OL, The Journal of Research
g;J|ON THE Lepidoptera

Volume 39

2000 (2006)

The Journal of Research on the Lepidoptera
ISSN 0022 4324

Published by:

Founder:
Edu'orial Staff:

The Lepidoptera Research Foundation, Inc.

9620 Heather Road

Beverly Hills, California 90210-1757

TEL (310) 399 6016

E-MAIL: Editorial: mattoni@ucla.edu

Business: mattoni@ucla.edu

Technical: jrl_lepidoptera@yahoo.com

William Hovanitz (1915-1977)

Andrew Warren, editor

Rudolf H.T Mattoni, assistant editor

Nancy R. Vannucci, managing editor

E-MAIL: jrleditor@yahoo.com
E-MAIL: mattoni@ucla.edu
E-MAIL: jrl_lepidoptera@yahoo.com

Associate Editors: The Lepidoptera Research Eoundation is in process of reorganization.

A board of associate editors will be established.

Manuscripts and notices material may at present be sent to the assistant editor, Rudolf H.T. Mattoni, E-MAIL:
mattoni@ucla.edu. The editor is in process of relocation, btu either the assistant or managing editors can be
contacted at the addresses given.

Please note the instructions to atuhors on the back inside cover of this Journal

M.vn.aginc editor at the address: Nancy R. Vannucci, Maza 3340, Moron B1708GOP, Buenos Aires, Argentina

The journal is sent to all members of the Eoundation.

Classes of membership: Regular (Individual)

Contributing

Student/ Retired-Worldwide
Stibscription Rate/ Instittitions
Life

$ 25.00 year (vol.)

$ 30.00 or more year (vol.)
$ 20.00 year (vol.)

$ 35.00 year (vol.)

% 300.00

STATEMENT OF OWNERSHIP AND MANAGEMENT

THE JOURNAL OE RESEARCH ON THE LEPIDOPTER\ will be published two times a year by the LEPIDOPTERA
RESEARCH FOUNDATION, INC. Publication and business offices are located at the Beverly Hills, California
address given above. The Foundation is a non-profit organization incorporated in the State of California in 1965.
The president is Rudolf H. T. Mattoni, the vice-president is Jeremiah Ceorge, the secretary-treasurer is Leona
Mattoni. The board of directors (2005-) is comprised of Konrad Fiedler, Dan Rubinoff, Jeremiah Ceorge, and
Rudolf H. T. Mattoni.

Past issues are .available .at our temporary website : www.doylegroup.harvard.edti/~carlo/JRL/jrl.html

jnumul of Research on the Lepidoptera

39; 1-7, 2()()() (2006)

Total sperm ejaculation in monandrous (Papilio machaon) and
polyandrous {P. xuthus) swallowtail butterflies (Lepidoptera:
Papilionidae) restricted to larval stage-derived nutrients

Mamoru Watanabe

Institute oi Biological Sciences, University of Tsuknha, 'Isukuba, Ibaraki 305-8572, J
K-mail: walanabe@kankyo.envr Isukuba. ar.jp

Taihei Kobayashi

Institute of Biological Sciences, University of Tsukuba, 'Isukuba, Ibaraki 305-8572, Japan

Abstract: Potential capacities of male swallowtail butterflies for ejaculation containing siDerin were
examined in laboratory studies in monandrous {Papilio machaon) and |5(dyandrons {P. xuthus)
species. Virgin males transferred a spermato|thore and accessor)' substances that corresponded to an
average of 2.4% and 2.1% of their body weight at eclosion in P. xuthus And P. machaon, respectively.
The spermatophore contained about 41 and 120 eupyrene sperm bundles and 247,000 and 202,000
apyrene spermatozoa for P. xuthus -And P. machaon, respectively. Individual eupyrene and apyrene
spermatozoa of P. machaon were smaller than those of P. xuthus. I lowever, when the males re mated
at two days after the first mating, they |troduced a spermatophore of 55% and 32% of full si/e
in P. xuthus and P. machaon, respectively. The ntunber of eupyrene sperm Irundfes in the second
spermatophore increased while the apyrene .s|)ermatozoa decreased in P. xuthus. I'he larger number
of eupyrene sperm Itundles and the relatively smaller number of a|tyrene spermatozoa in the first
and the second mating of P. machaon compared to P. xuthus are discussed from the viewpoint of
female monandr)' and polyandry.

Key words: apyrene sperm, eu]5yrene .S|)ei'm bntidle, Papilio machaon, Papilio xuthus, second mating,
spermatophore.

Introduction

In many bntterfly species, the male’s second sper-
matophore is significantly smaller than his first (Svard
& Wikhmd 1986, 1989; Oberhanser 1988). Before
production of a second spermatophore of full size,
a resting period of several days is required, during
which the male’s activity consists of feeding, flying,
avoiding predators, and so on (Watanabe & Hirota
1999). The copula duration of the second mating is
often significantly longer than that of the first, [irob-
ably due to the size of the ejaculate (Svard Sc Wikhmd
1986; Kaitala & Wiklimd 1995).

Among the yellow swallowtail butterflies in Japan,
females of Papilio xuthus Linne (Papilionidae) mate
three times during their life span while females of
P. machaon hippocrates C. et R. Felder are apt to be
monandrous (Watanabe & Nozato 1986). For poly-
androus species like P xuthus, each mating provides
the female with nutrients and some alkaloids with
both transferred to her via the male’s spermatophore
(LaMimyon & Eisner 1994). It has been suggested that

Recehied: 13 December 2003
Accepted: 14 February 2005

butterfly females can use such male-derived nutrients
for somatic maintenance, to enhancement of their
fecundity, and to increase fitness of their offspring
(Friedel & Gillott 1977; Watanabe 1988; Boggs 1990).
The potential importance of male-derived nutrients
might correlate positively with the size of the male
ejaculate and the degree of female polyandry (Svard
& Wikhmd 1989). When the male re mates, he must
jjroduce another spermatojihore of sufficient size
with a sufficient number of sperm in order to fertilize
as many eggs as possible. Therefore, the duration of
the male’s recovery time and the rate of lecovery in
terms of spermatophore size and sperm number are
important for males under sperm competition (Wa-
tanabe & Hirota 1999; Wedell & Cook 1999a). The
former correlates with the duration of the female’s
non-receptive refractory period following mating
(Kiiitala & Wikhmd 1995), and high sperm numbers
seemed to be advantageous in sperm competition
(Parker 1992). Rather than utilizing the sperm of all
her mates, the female uses predominantly those from
the last partner in P. dardanus (Simmons &: Siva-Jothy
1998). On the other hand, for monandrous species
like P. machaon, the female may not exercise such
postcopulatory sperm selection (Sims 1979). Intact

/. lifs.Lepid.

spermatophores have been observed in the bursa
co])nlatrix of aged females of inonandrous species,
such as P. machaon (impnblisbed) and Lycaena phlaeas
(Watanabe & Nisbimnra 2001), suggesting that these
females did not use male ejaculate to obtain the nu-
trients necessary for reprodtictive success.

Most butterfly species produce two types of sperm,
enpyrene and apyrene, which can often represent up
to 90% of the total sperm number (Cook & Wedell
1996; Watanabe et al. 1998a, b) . Males vary the size of
spermatophores produced as well as the mimber of
both types of sperm transferred de})ending on their
mating status (Cook & Wedell 1996; Wedell & Cook
1999b). Previously mated males of /I xu//m.v produce
spermatophores that are about half the size of those
they [jrodnced as virgins, when they used nutrients
derived only from their larval stages (Watanabe &
Hirota 1999). However, there have been no studies
on sperm competition in swallowtail butterflies. The
aim of this study was to compare male’s exploitation
of larval stage-derived nutrients for spermatophore
production between monandrous and polyandrous
swallowtail butterfly species in terms of copula dura-
tion, spermatophore mass, and the number of apyrene
and enpyrene spermatozoa produced.

Materials and methods
General methods

One and three-day-old males were mated with vir-
gin females of both P. xuthus And P. machaon. Sperm
transfer h'om the males to spermatophore was assessed
immediately after copulation. All animals were ob-
tained from a continuously breeding culture reared
in the laboratoiy at room temperature in the summers
of 2001 and 2003. Adults were collected in the held,
and females were allowed to oviposit on leaves of Citrus
unshu (Rutaceae) for P. xuthus'Awd those of Heraclmm
lanatum (Umbelliferae) for P. machaon. Larvae of both
species were reared in the laboratory under an 18-h
light (>h dark regime at 25°(i to avoid diapause. All the
pupae were placed in an emergence cage to ensure
normal eclosion. Adults were weighed on the day of
eclosion and given a unique mark on the hind wing
with a felt tipped pen. Then, males were maintained
in flight cages (30x40x45 cm) without intake of any
nutrients; i.e., they were fed on only water during
their life span. On the other hand, females were kept
in individual envelopes (14L10D light cycle at a tem-
perattire of 26-27°C) to maintain their virginity. The
mean weight of male at emergence was 453.5±63.6 mg
for P. xuthus And 588. 1±128.3 mg for P. machaon (±SD) .
One of the best indicators of nutrients obtained dur-

ing the larval stage is butterfly mass.

Because the mating ability of males on the day of
eclosion is low, we hand-paired one-day old males with
virgin females (one to three-days old) in the morning.
Hand-pairing is widely used in the laboratoiy rearing
of butterflies (West 1983; Scriber & Lederhouse 1988;
Watanabe & Hirota 1999). In this study, we brought
the tips of the male and female abdomens together,
squeezed the male in order to open his claspers, and
then joined him with the female. After joining, the
pair was removed to a small cage and the copula dura-
tion was recorded.

Males were returned to the flight cages immediate-
ly after their hrst mating and held to attempt a second
copulation three-day.s-later with virgin females. A total
of 26 mated males for P. xiith us and 24 mated males
for P. machaon were hand-paired. Spermatophore
mass and the number of enpyrene sperm bundles
and apyrene spermatozoa in the spermatophore were
determined. Becau.se the accessory substances were
ejaculated otUside the spermatophore into the bursa
copulatrix of the female, we could easily separate them
during the di.ssection of the bursa copulatrix.

Immediately after copulation, the females were de-
capitated and di.ssected under a stereo microscope to
measure the weight of ejaculates in the bursa copula-
trix. The intact spermatophore and gel-like accessory
substances were weighed (accuracy, 0.001 mg). Males
that had successfully mated twice were returned to
flight cage to attempt a third copulation.

Sperm counting procedure

The procedure used for sperm counting has
been described in detail iit Watanabe et al. (1998a,
b). We evaluated the number of both types of sperm
derived in females by counting the number of sperm
in the transferred spermatophore. Immediately after
copulation, the enpyrene sperm are packed in the
spermatophore in a bundle, whereas the apyrene
bundle has already dissolved, and thus tlie apyrene
sperm can be observed individually (Kiitsuno 1977).
Enpyrene sperm bundles are clearly visible at x40 mag-
nification and apparently uniform in size. Therefore,
we first mechanically disrupted the spermatophore
and directly counted the eiq^yrene sperm as a bundle
using a stereo microscope. Then, all of the ejaculate
was washed out into a small tube containing a known
volume of saline solution (Ringer’s solution for in-
sects). The tube was gently stirred for 1 min in order
to homogenize the sperm suspension. A total of six,
10 pi subsamples were removed from each sample us-
ing a Gilson autopipette and allowed to diy on slides
under dust covers. The dry slides were then dipped

39: 1-7, 2000 (2006)

3

for about 3 sec in distilled water and again allowed to
dry. Each subsample was examined under a darkfield
phasecontrast microscope (xlOO) to count the num-
bers of apyrene sperm and etipyrene single sperm.

The total number of apyrene spermatozoa in a
spermatophore was calculated by multiplying the
average sperm count in 10 pi of suspension by its dilu-
tion factor. All means are shown with their standard
errors. All weights were determined to the nearest
0.001 mg.

Results

All one-day-old (one day after eclosion) virgin
males of both species were successfully mated to
females by the hand-pairing. Each mean duration
for copulation of the first mating was about one
hour, without a significantly difference between the
species (Table 1). Although several mated males of
both species failed to copulate in the second mating,
the copula duration of successfully re mated males
was also about one hour, which was not significantly
different from that of the first matings. The reasons
for failure included inability to conjoin or failure to
transfer a spermatophore after long in copula dura-
tion. No males successfully mated a third time.

There was no relation between copula duration and
the spermatophore weight in virgin males. Table 2
shows that a virgin male of P. xuthus transferred a
spermatophore of about 5.5 mg with accessory sub-
stances of about 5.2 mg to the female bursa copulatrix.
Therefore, virgin males produced ejaculates of 10.7
mg, or 2.4% of their body weight at eclosion. Mated
males then produced a subsequent spermatopbore of
3.0 mg with accessory substances of 2.9 mg, both of
which weights were 55% of those from the first mat-
ing.

For P. machaon, on the other hand, a virgin male
transferred a spermatophore of about 8.0 mg with
accessory substances of about 4.1 mg to the female
bursa copulatrix, and both of these weights were sig-

nificantly different from those of 71 xuthus (Table 2).
Therefore, virgin males produced ejaculates of 12.1
mg, or 2.1% of the body weight at eclosion. Mated
males then produced a subsequent spermatophore of
2.6 mg, which was not significantly different from that
of mated males of 71 xuthus. This value was 32% of the
spermatophore size of the first mating. The mass of
accessory substances (1.3 mg) from tbe mated males
was also 33% of that from the virgin males.

There was no relationship l^etween the spermato-
phore weight and the copula duration in the mated
males (F=0.971 in P xuthus and F=().333 in P. mach-
aon). Since the copula duration at each mating for
each species was roughly one hour, the ejaculation
speed was a function of the mass when the copulation
terminated. Therefore, the speed at the first mating
was higher than that at the second mating, and the
difference of the speed between the first and the
second mating in P. machaon was longer than that in
P. xuthus. In addition, spermatophore weight of the
mated males depended on that of the first mating in
P. xuthus (F=10.365, P<0.()1). Males that produced
a larger spermatophore at the first mating also pro-
duced a larger one at the next mating, though the
mass of the second spermatophore was generally
half that of the first mating. By contrast there was no
trend in regard to spermatophore mass between the
first and the second mating in P. machaon (F=().333).

All spermatophores in this study included both
eupyrene sperm bundles and apyrene sperm. Since
the bundles were usually coming loose and meander-
ing widely under the microscope, we measured the
length of a free eiqjyrene spermatozoon instead of
the bundle. For P. xuthus, the length of a eupyrene
spermatozoa was 916. 7±44.() mm (10 males), and that
of an apyrene spermatozoon was 627.9±8.3 mm. For
P. machaon (10 males), the length of a eupyrene sper-
matozoon was 562.5±1 1 .0 mm, and that of an apyrene
one was 274.5±10.5 mm. Therefore, both eupyrene
and apyrene spermatozoa of 77 machaon were smaller
than those of P. xiithus (U=0.00, P<0.01 and U=0.00,

Table 1. Copula duration for P. xuthus and P. machaon males that were successful at the first and the second mating (min,
±SE, Mann-Whiteney U-test).

1st mating

2nd mating

P. xuthus

59.3±4.5 (n=15)

62.6±3.4 (n=15)

U=91.5, n.s.

P. machaon

55.3+2.7 (n=13)

56.3+3.3 (n=15)

U=92.5, n.s

U=87.(), n.s.

U=79.0, n.s.

n: number of males

/. Res.Lepid.

Table 2. Ejaculate mass transferred from P. xuthus and P. machaon males that were successful at the first and the second
mating (mg, ±SE, Mann-Whitney U-test).

Isi mating

2iid mating

.S|)einum)ph()re

It .XltlllU.'i

It machaon

5.5()±0.2.5-' (11=1.3)

8.():t±()..S3- (11= lit)

:t.()3±0.20'> (n=13)

2.6()±0.19'’ (n=16)

U= 5.0, P<().()1

U= 0.0, P<0.01

Accessoiy siibstaiice.s

It xuthus

It machaon

5.I7±().26' (ii=13)

4.()3±().,3.3‘ (n=i;t)

2.89±().16'‘ (n=15)

1..3.3±().22'‘ (n=16)

U=6.0, P<0.01

U=11.0, IMI.Ol

n: number of males
a: IJ=9.5, P<().()1
b; U=90.0, n.s.
c: LI=47.(), 0.()3>P>().()1
(1: U=I7.5, P<().()l

P<().()1, respectively).

Table 3 shows that virgin males of /■! xuthus trans-
ferred about 41 enpyrene sperm bundles in one
spermatopbore. There was no relation between the
number of bundles and the spermatophore weight
for the first mating (F=().()25), e.g. the number of
enpyrene sperm bundles did relate to spermatophore
size. In fact, the mated males transferred about (iO
enpyrene sperm bundles, which was significantly larg-
er than the number at the previous mating, though
the spermatophore weight at the second mating was
significantly lighter. There was no significant relation
between the number of bundles and the spermato-
phore weight at the second mating (F=().243).

For P. machuon, on the other hand, a virgin male
transferred about 120 enpyrene sperm bundles in a
spermatophore, and there was a positive correlation
between spermatophore weight and the number of
bundles (F=8.69(), ().()5>P>().01 ), ranging 71 to 192
bundles and (i.2 to 9.8 mg in weight. A virgin male
transferred about three times the number of bundles
in P. xuthus (Table 3). Then, the mated males of P.
machaon produced about 134 bundles, which was not
significantly different from the number of bundles at
the previous mating. There was no significant relation
between the numl)er of bundles and the spermato-
phore weight at the second mating (F=0.787).

4'here was no relation between the number of
apyrene spermatozoa and the spermatophore weight
for either the first (F=().()07) or the second (F=().534)
matings in P. xuthus. The number of apyrene sper-
matozoa at the first mating was significantly larger
than that at the second mating, suggesting that at the
second mating males tratisferred more enpyrene and

less apyrene spermatozoa irrespective of their second
mating ejaculate mass, but related to mating status.

Fhere was no relation between the number of
ajjyrene spermatozoa and the spermatophore weight
for either the first (F=1.212) or the second (F=().373)
matings in P. machaon. The number of apyrene sper-
matozoa at the first mating was signihcatitly smaller
than that in P. xuthus, but not significantly larger than
that at the second mating.

Discussion

Although copula duration in butterflies seems to
be under male control (Wickman 1985), one of the
possible costs of mating incurred by males in transfer-
ring spermatophores is time. Many butterflies try to
decrease copula duration against the risk of preda-
tion (Rutowski 1984), while prolonged mating has
a function other than just spermatophore transfer
irrespective of the size. In pierid butterflies differen-
tial transfer rates are due to behavioral competition
among males (Svard Sc Wiklund 1988; Watanabe et al.
1997). In this study, however, copula duration of the
second mating with small spermatophores in both P.
xuthus and P. machaon was not longer than the first,
unlike in other butterfly species (Sviird & Wiklund
1986; Kititala & Wiklund 1995).

Male P. xuthus produce full-size spermatophores
when virgins, or when they mate after several days of
abstinence (Watanabe & Hirota 1999). The present
study shows that, during an abstinence period, mated
males produce a spermatophore that is 55% {P.
xuthu.s) and 33% {P. machaon) of the size of a full-size
spermatophore. In many butterfly species, the second

39: 1-7, 2000 (2006)

5

Table 3. Number of eupyrene sperm bundles and estimated number of apyrene spermatozoa in a spermatophore transferred
from P. xuthus and P. machaon males that were successful at the first and the second mating (±SE).

1st mating

2nd mating

Eupyrene sperm bundtes

F. xuthus

P. machaon

41.3±.5.76'‘ (n=14)

1 19.8±13.7{)' (n=10)

.59.6±6.86'’ (n=14)

I33.7±l 4..3‘)'> (n=16)

U=.’i9.0, P<().05

U=72.5, n.s.

Apyrene spermatozoa

P. xulhus

P. machaon

247,43.6±30,I97^ (n=13)

202,143±.3.3,719' {n=10)

16r>,100±2(),8.W (n=14)

165,()98±2(),888" (n=l(i)

U=52.(), P<().05

U=61.(), n.s.

n: number of males
a: U=4.0, P<0.01
b: U=35.0, P<0.01
c: U=39.0, 0.05>P>0.01
d: U=86.0, ii.s.

spermatophore is smaller than the hrst (Sims 1979;
Oberhatiser 1988; Svard & Wikhmd 1989; Watanabe
et al. 1998a). Larger spermatophores induce longer
periods of sexual imreceptivity in some lepidopteran
females (Oberhaiiser 1989), during which females
lay eggs exclusively fertilized by that male. In CoUas
eurytheme, female fecundity has been shown to corre-
late positively with the size of the received ejaculate
(Rutowski et al. 1987). Therefore, a reduction in
spermatophore size may potentially be costly in terms
of a reduced female reproductive output and shorter
female refractory periods. In addition, He & Miyata
(1997) have shown that small spermatophores might
contain less eupyrene and apyrene spermatozoa than
large ones. In the present study, however, the number
of eupyrene spermatozoa in the second spermato-
phore produced by the male was more than that in
the hrst spermatophore for P. xulhus and identical
to that in the hrst spermatophore for P. machaon.
Males of both species were fed only water as adults,
meaning that they were using only larval nutrients to
produce sperm.

Cook& Wedell (1999) stated that in many butterfly
species there are typically 256 eupyrene spermatozoa
per bundle. Therefore, the range of approximately
10,000 (=41x256 in P. xulhus) to 34,000 (=134x256
in P. machaon) eupyrene spermatozoa was above the
level guaranteeing adequate fertilization, since a fe-
male of either species has fewer than 1000 eggs after
eclosion (Watanabe & Nozato 1986). Furthermore,
a single eupyrene spermatozoon might be used for
insemination of a single egg (Watanabe & Hachisuka
2005). Wedell & Cook (1998) pointed out that mated
males compensate by increasing sperm numbers.

resulting in high fertilization sncce.ss during sperm
competition.

For females, the mean lifetime mating frequency is
3 and 1 for P. xuthus 'and P. machaon, respectively (Wa-
tanabe & Nozato 1986). Svard & Wikhmd (1989) also
showed that the mean number of mating in the Euro-
pean P. machaon was 1.16. Assuming that every female
attempted to load a sufhcient number of eupyrene
spermatozoa for insemination of her eggs, the total
number of eupyrene spermatozoa in relation to her
fecundity must be similar among si)ecies, because the
number of eggs in the ovaries was roughly same. A
female ot' P. xuthus mated with 3 virgin males during
her life span would received 123 (=41x3) eupyrene
sperm bundles. This is comparable with the number
received at a single mating by a female of /f machaon,
though the eupyrene spermatozoon size of /I machaon
was smaller than that of /I xuthus. Lastly, a female of /I
xuthus mating 3 times must receive 750, 000 apyrene
spermatozoa.

For a male swallowtail butterfly second mating,
the probability of encountering non-virgin females
who accept copulation must be high in nature, due
to the long flying season and to contimtous genera-
tions during sitmmer. Hence for P. xuthus, it may be
numbers of eupyrene spermatozoa in spite of reduced
nutrient donations. The number of eupyrene sperm
bundles in the spermatophore significantly increases
for the second mating. During s])erm competition, an
increased risk with the degree of polyandry, produc-
tion of high numbers of sperm may be advantageous.
In Pieris rapae, males transferred significatitly more
eupyrene sperm on their second mating, although
the spermatophore size was reduced (Cook & Wedell

6

/. Kes.Lepid.

1996; Watanabe et al. 1998a; Wedell & Cook 1999a).
Since there might he little sperm mixing in P. xulhiis
dtie to the apyrene sperm in spermatheca (Watanabe
et al. 2000), it appears advantageous for males to
transfer an excess of enpyrene spermatozoa with an
exce.ss number of apyrene spermatozoa at the second
mating.

Male swallowtail hutterllies apparently lack means
to decrease sperm competition since they leave only
a small sperm plug after copulation. Further, males
have no provision for removing a rival male’s sperm.
Their penis cannot reach the spermatheca, where
eupyrene sperm is stored. The interval of mating and
the morphology of fertilization ducts allow etipyrene
spermatozoa from the second mating to be positioned
close to the exit. With ctil-de-sac spermatheca, the
sperm enter and leave by the same dtict, such that
sperm from the second mating should be the first
to leave, restilting in last-male priority in P. xuthus
(Watanabe et al. 2()()()). Last-male priority has been
established as the predominant pattern of sperm tise
in most butterfly species that have been sttidied (Sims
1979, Lederhotise 1981).

On the other hand, females of P. twac/mon seemed
not tise the spermatophore and accessory substances
for either energy or egg prodtiction (unpiiblislied)
and were rnonandrotis. In fact, they had larger fat
bodies in their abdomens after eclosion than P.
xuthus. This finding corresponds with body weights
of the males. Males of P. marhaon had clearly more
fat body than male P. xuthus, even though their body
sizes were similar. Tberefore, non-virgin females of
P. machnon nuist have reqtiired no further ejactilate
from males, and therefore avoided their conspecifics
and/or reftised mating when males were encoun-
tered. Since only \irgin females permitted copulation,
the operational sex ratio was biased toward males.
The probability of mating in males may be low in
the field, but the probability of re mating by mated
males must be zero if females are able to detect mated
males. Then, males of /^l machaon, particularly virgins,
might increase the number of eupyrene spermatozoa
as well as the maximum ejaculation mass as much as
possible when they coptilate. Consequently, sperm
size wotild decrease.

Althotigh the ntimber of apyrene spermatozoa in P.
xuf/uw significantly decreased for any second mating,
the life time production of apyrene spermatozoa was
higher than that in P. machnon, partictilarly in relation
to the number of eupyrene spermatozoa. Based on
the different amounts of eupyrene and apyrene spenn
in the testis (Hiroyoshi & Mitsuhashi 1999), males
independently vaiy the numbers of the two types of
sperm ejactilated into the spermatophore (Cook &

Gage 1995; Watanabe et al. 1998b; Wedell & Cook
1999a). The fact that apyrene sperm migrates earlier
than etipyrene sperm from spermatophore to sper-
matheca following copulation suggests that apyrene
sperm for the second mating can fill the spermatheca
with etipyrene sperm of the first mating in order to
prevent fertilization (Watanabe et al. 2000). There-
fore, it may be advantageous for P xuthus males to
provide polyandrous females with a certain number
of large apyrene spermatozoa at each mating to effect
sperm competition.

We have no information on the mating frequency
of male swallowtail butterflies in nature. We suspect
that males of both P xutlms and P. machaon mate at
intervals of more than a few days, extrapolated in part
by their 3-5 weeks life span in the laboratoiy Note
well that in the jtresent study P. xuthus'Awd P machaon
males were restricted to larval stage nutrients and
never mated more than three times. All males mating
only two times died the day after the second mating.
Fherefore, P xt/Z/ucs males produced a smaller number
of large eupyrene spermatozoa, while P. machaon
produced a large number of small eupyrene sper-
matozoa. However, there were more larger apyrene
spermatozoa in P. xuthus than in P machnon. Conse-
quently, P. xuthus males must provide for the extra cost
to produce eupyrene sperm and ignore the cost to
produce apyrene sperm for each mating under severe
sperm competition. By comparison P. machaon males
decrease the apyrene sperm mass under low sperm
competition in the reproductive organs of females.
Since survival, and thus mating frequency in nature,
should be supported by nutrients from nectar, field
studies comparing the mating behavior of P. xuthus
and P. machaon males are necessary.

Acknowledgements

We would like to thank M. Kamikubo of University of Tsnknba,
and A. Hachisnka and N. Dainone of Mie University, for their as-
sistance in rearing the swallowtail butterflies. Thanks are also due
to N. Uihi, Y. Nakatani and T. Yamamoto of the Kashihara City
Insectarinm, for their provision of the diet of the larvae. Comments
of two anonymous reviewers improved onr mantiscript. This work
was supported in part by a Grant-in-Aid for Scientific Research (C)
from the Ministry of Education, Science, Sports, and Culture of
Japati (No. 10640612) to M. Watanabe.

Literature cited

Boggs, C. L. 1990. A general model of the role of male-donated
nutrients in female in.sects’ reproduction. American Naturalist
136: 598-617.

Cook, P. A. & M. J. G. Gage. 1995. Effects of risk of sperm
competition on the numbers of eupyTene and apyrene sperm
ejaculated by the moth Ptodia inlerpwictetla (Lepidoptera:

39: 1-7, 2000 (2006)

7

Pyralidae). Beha\ioral Ecology and Sociobiology 36: 261-268.

Cook, P. A. & N. Wkdell. 1996. Ejaculate dynaniic.s in bntterllies:
a strategy for maximizing fertilization success? Proceedings of
the Royal Society of London B 263: 10471051.

. 1999. Non-fertile sperm delay female remating. Nature

397: 486.

Frif.del, T. & C. Gii.i.ott. 1977. Contribution of male-produced
proteins to \4tellogenesis in Mdatioljlus sangiiiniprs. ]onnr(i\ of
Insect Physiology 23: 145-151.

He, Y. & T. Mn’ATA. 1997. Variations in sperm number in relation
to larval crowding and spermatophore size in tlie armyworm,
P.seudaletia separata. Ecological Entomology 22: 41-46.

Hiroyoshi, S. & J. MmsuuASHi. 1999. Sperm reflux and its role in
multiple mating in males of a butterfly Folygonia c-aureum
Linnaeus (Lepidoptera: Nymphalidae). Journal of Insect
Physiology 45: 107-112.

K,‘\ii.ua, a. & C. WiKLUND. 1995. Female mate choice and mating
costs in the polyandrous butterfly Pieris napi (Lepidoptera:
Pieridae). Journal of Insect Behavior 8: 355-363.

Katsuno, S. 1977. Studies on eupyrene and apyrene spermatozoa
in the silk worm, Bombyx mori L. (Lepidt)ptera: Bombycidae).
II. The intratesticular behaviour of the spermatozoa after
emergence. Applied Entomology and Zoology 12: 236240.

Lamunyon, C. W. & T. Eisner. 1994. Spermatophore size as
determinant of paternity in an aectiid moth ( Utetheisa ornatrix) .
Proceeding of the National Academy of Science of the LISA
91:7081-7084.

Lederhouse, R. C. 1981. The effect of female mating frequency on
egg fertility in the black swallowtail, Papilio polyxenes asterius
(Papilionidae). journal of the Lepidopterists’ Society 35:
266-277.

Oberh.aliser, K. S. 1988. Male monarch butterfly spermatophore
mass and mating strategies. Animal Behavior 36: 1 384-1388.

. 1989. Effects of spermatophores on male and female

monarch butterfly reproductive succe.ss. Behavioral Flcology
and Sociobiology 25: 237-246.

P.ARKER, G. A. 1992. Marginal value theorem with exploitation time
costs: diet, sperm resen'es, and optimal ct)pula duration in dung
flies. American Naturalist 139: 1237-1256.

Rutowski, R. L. 1984. Sexual selection and the evolution of Initterfly
mating behavior. Journal of Research on the l.epidoptera 23:
125142.

Rutowski, R. L., G. W. Gilchrist & B. Terk/\ni.an. 1987. Female
butterflies mated with recently mated males show reduced
reproductive output. Behavioral Ecology and Sociobiology
20: 319-322.

ScRiBER, J. M. & R. C. Lederhouse. 1988. Hand-pairing of Papilio
glaucus glaucusdnd Papilio pilumnus (Papilionidae) and hybrid
survival on various food plants. Journal of Research on the
Lepidoptera 27: 96-103.

Simmons, L. W. & M. T. Si\'aJothy. 1998. Sperm competition in
insects: mechanisms and the potential for selection, pjt. 341-
434. In: Birkhead, T. R. & A. P. Moller (eds.) , Sperm competition
and sexual selection. AcademicPress, London.

Sims, S. R. 1979. Aspects of mating frequency and reproductive
maturity in Papilio zelicaon. American Midland Naturalist 102:
36-50.

SvARD, L. & C. WiKi.UND. 1986. Different ejaculate delivery strategies
in first versus subsequent matings in the .swallowtail butterfly
Papilio machaon L. BehaMoral Ecology and Sociobiology 18:
325-330.

. 1988. Prolonged mating in the monarch butterfly Danaus

plexippus and nightfall as a cue for sperm transfer. Oikos 52:
351-354.

. 1989. Mass and production rate of ejaculates in relation to

monandry/ polyandity in butterflies. BehaMoral Ecology and
Sodobiology 24: 395-402.

Watanabe, M., M. Bon’no & A. HAt:iiisuiv\. 2000. En|jyrene sperm
migrates to spermatheca after apyrene sperm in the swallowtail
butterfly, Papilio xitlhus 1,. (Lepidojitera: Papilionidae) . Journal
ofEtholog)' 18: 91-99.

Watanabe, M. & M. Hac:iii.suk.\. 2005. Dynamics of eupyrene
and apyrene sperm storage in ovipositing females of the
.swallowtail butterfly Papilio xuihus (Lejtidoptera: Papilionidae) .
Entomological Science 8: 65-71.

Watanabe, M. & M. Hiroia. 1999. Effects of sucrose intake on
spermatophore mass produced by male swallowtail butterfly
Papilio xuthusL. Zoological Science 16: 55-61.

Watanabe, M., Y. Nakantshi & M. Bon’no. 1997. Prolonged
copulation and spermatophore size ejaculated in the sulfur
butterfly, Colias erale (Lepidoptera; Pieridae) under selective
harassments of mated pairs by conspecific lone males. Journal
of Ethology 15: 45-54.

W.ataN/ABE, M. &c M. Nishimura. 2001. Reproductive output and
egg maturation in relation to mate-avoidance in monandrous
females of the small copper, Lycaena phlaeas (f.ycaenidae).
Journal of the Lepidopterists' Society 54: 83-87.

Watan.abe, M. & K. Nozato. 1986. Fecundity of the yellow swallowtail
butterflies, Papilio xuihus and P. machaon hippocrates, in a wild
environment. Zoological Science 3: 509-516.

Watan.abe, M., C. Wiktund & M. Bon’no. 1998a. The effect of
repeated matings on sperm numbers in successive ejaculates of
the cabbage white butterfly Pieris rapae (Lepidojitera: Pieridae) .
Journal of Insect Behavior 1 1: 559-570.

. 1998b. Ejaculation timing of eupyrene and a|tyrene sperm in

the cabbage white butterfly Pieris rapae (Lepidoptera: Piendae)
during copulation. Entomological Science 1: 15-19.

Wedei.l, N. & P. A. Cook. 1998. Determinants of paternity in a
butterfly. Proceedings of the Royal Society of London B 265:
625-630.

. 1999a. Strategic sperm allocation in the small white butterfly

Pieris rapae (Lepidoptera: Pieridae). Functional Ecology 13:
85-93.

. 1999b. Butterflies tailor their ejaculate in response to sperm

competition risk and intensity. Proceedings of the Royal Society
of London B 266: 1033-1039.

West, D. A. 1983. Hand-]>airing of Battus philenor (Papilionidae).
Journal of the Lepidopterists’ Society 37: 90.

WiCK.MAN, P.-O. 1985. Male determined mating duration in
butterflies? Journal of tbe Lepidopterists’ Society 39: 341-
342.

Journal of Ri'searrh on the I.(‘j>idol)lera

39: 8-17, 2000 (2006)

Flight Patterns and Mating Behavior in a Zephyrus Hairstreak,
Neozephyrus japonicus (Lepidoptera: Lycaenidae)

Michk) Imaflikh

Departnieiit of Zoology, (’iiaduate School of' Science, Kyoto L'niversity, .Sakyo, Kyoto, 606-8502 Japan
ima@ethol. zool. kyoto-u.ar. jji

Yasutoshi Matsui '■

Institute f’of Etivironinental Management Technology, National fnstititte of'Advaticed Industrial Science and Technology, 16-1 Onogawa,
Tsiiknba, Ibaraki, 305-8569 Japati

Hideko Matsui "

.5-102, 740Santiki, Ryngasaki, Ibaniki, 301-0033 Japan

' Present address: 7-106, 1-2-63 Enjimi. I ligashitnnrayama, Tokyo, 189-002 I Japati

Abstract: Elight patterns atid tnating behaviors of a /ephyrits hairstreak, Neozephyrus japonkus, were
itivestigated in a woodland in Rvngasaki (lity, Ja|)an. dtiritig its adult flight period in late June and
|nlv, Iroin 2000 to 2003. This species was most active around dusk, fiom 1 7:00 to 18:00 hrs, during
which males actively flew around the canopy and edges of the forest. (Circling flights performed by
two individitals were lre(|itently observed, as were chase flights, sometitnes cotnposed of mitlti|tle
individuals. (Circles flowti chiring circling flights averaged 14 can in diameter and 2.9 rotations/s,
tints the velocity of these flights averaged about 130 cm/s. Ehe directioti of rotatioti sometitnes
changed ditring circling llights. A third individttal occasionally joined a circling flight, which would
sometimes disrupt the flight, or rarely, would replace one of the origitial circlets, (lhase llights,
sometimes in a zig/ag pattei n, ttsualK' included a female, atid sometimes led to the initiation of
courtship. Uourtship atid copulation sometitnes occut red jitst after chase flights. Coitrtship time
was between 30 seconds and one minute, and coptilation tittle vai ied from 45 minutes to over three
hours. Examination of the bitrsa co|iulatrix of females of N. japonkus from the sttidy site indicated
the sporadic occut fence of imiltiple matings.

Key word.s: activitv titne, circling flight, chase flight, mating, Lycaenidae, sperniatophore

Introduction

Butterflie.s in the tribe Theclini, commonly called
“zephyrtrs haiustreaks” in Japan, are biologically
intere.sting for a variety of rea.sons. This relatively
small grotip shows great variation in adult and larval
morphologies, life histories, physiology, behavior
and ecology (Saigtisa 1988). In Japan, this group
includes 2,5 species in 13 genera such as CJirysozephy-
rus, NeozeJ)hynis and Favonius, whose phylogenetic
relationships have been sttidied, using morphologi-
cal (Shirozn & Yamamoto 1956) and molecular data
(Saigusa & Odagiri 2()()()) .

An intriguing aspect of the Theclini in Japan is
variation in wing colors; 13 species show sexttal di-
morphism, 1 1 species are nearly or perfectly mono-
morphic, and 1 species is in an intermediate situa-

Recewed: 10 June 200-4
Acrepled: 10 March 2005

tion. Some species of Chrysozephyrns, Neozephyrus And
Favonius sd\o\s^ the most conspicuous color differences
between sexes among butterflies occurring in Japan.
These color variations may be related to the behavior
and ecology of the various species. Males of some
species are territorial, and mate-locate by guarding
perches, whereas others mate-locate through patrol-
ling. Adult activity times vaiy among the different
species; some are active early in the morning, some
during midday, and others are most active at dusk
(Fuktida et al. 1984). However, published informa-
tion on the behavior and ecology of the zephyrus
hairstreaks is incomplete, and systematic stitches on
the ecology of these hairstreaks have not been previ-
ously conducted. This is partly because adults are
largely arboreal, and thus, makedifficult subjects for
behavioral or ecological studies.

Previously, the senior author observed mating
behaviors of one of the sexually dimorphic zephyrus
hairstreaks, Neozephyrus japonicus (Murray, 1845),

39: 8-17, 2000 (2006)

9

males of which have metallic green scales on the dorsal
surface of the wings, while females have mostly brown
or black wings above. Two attempts at courtship, and
one copulation event were witnessed, among 22 indi-
viduals released in a large cage (Imafukn et al. 2()()()).
Stibsequently, the junior authors observed copulation
events among wild pairs of TV. japonicus. Their results
were reported at the meetings of the Lepidoptero-
logical Society of Japan from 2()()() to 2002. Since
then, we have conducted a systematic investigation of
mating and related behaviors of a wild population of
TV. japonicus. Herein we present the results of those
studies, including details of mating behavior from the
initiation of courtship to final copulation, along with
information on diurnal activity, circling and chase
flights, and experimental results on male responses
to female models.

Materials and Methods

Study area

This study on the behavior oi' Neozephyrus japonicus
was conducted in a small wooded area in Ryugasaki
City, Ibaragi Pref., Japan (35T55’N 14()‘’1()’E), during
the butterfly’s adult flight season, from the end offline
to July, 2000 to 2003. The wooded area was divided

Fig. 1 . Study site in Ryugasaki City, Ibaragi Pref., Japan.
Patches of willow and alder trees were separated by a
path. A thermometer (T) and an illuminator (L) were
placed as indicated. A digital video camera (C) was
installed for observations of a fixed area, north of the
alder grove. A and B indicate sites at which individuals
of N. japonicus were marked (see Results). Wil. = willow;
Ches. = chestnut.

by a narrow lane into two sections; one composed of
tall (10-15 m) alder trees, Alnus japonica Steude], and
the other composed of willow trees (4-6 m), includ-
ing Salix subfragilis Andersson and S. cliaenomcloiclcs
Kimura (fig. 1). Alnus japonica is a local foodplant
for larvae oi' Neozephyrus jajwnicus. The study site was
surrounded by grasses such as the reed Phragmites
co?» wunisTrinius, the goldenrod Solidago allissima L.,
and also by cultivated fields.

Observation of diurnal activity

The number of butterflies observed in flight was
recorded every 15 to 60 minutes, from dawn to dusk,
by slowly walking along the lane. The air temperature
was recorded every 5 minutes with a temperature data
logger (Gemini Data Logger, temperature) settled in
the willow grove (T in fig. 1). The light intensity was
recorded by an illuminator situated in open space
(shown with L in fig. 1 ) , and data were stored every 5
minutes in a second data logger (Gemini Data Logger,
voltage) . These counts were conducted from June 30
to July 2, 2002.

Observation of flight activity using a video camera

Tliese sinweys were conducted on 4 days between
July 2 and 7, 2002. During the peak time of adult
activity, near dusk, many butterflies were observed
flying over the canopy and edges of trees, occasion-
ally performing chase flights and circling flights (or
the spiral flight described by Davies 1978). To record
a detailed temporal seqitence of changes in these
activities, a digital video catnera (Sony, DGR-rRV20)
was set at a fixed point (Ci in fig. 1) to record flying
bittterflies. The camera remained stationary, and
recorded individuals that appeared on the north
side of the alder woods. Numbers of butterflies, of
circling flights and of chase flights appearing every
10 minutes were recorded from 16;3() to 19:00 hrs.
Numbers of butterflies were estimated by treating
each individual appearing on the .screen, until the
time it disappeared from view, as separate individuals.
The air temperature and light intensity were recorded
with a thermometer and illuminator (Mother Tool,
LX-lOO).

Detailed observations and experiments

In order to document detailed patterns of circling
and chase flights, butterflies were recorded with a
hand-held digital video camera. Recorded images
were then incorporated into a personal computer
using “iMovie, 2.1.1” (Microsoft), and flight paths

10

J. Res.Lepid.

were plotted using “NIH images, 1 .62” (courtesy of
the National Institute of Health, USA). For draw-
ings in figs. 5 and 6, and in measurements of flight
velocity, a reduced scale was calculated by comparing
the image size of a butterfly oil the computer screen
to the average natural size of butterflies collected at
the study site. Two arbitrarily determined reference
points were also plotted to compensate for changes
in images caused by zooming and movement of the
camera during recordings.

As chase flights were inferred to occur on par-
ticipation of a female, two model experiments were
performed. In the first experiment (“female model,”
July 4, 2001), a female hairstreak was allowed to fly,
but was tied to the tip of a 3.3 m fiberglass rod with
60 cm of cotton string (#50) secured to her thorax,
between the fore-and hindwings. In the second
experiment (“female wing model,” July 7, 2001),
female wings were pasted on both sides of a plastic
figure (made from a plastic plate), shaped to mimic
a female hairstreak holding its wings open, which was
secured to a rod and rotated with a small motor. In
each experiment, models were situated in areas where
male hairstreaks were active.

In order to determine the sexes of butterflies
involved in chase flights, capture with an iirsect net
was attempted. For observations of mating behaviors,
binoculars and a tddeo camera were used, and detailed
\erbal descrijttions of various processes were recorded
on the latter.

Results

Diurnal activity

Adults of Neozephyrus japouicris were most active
from 17:00 to 18:00 hrs., just before dusk (fig. 2).
This period of peak activity was essentially the same
on clear days (e.g., June 30 and July 2, with the high-
est air temperature over 28”C and the maximtim light
intensity exceeding 60,000 lux) as it was on cloudy
days (e.g., July 1, with temperatures lower than 25.5‘’C
through the daytime and the maximum light intensity
lower than 20,000 lux). On clear days, some adult
activity was observed in the morning. Shortly before
the period of peak adult activity, around 16:00 hrs.,
females were frequently observed on grasses around
the woods, sometimes basking with open wings. As
time elapsed, they moved to higher sites in the trees.
Around 16:30 hrs., males started to fly over the sides
and canopy of the alder and willow groves, andocca-
sional circling flights by two males were performed.
As it became darker, chase flights occurred, in which
some butterflies appeared to be rapidly chasing other

individuals. With increasing darkness, numbers of
flying hairstreaks decrea.sed, and no butterflies were
observed by 19:30 hrs.

Evening flight activity

Observations recorded with the fixed video cam-
era during the evening period of peak activity are
shown in fig. 3. Flying butterflies first appeared at
16:55 brs., when the light intensity was 13,000 lux,
and numbers of flying butterflies gradually increased
with a peak around 18:15 hrs. (fig. 3b). The average
time of peak adult activity was calculated to be 18:07
hrs. The lightintensity at this time was about 4,000
lux. After this time, the number of flying butterflies
gradually decreased.

Circling flights started to occur as number of fly-
ing butterflies increased, and continued until flight
activity ended (fig. 3a) . The frequency of chase flights

Light int.
XIO^ lux

Time (hrs.)

Fig. 2. Changes in the number of flying individuals of
N. japonicus (thick line) on three days in 2002. Thin
and dotted lines indicate light intensity and temperature,
respectively. Two days (June 30 and July 2) were sunny,
while July 1 was cloudy.

39: 8-17, 2000 (2006)

11

Table 1 . Diameter and rotation speed of circling flights in
Neozephyrus japonicus.

Sample

Sampling

time

(s)

Diameter

(cm)

Rotation

.speed

(roimds/s)

#1

1..30

11

3.8

#2

l.-TO

12

2.9

#3

2.77

16

2.6

#4

2.93

15

2.4

#5

3.73

19

2.6

#6

2.07

12

3.2

#7

2.63

11

3.1

#8

2.63

20

2.1

#9

.3.13

12

3.3

#10

1.97

15

3.1

Average

Range

2.47

1., 30-3.73

14

1 1-20

2.9

2. 1-3.8

was greater in the later half of the evening activity
period. The greatest average number of circling
flights occurred at about 18:05 hrs., while the great-
est number of chase flights occurred slightly later, at
about 18:20 hrs.

When males were active in circling flights, captur-
ing attempts were made for butterflies straying around
a canopy or sub-canopy level, with a result that 4 out of
5 captured individuals were females. Flight behavior
of females appeared to be somewhat diflerent than
that of males, in that they flew slightly slower, and at
a lower altitude.

Circling flights

Circling flights involved two male butterflies, each
flying rapidly in a small circle. Most circling flights
were completed in a few seconds, whereas some lasted
longer. Durations of 46 circling flights observed from
July 12 to 17, 2002, are shown in hg. 4. Most circling
flights (72 %) were completed within 30 s. The long-
est circling flight lasted 6 min 18 s. The diameter of
the circle varied from 1 1 cm to 20 cm, with an average
of 14 cm (Table 1). Butterflies engaged in a circling
flight rotated 2. 1 to 3.8 times per second, 2.9 times on
average. Thus, their flight velocity in circling flights
was calculated to be about 130 cm/s.

A trace of a typical circling flight is shown in hg.
5b. Circling flights were frequently initiated when
males happened upon each other, either through
random movements (fig. 5a), or through a chase.
Rotation direction was not always fixed, and as shown
in fig. 5c, sometimes changed. In some cases, a third
individual joined into ongoing circling flights. Out
of 46 circling flights observed, 14 were invaded by a

third individual. Among these, six were continued
by the original circlers with the third butterfly fly-
ing away (fig. 5e), six were broken up, and two were
continued by the original and the newly arrived third
individual (fig. 5f) . At the end of circling flights, one
of the circlers flew away sometimes, chased by the
other (fig. 5d).

Time

Fig. 3. a. Temporal distribution of circling flights (gray bar)
and chase flights (solid bar) during the evening activity
period of N. japonicus at the study site. b. Temporal
distribution of flying butterflies at the study site. Light
intensity and temperature are shown with solid and dotted
lines, respectively.

Duration (sec.)

Fig. 4. Frequency of circling flights in N. japonicus, plotted
against their duration.

12

/. Res.I.epid.

Fig. 5. Traces of flight patterns related to circling flights of N. japonicus. In each figure, alphabetical letters (upper cases)
indicate individual butterflies and their positions in the first frame of the figure. Each point on the trace line represents the
position of the butterfly in successive frames, thus, every 1/30 s. A scale bar below each figure indicates 10 cm. a. Initiation
of a circling flight (17:12:55, July 1, 2002). b. Typical circling (18:09:44, July 1, 2002). c. Change of rotation direction; the
initial right rotation for the butterfly (left rotation for the observer below) changes to a left rotation (17:48:1 1 , July 1 , 2002). d.
End of circling, resulting in a chase of one butterfly (A) by the other (B) (18:14:52, July 1 , 2002). e. Approach and subsequent
retreat of a third butterfly (C) to a circling flight (1 7:34:21 , July 1 , 2002). f. Replacement of circlers. One of the initial circlers
(A) continues to circle with the invader (C), while the other circler (B) retreats (18:00:37, July 12, 2002).

Residency

(Circling flight.s or .spiral flights are usually shown
by territorial species (see Discu.ssion). Males of Ne-
ozrphyrii.s japonicus in the study area did not show clear
territorial behavior, but some males repeatedly flew
around a specific jjarts of the canopy. Therefore,
residerrcy was examined for orrr poprrlation. Two but-
terflies at points A atrd B on fig. 1 wer e captur ed and
marked with a black felt-tip pen on the irnder sides of
the wings at aborrt 18:00 hr s on Jrrne 29, 2002. The
birtterlly marked at point A was recaptur ed at the
same site arrd at similar times the following 3 days,
irntil obser vations at the site were terminated. The
brrtterfly rrrarked at point B was recorded from the
sarrre site the following 2 days, brrt not on the last

day of the study. Thrrs, butter flies in the study field
exhibited a certain degr ee of r esidency.

Chase flight

Chase flights were occasionally initiated immedi-
ately after circling flights, and sometimes involved
multiple individirals (fig. 6d). Some chase flights
were very persistent, covering a lar'ger distanceim-
mediately after cir cling flights. Rar ely, zigzag flights
involving two irtcliviclirals were obser ved. A tr ace of a
zigzag flight is shown in fig. 6b, whei'e one birtterfly
(A) appr oaches another (F), and proceeds to follow
it closely. The chasing birtter fly (A) showed a similar
tirrning pattern to that of the butter fly being chased
(F), with a delay of 2 to 3 frames (67 to 100 msec).

Fig. 6. Traces of flight patterns related to chase flights of N. japonicus. Presentations are the same as in Fig. 5, unless otherwise
noted, a. A persistent chase of a presumed female (F) by a male (A). Traces were interrupted because of tree branches; 12,
31 and 24 frames were lacked at circles 1 , 2 and 3, respectively. This chase continues at least 5.8 s (1 7:51 :26, July 2, 2002).
b. A zigzag flight. A presumed female (F) is chased by a male (A) which approaches from below (17:02:56, July 2, 2002). c.
Flight path of the presumed female prior to chase by a male, showing a “swinging” flight. The final point of this figure continues
to the start point of figure b (17:02:54, July 2, 2002). d. A group chase. A presumed female (F) is always at the lead. “The
male “C” appears in the figure after at the 12th frame (17:22:07, July 2, 2002). e. Chase flight by two males (A & B) to the tied
female model (M) (18:28:29, July 4, 2001). f. Simultaneous approach of a solitary male (A) and circling males (B & C) to the
female wing model (shown with a cross near the center) rotated by a motor (17:58:03, July 7, 2001).

