly aggressive towards other males; they occasionally promenade
over small circuits inside the woods, but are not much given to flying unless dis-
turbed, and are very difficult to see when landed in the shade. The females, likewise,
spend much time landed, or flying slowly through the forest at low to moderate
elevations. As they lay their minute white eggs in large rafts, they are only rarely
seen in actual inspection of their food-plant; but they are not normally encountered
more than a few dozen meters away from this plant.
In Areia Branca, the humid streamside woods are full of Ipecac flowers
(Cephaelis ipecacuanha, Rubiaceae), which are very attractive to the local heliconians
(which may be their principal pollinators), including erato venustus and transitions
to e. phyllis (Fabricius) (Plate III, Figure 20), melpomene penelope, and eurycleia.
Other local Heliconius species include wallacei flavescens, numata superioris,
and sara thamar. From Areia Branca southward and eastward (including south
of Uirapuru, only 20 Km. southeast), eurycleia is found together with m. penelope
and e. phyllis, in a fairly large and exceptional area of non-parallel variation of
these last two species. In many areas, other subspecies of aoede may be found on
both red (Gurania and Rubiaceae) and white (Compositae, Eupatoriae) flowers.
Unlike all other species of Heliconius examined, aoede and its close relatives
metharme Erickson and godmani Staudinger do not seem to collect pollen for
amino-acid nutrition (Gilbert, 1972).
The egg of eurycleia (Plate I, Figure 10), expressed from a female, is very
small, white, and subspherical, as with other aoede subspecies, and if expressed
hatches only rarely. The probable foodplant, as with H. metharme and other sub-
species of H. aoede, is Dilkea (probably Johannesii), a primitive sapling-like
passifloraceous plant which only rarely produces flowers or characteristic tendrils.
The subspecies name is a substantive in the feminine gender, in apposition
with the genus and species names, chosen to remember Odysseus' faithful nurse,
who was a constant companion of his wife Penelope during her nineteen years
3. Heliconius ethilla chapadensis K. Brown, new subspecies
(Plate II, Figures 11-12; Plate III, Figures 15-17; Text-Figures 5-6)
Rather similar to the Colombian and Venezuelan H. e. metalilis both
Text-Figure 4: H. a. eurycleia, Allotype 9 from the same locality,
bursa copulatrix, abdominal process, and mesopretarsus.
morphologically (Text-Figures 5 and 6) and in color-pattern (Plate II, Figures
11-13), being differentiated by complete geographic isolation (separated by at
least 1500 Km. from H. e. eucoma, see Plate II, Figure 14) and in the following
additional manners (see comparison in Figures 11-12 versus 13-14):
small (diameter of larger (diam.
second about 3 mm), of second 3-4
always only three mm), often 4
almost always conti- always discon-
nuous, or if narrow,
interrupted at vein
M:, by a broad trian-
gular black spot
angular or squarish
Hindwing always orange
median bar ventrally
rupted by black
ing vein M.3
eucoma form "numis-
maticus" Weymer, 1894
larger as in metalilis
(smaller to absent in
almost always disconti-
nuous as in metalilis;
in eucoma, broad and
continuous, with the
cubital spots reduced.
spots on ven
larger, well- usually larger,
marked sometimes yellowish
Text-Figure 5: Heliconius ethilla chapadensis, Holotype 3 from Buriti,
Chapada de GuimaIes, Mato Grosso, internal aspect of left genital valve; and allo-
type female, same locality, bursa copulatrix, abdominal process, and mesopretarsus.
Forewing 38-44 mm.; one distinct red basal spot on the under surface of
the hindwing, characteristic of all ethilla subspecies (K. Brown, revision in pre-
paration); marginal white streaks on the dorsal and ventral hindwing usually
HOLOTYPE 8 and ALLOTYPE 9 donated by the author to the Museu
Nacional, Rio de Janeiro; Colegio Evang6lico de Buriti, Chapada de Guimaries,
central Mato Grosso, 650 m. elevation, May 23, 1969 and June 9, 1970, respectively,
K. Brown leg.
PARATYPES: from the same locality, collected and retained by K. Brown:
one a 28-XI-68, five 8 and one 9 23-V-69, one pair 24-V-69, three S 25-V-69, one S
26-V-69, five S 27-V-69, one S 6-VI-70, one S 9-VI-70, one 6 16-VI-70, one a 19-VI-70,
one S 29-V-71, five S 30-V-71 (these will be opportunely distributed to major
Heliconius collections); from the same locality, one a 12-II-67 and one S 24-1-70
in the collection of Nirton Tangerini, Rio de Janeiro; one S 25-V-69 in the collection
of S. S. Nicolay, Virginia Beach, Virginia; and one $ and two 9 collected by C. L.