Zigzag flights observed at the study site are thought
to involve the chasing of a female by a male (see Dis-
cussion). Prior to the zigzag flight shown in hg. bb,
the presumed female (F) descended from the alder
canopy into an open space, where she flew in a swing-
ing flight (fig. 6c).

Occasionally, a hairstreak was observed to ap-
proach a circling flight involving two males, which
resulted in the chasing of the third individual by the
two males that had been circling. In cases such as this,
the approaching individual may have been a male
(as seen in figs. 5e and 5f), but in several cases it was
thought to be a female, because the resulting chase
by males was vei'y persistent.

The model experiments sought to test the idea
that chase flights are frequently composed of a male

(chasing) and female (leading) hairstreak, and the
reactions of males upon identifying a potential female
mate. Multiple males were attracted to the “female
model” tied with thread, and closely followed her
flight when she turned sharply (fig. be). In fig. be,
the chasing male followed the abrupt turn of the
tied female with a delay of about 2 frames (67 msec).
When the “female wing model” rotated by a motor
was revealed at the study site, one or more males were
frequently attracted (fig. bf). It should be noted that
males engaged in circling flights were sometimes at-
tracted by both of the models.

Finally, the capture of butterflies engaged in chase
flights was attempted, using an insect net. Among
each of the three successf ul captures, one female and
one male was involved.

14

/. Res.Lepid.

Table 2. Copulation statistics for Neozephyrus japonicus

(^ase

Date

Time

Temp.

Duration

Site (from the ground)

#1

l()-7-’()0

18:46

24.5“!:

on an alder leaf at 3.8 m

#2

12-7-’()l

18:20

27..6°C

ca. 3:20

on a goldenrod leaf at 1.2 m

#3

15-7-’()l

18:20

28..5°C

.3:18-21

on a goldenrod leaf at 0.9 m

#4

12-7-’()2

18:20

25.5°C

-

on an alder leaf at 4.0 m

#5

l.'i-7-’()2

18:13

28..5"C

3:07*

on a willow leaf at 1 .8 m

#6

17-7-’()3

17:13

21.6”C

3:43-4:27

on an alder leaf in 4-5 m

#7

17-7-'()3

17:48

21.6“C

5:12-6:08

on an alder leaf in 4-5 in

#8

18-7-’()3

17:22

20.9“C

3:40-4:38

on an alder leaf in 4.5 m

*Pairs were caretiilly broiiglit to a room (30.()"C) and continuously obsen'ed.

Courtship and copulation

During the study period from 2000 to 2003, eight
instances of copulation were witnessed (Table 2).
("opulation usually occurred on leaves, between 3 to 5
m above ground level. On July 10, 2000, at 18:46 hrs,
a chase flight involving four butterflies was broken up
when a male and a female split away from the group,
and lauded on an alder leaf 3.8 m above ground level.
Subsequently, tlie male situated himself at a right
angle to the female. About 30 seconds after laud-
ing, the male walked toward the female and aligned
himself next to the female, in a parallel position, so
that they were both facing the same direction. Five to
six seconds later, the male curved the tip of his abdo-
men toward the female’s genitalia, and after another
five to six seconds, succeeded in connecting with the
female, to assume a V-shaped body position. At 18:47,
the male straightened his figure, and the butterflies
faced opposite directions in a normal end-to-end
copulation posture. Then, they remained stationary
for several hours. At 4:25 hrs the following morning,
the pair was found separated. Anatomical examina-
tion of the female revealed one spermatophore in the
bursa copulatrix.

On July 15, 2001, acopiilation event was witnessed
after a pair of hairstreaks landed on a goldenrod
leaf 0.9 m above ground level, permitting close ob-
servations. Upon landing of the female, the male
barely clung to the leaf on which the female normally
perched. After the male climbed up the leaf, it ap-
proached the female from behind at an angle of 60°.
Once the male had oriented himself to be side by side
with the female, in a parallel position, he retreated
slightly, sharply curving the posterior part of his ab-
domen to contact her genitalia. About 30 seconds
elapsed between the time when the pair landed on the

goldenrod leaf and the time they assumed a normal
end-to-end copulation posture.

Pairs of hairstreaks landing together after a chase
flight sometimes did not couple. In three such cases
(17:45 and 17:55 on July 11, and 18:00 on July 12,
2000) copulation attempts ended in failure when the
female flew away, even though in one case the male
attempted copulation. In most of the copulation at-
tempts observed, males did not open or flutter their
wings. However, on July 12, 2002, one male slightly
but rapidly and repeatedly fluttered his wings, while
situated next to the female, prior to successful copula-
tion. As shown in Table 2, copulation times observed
for Neozephyrus varied from about 45 minutes to 3
hours and 20 minutes. To check for multiple matings,
the btirsa copulatrix of 8 females was examined. Five
of these females possessed one spermatophore, one
female had two, and two females had three.

Discussion
Activity time

Different species of zephyrus hairstreaks are most
active at different times during the day. Previously,
Neozephyrus japonicus has been described as being
most active around dusk (Fukuda et al. 1984), but
no detailed observations were presented. Herein,
we have documented that the peak activity period of
wild N. japonicus 'AdnXxs is between 17:00 hrs and 18:00
hrs.These results agree with those of an earlier study
on a captive population of N. japonicus, where adults
remained inactive until 16:00 hrs, and were most ac-
tive after 17:00 hrs (Imafuku et al. 2000).

Peak periods of activity for some other zephyrus
hairstreaks have been determined using quantitative
data. As shown by Koda (1982) for Chrysozephyrus

39; 8-17, 2000 (2006)

15

bnllnntinus (Staudinger, 1887) and by Takeuchi &
Imafuku (2005) ior Favonius taxila (Bremer, 1864),
some species are most active during the morning,
from 7:00 to 11;00 hrs. Hirowatari and Ishii (2001)
showed that Favonius cognatus (Staudinger, 1892) is
most active during midday, from 10:00 to 14:00 hrs.
According to Koda (1982), adults of Chrysozephyriis
smaragdinus (Bremer, 1864) and Favonius orienlalis
(Murray, 1875) are most active from 8:00 or 9:00 to
16:00 hrs, with suppression of activity during periods
of great light intensity. A bimodal pattern with activ-
ity peaks in the morning and evening is known for
Artopoetes pryeri (Murray, 1873) (Koda 1982), Antigius
attilia (Bremer, 1861) (Koda 1982; Hirowatari 8c Ishii
2001) and Favonius saphirinus (Staudinger, 1887)
(Hirowatari & Ishii 2001). As shown by Akiyama et
al. (1969) and Hirowatari & Ishii {2^01), Japonica
saepestriata (Hewitson, 1865) is most active during
the evening hours. Thus, times of peak adult activity
are very different among different species of zephy-
rus hairstreaks. Species-specific activity times might
be explainable in terms of predation pressures and
intraspecific mating behaviors.

The evening crepuscular activity of Neozephyrus
japonicus may be related to predator evasion. Birds are
known to be one of the most important predators of
butterflies, and their activities usually decline at dusk.
Swallows, however, were sometimes witnessed to attack
circling males ot' N. japonicus during our observations,
and some butterfly specimens collected during the
course of this study had “beak marks” (Johki 1985) on
their wings. Thus, it seems possible that the time of
peak activity for adult N. japonicus is sittiated between
the time in the evening when most local bird species
become inactive, and when it becomes too dark for
hairstreaks to see.

Circling flights

In N. japonicus, the diameter of circling flights
averaged about 14 cm (Table 1). Apparently, the
diameter of circles flown by other species of zephyrus
hairstreaks has not been measured. However, Fiijii
(1982) illustrated flights of Chrysozephyrus smaragdinus,
Favonius taxila, Iratsume orsedice (Butler, 1881), and
(Hewitson, 1865) in “small circles.” In contrast, the
diameter of circles flown by another lycaenid, Holo-
chila helenita (Semper, 1879) , extends up to 1-2 meters
(Sibatiini 1998). Among nymphalid butterflies, Davies
( 1978) depicted two individuals of Pararge aegeria (Lin-
naeus, 1758) spiraling in small circles, and Sibatani
(1998) observed a co-rotating flight of 20-50 cm in
diameter for Mycalesis gotama Moore, 1857. Sibatani
(1998) also observed Hyjwlimnas holina (Linnaeus,

1758) flying in a circle with a diameter of 50 cm to
1 meter. Bitzer & Shaw (1979) figured a trace from
the flight of a red admiral, Vanessa atedanta (L., 1758),
flying in an ascending helical path, as large as 9 meters
in diameter.

Circling individuals of N. japonicus occasionally
changed the direction of the spiral, as shown in fig.
5c. Such changes were also observed in circling flights
of M. gotama (Sibatani 1998). However, two interact-
ing individuals of V. atedanta reportedly circled “most
often in a counter clockwise direction as seen from
below” (Bitzer & Shaw 1979).

As for circling or spiral flights, some functions
have been proposed. A contest for occupation of
territories by males of territorial species has been
widely accepted (Baker 1972; Davies 1978; Bitzer &
Shaw 1979; Wickmann & Wiklund 1983; Takeuchi &
Imafuku (2005) . Neozephyrus japonicuswus described as
a territorial species by Fukuda et al. (1984). The two
individuals of N. jafx)7iicus that were marked during
our study and observed for several days persisted in
guarding the same perching sites, day after day.

As noted by Sibatini (1989), circling flights are
also known for non-territorial butterfly species such
as Holochila helenita, Furema blanda (Boisduval, 1836)
and Neptis sapfjho (Pallas, 1771). This suggests that
the behavior is not always related to the defense of a
territory, but that it may serve to aid in recognition of
potential mates (Scott 1974; Suzuki 1976).

A further possible explanation for circling flights
among males of N. jafmnicus could be to attract fe-
males. Such a scenario may parallel displays by two
males toward a female as seen in manakin birds (Snow
1963; Grzimek et al.l968), where ornamented males
show a specific synchronotis dance in front of a fe-
male.This idea is discussed in further detail below.

Chase flights

Chase flights between two individuals sometimes
attracted the attention of other hairstreaks, and small
groups were occasionally involved in chases. Capture
of hairstreaks engaged in chase flights revealed that
such flights included at least one female. Chase flights
were often terminated when two individuals break out
of the chase, and land near each other. This often led
to subsequent courtship or copulation. Thus, chase
flights are thought to represent the pursuit of a female
by males, and may be one of the steps in courtship.

Zigzag flights observed in the j^resent study (fig.
fib) seem to be a type of chase flight, where a female
is being chased by a male. A similar flight pattern
was observed in another zephyrns hairstreak, Chrys-
ozephyrus smaragdimis (Imafuku pers. obs.). Males of

Hi

/. Ri’s.Lipd.

C. smaragdinus have brilliant green wings above, and
are most active during midday, from 10:00 to 16:00
Ill's. Therefore, we could easily determine that zigzag
nights were composed of a male and a female. It
seems possible that females are exerting some selec-
tive choice on males, perhaps as judged through their
maneuverability, during zigzag flights.

Courtship and copulation

Courtship and copulation events between pairs
of N. japonicus were witnessed ihronghout the peak
activity period of adults, around dusk, and usually oc-
curred on vegetation between three and five meters
above ground level. The timing and placement of
these events may protect coupled pairs of N. Japonicus
avoid diurnal predators, and tiiay help them avoid
terrestrial predators, such as centipedes and carabid
beetles, through periods of relative inactivity duritig
the night.

Duration of copulation between pairs of A^. japoni-
ncs ranged from 3 to 6 hours (Table 2). This does not
largely differ from times observed among individuals
of N. japonicus in captivity, between 2.5 and 3.5 hours
(Imafnkn et al. 2000). Scott (1973) .summarized copu-
lation times for various butterfly species. Extremes
included 10 mituites for PapiUo xulhulus, two days
for some Parnassius and Picris species, and among
lycaenids, two to eight hours for Pallophyrus augustin us
(Westwood, 1852) and 1 hour or more for Nojdmannia
iiicis (Esper, 1779). According to Wlklund (2003),
copulation times varied from 10 min in Coenonympha
panipliilus (Linnaeus, 1758) to a week in Goncpteryx
rJtamni (Linnaeus, 1758).

Multiple matitigs were conhrmed among females
of N. japonicus. This phenotnenon is widespread
atnong butterfly species. Scott (1973) examined
females of various butterlly species and found that
about 95 % of those species sometimes engaged in
multiple matings. Suzuki ( 1988) studied the number
of spermatophores in females of 29 butterfly species,
and found that all of them possessed two or more
s])ermatophores, except Lycacna jddaeas (Linnaeus,
1761) with had only one. The maximum number of
spermatophores found in a single female butterfly
was 15 in a female oi' Danaus gilijprus (Cramer, 1775).
Seventeen papilionid species studied by Matsumoto
& Suzuki (1995) included female individuals that had
tnated more than once, atid some females of Papilio
Linnaeus, 1758 and /I bianorCA Amer, 1777 had
tnated 6 times. Thus, the phenomenon of multiple
matings seems to be general among butterflies. The
function of multiple matings for females is thought
to iuclude the production of heterogeneous offspring

(ref. Suzuki 1988), or the accjuisition of nutrients by
females from spermatophores contents (Boggs &
Gilbert 1979).

The function of male coloration deserves further
di.scussion. Since the fittest males in various animal
populations are often those who successfully compete
for mates (Krebs & Davies 1981), male coloration in
sexually dimorphic species may serve to attract or
court females. As early as 1874, Darwin (p. 505) pro-
j)osed this hypothesis when he noted that “when we see
many males pursuing the same female, we can hardly
believe that the pairing is left to blind chance — that
the female exerts no choice, and is not influenced by
the gorgeous colours or other ornaments with which
the male is decorated.” Scott (1973: 101) subsequently
noted that “this places selective pressure on males to
develop sexual characteristics to make the unrecep-
tive females receptive, . . . This may explain why males
have brighter colors...” Rntowski (1985) showed that
males of Colias eurytheme Boisduval, 1852 with strong
LIV reflectance patterns were more readily accepted
as mates by females than were males with weak or no
UV reflectance. However, Silberglied (1984) could
not confirm this for the nymphalid Anartia amathea
(Linnaeus , 1758), in which males preferred colorful
females but females apparently exhibited no choice
between colorful and colorless males. If male colora-
tion serves to attract or court females, it seems that
males should actively display their wings during court-
ship. During courtship betw'een pairs of the sexually
dimoi'phic lycaenid Pseudozizema maha (Kollar, 1844),
the male’s wings are held wdde open, as seen in photos
by Wago et al. (1976) and Wago (1978), or may be
violently fluttered, demonstrating the conspicuous
blue upperside coloration (pers. obs.). However, such
behavior was not observed among courting males of N.
japonicus, although slight fluttering of wings was seen
on one occasion. In onr study, presumed females of
N. ja/)onicus were occasionally observed to approach
a pair of circling males, resulting in the formation
of a chase flight. In such cases, the female may have
been attracted by male coloration, but at the moment,
no evidence to support this idea has been presented.
Eurther studies on the function of male coloration and
of circling flights in N. jafmnicus 'dr e needed.

Acknowledgements

We thank Dr. Andrew Warren ot the Department of Zoolog)’,
Oregon State University wlio elaltorately read and extensively
im|)roved the mantiscript together with valuable comments;
Messers Ktt/ito Matsubara and Katsnyoshi lino, the owners of the
sttidy site, for permitting tts to work there and also for arranging
the path to allow for easier observations; Tasnkn Kitamnra, of the
EthologicalLaboratory of Kyoto University, for assisting in onr

39: 8-17, 2000 (2006)

17

study; and the United States National Instittite ol 1 lealth lor use of
computer software to create images presented herein.

Literature cited

Akiv.vm-\, Y., T. Sekiguchi & N. Kovam\. 1969. Fluttering activity
of a lycaenid butterfly, yfl/;r)/;/c« saepcstriata Hewitson. New
Entomologist 18: 1-17. (In Japanese with English al)stract)
Baker, R. R. 1972. Territorial behaviour of the nymphalid
butterflies, Aglais utikae (L.) and Inachis io (L.). Tlie jotirnal
of Animal Ecology 41:453-469.

Bitzer, R. [. & K. C. Shaw. 1979 Territorial behavior of the red
admiral, Vanessa atalanta (L.) (Lepidoptera: Nymphalidae).
The Journal of Research on the Lepidoptera 18: 36-49.

Bt)GGs, C. L. & L. E. Gilbert. 1979. Male contribution to egg
prodtiction in butterflies: evidence for transfer of nutrients at
mating. Science: 206: 83-84.

Darwin, C. 1874. The descent of man and selection in relation to
sex. 2nd ed. John Murray, London. 1031pp.

Davies, N. B. 1978. Territorial defence in the speckled wood
butterfly (Pararge aegeria): the resident always wings. Animal
Behaviour 26: 138-147.

Fu|ii, H. 1982. Adtdt behavior of Theclini (Lepidoptera,
Lycaenidae). Yadoriga 107/108: 1-37. (In Japane.se)

Fukida, H., E. Hama, T. Kuzlwa, A. Tak/\h.ashi, M. Takahasiii, B.
Tanaka, H. Tanaka, M. Wakab.\a'asi!i & Y. Wat.anabe. 1984. 'Fhe
life histoiy of butterflies in Japan 3. Hoikusha, Osaka. 373pp.
(In Japanese with English stimmary)

Gr/.imek, B., W. Meise, G. Nietrammer & J. Sieinbaciier (Eds.) . 1968.
Grzimek’s animal life encyclopedia. Vol. 9 birds III. Van
Nostrand Reinhoid Company, New York. 648pp.

Hirowatari, T. & M. Isiiii. 2001. Observations on the daily activity
and feeding behavior of adult Zei>hyrus {I.epido|ttera,
Lycaenidae) in Mt. Mikusa. Scientihc Report of the Gradtiate
School of Agrictilture and Biological Sciences, Osaka Prefecture
University 53: 23-29.

Im.afuku, M., T. Ot.ani & T. T,\itEUc;iii. 2000. Copulation of Nemphyrus
japonkus (Lycaenidae) under captive conditions. Transactions
of the Lepidopterological Society of Ja|tan 52: 1-10.

JoiiKt,Y. 1985. Wingdamagesof butterflies and birds’ attacks. Tyo
to Ga 35: 202-207.

Koua, N. 1982. Obseiwation on the male activity of live Ja|)anese
species of zephyrus hairstreaks, with special reference to the
influence of the light intensity (Lepidoptera, Lycaenidae) . Tyo
to Ga 33: 29-39. (In Japanese with English sumtnai^)

Krebs, J. R. & N. B. D.avies. 1981 An introduction to behavioural
ecology. Blackwell Scientific Publications, Oxford. 292pp.
Matslimoto, K. & N. Suzuki. 1995. The nature of mating plugs and
the probability of reinsemination in Jaisanese Pa|tilionidae,
pp. 145-154 In: Scriber, J. M., Y. Tsubaki & R. C. Lederhouse
(eds.). Swallowtail butterflies: their ecology and evolutionary
biology. Scientific Publishers, Gainesville, FL.

RuTOWSKt, R. L. 1985. Evidence for mate choice iti a sitlphttr
butterfly (Colkis eurytbeme) . Zeitschrift fur Tierp.sychologie
70:10.3-114.

Saigli.sa, T. 1988. Comparative biology of the Zc/t/tyrt/.v hairstreaks.

Nature Insects 23(10): 2-12. (In Japanese)

Saigusa, T. & K. ODAGiRt. 2000. Recent advances in stitches of the
“Zephyrus” hairstreaks (Lepidoptera, Lycaenidae). Gekkan-
Mushi 348: 2-9. (In Japanese)

Sc.o'n , J. A. 1973. Mating of butterflies. The Journal of Research
on the Lepidoptera 11: 99-127.

Scorr, J. A. 1974. Mate-locating behavior of butterl lies. The
American Midland Naturalist 91: 103-1 17.

SiitRozu, T. & H. YaM/\moto. 1956. A generic revision and the
Ithylogeny of the tribe Theclini (Lepidoptera; Lycaenidae).

Sieboldia 1: .329-421.

StBATANt, A. 1989. Conspecific recognition in male butterllies.
Co-rotating and catenate llights. Rivista di Biologia - Biology
Forum 82:"l.5-.38.

StBAT.ANi, A. 1998. An instance of co-rotating llight in Mycalesis
golama Moore. Yadoriga 179:20-22. (In Japanese)

Sti.BERta.tED, R. E. 1984. Visital comnutnicalion and sexual selection
among butterllies, pp. 207-223 In: Vane-Wright, R., P. Ackery,
P. DeVries (eds.). The biology ol bitttet Hies. Acadetnic Press,
New York.

Snow, D. W. 1963. The display of the blue-backed manakin,
Chiroxiphia pareola, in Tobago, W. 1. Zoologica 40: 167-176.

Suzuki, Y. 1976. So-called territorial behaviour of the small copper,
Lycaena phlaens dahnio Seitz (L.epidoptera, Lycaenidae).
Kontyu, Tokyo 44: 193-204.

Suzuki, Y. 1988. Evolution of mating strategies in butterllies.
Special Bulletin of Lepido])terological Society ofjapan 6: 301-
321. (In Jajtanese with English abstract)

Takeughi, T & M. Im.afuku. (2005) Territorial behavior of Favnnius
taxild (Lycaenidae): ten itory size and persistency. The Jout tial
of Research on the Lepido])tera 38:59-66

W.AGO, H., K. Unno & Y. SuzuKt. 1976. Studies on the mating
beha\iorof the pale grass blue, Zizi'tria mahaarg'ia (Lepidoptera,
Lycaenidae). I. Recognitioti ol cons|)ecirtc individuals by Hying
males. Applied Entomolog)' and Zoology 1 1: 302-31 1.

Wago, H. 1978. Studies on the mating behavior of the pale grass
blue, Zizeeria mnha argia (Le]tidoptera, Lycaenidae). 111.
Olfactory cues in sexital discrimination by males. Ajtplied
Entomology and Zoology 13: 238-289.

WiCKMAN, P.-O. & C. WtKEUNt). 1983. Territorial defence atid
its seasonal decline in the speckled wood butterfly (Pararge
aegeria). Anitnal Behaviour 3 1 : 120()-I216.

WiKi uNt), C. 2003. Sexual selection atid the evohttion of butterfly
mating systems, ]>p. 67-90 In: Boggs, G. L., W. B. Watt & P. R.
Ehrlich (eds.). Butterflies: ecology and evolitlion taking flight.
The Lhiiversity of Ghicago Press, Ghicago atid London.

journal of Ri’scarch on the I .ej>idol)lera

39: 18-37, 2()0() (‘2()()6)

Feasibility of light-trapping in community research on mothsi
Attraction radius of light, completeness of samples, nightly fhght
times and seasonality of Southeast-Asian hawkmoths (Lepidoptera:
Sphingidae)

Jan Beck* and K. Eduard Linsenmair

Dept, of Animal Ecology' & Tropical Biology, Biocenter, University of Wnr/.bnrg, Am Unbland, 97074 Wnrzbnrg (Germany)
jan.beck@g)nx. net

Ab.stract: Elx|)erimental data and {|nantitative samples of Spbingid assemblages from tropical
Soiitheast-/Vsia were analyzed to investigate metbodologically relevant tojtics of light-trapping. Mark-
Release-Recaptnre experiments revealed differences between lepido|tteran families in the attraction
radius of a light sotirce, but no such differences cotild be fotmd between 18 species within the
familv S|thingidae. Attraction radii (for 50% return rate within 5 mintites) were generally below 30
meters, which confirms results from previotisly published studies. Arrival of Sjthingidae individuals
at a light source was .symmetrically distributed around midnight, and species differed significantly in
median arrival time. No evolutionan' hvpoihesis for such flight time differences (such as avoidance
of interspecific mating or an effect of body size) could be confirmed from oiir data. At appropriately
chosen sample sites (avoiding dense nndergrowth), all-night sampling with a 125 Watt mercnrv-
vapor lam|) yielded more than % of the expected species richness of S|3hingidae in an average of 5-6
sample nights. Seasonality and tem])oral changes of local assemblages can probably be neglected
for samples f rom largely non-seasonal regions like Borneo if data stem from a relatively short stndv
time of a few years. In conclusion, ( 1 ) there are no indications that light trapping ‘draws’ specimens
from distant habitats to the sampling site, (2) we did not find |)roof that species within the family
Sphingidae are differentially drawn to light, which would lead to biases if hght-tra|3ping data are used
as a measure of relative abundance in the habitat, (3) such biases, on the other hand, probably exist
between taxonomically or morphologically more diverse taxa (e.g. for different families) , and must
be considered for a proper interpretation of results, and (4) hgbt-trap|3ing is an effective means of
asse.ssing species composition and relative abundances of Spbingid assemblages in Southeast-Asia,
btit sam|)hng has to be carried out all night in order to maximize catch size and avoid biases due
to different llight times of species.

Introduction

The use of artificial light sources is a commonly
employed technique to attract night-active Lepidop-
tera for the study of taxonomy, biogeography and bio-
diversity (e.g. Holloway et al. 200 1 , Intachat & Woiwod
1999). While neither the physiological mechanism
(Spencer et al. 1997, Sotthibandhu & Baker 1979,
Hsiao 1973, Bowden 1984) nor the evolutionary
significance (e.g. Holloway 1967) of this well-known
attraction is known to a satisfying degree, it offers a
number of advantages over alternative methods such
as torchlight-transects (Birkinshaw &: Thomas 1999),
baiting with fruits, red wine (Sussenbach & Fiedler
1999, 2000), cheese or shrimp paste (S. Benedick &
J. Hill, pers. com.), malaise traps (e.g. Butler et al.
1999), suction traps, rotary traps or other methods
of passively sampling the air space (see Southwood

* Corresjmnding author

Received: 21 October 2004
Accef)ted: I February 2005

Sc Henderson 2000).

Light trapping yields a large number of specimens
with a minimum of effort (Holloway et al. 2001,
Fiedler &: Schulze 2004). This is particularly true for
automatic light-traps (see Southwood & Henderson
2000 for an overview of designs), which do not even
require the [jresence of the researcher during trap-
ping. However, some groups of Lepicloptera such as
Sphingidae, which are the main subject of this study,
have a tendency not to enter such traps in large num-
bers, but settle on the outside and in the perimeter
of the light source. A comparison between data from
automatic light-traps (Nasir Abd. Majid, pers. com.)
and hand-sampling at light (own data) showed a ca.
30-fold higher yield for the latter method in lowland
Borneo (see also Axmacher & Fiedler 2004, Brehm
2002). Thus, for the rest of this article, the term
‘light trapping’ refers to attracting moths with light,
but sampling them by hand or net. Light can be as-
sumed to sample the community more ‘neutrally’
than trapsbaited with food or pheromones, where
specializations are more likely to occur. Last but not

39: 18-37, 2000 (2006)

19

least, light trapping allows sampling and killing most
specimens relatively undamaged, an important re-
quirement for precise species identification, which is
not met by many ‘passive’ sampling techniques such
as malaise or rotary traps.

On the other side, a number of objections to light
trapping have been put forward, and Lepidopterists
regularly face skeptic criticism by journal reviewers
when publishing light-trapping research. The unre-
solved question of why moths actually come to light
(see above for references) might further erode the
scientific credibility of the method. The main argu-
ments against light trapping in community ecology
(see also Schulze & Fiedler 2003, Brehm 2002 for
a more detailed discussion), as well as common re-
sponses to these arguments, are:

1 ) Light traps sample communities selectively rath-
er than randomly. Besides measuring activity rather
than relative abundance (see e.g. Wolda 1992, Simon
8c Linsenmair 2001), not all species are attracted to
light to the same extent (Bowden 1982, Butler et al.
1999). Some species of hawkmoth are rarely or not
at all attracted to light, but can be caught in numbers
by other methods (e.g. Butler et al. 1999, Kitching &
Cadiou 2000) . Some species seem to be attracted only
in parts of their range (e.g. Daphnis nerii comes to light
in Africa, but not in Asia; I.J. Kitching, pers. com.).
Females are generally rarer in light catches of Lepi-
doptera than males (see e.g. Brehm 2002 for data),
which might reflect differences in activity as well as in
attraction to light between the sexes. Janzen (1984)
described arrival patterns of neotropical Sphingidae
and Saturniidae at light and hypothesized on behav-
ioral mechanisms leading to differences between taxa,
sexes and age groups. Thus, it can be suspected that
relative abundances of species at light are distorted
by differential attraction to light and different levels
of flight activity. While these potential problems are
undeniable, their actual effect on the results of stud-
ies on the community ecology of moths remains to
be quantified. Light trapping has often been proven
to produce readily interpretable and ecologically
meaningful results in studies on the biodiversity of
Lepidoptera (e.g. Holloway 1976, Schulze & Fielder
2003, Fiedler & Schulze 2004). Furthermore, the
constraints of light trapping are shared with any other
comparable sampling method (Schulze & Fiedler
2003, Southwood & Henderson 2000).

2) The effective attraction radius of light sources
might be so large that moths are drawn from other
habitats to a sampling site. Measurements of the attrac-
tion radii of light revealed distances of 3-250 meters
(depending on study method and species; Muirhead-
Thompson 1991, Bowden 1982), but attraction radii

are probably smaller than 30 meters in most situations
for ‘normal’ light sources in entomological research
(Butler & Kondo 1991, Muirhead-Thompson 1991).
Furthermore, studies on the stratification of moth
communities in the forest produced clear community
differences between strata at height differences of
20-30 meters (e.g. Beck et al. 2002, Schulze & Fiedler
2003, Beck & Schulze 2003, using 15 Watt blacklight
sources). Thus, while this is a common point of criti-
cism, there is actually little indication that ‘drawing’
specimens to light from distant habitats really under-
mines the interpretability of samples.

3) The abundance of specimens at light is influ-
enced by weather, lunar light and vegetation. While
the effect of vegetation density around a sample site
might be overestimated (Schulze 8c Fiedler 2003),
effects of moonlight and weather (temperature,
rain, fog, wind) have been clearly documented and
discussed (e.g. Persson 1976, Muirhead-Thompson
1991, Holloway et al. 2001, Intachat et al. 2001, Yela
& Holyoak 1997, McGeachie 1989, Brehm 2002).
Generally, warm, moist and moonless nights produce
highest specimen counts (as has been already noted
by R. A. Wallace in 1869, pp. 95-97). Thus, raw abun-
dances at light can never be used as reliable indicators
of absolute population sizes in a region, but must be
adequately ‘converted’ into figures which are compa-
rable across samples with regard to the initial question
of a study (Southwood & Henderson 2000).

Different light sources (power, wavelength) might
also influence sample size and species composition
(though this is mainly anecdotal; see Brehm 2002),
as does the time of the night during which a trap is
operated (e.g. Schulze 2000 found a steady decline in
Pyraloidea specimens during the first 3 hours of the
night in samples from tropical Borneo).

In an attempt to add new quantitative data to this
discussion, two sets of questions were experimentally
investigated in Sphingidae and other lepidopteran
families in Southeast-Asia:

1 ) What distances of light attraction can we obseiwe
under ‘real life’ research conditions in a tropical habi-
tat? This is particularly interesting for the Sphingidae
- very large and extremely fast- and far-flying moths,
which might thus be expected to exceed known fig-
ures of light attraction radii.

2) Is there evidence for differences in the attrac-
tion radius a) between Lepidoptera families with
largely differing size, body shape and flight ability, and
b) between species of the family Sphingidae? Such
differences would be a clear indication that relative
abundances of moths at light might present a distorted
picture of real abundances, even if species which do
not come to light at all (e.g. largely diurnal taxa such

20

/. Res.Lepid.

Table 1: List of sampling sites in north-eastern Borneo where release experiments were conducted. Note that latitude and
longitude are given in metric format. Additional sampling sites which were re-sampled and used for assessments of seasonality
are both on canopy platforms In primary dipterocarp forests at Danum Valley (DV1 - N4.96°, El 17.80°, Elev.: 220m a.s.l.), and
at Poring Hot Springs in Kinabalu Park (PORI - N6.04°, El 16.70, Elev. 570m a.s.l.).

Site

Nights

Released

Latitude

moths

Longitude

Elevation

I labitat
a.s.l. [m]

Region

CROl

t)

2S5

.5.44°

116.08°

1170

Ridge with road & planted
banana liclds, overlooking PF
valley & mountains

Crocker Range

Park

DV.S

5

.‘54

4.9()°

1 1 7.86°

220

selectively logged (1988),
opening tilong road

Danum Valley

DV4

6

1.S4

4.97°

117.84°

340

selectiveh logged (1988/89),
along road overlooking valley

Danum Valley

TORS

14

1.127

(5.03°

116.77°

.350

Local agricnltitral area near

village

Poring Hot

Springs

as Macroglossiim) were not considered.

Fnrtherinore, sampling data were analyzed to
assess an.swers to the following general problems of
sampling tropical insect po|)nlations:

3) What i^ercentage of a local tropical community
of hawkmoths can be sampled in a short-term light
trapping program? WTiile this (piestion is not par-
ticular to light trapping but to any time-constrained
ecological study, it is an important Itackgronnd figure
to interpret light trapping results.

4) How does the nightly trajiping time influence
specimen numbers and species composition of sam-
ples? It has been suggested (e.g. Kitching & Cadion
2()()(), Diehl 1982) that certain species can only be
successfully caught at certain times of the night, but
so far no qnantitati\’e data for whole local assemblages
were available.

5) How does .seasonality or other temporal change
in the Sphingid a.ssemblages of wet-tropical Borneo
affect the credibility of results from short-time sam-
pling of local a.ssemblages? While it is often inferred
from the climatological stability of tropical habitats
that seasonal changes in connmmities are minute
compared to temperate regions, significant effects of
weather and season (mostly defined by rainfall) on
insect populations have been shown (e.g. Schnlze &
Fiedler 2003, Sh.ssenbach 2003, Intachat et al. 2001,
Novotny & Basset 1998, Wolda 1978, 1988, Wolda
& flowers 1985, Tanaka & Tanaka 1982, Kato et al.
1995, Smythe 1985). Snch effects are often ignored in
ecological studies in the tropics, as time and logistic
constraints do not nsnally allow for year-round, long-
term sampling.

Methods

a) Field methods

Attraction radius experiments

During two periods of field work in 2001/2002
and 2003, four suitable sampling sites in north-eastern
Borneo (Sabah, Malaysia; see Table 1 for details) were
chosen for release experiments. Site characteristics
that influenced their choice were a high yield of
Sphingid specimens (known from previous sampling),
overall favorable logistic conditions and the existence
of a straight stretch of logging road of at least 120
meters length. A generator-powered merctiry-vapor
bulb (125 M/itt) was placed inside a white, cylindri-
cal gauze-bower’ with a height of approximately 1.7
meters from the ground. Sampling was carried out in
the period of reduced moonlight from a week before
new moon until a week after new moon to maximize
catch size. Nightly sampling was carried out from ca. ka
hour after sunset until V2 hour before sunrise, except
if logistic problems made this routine impossible.

All arriving Sphingidae were hand-sampled from
the light or nearby vegetation (<ca. 3 meter radius),
measured (forewing-length), identified (Holloway
1987, D’Abrera 1986, Kitching Sc Cadion 2000), and
marked individually with a waterproof felt-tip pen on
the dorsal forewing. Rare species were either killed
and taken for closer taxonomical examination or
stored inside the gauze cylinder for release at dawn,
whereas common species (e”14 specimens, see Table
2) were used for this study and released from random-

39: 18-37, 2000 (2006)

21

Table 2: List of all 24 species or PUs (1580 individuals) that were included in the release experiments. 18 species belong to
the family Sphingidae, five species to the Geometridae and one to the Noctuidae. Some of the non-Sphingid PUs could not be
reliably determined under field conditions and might refer to any of the species listed in the right column.

Family

Species

Individuals

Comments

Spliingidae

Achemntia ladiesis Fabriciu.s

42

Acosmer'jx anceus Stoll

16

Acosmeryx shmiillii Boi.sduval

70

Ainbulyx canescens Walker

37

Ambulyx moorei Moore

38

Ambulyx pryeri Distant

49

Ambulyx subslrigilis Westwood

20

Amplypterus panopus Cramer

19

Cechenma /tcfo/;.s' Walker

40

Daphnis hypolhous Cramer

211

Marumha juvencus Rothschild & Jordan

14

Megarormn obliquu Walker

20

Fsilogramma mencphron Cramer

147

Therctra clolho Drury

112

Therelra latreillii'W.S. Macleay

49

Theretra nessas Drury

184

Theretra rhesus Boisduval

260

Therelra silhelerisis Walker

18

Geometridae (Eimom.)

Biston 3sp. Leach

31

B. inouei 1 lolktway, puslulata

Warren or /w.s!t/«n',v Warren

Celerena signata Warren

22

Dalima subjlavala Felder & Rogenhofer

36

Fingasa chlora Stoll

44

Geometridae (Geomii.)

7'/!«to.s.vo<7ra-complcx (24sp.)

34

Genera Thatassodes, Felugodes or
Orolhalassodes

Noctuidae (Agaiiaiiiae)

Asola 4sp. Walker

67

/I. /;/««« Walker, albipirmis Swuhoe,
helicouin Linneaus or cgctrs Walker

Z Sphingidae

1346

Z Geometridae

167

Z Noctuidae

67

lychosen distances along a logging road (distances
in 5 meter steps, from 5 up to 120 meters, at one site
up to 130 meters). After preliminary trials moth were
transported to the release distance inside a plastic jar
(500 ml) immediately after marking and released by
turning the jar upside-down without giving the moths
an initial flight direction. Wliile this procedure carries
the risk of non-directional, panicked flight rather than
providing an ideal situation for deliberately choosing
a flight direction, it avoided the effect that moths kept
for a prolonged period after catching ‘cooled down’
after handling and often refused to fly at all when
released, sitting in the same spot for hours, (latch
time, release time and distance, and recapture time
at the light were noted for all individuals. Each moth
was released only once: after recapture it was stored

inside the gauze cylinder for release at dawn. Recap-
tures of marked specimens on following nights were
not considered at all, but their occurrence at a rate
of ca. 5 percent indicates that marking does not harm
the moths (see also Beck & Schulze 2000, Beck et al.
1999). One species, Daphnusa ocellaris, was excluded
from analysis as it was frecpiently observed trying to
escape by crawling rather than by flight after handling
(see discussion).

For a comparison between lepidopteran families,
several parataxonomic units {PUs, Krell 2004) from non-
Sphingid groups were chosen for their commonness
and easy identification under field conditions (Hol-
loway 1986, 1993, 1996). Three of these PUs contained
several species in a genus, impossible to separate alive
and in the field (see Table 2). These considerably

22

/. Rfs.Lepid.

smaller and more delicate moths had to be caught,
marked and handled with special care. Specimens
which were accidentally injured were excluded from
experiments.

Completeness of samples, flight time and
‘seasonality’ comparisons

Data for an assessment of the faunal completeness
of short-time, high-intensity light trapping stem from
an extensive sampling program in Southeast-Asia that
was carried out from early 2001 to early 2004. Except
for the four sites at which release experiments were
conducted (see above) all sampling schedules were
carried out inde])endently of weather or moonlight
conditions, so effects of these factors on abundance
or flight time of moths should be randomly distrib-
uted. Generally, Sphiugidae were hand-sampled (as
described above) all night long for three to nine con-
secutive nights in a block. Median nightly .sampling
time was 10.2 hours. Sites with samples of le.ss than 20
specimens within the first three nights were ignored.
For the purpose of assessing the completeness of
the applied sampling procedure we used data for 15
sites in north-eastern Borneo and one in Peninsular
Malaysia. Sampling habitats ranged from primary
forests through variously disturbed forest types to
open, agricultural landscapes, from lowlands up to
ahnost 1500 meters elevation. Sites were generally
situated either in open areas or in the forest canopy
(platforms or on cliffs or steep slopes) in order to
maxitnize Sphingid catch (see Schulze & Fielder
1997). Four sites iu Sabah (north-eastern Borneo)
were re-sampled up to four times during the 3 year-
study (see Tables 1 & 5, minimum 6 month between
re-samples). These re-samples were used to a.ssess
effects of temporal change, but were considered as
independent .samples for the purposes of an evalua-
tion of sample completeness, which raises the sample
size to 23 sampling sessions.

At 1 1 sites in Borneo and one in Peninsular Ma-
laysia, detailed arrival titiies of all specimens were
measured (iu 15 minute-steps: data from all sites were
pooled for this analysis).

b) Methods of analysis

Return times of the experimentally released
specimens ranged from a few seconds to more than
eleven hours; about 47 percent of the released moths
were not seen again at all during the night of release.
Some specimens obviously did not directly return to
the light, but flew arotmd in the area and were later
attracted to tlie light source again. Thus, we applied

the rule that only returns within five minutes from
release were cotmted as 'returns’ for analysis, while
any later arrivals were considered as 'non-returns’ .
From speed measurements (100 meters in less than
20 seconds for several Sphingidae species) and direct
olxservations of flight behavior we concluded that all
species should be able to reach the light in that time
interval even if they take some time to start or orient
after release. Return rates per minute dropped rapidly
within the first few minutes and reached a bottom level
after about eight minutes (when 50% of all return-
ing moths have come back to the light). Return rates
after this point fluctuated apparently randomly (on
a level of 0-2% return rate per minute) and probably
rejjresent released specimens which did not return
directly to the light, but flew around in the area and
entered the attractive radius of the light again at some
later time, as described above. Preliminai7 analyses
suggested that analysis with a five-minute return cri-
terion yields a higher statistical power than longer
return times (i.e., 8 min., 15 min.). After that time
directional movements towards the light can probably
not be expected any more.

Besides standard statistical procedures, the follow-
ing methods of analysis were employed:

Loglinear Model

A loglinear model was used to test for influential
factors on the frequency of returns vs. non-returns.
Release distances were grouped into six 20 meter
classes (5-2()m, 25-40m, 45-60m, 65-80m, 85-lOOm,
105-12()m), release distances >120m were not used for
this analysis, as they were not available from all sites.
A multi-dimensional contingency table, containing
the frequencies ret urns tls, well as those of suspected
influential factors (such as release distance class,
species identity), was constructed with all possible
interactions between these factors, and then tested
against the actual data (for a detailed description of
loglinear models see StatSoft 2003).

Logistic Regression

As a second mode of analyzing the release ex-
periment data we used logistic regressions (Trexler
& Travis 1993). While cariying the disadvantage that
not all data sets can be fitted well by logistic regression
(predicting return/non-returnheiiei' than random, see
below), they allow assessing attraction radii (as the
distance of 50 percent return-probability) in meters,
rather than just comparing them on a class level.

The k:)gi.stic equation (see e.g. Trexler & Travis
1993) was fitted to the return /non-retum (1/0) data

39: 18-37, 2000 (2006)

23

(original data in 5 meter intervals) . Regression valnes-
can be interpreted as probability for return (StatSoft
2003), the point of 50% return probability (x^ =
turning point oi the logistic regression for species i) is
used as a measure of attraction radius. The variance
of is calculated from the variance of the regression
parameters bp as

var{i,)=(i,)- •

For graphic display, 95% confidence intei-vals were
assessed as 1.96*(SD (x)). For a more rigorous test
of the hypothesis of a difference between two turn-
ing points, a z-test (StatSoft 2003) was used. Both the
loglinear models and the logistic regressions were
calculated with the computer program Statistica 6. 1
(StatSoft 2003).

Estimating total species richness

From the distribution of species in discrete sam-
ples an estimate of the total species richness at a site
can be assessed by several methods (see Chazdon et
al. 1998, Colwell & Coddington 1994, Colwell 2000,
Melo et al. 2003) . Of these, the non-parametric Chaol-
estimator was used as it has proven robust in pilot
studies (Chazdon et al. 1998, Peterson & Slade 1998)
and yielded realistic figures in studies on temperate
moths (Beck & Schulze 2003, Siissenbach & Fiedler
1999), where the total species richness is much better
known than in tropical regions. Assessments of the
species diversity of habitats by Chao l-esthm\tes are
often congruent to those with well established meth-
ods like Fisher’s a or rarefaction curves (e.g. Beck et
al. 2002, Schulze 2000). However, Brose & Martinez
(2004) have concluded from simulation studies that
in assemblages of species with variable mobility other
estimators might perform better. In order to account
for this finding, we additionally followed the suggested
procedure of finding the ‘optimal’ estimator for the
sample coverage at each site (calculated from the
means of ACIi, ICE, MMMmeans, Chaol, Chao2, 1'‘ order
Jackknive and 2'"' order Jackknive estimators; see Brose
& Martinez 2004) . All species richness estimates were
calculated with the computer program Estimates 5.01
(Colwell 2000).

The ‘False Discovery Rate’-control of Beujamini
& Hochberg (1995) was applied to avoid spurious
significances due to multiple tests from the same data
set, and all results which pass the criteria are marked

with an asterisk (*). However, it was not considered
necessary to control analyses of different data sets,
even if they overlap or are nested within another (see
also Moran 2003).

Phylogenetic independence

Correlations of species’ characters might not be
statistically independent because of their common
phylogenetic history (see e.g. Garland et al. 1999 for
a review). The phylogenetic signal in data was tested
with a randomization test (1000 runs), using the pro-
gram Phylogenetic Independence 2.0 (Reeve & Abouheif
2003, see also Abouheif 1999, Freckleton et al. 2002) .
Hawkmoth phylogeny was based on an updated ver-
sion of the systematics in Kitching & Cadiou (2000, 1.J.
Kitching, pers. com.), allowing for unresolved nodes
where applicable. To control for non-independent
data, ‘independent contrasts’ (Felsenstein 1985) were
calculated using the computer program Phyiip 3.61
(Felsenstein 2004; all branch length set to 1 except
unresolved nodes, which were set to 0.0001).

Ree^

la) Release experiments: Loglinear models

A sample size of 1527 released moths was available
for loglinear model analysis. The data structure was
tiot suitable to include all interesting variables (return
frequency, distance, species identity, family affiliation,
experimental site) into one model. Therefore, certain
variables were tested in separate models.

The first model analysis (see box 1 ) suggests that
the research site had no influence on return frequen-
cies, so data from different sites were pooled for
all further analyses. Family affiliation of specimens
had a clear influence on return frequencies (see 2"'*
model in box 1, figure 1), while for 18 species within
the family Sphingidae no significant effect of species
identity on return frequencies could be found (3'‘‘
model in box 1). All analyses show a significant ef-
fect of release distance on return frecpiencies. This
was expected, since the frequency of returns should
decrease with diminishing light intensity at larger
release distances.

lb) Release experiments: Logistic regression

For the three families, as well as for twelve Sphingid
species, logistic regression models could be construct-
ed, while for six Sphingid species the models did not
pass the -f^-test of a better-than-random prediction of
the data. These species were consequently excluded

24

/. Res.Ij'pid.

Box 1: I .oglinear models

Model 1: “Effects of'i csearch site" [site (4) x reliini (2) x release distance (6)] (N=1527)

Model 1 ^ (max, likel.) cff g_

Start model with 3 double-interactions 15.807 15 0.395

Best model (retnrn-dist., site-dist. interactions) 18.715 18 0.410

Already the optimized model does not contain interactions of research site Sc return frequency.

Conclusion 1: No effect of site on rettirn frequencies.

Model 2: “Effects of fainilv" [family (3) x return (2) x release distance (0)] (N=1527)

Model 2 ^ (max, likel.) df £_

Best model: 3fold- interaction fain. X return X dist 0 0 1

Ex|5. model 1 ; no 3-fold, but all 3 2-fold interactions 20.399 10 0.026*

Exp, model 2: no .S-fold and no fam.-retiirn interaction 54.360 12 <0.0001*

riie exclusion of the 3-fold interaction (“family influences distance-return interaction”) leads to a significant difference
between model predictions and real f requencies in efata. Further excltision of the family-return interaction brings another
significant loss of predictive |)ower of the model ( ‘’^,,2=34, p<0.0001*).

Conclii.sion 2: Families have different return rates from different distances, i.e. different attraction radii. Beyond that, families
differ in overall return rates (exp. model I vs. 2).

Model 3: “Effects of species" |s|)ecies (18) x return (2) x relea.se distance (6)]

(only Sphingidae, N=1352)

Model 3 (max, likel.) eff £_

Start model with 3 dotible-intei actions 68.625 85 0.903

Best model (onlv retnrn-dist interaction) 169.99 187 0.809

Exp, model without any interactions 298.73 192 <0.0001*

Conclusion 3: No effect of species identity (within the S|)hingidae) on return frequencies is evident: already the o]Jtimized
model does not include species. The exp. model onlv jrroves the es.sentially expected effect of release distance on return rates -
othenvise the experiments would have been senseless.

from analysis. Figure 2 shows an example of a logistic
regression for one species, figure 3 plots the attraction
radii of the light (measured as the ‘turning points’
of the regression) and their estimated confidence
intervals for the hawkmoth species.

On family level, 50% return rates vary between ca.
10-13 meters for Sphingidae and Noctuidae, whereas
negative values for Geometridae were calculated due
to veiy low overall return rates for Pingasn (5 returns
of 44 releases) and particularly the small Geometrinae
of the 77/rt//a.s,vocfev-group (5 of 34). Possibly handling
effects have affected results in this veiy delicate group,
although no obvious inability of flight was observed.
The other Ennominae species showed attracticin radii
comparable to that of Sphingidae or Noctuidae (data
not shown).

Turning points for Sphingid species vaiy between

26 meters and negative values (for species with very
low return rates). Confidence intervals are large and
indicate a high, unexplained variability in return
behavior. For two species no variance of parameters
could be calculated due to the structure of the data
matrix. Maximum confidence estimates range up to
60 meters, which is still a value in reasonable bounda-
ries of the literature for attraction radii of light (see
e.g. Muirhead-Thompson 1991). No significant dif-
ferences between species could be found for the 10
species for which testing was possible. Turning points
of species do not correlate with the average body size
of the species (N=12, r-=0,()31, p=0. 588), whereas Fie-
dler et al. (unpublished) have found effects of body
size on return rates in other, temperate Lepidoptera
families.

Thus, the results obtained by logistic regression

39: 18-37, 2()()0 (2006)

25

confirm the analyses with the loglinear models.

2) Completeness of samples

During an average of 5-6 nights per sampling ses-
sion, an average of more than % of the C/«ao /-expected
‘true’ species richness could be collected (see Table
3 for details). Sampling success is weakly related to
the number of sample nights as well as the number
of sampled specimens, but not to (observed species
richness or diversity (as Fisher’s a; see figure 4 for
test details).

Measurements of sampling success based on six
other selected estimators of species richness (see
methods) yield mostly quite similar figures, although
Kendall’s concordance coefficient for the seven esti-
mators is surprisingly low at 0.173 and a Friedman-
ANOVA indicates significant differences between
estimator ranks (N=23, |,=23.85, p<0.()()l*).

An application of the method suggested by Brose
& Martinez (2004) lead to the tise of ICE, V and 2'"'
order fackknive estimators, depending on the sample
coverage for each site. However, overall results are
very similar to C/taoi-based estimates and indicate an
average sample coverage of 77.3 percent.

Comparison with other samples from the region

In Table 4 the species richness (observed and ex-
pected) of the 23 standardized samples is compared
to a combination of data from our own samples,
published literature (Chey 1994, 2002, Holloway
1976, Tennent 1991, Zaidi & Chong 1995, Schulze
2000) and unpublished collections (Azmi Mahyudin
and J.D. Holloway, pers. com.). These data vary in
the use of different light types, sampling regimes
and specimen numbers (local samples <20 speci-
mens were not considered). Observed local species
richness ranges up to 50 species (see figure 4: only
samples with >1000 specimens contained over 40 spe-
cies). Thus, the highest C/moi-estimate of 68 species
(Table 4) is still in a realistic range. The slope of the
data in figure 6 suggests that tliis may be close to the
maximum local species richness that can be found
by light-trapping in this region. Schulze et al. (2000)
reported 59 night-active Sphingidae species from a
compilation of data from several sampling sites within
Kinabalu Park, Sabah. Higher records of local spe-
cies richness in Southeast-Asia can probably only be
found from continental regions (e.g. 67 species from
year-round sampling on a site in northern Vietnam,
T. Larsen, pers. com.), where regional species rich-
ness is higher than in insular Southeast-Asia (Beck &
Kitching 2004).

70

20 40 60 80 100 120

Distance class [m]

Figure 1 : Return rates within 5 minutes (in percent) of
three Lepidopteran families from six release distance
classes (note that ‘Noctuidae’ contains only the genus
Asota, see Table 2). Loglinear contingency table analysis
(see box 1 ) indicates significant differences in return rates
between the families.

0 20 40 60 80 100 120

Release distance [m]

Figure 2; Exemplar logistic regression fit of the binary
data {return/ non return) for the Sphingid species Ambulyx
canescens (N=37). Regression values (y-axis) range
between 0 and 1 and can be interpreted as probability
for ‘return’. The ‘attraction radius’ (turning point of the
regression curve = 50% return probability) is estimated
at 23.9m for this species. Note that several data points
may lie on the same position, which are not shown in the
graph but influence the slope of the curve.

3) Flight time during the night

Figure 7 shows the distribution of arrival times of
1450 hawkmoths at 6 sample sites. Arrivals are clearly

26

/. Res.Lepid.

Box 2: Z-tests for flilTerciiccs of logistic regression turning points (a) of
families (b) of tlie most extreme species. Values in bold indicate
significanl differences (z>1.9(i, p<0.0,'j).

a)

z-value

Sphin^dae

Noctuidae

Noctuidae

-(1.664

Geometridae

2.173

2.418

6)

z-value

A. shen>i/lii

/'. liilirilUi

T. rhesus

Asota 4sp.

T. latreillii

1 .469

r. rhesus

1 .624

-0.966

Asola 4sp.

(1.908

-l.l 12

-0.307

1). subfluvata

1.790

-0.29,6

1 .094

1 .436

Table 3. Mean values of 23 sampling sessions for the
number of individuals (N), species (S), sampling nights,
as well as the C/rao?-estimate of true species richness
(Colwell 2000), the proportion (in percent) of observed/
estimated species richness (% and the mean

number of individuals per sampling night. The median
is also given where distributions deviate from normality
(KS-test, p<0.05).

Mean ±SE

Median

Min.

Max.

N

220.9 ±-tX.5

1 16

25

847

^obs

22.1 ±1.5

10

38

Nights

5.5 ±0.3

3

9

Chaol

30.1 ±2.5

28

14

68

%

75.9 ±3.2

37.9

94.4

N / night

35.9 ±6.5

26.2

6.2

121

Table 4. Species richness (observed and estimated) of
local samples from Borneo and Peninsular Malaysia. See
text for data sources of “all data”.

Samples

Mean

SD

Min Max

S„bs

23

22.1

7.4

10 38

Chaol-est (own)

23

30.1

12.1

14 68

All data S ,

obs

60

20.2

10.4

5 50

symmetrically distributed around midnight, with a
steep rise in specimens in the third hour since sunset,
and a decline after eleven hours. This is in marked
contrast to smaller moths in Borneo or in temperate

regions (e.g. Thomas 1996, Schulze 2000, and own
observations), which considerably decline in numbers
after 2-3 hours past sunset.

Flight times clearly differ between species (figure
8). Median flight times also differ between Sphingid
subfamilies, with Smerinthinae flying on average
earlier in the night and Sphinginae later (KW-Anova:
H^if^^=20.27, p<0.0001*). However, median values for
subfamilies are still quite tightly clustered around
midnight (Smerinthinae 6h, Sphinginae 7.5h past
sunset). Median arrival times also differ between
sites (KW-Anova: 130.8, p<0.0001*), which is

most probably the effect of different moonlight and
weatlier conditions. Over the range of sample sites,
however, these differences are leveled out as figure 7
shows a veiy symmetric distribution.

Assuming flight times are adaptive, two hypo-
thetical factors can be tested with for an influence on
flight time differentiation: a) Avoidance of mating in
closely related taxa might be a reason for differences
in activity patterns. Species within genera should have
le.ss (light time overlap than average species couples,
b) With decreasing temperatures during the night,
larger species can maintain their flight muscle tem-
perature, hence their agility, more easily than small-
bodied species.

Pianka’s niche overlap of flight times was calcu-
lated for the 20 most commonly recorded Sphingi-
dae species _,p>12] as well as for intra-generic
comparisons [N^^.^^^^e”K)] within the genera Amhulyx
[5 spp.], llieretra [4 spp.] and Acosmeryx [2 spp.] (see
Southwood Sc Henderson 2000, computed with Pro-
grams for Ecological Methodology, Kenney 8c Krebs
2000). Mean values of niche overlap range around
0.57 for all Sphingidae and >0.7 for the intra-generic
comparisons. Thus, no indication for a lower overlap
for within-genera comparisons was found. There is a
tendency for smaller species to fly earlier than large
species (N=49, Pearson’s r‘-’=().109, p==0.021; body size
was measured as mean forewing length, which is a
good surrogate for body mass within a group of similar
body architecture; Loderetal. 1998, Schoener 1980).
However, a clear phylogenetic signal was detected
in body size data (randomization test: p=0.001* for
phylogenetic independence), whereas no signal was
detected in flight time data (p=0.430) . A correlation of
independent contrasts for body size and flight time is
not significant (N=48, rM).0l7, p=0.376), nor is a cor-
relation of contrasts for body size with ‘raw data’ for
flight time (N=48, r-=0.002, p=0. 769), using contrasts
for a neutral, star-like phylogeny (see also Rheindt et
al. 2004 for methods). These analyses indicate that
the weak relation between flight time and body size
(see above) must be considered spurious under the

39: 18-37, 2000 (2006)

27

Ac_sfterv. Ain _pryen Dajiypotfi Psjnenepfi. Thjatreilfii Th__t1fesus

Ain_canes. Ainp _panop. Me_obfiqua Tli_cJobto Thjiessus Th_silhet

Figure 3. Turning points’ of logistic regressions for
Sphingidae species, which indicate the distance of 50
percent return-probability (±1.96SD) after experimental
release. No significant differences can be found between
the species.

above described assumptions.

4) ‘Seasonality’ and temporal change

Comparisons of the ‘within-habitat’ diversity of
re-sampling sessions (figure 9) showed a remarkable
constancy of measures; only at one site (CROl) a
significant change in diversity could be observed,
even though the species inventory in both samples
was identical (Table 6). No inlluence of ‘seasons’ (see
Table 5) on diversity differences could be observed,
indicating that changes in the structure of hawkmoth
assemblages are either random or directional in a
longer time scale than one year (Beck et ah, 2006).
Measures of between-session similarity of sampled
assemblages are relatively high, particularly for NESS-
indices (Grassle & Smith 1976) which are not biased
(towards lower values) by incomplete speciesinvento-
ries as S0rensen-indices are (Southwood & Henderson
2000) . If common species are weighted high for calcu-
lation of NESS(OT=i ), it seems that primary forest sites
(DVl, PORI) are more stable than disturbed sites.
This pattern, however, breaks down if rare species
are weighted higher at m=21. All NESS(rrt=27)-values
are >0.82 for within-site comparisons of sessions (as
e.g. in Novotny et al. 2002). Multidimensional Scal-
ing (figure 10) of NESS(j«=2i)-values illustrates that
fatmal differences between seasons are mostly smaller
than those between sites. The sampled assemblages
were tested for spatial and temporal differences (see
Table 5 for classification) by the randomization test

Analysis of Similarity (ANOSIM), using PRIMER 5
(2002). In one-way designs, significant effects could
be found between sites (Global R=0.711, p^O.OOl*),
but not between seasons (Global R=-0.043, p=0.552)
or sampling year (Global R=0.026, p=0.397). Further-
more, in various two-way designs (including nested
designs) no temporal effects (season or year) could
be found.

Relative abundances of sampled species were
correlated to each other (Table 7) to assess how well
sampling in one session reflects the rank order of
species in other sessions at a site. All relevant cor-
relations are highly significant, though R’-values are
not particularly high. Within-site correlations have
significantly higher R'^-values than between-site cor-
relations (t-test; t_jj^^.j=7.47, p<().0()01*).

O

(D cn
Q. cn
iS) ^

"CD -£=
D O

CO

o

o

<D

CL ^
LQ ^

"Q) sz

100
90
80
70
60
50
40
30

3 4 5 6 7 8 9

Sample nights

100
90
80
70
60
50
40
30

0 200 400 600 800

J» • • • ♦

f • •

•

♦

Spearman rank correlation
N=23, R^=0.2ri,p=0 027

•

#

S

Spearman rank correlation:
N =23, R^=0,176, p=0.045

Number of specimens (N)

Figure 4. Correlations between the number of sample
nights (upper graph) and the number of sampled
individuals (lower graph) and the percentage of sampled
‘true’ species richness (based on Chao 7-estimates,
Colwell 2000). Each dot refers to a different sampling
session, not to sub-samples from the same sampling
session. No correlations were found with the observed
number of species N=23, R^=0.012, p=0.614)

and the species diversity (Fisher’s a: N=23, R2=0.084,
p=0.179).

28

/. Res.Lepid.

Sample nights

Figure 5: Exemplar species accumulation curves for a
‘good’ sampling site at Danum Valley, Borneo (DV1-I);
After 6 nights of sampling, 21 of 23 estimated species were
caught (91 .3% sampling success). is smoothened
by a 100-fold randomization of the sample night order
(Colwell 2000), Chao1 is the estimate of total species
richness at the respective number of (randomized)
sampling nights. The A4/W/Weans estimator (Colwell 2000),
which fits an asymptotic curve to the randomized species
accumulation curve, yields an estimate of 22 species for
this site. The total number of specimens at this sampling
session was 437.

Figure 6: Observed species richness of Sphingidae as
a log,(j-function of the number of sampled specimens
for 60 local light-trapping samples from Borneo and
Peninsular Malaysia (see text for data sources). The data
are significantly correlated (N=60, Pearson’s r2=0.839,
p<0.0001*).

Discussion

Attraction radius of light sources

In the experiinetital part of this sttidy we showed

for 18 species of Sphingidae, as well as some species
of the Geomelridae and Noctiiidae, that the effective
attraction radius of a 125 Watt MV-lamp is indeed
relatively low with mean attraction distances below
30 meters. This conhrms what most previous studies
found with different methods, species and habitats
(e.g. Muirhead-Thompson 1991, Bowden & Morris
1975, Onsager & Day 1973, Plant 1971, Meineke 1984,
Kovacs 1958). Even for hawkmoths, which probably
form the upper limit of lepidopteran flight strength
and speed, there is no indication that light trapping
draws moths from distances so great that investigations
on local habitats were ‘polluted’ by specimens from
far away. Occasionally specimens were found at light
sources far from their typical habitat (e.g. on ships
far off the nearest coast; I.J. Kitching, pers.com.), but
there is no indication whatsoever that stich specimens
were artificially drawn otit of their natural habitat by
the light. Due to dispersal and migratory behavior
some few individuals of a species will always show up
away from its breeding habitats, but sttch ‘strays’ are
not /icr.vcan artifact of .sampling.

Differences between taxa

A comparison between families with considerably
different body sizes and shapes revealed significant
differences in rettirn behavior and attraction radius.
Fiedler et al. (unptiblished data, pers. com.) found
in similar mark-release experiments in Germany that
body size has a signihcant effect on rettirn rates of
Geometroidea (though not in other taxa). Therefore,
results of light trapping sttidies from taxonomically
and/or morphologically very diverse samples must
be viewed very carefully, as relative abundances at
the light might not reflect relative abundances under
nattiral flight conditions acro.ss all taxa. Within the
relatively uniform taxon Sphingidae, on the other
hand, comparisons of 18 species did not show any
non-random differences in return behavior despite
a considerable sample size (over 1300 released speci-
mens, see Table 2). While it is known that some dif-
ferences between species must exist (e.g. some species
do not come to light at all; Butler et al. 1999 found
stibstantial differences between blacklight and malaise
traps for three North American heiwkmoth species),
these differences might often be diluted beyond sig-
nificance by a high variability in attraction radii (see
below) dtiring natural trapping conditions. It might
be argued that in this study the concentration on
relatively common species (for the obvious reason of
attaining sufficient sample sizes) might have neglected
systematic differences between common and rare
species. Generally, a lack of an ef fect is never proof of

39: 18-37, 2000 (2006)

29

its non-existence, but there is presently no indication
to assume a systematic difference in attraction radius
between rare and common species.

The species-abundance distributions in a large
number of light-trapping samples of hawkmoths from
Southeast-Asia fit the lognormal distribution very well
(Beck, 2005). This or similar mathematical distribu-
tions have been found in samples of a large number
of organisms (e.g. Tokeshi 1993), including light-trap
samples of moths (e.g. Robinson 1998) as well as data
from sampling procedures that are beyond any doubt
free of sampling errors (e.g. counting tree seedlings in
sample squares, Hubbell 2001 ). If the lognormal and
related distributions have any biological significance
(e.g. Hubbell 2001, Hengeveld & Siam 1978) and are
not merely a statistical characteristic inherent to any
heterogeneous data set, then massive biases of light
trapping would have distorted this relationship for
night-active Lepidoptera sampled in this way.

High variability in data: A methodological artifact?

The release experiments revealed a large varia-
tion in return behavior that was also evident from
observations during held work. Wliile some specimens
were seen flying in a straight line towards the light
after release at distances of up to 130 meters, other
specimens did not return at all. The failure to find
species-specihc differences in light attraction might be
attributed to this variability rather than homogenous
measures of attraction (see e.g. hgure 3). Therefore
it is a crucial question for the interpretation of results
to what degree such variation might be caused by han-
dling effects or other problems of the experimental
design. Obvious handling effects could repeatedly be
obseiwed in Daphnusa ocellaris, which walked away after
release rather than trying to fly, and was consequently
excluded from all analyses. However, it was not com-
pletely surprising to hnd odd behavior in this species,
as it is an ‘unusuar hawkmoth in a number of other
behavioral traits: In Borneo, it is the only species that
is frequently encountered in the undergrowth of the
forest, while all other species tend to fly in the open
airspace above the tree canopy (Schulze & Fiedler
1997). Furthermore, its relatively small thorax makes
it considerably weaker in beating its wing than species
of comparable size (pers. obs. during handling). The
observed behavior can probably be interpreted as a
predator escape tactic after an unsuccessful first at-
tack of a bat. In no other Southeast-Asian hawkmoth
species was such behavior observed.

Predation by bats around the light was high, and
occasionally released moths were caught by bats
just before returning to the light (pers. obs.). While

Figure 7: Frequency distribution of arrival times of
Sphingidae specimens in Borneo and in Peninsular
Malaysia as a function of sampling time. Despite a
symmetric distribution around midnight data do not fit a
normal distribution.

this reflects the situation at most trapping sites in
Southeast-Asia, it might mean that ‘physiological’
attraction radii are slightly larger than ‘ecologi-
cal’, bat-influenced measuies of the radii. Weather,
moonlight and temperature varied within as well as
between sampling nights and might be responsible
for a large part of variatioit in measured attraction
radii. All three parameters are known to influence
total catch size (e.g. Yela & Holyoak 1997) and might
be suspected to influence the flight activity of moths
as well as the attraction radii of light sources. Fiedler
et al. (pers. com.) found effects of temperature on at-
traction radii of Geometroidea in temperate Germany,
where temperatures vary much more than in tropical
lowland areas. However, Just like bat predation, such
variability only reflects realistic .sampling conditions
and is therefore not a methodological problem.
Whatever physiological differences there might be
to influence differences in response to light between
species, natural variation in conditions during realistic
sampling procedures seems to reduce them to a non-
directional ‘noise’ that will most probably not produce
any artifact results in ecological studies.

Completeness of samples

Using the C7t«oi-estimator (Colwell 2000) as a
measure of true species richne.ss, we found that short
term (<10 nights), high intensity light trapping (125W
MV-lamp, all night hand-sampling) can yield a surpris-

30

/. Ilt^s.Lepid.

ai

CO

c

7h

CL*

Q

CO

CO

Z5

5 H] ^

%p

I If,

>< 11] in II]
o^: a, in

III 11] c-

I- l'^ I
'a; .. I:*; ill
i ^(-1
t

I

'•w ^ l-J

't-r

LO LH

s'^

■S5

I t

■fl

-t J;
OK

u:i in o m in X in in in m c- ni m R] o ni

■■ ? 'i.E 9-5 5

m

Ri t- 0,1

Si?

0 t
I lo i"-

'■■’iS

di-t'ts'a" m i in S;* o'o S th

^ ^<3 oj’

in I r-' <■■
l__r^ Oji^
■^U] a.

u

m m m Ri in
:d e t
o;? o ^ a,

f 5 o o i-
^ ^ Q- 1 I

- ‘1 ill t
-| in iM

Figure 8: Median, earliest and latest arrival times of 50 Sphingid species in Borneo and Peninsular Malaysia (N=:1450). A KW-
Anova of the 24 species with individual numbers e”10 indicates that different species have clearly and significantly differing
flight times during the night (N=1375, ^23=509. 04, p<0. 00001*).

ingly complete picture of the species assemblage that
is present in a habitat at a particular time (excluding
the day-active taxa, of course). An average of% of the
expected species (and often much more) could be in-
ventoried with this method. S])hingidae are not a very
speciose taxon in Sontheast-Asia (e.g. species ntimbers
from Borneo: 1 13 Sphingidae (Beck & Kitching 2004),
compared to ca. 1000 described Geometridae (Hollo-
way 1993, 1990, 1997) ), but a relatively high degree of
available ‘background information’ even for tropical
species (taxonomy, distribution, host plants: Kitching
& Cadion 2000) renders them a very attractive group
for ecological research within the Lepidoptera.

It is difhcnlt to judge how reliable species richness
estimates really are. CV/r/o /-estimates increased with
increasing within .some sites (see Schulze & Fiedler
2003 for a similar effect on Fisher’s a of Pyralidae),
even though they often reached relatively stable values
with addition of the last 1-2 sample nights (see e.g.
figure 5). Alternative estimators (see results) came to
very similar figures of expected species, and figure 6
indicates that extrapolation estimates are generally in

a realistic range. Particularly, ‘optimal estimators’ ac-
cording to Brose Sc Martinez (2004) correlate well with
Cliaol (Spearman rank correlation: N=23, R-=0.525,
p<0.00001 *) and lead to an overall veiy similar assess-
ment of sampling success. Thus, estimates are consid-
ered credible, although only more empirical studies
on Colwell’s (2000) estimators can reallyjudge their
value as predictors of true species richness.

Moreno & Halffter (2000) used randomized spe-
cies accumulation curves to determine asymptotes of
species inventories for Neotropical bat assemblages,
which were suggested to be used for the comparison
of incomplete samples, across different sampling
methods or sampling efforts. This was criticized by
Willott (2001; but see Moreno & Halffter 2001),
who pointed out that (1) the number of sampled
individuals is a better measurement of sampling ef-
fort than the number of sampling units, and (2) the
method is not suitable for high diversity taxa where
only a small fraction of a local assemblage has been
santpled. The MMMeans method of richness estima-
tion (Colwell 2000, Chazdon et al. 1998) follows a

39: 18-37, 2000 (2006)

31

similar approach by fitting an asymptotic Michaelis-
Menten type curve to randomized species accumu-
lation curves. MMMmns-estimates were generally
very similar to C/woi-estimates in this study. Figure
5 shows an example of species accumulation curves.
Estimates of total species richness from Chaol were
in an acceptable range when viewing randomized
species accumulation curves, which adds credibility
to both approaches.

Flight times

The analysis of flight times shows that it is crucial
to stay out all night in order to sample Southeast-Asian
hawkmoth communities successfully and completely.
Not only does the greatest number of specimens come
around midnight to the light (and moths keep com-
ing until dawn) , but a shorter sampling period would
also systematically under-represent certain species,
as median flight times vary significantly between spe-
cies. The ability of Sphingidae to warm up their flight
muscles by shivering gives them the ability to make
their flight activity relatively independent of ambient
temperatures, at least under the moderate tempera-
ture changes of a tropical night. However, in other
tropical moth taxa empirical data prove that numbers
of moths arriving at night decrease after the first few
hours (Schulze 2000, Brehm 2002, Sfissenbach 2003).
Hence, the judgment stated above might not be valid
for such taxa, and the ‘coverage’ of the assemblage in
samples may be considerably higher even if sampling
is carried out only for parts of the night.

It remains to be seen from similar data on other
taxa, if ‘niches’ in flight time are a general feature
of moth assemblages. Anecdotal information points
into this direction - Southeast-Asian Saturniidae, for
example, are found in numbers only after midnight
(pers. obs., see also Janzen 1984). No ecological
reasons for temporal niche segregation could be con-
hrmed from our data, though early-flying taxa tended
to be smaller than late-flying ones. Studies on other
organisms suggest that partitioning of activity times is
only rarely caused by competition or predation, and
that endogenous rhythmicity may be an evolutionaiy
constraint (e.g. Kionfeld-Schor & Dayan 2003). We
did not find any decline in the number of Sphingidae
specimens (nor for other groups, though this was
not quantihed) under conditions of heavy rain. No
moths might be flying during the peak minutes of a
tropical rainstorm, but they still keep coming under
very unpleasant sampling conditions. Thus, breaks
from a sampling schedule due to heavy rain (which
are commonly reported in the literature) cannot
be excused by low sampling success, although more

Figure 9: Fisher’s a (±95% confidence interval) as a
measure of within-habitat diversity for re-samples of four
sites in north-eastern Borneo (see Table 1). Only at one
site (CR01) a significant difference in diversity between
re-samples within a site (dashed lines) can be observed.
There is no indication for systematic influences of season
(see Table 5) on diversity. Fisher’s a (±95%CI) for pooled
sample sessions are: DV1 : 5.98±0.99, CR01 : 8.15±1 .23,
PORI: 8.65±2.46, POR8: 7.97±1.05. All samples fit the
tofiiser/es-distribution sufficiently good to justify the use of
Fisher’s a (Southwood & Henderson 2000).

CN

C

o

to

c

<1»

E

Q

08

CROIJI (01)

0.6

POR8-II (02)

0 4

DV1-III (03) ^

° P0R8-I (01)

DV1-IV(03) O

0.2

A

0.0

DV1-I (01)

• POR8-III(03)

-0.2

CRori (01)

O

-0 4

PORl-j(02) DV1-II(01)

-O.B

POR1-II (03)

0

-20 -1.5 -10 -05 0,0 05 1,0 1 5 2.0

Dimension 1

o

A

'Dry season'

SW-monsoon

NE-monsoon

Figure 10: Multidimensional Scaling of NESS(m=27)-
values ordinate sampling sessions according to their
faunal similarity. Different symbols indicate season (see
Table 5 for classification), labels give sample acronyms
and the year of sampling (in brackets). Temporal
differences (season, year) do not produce meaningful
patterns, whereas samples are differentiated according
to site affiliation (see text for randomization test).

fragile taxa might by damaged beyond identification
by wet equipment.

32

/. I^s.Lepid.

Table 5. Timing of re-sampling sessions (4-9 nights each) at four sites in Sabah, Borneo. Numbers in brackets indicate ‘seasons’
based on rainfall measurements at Danum Valley Field Centre (see Marsh & Greer 1992), but our assessment of seasons is
based on means of 1 7 instead of 6 years of meteorological data collection (data not shown, G. Reynolds pers. com.): (1 ) = ‘Dry
Seasons’ February-April, July-September (<230 mm rainfall per month), (2) = ‘South-West Monsoon’ May-June, (3) = ‘North-
East Monsoon’ October-January (>250 mm). Climate patterns might differ between parts of Borneo island, see also Walsh &
Newbery (1999), Kato et al. (1995), Kitayama et al. (1999).

Site

Sample I

Sample II

Samjtle III

Sample IV

DVl

Jim 2001 (2)

Dec 2001 (3)

Mar 2003 ( 1 )

Dec 2003 (3)

CROl

May 2001 (2)

Nov 2001 (3)

PORI

Jan 2002 (3)

Keb 2003 ( 1 )

POR8

Jul 2001 (1)

Jan 2002 (3)

Feb 2003 ( 1 )

Table 6. Measures of between-sample diversity between sampling sessions at four sites in Borneo. All indices range from 0
(no species in common between sessions) to 1 (identical samples). While Sorensen-indices consider only presence-absence
data (Southwood & Flenderson 2000), NESS-indices (Grassle & Smith 1976) use quantitative data with increasing weight on
rare species with increasing m (see e.g. Brehm & Fiedler 2004 for an assessment of between-habitat diversity measures). For
comparison between-sample diversity of 10 samples (within 12 months) in seasonal northern Vietnam (Hoang Lien Nature
Reserve, data from T. Larsen, pers.com.) is presented.

Site

Re-.sam|)les

Soien.sen (Mean±SD)

NESS (w=/)(Mean±SD)

NESS (w=27)(Mean±SD)

DVl

4

0.83 ±0.18

0.9,5 ± 0.04

0.85 ± 0.09

CROl

9

1.00

0.80

0.95

PORI

9

0.67

0.86

0.94

POR8

3

0.87 + 0.12

0.78 ± 0.07

0.92 ± 0.07

HOA

10

0.47 ± 0.24

Seasonality and temporal change

We have found no evidence of ‘seasonality’ in
Spltingid assemblages from north-eastern Borneo,
though otir data were too limited for an application
of rigorotis tests of cyclic patterns (e.g. Wolda 1988).
Previous studies suggested that not only in tropical
habitats with clear wet and dry seasons (e.g. Frith &
Frith 1985, Jan/.en 1993), hut also in less seasonal
regions (Novotny & Ba.sset 1998, Intachat et al. 2001,
Wolda 1978) insect abundances can lluctuate con-
siderably as an effect of changes in precipitation.
Our data stiggests that while there are considerable
changes of relative and absolute abtmdances of indi-
vidual species between sampling sessions (see below),
measures ol'diversity (figtire 9) , community structure
(Table 6) and the rank order of species (Table 7) are
not dramatically different. A number of other studies
also concluded that temjioral changes did not affect
meastires of community structure for studies of ants
in Borneo (Briihl 2001) or Lepidoptera in Sulawesi
(Barlow 8c Woiwod 1993), New Guinea (Novotny et

al. 2002) and Borneo (e.g. Fiedler & Schulze 2004,
Schulze & Fiedler 2003) . Thus, in the aBsence of clear
seasonal patterns (see also Walsh & Newbery 1999),
we conclude that short-term samples probably give
reasonably good data for analyses of local assemblages
of Sphingidae in Borneo - even though it has to be
kept in mind that assemblages might change over
timescales of several years (Beck et al., 2006) . For a few
species ( Theretra rhesus, T latreillii, T insularis, Daph-
nis hypothoiis, Mnrumhn juvenciis, Enpinanga borneenis,
Cechenena Ihieosa) we have indications of larger long-
term poptilation fhictuations from the re-samples in
our data as well as in comparison with older literattire
(e.g. Holloway 1976, 1987, Tennent 1991).

Conclusions

From the experimental and empirical data that
we presented above, the following conclusions and
methodological advice can be drawn:

1) For complete and efficient sampling of Sph-
ingidae (at least in Southeast-Asia) , hand-sampling

39: 18-37, 2000 (2006)

33

Table 7: R-values for Spearman rank correlations of relative abundances of species across 1 1 (re-)sampling sessions in Borneo.
All correlations have p-values <0.05* (N=58), values in bold print mark correlations at p<0.0001*. The latter include all within-
site correlations of sessions (in boxes).

DVl-I

DVl-II

DVl-III

Dvi -rv

CROl I

CROl-II

PORl-I

PORl-II

POR8-I POR8-II

DVl-II

0.812

DVl-III

0.673

0.624

DVl-IV

0.650

0.701

0.576

CROl-I

0.555

0.572

0.465

0.515

CROl-II

0.493

0.586

0.379

0.410

0.740

PORl-I

0.603

0.650

0.577

0.525

0.657

0.621

PORl-II

0.598

0.643

0.504

0.425

0.525

0.549

0.676

POR8-I

0.359

0.430

0.272

0.274

0.419

0.388

0.309

0..393

POR8-n

0.425

0.510

0.395

0.409

0.422

0.458

0.538

0.474

0.767

POR8-III

0.373

0.441

0.370

0.400

0.478

0.449

0.453

0.534

0.775 0.842

at a light source during the whole night is necessary.
Shorter nightly sampling sessions miss a considerable
number of specimens, and - more importantly - might
specifically miss species with particular flight times,
which wotild bias data. However, as long as sampling
is kept standardized across habitats to be compared,
and provided that the samples are stifficiently large
and representative (e.g. by sampling during the peak
hours of activity around midnight) between-site com-
parisons might still yield valid results.

2) Under these conditions, a week of sampling will
usually yield over 100 specimens, which often repre-
sent more than % of the expected species in a habitat
(excluding day-active species). Pre-condition to this
is an adequate choice of sampling site, which shotild
sample from some open airspace rather than dense
vegetation (Schulze & Fiedler 1997). Seasonality of
assemblages can probably be neglected for practical
purposes, though care has to be taken when includ-
ing data from older sources or from more seasonal
regions.

3) There is no indication that a significant frac-
tion of specimens are not sampled locally, btit drawn
from some distant natural habitats to the sampling
site by the light. A 125 Watt mercury-vapor bulb has
a 50% attraction radius of less than 30 meters (which
confirms older measures from the literature) , so even
highly active, fast-flying taxa such as Sphingidae can be
very locally sampled. However, individual specimens
might fly far from their normal (breeding) habitat for
natural reasons (e.g., dispersal, migration).

4) No evidence was found that species within the
family Sphingidae differ significantly in their attrac-
tion to light. While this does not generally rule out
that such effects may occur, natural variation of sam-
pling conditions will effectively level out such stibtle

differences. We tentatively concltide that abundances
at light sources do largely reflect relative abundances
(or rather flight activity) in a habitat. However,
wherever there is au option of comparing relative
abundances at light with other measures of relative
species abundance (e.g., counts of caterpillars from
random samples) these should be employed and
critically discussed.

5) There is indication that attraction towards light
sources differs between higher taxonomic units such
as Lepidopteran families. Large differences in body
size or shape might be a key predictor for the dimen-
sion of such differences. Data for taxonomically or
morphologically diverse assemblages should be criti-
cally explored for potential biases resulting from dif-
ferent attraction radii of light. In biodiversity studies
that compare different habitats, for example, it should
be explored whether different taxonomic sub-units
follow similar patterns before they are presented as
one common trend from a pooled data set.

Acknowledgements

We thank Dr. Chey Vun Khen (Fore.st Research (lentre of .Sabah,
FRC), Dr. Jamili Naisand Maklarini bin Lakiin (Sabah Parks), Peter
Chong (DVMC, Innoprise Inc.) , Glen Reynolds and Dr. Roiy Walsh
(Royal Society), Intraca Woods Inc. and Henry Barlow ((tenting Tea
Estate) for productive cooperation and/or for granting |)erinission
to sample on sites tinder their respective control. Ftirthennore, we
thank the Economic Planning Unit (EPH) of Malaysia for granting
a Research Permit for Malaysia, and the British Royal .Society for
accepting this project as part of their Sontheast-Asian Rainforest
Research Project (SEARRP). Dayang .Siti Nortasha (Universiti
Malaysia Sabah) and Katrin Blassmann (I'niversitat Wiirzbtirg)
helped with the mark-release experiments. Dr. Ian |. Kitching
(NHM, London) helped with the identification of specimens
and methodological advice. Ftirthennore, we thank Dr. Konrad
Fiedler, Dr. Christian H. Schulze and Dr. Jereim D. Holloway for
helpful comments and discussions. Two anonymous reviewers gave

34

/. Res.Lef)id.

veiT helpful comments on an earlier draft ttf'the manuscript. The
study was financially sni>|jorted hy the ‘Clraduiertenkolleg’ of the
German Research Council (l)F(i, ‘Grundiagen des Arthropoden-
verhaltens’), the ‘Sy.s-Resource’ program of the European Union
and by a Ph.D.-scholarship of the German Academic Exchange
Service (DAAD).

Literature cited

AisomiEiF, E. 19f)9. A method for testing the assumption of
phylogenetic independence in comparative data. Evolutionary
Ecology Research 1 : 89.5-909.

Axmaciier, |. G. & Firni.KR, K. 2004. Manual versus automatic moth
samjtling at ef|ual light sources - a comparison of catches
from Mt. Kilimanjaro. |ournal of the Le|)ido[)terist Society
.58: 19(>-202.

B.vri.ovv. 11. S. & l.P. VVoiwon 1993. Seasonality and diversity of
Macrolepidoptera in two lowland sites in Dumoga-Bone
National Park, Sulawesi Utara. jtp 107-172. Itv. W.J. Knight (ed.),
Insects and the Rain Forests of Soiith-Fhist Asia (VVallacea).
Royal Entomological Society, London,

B.\s,sft, Y., P.M. H.vmmonu, II. Barrios, ).l). Hoi.i.ow.w & S. Mii.t.FR
1997. Vertical stratification of arthropod assemblages. |)p
17-27. In: N.E. Stoik, ). Adis & R.K. Didham (eds.), Cano|)y
Arthropods. Chapman & Hall, London.

Bfok, J. 200.5. Fhe macroecology of Southeast-Asian hawkmoths
( I.epidoptera: Sphingidae). PhD. -thesis. University of
Wur/hurg (Germany). Available online at http: / /opus.bihliothek.
u n l-w(’izl>ii >g. de/ (>pn.s/xn)lltextt'/2()()5/ 1 3()()/

Bfcr, |. & I.|. Kiiciiino 2004. Fhe Sphingidae of Southeast-Asia: a
comjrrehensive website on the disti ihution of hawkmoths from
Burma to the Solomon Islands (vmion 1.0). Internet website at
http://ximmi.sphingida<’-.'iea.hiozentrum.uni-wuei-zburg.de
.FOR, [., I. |. KuctiiNO & K. E. Linsfnmair 2006. Effects of habitat
disturbance can he subtle yet significant: biodiversity of
hawknu)th-a.s,semhlages (Lepidoptera: Sphingidae) in Southeast
-Asia, Biodiversity and Conservation 1.5, in press.

Bfc.r [. & C.H. .Sciii'i./F 2000. Diversity of ffuit-feeding butterflies
(Nymphalidae) along a gradient of tro|ucal rainforest
succession in Borneo with some remarks on the problem of
“pseudore|>licates”. [ournal of the le|)ido|rterological Society
of Jairan 5 1 : 89-98.

2003. Diversitiit von S|)annern (Lepido|)tera: Cieometridae)

in einem oherf iankischen Laiubwald. Nachrichten des
entomologischen Vereins Apollo, Frankfurt, N.F. 24: 131-140.
Bkcr. J., E. Muiii.FNiiKRo & K. Fifui fr 1999. Mud-puddling behavior
in trojrical butterflies: in search of protein or minerals?
Oecologia 119: 140-148.

Bfcr.J., G.l 1. Set it:i.zF, K.E. Lin,sknm.\ir K- K. FtFDi.FR 2002. From forest
to farmland: diversity of geometrid moths along two habitat
gradients on Borneo, |ournal of Tropical Ecolog)' 18: 33-.51.
Bfnjamini Y. &Y. flociitiFRO 1995. Controlling the false discoveiy rate:
a practical and powerful a|tproach to multiple testing. Journal
of the Royal Statistical Society (B) 57: 289-300.

Birrinsiiaw, N. & C.D. 4'i toM.vs 1 999. Torch-light transect surveys for
moths, journal of Insect Gon.servation 3: 1.5-24.

Bttwden, J. & M.G. Morris 1975. Fhe influence of moonlight on
catches of insects in light-traps in Africa HI: The effective
radius of a mercury-vapor light-trap and the analysis of catches
using effective radius. Bulletin of entomological Research 65;
30,3-348.

Bowden, J. 1982. An analysis of factors affecting catches of insects in
light-traps. Bulletin of entomological Research 72: 535-556.

1984. Latitudinal and seasonal changes of nocturnal

illumination with a hypothesis about their effect on catches
of insects in light-tra|ts. Bulletin of entomological Research

74: 279-298.

Brfiim, G. 2002. Diversity of geometrid moths in a montane
rainforest in Ecuador. Ph.D.-thesis, University of Bayreuth
(Germany). Available at http://www.opns.ub.uni-bayreuth.de/
v()lltexte/2()03/ 20

Brfum, G. & K. Fifdi.fr 2004. Ordinating tropical moth ensembles
from an elevational gradient; a comparison of common
methods, journal of Tropical Ecology 20: 16.5-172.

Brosf LI. & N.D. Martinez 2004. Estimating the richness of species
with variable mobility. Oikos 105: 292-300.

Bruiii., G.A. 2001. l.eaf litter ant communities in tro|)ical lowland
rain forests in Sabah, Malaysia: effects of forest disturbance
and fragmentation. Ph.D.-thesis, Llniversitat Wiirzdnirg
(Germany).

Bi'ti.fr, L. & V. Kondo 1991. Macrolepidopterous moths collected
Iw blacklight tra|r at Gooper's Rock State Forest, West Virginia:
a baseline study. Bulletin 705, Agricultural and Forestry
Fixperiment Station, West Virginia University.

BiTi.FR, L,, V. Kondo, E.M. B.vrrows & E,G. Townsend 1999. Effects of
weather conditions and trap types on sampling for richne.ss and
abundance of f orest Le]ridoptera. Environmental Entttmolog)'
28: 795-811.

Ghazdon, R.L., R.K. Goi.vvFt.t,, J.S. Densi.ow & M.R. Guariuu.vta
1998. Statistical methods for estimating species richness of
woody regeneration in |n imary' and secondary' rain forests of
northeastern Gosta Rica, pp 28.5-309. In: F. Dallmeier & j.A.
Gorniskey (eds.). Forest biodiversity research and modeling
- conceptual backgrounds and Old World case studies,
Smithsonian Institution, Washington DG, USA, published hy
LINESGO, Paris & The Parthenon Publishing Group.

Giify, Y'.K. 1994. Gomparison of biodiversity between plantations
and natural forests in Sabah using moths as indicators. Ph.D.-
thesis, University of Oxford (UK).

2002. Gom|)arison of moth diversity between lightly and

heavily logged sites in a tropical rainforest. Malayan Nature
journal 56: 23-41.

Goi.wfi 1., R.K. & j.A. GoDDiNcnoN 1994. Estimating terrestrial
biodiversity through extra|rolation. Philosophical Transactions
of the Royal Society (B) 345; 101-118.

Got.WFi.t., R.K. 2000. Estimates: Statistical estimation of species
richness and shared species from samples. Version 5.01 , User's
Guide and a|)plication published at http://vicerQy.eeh. uconn.
edu/ estimates

Difiii , E.W. 1982. Die Sphingiden Sumatras. Heterocera Sumatrana
1 - S]rhingidae. E. Glassey.

D'Ahrfr.\, B. 1986. Sphingidae Mundi - Hawkmoths of the World.
E.W. Glassey Ltd., Faringdon, L'K.

Ffi.senstfin, j. 1985. Phvlogenies and the comparative method.
American Naturalist 125: 1-12.

2004. PHYI4P (Phylogeny Inference Package), version 3,6,

Distributed by the author. Department of Genome Sciences,
University of Washington, Seattle,

Fifdi.fr, K. & G.H. Sciiui.zE 2004. Forest modification affects
diversity, but not dynamics of specious tropical pyraloid moth
communities. Biotropica 36: 615-627.

FRFCRt F i ON, R.P., P.H. Harvey & M. Pagel 2002. Phylogenetic analysis
and comparative data: A test and review of evidence. American
Naturalist 160: 712-726.

Frith, C.B. & D.W. Frith 1985. Seasonality of insect abundances in
an Australian upland tropical rainforest. Australian journal of
Ecology' 10: 237-248.

G.vriand, T.,jR., P.E. Midford & A.R. Ives 1999. .Vn introduction to
|ihylogenetically based statistical methods, with a new method
for confidence intenals on ancestral values, American Zoologist
39: 374-288.

Gr;\ssi.e, j.F. & W. Smhh 1976. A similarity measure .sensitive to
the contiibution of rare species and its use in investigation of

39: 18-37, 2000 (2006)

36

variation in marine benthic communities. Oecologia 2.5:13-

22.

Hengevei.d, R. & A.J. Siam 1978. Models explaining Fisher’s
log-series abundance curve. Proceedings of the Koninklijke
Nederlandse Akademie van Wetenschappen C 81: 415-427.

Hoi.i.oway, J.D. 1967. Studies and suggestions on the behavior of
moths at light. Proc. S. Loud. Ent. Nat. Hist. Soc.: 31-46.

1976. Moths of Borneo with special reference to Mt.

Kinabalu. Malayan Nature Society, Kuala Lumpur.

1986. The moths of Borneo - Part 1: Key to families;

Families Cossidae, Metarbelidae, Ratardidae, Dudgeoneidae,
Epipyropidae and Limacodidae. Malayan Nature Journal 40:
1-166.

1987. The moths of Borneo, part 3: Lasiocampidae,

Eupterotidae, Bombicidae, Brahmaeidae, Saturniidae,
Sphingidae. The Malay Nature Society and Southdene Sdn.
Bhd., Kuala Lumpur.

1 993. The moths of Borneo: Family Geometridae, Sultfamily

Ennominae. The Malayan Nature Journal 47. The Malayan
Nature Society, Kuala Lumpur.

1996. The Moths of Borneo: Family Geometridae, Subfamilies

Oenochrominae, Desmobathrinae and Geometrinae. The
Malayan Nature Journal 49. The Malayan Nature Society,
Kuala Lumpur.

1997. The Moths of Borneo: family Geometridae,

subfamilies Sterrhinae and Larentiinae. The Malayan Nature
Journal 51. The Malayan Nature Society, Kuala Lum|>ur.

Ho[.i.ow.ay,J.D., G. KiiiBY& D. Pkggik 2001. The families of .Malesiau
moths and butterflies. Fauna Malesiana Handbook 3. Brill
(Leiden, Boston, Koln).

Hslao, H. S. 1973. Flight paths of night-flying moths to light. Journal
of Insect Physiology 19: 1971-1976.

Hubbell, S.P. 2001. The unified neutral theory of biodiversity
and biogeography. Monographs in Population Biology 32.
Princeton University Press.

1ntac:hat, J. & l.P. Wttiwon 1999. Trap design for monitoring moth
biodiversity in tropical rainforests. Bulletin of Entomological
Research 89: 153-163.

Intachat, J., J.D. Hoi.i.oway & H. Seaines 2001. Effects of weather
and phenology on the abundance and diversity of geometroid
moths in a natural Malaysian rain forest. Journal of Tropical
Ecology 17: 411-429.

Janzen, D.H. 1984. Two ways to be a tropical big moth: Santa Ro.sa
saturniids and sphingids, pp 85-140. In\ R. Dawkins & M.
Ridley (eds.) , Oxford suiweys in evolutionary biology 1 . Oxford
University Press, Oxford.

Janzen, D.H. 1993. Caterpillar seasonality in a Costa Rican diy forest,
pp 448-477. In: N.E. Stamp & T.M. Casey (eds.). Caterpillars,
ecological and evolutionai7 constraints on foraging. Chapman
& Hall, New York.

Kafo, M., T. Inoue, A.A. H.v.mid, T. N.ao.amitsu, M.B. Meroek, A.R.
Nona, T. IriNtt, S. Ya.mane & T. Yumoto 1995. Seasonality anti
vertical structure of light-attracted insect communities in a
Dipterocaip forest in Sarawak. Research in Population Ecolog)'
37: 59-79.

Kenney, A.J. & CJ. Krebs 2000. Programs for Ecological Methodology,
Version 5.2. Published at http: / /xmuw.zoology.uhr.ca/~krcbs

Kitayama, K., M. Lakim & M.Z. Wahab 1999. Climate profile of Mount
Kinabalu during late 1995-early 1998 with special reference to
the 1998 drought. Sabah Parks NatureJournal 2: 85-100.

Kitghing, l.J. & J.-M. Cadioli 2000. Hawkmoths of the world. The
Natural History Museum, l,ondon. Cornell University Press,
London.

KovAcs, L. 1958. Quantitative Lhitersuchungsmethoden bei
Schmetterlingen. Acta Zoologica 4: 191-206.

Krei.i. E.-T. 2004. Parataxonomy vs. taxonomy in biodiversity
studies - pitfalls and applicability of ‘morphospecies’ sorting.

Biodiversity and Conservation 13: 79.5-812.

Kroneei.d-Sc:hor, N. & T. Dayan 2003. Partitioning of time as an
ecological resource. Aunual Review of Ecolog)', Evolution and
Systematics 34: 153-181.

Loder, N., K.J. G.A.STON, P.H. Warren & 1 l.R. Arnold 1998. Body size
and feeding specificity: macrolejridoptera in Britain. Biological
Journal of the l.innean Society 63: 121-139.

Marsh, C.W. & A.G. Greer 1992. Forest land-use in Sabah, Malaysia:
An introduction to Danum Valley. Philosophical Transactions
of the Royal Society, London (B) 335: 331-339.

McGeachie, W.J. 1989. The effects of moonlight illuminance,
temperature and wind speed on light-trap catches of moths.
Bulletin of Entomological Re.search 79: 18.5-192.

Meineki;, T. 1984. Untersuchungen zur Struktur, Dynamik und
Phiinologie derGroBschmetterlinge (Insecta, Lepidoptera) im
sudlichen Niedersachsen. Mitteilungen zur Fauna und Flora
Sud-Niedersachsens 6 (pp 1-453). Ph.D. -thesis, Universitiit
G(”)ttingen (Germany).

Mei.o, A.S., R.A.S. Pereir.\, A.J. Santos, G.J. Sherherd, G. M.achado,
H.F. Medeiros & R.J. Sawaya 2003. Comparing species
richness among assemblages using sample units: why not use
extrapolation methods to standardize different sample sizes?
Oikos 101: 398-410.

Moran, M.D. 2003. Argumenls for rejecting the sequential
Bonferroni correction in ecological studies. Oikos 100: 403-
405.

Moreno, C.E. & G. H.vi.FFri R 2000. A.s.se,ssing the completeness of
bat biodiversity inventories using sjtecies accumulation curves.
Journal of Applied Ecology 37: 149-158.

2001. On the measure of sam]5ling effort used in species

accumulation curves. Journal of Applied Ecolog)' 38: 487-
490.

Miiiriiead-Tiiomp.son.R.C. 1991. I'rap Responses of Flying Insects.
Academic Press, London.

Novotw,V. & Y. BA.s.sEr 1998. Seasonality of sap-sucking insects
(Achenonwncha, Hemi|)tera) feeding on Ficus (Moraceae) in
a lowland rainforest in New Guinea. Oecologia 1 15: 514-522.

Novotny, V., S.E. Miller, Y. B.\.s.set, 1.. Cizek, P. Drozd, K. D\rrow
& J. Lei’s 2002. Predictablv simple: communities of caterpillars
feeding on rainforest trees in Papua New Guinea. Proceedings
of the Royal Society (B) 269: 2337-2344.

Onsager, J.A. & A. D.av 1973. Efficiency and effective radius of
blacklight traps against Southern Potato Wireworm. Journal
of economical Entomology 66: 403-409.

Per.sson, B. 1976. Influence of weather and nocturnal illumination
on the activity and abundance of populations of Noctuids an
south coastal Queensland. Bulletin of Fhitomological Research
66: 33-63.

Peler.son, A.T. & N.A. Si.ade 1998. Extrajjolating inventory results
into biodiversity estimates and the importance of stopping
rules. Diversity and Distribution 4: 95-105.

Piaut, H.N. (1971). Distance of attraction of moths oi Spodoptmi
liltomlis to BL radiation, and reca|)iure of moths released at
different distances of an E.SA blacklight standard trap. Journal
of Economical Entomology 64: 1402-1404.

Primer 5 eor Windows 2002. Plymouth Routines in Multivariate
Ecological Research, Version 5.2.9. Primer-E Ltd., Plymouth
(UK).'

Reeve, J. & E. Abouiieif 2003. Phylogenetic Independence. Version
2.0, Department of Biology, McGill University. Distributed freely
by the authors on recjuest.

Riieindt, F.E., T.U. Gr/\fe & E. Abouiieif 2004. Rapidly evolving
traits and the comparative method: how important is testing
for phylogenetic signal? Evolutionary Ecolog)' Research 6:
377-396.

Robinson, G.S. 1998. Bugs, hollow curves and specie.s-diversity
indexes. Stats 21: 8-13.

3(3

/. Rfs.Lt'pid.

ScHOKNER. r.W. 1980. l.ength-vveight regressions in tropical
and temperate fbrest-nnderstorey insects. Annals of the
Entomological Society of America 73: 10(3-109.

Scniii.ZE, Ci.H. 2000. Answirkimgen anthropogener Storungen aiif
die Diversitiit von 1 lerbivoren - Analyse von Nachtfalterzdno.sen
entlang von Hahitatgradienten in Ost-Malaysia. Ph.D.-thesis,
Universitiit Bayreuth (dermany).

Sciiui.ZE, C.H. & K. Fiedi.er 1997. Patterns of diversity and
vertical stratilication in havvkmoths of a Bornean rain forest.
Mitteihmgen der Dentschen Gesellschaft fiir allgemeine mid
angewandte Entomologie, Band 1 1/Heft l-(3.

Schulze, (i.H. & K. Fiedler 2003. Vertical and temporal diversity
of a specie.s-rich moth taxon in Borneo, pp (39-88. hr. Y. Basset,
V. Novotny, S. E. Miller & R. L. Kitching (eds.), Arthropods of
tropical forests. S|)atio-temporal dynamics and resource ttse in
the canojty. damhridge Lhiiversity Pre.ss, Cambridge [EIK].

SciiLii.zE, C.ll., (i.L. M.Ui.ser & M. M.\RV,vri (2000). A checklist of
the havvkmoths (I.epidoittera: Sphingidae) of Kinabalu Park,
Sabah, Borneo. .Malayan Nature Join nal ,54: 1-20.

Simon, U. & K.F3. Lin.senm.vir 2001. Arthropods in tropical oaks:
differences in their s|>atial distributions within tree crowns.
Plant Ecology 153: 179-191.

Smvi iie, N. 1985. The seasonal abundance of night-dying insects in a
Neotropical forest, p]) 309-318. In: E.C. Leigh, Jr., A.S. Rand &
D.M. Windsor (eds.), The ecology ofa tropical forest: seasonal
rhythms and long-term changes, 2'“' ed. Smithsonian Tropical
Research Institute, Panama.

SormiB.VNDiiu, S. & R.P B.vker 1979. Celestial orientation by the
large yellow underwing moth, Noclua pronuba L. Animal
Behavior 27: 78(3-800.

SoutnwooD, T. R. E. & PA. Hender.son 2000. Fxological Methods
(3"' edition). Blackwell Science l.td., Oxford.

Si’ENCER, J.L., I,.J. Cew.vx, J.F). Keeler & J.R. Miller 1997.
Chemilnminiscent tags for tracking insect movement in
darkness: a|)plication to moth [thoto-orientation. The Creat
Lakes E.ntomologist 30: 33-43.

StatSoit, Inc:. 2003. S'l’A'riS'l'lCA.Version (3. Computer program &
mamial at www.statsoft.com

Slissenbac'.h, D. 2003. Diversitiit von Nachfaltergemeinschaften
entlang eines Hdhengradienten in Sndecuador (Lepidoptera:
Pyraloidea Arctiidae). Ph.D.-thesis, University of Bayreuth
(Cermany) . Available at http://www.opus. ub. uni-bayrcuth.de/
i'oUtexte/2()()3/

SussENii.uiH, D. & K. Fiedler 1999. Noctnid moths attracted to
fruit baits: testing models ands methods of estimating species
diversity. Nota lepidopterologica. 22: 1 1. 5-1 54.