Collenette, 17 to 30-VI-1927, in the British Museum (Natural History), numbers
1474, 1724, and 1493 (Joicey leg.).
From "Cuyaba" (probably also the Chapada area), three 5, in Drawer 1925
of the U. S. National Museum, Washington.
From "Amazons", collected by H. H. Smith (probably also Chapada),
one 5 in the British Museum (Natural History).
From "Matto Grosso", one pair, labelled "numata illustris" and numbered
(Staudinger/Bang-Hass??) 825 and 827, in the Allyn Museum of Entomology,
and one 5, numbered 717, in the British Museum (Natural History).
From Sao Jos6 da Serra (Cabeceira do Rio das Mortes), 102 Km. east of
CuiabA on the road to Rondon6polis, Mato Grosso, 650 m. elevation, one 5
22-V-69, collected and retained by K. Brown.
From Sao Vicente, 90 Km. east of CuiabA on the road to Rondon6polis,
600 meters elevation: one 9 28-V-69, one 9 18-VI-70, two 9 4-VI-70, one S 15-VII-72,
Text-Figure 6: Heliconius ethilla metalilis, Rio Negro, Meta, Colombia,
5, internal aspect of right genital valve, and 9, bursa copulatrix, abdominal
process, and mesopretarsus.
one 8 24-VI-72, all collected and retained by K. Brown; one 9 15-VII-72 from a
gallery woods at Km. 97, in the Departamento de Zoologia da Universidade
Federal do Parand, O. Mielke leg.
From 17 Km. south of the Posto Uirapuru, Km. 562 of the CuiabA-P6rto
Velho highway, one $ 13-VII-72; two further examples were seen here on 14-VII-72.
This specimen shows some characters transitional to H. e. eucoma, but still falls
near the normal range of variation for chapadensis from farther east.
A specimen of H. ethilla eucoma taken 70 Km. northwest of Vilhena, in
Rond6nia, shows very little if any intergradation towards chapadensis (Plate
II, Figure 14). The only example of the new subspecies presently known outside
of the Chapada de Guimaraes is that captured in Uirapuru, but it should eventually
be found elsewhere in central or northwestern Mato Grosso (Serra dos Parecis).
Only 200 Km. east of the Chapada de Guimaraes, on the Serra do Roncador (Royal
Society Base Camp, Mato Grosso Expedition), ethilla occurs monomorphically
as e. narcaea Godart 1819 (fide D. Gifford, Edinburgh), while a northeastern
Brazilian variant of e. eucoma has recently been collected by W. W. Benson in the
highlands of central Goids (see Map).
Most subspecies of ethilla, including chapadensis, are open-woods butterflies,
adaptable to second growth and even city gardens, but definitely preferring
forest edges in steep areas. The males choose small clearings treefalls, trails,
ridgetops, or cuts usually near water, to promenade back and forth during the
warmer hours of the day, often landing on moderately high leaves to survey the
area, or chasing other males which are promenading in the same area. The females
stay more in the shade, behaving much like the ithomiines (usually Tithorea,
Melinaea, and Mechanitis) which they resemble in color-pattern, and briefly
visiting favored flowers for nectar and pollen from early dawn through afternoon;
most oviposition occurs near midday, and the females and larvae accept a fairly
wide range of passion vines.
The probable foodplant of chapadensis in Buriti is Passiflora (Granadilla)
cornuta; larvae were readily bred through on P. (Distephana) glandulosa and would
probably accept most silvaniform/melpomene foodplants without difficulty.
The early stages (Plate III, Figures 15-17) are typical for ethilla (see Beebe, Crane,
and Fleming, 1960), although the egg is much larger than that of e. metalilis,
approaching the size of that of e. ethilla or e. narcaea. The mature larva (Figure 18)
is almost devoid of dark spots, as with essentially all ethilla larvae.
The subspecies name is an adjective in the feminine genitive singular,
derived from the name of the geographical region where the form occurs.
4. Dione juno suffumata Hayward, 1931: NEW STATUS
(Plate III, Figure 18)
This subspecies is readily distinguished from the nominate form by the
extensive invasion of the red-orange areas on the upper surface by the black borders,
producing a very dark forewing apex and broad black hindwing margin. The useful
parameters are as follows:
width of black border (tangent to minima in adjacent cells) along:
25 juno juno (Rio de Janeiro 16 juno suffumata
and Mato Grosso) (Brasilia area)
FW vein M 1.5-4.3 mm 4.2-6.0 mm (males)
(latter in females) 9.5-11.0 mm (females)
FW vein Cu, 1.0-2.5 mm 1.7-3.3 mm
HW vein M, 3.0-4.5 mm 4.2-6.0 mm
HW vein Cu., 3.2-5.2 mm 5.0-7.2 mm
The red-orange color of the underside is also of a deeper red than in most
other juno, and the underside is more contrasted in color (yellow/dark scaling)
than in specimens from other regions.