SussENB.vcH, D. & K. Fiedi.er 2000. F'atmistische Ergebnisse einer
Kdderfangstndie an Enlenfaltern (Noctnidae) im Obermain-
Hiigelland. Berichte der natnrwissenschaftlichen Cesellschaft
Bayreuth 24: 25.5-272.

Toreshi, M. 1993. Species abundance patterns and community
structure. Advances in Ideological Research 24: 111-186.

Tanak/V, L.K. & S.K. Tanara 1982. Rainfall and seasonal change of
arthropod abundance on a tro|tical oceanic island. Biotropica
14: 114-123.

Tennent, W.J. 1991. Notes on some Borneo hawk moths
(Lpeidoptera: Sphingidae) including Eupanacra hoUowayi sp.
n. and Macro^hissum amoenum Rothschild & Jordan, new to
Borneo. Entomologist’s Record 103: 223-235.

Thomas, A.W'. 199(3. Light-trap catches of moths within and
above the canopy of a northeastern forest. Journal of the
Lepidopterist’s Society 50: 21-45.

Trexi.er, J.C. &■ J. Tr.\vis 1993. Non-traditional regression analyses.
Ecology 74: 1(329-1(337.

Wai.i.ace, A.R. 18(39. 4'he Malay Archipelago. Oxford in Asia
Hardback Re])rint.s (198(3), Oxford University Press, Oxford.

Wai.sh, R.P.D. & D.M. Newberv 1999. The ecoclimatology of Dannm,

Sabah, in the context of of the world’s rainforest regions,
with [>articidar reference to dry periods and their imjjact.
Philosophical Transactions of the Royal Society (London) B
.354: 18(39-188.3.

WiLLorr, S. J. 2001. Species acennudation curves and the
meastiremeiu of sampling effort. Journal of Applied Ecology
38: 484-48(3.

Woi.DA, H. 1978. Fluctuations in abtmdance of tropical insects.
American Naturalist 112: 1()17-1()45.

1981. Similarity Indices, Sample Size and Diversity.

Oecologia 50: 29(>302.

1988. Insect seasonality: why? Annual Review of Ecology

and Systematics 19: 1-18.

1992. Trends in abundance of tropical forest insects.

Oecologia 89: 47-52.

Woi.DA, 11. & R.W’. F'lowers 1985. Seasonality and diversity of
mayfly adults (Ephemeroptera) in a ‘nonseasonal’ tropical
environment. Biotropica 17: 3.30-.3.35.

Yela, J.L. & .M. Holyoak 1997. Effects of moonlight and
meteoi'ological factors on light and bait trap catches of nocttiid
moths (Le|tido|5tera: Noctnidae). Environmental Entomology
2(3: 1283-1290.

Zaidi. M.l. & M.S. Chong 1995. Moth fauna of Sayap - Kinabalu
Park, Sabah: A preliminary stiney and assessment, pp. 20 1-209.
In. Cj. Ismail & L. bin Din (eds.), A scientific jotirney through
Borneo. Sayap-Kinabalu Park, Sabah. Pelanduk Ptiblications,
Selangor Darul Ehsan.

Journal of Researrh on the LepidoJHera

39: 37-79, 2()0() (2006)

A review of the genus Manerebia Staudinger

(Lepidoptera: Nymphalidae: Satyrinae) in the northern Andes

Tomasz W. Pyrcz

Zoological Museum of thejagielloiiian University, Ingardena 6, 30-060 Krakow, Poland
j)')m'ztomasz@hotmail.com

KkilTH R. WiLLMOTT

Florida Museum of Natural History, University of Florida, Oaine.sville, Florida, USA
kmnllmolt@holmail.com

Jason R W. Hall

National Museum of Natural History, Smithsonian Institution, Washington D.C., USA
hallja@si.edu

Angel L. Viloria

Uentro de Ecologi'a, Instituto Venezolano de Investigaciones CientiHcas, Apartado 21827, Caracas 1020-A, Vene/uela
aviloria @oikos. ivic. ve

Abstract: The taxonomic limits of the neotrojrical montane satyrine genus Manerebia Staudinger
are defined, with the names Penrosada Brown and Po.steuptychia Forster treated as junior subjective
.synonyms. The taxonomic history of the genus is discussed and the current knowledge on the
distribution, ecology and behavior of all tiorth Andean species is sumtnarised. A specie.s-level
taxonomic review of north Andean Manerebia is pre.sented that inchtdes 23 species and 37 taxa. Of
these, 10 new species and 13 new subspecies are described here: M. gennaniaen. sp., M. golondrina n.
sp., M. magnifica n. sp., M. manimulhus n. sp., M. pervaga n. sp., M. phiviosa n. sp.. A/, prattorum n. sp.,
M. rufanalis n. sp., M. seducta n. sp., M. undulala n. sp., M. franciscae rodnguezi n. ssp., M. germaniae
vitaki n. ssp., M. inderena antioquiana n. ssp., M. inderena clara n. ssp., M. inderena jina n. ssp., M.
inderena leaeniva n. ssp., M. inderena mirena n. ssp., M. inderena similis n. ssp., M. leaena gonzak'zi ii.
ssp., M. rufanalis fernandina ii. ssp., M. .salura lamasi n. ssp., M. satura pauperata n. ssp., M. undulala
milaena n. ssp. A lectotype is designated for M. leaena to stabilise future nometiclature and Penrosada
lanassa f. negkcta is placed as a subs|)ecies of Manerebia ignilineata n. stat. Accounts are |)resented
for each species, discussing identification, taxonomy and ecolog)’, and listing known geogra|rhic
and elevational range data. Adult specimens, drawings of male genitalia and distribution maps are
figured for all taxa where possible and the location of type material is given.

Keywords: cloud forest, Colombia, ecotone, Ecuador, elevational range, elfin forest, identification,
M. franciscae rodriguezin. ssp., M. gennaniaen. sp., AT germaniae vitaki n. ssp., AT golondrina n. sp., M.
indeiena antioquiana n. ssp., AT inderena clara n. ssp., M. indere?ia pna n. ssp., M. inderena kaeniwi n.
ssp., M. inderena mirena n. ssp., M. inderena similis n. ssp., M. seducta n. sp., AT leaena gonzalezi n. .ssp.,
AT magnifican. sp., M. rnammuthusn. sp., M. pemaga n. sp., M. pluvio.sa n. sp., M. jnattorum n. sp.. AT
rufanalis n. sp., AT rufanalis fernandina ii. ssp., M. .salura lamasi n. ssp., AT satura pauperata ii. ssp.,
AT undulala n. sp., AT undulala milaena n. ssp., paramo, Penrosada, Peru, Po.steuptychia, .sy.stematics,
taxonomy, tree-line, Venezuela.

Introduction and methods

The taxonomy of the neotropical Satyrinae re-
mains the most poorly understood of any nymphalid
butterfly group. The high levels of endemism and
diversity in the tropical Andes make understanding
the systematics of its fauna particularly challenging,

Received: 21 December 2004
Accepted: 28 May 2005

and this is certainly true of the speciose genus Maner-
ebia Staudinger. A notable feature of the genus is the
external similarity of many species, which can only
be identified with certainty through dissection, and
this fact has led to much confusion in the literature
and resulted in a significant underestimation of the
true taxonomic diversity of the genus. In this paper
we review the Manerebia fauna of the north Andean
region, where the highest number of undescribed
and taxonomically confusing taxa occur. The genus is

38

/. Res.Lepid.

remarkable for only 13 of the 23 known north Andean
species, and 14 of the 37 known north Andean taxa,
having been described to date, clearly illustrating the
importance of continuing collection activity in this
region. We aim to establish the identity of all historical
names applying to taxa from this region and describe
all unnamed taxa known to us, hopefully creating a
firm foundation for future taxonomic revision of the
entire genus.

The authors have been involved in studying the
diversity and taxonomy of the Andean satyrine but-
terflies of Venezuela, (Colombia and Ecuador for more
than a decade (e.g. Pyrcz et ai, 1999; Pyrcz & Viloria,
2004; Viloria, 2001). The region between southern
Ecuador and northern Peru forms a prominent
biogeographic divide for many montane butterfly
taxa (Willmott, Hall, Pyrcz, unpublished data), and
the majority of north Andean Manerebia species do
not occur further to the south. We therefore define,
for the purpose of this paper, the north Andes as
the area north of the Rio Chamaya - Maranon valley
in northern Peru, also known as the Huancabamba
deflection (Dillon, 1994).

Original descriptions were consulted for all pub-
lished Manerebia names and attempts made to locate
type material. specimens were examined in

collections in Europe, North and South America, as
listed below. One problem faced by earlier authors was
a lack of reliably labelled material in collections. How-
ever, extensive fieldwork throughout the region by
ourselves and others has provided a large amount of
new information on habitat preferences, geographic
and elevational distribution and phenotypic variation.
This information has been extensively used in propos-
ing new relationships between allopatric taxa, but we
have nevertheless had to make a number of arbitrary
assignments of taxonomic rank, based on incomplete
information. In general, where there are no appar-
ent differences in either genitalic characters, or wing
pattern cliaracters that do not vary intraspecifically
elsewhere in the genus, and closely allopatric taxa
have similar elevational ranges and occur in similar
habitats, we have regarded them as conspecific.

Dissections were made of males only, due to the
rarity of females in collections, and where necessary,
type specimens were dissected to confirm their iden-
tity. Abdomens were soaked in hot 10% KOH solution
for 15 min and subsequently stored in glycerol for
study under binocular microscope. Morphological
terms for genitalia largely follow Klots (1956) and for
venation follow Comstock & Needham (1918). The
following collection codens are used in the text;

AFEN: Collection of Andrew F. E. Neild, London, UK
AME: Allyn Mnseiini otFintoinology, Sarasota, USA (now at

McGuire Center for Lepidoptera and Biodiversity, Gainesville,
U.SA)

BMNH: The Natural History' Museum, London, UK (T=Type
coll.. R=Rothschild coll., A&B=Adams & Bernard coll.)

GR: Collection of Gabriel Rodriguez, Medellin, Colombia
IMLT: Fundacidn e Institiito Miguel Lillo, Tucuman, Argentina
JFLC: C lollection of Jean-Frangois Le Grom, Bogota, Colombia
KWJH: C lollection of Keith Willmott and Jason Hall, Gainesville,
USA

MALUZ: Mirseo de Artropodos de la Universidad del Zulia,
Maracaibo, Venezuela

MBLI: Collection of Maurizio Bollino, Lecce, Italy
MECN: Museo Ecuatoriano de Ciencias Naturales, Quito,
Ecuador

MHNUN: Mirseo de Historia Natural de la Universidad
Nacional, Bogota, Colombia

MHNUP: .Museo de Historia Natural de la Universidad
Redagdgica, Bogota, Colombia

MI2^: Museo de F.utomologi'a de la Universidad Central,

Maracay, Venezuela

MZPAN: .Muzeuin i Instytut Zoologii Polskiej Akademii Nauk,
Warsaw, Poland

MUSM: Museo de Historia Natural, Universidad Nacional Mayor
de San Marcos, Lima, Peru

MZUJ: Muzeum Zoologiczne Uniwersytetu jagiellohskiego,
Krak(')w, Poland.

PB: Collection of Pierre Boyer, Le Puy Sainte Reparade, France
PUCE: Museo de Fnitomologi'a, Pontificia Universidad Catolica
del Ecuador, Quito

PUJ; Dep: irtainento de Biologfa, Pontificia Lbiiversidad
Javeriana, Bogota, Colombia

SMTD: Staatlische Museum fiir Tierkunde, Dresden, Germany
TWP: Collection of Tomasz Wilhelm Pyrcz, Warsaw, Poland
USNM: United States National Museum - Smithsonian
Institution, Washington, USA

ZMHU: .Museum fur Naturkunde der Humboldt Universitat,
Berlin, Cermany

DHW, VHW, DFW, VFW: dorsal hindwing, ventral hindwing,
dorsal forewing, ventral forewing

Systematic overview

The name Manerebia was initially proposed by
Staudinger (1897) for five new, closely related Boliv-
ian and Peruvian species (M. cyclopina, M. cyclopeUa,
M. Cyclops, M. lyphlops and M. thyphlopsella) . Subse-
quent taxa described in the genus were also typically
from the sonthern tropical Andes (Schaus, 1902;
Hayward, 1949; Eorster, 1964), with the exception of
the Colombian M. nevadensisYechg^Y (Kriiger, 1925)
and Ecuadorian M. knadialeuka Hayward (Hayward,
1968). Brown (1944) introduced the generic name
Pen rosada for a cluster of mainly north Andean species
formerly often placed in Lytrertwo/toz/a Westwood, 1851,
including P. leaena (Hewitson), P. apiculata (C. & R.
Felder), P. lanassa (C. & R. Felder), P lisa (Weymer,
1911), P. sat lira (Weymer), P. cillutinarcn (Weymer,
1912, a synonym of M. zoippus (H. Druce, 1876) ) , and
P keithi (Dyar, 1913, a synonym of M. satura) . Adams &
Bernard (1979, 1981) and Adams (1986) described a
further four species of Penrosadabom Colombia and
Venezuela, and historically authors have continued

39: 37-79, 2000 (2006)

39

to regard both Manerebia and Penrosada as distinct
genera, without further discussion of the taxonomic
relationships of their members (e.g., Forster, 1964;
Miller, 1968; Adams & Bernard, 1977, 1979, 1981; Ad-
ams, 1985, 1986; D’Abrera, 1988; Racheli & Racheli,
2001).

Having examined the head, thorax, wing vena-
tion and pattern, and male genitalia of all species
formerly placed in Penrosada Brown and Manerebia
we have found no synapomorphies that distinguish
the two genera, and we therefore follow Lamas &
Viloria (2004) in considering the former as a sub-
jective junior synonym of the latter. Brown’s (1944)
original description of Penrosada made no reference
to Manerebia, and his morphological diagnosis of the
genus applies fully to the species originally treated in
MrtncrcZ^t’rtby Staudinger (1897), namely, the hindwing
vein M1-M2 is characteristically short (shorter than in
Lymanopoda Westwood) and gently curved, the root
of vein M3 is much closer to vein Cul than M2, the
hindwing is slightly incised near the anal angle, the
ocellus in cell 1A-Cu2 of the forewing and hindwing
is usually fully developed, and the walking legs are
yellowish. The conspicuous obliqtie, straight yellow or
whitish band on the hindwing underside present in
most Penrosada of Brown and absent in the hve origi-
nal Manerc/^irt of Staudinger almost certainly does not
define a monophyletic group, being highly variable
between and even within populations. Some species
of Manerebia, as defined here, such as M. ignilineata,
M. interrupta and M. apiculata, are polymorphic, with
the band shortened, discontinuous or even entirely
absent. Other typical features of the genus Manerebia
include male genitalia characterised by a long, arched
uncus, fully developed subuncus, and slender valvae
with a strongly dentate dorsal edge.

We also consider the genus Posteiiplychia Forster a
subjective junior synonym of Manerebia, where it was
placed by Lamas & Viloria (2004). Forster (1964)
erected the monotypic Posteiiplychia for Pronophila
mycalesoides C. & R. Felder on the strength of its
slightly unusual male genitalic morphology, but did
not notice that species he treated in Manerebia were
also similar in this respect, and shared similar wing
color pattern and venation. Indeed, M. mycalesoides
is perhaps genitalically most similar to M. nevadensis,
one of the few north Andean species actually originally
described in Manerebia.

Miller (1968) placed Manerebia in the tribe
Pronophilini Reuter, a neotropical section of the
Satyrinae. Adams & Bernard (1977, 1979, 1981),
Adams (1986) and Pyrcz (1999) did not question this
decision, but Viloria (2001) suggested that Manerebia
does not belong in the Pronophilina, but in the pre-

dominantly Holarctic Erebiina (following Harvey’s
(1991) arrangement, downranking former tribes to
subtribes of the Satyrini). Viloria (2001) listed three
putative synapomorphies of the Pronophilina absent
in Manerebia: setose eyes, hindwing cross vein M1-M2
curved basad into the discal cell and hindwing discal
cell equal or shorter than the maximum length of the
hind’wing. The absence of setae on the eyes also occurs
in the Erebiina, where the genus is currently placed
(Lamas & Viloria, 2004), as well as in the Palearctic
Satyrina. We believe the character evidence in stipport
of this subtribal position is currently weak, and a thor-
ough cladistic analysis of the tribe Satyrini is clearly
needed, hopefully including additional characters,
especially those from the early stages (e.g., Harvey,
1991) and molectilar sequence data. Pyrcz (1995) and
Viloria (2001) suggest Idmania Pyrcz or Idioneunila
Strand as possible sister genera of Manerebia.

The species-level systematics of Manerebia are very
complex and have not been rendered easier by mul-
tiple errors made by earlier workers (see Appendix
1). Three publications are especially significant for
the description of new taxa: Brown (1944), Forster
(1964) and Adams (1986). In addition to describing
the generic synonym Penrosada, Brown (1944) also
attempted to treat all Ecuadorian species, though he
seemed unaware of two species already described at
that time from the cotmtry, M. trimaculala and M. ig-
niUnenln, redescribing the latter as forms of “Penrosada
Innassri'. He also mistakenly applied the names leaena,
lanassa and apiculata to various species and described
several infrasubspecific forms differing in the expres-
sion of the hindwing band as new taxa. In the absence
of any figures of specimens, identification of the taxa
that Brown was really treating mtist be based largely
on the imprecise genitalic illustrations. Eorster (1964)
clarified, to some extent, the classification of the Boliv-
ian species and illustrated the most poorly known taxa
described by Staudinger ( 1897) . LInfortunately, he did
not examine the genitalia of any species, and hence
he also did not notice that Penrosada ‘And Posleuplychia
were morphologically similar to other Manerebia.
Einally, Adams (1986) treated most of the species of
Manerebia (under the name Penrosada) occurring in
Colombia and Venezuela, jrroviding valuable distri-
butional data and describing a number of new taxa,
but also made some errors in identification similar
to earlier authors. Lamas & Viloria (2004) provided
a synonymic checklist for the entire genus, thus cor-
recting previotts taxonomic errors.

Morphology

The wing patterns of all Manerebia species are very

40

/. Res.Lcpid.

simple (Figs. 1-9). Both dorsal and ventral surfaces are
generally uniformly brown (though M. /tsr/ and two
new species described here have a coloured dorsal
band), and the most visible character, the ventral
hindwing postdiscal band, is also one of the most
variable. In addition to simple local and geographic
variation in width, orientation and color, the band
may be absent altogether, split into spots, or reduced
to small sections, all within a single population. The
evolutionary basis for the polymorpliism within this
hindwing band remains unknown, but it occurs in
several species that are apparently not a monophyletic
group. The only other apparent pattern elements are
fine, darker postdiscal and submarginal lines, and oc-
casional submarginal ocelli in cells Cul-Cu2 on the
forewing and cells Cail-Cn2 and IA-Cu2 on the hind-
wing, and these subtle characters prove to be some of
the most useful for distingnishing species.

The male genitalia (Figs. 10-14) are also taxonomi-
cally extremely useful, for although also simple in
structure they may differ significantly between species
that are externally almost indistinguishable. In the
north Andean fauna, taxa frequently fall into one
of two principal morphological groups (that may or
may not be monophyletic); one in which the nncns
is strongly arched, the subunci relatively long, the
‘teeth’ at the distal tip of the valva often extending
anteriorly along the dorsal edge, and the dorsal base
of the valva armed with a projection with numerous
small ‘teeth’ (M. leacna, M. gennaniae, M. pluviosa n.
sp., M. apiciilata, M. iiavarrae, M. Jmnriscae, M. mam-
muthusn. sp., M. satura), and one in which the nncns
is slightly curving or straight, the snbunci relatively
short, the ‘teeth’ confined to the distal tip of the
valva, and the dorsal base of the valva armed with a
simple projection only (A/, inderena, M. golondrina n.
sp., M. prnttorum n. sp.. A/, trimaculatn, AT undulata
n. sp., M. inlerrupta) . Within these two groups, more
subtle characters in the shape of the uncus, subunci
and valvae are often consistent within and between
populations, and provide clues to the relationships
of allopatric taxa.

Diversity, distribution, ecology and behavior

Lamas & Viloria (2004) recognise a total of 45
Manerebin species, including both described and un-
described species. The genus is exclusively Andean,
including the peripheral ranges of Sierra Nevada
de Santa Marta and the Venezuelan Cordillera de la
(iosta (Figs. 15-20), and local faunas in the central
and northern Andes comprise three to eight species
(Appendices 2, 3) . Some species ot' Manerehia occur in
premontane forests as low as 800 m ( A4. niycalesoides, M.

mngnifica n. sp. and some Bolivian species) . The genus
is, however, most diverse in middle and upper eleva-
tion cloud forest and elfin forests from 2300-3000m.
A few species occur in the paramo grassland above
the tree line up to 4000m (M. kvana, M. ignilineala,
M. sedurta n. sp.).

Most north Andean Manerehia occur in low popula-
tion densities. This is particularly true for species in
Venezuela and northern Colombia, where they are
among the rarest cloud forest butterflies. Manerehia
mycalesoides, the sole representative of the genus in
the Venezuelan Cordillera de la Costa, remained
unknown there until 1999, despite this being prob-
ably the best sampled mountainous area in South
America. In Colombia and Ecuador Manerehia s\icc\c^
abundance remains relatively low in comparison to
that of other sympatric cloud forest satyrine genera,
such as Pedaliodes Sutler, 1867, PronophilaMouhXc&Ay,
[ 1 849] , or Lynianopoda. Aggregations at water seepages
of more than a couple of individuals are a rare sight,
but, surprisingly, towards the southern tropical Andes
the reverse may be true. In the cloud forests of Bolivia
certain species o^ Manerehia Are occasionally extremely
abundant, and literally hundreds of individuals can
be i'requently observed along roadsides and forest
trails (Pyrez, pers. obs.). Seasonal fluctuations in
abundance have also been noted (Adams & Bentard,
1981) I'or some north Andean species, especially those
occurring in the paramo. Adults of some species are
observed almost exclusively during the wet season
{M. Jranciscae, M. inlerrupta) , whereas others seem to
be tnnch more abundant during the dry season {M.
undulata n. sp., M. inderena).

Several Manerehia species are confined to very
narrow ecological zones, such as the forest-paramo
ecotone {M. inlerrupta), while other north Andean
Manerehia, although wider ranging, are inconspicuous
butterflies of dense cloud forests, usually overlooked
by lepidopterists. These facts have resulted in their
poor representation in major collections. Fortunately,
however, all of the cloud forest species o^ Manerehia Are
readily attracted to decomposing organic matter, es-
pecially dung and carrion, and the use of baited traps
provides a rapid and reliable method of sampling.
Paramo species also come to bait, but less readily, and
sampling for species above the tree-line must also be
done with hand-nets.

Some species of Manerehia demonstrate notable
wing pattern similarities to unrelated, microsympat-
ric pronophiline satyrines. Adams & Bernard (1979)
remarked on the similarity of tw'O Colombian species,
M. quinterae Aud M. navarrae, to two respectively sym-
patric Eretris species, and stated that there was “little
doubt” that mimicry was occurring. A further, even

39: 37-79, 2000 (2006)

41

more striking example concerns M. pluviosa n. sp.,
which has a very unusual wing pattern for the genus,
a wide postdiscal orange band on both wings. This
species is microsympatric with a similarly patterned
Pedaliodes species (Pyrcz & Viloria, in press). To date
there is no evidence for unpalatability in any of these
species, and it remains to be demonstrated that these
similarities result from mimicry.

Very little is known about the early stages of Maner-
ehin. Adams & Bernard (1981) recorded the montane
bamboo Chusquea (Poaceae) as the food plant for
Manerebia frandscae. Greeney (Greeney & Pyrcz, in
prep.) observed second to fourth instar larvae of M.
inderena daran. ssp. inYanayacu (Ecuador, Napo) also
on a species of Chusquea. This food plant is also that
of most Neotropical cloud-forest satyrines and will
probably apply to the entire genus.

Species accounts
Manerebia Staudinger, 1897

Manerebia Stimdinger, 1897; 139. Type-specie.s Manerebia cyrlopi-
Staudinger, by sub.seqiient designation (Heinining, 1943: 24).
= Penrosada Brown, 1944: 25.6. Type-species Lymanopoda leaena
Hewitson, by original designation.

= Posleuptychia Forster, 1964: 137. Type-species Pronophila my-
calesoides C. & R. Felder, by monotypy.

[NOTE: only north Andean taxa are included here, with
the exception of M. satura satura; - indicates a subspecies, —
indicates a synonym]

leaena (Hewitson, 1861)

-lanassa (C. & R. Felder, 1867)

-gonzalezi Pyrcz & Viloria, ii. ssp.
germaniae Pyrcz & Hall, n. sp.

-vitalei Pyrcz & Willmott, n. ssp.
pluviosa Pyrcz & Viloria, n. sp.
apiculata (C. & R. Felder, 1867)

— rurvilhiea Weymer, 1912
frandscae (Adams & Bernard, 1981)

-rodriguezi Pyrcz & Willmott, n. ssp.
mammuthus Pyrcz & Willmott, n. sp.
satura (Weymer, 1911)

-paupirata Pyrcz & Willmott, n. ssp.

-laniasi Pyrcz & Willmott, n. ssp.
navarrae (Adams & Bernard, 1979)
quinterae (Adams & Bernard, 1979)
inderena (Adams, 1986)

-antioquiana Pyrcz & Willmott, n. ssp.
frna Pyrcz & Willmott, n. ssp.

-sirnilis Pyrcz & Willmott n. ssp.

-rlara Pyrcz & Willmott, n. ssp.

-leaeniva Pyrcz & Willmott, n. ssp.

-mirena Pyrcz & Willmott, n. ssp.
golondrina Pyrcz & Willmott, n. sp.
prattorum Pyrcz & Willmott, n. sp.

Irimaculata (Hewitson, 1870)
undidata Pyrcz & Hall, n. sp.

-milaena Pyrcz & Willmott, n. ssp.
interrupta (Brown, 1944)

— keradialeuka (Hayward, 1968)

rufanalis Pyrcz & Hall, n. sp.

-jernandina Pyrcz & Willmott n. .s,sp.
ignilineata (Dognin, 1896)

-neglecta (Brown, 1944) n. stat.

— disrontinua (Brown, 1944)
seducta Pyrcz & Willmott. n. sp.
mycalesoides (C. & R. Felder, 1867)

— (Butler, 1867)
magnifica Pyrcz & Willmott, n. sp.
neiiadensis Kriiger, 1 925
levana (Godman, 1905)
pervaga Pyrcz & Viloria, n. .sp.

Manerebia leaena (Hewit.son, 18(51)

The identity of leaena, the oldest available name
in Manerebia, is crucial for establishing a stable no-
menclature for this genus in the northern Andes.
The original description is concise (“Upperside dark
brown from the base to the middle, rufous-brown [be-
yond. Underside as above, except that the posterior
wing is crossed transversely beyond the middle, from
margin to margin, by a regular, nearly straight band
of pale yellow. Exp. 19/20 inch”) and accompanied
by a black and white figure of the ventral surface
(Hewitson, 1861). Hewitson did not give any type
locality, sex or number of sjjecimens examined, which
makes the correct identification of leaena difficult.
However, several other pronophiline satyrines were
described in the same paper (e.g., Lymanopoda laclea,
Lymanopoda labda, Lymanopoda aUmcincta) , all of which
were from Colombia.

Kirby (1879), in his catalogue of butterflies in
the collection of Hewitson in the BMNH, mentions
seven specimens of leaena from Ecuador. Four Hewit-
son specimens labelled as leaena were located at the
BMNH, including two labelled Ecuador and two
without any locality. Since these specimens actually
represent two different species, M. leaena and M.
imdulala n. sp., a lectotype designation is necessary.
The specimen selected as the lectotype of leaena very
closely matches the original figure, in particular in
the precise shape of the pale hindwing ventral surface
band, which is slightly notclied along the basal edge at
the base of vein M3 and tapers at the tornus, and in its
uniform ground colour and lack of ocelli. Given the
accuracy of Hewitson figures, we believe it is probably
the specimen on which the illustration was based. The
specimen has a typed label, which would have been
added subsequent to the specimen’s accession to
the BMNH, which has “PEcuaclor” written on it, with
“Quito” written over the top. Presumably the original
specimen lacked locality data and a guess was made
as to its origin based on other similar Hewitson speci-
mens. However, a very careful comjjarison of leaena-
like specimens from all samjjled localities in Ecuador

42

/. lifs.Lepid.

(specified in this paper) and Colombia reveals that
the wing shape and colour pattern of the lectotype
matches perfectly only those found in the Colombian
(iordillera Oriental east of Bogota. Adams (1986) col-
lected a series of specimens of M. leaena recently in
that area. The genitalia of the lectotype confirm this,
being typical of (Colombian I me na 'ds treated here. It
is not the only case when a pronophiline butterfly
described by Hewitson bears incorrect “Ecuador” lo-
cality data. Eretris pliyllalia (Hewitson), whose syntype
specimen shares the same type of label as the lectotype
of leaena, is a synonym oi' Erelris ajndeja apiileja (C. & R.
Felder, 1867) found in the vicinity of Bogota.

Having no access to the British type material,
Brown (1944) misidentified M. leaena dnd applied
the name to an uncertain number of Ecuadorian
species. Adams (1986) correctly recognised that M.
leaena is found in the Bogota region, but implied
that it was sympatric with M. 1. lanassa, which he
considered a se])arate species. He treated M. leaena
as a species occurring thronghont the eastern Andes
from Colombia to Pern, and figured the genitalia of
a Peruvian specimen of Manerebia haywardi (Pyrcz,
2004) that is similar externally but actually has very
distinct genitalia. Hence, the genitalia illustrations of
y\/. leaena dnd most other taxa in Brown (1944) and
Adams (1986) are incorrect. Further misidentifica-
tions are found in D’Abrera ( 1988), whose treatment
was based on the arrangement of the main collections
in the BMNH, and corrections to names of figured
specimens are provided here (Appendix 1). In fact,
M. leaena, although externally similar to many other
species, has distinctive male genitalia, which have a
toothed projection near the base of the ampulla and
a strongly arched uncus, similar to M. germaniae n.
sp., M. pluviosa n. sp., M. apirulala, M. navarrae, M.
satura, M. fra nciscae and iM. mamniuthus n. sp. Within
this group, the short, squat valvae, with numerous
spines at the tip and pronounced spines on the am-
pulla, closely resemble only those of M. germaniae n.
sp. However, M. leaena consistently differs from M.
germaniae in having fewer spines on tlie dorsal edge
of the valva near the ba.se. In addition, the thin dark
brown submarginal lines on the ventral surface are
undulate in M. leaena, but almost straight in M. ger-
maniae. Although these two taxa are not known to be
sympatric to date, the wing pattern and genitalic char-
acters that distinguish them are consistent throughout
their respective ranges (with the exception of a single
specimen of M. leaena leaena discussed ixnder that
taxon). In addition, M. germaniae nrdy actually prove
to be the sister taxon of M. pluviosa (see under that
species), which also has straight ventral submarginal
lines, and which occurs .sympatrically with M. leaena

gonzalezi n. ssp. in the Venezuelan Sierra de El Tama.
There is slight subspecific genitalic variation in M.
leaena, in the number and distribution of spines along
the dorsal edge of the valva.

Manerebia leaena leaena (Hewitson, 1861)

Figs. lA.B, lOA, 15

Lymanopoda leaena Hewitson (1861: 156, jil. 9, fig. 1). TL:
[doloinhia, Ciorclillera Oriental]. LT male (here designated):
“PF.cnador, Quito”, Hewitson (loll. 79-69. B.M. Type No. Rh. .S9.^5.
BMNH(T) [examined].

Penmsada leaena (Hewitson); Brown (1944: 256) (inisidenti-
lication).

Manerebia leaena (Hewitson): Lamas & Viloria (2004: 215).

Diagnosis: In tlie nominate subspecies (Figs. lA, B) the VHW
postdiscal band is ajtproximately twice the width of that in M.
leaena lanassa, ;md also broader than in A/. /. gonzalezi (described
below) . This taxon is similar in wing pattern to M. inderena leaeniva
(described below), but theV'HW band is slightly contracted basally
at the anal margin and a richer yellow than in M. inderena leaeniva
or A/, inderena clara (described below). In addition to the characters
given in the species account above, A/, germaniaen. s|r. is also gener-
ally larger, has narrower \'HW postdiscal bands and a characteristic
convex FVV outer margin, that is straight in M. leaena. A single male
specimen in the BMNH(.A&B) from the Colombian Sierra Nevada
del Cocny has straighter ventral submarginal lines, similar to Af.
germaniae, but the genitalia, size and other wing pattern characters
suggest that it is a slightly aberrant specimen of Af. leaena.

CommeiiLs: The nominate subspecies occurs in the Cordillera
Oriental in Colombia (Cnndinamarca. Boyaca) on the eastern
slopes of the Andes, in Cacheta at 2500 m (Adams, 1986) and Cruz
Verde, over 2700m (Kruger, 1925). Remaining literattire localities
must be considered less reliable, with some probably applying to
A/, inderena (Adams, 1986). Pyrcz’s (1999) report of a specimen of
Af. leaena collected by Kruger in the Cordillera Central refers to
Af. germaniae n. s|). (described below).

Manerebia leaena lanassa (C. & R. Felder, 1867), stat.
rest.

Figs. 1C, 1), lOB, 15

Pronophila leaena \c\v. lanassa C. & R. Felder (1867: 474). TL:
Colombia, Cnndinamarca, Bogota. HT male: Colombia, Cimdi-
namarca, Bogota, I.indig, Feldei coll. BMNH(R) [examined],

Lymanopoda leaenaiorm lanassa (C. & R. Felder); Weuner (1911:
pi. 52, rowf; 1912: 249).

Penmsada lanassa (C. & R. Felder); Brown (1944: 258) (misi-
dentified); Adams (1986: .Af)5).

Manerebia leaena lanassa (C. & R. Felder); Lamas & Viloria
(2004: 215).

Diagnosis: Male genitalic and wing pattern characters show that
this taxon is very closely related to M. leaena leaena. The male geni-
talia (Fig. lOB), inclutling that of the holotype, share the characters
of other Af. leaena subspecies, an arched uncus atid s|tines at the
dorsal edge of the valva near the base. However, the latter are less
prominent, there are no 'teeth' between the shotilder and the apex,
the a]tical 'teeth' are fewer and the valva is more slender. The VHW
yellow band is half the width of nominate Af. leaena and slightly
narrower than in Af. 1. gonzalezi n. ssp. Adams (1986) considered
Af. leaena and lanassa to be separate species, and stated that the
latter taxon ttccurred on both slopes of the Colombian Cordillera
Oriental. However, we have found no e\idence to support the latter
statement, and a careful examination of Adams material deposited
in the BMNl 1, and material in collections in Bogota did not reveal
any s|tecimen of lanassa from the east slopes of the Cordillera.
Similarly, in other consulted collections (TWP, JFL.Ci, MHNUP,

39: 37-79, 2000 (2006)

43

PUJ) this taxon is represented only by specimens collected west
and northwest of Bttgota. Since lanassa therefore seems to be the
western slope replacement of M leaena leaena, we consider the two
taxa conspecific (Lamas & Viloria, 2004).

Comments: Maneivhm leaena lanassa is restricted to the western
slope of the Cordillera Oriental in Colombia (Ciindinamarca,
Boyaca). Adams (1986) reports that it occurs from 2250-2700 m,
while TWP has collected it at 2900 m near Facatativa.

Manerebia leaena gonzalezi Pyrcz & Viloria, n. ssp.

Figs. IE, IOC, 15

Manerebia leaena n. .ssp. (Pyrcz & Viloria); Lamas & Viloria
(2004: 215).

Diagnosis: This taxon differs from the nominate subspecies
primarily by having narrower VHW postdiscal bands, that are,
however, not as narrow as in M. leaena lanassa. The VHW postdiscal
band is slightly arched in comparison with the straighter bands of
M. /. lanassa or the nominate subspecies. The genitalia (Fig. IOC)
are also distinct in having additional ‘teeth’ along the whole dorsal
edge of the valva.

Description: MALE (Fig. IE): Head, f/joravand abdomen-, same
as in other subspecies. Wings: Forewing (length: 18.5-19 mm, mean:
18.75 mm, n =4) triangular, torniis obtuse. Hindwing rotinded,
anal margin very slightly excavated near angle. DFW' uniform cof-
fee brown. DHW coffee brown; reddish scales sparsely scattered
towards submarginal and marginal areas; basal half of hindwing
with dense, long hairs. VFW dark brown; reddish scales scattered
along costal margin and distal half of wing, more heavily towards
apex and marginal area; submarginal dark chocolate brown zigzag
line; vestigial postdiscal white dot in cell Ciil. VHW ground colour
chestnut, except for veins which are covered with chocolate brown
scales; narrow yellow band (c. 1 mm wide) from apex to anal margin
close to torna! angle, crossing just at extremity of discal cell but
not entering into it, dark brown submarginal line lightly undulate.
Male genitalia (Fig. IOC): similar to other subspecies except with
additional ‘teeth’ along dorsal edge of valva.

FEMALE: Unknown.

Etymology: We dedicate this new subspecies to Luis Alfonso
Gonzalez, a forest guard of the Venezuelan Parque Nacional E4
Tama, one of the few in the area who knows every detail of its geogra-
phy and nature, in gratitude for his cooperation and hospitality.

Types: Holotype ma\e: VENEZUELA: Tachira, Parqtie Nacional
El Tama, selva nublada ariiba de Betania, 2400-2480 m, 24. VI 11. 1996,
A. Viloria leg., MALUZ; Paratypes (3 males): VENEZUELA: 1 male:
Tachira, Parque Nacional El Tama, via La Lfnea, 2300-2700 m,
17.rV.1996, T. Pyrcz leg., TWP; 1 male: Tachira, via Bramon a De-
licias, Mata Mula, 1850-2020 m, 18. IX. 1997, A. Neild leg., Af’EN;
COLOMBIA: 1 male: Norte de Santander, Parque Nacional Natural
El Tama, Henan, 2500 m, X1L1993, J. F. Le Crom /eg., JFLC.

Comments: Manerebia leaena gonzalezi is known from a few indi-
viduals seen and collected in the cloud forest at elevations between
2300-2700 m. It flies high around bamboos (Chnsquea) along with
Lymanopoda lebbaea C. & R. Felder, 1867, ssp. (Pyrcz & Viloria, in
press), Lymanopoda albocinrla (Hewitson, 1861 ) and Erelris porphyrin
(C. & R. Felder, 1867) ssp. (Pyrcz & Viloria, in press), but it is not
such a fast flying insect as most of the butterflies found with it.

Manerebia germaniae Pyrcz & Hall, n. sp.

This species is superficially very similar to several
sympatric Manerebia, especially M. inderena. Manerebia
inderena is best distinguished by the usual presence of
submarginal ocelli on the VHW that are most visible
in cells Cu2-Cul and Cul-M3, while in M. germaniae
the pale dots representing the submarginal ocelli

are most visible in the middle of the wing. Also, in
most M. inderena Ldxix the thin dark siibmarginal line
is undulate, but straighter in M. germaniae. The male
genitalia of M. germaniae, however, which has a toothed
projection near the base of the ampulla and strongly
arched uncus, readily helps distifiguish M. germaniae
from all other externally sitnilar species. The genitalia
is most similar to that of AT leaena, M. pluviosa n. sp.,
M. apiculata, M. navarrae and M. salura. Within this
group, the short, squat valvae, with munerous spines
at the tip and pronounced spines on the ampulla,
are most similar to proliable close relatives M. leaena
and M. pluviosa, both of which also occupy similar
habitats and occur allopatrically. In comparison with
M. leaena, M. germeiniaehns more ‘teeth’ at the dorsal
edge of the valva, is larger, and the submarginal lines
on the ventral surface are nearly straight (rather than
zigzag) , running close to the outer margin. The latter
character also occurs in AT pluviosa, which is easily
distinguished by wing pattern atid which is sympatric
with AT leaena in the Venezuelan Sierra de El Tama.

There is considerable infraspecific variation in
M. germaniae involving the width of the VHW yellow
postdiscal band, but only one geographic population
is differentiated enough to deserve subspecific status.
However, when additional data become available,
more fine-scale patterns in geograijhic variation iti
AT germaniae iWMf become apparetit.

Manerebia germaniae germaniae Pyrcz & Hall, (n.
sp.)

Figs. 1F,G, lOD, 15

Manerebia n. sp. (Pyrcz, Willmotl & Hall); Lamas & Viloria
(2004: 216, n. 1 123).

Diagnosis: The nominate snlrspecies differs fnnn M. g. vilalei in
having a narrower yellow postdiscal band on the \'HW.

Description: MALE (Fig. IF): Head: ffons with a tuft of dark
brown hair; eyes blackish-brown, smooth; labial pal])i covered with
blackish-ftrown hair; antennae dorsally chestnut, ventrally rufous,
with white scales at the base of each .segment, club dorsally slightly
darker than shaft. Thorax, dtnsal and ventral surface dark brown;
legs paler yellowish-brown. Abdomen: dorsal and ventral surface
darkt)rown. W7ngs; forewing (length: 17.5-20.5 turn; mean: 19 mm:
n=12) distal tnargin convex, ajtex rounded to subacute; liindwing
rounded with w'eaklv pronotinced notch at tornus. DF\V medium
brown. DHW medium brown, slightly darker towards base, with
pale postdiscal band from apex to tornus indistinctly showing from
ventral surface. \'FW ground colour medium brown, becoming
slightly paler from base to apex; very faint, thin, straight, darker
fjrown postdiscal line, panillel to distal margin, in cells Cu2-Cul
to M1-R5; a row of faint, pale |)ostdiscal spots in center of cells
Cn2-Cul to M2-M] gently curved basally, in sttme individuals outei
spots indistinct; thin, faint, uneven (btit not zigzag), darker brown
submarginal line from apex to tornus. VHW medium brown, slightly
paler towards distal margin; stniiglit, cream-colored postdiscal
band from costa to tornus through base of cell Cul-M3; thin, faint,
slightly uneven (but not zigzag), darker brown submarginal line
from apex to tornus. Male genitalia (Fig. lOD): uncus smoothh
arching, subunci relatively long, valvae short and squat with uu-
merous ‘teeth’ at distal tip, some extending towards base, and

44

/. Ih’s.Lepid.

dense c luster of 'teeth' at dorsal edge of valva near base; aedeagus
shallowly curving dorsally, with several short, jrosteriorly directed
spines laterally neat middle.

FEMAl.E (Fig. IG): Similar to male except slightly larger
(t'orewing length 21 mm) and lighter coloured on both wing
surfaces.

Types: Holotype male: FIGU/VDOR: Gotojtaxi, Pilalo, 27,5()-S()()()
m, Vlf.l99b. 1. Aldas /eg., MZUJ; Allotype female: same data as the
holotype except 1997, FW’l’; Paratypes (76 males and 6 females):
ECUADOR: 3 males: Fichincha, San |uan-La Victoria road, 3200
m, 30. 1.2002, T. Pyrez /eg., I'VV'P (2), MUS.Vl (1); 1 male: same
locality, 2900-29.50 m, 01.11.2002, F. Pyrez /eg, TWP; 2 males: same
locality, 33,50 m. 11.2002, I. Aldas /eg. 'FWP; 2 males, Pichincha,
Aloag-Tandapi road, km 13, 3000-3050 m, 11,2002, I. Aldas /eg,
TWP: 2 males: Pichincha, Aloag-'l’andapi road. Sector Los Alpes,
2700-2750 m, 20.1.2004, T. Pyrez & R. (farlacz /eg, FWP; 1 male:
Pichincha, Volcan Pasochoa 2000-2750,. 22.V11I.2003, T. Pyrez /eg,
TWP; 2 males: Cotopaxi, Pilalo, 2750-3000 m, \'I1.1990, 1. Aldas /eg,
TWP (I), MECN (I): 1 male: Pichincha, old Quito - Sto. Domingo
rd. 2700m. 12.\'lll. 1993, |. I fall /eg., KVy[ll; 1 male, Pichincha,
Yanacocha, V'olcan Pichincha, 3500 m, IS. IX. 1997, K. Willmott /eg.,
KW|I 1: 4 males: Fungurahna, Triimfo - Patate, El Tablon, 3000 m.
21. XI. 1998, P. Royer /eg., I VVP ( I ), I’B (3); 2 males: I'ungurahua,
Friunfo - Patate, El Fablon, 3000 m, 1.1999. I. .Aldas /eg, FWP; 3
males: Fungurahua, Banos. Rimtiin, 3000 m, 111.1999, 1. Aldas /eg,
FWP; 2 males: Fungurahua, Banos, Pondoa, 3365 m. 19. 1. 2002, J.
Wojuisiak & R. Garlacz /eg., FWP; 32 males: Tungurahua, Banos,
E,1 Tabl<')n, 3000 m, 111.1999, 1. .Aldas /eg MBl.l; 5 males: 'Fungu-
rahua, Banos, .3000 m. III. 1999, I. Aldas /eg, MBl.l; 1 male: Mo-
rona-.Santiago, Gualaceo-Lim(')n road, 3100 m, 09.11.2002, 1. Pyrez
/eg., TWP; 1 male: same locality, .3100 m, 09.111.1998. I’. Boyer /eg,
FWP; 1 male: Carchi, Fulcan-Maldonado, Volc;'m Chiles, 3000-3050
m, 27. VIII. 2004, T. Pyrez /eg, FWP; 1 male: Carchi, near lluaca,
2900-3200 m, HI. 1999, 1. Aldas /eg, MBl.l; 1 female: Pichincha.
San jttan-La Victoria, 3,300-3400 in', 30.1.2002, T. Pvrez /zg., FWP; 1
female: 'Fungurahna, I’riunfo-Patate, F4 Fablon, .3000 m, 2 1 .IX. 1998,
P. Bover /eg, FWP; 1 female: Fungurahua, Banos, Runtnn, 3000m.
\'III. 1998, 1. .Aldas /eg, FWP; 1 female: Fungurahua. Runtun, 2900-
2950 m, 21.1.2002, |. Wojtusiak, /eg, FWP; 1 female: Tungurahua,
Banos, El Fablon, 3000m, 111. 1999, 1. Aldas /eg, MBl.l; CGI.OMBIA:
5 males; Cauca, AAilcan Purace, Fermales de San Juan, 3150-3200
m, 28-30.111.1996, F. Pvrez /eg, FWP; 1 male; Caldas. Paramo de
I.etras. 2.3.\'II.1993, ]-F. l.e Crom /eg, FWP; 2 males: Antioquia,
El Retiro, 2300-2800 m, 2fV30.\'II1.2003, G. Rodriguez /eg, TWP;
1 male: .Antioquia, El Retiro, Reserva San Sebastian, 2500-2800
111. 12. IX. 200.3, T. Pyrez /eg, FWP: 1 female: Antioquia, El Retiro,
2300-2800 m, 18-25A'II.200.3, G. Rodriguez, /eg, FWP.

Etymology: Fliis species is named after the Ecuadorian ento-
mologist, Ciermania Estevez, formerly of the Museo Ecuatoriano
de Ciencias Natnrales in Quito, in recognition of her help during
our research in Ecuador.

Comments: Manerebia germaniae occurs at high elevations,
near or at the tree-line, and has been recorded in Ecuador from
2700-3500 111. The nominate subspecies is widespread along the
central .Andean ridge from Colombia (Antiocpiia) to south-central
Ecuador, where it occurs on both Andean slopes (Fig. 15). It is
associated with intact cloud Forest and remnant scrub. In western
Ecuador (Pichincha, Cotopaxi) it generallv replaces the local
subspecies of A/, inderena at higher elevations, but there is a nar-
row elevational band where the two occur in sympatiy, generally
between 2800-3100 m.

Manerebia germaniae vitalei Pyrez & Willmott, n. ssp.

Figs. Ill, lOE, 15

Manerebia n. sp., n. ssp. (Pyrez, Willmott & Flail); l.anias &•
Viloria (2004: 216, n. 1 1231)).

Diagno.sis: In this taxon the VFIW yellow postdiscal band is

a|)proximately 1 .5 times the width of that of the nominate subspe-
cies.

Description: MALE (Fig. IH): Head, thorax and abdomen: as in
the nominate subspecies. Wings: Forewing (length: 19.5 mm; mean:
19.5 mm; n=2) distal margin veiy slightly convex, ajiex subacute;
hindwing rounded with weakly pronounced notch at tornus. DF'W
surface medinm brown. DHW medium brown; pale postdiscal band
from apex to tornus indistinctly showing from ventral surface. VFAA'
ground colour medium brown, slighth lighter in snbmarginal area;
very faint, thin, darker brown postdiscal line, straight and parallel
to outer margin Irom vein R5 to Cu2, then curving towards tornus;
a row of faint, barely visible pale yellow postdiscal spots in center
of cells Cu2-Cul to M2-M1; thin, faint, uneven (but not zigzag),
darker brown submarginal line from a|)ex to tornus. VI IW medium
brown, slightlv paler towards distal margin; straight, approximately
2mm wide, cream-colored (with an orange shade towards anal
margin) postdiscal band from costa to tornus through base of
cell Gu 1-M.3; occasional 3 to 4 minute pale yellow postdiscal spots;
faint, slightly uneven (but not zigzag), darker brown submarginal
line from apex to tornus. Male genitalia (Fig. lOE): similar to that
of the nominate stibspecies.

FF.M.ALE: Unknown.

Type.s: Holotype male; ECU.ADOR: F.oja, 0.8ld road Loja
- Zamora. 2500 in, XI-XIL1999, 1. Aldas leg., ex MBl.l, MZUJ.
Paratype male; same data as the holotype, MBLl.

Etymology: Fhis stibspecies is dedicated to the Italian lepidop-
terist Fabio A'itale, from Lecce, who specializes in the Ithomiinae.

Comments: Fhis subspecies consistently differs from the nomi-
nate throughout its range in the width of the VFIW postdiscal band.
It is the souihei nmost representative of yV/. germaniae, and is known
to date only from the eastern slopes of the Ecuadorian Andes in
the province of Zamora-Chinchipe (Fig. 15), though it may also
extend into extreme northern Peru.

Manerebia pluviosa Pyrez & Viloria, n. sp.

Figs. 2A, I OF, 15

Manerebia n. sp. (Pvrez & A'iloria); Lamas & Viloria (2004: 216,
n. 1 120).

Diagnosis: This sjtecies is easily distinguished from any conge-
ner by the wide yellow postdiscal band on both fore- and hindwing
dorsal and ventral surface.

Description: MALF) (Fig. 2A): Head: frons with a tuft of dark
brown hair; labial palpi covered with short dark brown hair; eyes
dark brown, smooth; antennae dark brown dorsally, chocolate
brown ventrally with white scales at base of each segment, chib twice
as wide as shaft, of same colour. Thorax, dorsally dark brown, haii^,
ventrally |)ale brown; legs paler yellowish-brow n. Abdomen: dorsally
dark brown, laterally and ventrally pale browm. Wings: Forewing
(length: 19 mm, n=2) triangular; tornus slightly obtuse. Hindwing
rounded, anal margin slightly excavated near tornus. Dorsal ground
colotir itnifortn chocolate brown, except for entire postiscal yellow
band on both wings, more distinct on hindwing; basal half and anal
region of hindwing with long thin brown hairs. VFW colour pattern
similar to dorsal surface but in general more intense; |)ostdiscal
band more distinct and wider, reaching torntts; snbmargitial and
marginal regions lighter brown. VHW similar to dorsal sttrface but
more intense in coloitr; postdiscal band slightly broadened towards
central portion; series of four submarginal white dots in cells M2-M.3
to Cit2-1 A, respectivelv. the first one half diameter of others, which
are ca. 0.8 mm. Male genitalia (Fig. lOF): uncus arched, valva with
spitiy dorsal proce.ss near base, aedeagus short.

FF.MALE: Unknown.

Type.s: Holo/ypemale: VENEZUEIA: Tachira, Parqtie Nacional
El Tama, entre Betania y F.a Banderola, 2810 m, 23.VIIL1996. A.
Viloria leg., MALLIZ. Para/ypemale: VENEZL’ELA: Tachira, Parqtie
Nacional El Tanui, Fundo Piedra Blanca, San Vicente de la Revan-
cha, 2400 m, 9-12.XII.1997, F. Rey leg., TWP.

39: 37-79, 2000 (2006)

45

Etymology: The name is the feminine form of the Latin atljec-
tive “pluviosiis”, meaning rainy. The male liolotype was collected
while flying veiy weakly during heavy rain in the upper cloud forest
of the Sierra de El Tama .

Comments: The male genitalia of this species are most similar to
those of yV/. leaena Mid M. germaniae, with an arched tmcns and spiiu'
process at the dorsal base of the valva. In M. pluviosn the narrow
dark VHW snbmarginal line is smoothly curving, not dentate like
M. leaena, and in this respect resembles that of M. germaniae, which
may prove to be its sister species. Manerebia pluviosa occurs in the
Sierra de El Tama in habitats similar to those of M. germaniae, close
to the tree-line, and possibly replaces M. leaena gonzalni at higher
elevations. The paratype was collected in disturbed habitat in San
Vicente de la Revancha at an elevation that is probably beneath that
at which the species usually occurs. Manerebia pluviosa is extremely
similar (though about half the size) to the microsympatric pro-
nophiline Pedaliodes sp. Pyrcz & Viloria (in press), though it is not
clear whether this similarity results from mimiciy, and if so, what the
basis for this mimicry might be. Some 30 years ago M. pluviosa wds
also captured by j. Bechyne, at 3100 m in the paramo of a remote
mountain range in the north Colombian department of Norte de
Santander (Cerro Oroque), where no other butterflies have since
been collected. That specimen shows some distinctive characters
and might represent a separate subspecies, so it is therefore ex-
cluded from the type series. This species seems to be a rare insect
and has only been recorded during the wet season.

Manerebia apiculata (C. & R. Felder, 1867)

Figs. 2B,C, llA, 15

Pronophila apirulataC. 8c R. Felder (1867: 474). TL: Colombia,
Cundinamarca, Bogotii. ST male: “apicu-lata Feld/Felder Colin./

Rothschild Bequest B.M. 1939-1/4/Bogota Lindig type/Type”.
BMNH(R) [examined].

= Lymanopoda apiculata form curvilinea ’Wcymer (1912: 248).
TL: Colombia, East Cordillera, Cundinamarca, Boquerdn de
C(h)ipaque. ST male(s): ZMHU? [not located].

Lymanopoda apiculata (C. & R. Felder); Weymer (1911: pi. 53,
row a (apkkulata [sic]); 1912; 248).

Penrosada apiculata (C. & R. Felder); Brown (1944; 257, |>1. 1,
fig. 1622) (male genitalia); Pyrcz (1999: 367).

Manerebia apiculata (C. 8c R. Felder); Lamas & Viloria (2004:
215).

Diagnosis: The male genitalia (Fig. 1 lA) of this species are
similar to those of M. leaena. The species is easily distingtiished
externally from all other sympatric Manerebia in the Colombian
Cordillera Oriental by the acute forewing apex. The expression of
the VHW band is variable, exhibiting a similar polymorphism to
the Ecuadorian M. interrupta (Fig. 6E,F,G) and M. ignilineata (Fig.
7D,E,F) . The syntype specimen in the BMNH lacks any VHW jtost-
discal band, while the form in which the band is fully develo|)ed
flies with typical specimens (Adams, 1986) and was named form
curvilinea by Weymer (1912) (Lamas & Viloria, 2004). Although
we have not located any types of curvilinea, the desci iption clearly
applies to tliis form of M. apiculata, which cannot be confused with
any other Mnwcrcfoift species from the region of the type locality.

Comments: This species inhabits the forest/ paramo ecotone
on both slopes of the central Cordillera Oriental in Colombia
(Cundinamarca: Sibate, Fusagasuga, Facatativa), from 3150-3300
m, where it occurs with M. levana (Adams, 1986).

Manerebia franciscae (Adams & Bernard, 1981)

This species is most closely related to M. mam-
muthus n. sp., M. satura and M. navarrae, and all four
species are allopatric. The males of all of these species

are unique among north Andean Manerebia in having
a distinct dark brown patch of androconial scales
on the DFW^ in the posterior third of the discal cell
and basal part of cells 1A-Cn2 to M2-M1 (Figs. 2, 3).
The male genitalia are also similar, with an arched
linens, relatively long snbimci and valva with a spiny
dorsal process near the base, similar to M. leaena and
related species (Figs. 10, 11; see Discussion under
that species). Manerebia /rancbcrtc differs from the
two neighbouring species M. navarrae and M. mam-
muthus in having an almost smooth aedeagus, with
only a few spines on the left side. Other differences
are discussed under those two species. It differs from
M. satura in the narrower, distal process of the valva
being relatively longer, with the mid-ventral edge of
the valva more indented.

There is geographic variation in the presence or
absence of the pale VHW postdiscal band, as in M.
satura, and two subspecies are recognised.

Manerebia franciscae franciscae (Adams & Bernard,
1981)

Figs. 2D, E, IIB, 16

Penrosada franciscae Mhwws 8c Bernard (1981: 365, tigs 1 I, 25,
26). TL: Venezuela, Merida, above La Montana, S. of Merida. HT
male: BMNH(A&B) [examined].

Manerebia franciscae (Adams & Bernard); Lamas & Viloria
(2004:215).

Diagnosis: This subspecies is .sujjerficially most similar to M.
inderena inderena, with which it occtirs, btit the large tortial and
a|)ical ocelli on the VHW, with those in the middle of the wing
reduced, are diagnostic.

Comments: Contraiw to the sttitemeut of Adams & Bernard
(1981 ), this taxon is not restricted to the Venezuelan Cordillera de
Merida, but also occurs in the Sierra de El Tama and on the western
slopes of the Cordillera Oriental in Colombia (Racho, LWP, SMTD)
(Fig. 16). Adams & Bernard (1981) report that M. /. franckcae dies
with a skipping action, usuallv more than 2 m above the ground,
and rests on foliage. Eggs are laid on the brownish |3etioles of young
leaflets, near the nodes of Chu.»iuea bamboo canes, from 1-4 m
above the ground. The species was seasonally common at the type
locality, and Adams & Bernard ( 1981 ) collected it from 23(K)-26()()
m. We have also found the species to be seasonal, only observing
and collecting it in the Cordillera de Merida in the rainy .season
from June to September (Pyrcz, pets. obs.).

Manerebia franciscae rodrigiiezi Pyrcz & Willmott, n.
ssp.

Figs. 2F,G, 1 1C, 16

Diagnosis: This taxon differs from nominate At. franciscae pi i-
marily by the absence of VHW yellow postdiscal band. The apical
ocelli on the VHW are also reduced, the yellowish rings surrounding
the VHW ocelli are indistinct, and the dark, thin submarginal lines
on the VFW and VHW are slightly reddish.

Description: .MALE (Fig. 2F): Head, thorax And abdomen sAme
as in the nominate subspecies. Wings: Forewing (length: 21,5-22
mm; mean; 21,7; n=3) distal margin straight to slightly convex,
apex rounded; hindwing rounded with almost no notch at tornus.
DFW^ chocolate brown; darker brown androconial scales occupying
posterior one-third of discal cell, basal half of cells 1A-Cu2, basal
third of cell Cul-M3, and basal (|uarter of cells M3-M2 and M2-M I .

46

/. lies.Lepid.

DHW unitonii cliocolate l^rown; occasionally one small black sub-
marginal white oval clot ringed with black in Cul-Cu2 and two, even
smaller, in 1 A-(]u2. VFW ground colour medium brown becoming
jralei and lighter towards distal half; thin, irregtilar reddish brown
submarginal line, jrarallel to distal margin and very thin, reddish
brown marginal line jrarallel to outer margin, from apex to tornus.
VHW uniform medium brown; veiy faint, darker brown, straight
postdistcal line from apex to anal margin near tornus; uneven,
faint stibmarginal reddish blown line and reddish brown marginal
line; small, white, oval submarginal dots in cell Cu2-Cail, and two
in cell 1A-Cai2, ringed with black. Male genitalia (Fig. 1 1C); uncus
long and arched; subunci long and thin; numerctus small spines
on distal tip of valva extending anteriorly along dorsal edge, dorsal
base of valva with a projection with two ‘teeth’; aedeagus length of
valva, slightly curved.

FEMAL.Fl (Fig. 2G): .Slightly larger than male (forewing length
23 mm), with more prominent white dots on ventral surface oc-
casionallv taking shape of fulb develo]>ed ocelli, particularly in
cell Cul-M3 on VFW, showing through on dorsal surface, and
additionally with mintite submarginal white dots presetit in tnost
cells of fore and hindwing.

Types: Holotype male: COLOMBIA; Antioeptia, Guarne, 2600-
26,50 111, 14.111.1097, G. Rodriguez leg., MZL'J; Allotype female;
Aiitioqiiia, K1 Retiro, 1800 m, I9.V111.2002, G. Rodriguez leg., FVVP;
Paratypes (4 males and 2 females); COLOMBIA; 1 male; Antioquia,
no data, G. Rodriguez leg., GR; 1 male; Antioquia, El Retiro, Rio
Agtidelo, 2400 m, VIE 1999, G. Rodiiguez leg., GR; 1 male; Antio-
quia, El Retiro, 2800 ni, 24. VI. 2001 , G. Rodriguez leg., GR; 1 male;
Antior|uia, El Retiro, 2200-2600 m, 20.V111.2003, G. Rodnguez.
leg., MZUJ; 1 female; Atitioc]uia, Envigado, 2600 in, 29.XI.2002,
G. Rodriguez leg. GR; Antioquia, Santa Elena, Represa de Piedras
Blancas, 2600m, 1, 5. IX. 2003, Rodriguez leg., GR.

Etymology; Phis subspecies in named after its first collector,
Gabriel Rodriguez from Medellin.

Comments; Ellis taxon is associated with M. jranrisene based
on similar male genitalia and VI IW |)attern, with characteristic
stibmarginal white dots in cells Cu2-Gul and 1A-Cu2. Maneiebia
fmnchcaerodriguezih apparently endemic to the northern Cordillera
Central in Colombia, where it occurs in mid-elevation cloud forest,
similar to the nominate subspecies,

Manerebia mammiithus Pyrez & Willmott, n. sp.

Figs. 211, I in, 16

Manerebia n. s]t. (Pyrez, Willmott &■ Hall); Lamas & Viloria
(2004; 216, n, 1127).

Diagnosis; Phis sirecies is distinguished from its ckxsest relatives,
M. franci.vae and M. satura (see Discussion tinder M. franrisrae) , by
its large size, the virttial absence of any ventral stibmarginal ocelli
and characters of the male genitalia. The aedeagus has a large
patch of dense, tiny sjiine.s on both sides, while M. franei.'irae and
M. .satura have at most a single lateral line of spines on each side
or are smooth. 4 he narrower distal portion of the valva is relatively
sliorter than in M. franrheae, and the mid-ventral edge of the valva is
indented, like M. jranciscaehui not M. satura, in which it is flatter.

Description; MALE (Fig. 211); Head: Irons with a tuft of dark
brown hair; eyes blackish-brown, smooth; labial palpi covered with
blackish-brown hair; antennae dorsally chestnut, ventrallv riifotts,
with white .scales at the base of each segment, club dorsally slightly
darker than shaft. Thorax, dorsal and ventral surface dark brown;
legs paler yellowish-brown. Abdomen: dorsal and ventral surface
dark brown. Wings: forewing (length; 22.5-24 mm, mean; 23.3 mm,
11=2) distal margin convex, ajrex rounded; bindwing rotmded with
almost no notch at torntts. DP’W medittm brown; darker brown
androconial scales occtipying posterior half of discal cell, basal
half of cells lA-Gtil, basal third of cell Gul-M3, and basal quarter
of cell M,3-M2. DIIW medium brown, slightly darker towards base,
with pale |tostdiscal band from apex to torntts indistinctly show-

ing from ventral surface. VFW^ ground colour meditim brown,
becoming slightly paler from base to apex; veiy faint, thin, straight,
darker brown postdiscal line, parallel to distal margin, in cells lA-
R5; three whitish postdiscal dots in centers of cells Giil-Ml; thin,
faint, uneven, darker reddish brown submarginal line from apex
to tornus; margin thinly lined with darker brown. VHW meditim
brown; thin, straight, yellowish cream-colored postdiscal band from
costa to torntts through base of cell Ctil-M3; thin, faint, zigzag,
darker reddish brown stibmarginal line from apex to tornus; tiny
white stibmarginal dots in centers of cells 1A-M2. Male genitalia (Fig.

1 ID); uncus long and smoothly arching; subunci relatively long;
valvae thinning gradually at middle towards posterior tip, dorsally
grooved, with 4-6 ‘teeth’ at distal lip and several additional ‘teeth’
extending basally along inner etlge; thin, pointed projection at
dorsal edge of valva near base, with numerous tiny sjrines; aedea-
gus curving dorsally, with patches of nttmerotis short, posteriorly
directed spines laterally near middle, on both sides.

FliMAPT; Lhiknown.

Type.s; Holotype male: EGUADOR; Sucttmbios, km 9 La Bonita-
Ttilcan rd., El 1 ligtierdn, 2200 m, 10.XI.1997, K. Willmott leg., to be
deposited in AME; Paratypes (3 males); EGUADOR; 1 male; same
data as the holotype, KWJH; COLOMBIA; 1 male; same locality as
the holotvpe, 2000-2400 m, X.200L L Aldas leg., ex MBl.l, TW'P; 1
male; same data as preceding, MBLl.

Etymology; The specific name is a nottn in apposition derived
from the elephant genus Mamniuthus, with reference to this spe-
cies’ large size, almost uniform brown coloration and elongate
uncus and subunci, resembling the trunk and tusks of these extinct
creatures.

Comments; We deliberated as to whether to treat this taxon as a
sub.s|)ecies of either M. satura or M. franrisrae. However, the number
of consistent wing pattern and genitalic differences between these
three sjtecies suggest they should be maintained. In addition, it ap-
pears to occur at higher elevations than M. satura. This species is
known from onlv four specimens, all collected in far northeastern
Flcitador along the valley of the Rio Chingual along a trail through
secondai'y growth, with large tracts of tmdistiirbed forest within 50
111 distance, from 2000-24()() in. The two males collected by KW were
encountered on the same dav puddling on wet sand and feeding
on horse dung. Despite a inimber of other \isits to the same locality
the species has onlv been seen on tw'o occasions.

Manerebia satura (Weymer, 1911)

Fig.s. 3A, IIG

Lymancjpoda leaenavar. saturaWeymer (1911; pi. 52, rowt; 1912;
249). TL; Peru, Cuzco; Colombia, Tolima, Qtiindiu Pa,ss. ST male(s);
ZMHLl? [not examined].

Penrosada satura (Weymer); Brown (1944; 258) (stattis uncer-
tain).

Manerebia satura (Weymer); Lamas & Viloria (2004; 216).

This is a large species, closely related to the allo-
patric M. franriscae, M. mam.muthus and M. navarrae,
as discussed tinder M. franciscae. The male genitalia
are distinctive, with the distal, narrower portion of
the valva being relatively short, and often only bear-
ing a few large spines, rather than numerous smaller
spines. Unlike M. mammiithus, the aedeagus is smooth
or has only a single lateral line of spines. It is the
most widespread member of the genus and it is fairly
polytypic, occurring along the eastern slopes of the
Andes from Bolivia to Ecuador in lower to mid-eleva-
tion cloud forests.

Manerebia saturawas described based on specimens

39: 37-79, 2000 (200B)

47

from two localities, Cuzco (Peru) and Quindiu Pass
(Colombia), representing two distinct species. The
specimen figured in the original description repre-
sents the species as treated here, while those from
Colombia are actually M. wdcrcna Adams (see Adams,
1986) , which also has a white VHW band but is smaller,
with smaller VHW ocelli, and is not closely related to
M. satura. The syntype specimen (s) should be in the
ZMHU (G. Lamas, pers. comm.), but despite some
searching we have been unable to locate any. For
the present we do not designate either a lectotype or
neotype, since the type specimens may yet be found,
but our usage of the name preserves nonienclatural
stability, as should any future type designation. The
nominate subspecies, distinguished by the wide white
VHW postdiscal band and large tornal and apical
ocelli, particularly in cell Cu2-Cul (Fig. 3A), occurs
in southern Peru only (Cuzco, Puno, perhaps to
northern Bolivia) . An undescribed subspecies of M.
satura, characterised by a yellow VHW band, occurs
from central Peru (Junin, Pasco) to northern Peru
(San Martin, Amazonas) (Pyrcz, in prep.) . In extreme
northeastern Peru in the Cordillera del Condor, and
in eastern Ecuador, occur two distinct subspecies,
described below.

Manerebia satura lamasi Pyrcz & Willmott, n. ssp.

Figs. 3B, IIF, 16

Diagnosis: This subspecies is larger than the similar and neigli-
bouring M. s. paupmita (described below), darker brown on both
wing surfaces, with smaller ocelli on the VHW and a small ocellus in
cell M1-M2 on the VFVV'. There is no slightly paler brown postdiscal
line and the narrow, dark submarginal lines are less marked.

Description: M/\LE (Fig. 3B) : Head, thoraxiind abdomen as in M.
satura paiiperata. Wings: forewing (length: 22.5-23.5 mm; mean: 23
mm; n=3) distal margin straight, apex rounded; hindwing rounded
with very weakly pronounced notch at tornus. DFW' dark brown;
androconial scales occupying posterior half of discal cell, basal half
of cells lA-Cul, basal third of cell Cul-M3, and basal quarter of
cell M3-M2. DHW uniform dark brown, with two small suhmarginal
black ocelli thinly ringed with yellow in cells 1 A-Cu2 and Cu2-(ai; a
barely visible subniarginal blackish line. VYW ground colour dark
brown, slightly lighter along outer margins; tiny black submarginal
ocelli with white pupils in cells M2-M1, Cu2-Cul (and in M1-R5 in
one individual); faint, black submarginal line from tornus to apex;
two thin, straight, dark brown marginal lines. VHW uniform dark
brown; barely visible straight, thin, blackish postdiscal line from
tornus to apex, joining with a better marked, blackish submarginal
line that is smoothly curving and parallel to distal margin; black
submarginal ocelli, ringed with dark yellow, with white piqtils, as
follows: two small in cell 1 A-Cu2, one larger in cell Cu2-Cul (half
width of cell), one in cell M2-M1, one small in cell Ml-Rs, and tiny
white dots in cells Cu2-M3; thin, blackish marginal line close to
and parallel with distal margin. Male genitalia (Fig. IIF): uncus
long and arched; subunci relatively long; valva thinning gradually
throughout from base to posterior tip, dorsally grooved, with 6
‘teeth’ at distal tip; wide, toothed projection at dorsal edge of valva
near base; saccus shallow; aedeagus short, slightly curved dorsally
and smooth.

FEMALE; Unknown.

Types: Holotype male: PERU: Amazonas, Cordillera del Con-

dor, 2-3 km N PV3 (Alfonso Ugaite), 0345/7826, 1600-1750 m,
22. VI 1.1 994, G. Lamas leg., MUSM; Paratypes: 2 males: same data
as the holotype, MUSM.

Etymology: This subspecies is named for Gerardo Lamas, the
collector of the type series, in gratitude for all his help and cor-
respondence over many years.

Comments: Manerebia satura lamasi appears to be endemic to
the Cordillera del Condor. Other pronophilines that are also a])par-
ently endemic to this mountain range include Manerebia magnifica
n. sp. (described below) , which occurs in the same area but at lower
elevations, and an unde.scribed sjjecies of Panyapedaliodes Forster.
1964 (Viloria & Lamas, in pre]t.).

Manerebia satura pauperata Pyrcz & Willmott, n. ssp.

Figs. 3C,D, HE, 16

Euptychia jovita (C. & R. Felder); D’Abrera (1988: 778) (misi-
dentification).

Penrosada sp.: D’Abrera (1988; 824) (misidentification).

Manerebia satura n. ssp. (Pyrcz, Willmott & Hall); Lamas &
Viloria (2004: 216).

Diagnosis: Males of this sub.s|>ecies are distinguished from all
other M. 5«/Mra subspecies, except A/, .s. lamasi, by the absence of the
pale VHW postdiscal band. In females the band is well-developed
and chalky white, but specimens may easily be identified by the
large ocellus in cell Cu2-Cul that is visilrle on the dorsal surface
of bt)th wings. The characters that distinguish this subspecies from
A/, s. lamasi dre discussed under that taxon.

Description: MALE (Fig. 3C): Head: Irons with a tuft of dark
brown hair; eyes blackish-brown, smooth; labial palpi covered with
blackish-brown hair; antennae dorstilly chestnut, ventrally rufous,
with white scales at the base of each segment, club dorsally slightly
darker than shaft. Thorax, dorsal and ventral surface dark hrown;
legs paler yellowish-brown. Abdomen: dorsal and ventral surface
dark hrown. W7ngs: forewing (length: 19.5-22 mm; mean: 20.7 mm,
11=10) distal margin straight, apex rounded; hindwing rounded
with very weakly pronounced notch at tornus. DFW medium
brown; darker brown androconial scales occu|jying posterior half
of discal cell, basal half of cells lA-('uL basal third of cell Cul-.M3,
and hasal quarter of cell M3-M2. DHW medium brown, slightly
darker towards base, with two submarginal black ocelli ringed with
yellow in cells 1A-Gu2 and Gti2-GtiL VFW ground colour medium
hrown, hecoming slightlv paler from base to distal margin; thin,
straight, darker brown postdiscal line, parallel to distal margin, in
cells 1 A-R5; a tiny black subniarginal .s|K)t ringed with yellow in cell
Gu2-Gul; distinct, uneven, darker brown submarginal line from
tornus to apex, becoming more imdidate towards apex; two thin,
straight, dark brown marginal lines. VHW medium brown, darker
in basal half; distinct, thin, darker brown jiDstdiscal line from
torntis to apex, joining with a distinct, dat ker brown submarginal
line that is stnoothly curving and parallel to distal tnargin; black
subniarginal ocelli, ringed with dark yellow, with white pupils, as
follows: two small in cell 1 A-Cu2, one large in cell Cu2-Cul (almost
width of cell), one medium in cell M2-M 1 , one small in ceil M 1-Rs,
and tiny white dots in cells Cu2-M3: thin, dark brown marginal line
close to and parallel with distal margin. Male genitalia (Fig. HE):
uncus long and smoothly architig; subunci relatively long; valvae
thinning gradually throitghout from base to posterior tip, dorsally
grooved, with 4-6 ‘teeth’ at distal tip; thin, rounded projection at
dorsal edge of valva near base; aedeagus curving dorsally, with a
row of several short, posteriorly directed spines laterally near mid-
dle, on each side.

FEMALE (Fig. 3D): Differs from male as follows: larger (slightly
length: 22-23 mm; mean: 22.5 mm, n=2) paler brown throughout.
DFW lacking androconial scales, with large submarginal black,
yellow-ringed ocellus with white piqril in cell Cu2-CuL DHW with
ventral ocelli visible in cells 1A-Gu2, Gu2-Gul, Cul-M3, and M2-
Ml. VFW' with similar large ocellus in cell Gu2-Gul. with yellow

48

/. Res.Lepid.

ring extending into cells anterior and posterior, and tiny white
stibinarginal dots in cells ('.til-M 1 . V^HVV with broad, chalky white
postdiscal hand posterior to dark brown postdiscal line, broadest
in cells 1A-Cti2 and (in2-C'.nl, ocellus in cell Ml-Rs larger.

Types: Holotype male: EdUADOR: Zaniora-Chinchipe, Loja
- Zamora km 40,’ 1500 m, 31.V111.1990, P. Gros & S. Attal, MZUJ;
Allotype female: same data as the holotype, MZUJ; Paratypes (10
males and 2 females): EGUADOR: 1 male and 1 female: Zamora-
Chinchipe, Loja-Zamora i d., nr. Sabanillas, Qiiebrada San Ramon,
1700 m, 27-29. X. 1997, K. Willmott leg., KWJH; 2 males: Zamora-
Ghinchipe, Zamora-Loja rd., 1800 m, 9.X1.1996, K. Willmott leg.,
KWin {1),MEGN (1);1 female: Zamora-Chinchipe, Loja- Zamora
km 40, 1500 m, 01 .IX. 1990, P. Gros & S. Attal, TWP; 3 males: same
data as the holotype, TWE; 1 male: same data but 03. IX. 1990, TWP;
1 male: same data but 08. XI. 1990, TWP; 1 male, Zamora-Chinchipe,
West ofVallatlolid, 1800 m, 22.1V. 1997, A. Jasihski /r'g.,TV\'P; 1 male:
Zamora-Chinchi[te, Vtilladolid, 2.5.V11.1992, G. Estevez, MECN; 1
male: Timgnrahna, Machay, 17.X.I994, 1. Aldas leg., EWP

Etymology: Ehe name is the feminine form of the Latin adjec-
tive “pan|3erattis”, metming impoverished, with reference to the
absence of the ptile ventral hindwing prtstdiscal band in males, in
comparison with the nominate subspecies.

Comments: This taxon was illustrated twice by D’Abrera { 1988)
as Fenrosadd up. (p. 824) and as Euptyehin jovita (p. 778). It ajtpar-
ently occttrs from northeastern (Napo - one male in coll. TWP from
Cordillera de 1 Inacamtiyos, no date, A. Jasihski leg., Tnngtirahna)
to sontheastei It (Zamora-Chinchi|)e) Plctiador (Fig. 16), although
it is more common in the south, and has been collected in cloud
forest habitats in a narrow elevational band from 150()-18()() m.
The s])ecimen from Cordillera de los lliiacamayos is exchided
irom the type series since it shows slight dilferences to southern
specimens. We have found males puddling along forested streams,
and a single female flying low to the ground in a small field at the
forest edge.

Manerebid ruivarrae (Adams & Bernard, 1979)

Fig.s. 3E, 1 IH, 16

Penrosmla /t«w/nr«' Adtims & Bernard (1979: 1 14. tigs 11, 32).
TL: Colombia, Cesai, Serrania de Valledtipar, 2300 m. HT male:
BMNHfA&B) [examined].

Manerehia navamie (Adams & Bernard); Lamas & Viloria
(2004: 215).

Diagnosis: This s|3ecie.s sttperficially resembles several other
Manerebid that lack the jtale \'l IW postdiscal band, but may be dis-
tingthshed by the orange wedge extending basally from the VHW
postdiscal line at the tornus, and by the ocellus in cell Cu2-CuL
which is absent in tbe sttperficially similar M. quinterae. The male
genitalia (Fig. 1111) are distinctive in the dorsal process at the base
of tbe valva being more elongate than in other species, while the
thickened, strongly curved aedeagtis with a dense patch of spines
on the right side only is unique.

Comments: This sjDecies appears to be most closely related
to M. satiira, M. nunnniullni.s irnd M. franciscae, as discussed tinder
the last of these species. It is known only from the type locality in
the Serrania de Valledu|tar in the Sierra de Perija range on the
Colombia/Veneztiela border, at 2300 m. Adams & Bernard (1979)
state that the entire type series was caught before 10:00 hrs, flying
around a single patch of bamboo.

Manerehia quinterae (Adams & Bernard, 1979)

Figs. 3F, 111, 16

Penrosddd quinterae Adiims & Bernard (1979: 115, figs 12, 33,
34). TL: Venezuela, Ztilia, Serrania de Valledtipar, 3050 m. HT
male: BMNH(A&B) [examined].

Manerehia quinterae (Adams & Bernard); Lamas & Viloria
(2004: 215).

Diagnosis: The ventral stirface of M. quinterae (Fig. 3F’) some-
what resembles that of M. navarrae, but M. quinterae may be easily
distinguished by the irregular dark postdiscal line, lack of any well
developed submarginal ocelli, and instead a stibinarginal line of
whitish spots in cells Cti2-ML The male genitalia (Fig. Ill) are
distinctive in lacking a spiny dorsal process at the base of the valva,
unlike all |3receding species, but notably have an arched uncus,
like many of the preceding species. The affinities of this species
are therefore unclear. The small, lateral patches of spines on both
sides of the aedeagtis near its posterior tip are distinctive.

Comments: Adams & Bernard (1979) reported that this species
occurred f rom 2750-3050 m in the Serrania de Valledtipar, where
it could be locally and seasonally common. It is known to date only
from the type locality.

Manerebia inderena (Adams, 1986)

The male genitalia of this species do not differ from
those of M. golondrinan. sp., which can be immediately
recognised by the lack of a VHW postdiscal yellow or
white band, and are also similar to M. pratlorum n. sp.
These three species can be distinguished from other
species by the simple valva (lacking a prominent spiny
process at the dorsal edge near the base), the uncus
which is strongly “bent” near the base (almost straight
in M. undulatan. sp., M. trimarulat a and M. interrupta) ,
then approximately straight (smoothly curving in M.
rufanaUs), and in usually liaving several tiny ‘teeth’ on
the lefthand side of the aedeagus. The VHW narrow
dark submarginal line is irregularly zigzag, but never
as undulate as in M. undidata or parallel to the distal
margin as in M. germaniae.

This is a widespread and polytypic species. The
various taxa are grouped together on the basis of
similar genitalia, habitat and elevation, and close
range allopatiy Reasons for considering M. golondrina
a distinct species are discussed under that species.

Manerebia inderena inderena (Adams, 1986)

Fig.s. 3G,H, 12A, 17

Penrosada inderena Xd'Amf, (1986: 305). TL: Colombia, Tolima,
south above Cajamarca. HT male: BMNH(A&B) [examined].

Manerebia inderena (Adams); Lamas & Viloria (2004: 215).

Diagnosis: The nominate subspecies (Fig. 3G,H) is character-
ized by medium wide, whitish VHW postdiscal band, that is yellow
and usually thinner in M. i. antiuquiana, white and much thinner
in M. i.Jina, whitish and wider in M. i. clara, wider and yellow in M.
i. mirena, M. i. leaeniva and M. i. similis. The nominate subspecies
and M. i. antioquiana may also be distinguished from other similar
sjiecies (M. germaniae, M. leaena, M. franciscae) by the three marked
submarginal ocelli in cells 1A-Cu2 and Cu2-Ctil only. Male genitalia
as illustrated (Fig. 12A).

Comments: Adams (1986) stated that this taxon cotild be found
skipping around clumps of bamboo, resting on the foliage, and
occasionally descending to feed on excrement on the ground. In
Colombia it has been found in the Cordillera Central in Tolima
and Cauca (Volcan Purace) where it occurs from 2450-3100 m
(Pyrcz, 1999) (Fig. 17). It is locally sympatric with M. germaniae n.
s]). (described below) at the highest reach of its altitudinal range.
Specimens of M. inderena from extreme northern Ecuador on the
eastern slopes of the Andes (Sucumbi'os, Carchi) are considered

39: 37-79, 2000 (2006)

49

for the |)resent as representing tlie nominate snbs|)ecies, although
in general they are sliglitly larger, with a slightly broader pale VHW
band, and generally reduced VHW ocelli, in some specimens ap-
proaching M. i. fina.

Manerebia inderena antioqidana Pyrcz & Willmott n.
ssp.

Figs. 4A,B, 12B, 17

Manerebia inderena n. ssp. (Pyrcz, Willmott & Hall); Lamas &
Viloria (2004: 21.5).

Diagnosis: This subspecies has yellow VHW bands, instead of'
white as in the nominate subspecies, and is typically smaller than
the latter.

Description: MALE (Fig. 4A): Head: eyes, labial palpi and
antennae same as in the nominate stibspecies. Thorax, dorsal and
ventral surface dark brown; legs paler brown. Abdomen: dorsal and
ventral surface dark brown. Wings: forewing (length: 16.5-18.1
mm, mean: 17.3 mm, n=4) distal margin straight, apex rounded;
hindwing rotinded with veiy weakly pronounced notch at tornns.
DFW' meditim brown; darker brown in discal area. DHW medium
brown, slightly darker towards base, with two snbmarginal black
ocelli ringed with dark orange in cells 1A-Cti2 and Cn2-CuL VFW^
ground colour medium brown, becoming slightly paler from base
to apex; thin, straight, darker brown postdiscal line, parallel to
distal margin, in cells 1A-R5; a tiny black snbmarginal spot ringed
with dark orange in cell Cu2-(ai1; distinct, zigzag, darker brown
snbmarginal line from torims to apex; thin, straight, dark brown
marginal line. VHW medium brown; pale yellowish postdiscal band
from apex to tornns, passing through base of cell Cul-M3; distinct,
darker brown, strongly zigzag submarginal line; tw'o small black
submarginal ocelli, ringed with dark orange-brown, with white
pupils, in cells 1A-Cu2 and (ai2-Cnl; thin, veiy faint, dark brown
marginal line close to and parallel with distal margin, Male genitalia
(Fig. 12B): similar to nominate subspecies.

FEMALE (Fig. 4B): Similar to male, except VHW postdiscal
band whitish, instead of yellow, and ventral ocelli better marked.

Types: Holotype male: COLOMBIA: 1 male: Antioquia, San
Felix, Las Antenas, 2700-3050 m, 13.1X.2()03, T. Pyrcz leg., MZUj;
Allotype female: COLOMBIA: Antioquia, Medellin, El Retiro,
Reserva San Sebastian, 2500-2800 m, 12. IX. 2003, T. Pyrcz leg., TWP;
Paratypes ( 14 males): COLOMBIA: 1 male: Antioqtiia, Medelh'n, El
Retiro, Reseiwa San Sebastian, 2500-2800 m, 12. IX. 2003, T. Pyrcz
leg., TWP; 1 male: same locality, 2600m, 09.XL2002, C. Rodriguez
leg., TWT; 1 male: same locality, 2300-2800 m, 1 8-25. VII. 2003,
C. Rodriguez leg., TWP; 2 males: same locality, 2600-2800 m,
20.Xn.2002, C. Rodriguez leg, (1 TW'P, 1 BMNH); 1 male: same
locality, 2600-2800 m, 26-30.VI1L2003, G. Rodriguez leg, TWP; 1
male: same locality, 2200-2600 m, 20.V111.2003, C. Rodn'gtiez leg.,
TWP; 2 males: Antioquia, Medelh'n, El Retiro, 2700m, 0LV11I.1993,
J-F. Le Crom leg, TW'P; 1 male: Antioquia, Polmifor, 24.V11,200L C.
Rodriguez leg, 'TWP; 1 male: Antioquia, Envigado, 2600-2800 m,
06. VI. 2004, G. Rodriguez leg., TWP; 1 male: Antioquia, Los Llanos,
via a San Andres km 10-14, 2600-2750 m, 14.XI.2003, T. Pyrcz leg,
TW'P; 2 males, Antioquia, El Retiro, 2600-2700 m, 18.X11.2001, G.
Rodrfguez leg., MBLl.

Etymology: This taxon is named after the Colombian depart-
ment of Antioquia, where the type locality is situated.

Comments: To date this taxon is known only from the northern
part of the Colombian Cordillera Central, from 2500-2800 m, where
it appears to be locally not tmcommon.

Manerebia inderetta fina Pyrcz & Willmott, n. ssp.

Eigs. 4C,D, 12C, 17

Manerebia n. sp., n. ssp. (Pyrcz, Willmott & Hall); Lamas &
Viloria (2004: 216, n. H24b).

Diagnosis: This subspecies differs from other Ecuadorian sub-

sjtecies by having a narrower VHW postdiscal band. Along with M.
i. simili.s, it may also be distinguished from all other subspecies by
the dark, thin snbmarginal line on the VtlW being more sinoothh
curving, rather than dentate as in other taxa. The nominate sub-
species and M. i. anlioqHiana also differ in having marked VHW
snbmarginal ocelli, especially in 1A-Cu2 and Cu2-Cul, that are
usually absent or much reduced in A/, i.fina. The w'idth of the band
in M. i.Jinais about the same as in M. gennaniae germaniae, which is
locally sympatric, though generally occurring at higher elevations.
The latter is distinguished externally with diffictilty, by several sub-
tle characters (see under A/, gennaniae). Another similar (though
not .sympatric) taxon is the Colombian M. leaena lanassa. Both M.
leaena And M. gennaniaeiu c most reliably distinguished by the male
genitalia, which have a curving iincits, relatively long subnnci and
‘teeth’ at the dorsal edge of the valva near the base.

Description: MAl.E (Pig. 4C): Head: frons with a tuft of dark
brown hair; eyes blackish-brown, smooth; labial palpi covered with
blackish-brown hair; antennae dorsally chestnut, ventrally rtifotis,
with white scales at the base of each segment, club dorsally slightly
darker than shaft. Thorax-, dorsal and ventral surface dark brown;
legs paler yellowish-brown. Abdomen: dorsal and ventral surface dark
brown. Wings: forewing (length: 18.5 mm, mean: 18.5 mm, n=2)
distal margin almost straight, a|)ex rounded; hindwing with distal
margin rounded, with tornal notch almost absent. DFW' medium
brown, darker brown towards base. DHW’ medium brown. VFW'
ground colour medium brown, basal half slightly darker, bordered
distally by very thin, indistinct, darker brown postdiscal line, that
is slightly inclined towards apex, and curves slightly distally on
approaching costa; faint, slightly wary, darker brown snbmarginal
line from tornns to apex; veiy thin, straight, dark hrown marginal
line. VHW’ medium brown; thin (r. 1.5 mm) veiy pale yellowish
(almost white) postdiscal band from apex to tornns, veiy slighth
concave, passing through base of cell C;ul-.M3; faint darker brown
submarginal line, smoothly curving from tornns to cell (ail -M3,
then slightly undulate to apex; white submarginal dots in cells
Cu2-Cul and Cul-M3; thin, dark brown marginal line close to and
parallel with distal margin. Ma/egetitta/ta (Fig. I2G): uncus curving
slightly ventrally near middle, and bent more sharply near base,
snbunci of medium length; valvae thinning sharply at middle and
tapering posteriorly, dorsally grooved, with 5-7 ‘teeth’ at distal tip;
aedeagns curving dorsally, with a couple of tiny ‘teeth’ on left hand
side near middle.

FEMALE (Fig. 41)): Differs from male as follows: slightly larger
(forewing length: 19 mm, n=l). Both wing surfaces paler brown.
VHW’ postdiscal band wider.

Type.s: Holotype male: ECU.\DOR: Fichincha, Aloag-Tandapi
rd., km 18, Sector Los Alpes, 27()l)-275() m, 26.1.2004, T. Pyrcz & R.
Garlacz leg, MZL[J; Allotype female: Fichincha, Volcan Pasochoa,
nr. Amaguaha, 3500 m, 7.X.1997, K. W'ilhnott leg, KW'fH: Paratypes
(23 males): ECL’ADOR: 4 males: Fichincha, Reserva Geoho-
tanica Fuhilahua, 2300-2600 m, 11-12.11.2002, T. Pyrcz leg. TW’P;

1 male: Fichincha, Chilk) Gallo, San Juan-La Victoria, 3300-3400
m (unreliable), 30.1.2002, T. Pyrcz leg., 'TV\'P; 1 male: Pichincba,
Volcan Fichincha, 3000-3050, X.2002, 1. Aldas leg, TW’P: 2 males:
Fichincha, Aloag-Tandapi rd., 2800 m, 11.2002, 1. Aldas leg. FW’P;

1 male: Fichincha, Nono - Nanegalito km 15/20, 2100 m, F. Boyer
leg, MZUJ; 1 male: Fichincha, Volcan Pasochoa, 3000-3200 tn,
07. X. 1997, A. Neild leg, FW'P; 1 male: Fichincha, Volcan Pasochoa,
2600-2700 m, 28.1.2002, T. Fvrcz leg., TW'P; 5 males: Fichincha,
/Vloag-'Eandapi rd., km 1 8, Sector Los .\lpes, 2700-2750 m, 26.1.2004,
T. Pyrcz & R. Garlacz leg, TW'P; 1 male, same data, FB; 1 male,
Fichincha, Nono-Nanegalito, 2700-3000 m, 10.V.I999, F. Boyer
leg., PB; 1 male: Fichincha, environs de Nono, 2600m, 15.111. 1998,
F. Boyer leg, PB; 1 male: Fichincha, Nanegalito, 1800 m, IX. 1996,
P. Boyer leg, FB; 1 male: Fichincha, km 26 Nanegalito-Qnito rd.,
Quebrada Molino, 2400 m, 16.X.I996, K. W’ilhnott leg., KW'[H: 1
male, Fichincha, km 7 Aloag-Tandapi rd., 2500 m, 13.V1.1994, [.

50

/. Res.Lepid.

Fig. 1 . Adult Manerebia, left dorsal view, right ventral view. Double arrows above names indicate image is reflected in vertical
plane, so figured wings are righthand pair. A, M. leaena leaena male; B, M. leaena leaena female; C, M. leaena lanassa male;
D, M. leaena lanassa female; E, M. leaena gonzalezi n. ssp. male; F, M. germaniae germaniae n. sp. male; G, M. germaniae
germaniae n. sp. female; H, M. germaniae vitalei n. ssp. male. See Appendix 4 for specimen data.

39: 37-79, 2000 (200(i)

51

M. apiculata

M. pluviosa

M. apiculata

M. franciscae franciscae

M. franciscae rodriguezi

M. mammuthus

Fig. 2. Adult Manerebia, left dorsal view, right ventral view. Double arrows above names indicate image is reflected in vertical
plane, so figured wings are righthand pair. A, M. pluviosa n. sp. male; B, M. apiculata male; C, M. apiculata female; D, M.
franciscae franciscae male; E, M. franciscae franciscae female; F, M. franciscae rodriguezi n. ssp. male; G, M. franciscae
rodriguezi n. ssp. female; FI, M. mammuthus n. sp. male. See Appendix 4 for specimen data.

M. franciscae franciscae

M. franciscae rodriguezi

52

/. Rps.Lf'pid.

1 lall li% MEC.N; 1 male: Pichinclia, kin 5 /Vloag-Tandapi rd., 27()()ni,
13.VI.I994, |. Hall % KVV;|I1:

Etymology: The name is tlie leiiiitiine form of the Latin adjec-
tive “fiiitis”, meaning narrow, with reference to the thin jtostdiscal
band on the VI IW.

Comments: Tliis subspecies occurs on tlie w'estern .slo|tes of tlie
.Viides in northern Ecuador (Pichincha, Imbahura) and possibly
also in southern Colombia (Nariho). It has been recorded from
2100-3200 m, and once at 3500 m (Volcan Pasochoa), though it
is most common between 2200-2600 m. It is generally replaced at
higher elevations by A/, ffrtnmiiae. Males are often encotuUered
]juddling along roads or streams, and are also attracted to rotting
fish. We have also found males hilltopping on Loma La Palmira,
flying low’ over the stunted bushes growing on the stunmit during
bright sunlight, after 09:30 hrs.

Manerebia inderena similis Pyrez & Willmott n. ssp.

Figs. 4E,F, 121). 17

Manerebia n. sp., n. ssjt. (Pyrez, Willmott & Hall); Lamas &
Viloria (2004: 216, n. 1 124c).

Diagiio.sLs: fhis subspecies differs from the neighbouring M. i.
fina to tbe north by the broader, pale \'HW postdi.scal band which
is more strongly yellow especially towards anal margin. It is very
similar to the east Andean A/, i. leaeniva, but both west Andean
subspecies (A/, i. similis mu\ M. i. fina) are distinguished from it
and all others by the thin, dark submarginal line on the VHW be-
ing in most examined individuals more smoothly curving, rather
than dentate.

Description: MAl.F. (Fig. 4E) : Head: eyes, labial jtalpi and anten-
nae as in nominate stibspecies. Thorax, dorsal and ventral surface
dark brown; legs palei brown. Abdomen', dorsal and ventral surftice
dark brown. Wings: forewing (length: 19-20 mm; mean: 19.3 mm;
11=3) distal margin almost straight, apex rounded; hindwing with
distal margin rounded, with tormil notch almost absent. DFW itiedi-
um brown, darker brown towards base. DHW medium brown. VFW
ground colour medium brown, basal half slightly darker, bordered
distally by very thin, indistinct, darker brown postdiscal line, that
is slightly inclined towards apex and curves slightly distally on ap-
proaching costa; faint, slightly wavy, darker brown stibmarginal line
from tornus to apex; veiw thin, straight, dark brown marginal line.
VHW medium brown; broad (r. 2.5 mm) yellowish (more intense
towards tornus) postdiscal band f rom ajjex to toriuts, straight and
of even width exce|tt tapering slightly at tornus. passing through
base of cell (ai I-M3; faint, uneven darker brown submarginal line,
more undulate in cells Cu2-Cu 1 to M3-M2; white submarginal dots
in cell (ail -M3; tiny black submarginal ocellus, ringed with dark
orange, with white pujjil. in cell (ai2-(ail. a black spot in anterior
half of cell lA-(ai2; thin, dark brown marginal line close to and
parallel with distal margin. Male genitalia (Fig. 12D): uncus curving
slightly ventrally near middle, and bent more sharply near base,
subunci of medium length; valvae thinning sharply at middle and
tapering posteriorly, dorsally grooved, with 5-7 'teeth' at distal tip;
aedeagus curving dorsally, with a couple of tiny ‘teeth’ on left hand
side near middle.

FEMAf.E (Fig. 4F): Similar to male but slightlv paler brown
ventrallv.

Type.s: Holotyfie male: EOLADOR: Bolivar, Balzapamba, ar-
riba de Santa Lucia, 2206-2250 m, 03. IX. 2003, T. Pyrez leg., MZUJ;
Allotype female: same data as the holotype, TWP; Paratypes (34
males and 3 females): ECiUADOR: 3 males: Bolivar, Balzapamba.
arriba de Santa Lucia, 2600-2650 m, 03. IX. 2003, T. Pyrez /eg., TWP;
12 males: Bolivar, Balzapamba, arriba de Santa Lucia, 2200-2250
m, 03. IX. 2003, T. Pyrez leg., FWP (II), BMNH (2): 5 males and
1 female: Bolivar, Balzapamba, arriba de Santa Lttcfa, 2200-2250
m, 05. IX. 200 1, T. Pyrez. leg, FWP; 1 male: same locality and date,
2400-2450 m, T. Pvrez. leg, 'FWP; 1 male: .same locality and date,
2500-2550 m, T. Pyrez leg, IW'P; 2 males: Bolivar, Balzapamba, Rio

Alcacer, 2700m, 04. XL 1996, S. Altai kg, MZUJ; 1 male: Bolivar, old
Uuaranda road, VIII. 1997, I. Aldas leg, TWP; 3 males: Uotopaxi,
above Pilalo, 3000-3050 m, 03. IX. 2004, T. Pyrez leg., TWP; 6 males
and 2 females: Azuay, Uuenca - Naranjal road, Molleturo, 2600-
2650 m, 01 .IX.2003, T. Pyrez leg, TW'P.

Etymology: The name is derived from the Latin “similis”,
meaning similar, with reference to this taxon’s resemblance to
the subspecies M. i. leaeniva, which occurs on the east slojtes in
central Ecuador.

Comments: I’liis suhs|tecies occurs in central western to
southwestern Ecuatlor, where it has been recorded above Pilalo
(Colo|taxi), .Santa Lucia (Bolivar) and on the western .slo|>e.s of
the Cajas massif near Molleturo (Azuay) from 2200-3000 m. In all
these localities A/, i. similis \s the upper parapatric replacement of
A7. nndnlata n. sp.

Manerebia inderena clara Pyrez & Willmott, n. ssp.

Figs. 4G,H, 12E, 17

Manerebia n. sjt., n. ssp. (Pyrez, Willmott & Flail); l.amas &
Viloria (2004: 216, n. I124d).

Diagnosis: I his subspecies differs from all others by the broad,
pure whitish VI IW postdiscal band, which is yellow in neighhoring
A/, i. leaeniva to the .south and yellowish posteriorly and tajiering
anteriorly in the similar A7. i. similis from western Ecuador. The
thin, dark submarginal line on the VHW is noticeably dentate,
slightly more so than in M. i. leaeniva 'dnd noticeably more so than
in M. i. mirena.

Description: MALE (Fig. 4G): Head: as in the nominate sub-
species. Thorax, dorsal and ventral surface dark brown; legs paler
brown. Abdomen: dorsal and ventral surface dark brown. Whigx.
forewiiig (length: 17. .5-19 mm; mean: 18.2 mm; n=19) distal margin
slightlv convex, apex rounded; hindwing with distal margin slightly
angled at vein M3, with tornal notch almost absent. DFW medium
brown; darker brown in discal area. DHW medium brown, slightly
darker towards base, with small black submarginal ocellus ringed
with dark orange in cell Cu2-CuL VFW- ground colour medium
brown; thin, very indistinct, sligbtly curv'ed, darker brown |30stdis-
cal line, in cells Gu2-M2; a minute white submarginal dot in cells
Gul-M3; indistinct, zigzag, darker brown submarginal line from
tornus to apex; distal margin lined indistinctlv with dark brown.
\'I IW medium brown; whitish postdiscal band from apex to tornus,
passing through base of cell Cul-M3, straight and of even width,
except ta|)ering slightly in cell 1 A-Cu2; faint, darker brown, strongly
zigzag submarginal line; a small black submarginal ocellus, with a
white |5upil, faintly lined with dark orange, in cell Gii2-Cul, with
minute white submarginal dots in cells Ciil-M3 and M.3-M2; distal
margin lined indistinctly with dark brown. Male genitalia (Fig.
12F)): uncus bent near base then flat, subunci of medium length;
valvae thinning sharply at midtlle and tapering posteriorly, dorsally
grooved, with 4-6 ’teeth’ at distal tip; aedeagus ciining dorsally, with
a couple of tiny ’teeth’ on left hand side near middle.

F’EMALE (Fig. 4H) : Differs from male as follows: larger (forew-
ing length 20.5 mm) ; ventral surface paler brown, with dark .submar-
ginal lines more distinct; a large ocellus in cell Cu2-Cul on VFW.

Types; Holotype male: ECUADOR: Napo, Baeza, Rio Florituy-
acu [OritoyacuJ, 1800 m, 08.VI.1999, T. I^tcz & J. Wkijtusiak leg.
MZUJ; AZ/o/y/te female: ECUADOR: Napo, Hda. San Isidro, 2000 m.
18. XII. 1996, P. Boyer leg, PB; Paratypes (31 males): ECUADOR: 2
males: same data as the holotype, TWP; 1 male: Napo, Quito-Baeza
rd., east of pass, 2300 m, 17.VI.1994, J. Hall leg, KVyjH; 1 male:
Napo, Baeza - Papallacta, 2100 m, 07.IAa998, A. Neild kg, IWP;
1 male: Napo, SE of Cosanga, Rfo Chonta, 2000 m, 18. X. 1996, K.
Willmott kg, KWJH; 1 male: Napo, Baeza - Papallacta, 2100 m,
07. IV. 1998, A. Neild kg, FWP; 5 males: Napo, Baeza area, 2050-2200
m, 02.x. 1995, A. Neild kg, TWT (4), MECN (1); 2 males: Napo,
Baeza, 2000-2200 m, 19. X. 1996, A. Neild kg. TWP; 7 males: Napo,
Baeza, 1800 m, IX. 1996, P. Boyer kg, TWP (5). BMNH (1), MZLJ

39: 37-79, 2000 (2006)

53

(1); 1 male: Napo, Baeza, 1800 ni, 10. XII. 1996, P. Boyer leg., rWP;
1 male: Napo, San Isidro, 2000 ni, 08. XII. 1996, P. Boyer hg., TWP;

1 male: same data but 18. XII. 1996, TWP; 1 male: Napo, Closanga,
1600 m, 06.XI.1996, P. Boyer kg., PB; 1 male: Napo, Baeza - Tena
km 19, 2100 m, 04.XII.1997, P. Boyer leg, TWP; 3 males: Napo,
Reserva Yaiiayacu, 2100-2150 m, 06-07. IX. 2003, T. Pyrcz leg, TWP;
3 males: same locality, no date, H. Greeney leg., MBLI.