Although this subspecies was originally described as an aberration from
a single specimen from Paraguay, it occurs essentially monomorphically in the
Brasilia area and may eventually be found in other parts of central or south-central
Brazil. In the CuiabA area of Mato Grosso, populations show occasional dark forms
near suffumata, but are mostly j. juno (Plate III, Figure 19), with occasional light
specimens approaching the far v'estern j. andicola Bates, 1864 also present.
Specimens examined (collection of K. Brown, to be distributed opportunely):
Parque do Gama, DF, three a and one 9 9-VI-66, three $ and one 9 14-V-69,
two $ and one 9 17-V-69, two 8 22-V-71.
Parque Zool6gico, Brasilia, one S 3-11-67.
Fercal, Ribeirao da Contagem, DF, one & 8-11-70.
Brasilia Country Club, DF, one a 15-V-69, one a 8-11-70.
The author is grateful for discussions, information, hospitality, and field
assistance, to the Col6gio EvangBlico de Buriti (especially Ary Antonio Nogueira,
Dale Tutje, Gordon Trew, Adail Sandoval), E. W. Schmidt-Mumm (Bogota)
Woodruff W. Benson (Rio), Olaf H. H. Mielke (Curitiba), Stanley S. Nicolay
(Virginia), Nirton Tangerini (Rio), Karl Ebert (Rio Claro, Sao Paulo), Michael G.
Emsley (Virginia), Alfredo do Rlgo Barros (Museu Nacional, Rio), Lee D. Miller
(Allyn Museum, Sarasota, Florida), Helmuth Holzinger (Vienna), R. I. Vane-Wright
and T. G. Howarth (British Museum), P. Viette (Paris Museum), W. Forster (Munich
collection), F. M. Brown (American Museum), and W. D. Field (U. S. National
Museum). Financial support is gratefully acknowledged, from the Conselho Nacional
de Pesquisas, the Banco Nacional do Desenvolvimento Econ6mico, the Ministerio
do Planejamento (FINEP/FNDCT), the National Science Foundation (GB 5389
XI), and the CPEG of the UFRJ.
Beebe, W., J. Crane and H. Fleming, 1960. A comparison of eggs, larvae, and pupae
in fourteen species of heliconiine butterflies from Trinidad, West Indies.
Zoologica (New York), 45: 111-154.
Brown, K. S., Jr. and 0. H. H. Mielke, 1972. The Heliconians of Brazil (Lepidoptera:
Nymphalidae). Part II. Introduction and general comments, with a supple-
mentary revision of the tribe. Zoologica (New York), 57: 1-40.
Emsley, M. G., 1965. Speciation in Heliconius (Lep.: Nymphalidae): morphology
and geographic distribution. Zoologica (New York), 50: 191-254.
Gilbert, L. E., 1972. Pollen feeding and reproductive biology in Heliconius
butterflies. Proc. Nat. Acad. Sci., USA, 69: 1403-1407.
Collenette, C. L. and G. Talbot, 1928. Observations on the bionomics of the Lepidop-
tera of Matto Grosso, Brazil. Trans. Royal Ent. Soc., London, 76: 391-416
Hayward, K. J., 1931. Catalogo sinonimico de los rhopaloceros Argentinos,
excluyendo "Hesperiidae". Acta Zool. Lilloana, 9: 85-281
GO 9 IUIb 50 I 1 000 -
Heliconius aatraea rondonia Helieonius aoede euryeeia_
0 Known collecting points V Known collecting points
%; Probable range limits Probable range limits
Heliconius ethilZa chapadensis Dione juno suffumata
* Known collecting points Known collecting points
*..**Probable range limits .ewii Cities and towns
,H.e.narcaea northwestern ---- State boundaries
Points and range limit -- National boundaries
/, .npr. southern points -m Cuiabi-P6rto Velho hi
and probable range limit
Plate I, Heliconius. All butterflies slightly less than life size (about 0.85x),
colored black, yellow and red; eggs 17x life size, creamy white. Fig 1, H. astraea
rondonia subsp. nov., Holotype 3 (upper figures) and Allotype 9 (lower figures),
dorsal (left) and ventral (right) surfaces; Km. 70, Vilhena-Pimenta Bueno highway,
Rondonia, October 19, 1971 (Museu Nacional, Rio de Janeiro, K. Brown leg.).