Etymology: The name is derived from the Latin “darns”, mean-
ing pale, with reference to the pale VHW postdiscal band.

Comments: Manerehia inderena clam has been recorded only in
the Baeza area, in the valleys of the Rios Papallacta and Cosanga,
sotith to the Cordillera de los Htiacamayos, from 1700-2400 m. It is
not uncommon, and males can be found puddling along mountain
streams in cloud forest, or feeding on horse dung or rotting lish.

Manerebia inderena leaeniva Pyrcz & Willmott, n. ssp.

Figs. 5C, 12F, 17

Manerebia n. sp., n. ssp. (Pyrcz, Willmott & Hall); Lamas &
Viloria (2004: 216, n. 1 124a).

Diagnosis: The yellow VHW band is twice as wide in this sul>
species as in the nominate, and the VHW stibmarginal ocelli are
reduced, similar to AT i.Jiria, M. i. iiw/fo and M. i. clam. In AT i.
clara the VHW postdiscal band is as wide as in AT i. leaeniva, but
white, while in AT i. similis from the west Andes it is as wide hut a
more intense yellow, and the VHW submarginal line is smoothly
undulating rather than dentate. The southern subspecies M. i.
mirena has a VFV\^ ocellus in cell Cu2-Cul, two to three ocelli in
cells 1A-Cu2, Cu2-Cul on the VHW and an ocellus in cell Cu2-Cul
on the DHW.

Description: MALF. (Fig. 5C): Head: frons with a tuft of dark-
brown hair; eyes blackish-brown; labial palpi covered with dark-
brown hair; antennae dorsally pale brown, ventrally light beige,
club laterally rufous, white scales at base of each segment. Thorax.
dorsal and ventral surface dark brown; legs paler brown. Abdomen:
dorsal and ventral surface dark brown. Wings: forewing (length:
17-20.5 mm; mean: 19 mm, n=ll) distal margin slightly convex,
apex rounded; hindwing with distal margin slightly angled at vein
M3, with tornal notch almost absent. DFW' medium brown; darker
brown in discal area; androconial scales not apparent. DHW me-
dium brown, slightly darker towards base, with two minute white
submarginal dots in cells Cu2-Gul and Cul-M3. VFW ground
colour medium brown, becoming slightly paler from base to apex;
thin, indistinct, slightly curved, darker brown postdiscal line, in
cells Cu2-M2; a row of minute white submarginal dots in cells
Cai2-Cu1 to M2-M1; indistinct, zigzag, darker brown submarginal
line from tornus to ajrex; very thin, straight, dark brown marginal
line. VHW medium brown; pale yellowish postdiscal band from
apex to torntis, passing through base of cell Gul-M3, straight and
of even width; distinct, darker brown, strongly zigzag submarginal
line; a small black submarginal ocellus, with a white pupil, in cell
Cai2-CuT, tiny white submarginal dots in cells Cul-Ml; thin, very
faint, dark brown marginal line close to and parallel with distal
margin. Male genitalia (Fig. 12F): uncus flat and bent near base,
suhunci of medium length; valvae thinning sharply at middle and
tapering posteriorly, dorsally grooved, with 4-6 ‘teeth’ at distal tip;
aedeagus curving dorsally, with a couple of tiny ‘teeth’ on left hand
side near middle.

FEMALE: Unknown.

Type.s: Holotype maie: F.CUADOR: Tungurahua, Chinchin, 2000
m, 06.11.2004, T. Pyrcz kg, MZUJ; Paratypes (15 males): ECUAFXDR:

2 males: Tungurahua, Rio Verde Chico, 2100 m, 08.X.1995, A.
Neild kg., TWP; 2 males: same data as preceding but 24. IX. 1995,
TWP; 2 males: Tungurahua, Banos, Runtun, A. Jasinski kg., TWP;
2 males: Tungurahua, Viscaya, 2100 m, 17. XI. 1996, P. Boyer kg,
'TWP; 5 males: Ttingurahua, Triunfo - Patate, El Tablon, 3000m, P.
Boyer kg, TWP (2), PB (3); 1 male: Tungurahua, Banos, El Tahlon,
3000m, III. 1999, 1. Aldas kg, MBLI; 1 male: Tungurahua, Runtun,

2900-2950 m, 22.1. 2002, J.WoJiusiak kg, TWP.

Etymology: The subspecific name is derived from “leaena”,
the name of the species with which this taxon has been most ofteti
confused.

Comments: This subspecies is known from the upper valley of
the Rio Pastaza (Titngurahita) soitth to Morona-Santiago, where
it has been recorded from 2100-3000 m, although it is more com-
mon in the lower part of this elevational range. Specimens from
Morona-Santiago (Gualaceo-Limon road, via Las Chacras, 2600-
2850 m, TWP, KW|H) apparently represent this subspecies, bttt
are excluded frotn the type series hecause we have been unable to
examine sufficient tnaterial to reliably assess variation.

Manerebia inderena mirena Pyrcz & Willmott n. ssp.

Figs. ,5A,B, 12G, 17

Diagnosis: This subspecies generally has a slightly wider VHW
postdiscal band than AT. /. leaeniva, a slightly tnore reddish brown
ground colour, and promitient ocelli, in most individuals, iti cell
(’At2-Cul on the DHW and occasionally on the VFW and VHW itr
cell Cu2-Cul. The uncus in AT i. mirena is usitallv more noticeably
ctirved ventrally in the middle than iti rtther subspecies, and the
saccus is distinctive iti always being swollen anteriorly.

Description: MALE (Fig. .5A): Head: frons with a tuft of dark
brown hair; eyes blackish-brown, smooth; labial palpi covered with
blackish-brown hair; antetinae dorsally lirown, ventrally dirty yellow,
club slightly darker than shaft. Thorax, dorsal and ventral surface
dark brown; legs yellowish-brown. Abdomen: dorsal and ventral sur-
face dark brown. Wings: forewitig (length; 18-19.5 mm; mean: 18.6
mm; n=3) distal margin slightly convex, apex roitnded; hindwing
with distal margiti rounded, with tornal notch almost absent. DFW
medium brown; darker browti iti discal area. DHW mediitm brown,
slightly darker towards base; a mitmte white sitbmarginal dot in cell
Gu 1-M3, a small black ocellus ritiged with dark orange with a white
jiupil in cell 1 A-Gu2, a larger sitnilar ocelhts in cell (’At2-Cul . VFW
groitnd coloitr meditttn brown, liecotning slighth jialer from base
to apex; thin, indistinct, straight, darker brown postdiscal line iti
cells ( At2-M2: a row of minute white submarginal dots in cells ( ai2-
Gttl to M2-M1; in most individuals the dot in Cu2-Gttl is replaced
by a black ocellus, ringed with orange and with a white pupil, of
variable size; indistinct, undulate, darker brown sitbmarginal line
from tornus to apex; veiy thin, straight, dark brown marginal line.
\'HW medium brown; a wide pale yellowish (tnore inten.se at tornns
and costa) postdiscal batid from apex to tornus, passing through
base of cell Cul-M3, straight and of even width; tnost individuals
have one or two simill black siibtnarginal ocelli, with white jmpils,
in cell 1A-Cu2, and occasiotially a larger ocellus in cell Cu2-Gttl,
and white submarginal dots in other cells; darker brown, zigzag
siibtnarginal line; thin, dark brown marginal line close to and
parallel with distal margin. Male genitalia (Fig. 12G): uncus curving
slightly ventrally near middle, and bent more sharply near base,
subunci of medium length; valvae thinning sharply at middle and
tapering posterioi ly, dorsally grooved, with 4-6 ‘teeth’ at distal tip;
aedeagus curring dorsally, with a couple of tiny ‘teeth’ on left hand
side near middle.

FEMALE (Eig. 5B): Similar to male but paler with a fainter
|)attern on both dorsal and ventral wing surfaces.

Types: Holotype male; EGLIADOR: Zamora-Ghinchipe, Val-
ladolid, Quebracia de los Muertos, 2550 m, XI. 1999, I. Aldas kg,
MZUJ; A//ofype female: EGUADOR: L.oja, Loja-Zumba, km 9,5-l()0,
2500-2600 m, 27. XI. 1998, R Boyer kg., PB; Paratypes (59 males and
1 female): EGLIADOR: 2 males: Zamora-Ghinchipe, Valladolid,
no other data, P. Boyer kg, f’B; 2 males, Zamora-Ghinchijte, km
34 jimbura-San Andres rd., 2900 m, 23. IX. 1997, K. Willmott leg.,
RWJH (1), MEGN (I); 41 males: Zamora-Ghinchi|5e, Valladolid,
Quebracia de los Muertos, 2550 m, XI. 1999, I. Aldas kg, TWP (8),
MBLI (33); 2 males; Zamora-Ghinchipe, L()ja - Zamora 2600m,
22.XI.1996, P. Boyer kg., PB; 1 male: l.oja, Loja-Zumba, km 95-100,

54

/. Res.Lepid.

M. satura satura

ftf. satura lamasi

M. satura pauperata

M. satura pauperata

M. navarrae

M. quinterae

M. inderena inderena

M. inderena inderena

Fig. 3. Adult Manerebia, left dorsal view, right ventral view. Double arrows above names indicate image is reflected in vertical
plane, so figured wings are righthand pair. A, M. satura satura male; B, M. satura lamasi n. ssp. male; C, M. satura pauperata
n. ssp. male; D, M. satura pauperata n. ssp. female; E, M. navarrae male; F, M. quinterae male; G, M. inderena inderena male;
H, M. inderena inderena female. See Appendix 4 for specimen data.

39: 37-79, 2000 (2006)

Fig. 4. Adult Manerebia, left dorsal view, right ventral view. Double arrows above names indicate image is reflected in vertical
plane, so figured wings are righthand pair. A, M. inderena antioquiana n. ssp. male; B, M. inderena antioquiana n. ssp. female;
C, M. inderena fina n. ssp. male; D, M. inderena fina n. ssp. female; E, M. inderena similis n. ssp. male; F, M. inderena similis
n. ssp. female; G, M. inderena clara n. ssp. male; H, M. inderena clara n. ssp. female. See Appendix 4 for specimen data.

56

J. lii's.Lefnd.

25()()-26()()ni, 27. XI. 1998, 1’. Boyer PB; 2 males: Loja, Barque
Nacional Podoearpus, (^ajanunia, 27()()ni, 10. XI. 1996, A. Neild,
TVVP; 1 male: Loja, Loja-Zaniora, K1 Basurero, 260()m, 22. XI. 1997,
P. Bover leg., PB; 2 males: Loja, Old road L()ja - Zamora, 2800
m, XL 1999, 1. Alda.s leg, .VIBLI; 4 males, 1 female: .same data as
preceding but 2500 m, MBLI; 1 male: Loja, Cenen Alto, 2800m,
XL 1999, 1. Aldas leg, MBLI. PERL': 1 male: Cajamarca, Tabaconas,
I. Aldas leg, LWP. '

Etymology: Ebe name “mirena” is composed from the subspe-
cific names “milaeua” and “inderena”.

Comments: This stibspecies occurs lrt>m southeastern Ecuador
on both slopes ol the Andes (Zamora-Chincbipe, Lttja) to north-
eastern Peru (Tal)aconas). It has been recorded in cloud forest
within ;i narrow elevational band, from 2500-3000 m, where it may,
however, he locttllv common.

Manerebia goloudrina Pyrcz & Willniott n. sp.

Figs. 5D, 1 21 1, 17

Manerebiii n. s|t. (Pyrcz, VVillmoti & Hall); Lamas & Viloria
(2004:216,11.1128).

Diagnosis: fins species lacks die pale VI IVV postdiscal hand that
occtirs in oilier A/«)im7w/ species with similar genitalia (M. inderena,
M. /nallornm n. sp.) . The wing shape, ocelli on the VHW and ocellus
on the f)I IW in cell Ctt2-Cul are somewhat similar to M. inderena.
fhe male genitalia are indistiuguishahle from those of A/, inderena,
excejit for lacking latertil teeth' on the aedeagtis as jiresent in A/.
inderena, a character that varies within other species.

Description: MALE (Fig. 51)): Head. Irons with a ttift of long,
dark lirown hair; eyes glahrotis, dai k brown; labial palpi covered
with black and brown hair ventrally and dorsally, laterally with
shot t black and light brown .settles, last segment covered with light
brown scales and ventnilly with short brown hair; antennae blown
with white .scales at the base of each segment, club only slightly
thicket than shaft. Thorax- dorsal and ventral stirface dark brown;
legs paler btown. Abdomen: dorsal and ventral surface dark brown.
Wings: forewing (length: 19-21 mm, mean: 20.1 mm, n=10) distal
margin slightlv convex, apex rouiuled; bindwing with distal margin
sliglulv angled at vein M.3, with tomal notch almost absent. DRV
medittm brown; darker brown in distal area, androconial scales not
apparent. DHVV medium brown, with small black sttbmarginal ocel-
Itis ringed with dark orange in cell Cti2-CuL \'FVV ground colotir
medium brown, dai kei brown in basal half; very indistinct, darker
brown postdiscal line, in cells (’.ti2-(atl to costa; indistinct, slightly
tmdtilate, darker brown submarginal line from torntis to apex; distal
margin lined indistinctly with dark brown. V'HW medium brown,
dai ker towards base; slightly paler brown, thin, straight postdiscal
line from apex to tornus; faint, darker Itrown, undulate submarginal
line; a black sttbmarginal ocellus, ringed with dark orange, with
a white pupil, in cell ('.u2-('.u 1 , a similar but smaller ocellus in the
anterittr half of cell lA-(at2, and a black dot in the posterior half
of the same cell; distal margin lined indistinctly with dark brown.
Male genitalia (Fig. 1211): uncus bent near base then flat, subtmei
of medium length; valvae thinning sharply at middle and tapering
postei iorly, dorsally grooved, with 4-6 'teeth' at distal tip; aedeagtis
curving dorsally, laterally smooth.

FEMALE: Unknown.

Types: Holotype male: EULfVDOR: Uarchi, Reserva Forestal
Las C'.olondi inas, 2350 m, 20.V.1999, T. Pyrcz & J. Wojttisiak leg.
MZL'|; Paratypes (29 males): ECUADOR: 1 male: same data as the
holotvpe btit 26()0m. 22.VI.1999, fWP; 3 males: same data but 2200
tn, 22.VI.I999, IVVP; 2 males: same data but 2150 m, 23.VL1999,
'RVP; 3 males: .same data btit 2000 m, 23.\'L1999, TWP; 1 male:
same data Imt 2550 m, 02.VIL1999, fWP. 1 male: .same data bttt
2300 111, 23.VI. 1 999, IVVP; 1 male: .same data btit 2200 in, 22.VI. 1999,
TVVP; 1 male: same data but 27.VL1999, TWT; 1 male: same data
but 2150 111, 21.\'L1999, BMNH; 2 males: Carchi, nr. La Carolina,
Reserva Las Colondrinas, Santa Rosa, 1700 in, 05. IX. 1996, K. Will-

niott leg., KVy|H; 1 male: Carchi, Tulcan - Maldonado, 1 300-1600 m,
24. V. 1997, A.Jasinski leg, TVVP; 1 male; Carchi, Tulcan-Maldonado
km 40 a 50, 2800-3200 in, P. Boyer leg, PB; 2 males: same locality,
2450 111, 27.V^I1L2004, T. Pyrcz leg., TVVT; 9 males: hnbabura. La
Carolina, Route de Btienos Aires km 25, 2600ni, 05.V.2000, P. Boyer
leg, PB (7),TW'P (2).

Etymology: The specific name is derived from the name of the
private cloud forest reserve where most of the individuals, includ-
ing the holotype, were collected, the Reserva Las Colondrinas,
managed by Fundacidn Golondrinas.

Comments: This sjiecies is most closely related to M. inderena,
with which it is currently not known to be sympatric. We treat it
as a distinct species because it occurs at notably lower elevations
than AT inderena Jina, and because of the phenotypic similarity of
A/, inderena fina and AT inderena inderena from northeastern and
western Ecuador. In addition, specimens of AT inderena fina are
known from Cotacachi at 30()()m, approximately 40 km sotith of
the nearest locality of M. golondrina, between which there are no
obviotis geographical barriers. This species is known to date only
from northwestern Ectiador (Fig. 17), an area that is a local centre
of endemism forclotid forest satyrines (e.g., Lasiophila phalaesia (dee
Pyrcz, CoradesviolaeeaVxYcv., Pedaliodes phrasicla inmaeidalaVyrcv.) . It
is found on the south (Buenos Aires) and north (Las Golondrinas)
banks of the Rio .Mira. It should certainly also occur in sotithwestern
Golombia south of the Rio Narifio valley. The two males collected by
KRVV were both attracted to rotting fish. Other individuals were cof
lected in traps baited with excrement. Manerebia golondrina occurs
in middle elevation cloud forests from 1600-2600 in, but sani|iling
with baited trajis carried otit by the first author in the Reserva I.as
Golondrinas, along an elevational transect, indicated the species
to occur most commonly from 2000 to 2600ni.

Manerebia prattorum Pyrcz & Willniott n. sp.

Eig,s. .5E, 121, 17

Manerebia n. s|). (Pyrcz, Willniott & Hall); Lamas & Vdloria
(2004:216,11.1129).

Diagiio,si.s; This species is easily distinguished from its congeners
in the northern Andes by the postdiscal orange band on the DHVV’
only, fhe only species with somewhat similar pattern is Manerebia
lisa (VV'eynier), occurring in central Peru, but in that taxon the
btiiid is darker, with blurred edges, narrowing gradually from anal
to costal margin. The male genitalia of AT lisa show that the two
species are not closely related, however, being similar to AT salura
(see cbaracters grouping M. salura with M. franciscae, tinder the
latter species) . The size and wing shape of AT prattorum are similar
to AT inderena mirena n. .ssp. and AT undulata milaena n. .s|i, n. ssp.
Manerebia prattorum, however, lacks the ventral magenta or greyish
sheen in the distal marginal areas that characterises M. undulata,
and does not have the DHVV stibinarginal ocelli that occur in AT
inderena mirena. The male genitalia differ from AT inderena and
AT undulata in having only 2-3 'teeth' at the distal tip of the valva,
and from AT rufatmlis in having a straighter uncus and dif ferently
shaped valva (see diagnosis of the latter species).

Description: MALE (Fig. 5E): Head: Irons with a tuft of short,
brown hair; eyes chocolate brown, smooth; labial palpi covered
with short, meditmi brown bair; antennae medittm brown, slightly
lighter on ventral surface. Thorax- dorsal and ventral stirface dark
brown; legs paler brown. Abdomen: dorsal and ventral stirface dark
brown. Wings: forewing (length: 1 7-1 7.5 mm; mean: 1 7.2 mm, n=4)
distal margin almost straight, apex rotinded; hindwing with distal
margin slightly angled at vein M3, tornal notch almost absent. DRV
medium brown, darker brown towards base. DHVV' ground colour
medium brown; a broad, orange postdiscal band extending from
costa to anterior half of cell 1 A-Cu2, basal edge straight, distal edge
kinked at vein M3, so that widest point of band is at vein M3. V’RV
ground colotir mediimi brown, basal half slightly darker, bordered
distally by veiy thin, indistinct, darker brown postdiscal line, that

39: 37-79, 2000 (2006)

57

is sliglitly inclined towards afrex; faint, slightly wavy, darker brown
siibmarginal line t'roni torniis to apex. VHW medium brown; broad
whitish postdiscai band from apex to tornus, tapering towards
costa and tornus and widest in cells Cu2-Cul and Cai1-M3, passing
through base of cell Cnl-M3; faint, slightly undulate darker brown
submarginal line. Male genitalia (Fig. 121): uncus nearly straight
except where bent sharply near base, subunci of medium length;
valvae thinning sharply at middle and tapering posteriorly, dorsally
grooved, with 2-3 ‘teeth’ at distal tip; aedeagus curving dorsally.

FEMALE: Unknown.

Types: Holotype male: PERU: Piura, arriba de Canchaciue,
2100 m, 0522/7934, 05.V1.2000, G. Lamas kg., MUSM; Paratypes
(10 males): PERU: 2 males: same data as the holotype, MLISM; 2
males: same data as the holotype except R. Robbins kg., USNM;
1 male: same data as preceding except 07.VI.2000, LISNM; 1 male:
same data as preceding except G. Lamas kg., MUSM; 4 males: West
slopes of Andes, N. Peru, 10 ()()() ft., June 1912, Pratt,Joicey Bequest,
Brit. Mus. 1934-120, BMNH.

Etymology: This species is dedicated to Antwerp Edgar Pratt and
his son Felix Pratt, who first collected it almost a centuiy ago.

Comments: Manerebia prattorum occiiva at 2100-2600 m on the
western slopes of the Andes in northwestern Peru, on the west slopes
of the Andes, east of the locality of Ganchaque and possibly in the
valley of Huancabamba (Piura). Specimens potentially represent-
ing another unde.scribed subspecies have also been collected in
northwestern Peru (Cajamarca) , west of Chiclayo, above the locality
of La Florida (Fig. 17).

Manerebia Irimaciilata (Hewitson, 1870)

Figs. 5F,G,H, 13A, 18

Lymanopo/la trimarulala Mewitson (1870: 159). TL: Ecuador,
Morona-Santiago, St. Rosario. ST male: BMNH(T) [examined].

Manerebia trimaculala (Hewitson); Lamas & Viloria (2004:
216).

Diagnosis: Manerebia trimaculata dnd the related species, M. iin-
dulala n. sp. and M. interrupla, are all characterised by a light greyish
or magenta marginal sheen along the distal quarter of the ventral
surface of both wings, and a thin, dark brown, undulate line passing
through the center of the VHW discal cell. The male genitalia (Fig.
13A) of all three species are distinctive in the uncus being almost
straight, so that the dorsal edge of the tegumen and uncus form a
smoothly curving line, the subunci are very short and the valva is
sharply constricted in the middle to |)roduce the attenuated distal
half. This species might arguably be considered conspecific with
the west Andean M. undulala, but our reasons for keejring them
separate are discu.ssed under that species. The VHW yellow band in
specimens of the syntypic series is reduced to three spots near the
tornus (Fig. 5F), but in other specimens it can be fully developed,
but rather irregular at the basal edge (Fig. 5G). Manerebia trimaeu-
lata is readily distingidshed from other similar species by the two
or more well developed submarginal ocelli in cells Gu2-Gul and
Cul-M3 on the DHW, and usually by the submarginal ocellus in
cell Cu2-Cul on the VFWf Some specimens of M. interrupla have
similar but smaller ocelli and are typically smaller in size.

Comments: This species is confined to southeastern Ecuador
(Morona-Santiago and Zamora-Chinchipe) (Fig. 18), where it is
rather uncommon, and it was omitted entirely by Brown (1944).
Manerebia trimaculala occurs in relatively intact cloud forests from
2500-2800 m, and in contrast to the closely related M. interrupla, it
frequents openings within the forest, rarely straying into cleared
areas. A single specimen was collected in the Cordillera del Condor
on the border between Ecuador and Peru (Camp Achupalla, 15
km E. Gualaquiza, 2100-2200 m, MUSM).

Manerebia undulata Pyrcz & Hall, n. sp.

Manerebia undulatais characterised by a thin, dark

brown, undulate line passing through the center of
the VHW discal cell, a character only otherwise oc-
curring distinctly and consistently in M. inlerrupln
and M. trimaculala. Also distinctive is the paler brown
or purplish shading around the distal margin of the
VHW and apical area of the VFW' (also occurring in
M. trimaculata and M. inlerrupta) , and the markedly
undulate dark brown VHW submarginal line. The
first two of these characters distinguish M. undulala
from most other west Ecuadorian species (M. inderena,
M. germaniae, M. ignilineala) . The last two characters
also occur in M. ruf'analis n. sp., but that has distinct
male genitalia, with a smoothly arching uncus (not
straight), longer subunci and a smooth ventral edge
to the valva, which gradually tapers posteriorly. The
most closely related species, as indicated by wing
pattern and the male genitalia (see discussion un-
der M. trimaculata) , seem to be M. interrupta and M.
Irimaciilata. The former occurs at higher elevations
and in drier habitats on the western slopes and may
be distinguished by its smaller size, the more pointed
forewing apex, DHW submarginal ocelli, and uneven
VHW postdiscai band (in forms with a full band).
Manerebia undulata is not known to be sympatric with
the eastern slope M. Irimaculala, with which it might
be considered conspecific, but given that M. undulala
is much commoner at substantially lower elevations,
not (or only exceptionally) |)olymorphic in expression
of the hindwing band, and lacks prominent ocelli on
the DHW in any subspecies, we treat the two taxa as
distinct species.

Unlike M. trimaculata And M. interrupta, polymor-
phism of the pale VHW postdiscai band in M. undu-
lala is rare or absent (but see discussion under M. u.
undulata). Two subspecies are recognised.

Manerebia undulata undulata Pyrcz & Hall, (n. sp.)

Fig,s. 6A,B,C, 13B, 18

Manerebia ii. sp. (Pyrcz, WillmoU & Hall); l.amas & Viloria
(2004:216,11.1125).

Diagnosis: The nominate subspecies has narrower, although
somewhat variable, pale VHW |io.stdiscal bands and is slightly larger
than A'/, undulata milaena (described below).

Description: MALE (Fig. 6A): Head: Irons with a tuft of butwn
hair; eyes dark brown, smooth; labial palpi covered with brown
hair; antennae dorsally greyish-brown, ventrally orangeish, with
whitish scales at base of each segment, tip darker brown. Thorax.
dorsal and ventral surface dark brown; legs paler brown. Abdomen:
dorsal and ventral surface dark brown. Wings: forewing (length:
17.5-19 mm; mean: 18 mm; n=12) distal margin slightly convex,
apex rounded; hindwing with distal margin slighth' angled at vein
M3, with tornal notch almost absent. DFW medium brown, darker
brown in basal half. DHW medium brown, slightly darker towards
base; mintite black submarginal ocellus with white pupil in cell Cu2-
Cul. VFW ground colour medium brown, slightly darker towards
base; thin, distinct, almost straight, darker brown postdfscal line
from anal margin to near costa; distinct, tmdulate, darker brown
submarginal line from tttrnus to apex; very thin, indistinct, dark

58

/. lit's.Lrpid.

Fig. 5. Adult Manerebia, left dorsal view, right ventral view. Double arrows above names indicate image is reflected in vertical
plane, so figured wings are righthand pair. A, M. inderena mirena n. ssp. male; B, M. inderena mirena n. ssp. female; C, M.
inderena leaeniva n. ssp. male; D, M. golondrina n. sp. male; E, M. prattorum n. sp. male; F, M. trimaculata male (form); G,
M. trimaculata male (form); FI, M. trimaculata female. See Appendix 4 for specimen data.

39: 37-79, 2000 (2006)

59

Fig. 6. Adult Manerebia, left dorsal view, right ventral view. Double arrows above names indicate image is reflected in vertical
plane, so figured wings are righthand pair. A, M. undulata undulata n. sp. male; B, M. undulata undulata ? male; C, M. undulata
undulata n. sp. female; D, M. undulata milaena n. ssp. male; E, M. interrupta male (form); F, M. interrupta male (form); G, M.
interrupta male (form); FI, M. interrupta female. See Appendix 4 for specimen data.

60

/. Res.Le/ml.

brown inarsi;inal line; |)ale brownish scaling extending in from
distal margin to siirroimd dark brown siibmarginal line, from cell
Cnl-M3 to apex. VI fW medium brown; indistinct, undtilating dark
brown line f rt)m costa to anal margin through middle of discal cell;
thin, whitish postdiscal band from apex to tornus, passing through
base of cell C,ul-M.3, slightly convex and thinning slightly towards
costa; darker hrown, strongly zigzag submarginal line; minute
black submarginal ocelli, with white pupils, in cell lA-(ai2 (two)
and cell (.u2-(ail, white submarginal dots in cells Cul-M3 and
M3-M2; pale brownish scaling extending in from distal margin to
surround dark brown submarginal line from tornus to apex; veiy
faint, thin, dark brown marginal line close to and parallel with
distal margin. Male genitalia (Fig. 1.3B); uncus straight, with dor.sal
edge and dorsal edge of tegumen forming a smooth cune, subunci
short; valvae thinning shaiply at middle and tapering posteritn ly,
dorsally grooved, with 4-6 'teeth’ at distal tij); aedeagtis curving
dorsally, laterally smooth.

FEMALE (Fdg. 6(1); Differs from male as follows; ventral surface
lighter and duller; an ocelltis in cell 1 A-(lu2 ttn the forewing.

Type.s: Hololype male: Edl'ADOR: Bolivar, Balzapamba, arriba
de Santa Lucia, 1700-1 7.50 m, 03.1X.2003, T. Pvrcz leg.. MZUJ; Al-
lotype female; same data as the holotvpe, MZL’J; Parafypes (55 males
and 2 females): EdL’ADOR: 31 males and 1 female: same data as
the holotvpe, TWP (29), BMNl I (2 males); 2 males: Bolivar, Balza-
])amba, arriha de Santa Lucia, 1400-1450 m, 03. IX. 2003, T. Pvrcz
leg, I'WP; 2 males; same locality, 1600-1650 m, 05. IX. 2004, T. Pvrcz
leg, FVVP; 9 males: Loja, Zamhi, 2200-2.300 m, 08. III. 1998, P. Boyer
leg, I'VX'P (4), PB (5); 2 males: Loja, dtiayqtiichuma, 28.V.1996, S.
Attal & 1. Aldas leg, FWP; 1 male: same data but 2LV.1996, FWP;
4 males: Loja, duaycjtiichuma. 1800-2000 m, 15.1V. 1997, A. Jasihski
leg, FVVP (3), Ml INLld ( 1 ); 1 male: Zamora -dhinchiite, Zambi, A.
[asihski leg, FW'P; 3 males: dotopaxi, Pilald, 2500-3000 m, VII. 1996,
I. Aldas leg, 4 WP; 1 male: Pichincha, .Vloag-Sto. Domingo rd., Tan-
dapi, 1550 m, 10. VIII. 1993, |. Hall leg, MEdN; 2 males: same data
as preceding except 1700 m, 3.\'1II. 1996, R. VVillmott leg., KV\']H;
1 male: Pichincha. old (.)uito-Sto. Domingo rd., nr. dhiriboga, Rio
Las Palmeras, 1900 m, 14.VIIL 1993, |. Hall leg, R\y|H; 1 female:
f.oja. Zambi, P. Boyer leg, PB.

Etymology: I'he name is the f eminine form of the Latin adjec-
tive “nndulattis”, meaning undulate, in reference to the undulate
VHVV submargiual line.

Comments: VVe have examined two specimens of A/, undulata
in the TWP recentlv collected by Ste|thane Attal in southwestern
Ectiador (Loja) and lahelled "\angana. route de Valladolid, 2600m,
23.V.2000” which have a hroken jtale \41W postdiscal band (Fig.
6B). A single male in the BMNH from “Ectiador” is similar. These
specimens are associated with M. undnlatavaxher thaw M. inlerriipla
tm the basis of their larger size (30% larger than A/, interrupta) and
rounded forewing apex. It is possible that these specimens re]3resent
a local population in which the hroken band is monomorphic, given
the lack of known specimens with a comjrlete hindwing band from
this area and phenotyjric stability of the species elsewhere, but more
material is retpiired to confirm this.

The nominate subspecies occurs on the western and south-
western slopes of the Ecuadorian Andes, where it may be locally
common. It has been recorded from 1400-2300 m.

Manerebia undulata milaena Pvrcz & Willniott, n. ssp.

Eigs. 6D, 13C, 18

Manerebia n. sp., n. ssp. (Pvrcz, VVillmott & Hall); L.amas &
Viloria (2004: 216, n. 1125b).

Diagnosis: This subspecies is smaller than the nominate. The
VHVV pale postdiscal hands are wider, and broaden from the
costa to tornus. Fhe ventral distal marginal sheen is wider and
a purplish, rathei than jtale brown, colour. The dor.sal surface
ground colour is also paler, so that the VHVV postdiscal band is
faintly visible. The male genitalia do not differ consistently from

AT undulata undulata.

Description: MALE (Fig. 6D): Head: eyes, lahial palpi and an-
tennae as in the nominate subspecies. Thorax, dorsal and ventral
surface dark brown; legs paler brown. Abdomen: dorsal and ventral
surface dark brown. Wings: forewing (length: 17-18.5 mm; mean:
17.1 mm, n=3) distal margin angled slightly at vein ML apex
rounded; hindwing outer margin almost |jerfectly rotmded with
a very slight angle at vein M3, and a slight tornal notch; fringes
light brown. DF'W medium brown, darker brown in basal half,
androconial scales not a|tparent; a faint, darker, tmdulate suhmar-
ginal line. DI fW medium brown, slightly darker towards base; very
faint, |)ostdiscal band of pale .scales reflecting VHW band, stronger
at costa; slightly paler yellowish brown, sparse scales in marginal
area; darker, zigzag submarginal line. VFW^ ground colour medium
brown, darker towards base; thin, distinct, almost straight, darker
brown [tostdiscal line from anal margin to near costa; distinct,
slighih undulate, darker brown submarginal line from tornus to
ajtex; very thin, indistinct, dark brown marginal line; pale purplish
grey scaling extending in from distal margin to surround dark
brown snbmarginal line, f rom cell CuTM3 to a|3ex; row of tiny white
submarginal dots in cells Cat2-Cti 1 to M2-M 1 . VHVV' medium brown;
indistinct, undulating dark brown line from costa to anal margin
through middle of discal cell; thin, yellowish white postdiscal band
from apex to tornus, passing through base of cell Cail-M3, ap|)roxi-
mately straight and broadening continuously f rom costa to torntis;
darker brown, strongly zigzag submarginal line; pale purplish grey
scaling extending in from distal margin to surround dark brown
submarginal line from tornus to cel Ml-Rs. Male genitalia (Fig.
13(1): tmeus straight, with dorsal edge and dorsal edge of tegumen
forming a smooth ctirve, stibunci short; valvae thinning sharply at
middle and ta|tering posteriorly, dorsally grooved, with 4-6 'teeth'
at distal tip; aedeagtis curving dorsally, laterally smooth.

FEMAI.E: Unknown.

Types: Holotype male: F.dLlADOR: Loja, Cordillera de Lagunil-
las, Jimhura- Laguna Negra, 3000-3200 m, 15.V.1998, A. Jasihski leg,
.MZLJ; Paratypes: 2 males: same data as the holotype, TVV'R.

Etymology: The name is a composite of the names of two other
related Ecuadorian Manerebia taxa; (AT inderena) mirena and (AT
i n derena) leaen iva.

Comments: Unlike A/, undulata undulata, which is a cloud forest
species of intermediate elevations, the type series of A/, undulata
milaena was collected at the tree-line in the Cordillera de Lagtmillas
range, at 3000-3200 m. Unforttmately, there is virtually no natural
vegetation remaining in sotithwestern Ecuador helow 2800m. so
it may he impossible to locate the true natural lower elevational
limit. A single individual of (apparentlv) AT undulata milaena (in
the MLI.SM), collected ftirther south on the west slopes of the
Andes in Peru, east of Chiclayo (Cajamarca), at 2400-2600 in, is
exchided from the type series. Specimens of AT prattorum irom the
same locality (see southernmost data point on Fig. 17) re|)resent a
stibspecies different from that in sotithern Ecuador, and the same
might apply to M. undulata.

Manerebia interrupta (Brown, 1944)

Eigs. 6E,F, G, H, 13D. 18

Penro.sada apicutata form interrupta Brown (1944: 257, male
genit. Fig. 1618). TL; Ecuador, Aztiay, Seville de Oro. HT male:
Ecuador, Azuay, Seville de Oro, 2500 m, 15.11. [19] 39. AME [pho-
tograph examined].

= Manerebia keradialeuka Hayward (1968; 205, Hgs. 4, 8). TL: Flc-
uador, Aztiay, Tarqui. HT male: Ecuador, Azuay, Tarqni, 08.\'. [ 19]65,
L.tiis Pena teg. IMLT [examined].

Penro.sada apicutata form ciDiiilinea (W'eymer); Brown (1944:
258) (misidentifkation).

Manerebia interrupta (Brown); Lamas & Viloria (2004: 215).

Diagnosis: As in AT trimaculata, the VHVV' yellow band is variably
expre.ssed, and varies from complete absence (Fig. 6G), through

39: 37-79, 2000 (2006)

61

a series of semicircular spots (Fig. 6F), to being complete, with an
uneven distal edge (Fig. 6E). Manerebia interrupta is much smaller
than M. trimcinilata, has a more acute forewing apex, and lacks an
ocellus on the VFW in cell Cul-Cu2. Brown (1944), and presum-
ahly earlier workers, misidentified this species in Ecuador as the
superficially similar but genitalically distinct Colombian M. apicu-
Idtn, and named a form with the band broken into spots as form
interrupta. Hayward (1968) subsequently named a form without any
hindwing band as Manerebia keradialeuka. Brown’s name interrupta
thus becomes the first available name for this species, to which M.
keradialeuka Hayward is a junior synonym (Lamas &Viloria, 2004).
Male genitalia as illustrated (Fig. 13D).

Comments: Manerebia interrupta appears to be most closely re-
lated to M. undulata'dnd M. trirnaculata (see discussion under those
species). Although it has not been reported from the same sites as
either M. trirriarulataor M. undulata, its closest relatives, it appears to
replace each locally at higher elevations and in drier habitats. This
species occurs from south-central (Morona-Santiago: Gulalaceo-
Chiguinda rd.; Gualaceo-Limon road,) to southern Ecuador (Loja:
above Catamayo; Cerro Palma, Loja-Zamora rd.) and in northern
Peru (Piura: entre Las Minas y El Tambo) on the western slopes of
the Andes. It occurs from 2400 m up to the tree-line around 3200
m. Males were found flying low to the ground in areas of recent
bamboo regrowth on landslips in elfin forest/ paramo mosaic, and
also in a dry river gulley through desert scrub on the southwestern
slopes. The species is most commonly encountered in drier habitats,
such as those of the inter-Andean valleys. We have also observed
males hilltopping at Cerro Palma, and occasionally puddling at
damp sand. The species appears to be highly seasonal; whereas
it was very common along the Gualaceo-Limon road in February
(wet season), no individuals were observed in the same locality in
August, during the dry season.

Manerebia rufanalis Pyrcz & Hall, n. sp.

This species is clistinguislied from all others by the
rusty suffusion and submarginal ocelli at the tornus of
the DHW. The light magenta sheen along the distal
margins on the ventral surface, especially at forewing
apex and on the hindwing, is also distinctive. The VFW'
ocellus in cell Cul-Cu2 is generally, but not always,
well developed and is occasionally also apparent in
adjacent cells. The VHW^ yellow band is variable and
appears to be relatively wider in smaller specimens.
The genitalia are distinctive within the genus, and
differ from those of M. inderena and M. undulata by
having an arched uncus and relatively long subunci,
similar to species such as M. fe/maand M. salum, while
the elongate valva, which lacks ‘teeth’ at the dorsal
edge near the base, similar to M. inderena and M.
trirnaculata, has a smoothly upwardly curving (rather
than “stepped”) basal edge. In addition, the valva
always has relatively few (usually 2-3) large ‘teeth’ at
the distal tip.

Manerebia rufanalis rufanalis Pyrcz & Hall, (n. sp.)

Figs. 7A, B, 1,3E, 19

Manerebia n. sp. (Pyrcz, Willmott & Hall); Lamas & Viloria
(2004; 216, n. 1126).

Diagnosis: The nominate subspecies is distinguished from M. r
femandina ds disussed below. There is some variation in this taxon.

and some individuals have an additional ocellus on the DHW in
cell Ctil-M3.

Description: MALE (Figs. 7A): Head: Irons with a tuft of dark
brown hair; eyes chocolate brown, smooth; labial jtalpi covered
with dark brown hair; antennae dorsally chestnut, ventrally rufous,
with white scales at the base of each segment, club same colour as
shaft. Thorax, dorsal and ventral stirface dark brown; legs |)aler
brown. Abdomen: dorsal and ventral surface dark brown. Wings:
forewing (length: 18-21 mm; mean; 19.9 mm; n=8) distal margin
slightly angled at vein M2, ajtex rounded; hindwing slightly angled
at vein M3, with slight tornal notch. DEW medium brown, darker
brown in basal half. DHW medium brown, slightly darker towards
base; a small submarginal black ocellus ringed with orange, with
a white pupil, in anterior half of cell 1 A-Cu2, a larger similar ocel-
lus in cell Cu2-Cul; orange-brown scaling alotig the anal margin,
broadening into tornus to extend to edge of ocellus in cell 1 A-Cu2.
VFW ground colour medium brown, darker in basal half; thin, very
faint, straight, darker brown |)ostdiscal line in cells Cu2-Ml; black
submarginal ocellus ringed with dark yellow, with a white pupil,
in cell Cu2-Cul; three white submarginal dots in cells Cul-M3 to
M2-M1; indistinct, unditlate, darker brown submarginal line from
tornus to apex; thin, very faint, dark brown marginal line; pale
greyish scaling extending from distal margin to surround dark
brown submarginal line in cells M3- .M2 to costa. VHW medium
browti, scattered with very .sparse red-brown scales in basal two-
thirds, denser along anal margin and costa, particularly at apex;
pale yellow postdiscal band (becoming white at basal edge) from
apex to tornus, passing throngh base of cell Cul-M3, straight and
of even width; faint, darker brown, zigzag snbmarginal line; three
small black submarginal ocelli, with a white pupil, in cells I A-Cn2
and Cu2-Cul; a white submaiginal dot in cell Cul-M3; thin, dark
brown marginal border; pale greyish scaling sparsely extending in
Ifom distal margin to just past dark brown submarginal line from
tornus to cell M.3-M2. Male genitalia (Fig. 13E); uncus smoothiv
arching, subunci relatively long; basal edge of valvae smootbly
curving, dorsally grooved, with 2-3 ‘teeth’ at distal tip; aedeagus
thin and shallowly curving dorsalK, latei ally smooth.

FEMALE (Fig. 7B); Similar to male but lighter on both wing
surfaces.

Types: Holotype male: ECL'ADOR: Tungurahua, Banos, Rnntun,
26()()-300() m, Ob.VllI. 1998, T. Pvrcz leg., MZL(J; Allotype female-. EC-
UADOR: Loja, Old road I.oja-Zamora, 26()()m, XL 1999, 1. Aldas,
leg, MBL.I; Paratypes (32 males and 1 female): ECUADOR: 1 male:
Tungurahua, Banos, TWP; 1 male: Tnngnrahua, Banos, V. 1995, I.
Aldas leg, TWP; 1 male, Tungurahua, Run tun, 3()()()m, 21 .XL 1998,
P. Boyer leg., PB; 1 male: rungurahua, Bahos-Puyo rd., Rio Machay,
1700 m, 19-20.VI11. 1993, J. Hall leg, MECN; 1 male: .same data as
preceding except 12. IX. 1993, KW|H; 1 male: Napo, Baeza, 1800
m, IX. 1996, P. Boyer leg., PB; 1 male; Zamora-Chinchi|)e, San
Andres, 2200 m, 1 3.VI11. 1998, T. Pyrcz leg., fWP; 1 male; Zamora-
Chinchipe, region de Valladolid, 2500 m, 25. XL 1993, B. Mery &
S. Attal leg., TWP; 1 male; Zamora-Chinchipe, Ri'o San Francisco,
1400[?] m, 1 1 .XI. 1989, A. Crosson-dn-Cormier & S. Attal leg, TWP;
1 male; Zamora-Chinchi|5e, San Andres - Calderon, 2200-2600 m,
20.V.1998, A. Jasihski leg., TWP; 1 male; Zamora-Chinchipe, nr.
Valladolid, Quebrada Tapichalaca, 1950 m, 4. XI. 1997, K. VV'ill-
mf)tt leg, KWJH; 1 male: Zamora-Chinchi|5e. km 34 Jimbura-San
Andres rd., 2900 m, 23.IX.I997, K. Willmott leg, KVy|H; 1 male:
Zamora-Chinchipe, Zamora-Loja i d., nr. Sabanillas, Quebrada San
Ramon, 1700 m, 27-29. X. 1997, K. Willmott leg., KVy[H; 13 males:
Zamora-Chinchipe, Valladolid, Quebrada de los Muertos, 2550 m,
XL 1999, 1. Aldas leg, TWP (4), MBLI (9); 3 males: Loja, Old road
Loja-Zamora, 2500-2600m, XI-XII.1999, 1. Aldas, leg, .MBLI; 1 male;
Tungurahua, El Tablon, .3000m, III. 1999, 1. Aldas leg, IVIBLI; 1 male;
Tungurahua, Banos, Rio Verde, 2300 m. III. 1999, G. Velastegui leg,
MBLI; 1 female; Loja, Old road Loja-Zamora, 2600m, XL 1999, 1.
Aldas, teg., MBLI. PERL': 1 male; Cajamarca, faltaconas, VIII. 1996,

62

/. Rn.Lepid.

M. ignilineata ignilineata M. ignilinaata ignilineata

Fig. 7. Adult Manerebia, left dorsal view, right ventral view. Double arrows above names indicate image is reflected in vertical
plane, so figured wings are righthand pair. A, M. rufanalis rufanalis n. sp. male; B, M. rufanalis rufanaiis n. sp. female; C,
M.. rufanalis fernandina n. ssp. male; D, M. ignilineata ignilineata male (form); E, M. ignilineata ignilineata male (form); F, M.
ignilineata ignilineata male (form); G, M. ignilineata ignilineata female. See Appendix 4 for specimen data.

I. Alcki.s leg.. TWF.

Etymology: Tlie .s|)ecifs name is derived from the Latin “rufus”,
reddish brown, referring to tlie distinctive color in the anal area
of the DHW.

Comments: fhis subspecies is known from the east Andean
slopes in central Ecuador to far northern Peru, and it is one of
the most common Manerebia, occurring in fairly intact cloud for-
est habitats from 17()()-29()() m. Males may often be encountered
]mddling at stream and river banks, particularly at urine, and are
attracted to rotting fish. We have also observed males hilltopping

on open, grassy summits with low bushes, along the Jimbura-San
Andres road.

Manerebia rufanalis fernandina Pyrez & Willmott, n.
ssp.

Figs. 7C, 13F, 19

Manerebia n. sp., n. ssp. (Pyrez, Willmott & Hall); Lamas &
Viloria (2004: 216, n. 1126b).

Diagnosis: This sitbspecies is slighdy smaller than the nominate.

39: 37-79, 20()() (2006)

63

M. ignilineata neglacta M. ignilineata neglacta

M. ignilineata neglecta

M. seducta

M. ignilineata neglecta

M. seducta

M. magnifica

M. mycalesoldes

M. mycalesoldes

Fig. 8. Adult Manerebia, left dorsal view, right ventral view. Double arrows above names indicate image is reflected in vertical
plane, so figured wings are righthand pair. A, M. ignilineata neglecta male (form); B, M. ignilineata neglecta male (form); C, M.
ignilineata neglecta male (form); D, M. ignilineata neglecta male (form); E, M. seducta n. sp. male; F, M. seducta n. sp. female;
G, M. magnifica n. sp. male; H, M. mycalesoldes male; 1, M. mycalesoldes female. See Appendix 4 for specimen data.

/. Res.Lepid.

()4

has reduced reddisli-Irrowii scaling on the DI IW anal area and the
ocelli are smaller on both wing surfaces.

Description: MALE (Fig. 7(1); Head, frons with a tuft of dark
brown hair; eyes chocolate brown, smooth; labial palpi covered
with long, dark brown hair; antennae dorsally chestnut, ventrally
orange, with white scales at the base of each segment, chib same
colour as shaft. Thorax, dorsal and ventral surface dark brown;
legs paler brown. Abdomen: dorsal and ventral surface dark brown.

forewing (length; 18 mm. mean: 18 mm, n=l) distal margin
slightly angled at vein M2, apex rounded; hindwing slightly angled
at vein M3, with slight tornal notch. DEW medium brown, darker
browti in basal ball. 1)1 IW meditim brown, slightly darker towards
base; two small snbmarginal black ocelli ringed with orange, with
white jttipils, in anterior half of cell 1A-Cln2 and in cell Cn2-Clnl;
orange-brown scaling in the tornns, not extending to ocellus in
1 A-(ai2. VFW ground colour medium brown, darker in basal half;
thin, very faint, straight, tlarkei brown postdiscal line in cells (;tt2-
M 1 ; small black submarginal ocellus ringed with dark orange, with
a white pu[)il, in cell C:n2-( ai I ; three white subtnarginal dots in cells
C'.ul-M3 to M2-.M1; distinct, tmdulate, darker brown submarginal
line from tornns to apex; thin, very faint, darker brow'ii marginal
line; pale purplish grey .scaling extending from distal margin to
surround dark brown submai ginal line in cells Cu2-Cul to costa.
\'I IVV medium brown, scattered with very sjtarse red-brown scales
in basal two-thirds; |)ale yellow |tostdiscal band (becoming white
at basal edge) from ajtex to tornns, passing through base of cell
('.ul-M3, slightly concave and of even width; faint, darker browti,
zigzag submarginal line; two small black snbmarginal ocelli, with a
white pupil, in anterior half of cell I A-(iu2 and in Cu2-C'.u 1 ; a w'hite
snbmarginal dot in cell (lul-M.S; ibin, dark brown marginal border;
pale ptirplish grey scaling .s|)arsely extending in from distal margin
to jtist jiast dark brown stibmarginal line from tornns to cell M3-M2.
Male genitalia (Fig. 13F); tmens smoothly arching, subnnei relatively
long; basal edge of valvae smoothly curving, dor.sally grooved, with
a single “tooth" at distal tip; aedeagns thin and shallowly curving
dorsally, laterally smooth.

FEMALE; Unknown.

Types: Holotypemale: FULIADOR: Aztiay, Clirdn, San Fernando,
2500 m, 08.\f 1998, F. Krdl leg.. MZUJ; Paratypes: 2 males: same data
as the holotyjje, EWR

Etymology: The name is derived from that of the tvpe locality,
San Fernando.

Comments: I'he three types of A/. rufannlA femandina me the
only known specimens of this species from the western Andes.
The tvpe locality of fernandina (San Fernando) is in the Cuenca
valley, an area of endemism in clotid forest satyrines, generally at
the stibspecibc level (e.g., Lasiophiln phalaesia krnWPyrcv., Fedaliodes
xan/ho.sphenisra ssp. I’yrcz, in prep.). Since a number of other east
Andean cloud forest taxa cross o\er to the western slopes in south-
ern Ecuador (Willmott & 1 bill, in jti ep.), this taxon may also occtir
further south, where little habitat remains at a suitable elevation.

Manerebia igpiilineata (Dogniii, 1896)

Tills is the smallest species of Manerebia. and is
easily distinguished from all other species by the
elongate, rounded wings (a character shared with
M. pervaga), the irregular VHW stibmarginal line
placed relatively far from the distal margin, and the
male genitalia. The latter are markedly distinct from
all other north Andean species, except M. seducta, in
the short, squat valva, which is enlarged in the basal
half, lacks a dorsal groove, has very large, sparse distal
‘teeth’, and whose tips flare otitwards in ventral view.

Like M. trimaculata and M. interrupla, this species is
polymorphic in the expression of the VHW postdiscal
band, and occurs in three fairly discrete forms; one
form has a fttlly developed VHW band (Fig. 7D), the
second has the band broken into a row of quadrate
spots by dark brown lines on the veins (Fig. 7E), and
the last has the band completely absent (Fig. 7F).

Manerebia ignilineata ignilineata (Dognin, 1896)

Figs. 7D, E, F, C, 14A, 20

Lymanopoda ignilineataliognm (1896: 134). TL: Ecuador, Ltija,
Loja. HT male: Ecuador, Loja, Environs de Loja, 1890. BMNII
[examined]

Penrnsada ignilineata (Dognin); D’Abrera (1988; 824).

Manerebia ignilineata (Dognin); Lamas & Viloria (2004; 215)
(in part).

Diagnosis; 4 be nominate subspecies differs from M. i. neglerta
as discussed under that taxon. The expression of the VHW jiale
postdiscal band is polymorphic in both sexes (Figs. 7D, E, F, (1).
Male genitalia as illustrated (Fig. 14A).

Comments; Dognin’s (1896) description of this species is
concise but clear, and the holotype male, which has an indistinct,
broken \'HW |iostdi.scal band intermediate between the specimens
ligured in Figs. 7E and 7F, is in the BMNH. Manerebia ignilineata
ignilineata occurs at the cloud forest/paramo ecotone, and in the
lower paramo, in southern Ecuador (Morona-Santiago; Gualaceo-
Limbn; Loja; Jimbura-San Andres rd.; Loja-Zamora rd.) on the
eastern, and in far south, western slopes of the Andes, from 2700-
3300 m. Lite species seems to be veiy local and seasonal. Males fly
usually 1-3 m above dense stands of bamboo growing in paramo
intermixed with elfin forest. Manerebia ignilineata occurs also in
northern Peru, east of the Rio .Marahon, as a distinct subspecies,
M. ignilineata jalra Pyxxr (2004).

Manerebia ignilineata neglecta (Brown, 1944), n. stat.

Figs. 8A, B, C, D, 14B, 20

Penrnsada lanassa form neglecta Brown (1944; 260, male genit.
fig. 1619). TL; Ecuador, Tungurahua, Minza Chica. HT male; Ec-
uador, Tungurahua, Minza Chica, 08.1V.[19]39, 3200 m, leg. F. M.
Brown, No. B1619. AME [jrhotograph examined], n. stat.

= Penro.sada lanassa form disrontinua Brown (1944; 260, male
genit. fig. 1617). TL; Ecuador, Tungurahua, Minza Chica. HT male;
Ecuador, I'ungurahua, Minza Chica, 08.1V.[19].39, 3200 m, leg. F.
M. Brown, No. B1617, AME [])hotograph examined].

[Penro.sada lanassa (C. & R. Felder); Brown (1944; 258)]

Manerebia ignilineata (Dognin); Lamas & Viloria (2004; 215)
(in |)art).

Diagnosis; This subspecies differs from the nominate in the
forewing distal margin being nearly straight, except angled slightly
at vein M2, instead of convex, and in the hindwing being slightly
angled between M I and M2 and at M3, instead of almost perfectly
rounded. It also lacks the shining magenta colour on the distal
margin of the VHW and apex of the VFW, and the VHW submar-
giiial line is more zigzag than in the nominate.

Redescription; MALE (Figs. 8A, B, C); Head: frons with sparse
blackish hair; eyes blackish-brow'n, smooth; labial palpi covered
with black hair; antennae dorsally brown, ventrally dirty yellow,
club formed of 10-11 segments, nrice width of shaft. Thorax, dorsally
blackish-brown, ventrallv medium brow'n. Abdomen: dor.sally black-
ish-brown, laterallv and ventrally medium-brown. Wings: forewing
distal margin nearly straight excejit angled slightlv at vein M2, apex
rounded; hindwing slightly angled between Ml and M2 and at
M3, tornal notch absent. Fringes of both fore and bindwings light-
brown. DfAA' medium brown, darker towards base. DHW almost
uniform metlium brown, except for faint trace of postdiscal band

39: 37-79, 2000 (2006)

65

Fig. 9. Adult Manerebia, left dorsal view, right ventral view. Double arrows above names indicate image is reflected in vertical
plane, so figured wings are righthand pair. A, M. nevadensis male; B, M. nevadensis female; C, M. levana levana male; D, M.
levana levana female; E, M. pervaga n. sp. male; F, M. pervaga n. sp. female. See Appendix 4 for specimen data.

of pale scales reflecting VHW band, noticeable only towards costal
margin. VFU' grotind colour medium brown, slightly lighter in distal
third; two or three tiny, barely noticeable yellow submarginal dots;
rufous-brown, slightly undulate submarginal line. VHW medium
brown; indistinct, “S”-shaped postdiscal line, a thin (r. 1 mm) pale
yellow postdiscal band from apex to tornus, nearly straight, ]>ass-
ing along distal edge of discal cell at vein M3, in some individuals
broken into a series of spots or completely absent; a thin, wasy
submarginal darker brown line, area distal to it and immediately
basal pale grey with a light magenta sheen. Male genitalia: as il-
htstrated (Fig. 14B).

FEMALE (Fig. 8D): Similar to male except with a paler ventral
surface.

Comments: Brown (1944) introduced the names discontinua
and neglecta ior two individual forms of kinassa". The original
illustrations of the holotype male genitalia of each, though |rooi ly
drawn, show the sparse, large dorsally directed spines at the tijr ol
the valva, the very broad base of the valva, and the short, curved un-
cus and stibunci that occtir only in M. ignUinmta. yi’e have examined
photographs of the holotypes of neglecta and discontinua (provided
by Gerardo Lamas), now deposited in the AME, and both names
apply to a population of M. ignilineata occurring in central Ecuador
that a|)pears to be consistently distinct from the nominate. Lamas
& Viloria (2004) placed both names as synonyms of M. ignilineata,
and since no other atithors have dealt with Brown's names, we select
neglecta as the name for this taxon (n. .stat.). Manerebia ignilineata

/. lies.Lepid.

6(i

Fig. 10. Manerebia male genitalia, lateral view; v = valva ventral view, a = aedeagus lateral view, ad = aedeagus dorsal view.
M. leaena, M. germaniae n. sp., M. pluviosa n. sp. See Appendix 5 for specimen data.

neglecta is known from (ianai : Pimo. Gun, Zliud-Alansi km 2, 3100
m; rnngurahua: Minza Ghica, 3200 m; Napo: Fairallacta-Archidona
trail; Picliincha: Quito-Rio Toachi trail; Gliimborazo: Huigra; Hda.
Licay, above Ilnigra; Bolivar; Pilald, 3100 m sttnth to Ganar, from
3100-3200 m. It is locally common in the forest-paramo ecotone
near Zluid (A/.uay), where it flies with Neopedaliodes parrhoebia n.
ssp. and Lymanopoda hazflana n. ssp. (Pvrcz, in prep.).

Manerebia seducta Pyrcz & Willniott, n. sp.

Figs. 8E,F, 14G, 20

Manerebia n. sp. (Pvrcz, Willniott & Hall); Lamas & Viloria
(2004; 216, n. 1 122).

Diagnosis; Phis species is recognised by the elongate shape of
the wings, the white VHW hand unitjuelv lying distal of the disco-
cellidars, and by the small single ocelli on the fore and hindwing
(absent in the female). Phe male genitalia are most similar to A/.
igniUneala, which may he the sister species, and are characterised by
the broad base to the valva, with few terminal 'teeth’, and elongate
aedeagus. Manerebia seducta differs principally from M. igniUneala
by the less reddish ground colotir to the wings and more pointed
forewing apex. The two species are microsympatric.

Description; MALE; (Fig. 8F1); Head: Irons with a tuft of brown
hair; labial |ial|)i covered with dense and long black hair; eyes

blackish, smooth; antennae dorsally brown, ventrally chestnut,
white scales at base of each segment. Thorax, dorsal and ventral
surface dark brown; legs paler brown. Abdomen: dorsal and ventral
surface dark brown. PFmgs; forewing (length: 23 mm, mean; 23
mtn, n=2) elongate, triangular, with almost straight distal mar-
gin and sharply pointed apex; hindwing elongate and smoothly
rotmded, with no tornal notch. DFW' uniform medium brown.
14HW tmiform medium brown. VFW ground colour medium
brown, somewhat variable, slightly lighter towards outer margin;
irregular dark brown submarginal line; area between it and outer
margin suffused with magenta; a minute black ocelltis with white
pupil in cell Gtt2-Cu 1 . VI IW medium brown; barely visible, uneven,
dark brown line through middle of discal cell, fading near costa
and anal margin; thin, milky white postdiscal band from apex to
tornus, with somewhat irregular inner and sharp otiter edge; thin,
smoothly curving darker brown submarginal line, parallel to distal
margin, from tornus to near apex; thinner dark browti marginal
line; the area between submarginal line and otiter margin suffused
with magenta, twice as wide as on the forewing. Male genitalia: (Fig.
14G) tegumen slender; uncus arched, subuncus half the length of
uncus; valvae stout, sharply thinning at middle with three promi-
nent distal 'teeth' and grooved dorsal surface; aedeagus long, thin
and smooth, with a pronounced 'collar' at junction of anterior and

39: 37-79, 2000 (2006)

67

Fig. 1 1 . Manerebia male genitalia, lateral view; v = valva ventral view, a = aedeagus lateral view, ad = aedeagus dorsal view.
M. apiculata, M. franciscae, M. mammuthus n. sp., M. satura, M. navarrae, M. quinterae. See Appendix 5 for specimen data.

posterior portions. uniform medium brown. VFW' ground colour medium brown; basal

FEMALE: (Fig. 8F) Head, thoraxAnd abdomenasin male. Wings: two-thirds separated from distal third by sharp vertical boundar)',

forewing (length: 17.5 mm, n=2) similar to male. DFW and DHW basal two-thirds darker brown; minute jjaler submargiual dots iii

/. lies.Lejnd.

(i8

Fig. 12. Manerebia male genitalia, lateral view; v = valva ventral view, a = aedeagus lateral view, ad = aedeagus dorsal view.
M. inderena, M. golondrina n. sp., M. prattorum n. sp. See Appendix 5 for specimen data.

centers of cells (4i2-{'.nl and (ail -M3; thin distal marginal border
of sjtarse, pale greyish .scales. VHW niedinin brown; basal half with
sparse, long brown hairs and scattered with very sparse paler brown
scales; convex, tnieven, dark brown line tbrongb middle of discal
cell, fading near costa and anal margin; thin white postdiscal band
from apex to tornns, broadest in cells ('ai2-('ail to M3-M2, tapering
to a point at costa anti anal margin, lying distal of cliscocelltilars;
thin, stnootbly citrvitig darker brown snbmarginal litie, parallel to
distal margiti, from toi tins to neat apex; marginal border distal of
snbmargitial litie dusted with sitarse pale greyish scales.

Types: Holotypema\e: PERU. San Martin, Pnerta del Monte, ca.
fiO km NE Los Absos, 32.50 tn, 22.VI11. 198 1 , L. j. Barkley leg., MUSM;

Allotype female: ECUADOR: Loja, km 20 Jimbnra-San Andres rd.,
3300 m, 24. IX. 1997, K. Willmott leg, KWjH; Paratype male: PERU:
San Martt'ti, Parqtie Nacional Abiseo, Hnicimgo, Pnerta del Monte,
3190-3250 m, 19.V1I.1990, M. Medina leg, MUSM.

Etymology: The name of this species is the femitiine form tif
the Uatiti adjective “sediictiis”, meaning remote or distant, with
referetice to this species isolated range and rarit)'.

Comments: The two males of this species were collected in
the Central Cordillera in northern central Pern, whereas the only
ktiown female specimen was collected in the Cordillera de l.agn-
nillas iti extreme soittbern Ecuador. Manerebia seducla tints has a
wide, and as yet only rongblv defined range, ft is apparently a lower

39: 37-79, 2()()() (2006)

69

Fig. 13. Manerebia male genitalia, lateral view; v = valva ventral view, a = aedeagus lateral view, ad = aedeagus dorsal view.
M. trimaculata, M. undulata n. sp., M. interrupta, M. rufanalis n. sp. See Appendix 5 for specimen data.

paramo gra.ssland species, similar to A/, ignilineata, wltich occurs
sympatrically and much more abundantly in the same habitat.
The elongate wings of the female of this species are similar to M.
Iniana, and those of other paramo butterflies (e.g., the satyrine
genus Lymaiiapoda, see Pyrcz rl al, 1999), and may be an ada|>tation
to flying or resting in the strong winds that are fretiuent in these
high elevation grasslands.

Manerebia 7uycalesoides (C. & R. Felder, 1867)

Figs. 8H, l,'l4D, 20

Pronophila ni'icaksoidesV.. & R. Felder (1867: 473) . TL: Colombia,
Cimdinamarca, Bogota. ST male: BMNH(R) [examined]

= Euptychki iHhe Butler (1867: 46.6). TL: Venezuela. ST males:
BMNH(R) [examined]

“Penrosada" ktlie (Butler); D’Abrera (1988: 824, fig.).

Pedaliodes mycalesoides (C. & R. F’elder); Thieme (1905: 69).

Eiiptychia mycalesoides (C. & R. Felder); Weymer (1911: 224).

Postetiplychia mycalesoides Forster (1964: 137, fig. 171) (male
genitalia).

Manerebia mycalesoides (C. & R. Felder) ; Lamas & Viloria (2004:
215).

Diagnosis: Manerebia mycaksoides is easily distinguished from all
other congeners, except M. magnifica, by the large ventral ocelli on
both fore and hindwing and wavy, dark, prominent postdiscal line
on both VPAV and VHW (Figs. 8H, 1). The genitalia (Fig. 14D) are
distinctive in the elongate distal part of the valva which is strongly
curved upwards, ending with several short ‘teeth', being most
similar to M. magnifica and M. nevadensis. Manerebia magnifica is

distinguished under the account for that species.

Comments: Pronopbila mycalesoides was described from Bogota
(Colombia) (C. & R. Felder, 1867), but the ]trecise type locality' is un-
known and the species was not reported by Adams ( 1 986) . Shortly
afterwards in the same year the same taxon was also de.scribed as
Euptychia leihe by Butler (1867), from an unspecified Venezuelan
locality. We have examined the .syntypes of both names at the BMNl 1
and lelbe IS a junior subjective synonym of mycalesoides (l.amas &
V'iloria, 2004). Manerebia mycalesoides seems to be a veiy rare spe-
cies. It is found in lower cloud forest on the foothills of the Sierra
de El Tama at 1000 m. Nothing was known about the behavior or
ecology (tf this species until Andrew Neild (|5ers. comm.) observed
and collected it at l.oma del Viento, Tachira, in 1997, where it
flies in an open, windswe|rt area, in association with bamboo. Tbe
range of the species has been extended by recent collecting, with
records in the Venezuelan Cordillera de la Costa (San Antonio de
Los Altos, Colouia Tovar), Sierra de El Tama (San V'icente de l.a
Revancha, Chorro El Indio), and the west Colombian Rio Cauca
valley (l’o|tayan). Manerebia mycalesoides occurs in premontane rain
forest habitats at around 1000 m, where it is a shy inhabitant of
shadv places within dense forest. Usually only single individuals
are encountered, but Pierre Boyer (pers. comm.) observed on one
occasion a large group of over ten males in a forest clearing in the
Avila range above Caracas.

Manerebia magnifica Pyrcz & Williiiott, n. sp.

Eigs. 8G, 14E, 20

Diagnosis: I bis species differs from its closest relative, M.

70

/. Res.Lepid.

Fig. 14. Manerebia male genitalia, lateral view; v = valva ventral view, a = aedeagus lateral view, ad = aedeagus dorsal view. M.
ignilineata, M. seducta n. sp., M. mycalesoides, M. magnifica n. sp., M. nevadensis, M. levana. See Appendix 5 for specimen
data.

mycalesoides, in tlie larger size of the VHW suhmarginal ocelli, with
tin additional ocellus in cell (in I -M3, and in their shape being oval
instead of rounded. The distal tip of the valva is also less enlarged
and recurved.

Description: MALE (Fig. 8G): Head, frons with a tuft of black
hair; eyes black, smooth; labial |talpi covered with long, black hair;
antennae dorsally dark brown, ventrally chestntit, with white scales
at the base of each segment, club same colour as shaft. Thorax.
dorsal and ventral sitrface blackish brown; legs pale brown. Abdo-
men-. dorsal surface blackish brown, ventrally grey. Wings: forewing
(length: 23-24.5 mm, mean: 23.8 nun, n=3) distal margin straight,
apex rotmded; hindwing with distal margin vety slightly angled at
vein .M3, tornal notch absent. DEW uniform dark brown; andro-

conial scales not apparent; faint blackish submarginal line. DEfW
dark brown, slightly paler towards distal margin, faint blackish
submarginal line. VFVV' grotmd colotir meditim brown; indistinct,
darker brown discal line rtmning across discal cell, to base of vein
Cu2, continuing towards anal margin; postdiscal line of same
colotir, bent distally in cell 1A-Cti2 to join submarginal line at
tornus; submarginal line, and two dark brown thin marginal lines
straight and parallel to otiter margin; large (nearly width of cell),
rotmded, submarginal black ocelhis with a white pupil and ringed
with orange in cell (ai2-Cul; another submarginal ocellus, half
its size in cell M2-M1. VHW uniform medium brown; dark brown
discal line from costa to inner margin, roughly parallel to otiter
margin, passing through base vein Cu2; dark brown postdiscal line.

39: 37-79, 2000 (2006)

71

• leaena leaena
▲ leaena lanassa
■ leaena gonzalezi
0 germaniae germaniae
germaniae vitalei
□ pluviosa
apiculata

• franciscae franciscae
■ franciscae rodrignezi
^ mammuthiis
0 satura pauperata
ffl satura lamasi
▲ navarrae
c{)3 quinterae

Fig. 15. Locality records for Manerebia in the northern
Andes: M. leaena, M. germaniae, M. pluviosa and M.
apiculata

Fig. 16. Locality records for Manerebia in the northern
Andes: M. franciscae, M. mammuthus, M. satura, M.
navarrae and M. quinterae.

nearly straight, with the extremities ctirving slightly distally at apex
and tornus to merge with a thinner siibmarginal dark brown line;
siibmarginal line and two marginal lines parallel to outer margin;
a series ol black stibmarginal ocelli, four of them oval, two ocelli in
cells 1A-Cti2, one each in C]u2-Cnl and Ciil-M3, one rounded in
cell M2-M1, the biggest of which is ocellus in Cai2-Ciil, extends to
entire width of cell and nearly half of its length, remaining ocelli
diminishing in size anteriorly and posteriorly, with tornal ocellus
smallest; all ocelli ringed with orange, pupils in cells 1A-Cu2 and
Cu2-Cul white, those in cells Ciil-M3 and M2-M1 blue; blue sub-
marginal dot in cel! M3-M2. Male genitalia (Fig. 14E): tegumen
slender; uncus long and arched; subuncus rather short; extended
apical part of the valva strongly ctirved upwards, ended with two
‘teeth'; aedeagtis straight with a pronounced ‘collar’ at junction
of anterior and posterior portions

FEMALE; Lhiknown.

Types: Holotype male: PERLl: Amazonas, Cordillera del Con-
dor, alto Rio Comaina, PV22, falso Paqtiisha, 800 m, 25. X. 1987, G.
l-amas leg., MLISM; Paratypes: 2 males: same data as the holotype,
MUSM. '

Etymology: This species name is the feminine form of the Latin
acljective, “magnificus”, meaning magnificent, with reference to the
impressive submarginal ocelli on the ventral surface.

Comments: This species is clearly most closely related to M.

mycalesoide,s, but the large apparent range disjunction between the
two species and slight differences in the male genitalia stiggest the
two shotild be treated as distinct for the present. Manerebia magnifica
is known so far only from the eastern slopes of the Cordillera del
Condor in Perti, where it occurs in premontane forest. Its apparent
absence in the heavily collected Zamora valley further west suggests
it may be endemic to this isolated mountain range.

Manerebia nevadensis Kruger, 1925

Figs. 9A, B, 14F, 29

Manerebia nevadensis Kruger (1925: 25). TL: Colombia, Sierra
Nevada de Santa Marta, 2600m. LT male (designated by Pyrcz,
1999: 351): San Lorenzo, Sierra Nevada de Santa Marta, 06.1X.19I9,
2600m, E. Kruger leg. MZPAN [examined].

Manerebia nevadensis Kniger, Adams &; Bernard ( 1977: 273, lig.
17, male genit. fig. 5); Lamai & Viloria (2004: 215).

Diagnosis: Manerebia nevadensis superficially resembles some
subspecies of M. satura in wing pattern, but the male genitalia (Fig.
I4F) are strongly distinct from that sjtecies, instead indicating a
relationship with M. myealesoides AnA M. magnifica. AW three of these
species have a pronounced ‘collar’ at the junction of the anterior
and posterior portion of the aedeagtis, upturned tijt to the valva
and short subuncus, similar to certain southern Andean Manerebia

72

/. Ih'S.Lepid.

Fig. 17. Locality records for Manerebia in the northern
Andes: M. inderena, M. golondrina and M. prattorum.

(e.g., M. ryrlopiiiaSVdiuVmgev).

Comments; I'hi.s .sjjecie.s is endemic lo tlie Sierra Nevada de
Santa Marta, vvliere Adams & Bernard (1977) and Pvrcz (1999)
report tliat it occitrs Irotn 2,5()()-3()()() m. It is an elttsive bittterfly,
with a skipping flight, pret'erring to retnain inside dettse bamboo
clttmps atid seldom comitig to the edges.

Mnnen'bid lenana ((ioclnian, 1905)

Figs. 9C, D, 14G, 20

Lywanopada Codman (1905: 188. pi. 10, fig. 10). TL:

Colombia, Cimdinatnarca, Bogota. ST male; Colombia, Ctmdi-
namarca, Bogota. BMNIl [examined].

Penrosada Imana (Codman); Adams (1986: 307); Pvrcz (1999:
367).

Manerebia levann (Godman); Lamas & Viloria (2004: 215).

Diagnosis: This is a small, very distinctive species, .sti|3erficially
resetnbling only M. pemaga and to some extent A7. navarrae. The
VHVV postdiscal band is yellowish, indistinct, oblique and marked at
its distal edge by a thin, dentate atid dark brown line, dividing the
wing into an yellow-orange area basally and a chestnut area distally.
The ventral siibmarginal ocelli are mttch redttced, with only a small
one in cells 1 A-Cn2 on the VHW and VFM'. Further distingitishing
characters are discussed under M. pervaga. The male genitalia
resemble only tho.se of A/, peivaga. The.se two species are unique
within Manerebia in having a highly elongate distal portion of the
valva atid elongate aedeagns, bearing two dorso-lateral patches of
spines in the tniddle of the posterior section. The relationships of

Fig. 18. Locality records for Manerebia in the northern
Andes: M. trimaculata. M. undulata and M. interrupta

the sjiecies to other congeners are tmcertain. M. hvana is geograjihi-
callv variable and it remains possible that distinct sttbspecies will be
recognised in ftittire. Male genitalia as illustrated (Fhg. 14G).

Comments: .Mthoiigh Adatns (1986) reports this species from
Panama, based on specitnens iti the BMNH, these are definitely
tnislabelled. The species is known from both slopes of the Colom-
bian Cordillera Oi iental in the Bogota region (Cerro Monserrate)
f rom 27()()-33()() m, and Adams (1986) found it in jiaramo grassland
between bamboo-filled gullies at the tree-line. Kruger (1924) de-
scribed the female of this species (Pyrcz, 1999).

Manerebia pervaga Pyrcz & Viloria, n. sp.

Figs. 9E, F, 14M, 20

Manerebia n. .s]i. (Pvrcz & Vhloria) ; l.amas & Viloria (2004: 216,
n. I 121).

Diagnosis: This species differs from A7. levana in several char-
acters. The VHW is more utiiformly coloured in the male, while
the female is tmiformly brown, lacking the yellowish colouring
of A/, levana. Both sexes lack tornal ocelli on both wings and the
dark postdiscal line on the VHW is more basally |rositioned, be-
ing |)resent in cell 2A-Cit2 (absent in M. levana). The female has
a strongly dentate thin brown marginal line that is more basally
irositioned, and the lot ewing, and to a lesser extent the hindwing,
are mttch more rottiided than in M. levana. Finally, both sexes
have a dark discocellular streak between the bases of veins M2 and
Ml on the hindwing (also on the forewing in the female) that is
uniqtie in tlie genus.

Description: MAl.E (Fig. 9E): Head: frons with a tuft of dark
brow'll bair; eyes dark coffee brown, smooth; labial palpi slightly
longer than head, with light brown and black hairs; antennae with
club twice as broad as sliaft, orange brown, darker dorsally. Thorax.
moderately hairy, more densely on ventral surface, dorsally black-
ish brown, ventially brown; legs medium brown. Abdomen: dark
brown, lighter on ventral surface, especially at posterior tip. fTings.-
forewing (length; 16.5-17.5; mean: 17 mm; n=3) triangular, tornus

39; 37-79, 2000 (2006)

73

Fig. 19. Locality records for Manerebia in the northern
Andes: M. rufanalis

obtuse; hindwing rounded, tornirs moderately pointed, anal niargiti
straight; dorsal surface of both wings hairy in basal half and along
anal margin. Dorsal surface ground colour coffee brown; diffuse
orange patch on DHW tornns. VFW ground colour chocolate
brown, slightly darkened at base; costal margin lighter; distal mar-
gin, apical, and sitbapical region ochraceous brown (with reddish
tone towards atiterior edges) , as well as base of costal margin region.
VHW ground colour yellowish brown, darker at basal region; light
suffusion of orange towards posterior postdiscal region, forming
diffuse irregular wedge bordered distally with thin dark brown den-
tate postdiscal line, fading anteriorly and posteriorly and parallel to
distal margin, in cells 2A-Cu2 to M2-M1. Male genitalia (Fig. 14H):
similar to M. Itniana v/nh elongate distal tip to valva, except dorsal
‘teeth’ at distal tip of valva more extensive, extending anteriorly;
aedeagus similar to M. levaria, elongate, with two dorso-lateral
patches of spines in middle of posterior section.

FEMALE (Fig. 9F'): Head, frons with a tuft of brown hair; eyes
medium brown, smooth; labial palpi 2.5 times length of head, with
long brown hair; antennae with club three times as thick as basal
segments, shaft dorsally and ventrally light brown, club vetitralh
orange, dorsally blackish brown. Thorax, moderately baity, more
densely on ventral surface, dorsally blackish brown, venti ally brown;
legs medium brown. Abdomen: dorsally blackish brown, laterally and
ventrally medium brown. W7«gs; forewing (length; 16.5-17.5 mm;
mean; 17 mm; ti=2) costa slightly arched, apex bhtnt, distal margin
and tornus rounded, Hitidwing overall triangular, with apex, tornus
and margins smoothly rounded. Dorsal surface unilorm medium
brown; fringes light grey. VFW greyish brown, darker brown in
basal half; darker brown streak over discocellulars between bases
of veins Ml and M2. VHW greyish brown, darker brown at base
and posterior of discal cell; darker brown streak over discocellulars
between bases of veins Ml and M2; postdiscal dark brown line
composed of lunular streaks incurved basally in cells M2-M1 to
1A-Cit2, approximately jrarallel to distal margin; faint dark brown

Fig. 20, Locality records for Manerebia in the northern
Andes: M. ignilineata, M. seducta, M. mycalesoides, M.
magnifica, M. nevadensis, M. levana and M. pervaga.

suhmarginal 'V’-shaped streaks in cells Ml-Rs to IA-Ctt2.

Material examined: Holotype male: VFINEZLIELA: Tachira,
Parqiie Nacional El Tama, Venezuela, 3100-3350 m, 16-18.11.1992,
A. Viloria &J. Camacho leg., MAI.UZ; A/fo/y/te female: COLOMBIA:
Norte de Santander, Cerro Orotjtte, 3850 m, 12.VI.1965, j. Bechyne
leg, MIZA; Paratypes (3 males) : VENEZLT'TA: 1 male: same data as
tl'ie holotype, MALUZ; 1 male: Paramo El Tama, 2600m, 12.11. 1983,
M. Vivas leg, MIZA; COLOMBIA: 1 male: Norte de Santander,
Paratno El Tama via Henan , XII. 1993, J. F. Le Crom /eg., JFLC: 1
female: same data as allotype, MIZA.

Etymology: The name is the feminine form of the Latin adjec-
tive “pervagus”, meaning wandering, in reference to the flight of
this species in the paramo grassland.

Comments: Manerebia pewaga is known from the Sierra de El
Tama on the Venezuela/Colomhia border and the Cerro Oroque
iti Norte de Santander in Colombia. It occitrs in o|)en paramo
covered with low growing fl/ttcvr/Mra thickets from 3200-3850 m. This
is higher than that reported for any other congeners, even paramo
species like M. seducta, M. ign Hi neat a and M. Iniana. Manerebia per-
vaga shares with the last two of these species its small size and dull,
ciyptic colours, which may be adaptations to paramo habitats and
not necessarily indicate any close affinity. Phis species seems to be
seasonal and has only been found Hying during the diy season, on
days with bright, direct sunlight.

74

/. lies.Lepid.

Acknowledgements

We thank l’hili|) Ackeiy for giving ns access to the collections at
the BMNH and lor permitting ns to make crucial dissections of type
material. We thank Gerardo Lamas for photographs and informa-
tion for a number of type s|5ecimens, permission to examine and
loan of material from the MLkSM, and Pierre Boyer, Andrew Neild,
Gabriel Rodn'gttez, Artur Jasihski, Jean Francois Le Crom, Piotr Krol
and Piotr l.os for additional material. We are grateful to Manrizio
Bollino for jiroviding certain genitalia drawings and information.
I'WP thanks [annsz Wojtnsiak, Szczepan Bilihski and Rafal Garlacz,
for supporting in many ways the research on Andean Le|)idoptera
carried out at the Zoological Miisetim of ihejagiellonian Lhiiversity
in Krakow and for their company in Ecuador, Colombia, Venezuela
and Peru. We also thank the late Dr. Alvaro Jose Negret, director
ol the Mnseo de Historia Natural de la Universidad del Caitca in
Popayan fbi his co-operation in Colombia. We thank F.liza Manteca,
Piet Sabbe atid 1 larold Greeney for accommodation and permission
to collect in their private reserves. Las Golondrinas and Yanayacit
res|tectively. Permits for research and collection in Ecuador were
provided by INFT’AN and tbe Ministerio del Ambiente, Direccidn
de Bosqites y de Areas Naturales Protegidas, through the Mttseo
Ecuatoriano de Ciencias Naturales in Qtiito, with the helj) of Ger-
matiia Estevez and Marfa de los Angeles Simbaha, and tlnoitgh the
Pontilicia Lhiiversidad (iatolica, with the helj) of Giovanni Onore,
who also provided institutional support for I'WP. lustitittional
sitpport in Pent was provided by Gerardo Lamas and collecting
permits weie gratited by INRF.NA. Fieldwork of TWP in Colombia
in 1997 atid Ffcnadttr in 1998, 1999, 2002 and 2003 was stt|5]5orted
by research gratits of the Institute of Zoology of the Jagiellonian
I'niversity BW atid the Polish Committee for Scientific Research
(KliN Grant 014()/P04/2003/2). The followitig also assisted KRW'
and |PWH with the costs of held work in Ectiador: (1993-94) Mr.
I Willmott. Mrs. M. Willmott, Christ’s College Cambridge L'niv.,
.\lbert Reckitt Charitable Trtist (C. T), Poulton Fund Oxford L'niv.,
Round Table Trttst, Lindeth C. T, Catherine Cookson Fotmdation,
Morton C. T. Royal Entomological Society, Butler C. T, Mr. D. Exell.
Peter Nathan C. T. ffarry Crook Fotitidatioti, Douglas Heath Fives
C. T, R. & M. Foreman C. T, Northern Bank, Banbridge Academy,
C. Britce, Hickley Valtone Ltd., Vera Trinder lad., Agfa. Phoenix
Mottntaineering, Balfonr-Browne Ftttid, Worts Fund (KRW), Sigma
Xi the Scientific Research Society (JPWH, 1995-6; KRW, 1996) and
Eqtiafor; field and musenm research iti 1997-2000 was funded by a
National Geographic Society Research and Exploratioti Grant (No.
5751-96), and from 2002-2004 by the National Science Foundation
(BS&l gratit#0l0374(i).

Literature cited

Aiiam.s, M. |. 1985. Speciation in the Protiophiline Butterflies
(Satvridae) of the Northern Andes. Jtiitrnal of Research on
the Lepidoptera, 1985, Stip|)lement No.L 3,3-49.

. 1986. Pronophilitie Initterflies (Satyridae) of the three

Andean Cordilleras of Colombia. Zoological Journal of the
Linnean Society, 87: 235-320.

Ati.vMS, M. J., & C. 1. Bern.vro. 1977. Protiophiline butterflies
(Satyridae) of the Sierra Nevada de Santa Marta, Colombia.
Systematic EtUotiiology, 2: 26,3-281.

. 1979. Protiophiline butterflies (Satyridae) of the Serranfa

de Valledttpar, Colombia-Venezttela border. Systematic
Entomolog)', 4; 95-1 18.

. 1981. Pronophiline butterflies (Satyridae) of the Cordillera

de Merida, Venezuela. Zoological Journal of the Linnean
Society, 71: 343-372.

Brown, F. M. 1944.- Notes on Ecuadorian butterflies. IV. The

genus Penrosada, new (L.epifloptera, Satyridae). Annals of the
Fintomological Society of America, 37: 25,5-260.

Butler, A. C. 1867. A monograph of the genus Euplychia, a
numerous race of butterflies belonging to the family Satyridae;
with descriptions of sixty species new to science, and notes to
their affinities, etc. Proceedings of the Zoological Society of
London, 1866(3): 484-504, pis. 39-40.

CoM.STOCK, J. H., & J. G. Nekdh.am. 1918. The wings of insects.
American Naturalist, 32: 253-257.

D’AitRKR.\, B. 1988. Butterflies of the Neoptropical Region, part
V, Nymphalidae (Concl.) & Satyridae. |t|). 680-887. Victoria:

1 fill House.

Diit.oN, M. O. 1994. Bosques humedos del norte del Peru. Artialdoa
2(1): 29-42.

Do(,nin, P. 1896. I.ejtidopteres notiveatix de Loja of London, (B)
12:23-.30.

Hkwit.son, W. C. 1861. De.scri|)tions of new diurnal Lepidoptera.
Journal of Entotnology, 1(3): 155-158.

. 1870. Descriptions of twenty-two new species of Equatorial

Lepidoptera. Transactions of the Entomological Society of
l.ondon. 1870(2): 153-163.

KiRiiv, W. F. 1879. Catalogue of the Collection of Diurnal
Lepidoptera Formed by the I.ate William Chapman Hewitson
of Oatlands, Walton on Lhames; and Bequeathed by him to the
British Mtiseum. London: John Van Voorst, iv-i-246 pp.

Kiors, A. B. 1956. Lepidoptera, pp. 1 15-130. In: Tnxen. S. L. (ed.).
Taxonomists’ Glossary of Genitalia in Insects. Copenhagen:
Munkssgard.

Kruofr, E. 1 924. Beitrage zur Kenntnis der columbischen Satyriden.
Entomologische Rundschati, 41: 7,9-10, 16, 19-20,23-24,27-28,
31-32, .35, 38-39, 41-42, 46-47.

. 1925. Beitrage zur Kenntnis der columbischen Satyriden.

Entomologische Rtmd.schau, 42(3): 10-12.

L.A.MAS, C., & A. L. Vii.ORiA. 2004. Nymi^halidae. Satyrinae. I’ribe
Satvrini. Subtribe Erebiina, it|3. 2L5-216. In: l.amas, G. (Ed.),
Checklist: Part 4A. Hesperioidea-Pa|5ilionoidea. In: Heppner,
|. B. (Fid.), Atlas of Neotro|jical Lepidoptera. Volume 5A.
Gainesville, As.sociation for Tropical l.epidoptera; Scientific
Publishers.

Miller, L,. D. 1968. The higher classification, phylogeny and
zoogeogra|3hy of the Satyridae (Lepidoptera). Memoirs of the
American Entomological Society, 24: [6] + iii + 174 pp.

fb'RCZ, T.W. 2004 Pronophiline butterflies of the highlands of
Chachapoyas in northern Peru: faunal survey, diversity and
distribution patterns (Lepidoptera, Nym|thalidae, satyrinae).
Genus, 15(4): 45.5-622

FhRCZ, T. W. 1995. A new genus, and a new species, Tamania

jacquelinae, from the Tama range, Venezuela Colombia border,
and some thougbts on the diagnosis of tbe tribe Pronophilini
(Nvmphalidae: Satv'rinae). Lambillionea, 99(4): 519-525.

. 1999. The E. Kriiger collection of pronophiline

butterflies. Part 11: genera Manerehia to Thiemeia (Lepidoptera:
Nympbalidae: Satyrinae). Lambillionea, 99(3): 351-376.

Pyrcz, T. W., WiLLMon , K. R., &J. P. W. H.vi.i,. 1999. Contribution to
tbe knowledge of Ecuadorian Pronophilini. Part HI. Three new
species and live new subspecies of Lymanopoda (Lepidoptera:
Nymphalidae: Satyrinae). Genus, 10(3): 497-522.

INrcz, T. W., & A. ViLORtA. 2006. Cloud-forest satyrine butterflies
in the Sierra El Tama, Venezuela - Colombia border. Tropical
Lepidoptera, [in press].

R.\c:heli, T, & L. Ru:iiell 2001. An annotated list of Ecuadorian
butterflies (Lepidoptera: Papilionidae, Pieridae, Nvinphalidae).
Fragmenta entomologica, .33(2): 213-380.

ScHAUs, W. 1902. Descriptions of new American btitterflies.
Proceedings of the Lhiited States National Museum, 24(1262):
38.3-460.

.SiAUniNGER, O. 1897. Netie sudamerikanische Tagfalter. Detitsche

39; 37-79, 2000 (2006)

75

Entomologische Zeitschrift “Iris”, 10(1): 123-151.

Thieme, T. a. O. 1905. Monograi)hie der Gattung Fedaliodes
Biitl. (Lepidoptera Rhopalocera. Satyridae). Berliner
Entomologische Zeitschrift, 50(1/2): 43-141.

Vii,()Ri.\, A. L. 2001 . Studies on the systematics and biogeography of
sttme montane satyrid butterllies (Lepidoptera). Unpublished
Ph. D. dissertation.

Weymer, G. 191 1-1912. 4. Eamilie: Satyridae. In: Seitz, A. (ed.). Die
Gross-Schmetterlinge der Erde. Stuttgart; A. Kernen. 5: 224,
pis. 52, 53 (1911), 248-2,50 (1912).

APPENDIX 1. Errors in main references on north Andean Manei'ebia
Brown (1944):

Penrosada leaena = Manerebia mderena Imeniva
Penro.sada apiculata = Manerebia interrupta
Penrosada lanassa = Manerebia ignilineata

D'Abrera (1988):

Penrosada kaena (dorsal surface) = Manerebia undulata

Penrosada leaena (ventral surface) = Manerebia n. sp., Peru (Pyrcz, in prep.)

Penrosada lanassa (dorsal surface) = Manerebia rufanalis rufanalis

Penrosada lanassa (ventral surface) = Manetebia satura sainra

Penrosada sp. = Manerebia satura pauperata

Euptychia jovita = Manerebia satura pauperata

Penrosada lethe = Manerebia mycalesoides

Adams (1986);

Penrosada mderena male holotype = Penro.sada inderena female [taratype

APPENDIX 2. Distribution of taxa along hypothetical elevational transects

Approximate
elevational range

Ecuador, Zamora-
Chinchipe, E. slope

Ecuador, Pastaza, E. slope

Colombia, Choachi,

E. cordillera, E slope

Colombia, El Tama

E. cordillera, NE tip

1000-1 400m

magnifica

_

mycalesoides

mycalesoides

1400- 1800m

sahira pauperata

satura pauperata

franciscae franciscae

fra n ciscae pan ciscae

1800-2200m

rufanalis rufanalis

ru fa n a Us rufan alis

-

-

2200-2600m

trimandata

-

-

-

2600-3000m

inderena mirena

i n deren a leaen iva

leaena leaena

leaena gonzalezi

2800-3200m

geiinaniae vitalei

gemaniae germaniae

apiculata

plinuosa

3000-3400m

ignilineata ignilineata

ignilineata neglerta

leuana

peraaga

3200-3600m

seducta

-

-

pemaga

Total species

8

5

5

5

APPENDIX 3.

Distribution of taxa by country

Species

Subspecies

Ecuador

Colombia

Venezuela

North Peru

kaena

leaena

+

.

-

lanassa

-

+

-

-

gonzalezi

-

+

+

-

geiinaniae

germaniae

+

+

-

+

vitalei

+

-

-

-

pluviosa

-

+

+

-

apiculata

-

+

-

-

navarrae

-

+

-

golondrina

+

+

-

-

satura

pauperata

+

-

-

-

lamasi

+

-

-

+

76

/. Res.Lepid.

APPENDIX 3 (Com)

Species

Subspecies

Ecuador

Colombia

Venezuela

North Peru

mammuthus

+

+

-

-

1 l int rise (le

Imnciscae

-

+

-

rodrigitezi

-

+

-

-

igniUneata

igiti lilted la

+

-

-

+

itegleila

+

-

-

-

inderena

iiidereita

-

+

-

-

aitliiiijtiiana

-

+

-

-

leaenivd

+

-

-

-

(laid

+

-

-

-

siiiiilis

+

-

-

-

pita

+

+

-

-

mirena

+

-

-

+

jimllorum

-

-

-

+

trimandata

-

-

+

inidulala

tt ltd Ilia la

+

-

-

-

iiidaena

+

-

-

+

inlirruptii

+

-

-

+

nipnifdis

nilaiidlis

+

-

-

4-

jiTitaitdiita

+

-

-

-

quintmie

-

+

+

-

nnmdensis

-

+

-

-

In'inin

-

+

-

-

pm' ago

-

+

+

-

myiali’soidt's

-

+

+

-

mdgnijka

+

-

-

+

si’diicla

+

-

-

TOTAL 23

37

spl2/ssp22

spl4/ssp20

sp7/ssp7

spll/sspll

APPENDIX 4. Figured specimens

Species

Subspecies

Locality

Sex

Type

Coll.

Fig.

leaena

leaena

Colombia, (iundinamarca: Ciiasca-Gacheta

,M

BMNH (A&B)

lA

leaena

Colombia, Cimdiiiamarca; “Bogota”

F

TWP

IB

laitassa

Colombia, Boyaca: \V below Arcabiico

M

BMNH (A&B)

1C

lanassa

Colombia, Boyaca: W below Arcabuco

F

BMNH (A&B)

ID

gonzalfzi

Venezuela, I'acbira: Siena de El 'Fama

M

PT

TW'P

IE

germaniae

germaniae

Ecuador, Cotopaxi: Pilalo

M

HT

MZUJ

IF

germaniae

Ectiador, Cotopaxi: Pilalo

F

AT

TW’P

1C

vilalei

Ectiador, Zamora: Loja-Zamora old rd.

M

HT

MZU)

IH

plmdosa

Venezuela, Tacbira: San Vicente de la Revancha

M

PT

TWT

2A

apiculata

Colombia, Cundinamarca: “Bogota”

M

BMNH

2B

apictdala

Colombia, Cundinamarca: “Bogota”

F

MIZPAN

2C

fraiiciscae

franiiseae

Veneztiela, Merida: above La Montana

M

P4’

BMNH (A&B)

2D

franciscae

Venezuela, Merida: above La Montana

F

PT

BMNH (A&B)

2E

wdiiguezi

Colombia, Antioquia: Cuarne

M

HT

MZUJ

2F

ludrignezi

Colombia, Antioquia: El Retiro

F

AT

TMT

2C

mammuthus

Ecuador, Sucttmbi'os: El Higtierdn

M

HT

AME

2H

satura

sal lira

Perti. latino: Carabaya, Santo Domingo

M

BMNH(R)

3A

lamasi

Peru. Amazonas: AUbnso Ugarte

M

HT

MUSM

3B

pduperala

Ecuador, Zamora: Loja-Zamora rd. km 40

M

H'F

MZUJ

3C

paiipiTala

Ecuador, Zamora: Lxtja-Zamora rd. km 40

F

AT

TWP

3D

navarrae

Colombia, Cesar: S. de Valledtipar, Finca Altamira

M

PT

BMNH (A&B)

3E

quinterae

Venezuela, Ziilia: E. above Manaure

M

PT

BMNH (A&B)

3F

iiiderena

iiiderena

Colombia, Tolima: S above Cajamarca

M

HT

BMNH (A&B)

3G

iitderena

Colombia, Tolima: S above Cajamarca

F

PT

BMNH (A&B)

3H

aiiltoquiana

Cokmibia, Antioquia: El Retiro

M

HT

MZUJ

4A

anlioqiiiana

Colombia, Antioqtiia: El Retiro

F

AT

TWP

4B

39: 37-79, 2000 (2006)

77

APPENDIX 4 (Cont)

Species

Subspecies

Locality

Sex

Type

Coll.

Fig.

firm

Ecuador, Pichincha: Aloag-SaiUo Domingo rd.

M

PT

KWjll

4G

fin a

Ecuador, Pichincha: Volcan Pasoclioa

F

PT

Kwjll

4D

sitnilis

Ecuador, Bolivar: Balzapamba, arriba de Sta. Lucia

M

HT

MZUJ

4E

similis

Ecuador, Azuay: Molleturo

F

PT

FWP

4F

clara

Ecuador, Napo: Baeza

M

PT

I’WP

4G

clam

Ecuador, Napo: Hda. San Isidro

F

AT

PB

4H

mireyia

Ecuador, Loja: Cajanuina

M

PT

TWP

5A

mirena

Ecuador, Loja: km 95-100 L(ja-Zumba rd.

F

AT

PB

5B

leaeniva

Ecuador, Tungiirahua: El Tablon

M

PT

TWP

5G

golondrina

Ecuador, Carchi: Santa Rosa, Las Golondrinas

M

PT

KWjH

5D

prattorum

Peru: “West slopes of Andes”

M

P'F

BMNH(R)

5E

trimaculata

Ecuador, Zamora-Ghinchipe: Rio San Francisco

M

FWP

5F

Iriniaculata

Ecuador, Zamora-Ghinchipe: Loja-Zamora rd.

M

FWP

5G

trimaculata

Ecuador, Morona-Santiago: km 37 Limon-Gualaceo

F

KWJH

514

undulata

undulata

Ecuador, Pichincha: Tandapi

M

PT

KWJH

6A

undulata ?

Ecuador, Loja: Yaugana, Valladolid rd.

M

d’WP

6B

undulata

Ecuador, Bolivar: Balzapamba, arriba de Sta. l.ucia

F

AT

TVM’

6C

rnilaena

Ecuador, Loja: Jimbura-Laguna Negra rd.

M

HT

MZUJ

6D

interrupla

Ecuador, Azuay: Sigsig-Granadillas

M

TWP

6E

intmrupa

Ecuador, Azuay: Sigsig-Granadillas

M

TWP

6F

interrupta

Ecuador, Azuay: Sigsig-(iranadillas

M

TWP

6G

intrrrupa

Ecuador, Azuay: Sayausi

F

TWP

6H

rufanalis

rufanalis

Ecuador, Tungiirahua: Rimtiin

M

H'F

MZUJ

7A

rufanalis

Ecuador, Zamora-(ihinchi|te: Loja-Zamora old rd.

F

AT

MBLI

7B

feniandina

Ecuador, Giron: San Fernando

M

HT

MZUJ

7G

ignilineata

ignilineata

Ecuador, Moroua-Santiago: Gualaceo-Limou rd.

M

FWP

7D

ignilineata

Ecuador, Loja: Loja

M

’FWP

7E

ignilineata

Ecuador, Loja: Loja-Zamora rd.

M

'FWP

7F

ignilineata

Ecuador, Morona-Santiago: Gualaceo-Limou rd.

F

TWP

7G

neglectn

Ecuador, Bolivar: Pilalo

M

TWP

8A

neglecta

Ecuador, Ganar: Zhud

M

TWP

«B

neglecta

Ecuador, Bolivar: Pilalo

M

TWP

8C

neglecta

Ecuador, Cahar: Zhud

F

TWP

8D

seducta

Peru, San Martin: Puerta del Monte

M

HT

MUSM

8E

seducta

Ecuador, Loja: Jimbura-San Andres rd.

F

AT

KWJH

8F

magnifica

Peru, Amazonas: Falso Paquisha

M

HT

MUS.M

8G

mycalesoides

Venezuela, Miranda: Altos de Pipe

M

AFEN

814

mycalrsoides

Venezuela, Tachira: Loma del Vieuto

F

AFEN

81

nevadensis

Golombia: S. Nevada de Santa Marta, El Gampano

M

BMNH(A&B)

9A

nevadensis

Colombia: Sierra Nevada de Santa Marta

F

PIT

MIZPAN

9B

levana

Colombia: “Interior of Colombia”

M

BMNH

9G

Imana

Colombia, Tolima: Rio Chili

F

BMNH

911

pervaga

Venezuela, Tachira: Sierra de El Tama

M

HT

MALUZ

9E

peraaga

Colombia, Norte de Santander: (ierro Oroque

F

P'F

MIZA

9F

APPENDIX 5. Male genitalic dissections

Species

Subspecies

Locality

Type

Coll.

Dissection# Fig.

leaena

leaena

Colombia, Cundinamarca: “Env. Bogota”

BMNH

6443

leaena

Colombia, Cundinamarca: Ghoachi

IWP

02/02.05.1999 lOA

leaena

Colombia, Cundinamarca: Ghoachi

B.MNH

6441

leaena

Colombia, Cundinamarca: Guasca-Gacheta

BMNH

6442

leaena

Colombia, Boyaca: Sierra Nevada del Gocuy

BMNH

leaena

No data: “PEcuador/Quito”

IT

BVINH

lanassa

Golombia, Cundinamarca: Facatativa

FWP

04/07.04.1999 lOB

lanassa

Colombia: Santander

TM'P

lanassa

Colombia, Cundinamarca: “Bogota”

HT

BMNH

29877

gonzalezi

Venezuela, Tachira: Sierra de Fd Tama

PT

FWP

05/07.04.1999 IOC

78

/. lies.Lepid.

APPENDIX 5 (Cont.)

Species

Subspecies

Locality

Type

Coll.

Dissection#

Fig.

germaniae

germaniae

C()loml)ia, C’aiica: Purace

PT

TWP

germaniae

Ecuador, I’ichiiicha: Quito-Sto. Domingo old rd.

PT

KWjH

PENROS 12

germaniae

Ecuador, Pichiiicha: Yanacocha

P'P

KWJH

germaniae

Ecuador, Pichiiicha: above (ihiriboga

PP

BMNH

645 1

germaniae

Ecuador, Cotopaxi: Pilalo

PI

TWP

01/26.01.1999

lOD

germaniae

Ecuador, Tuiigurahua: El Pablon

PT

MBLl

vitaln

Ecuador, Zainora-Chincliipe: Ltija-Zamora rd.

PT

MBLI

2003-12-23

lOE

pliwiosa

Venezuela, Tachira: Sierra de El Pama

PT

TWP

lOF

apiculata

Colombia, Cundinamarca: Zipacjuira

TWP

05/30.03.1999

llA

fraticiscae

franciscae

Venezuela, Merida: .Merida

PT

BMNH

6436

[mnciscae

Venezuela, Tachira: Sierra de El Tama

TWP

01/02.05.1999

IIB

jmnciscae

Colombia, Cundinamarca: Pacho

PWP

mdriguezi

Colombia, Antioquia: El Retiro

PI’

'PWP

01/21/12/2003

lie

mammiithus

Ecuador, Sttctttnbt'os: El Iligtterdn

HT

KWIPl

PENROS 8

1 ID

mammuthiis

Ecuador, Sttctttnbios: El Higtterdn

PT

KW'fH

MAN 1

safura

pauperata

Ecuador, Najio: Sierra de los 1 hiacamayos

PT

TWP

02/31.03.1999

HE

pfiuperata

Ecitador, Pastaza: “Env. d'Ambato”

P'P

BMNH

6455

paupmUa

Ecuador, Zatnora-Chincbipe: Loja-Zamora rd.