2, H. a. astraea, Sao Paulo de Olivenca, Amazonas (Museu Nacional, Rio). 3,
H. egeria hyas, Mauls, Amazonas (Museu Nacional). 4, H. astraea rondonia, egg,
from Allotype 9. 5, H. aoede eurycleia subsp. nov., Holotype 3 (upper figures) and
Allotype 9 (lower figures), dorsal (left) and ventral (right) surfaces; Areia Branca,
Km. 575, Cuiabd-P6rto Velho highway, Mato Grosso, October 16, 1971 (Museu
Nacional, Rio de Janeiro, K. Brown leg.). 6, H. aoede lucretius, Mitu, Rio Vaupes,
COLOMBIA (collection K. Brown: J. Glassberg leg.). 7, H. aoede faleria with
visible eurycleia infusion, "CuyabA-CorumbA River System" (probably southeast
Rond8nia) (Allyn Museum of Entomology). 8, H. aoede eurycleia, three Paratypes,
dorsal surface; upper two topotypical, 23.vi.1971, lower from trail from Salto do
C6u to Rio Vermelho, Alto Rio Cabacal, western Mato Grosso (collection K. Brown).
9, H. melpomene penelope, Areia Branca (collection K. Brown). 10, H. aoede
eurycleia, egg, from Paratype Q.
Plate II, Heliconius. All butterflies 0.85x life size; orange, yellow and black.
Fig. 11, H. ethilla chapadensis subsp. nov., Holotype 8 (upper figures) and Allotype
9 (lower figures), dorsal (left) and ventral (right) surfaces; Buriti, Chapada de
Guimaraes, central Mato Grosso, May 23, 1969 and June 9, 1970 (Museu Nacional,
Rio de Janeiro, K. Brown leg.). 12, H. ethilla chapadensis, five Paratypes from
Buriti and Sao Vicente, Mato Grosso, 1969-1972 (collection K. Brown). 13, H.
ethilla metalilis, three variants from the Rio Negro, Meta, COLOMBIA (collection
K. Brown). 14, H. ethilla eucoma. Upper: typical individual from Km. 70, Vilhena-
Pimenta Bueno highway, Rond6nia, June 22, 1971; center: frequently captured form,
Obidos, Pard, September, 1970; lower: form "numismaticus", (bidos, August,
1967 (collection K. Brown).
Plate III, Heliconius and Dione. Egg 17x life size, yellow; third instar
larva 4x, creamy; fifth instar larva 2x, white, both with yellow-orange head;
butterflies 0.85x life size, orange and black (18-19) or black, yellow and red (20).
15, H. ethilla chapadensis, egg from Paratype Q. 16, Same, third instar larva,
lateral view. 17, Same, mature (fifth instar) larva, lateral view. 18, D. juno
suffumata, 8 (upper figures) and 9 (lower figures), dorsal (left) and ventral (right)
surfaces, Parque do Gama, Distrito Federal (collection K. Brown). 19, D. j. juno,
dorsal (left) and ventral (right) surfaces, Sio Vicente, 90 Km. east of Cuiaba,
Mato Grosso (collection K. Brown). 20, Variants of Heliconius erato from the hy-
bridized population at Areia Branca, Km. 575 of the Cuiabi-Pbrto Velho highway,
Mato Grosso, all on October 16, 1971 (collection K. Brown).
*~ .~ .4 4 4
Plate IV. Mimetic complex of dennis-rayed heliconians in southeastern
Rondonia (between Vilhena and Riozinho), in the Bolivian/Madeiran blend zone.
All 0.4x, black, yellow and red (collection K. Brown). A, Heliconius astraea
rondonia, $ (left) and 9 (right), Km. 70, Vilhena-Pimenta Bueno. B, H. elivatus
aquilina Neustetter, Riozinho. C, H. elivatus perchlora Joicey & Kaye, Riozinho.
D, H. b. burneyi, Km. 70. E, H. xanthocles nr. paraplesius, Km. 70. F, H. xanthocles
melete, Km 70. G, H. melpomene madeira, Riozinho. H, H. d. doris, form "delila",
Km. 70. I, H. aoede faleria x lucretius, 3, Riozinho. J, H. a. faleria, 9, Riozinho. K,
Eueides tales pythagoras Kirby, near form "barcellinus" Zikan, Km. 70. L,
H. erato, three variants from Km. 70; center is near typical vetustus. M, H. demeter
eratosignis, a (upper figure) and two 9 variants, Riozinho.