PT

KWIH

PENROS 9

lamasi

Peru, Ama/ona.s: Alf'on.so Ugarte

PT

MUSM

06/27.04.2002

HE

ssp

Peru, Cajamarca: La Balsa

MUSM

satin a

Peru, Puno: Santo Domingo

BMNH

6454

IIG

Saturn

No data: “Cauca valley, Colombia” - error

BMNH

6437

fiavatTae

Colombia, Cesar: Sierra de Perija

P'P

BMNH

29913

IIH

quiiiterae

Colombia, Cesai : Sierra de Perija

P'P

MA

(punteme

Venezuela, Ztilia: E. above Manatire

P'P

BMNH

6445

111

inderena

inderena

Colombia, Caldas: Paramo de Letras

TWP

inderena

Colombia, Tolima: S. above Cajamarca

BMNH

6453

inderena

Colombia: Ruiillo

TWP

09/30.03.1999

12A

inderena

Ecttador, Suctimbios: Qbda. de Piedras

KWjH

PENROS 13

inderena

Ecttador, Suctimbios: El Higueron

Kvyjn

PENROS 19

antiotpiiana

Colombia, Antioquia: El Retiro

PT

PWP

05/11.12.2003

12B

Jina

ticuador, Pichincha: Nono-Nanegalito rd.

PT

PWP

03/31.03.1999

12C

pna

Ecttador, Pichincha: Qbda. Molino

PT

KWjH

PENROS 3

fin a

Ecuador. Pichincha: .-Moag-Sto. Domingo rd.

P'P

KWIH

PENROS 6

fina

Ecuatlor, Pichincha: ,Aloag-Sto. Domingo rd.

PT

KvyiH

PENROS 4

fin a

Ecuador. Pichincha: Volcan Pasochoa

PT

'nvp

similis

Ecuador

BMNH

6439

similis

Ecuador, Bolivar: Balzapamba

P'P

lAVP

04/25.01.1999

121)

clara

Ecuador, Najio: Rio Chonta

P'P

KWJH

clara

Ecuador, Napo: Baeza

PP

TWP

06/25.01.1999

12E

leaeniva

Ecttador, Tuiigurahua: Volcan Ttingurahua

PP

TW'P

08/07.04.1999

12E

leaeniva

Ecuador, Tuiigurahua: Rio Verde Chico

PP

TWP

leaeniva

Ecuador, Tungurahua: Banos

P'P

BMNH

6457

leaeniva

Ecuador

PP

BMNH

6456

leaeniva

Ecuador, Mtirona-Santiago: Limon-Gualaceo rd.

KVyjH

PENROS 2

mirena

Ecuador. Zamora-Chinch.: limbura-S. Andres rd.

PP

KWJH

PENROS 14

mirena

Ecuador. Zamora-Chinch.: Jimbura-S. Andres rd.

PP

'TWP

mirena

Ecuador, Zamora-Chinchipe: Valladolid

PI

MBLI

mirena

Ecuador, Zamora-Chinchipe: Valladolid

PP

MBLI

mirena

Ecuador, Loja: Cajanuma

PP

TW'P

01/31.03.1999

12G

mirena

Peru, Cajamarca: Tabaconas

PT

TWP

mirena

Peru, Cajamarca: Hacienda Udima

P'P

MUSM

golondriiia

Ecuador, Carchi: Santa Rosa, Res. Goloiidrinas

PT

KVyjH

PENROS 10

12H

prattorum

Peru: “W. Slopes ol' Andes”

PT

BMNH

6448

pmUorum

Peru: “W. Slopes of Andes”

PT

TWP

03/02.05.1999

121

praltorum

Peru. Piura: Canchaque

PT

MUSM

trimacidata

Ecuador, Zamora-Chinchipe: Loja-Zamora rd.

TWT

03/07.04.1999

13A

trimanilala

No data

BMNH

6447

undidata

undidata

Ecuador, Loja: Giiayquichuma

PT

TWP

09/07.04.1999

13B

undidata

Ecuador, Pichincha: Rio Las Palmeras

PT

KWJH

PENROS 7

undidata

Ecuador, Pichincha: Tandapi

P'P

KWJH

PENROS 18

undidata

Ecuador, Cotopaxi: Pilalo

PT

milaena

Ecuador, Loja: Jimbura-San Andres rd.

PT

TWP

03/25.01.1999

13C

39: 37-79, 2000 (2006)

79

APPENDIX 5 (Cont.)

Species

Subspecies

Locality

Type

Coll.

Di.ssection#

Fig.

milaena

Peru, Cajaniarca: Hacienda Udinia

MUSM

interrupta

Ecuador, Azuay: Gualaceo-Cliiguinda rd.

KW|11

PENROS 1.6

inteiTupta

Ecuador, Loja: Loja-Ciuenca rd.

TMd’

()4/3().()3.1999

LSD

rufaiialis

rufanalis

Ecuador, Tuiigurahua: Rio Machay

PT

KWJH

PENROS 1

rufanalis

Ecuador, Loja: (iajanuma

PT

TWP

rufanalis

Ecuador, Zamora-Chincliipe: Qbda. San Ramon

PT

KWjH

PENROS 17

rufanalis

Ecuador, Zamora-Chinch.: )imbura-S. Andres rd.

PT

KW|H

PENROS 16

rufanalis

Ecuador, Zaniora-Chincliipe: “Zuniba”

PT

TWP

rufanalis

Ecuador, Zamora-Cihinchipe: San Ancire.s

PT

TWP

rufanalis

Peru, Cajaniarca: Tabaconas

PT

TWP

07/31.03.1999

13E

femandina

Ecuador, Azuay: Giron

PT

TWP

10/25.01.1999

13F

ignilineata

ignilineata

Ecuador, Loja: Jimbura-San Andres rd.

KWjH

PENROS 20

14A

neglecla

Ecuador, Cotopaxi: Pilalo

TWP

10/07.04.1999

14B

seducta

Peru, San Martin: Abiseo

PT

MUSM

01/22.06.2002

14C

mycalesoides

Venezuela, Barinas: Uribante

JFLC

74/1996

14D

mycalesoides

Venezuela, Tacbira: Sierra de El Tama

TMT

mycalesoides

Venezuela, Miranda: Altos de Pipe

AFEN

NEILD 01

magnifica

Peru, Amazonas: Falso Paquisba

PT

MUSM

07/27.04.2002

14E

nevademis

Colombia: Sierra Nevada de Santa Marta

MA

nniadensis

Colombia: Sierra Nevada de Santa Marta

BMNH

6444

14F

levana

Colombia, Cundinamarca: “Bogota”

TWP

levana

Colombia, Cundinamarca: “Env. Bogota”

BMNH

644(i

14G

pervaga

Venezuela, Tacbira: Sierra de El Tama

HT

MALUZ

14H

journal of Research on the Lej)idol)l('ra

39; 80-85, 2000 (2006)

Phylogenetic, habitat, and behavioural aspects of possum behaviour
in European lepidoptera

Tim G. Shref:vp:

Oxford Brookes University, De|)artinent of Biological and Molecular Sciences, Headinglon, Oxford 0X3 OBP, U.K.
lgshrepiie@bronkes. nc. uk

Roger L.H. Dennis

Oxford Brookes L'niversity, De|rartinent of Biological and Molecular Sciences, Headington, Oxford 0X3 OBP, U.K.
NERC Centre for Ecology and Hydrology', Monks Wood, Abbots Ripton, Huntingdon, Cainbridgesbire PE28 2ES, U.K.
rlhden nis@aol. com

Andrew Waki-.ham-Dawson

rite International Coininission on Zoological Nonienclatnre, c/o The Natural History Mtisentn, Cromwell Road, London SW7 5BD,
U.K.

andrw@nhm.ac. ult

Abstract: We describe the behavioin of |tlaying possum, or tbanatosis, in mate rejection by non-
receptive female btitterflies of the Satvrinae of the Palearactic. In this behaviour females feign
death with closed wings and release themselves from the stibstrate on which they are settled. This
behaviour only occurs with extreme male ]tersistence and is the final part of a mate-rejection
behaviotiral setpience. We suggest that this behaviour may be relatively rare, possibly restricted to
the tribes Elvmiini and Maniolini. There are potential associations with female mating freqtiency,
male mate-locating mechanisms and the physical strttctnre of habitats wbere attempted mating
occurs. We stiggest that the behaviotir occtirs in species where females occasionally mate more than
once, where the |)redominant male mating strategy is a perchitig sit-and-wait tactic and the species
occupy woodland strnctnres. In stich circtimstances males have relatively few opportunities to mate,
male-female encounter rates may be relatively infrequent and the physical strncttire of the habitat
allows females that adojrt posstim mate-rejection to escape from males bv dropping into vegetation.
W'e encourage further observations on this behaviour to allow a thorotigh analysis of its frequency
amongst species in order to allow a phylogenetic analysis.

Key words: |5redation, mate location, mate rejection, Pararge, La.siommata, Kirinia, Esperarge,
Maniola.

Playing possum (tbanatosis: feigning deatli) in
butterflies bas been described for two very differejit
circumstances. It was first described as a predator
escape mecbanism in Govepteryx rhamni L. Pieridae
(Dennis 1984) and Inachis io L. Nymphalidae (Den-
nis 1998). More recently, it bas also been shown to
have a role in mate refusal by females of Pararge And
iMsiommata species (Satyrinae) (Sbreeve 1985, Den-
nis 2003) and bas l)een described for Esperage cUmene
(Satyrinae) (Wakeliam-Dawson et al. 1999). As far as
we are aware it bas not been described in mate rejec-
tion for any other liutterfly taxon apart from Satyrinae
(nomenclature as in Ktirsbolt & Ra/.owski 1996) . Here,
we summarise the bebaMonral, morphological, habitat
and phylogenetic attributes for these behaviours as

Received: 3 September' 2()(H
Accepted: 21) June 2005

currently known; in doing so we focus attention on
possum behaviour in mate rejection, the objective be-
ing to encourage further observations (Table 1).

In the Lepidoptera possum behaviour is a wide-
spread secondary defence mechanism (Scoble 1992)
and in the butterflies it has been recorded in species
of Nymphalinae and Pieridae. On being caught by the
wings, these species undergo tbanatosis and effectively
play dead. When released they will lie on their side
inert for some time before taking flight (Dennis 1984,
1998). Species which adopt this behaciour tend to be
long-lived and have relatively robust wings. As manipu-
lation by birds can lead to wing damage, evidenced in
tears and scored beak marks across wings of butterflies
caught by birds (Collenette 1935, Bowers & Wiernasz
1979, Bengston 1981, Dennis et al. 1984), this would
give butterflies with tougher wings an advantage over
those with weaker structures. Butterflies described

39: 80-85, 2000 (2006)

81

displaying possum behaviour in relation to predation
typically hibernate as adults and have under-surfaces
and wing shapes that characteristically mimic dead
leaves. They spend much of the later summer feeding
avidly on nectar and in such situations are particu-
larly vulnerable to predation. G. rhamni are often so
engrossed in feeding that they can be picked off the
flower heads with ease (Dennis 1984). Experiments
on possum behaviour indicate that the behaviour may
be triggered more easily in autumn brood adults in-
tent on accumulating resources for hibernation than
in early summer broods actively engaged in mating.
Although described thus far for G. rhamni 'And /. io, it
is expected that this behaviour will extend to closely
related species with similar life histories, morphology
and behaviour (e.g., Aglais uriicae).

Possum behaviour in mate rejection is the hnal
stage of a behavioural sequence, with common ele-
ments to more normally widespread mate rejection
behaviour. First, when harassed by a male attempt-
ing to copulate, a settled female may simply rai.se its
abdomen to prevent copulation (e.g.. Finis napi L.
Pieridae; Plehejus argus, L. Lycaenidae; Wilcockson
2002, Dennis, pers. obseiwation). This mayor may not
involve wing fluttering as well. This behaviour, which
may involve exposure of the genitalia with j)ossible
chemical signalling, is common to many butterfly taxa
(Obara 1982, Obara 1984), but not all. In butterflies
that adopt abdominal raising, there may be a second
stage of mate-rejection in which a non-receptive fe-
male may attempt to prevent mating by completely
closing its wings if the initial abdominal-raising proves
unsuccessful. In butterflies that do not adopt abdomi-
nal raising, wing closing is usually the first stage to
prevent mating by a persistent male. The genitalia
of a closed wing female are then completely inacces-
sible to the courting male. Where possum behaviour
differs from ordinary wing closing is in the subsequent
stages. In possum-playing females the harassed female
may lean over onto the substrate, effectively playing
dead. Should the male persist then the female may
then release its tarsal claws from the substrate on
which it is settled and drop to lower vegetation or
even the ground. In most instances when this occur s
in Fararge aegeria L., and in the described behaviorrr
oi Esperage climeneEspev, the male will lose the fenrale
and fly off (Shreeve 1985, Wakehanr-Dawsotr et al.
1999). In other instances the male will pursite the
female to the lower substrate or gr oirnd and persist
in trying to enforce copulation befor e eventirally giv-
ing rtp and flying off. This persistence may be lorrg
(max 65 sec. in Fararge aegeria, Shreeve 1985). In no
instance has srtccessful copulation of the appar'ently
‘dead’ female been recorded, either before or after

falling from a srrbstr ate. W’hen a r ef usirrg female plays
possitm it will r emain qrtiescetU after the irrale leaves
before resuming its activities (flight or baskirrg). The
two stages of orientatiorr and dropping make possitm
behaviour unique. It appears to be part of ferrrales’
mate-rejection repertoires when the total dur ation of
male har'assment is very long (Dermis 2003).

Recorded instances (Table 1 ) of females inlay-
ing possum behavioirr in response to attempted
copulation by males is possibly restr icted to the tribe
Elymniini and one member of the tribe Maniolini,
Maniola jurtina L. (Satyrirrae, Maniolini). In tbe
case of M. furtina, this behavioirr has been recorded
in 4 instances out of 27 attempted courtships in dry
Mediterranean scr ub/woodland in Provence, Fr ance
obsei'ved between 27July and 1 1 August 1988 by one
of us (TGS) . As far as we ar e awar e possum behavioirr
linked to mate-rejection has not been r ecor ded in any
higher taxa. Instances of similar', but not identical,
behaviour are described in Garterocephalus palaemon,
Hesperiidae (Ravenscroft 1994: 1 185) and females of
other species will reject males by remaining quiescent
during attempted courtships (e.g., Leptidea sinapis
L. Disrnorphiinae; Wikhrnd 1977). In the case of
G. palaemon unreceptive females dro]) to the ground
with closed wings when pursrred by a flying male
and the behavioirr of 1.. sinapis is different from true
possirrn behavioirr in that irnreceptive flying females
detected by flying males sail to the vegetation and r est
quiescent with closed wings. Restr icting the analysis to
the Satyrinae reveals that there are jjotential associa-
tions with female mating frequency, male mate-locat-
ing mechanisms and the physical str uctur e of habitats
wher e attempted mating occirr s. Po.ssirrn behavioirr
has not been r ecorded in those taxa wher e male access
to already mated females is physically obstrrrcted by a
sphragis (mating plirg) (e.g., I lipparchia, / leteronympha:
Orr 2002). On the other hand, it has been recorded
in those members of the Elymniini {Fararge aegeria,
F xiphia Fabriciits, F. xiphiodes Standi nger) and in
Maniola jurtina, (Shr'eeve 1985, Dennis 2003, J. Ten-
nent, per'sonal communication) wher'e females may
occasionally mate mor e than once {Faegeria 4- 10%
Wickman & Wiklund 1983, Shreeve 1985; F.xiphiab%
and Fxiphioides 8%, Shreeve unpirblished; AT jurlina
4%, Maier, 1998) The mating fr equency of Esperage
climene, in which possirrn behaviour occurs, is not
known. With one exception, female jrossirm behavioirr
in mate-rejection has only been recor ded in species
in which the primary rnecharrisrn of mate-locating
by males is perching and where perching occurs in
predictable woodland strirctures, with males prima-
rily settled on vegetation above the gr ound. Partial
possum behaviour (in which females do not release

Table I.The occurrence of possum behaviour in relation to habitat structural associations, mate locating methods, female mating frequencies and congruence of hostplants with mating
areas in Palearctic butterfly taxa

82

/. Res.Lepid.

5 -

V Qj

biO

X

03

.Z ^

>. to z

03 r- U

-T biD

be 2 o

w e o

bc bc bc be be be

X 5 ir XXX

MMbc bCMbcbebe be

XXX SxSSS X

OJ

s

5

XX 5 j

<u

coo
-J J ^

be be bC be bJD be

X S X X 5 S

V

"be

<u

'(7) c75 S

s s s

qj be

^

Cl, C C

be- be be be be

^ § b -

o o ’7. ^

h h u! ^

03 Qj u

C- C- C n..

O Cj

p

C I'
c

s .

O O

K

S ^

£ J3

S H

S>o

Cb a;

cL qj G

S ^

£ •£ •-

^ V "3

Q ii: U

a.

-£ 0;

i-J “

ca Jd u
C 5 ^

p ri ;a

■ S ^ a.

J U < c/^

GJ

X

pioneer, scrub

I,.petropilana - 0]>en grassland Perching ? High High

L.paramegaera - Open grassland Perching ? High High

L.maera - Open grassland Perching ? High High

Super- Family Subfamily Tribe/Species Possum Habitat structure Primary male Female Egg-laying and Hostplant location

family behaviour mate-locating mating mating location selectivity by laying

mechanism frequency congruence females

39: 80-85, 2000 (2006)

83

bC

S

5

C

o o

? ^
o c o

^ J J

bo bO bO bJD bo bO

X S S S X £

bo bo
X X

o

bo bo
C C

bo

C

p

o-

£

s s

be ^
C XJ
oj O
Q- P

O ^

Oh

O

bo

O

j::

O

o ^
u X

.. I

bo bo
C G

C C
u u

Cl. O-

'V

c? cS
So So
c c

O (U
rs /-t

o o

■V. ^

bO

X .

D O
J J

bo

bo

o c

O 0

^ bO

'c 'p
Cl C-

c

OJ

a,

o

c)o O ^

bo bO
C G

be bo bo
G .G S
CO ^ «

bo bo bo bo
G G G G .

3j 3j O r3
G,

-Z u

3 "5 = 3 3 2

y = S S 5

u ^ O ^ ^

S o o

- P s

^ M

.0 ^ g o

^ O CO

c u

c =

3. c- <;

.£

4- ^

faC c o

■ S- o

2 -s = S -S

3 S 2 E s
^ b. ^ -2-^

s :G

GJ

G ^

■C s

•S

S-.

CQ O

o G

C3

"g;

Multiple female mating frequency = known instances of females mating more than once

8-1

/. Rn.Lepid.

f rom the substrates they arc on) occurs within the La-
siomniaUi genus, but in all instances occurrences have
taken place on bare ground substrates (T.G. Shreeve,
jjersonal observation). Recorded instances of possum
behaviour by females of Maniola jurtina have been
in woodland structures in southern Europe where
summer aestivation occurs. With the exception of
Maniola jurtina, females that ado]Dt jDossnm behaviour
are also highly selective of the locations in which they
lay eggs, even if plant species on which larvae feed are
themselves widespread.

Whilst information on the details of mate rejec-
tion by females of the majority of species is extremely
scant, the provisional analysis that we supply here is
it.self revealing of a possible phylogenetic component
with links to the overall behaviours of both sexes and
the physical structure of the habitats in which species
occur. The occurrence of female possum behaviour
as a male rejection behaviour can be explained by
a common set of characteristics; females potentially
mating more than once, occupation of specific wood-
land structures, the occurrence of perching within the
male locating repertoire and selectivity by females for
specific egg-laying locations. y\ common element of
female possum behaviour is that it only occurs after
other mate-rejection postures, such as abdominal rai.s-
ing, wing fluttering and wing closing, have failed to
deter the male. It is also potentially time consuming
for both the male and even more so for the female,
which will tend to remain (|uiescent after the courting
male has left.

If females easily and unecjuivocally communicate
their non-receptivity then it is jiointless for any male
to spend time attempting to court a non-receptive fe-
male. In such circumstances it is unlikely that female
po.ssum behaviour will be employed or even develop.
Thus the behaviour is absent where there is the clear
signal of a sphragis in single mating females (e.g.,
Hi/)j>airhia-, Satyrinae). It is also absent where there is
multiple mating, but unreceptive females are plugged
with a sphagris and males have limited resources to
make plugs (e.g. Helnonyniplia species) or where ef-
fective chemical signalling associated with abdominal
raising occurs in potentially multiple mating females
(e.g. /-’/cm spp.; Pieridae, Ohara & Hidaka 1964). It
is also a|)parently absent in species suin which recep-
tive females have specihc behaviours to advertise their
receptivene.ss to males such as the jumping flights of
virgin Aphantopus hyperantus L. (Maniolini) (Wikhmd
1982). Efficient communication of receptivene.ss is
advantageous to females since they will not spend
time harassed by males. When males are given an
unambiguous early cue of non-receptivity then they
are unlikely to harass, so more complex rejection

behaviours may not be developed.

In species in which female possum behaviour
occurs the primary male mate locating mechanism
is perching. Perching males sit and wait for females
and their persistence with individual females may
be high to try and enforce copulation with relatively
scarce females. If mated females fly into areas where
perching males detect them they do so because they
are searching for other resources (e.g. egg-laying
sites, nectar sources). In the case of Pamrgc species,
females are highly selective of egg-laying sites (e.g.
Shreeve 1985, 1986, Shreeve & Smith 1992) and pos-
sum behaviour may be the only sure mechanism of
deterring a courting male if the female is in a suitable
egg-layiug location. Time spent playing po,ssum may
be less than time spent searching for alternative re-
■sources elsewhere. In the case of Pararge aegetia, the
maximum recorded time playing possum (65 seconds)
is within the range of times spent on searching flight
between egg-laying locations (Shreeve 1985).

Eemale possum behaviour may be absent in strictly
patrolling species. We suggest that it has not devel-
oped in this circumstance because in such species
females are readily a])]iarent to males by colouration
(Dennis & Shreeve 1988) or are abundant. Thus it
may be unproductive for males to persist with an im-
receptive female since others will be readily detected
or encountered. Similar arguments apply to lekking
species (e.g. some Theclinae) ; the majority of females
entering a lek will be receptive.

Although we have limited data, it is possible that
female possum behaviour may be restricted to spe-
cies that use a specific set of woodland structures,
though not universally so. Eor example, it has not
been recorded for species that locate mates primarily
within the woodland canopy (e.g. Argynnis species,
Nymphalidae; Neozephyrus species, Lycaenidae) or on
the ground layer of woodland (e.g. Arelhiisena species,
Nymphalidae).. One key characteristic of those spe-
cies that use this form of possum behaviour is the re-
semblance of the wing underside to dry leaves, which
are predictable components of the ground layer. As ex-
amples, Argv?n? A and Neozephyrus species may be highly
conspicuous to conspecifics and predators when on
the ground layer and adoption of possum behaviour
may increase predation risk. The absence of possum
behaviour from multiple mating species of grassland
and more open structures such as rock slopes may also
be related to wing and substrate colouration, with the
latter tending to be more variable than in woodland.
In such circumstances any individual playing possum
may render itself conspicuous and vulnerable to pre-
dation. We suggest that contrast of underside wing
colouration with background is itself a constraint on

39: 80-85, 2000 (2006)

85

the development of possum behaviour.

Whilst the information on this behaviour is rather
limited there are indications that it may also have a
phylogenetic component. The adoption of possum
behaviour in mate-rejection appears restricted to
one or perhaps two tribes of the Satyrini. It may also
be a derived trait associated with their adoption of
woodland structures from ancestral open grassland
structures (Dennis & Shreeve 1988) . Evidence for the
possible restriction of possum behaviour to the tribe
Elymiini within the Satyrinae comes from its appar-
ent absence from the behavioural repertoire of Lethe
(liana, which also occupies woodland structures (Ide
2002) . Where possum behaviour occurs in other taxa
(e.g. Nymphalinae and Pieridae) , as indicated above,
it is a primary response to predator attack and the
species which adopt such a behaviour are long lived
and have relatively thicker and more robust wings than
the Satyrini (Table 1 ) . We suggest that a more detailed
examination of possum behaviour would reveal much
about the interrelationship of behaviour with habitat
structural predictability and costs and benefits of in-
dividual behavioural traits. In particular, information
on the extent of its occurrence is required for a full
analysis and a proper test of hypotheses to account for
its occurrence. We would welcome observations.

Acknowledgements

The authors thank John Tennent for his observations on
Pararge xiphia in Madeira and two anonymous referees for their
useful comments.

Literature cited

Benoston, S. a. 1981. Does bird predation influence the spot-
number variation in Maniola jurtina? Biological journal of the
Linnean Society 15: 23-27.

Bowers, M. D. & D.C. Wiernasz 1979. Avian |)redation on the
palatable butterfly, Cercyonis pegala (Satyridae). Ecological
Entomology 4: 20.5-209.

CoLENEEi E, C. L. 1935. Notes concerning attacks by British birds on
butterflies. Proceedings of the Zoological Society of London
1935:201-217.

Dennis, R. L. H. 1984. Brimstone butterflies (Gonepleryx rharnni) (L.)
playing po.ssum. Entomologist’s Gazette, 35: 6-7.

1998. Thanatosis in Inachis io (L.) Nymphalidae.

Entomologist’s Record & Journal of Variation 110: 1 15-1 16.

. 2003. Playing possum as an alternative to mate-refusal

posture in Pararge aegeria (L.), Satyrinae. Entomologist’s Record
& Journal of Variation 115: 293.

Dennis, R. L. H., K.Porter & W.R.Wiii.iams 1984. Ocellation in
Coenonynmpha tullia (Midler) (Lepidoptera: Satyridae). 1.
Structures in correlation matrices. Nota Lepidopterologica
7: 199-219.

Dennis, R. L. H. &T.G.Siireeve 1988. Uostplant-habitat structure and
the evolution of butterfly mate-location behaviour. Zoological
Journal of the Linnean Society 94: 301-318.

Ii)E,J. Y. 2002. Mating behaviour and light conditions cause seasonal

changes in the dispersal |tattern of the satyrine butterfly Lethi’
(liana. Ecological Entomology 27: 33-40.

Kvrsholt.O. & J.R\zowski 1996. The Lepidoptera of Europe. Apollo
Books, Svenborg.

M.vier, C. 1998. The behaviour and wing moiphology of the meadow
brown butterfly {Maniola jurlinaL.) in Britain, the influence of
weather and location. PhD Thesis, Oxford Brookes University,
Oxford.

Obar/V, Y. 1982. Mate refusal hormone in the cabbage white
butterfly? Naturwissenschaften 69: 551-552.

. 1984. Key stimuli eliciting the mate refusal |50sture in

the mated female of the cabbage white butterfly, Pieri.s rapae
crudvora. Proceedings of thc ja|)an Academy, 60: 145-148.

Orr, a. G. 2002. The sphragis of I lol('rnnympha Waterhouse

(Lepidoptera: Satyridae): its structure, function and role in
s|>erm guarding. Journal of Natural History 36: 185-196.

RwENSCRorr, N. O. M. 1994. Environmental influences on mate
location in male chequered skipper butterflies, Carterocephahis
palaemon (Lepidoptera: 1 lesperiidae). Animal Behaviour 47:
1179-1187.

ScoBi E, M. J. 1992. fhe Lepidoptera: form, function and diversity.
Oxford University Press, Oxford.

SiiREEVE, T. G. 1985. I'he po|)nlation biology of the Speckled Wood
butterfly Pararge aegeria (L.) (Lepidoptera: Satyridae). PhD
Thesis (CNAA), Oxford Polytechnic, Oxford.

. 1986. Egg-laying by the Speckled Wood butterfly (Pararge

aegeria): the role of female behaviour, hostplant abundance and
temperature. Ecological Entomology 1 1: 229-236.

SiiREEV'E, T. G. & A.G.Smiiii 1992. The role of weather-related habitat
use on the impact of the Enrojtean sjteckled wood Pararge
aegeria'And the endemic Parage xiphia o\\ the island of Madeira.
Biological Journal of the Linnean Society 16: 59-75.

WakI'.iiam-Dawson, a., T.Benton & V.Barniiam 1999. Butterflies
and dragonflies in northei ii Greece, 27 June - 9 July 1997.
Entomologist’s Record &Journal of Variation 111: 121-128

Wickman, P.O. & G.Wikiunii 1983. Territorial defence and its
seasonal decline in the speckled wood butterfly (Pararge
aegeria) . Animal Behaviour 31:1 206- 1 2 1 6.

WiKi.iiNi), C. 1977. Courtship behaviotir in relation to female
monogamy in Leplidea sinapis (Lepidoptera). Oikos 29: 27.5-
283.

. 1982. Behavioural shift from courtship soficitation to mate

avoidance in female ritiglet butterflies (Aphantopns hyperantus)
after copulation. Animal Behaviour 30: 790-793.

Wtt.cocKSON, A. 2002. Phe ftinctional significance of wing
morjtholog)' variation in the green-veined white butterfly (Pieri.s
napi (L.)). PhD Thesis, Oxford Brookes L’niversity, Oxford.

foHriKil of Research on the Lepidoptera

39: 86-89, 2000 (2006)

Notes

Observations on the shelter building behavior of some Asian
skipper larvae (Lepidopterai Hesperiidae)

Introduction

The larvae of many lejiidopteran families modify
their immediate environment by cutting, folding,
rolling, and/or silking portions of their foodplant
into a shelter (DeVries 1987, 1997, Scoble 1992, Stehr
1 987) . While only a few species have been investigated
in detail, the possible functions of these structural
retreats include reduced predation (Eubanks et al.
1997, Jones et al. 2002), prevention of dislodgment
(Loeffler 1996), and creation of a more favorable mi-
croclimate (Henson 1958). Among the most prolihc
groups of shelter building larvae are the Hesperiidae,
most of whom, with only a few exceptions (eg. Moss
1949, Scudder 1889), have been reported to construct
shelters. A recent review of hesperiid larval shelters
found there to be 10 basic shelter types constructed by
skipper larvae, and pointed to the possible importance
of shelter form for phylogenetic analysis (Greeney
& Jones 2003). Here we use the classification and
terminology proposed by Greeney & Jones (2003) to
describe the shelters built by 4 species of hesperiids
from China and the Philippines. Only shelters built
in nature were considered.

Species accounts

Capila Iranslucida (Leech 1893). Larvae were
reared on Cinnaviomum camphorn (Lauraceae) in Ru
Yang, Peoples Republic of China. Larvae of all instars
rest upside down inside their shelters, attaching their
crochetts to a pad of resting silk laid down on the
inside of the shelter lid. No frass accumulated within
the shelters, and larvae of all instars forcibly eject frass
from the anus. First instars were found to build Group
II Type 5, center-cut folds. These are built by cutting
into the middle of the leaf, away from the margin. The
cut is nearly circular, with the proximal and distal ends
converging outwards from the circular portion to give
the cut an overall paddle shape. This round flap is
then folded over at the narrow shelter stem to create
a man-hole-like retreat (Fig. la). Later instars build
Group III Type 10, two-cut stemmed folds. These are
roundly triangular portions of the leaf cut from the
margin using two cuts. The cuts curve towards each
other near their distal ends and finally run parallel

to create a narrow shelter stem. This narrow stem is
heavily silked to pull the flap over to meet the leaf
surface (Fig. lb). Often a minor cut is made on the
edge of the shelter lid (Fig. Ih). Opposite sides of
this small cut are then silked together to pinch the lid
into slightly peaked or tented form. This modifica-
tion is also occasionally accomplished without a cut,
by sim[)ly laying silk along the lid margin to pinch it
upwards into a peak. These modifications presum-
ably give the larvae more room to maneuver inside
the shelter. Final instars may build Type 10 shelters or
may construct Group I Type 4, two-leaf shelters. Type
4 shelters are made by silking two leaves together to
create a shallow pocket. The first instar shelter built by
C. transiucida is very similar to that recently reported
for first instars of the pyrgine, Noctuana haematospila
(G. Felder & R. Felder 1867), in Ecuador (Greeney
& Warren 2004) . The ontogenetic changes in shelter
type, however, are quite different. For a description of
the egg, larvae, and pupa of C. Iranslucida see Young
& Chen (1999).

Idgiades Utigiosus litigiosus (Moschler, 1878). Lar-
vae were reared on Dioscorea fordii (Dioscoreaceae) in
Kiui Tam Tso, Wu Kau Tan, and Tai Po, Hong Kong.
Larvae of all instars rested upside down on the shelter
lid, and kej)t their shelters clean by ejecting frass away
from the shelter. First through fourth instars built
Group III Type 9, two-cut unstemmed folds. These
were built by making two cuts originating at the leaf
margin which converged slightly towards their distal
ends to create a broad shelter stem over which the re-
sulting flap was folded onto the leaf surface. This cre-
ated a broadly folded, roundly rectangular or square
shelter (Fig. Ic). Heavy resting silk on the inside of
the shelter lid caused the shelter to bow! slightly, giv-
ing the larvae more room inside their retreat. Often
the bridge (portion where the stem is folded) was
scored along the inside by cutting only through the
leaf epidermis to weaken the folding point. Final
instars either constructed a larger version of the Type
9 shelters described above, or made Group 1 Type 4,
two-leaf shelters by silking two adjacent leaves into a
shallow pocket. The ontogenetic changes in shelter
type, as well as the individual shelters tised by each
instar were very similar to the shelters described for
larvae of the pyrgine Eantis thraso (Hiibner, 1807) in

39: 86-89, 2000 (2006)

87

Fig. 1. Shelters built by larvae of Asian skippers: Shelters not drawn to scale, stippled areas represent the area hiding the
larvae from view. A) First instar of Capila transiucida on Cinnamomum camphora, Ru Yang, Peoples Republic of China. B)
Fourth instar of C. transiucida on C. camphora, Ru Yang, Peoples Republic of China, C) First instar of Tagiades litigiosus
litigiosus on Dioscorea fordii, Hong Kong. D) Final instar of Coladenia agnioides on Eriobotrya fragrans, Ru Yang, Peoples
Republic of China. E) First instar of Bibasis sena paiawana on Hiptage bengalensis, Los Banos, Philippines. F) Final instar
of B. sena paiawana on H. bengalensis, Los Banos, Philippines. G) Final instar of B. sena paiawana on H. bengalensis, Los
Banos, Philippines showing isolation of shelter by feeding damage. H) Secondary cut to allow tenting of shelter lid by fourth
instar of C. transiucida. I) Externally visible tying silk, securing shelter lid of final instar of C. agnioides.

Ecuador (Greeney & Warren 2003).

Coladenia agnioides (Elwes & Edwards, 1897). Lar-
vae were reared on Eriobotrya fragrans (Rosaceae) in
Ru Yang, Peoples Republic of China. Frass was never
observed to accumulate inside the larval shelters.
Final instar larvae built Group III Type 10, two-cut
stemmed folds. The two major cuts, initiated from
the leaf margin, angled towards each other then ran
parallel for some distance to create a narrow stem. At
their distal ends, however, the cuts separated slightly
to create a bridge which was broader than the stem.
Larvae scored the inside of the bridge, presumably
to facilitate folding. After constructing this long-
stemmed, triangular shelter, larval feeding damage
created large channels cut in from the margins of the
shelter lid, giving the lid margins ajagged appearance.
The final product, therefore, took on a form similar
to a Christmas tree (Fig. Id) . Strong lines of tying silk
attached the rim of the shelter lid to the surrounding
leaf (Fig. li). A heavy pad of resting silk, laid in a

roughly circular pattern on the inner surface of the
shelter lid, caused the lid to bowl slightly. The chan-
nels, by weakening the structural integrity of the lid,
likely aided in this process.

Bibasis sena paiawana (Staudinger, 1889). Larvae
were reared on Hiptage bengalensis (Malphigiaceae) at
Mount Mikiling, Los Banos, Philippines. Early instars
built Group III Type 9, two-cut unstemmed folds. One
primary cut, beginning at the leaf margin, was long
and arcing, curving back towards the leaf margin. The
second primary cut was much shorter, approaching
the distal end of the larger cut directly from the leaf
margin, but leaving a broad shelter bridge with no
stem. The resulting flap was then folded to the leaf
surface. This created a distinctly shaped shelter lid,
rounded on one side, and straight along the portions
consisting of the leaf margin and along the shelter
bridge (Fig. le). The lid was not bowled or tented
in any way and remained tighly appressed to the leaf
surface. This shelter was then modified with several

88

/. lies.Lepid.

peribrations in the shelter lid, created by larval feed-
ing damage. Later instars built a Group 11 Type 6,
one-cnt fold by making a single large cut near the
distal portion of the leaf. This cut extended from
the leaf margin, directly to the midvein. This side of
the cut was then drawn together with the opposite
leaf margin creating a large, flattened pocket (Fig.
If). Heavy feeding damage around the area where
the larvae rested often resulted in this portion of the
shelter being nearly isolated from the surrounding
leaf (Fig. Ig). If the original shelter had not been
seen, the hnal product could easily be mistaken for a
Group 11, Type 9, two-cut unstemmed fold.

VA’hile there has been too little published concern-
ing the details of larval shelter construction to make
ct)mixirisons between and among taxa, these data pro-
vide evidence that the classification system provided
by Greeney &Jones (2003) is a useful tool in a variety
of geographic regions. We hope this note encourages
others to publish similar details so that such architec-
tural features may be used in the development and
testing of phylogenetic hypothesis.

Acknowledgements

We thank A. D. Warren for thonghlf'nl cnnnnent.s on earlier
versions ol this inatinscript. I IFG wishes to thank Ruth Ann and
|ohn \t Moore for their support throtigh the Population Biology
Frtnndatioti. the PBNI IS spotisored atid eticonraged onr interest
in natural histoiy This is publication tiittnher 32 of the Yatiayacn
Natural History Research Group atid is dedicated to Alexander
F. Skntch.

Literature cited

DicVRttcs, P. |. 1987. The bittterflies of Costa Rica and their tiatitral
history, Vol. 1 ,

Papiliotiidae, Pieridae, Nyinphalidae. Princeton Univ. Press,
Priticeton, New Jersey. 327 pp.

DEVRtt'.s, P. [. 1997. The butterflies of Ctrsta Rica, Vol. 2, Riodinidae.
I’riticetoti Utiiv. Press, Princeton, Newjersey. 288 p|5.

Eiibanks, M.D., K.A. Nr.sct, M. K. Petersen, Z. Liu & H. B. S,\nche/,.
1997. The exploitation of an ant-defended host plant by a
shelter building herbivore. Oecologia 109: 454-460.

(iREENEV, H. F. &• M. Jones. 2003. Shelter building in the Hesperiidae:
A cla.ssihcation scheme for larval shelters. Journal of Research
on the Lepidoptera. 37: 27-36.

Greeney, H.F. & A. D. W.arren. 2003. Notes on the natural history
of Kantis Ihraso (Hesperiidae: Pyrginae) in Ecuador. Journal of
the l.epidopterists Society 57:43-46.

Greeney, H.F. & A. 1). W.arren. 2004. Natural histoiy and shelter
building behavior of Nocluana haematospUa (Hesperiidae) in
Ecuador. Jotirnal of the Lepidopterists Society (in press)

Hensen, VV'. R. 1958. Some ecological implications of the leaf
rolling habit in Compsolerhia nhwopuhidla. Canadian Journal
of Zoology 36: 809-818.

Jones, M. T., I. CA.srEi.iANOs, & M. R. Wei.ss. 2002. Do leaf shelters
always protect cateipillars from invertebrate

predators? Ecological Entomology 27:753-757.

Loeeit.er, C. C. 1996. Caterpillar leaf folding as a defense
against predation and dislodgment: staged encounters using
Dkhomeris (Gelecbiidae) larvae on goldenrods. Jottrnal of the
Lepidopterists' Society 50: 245-260.

Moss. A. M. 1949. Biological notes on .some “Hesperiidae” of Para
and the Amazon. Acta Zoologica Lilloana 7: 27-29

Scoisi.E, .M.J. 1992. Fhe Lepidoptera. Form, function, and diversity.
Oxford Univ. Pre.ss, New York. 404 jtp.

SounuER, S. H. 1889. Fhe butterflies of Eastern United States with
special reference to New England. Vols. 1-3. Cambridge,
Mass.

SiEiiR, F. W. 1987. Immature insects, Vol. 1. Kendall/Hunt
I’tiblisliing Com|)any, Dubuc|ne, lA.

Yoiino, ]. J. 1997. A brief note on the life histoiy of Celaenorrhinus
leucocera (Kollar) in Hong Kong, China (Lepidoptera:
Hesperiidae). Entomological Reports of Futao-Km no. 26.

Youno, J. J. & X. C. CtiEN. 1999. A brief note on the life history of
Capihi traiislucida (Leech) (Lepidoptera, Hesperiidae) in Ru
Yang, Gitandong, China. Entomological Reports of FUTAO-
lv\l no. 31 .

H.\rold R Greeney, Yanayacu Biological Station &
Outer for Cireative Studies, Gosanga, Ecuador c/o
721 Foch y Amazonas, Quito, Ecuador.

E-mail: revmmoss@ynhoo. com

James J. Young, 4950 Connaught Drive, Vancouver,
B.C. Canada V6M 3E9

An interspecific mating attempt between a male Siproeta epaphus
Latreille and a female Anartia amathea Linnaeus (Lepidoptera;
Nymphalidae).

Intersjjecific mating attempts between lejiidopter-
ans are not uncommon (eg. Davies el al. 1997, Deer-
ing & Scribner 2002). Here we report an attempted
copulation between two nymj>halid species, Siproeta
epaphus (Latreille 1819) and Anartia amathea (Linn.
1758) in western Ecuador. Observations were made

a the Sachatamia Lodge (0.01. 35 S 78.45.34 W) near
the town of Mindo, Pichincha Pro’vince, at 1700 m

elevation.

On OJuly 2004, at approximately 1 1:30 am, a male
S. epaphus WAS seen and photographed in copula with
a female A. amathea (Figure 1 ) . Uj3on closer examina-

39: 86-89, 2000 (2006)

89

Fig. 1. Male Siproeta epaphus in copula with a female
Anartia amathea at 11:30 am, northwestern Ecuador.
Photo by H. F. Greeney.

Fig. 2. Close up of coupling of a male Siproeta epaphus
and a female Anartia amathea in northwestern Ecuador.
Inset shows resulting damage to female abdomen. Photo
by H. F. Greeney.

tion, it could be seen that, rather than beingjoined at
the genitalia, the male’s claspers were attached to the
side of the female’s abdomen around segment 7 or
8. The exoskeleton of the female was ruptured, and
fluids and internal organs were pushing out around
the male’s claspers (Figure 2). Damage was so severe,
it is likely that the female had lost her reproductive
capacity, and we doubt sperm transfer was possible.
Both individuals were collected and deposited in the
Museo Ecuatoriano de Ciencias Naturales in Quito.

The close relationship between Anartia and Siproeta
is well supported (Freitas & Brown 2004, Wahlberg
et al. 2005), and both genera feed on plants in the
family Acanthaceae (DeVries 1987). As plant-derived
compotmds may be used to derive intraspecific com-

munication pheromones (eg. Conner et al. 1981,
Schultz et al. 1993), similarities in sex attractants may
have resulted in the observed mistake.

Acknowledgements

We thankjane Lyons and Robert Johnsson for facilitating our
trip to Ecuador and the staff and owners of Sachatainia Lodge for
their liospitality. The work of HFG is supported in part by the
Hertzberg Family Foundation, the Population Biologv' Foundation,
Nattire & Ctilture International, and a Rufford Small Grant. Thank
you to the PBNHS for their continued support. This is publication
number 69 of the Yanayacu Natural Histoiy Research Group.

Literature cited

CoNNF.R, W. E., T. Eisner, R. K. V.wder Meer, A. Guerrero, D.
Ghiringelli, & J. Meinw.^i.d. 1981. Precopulatory sexual
interaction in an arctiid moth [Utetheisa ornalrix): role of a
pheromone derived from dietaiy alkaloids. Behavioral Ecology
and Sociobiolog)' 9: 227-235.

D.wies, N., a. Aiei,i.o,J. M.vllet, A. Pomlvnkow.ski, & R. E. Sii.BFRca.iED.
1997. Speciation in two Neotropical butterflies: extending
Haldane's Rule. Proceedings of the Royal Society of Loudon
B 264: 845-851.

Deering, M. D. & |. M. Scribner. 2002. Field bioassays show
heterospecific mating preference asymmetry between
hybridizing North American Papilio butterfly species
(Lepidoptera: Papilionidae) .

Dei'Vries, P. J. 1987. The butterflies of Costa Rica. Volume 1.
Pa|tilionidae, Pieridae, and Nymphalidae. Princeton University
Press, Princeton, N. J.

Freit.vs, a. \". L. & K. S. Brown )r. 2004. Phylogeny of the
Nvmphalidae (Lepidoptera). Systematic Biology 53: 363-
383.

Scnui.TZ, S., W. Fr\nk, M. Bofrre, T. Eesner, & J. Meinw.u.d. 1997.
Insect pheromone biosynthesis: stereochemical pathway of
hvdroxydanaidal production from alkaloidal jtrecttrsors in
CreaUmotus transiens (Lepidojttera: Arctiidae). Proceedings of
the National Academv of Sciences, USA 90: 6834-6838.
W.-uiEBERG, N., A. V. Z. Brower, & S. N\ein. 2005. Phylogenetic
relationships and historical biogeography of tribes and genera
in the snbfamilv Nymphalinae (Lepidoptera: Nymphalidae).
Biological Journal of the Linnean Society 86: 227-251.

Harold F. Greeney, Yanayacu Biological Station
& Center for Creative Studies, Cosanga, Ecuador
c/o 721 Foch y Amazonas, Quito, Ecuador. E-mail:
revmmoss@yahoo.com

Wanda Dameron, 23424Jonathan Street, Los Angeles,
California 91304, USA

Hank Brodkin, 3050 East Carr Canyon Rd, Hereford,

Arizona, 85615, USA

90

J. Res.Lepid.

%

Instructions to authors

Submission of Manuscripts. Papers must be written in English. Your initial manuscript must be submitted as an
electronic version by e-mail to the editor, preferably as a WORD document, double-spaced. Do not format the text
and do not justify the right margin. Use only the regular and italic fonts, with the italic font only for the scientific
names of the genus and species. Do not use boldface or vaiy type size (12 point preferred) for any sections or
headers. Put returns only at the end of paragraphs, not at the end of each line. Use one tab to indent each
paragraph. At the time of initial submission, all images (black and white or color) should be sent in digital form,
preferably in medium-high quality JPEG format.

Upon acceptance of a paper, higher quality images will be requested. Files may be submitted by e-mail, or CD. At
this stage digital images should be TIFF, or maximum qualityJPEG files with at least 2700 x 4000 pixels resolution.
Should you have further queries on this subject, please contact the technical editor.

Title material: All papers must include complete title, author’s name(s), institution (s), and address for
correspondence, including e-mail address. A family citation must be given in parentheses (Lepidoptera:
Hesperiidae) for referencing.

Abstract: Max. 300 words. No citations or figure references.

Keywords: Max. 10 words in addition to those in the title.

Text: The text of a regular research paper should be clearly structured: e.g. introduction, material and methods,
results, discussion, etc. with acknowledgements and literature cited at the end. Papers to be considered as Notes,
Opinion pieces, or Book Reviews need not follow this structure. A note with four or fewer references should have
these cited in the body of the text.

Name citations and Systematic works: The first mention of any organism should include the full scientific name
with unabbreviated name of author(s) and year of description. Taxonomic descriptions must comply with the rules
of the ICZN (4th edition).

Tables: Present tables in the simplest form possible. Tables must be numbered serially with Arabic numerals
independent from illustrations. Tables should be provided at the end of the paper on separate pages and not
embedded in the body of the text. Put the legends for tables on a separate page. Tables should not repeat
information that is already given in the text of illustrations. When formulating tables, keep in mind that the final
table will fill 1 column (width 8 cm or 2 columns (16,5 cm).

Illustrations are numbered serially with Arabic numerals. References to figures in both text and captions should
be as fig. and Fig. respectively. Do not spell out. Figure captions should be listed on a separate page. Maps are
considered figures and should be so captioned. Do not use plate designations; multiple figures in a single grouping
may be individually numbered or subdivided alphabetically (e.g. fig la, lb, etc). Line drawings in black and white
should include a metric scale. Both color and black and white images in high quality may be inchided. When
arranging your plates consider that they may appear either in 1 column (width 8 cm) or in 2 columns (17 cm).

References: All citations in the text must be alphabetically and chronologically listed under Literature Cited. In
the text, references should be cited in the form “... Wilson (1998a)...”, (Wilson 1988a), (Southwood et al. 1979),
or (Thomas 8c Thomas 1994). Please give full name of journals and conform to the following style:

Atkins, A. F. 1975. The life history of Anisv'^tn Waterhouse and Lyell (Lepidoptera:Hersperiidae:

Trapezitinae). Australian Entomological Magazine 2: 72-75
— 1987. The life histoi'y of Trapeziies iacclioides Waterhouse and Trapezites phigalioides Waterhouse
(Lepidoptera: Hesperiidae: Trapezitinae). Australian Entomological Magazine 13: 53-58.

Larsen, T. B. 1990. The butterflies of Egypt, Apollo Books, Svendborg. 112pp. Figurny-Puchalska E., R. M. E.
Gadeberg & J. J- Boomsma. 2000. Comparison of genetic population structure of the large blue butterflies
Macidinea nausithoiis and M. teleius. Biodiversity Conservation 9: 419-432.

Thomas, J. A., R. T. Clarke, G. W. Elmes & M. E. Hochberg. 1998a. Population dynamics in the genus

Maculinea (Lepidoptera: Lycaenidae), pp. 262-290. In: Dempster, J. P. & I. F. G. McLean (eds.). Insect
populations. Kluwer Academic Puhlishers, Dordrecht.

Reprints: We will provide authors with a PDF file of their paper by e-mail. Authors can produce an arbitrary
number of reprints identical to the printed manuscripts using these fi'f's.

The Journal of Research

ON THE LePIDOPTERA

Volume 39 2000 (2006)

IN THIS ISSUE

Date of publication: May 30, 2006
Papers

Total sperm ejaculation in monandrous (Papilio machaon) and polyandrous
(R xuthus) swallowtail butterflies (Lepidoptera: Papilionidae) restricted to
larval stage-derived nutrients

Mamorii Watanabe and Taihei Kobayashi 1

Flight Patterns and Mating Behavior in a Zephyrus Hairstreak, Neozephyrus
japonicus (Lepidoptera^ Lycaenidae)

Michio Imafukii, Yasutoshi Matstii and Hideko Matsui 8

Feasibility of light-trapping in community research on moths; Attraction radius
of light, completeness of samples, nightly flight times and seasonality of
Southeast-Asian hawkmoths (Lepidoptera: Sphingidae)

Jan Beck and Eduard Linsenmair 18

A review of the genus Manerebia Staudinger

(Lepidoptera: Nymphalidae: Satyrinae) in the northern Andes

Tomasz W. Pyrcz, Keith R. Willmott, Jason R M". Hall and Angel L. Viloria 37

Phylogenetic, habitat, and behavioural aspects of possum behaviour in
European lepidoptera

Tim G. Shreeve, Roger L. H. Dennis and Andrew Wakeham-Dawson 80

Notes 86

SMITHSONIAN INSTITUTION LIBRARIFa

3 9088 01387 2528

("over: Therelra nessus {Druvy, 1773) ©Jan Beck, 2005.